ORAL CANCER

CONTENTS

Preface
Sanjay P. Reddi
Oral Premalignant Lesions: Management Considerations
Sanjay P. Reddi and Adam T. Shafer

xi

425

Oral cancer is an insidious disease that afflicts about 30,000 individuals in the United
States annually. Despite advances in diagnostic modalities, surgical techniques, adjuvant
therapies, and postoperative care, the 5-year survival rate for oral cancer collectively remains at about 50%. Most oral cancers are squamous cell carcinoma and usually originate from asymptomatic premalignant lesions. Leukoplakia is the most common oral
premalignant lesion, followed by erythroplakia and lichen planus in the United States.
This article discusses management of these three types of lesions.

Evaluation and Staging of Oral Cancer

Vishtasb Broumand, Teresa E. Lozano, and Jesus A. Gomez

435

Oral cancer is a prevalent disease in the United States and accounts for the death and
deformity of many patients every year. Adequate evaluation and staging are paramount
in providing patients with the best chance of cure and survival. During diagnosis and
treatment, it is recommended that a multidisciplinary team discusses each particular
case to integrate the opinion and expertise of each specialty involved. These opinions
help to find the best possible solution regarding a patients needs given what is available
at the time. Modern medical technology gives us valuable means for this purpose and
will continue to evolve and provide even more possibilities to help our patients. It is important to continue to stress prevention, screening, and adequate diagnosis and treatment modalities for oral cancer.

Imaging of Oral Cavity Squamous Cell Carcinoma

Joseph M. Aulino, Megan K. Strother, and Jason L. Shipman

445

Imaging plays an invaluable role in the staging of oral cavity squamous cell carcinoma.
The mucosal extent is readily appreciated by physical examination, but submucosal
spread of larger tumors should be evaluated by cross-sectional imaging. CT may be useful to evaluate the extent of bone invasion but may be limited by dental amalgam artifacts. MRI generally suffers less from metallic artifact and is superior to define local soft
tissue invasion. The role of positron emission tomography is still emerging, and general
recommendations are made.

VOLUME 18

NUMBER 4 NOVEMBER 2006

Advances in the Detection and Diagnosis of Oral Precancerous

and Cancerous Lesions
John R. Kalmar

465

This article provides an overview and update on the clinical and laboratory diagnosis of
oral precancerous and cancerous disease. After a brief summary of the typical clinical
features, diagnostic aids designed for the dentist and dental specialist, including recently
available commercial products, such as brush cytology and tissue fluorescence units, are
discussed. Diagnostic techniques, such as scalpel biopsy, fine-needle aspiration cytology,
and sentinel node biopsy, are examined with an emphasis on recent morphologic, chemical, and molecular studies that relate to lesion behavior. The future of molecular techniques and their impact on the diagnosis and management of oral precancerous and
cancerous lesions are also presented.

Molecular Biology and Clinical Behavior of Oral Cancer

Brian L. Schmidt

483

Our understanding of the molecular biology of oral squamous cell carcinoma has progressed significantly over the past decade. In this article, the traditional histopathologic
features that have been used to treat patients with oral cancer are reviewed. Some of the
more recent molecular studies and technologies that we, as surgeons, might be using in
the future to tip the balance in our patients favor are then presented.

Oral Cancer Prevention

M. Abraham Kuriakose and Rajeev Sharan

493

Despite advances in modern multimodality treatment, the survival rate of oral cancer
has improved only modestly to approximately 60% in the recent past. The morbidity
of oral cancer treatment is proportional to the stage of disease at diagnosis. Measures
to downstage the disease or prevent it will have a significant benefit to society.
Advances in our understanding of oral carcinogenesis, methods for early detection,
and development in chemoprevention have made oral cancer prevention a clinical reality. This article reviews the current status of early detection and chemoprevention measures in oral cancer.

Verrucous Carcinoma
Rocco R. Addante and Samuel J. McKenna

513

Oral verrucous carcinoma (VC) is a distinct low-grade variant of squamous cell carcinoma of the oral cavity. It is characterized by a verrucous or wart-like papillary surface,
slow growth, lack of ulceration, and a propensity for local invasion rather than metastatic spread. With progression, it can assume an exophytic fungating quality; despite
its deceptively benign microscopic appearance, VC can extensively infiltrate and destroy
adjacent tissue, including muscle, cartilage, and bone. It can present in concert with proliferative verrucous leukoplakia and invasive squamous cell carcinoma and may represent a stage in the continuum from verrucous hyperplasia to invasive squamous cell
carcinoma. The preferred treatment is wide local excision.

Special Considerations with Floor of Mouth and Tongue Cancer

Oneida A. Arosarena, Matthew Madsen, and Richard Haug

521

Oral tongue and floor of mouth cancers comprise most cases of oral cavity malignancies
and demonstrate locally aggressive behaviors related to lack of anatomic barriers to
spread and a propensity for early cervical metastases. Current guidelines for oral cavity

vi

CONTENTS

cancer management include surgery for stage I and II disease and combined surgery and
postoperative radiotherapy for stage III and IV disease. The surgical management of oral
tongue and floor of mouth cancers can result in significant functional impairment. Surgical defects resulting from resection of advanced tumors are best reconstructed with microvascular free tissue transfer. Large multi-institutional studies are needed to assess the
value of primary tumor characteristics as indicators of metastatic potential and for refinement of methods for detecting micrometastases.

Surgical Management of the Neck in Oral Cancer

Eric R. Carlson and Ivo Miller

533

Surgical management of the cervical lymph nodes is an essential aspect of comprehensive surgical care for patients with cancer of the oral cavity. The histopathologic status
of these lymph nodes remains the main prognostic index for patients with this disease
process. As such, it is incumbent on the surgeon to apply a scientifically valid approach
to neck surgery so as to improve the chances of survival of patients who have oral cancer. This philosophy is applicable to the clinically positive neck (N+) and the clinically
negative neck (N0). This article reviews the history and practical aspects of this fascinating and thought-provoking discipline.

Sentinel Lymph Node Biopsy in the Staging of Oral Cancer

Thomas D. Shellenberger

547

Effective management of patients with squamous cell carcinoma of the oral cavity
depends on accurate staging to determine the prognosis and to select appropriate therapeutic strategies. The accuracy of the currently available means of staging the clinically
negative neck in oral cancer is limited, however, resulting in most patients undergoing
pathologic staging by elective neck dissection. Sentinel lymph node biopsy (SNB) is under investigation in the staging of oral cancer as a potentially more accurate and less invasive approach than neck dissection. The rationale for SNB is discussed, along with the
premise for its use in oral cancer. The technique for SNB in oral cancer is presented, along
with a review of its diagnostic efficacy and remaining obstacles to general acceptance.

Access Surgery for Oral Cancer

Rui Fernandes and Robert Ord

565

The most important task in managing oral cancer is the resection of tumors without positive margins. To accomplish this goal, the surgeon is often faced with placing facial skin
incisions to improve access to the oral cavity. This article has reviewed some of the most
commonly used approaches and highlighted techniques used by the authors to resect tumors with minimal postoperative scarring. Although esthetics are important to the patient, they remain a distant third consideration to the resection of the tumor and
functional reconstruction. The surgeon and the patient should not lose sight of this fact.

Postablative Reconstruction Techniques for Oral Cancer

D. David Kim and G.E. Ghali

573

Cancer of the oral cavity can leave devastating defects after surgical ablation. The use of
new and conventional reconstructive techniques can provide functional and aesthetic restoration of the affected areas. Notably, microvascular reconstructive techniques can be
useful in providing replacement tissues that are tailored to the defect while minimizing
donor site morbidity for even the most complex ablative defect.

CONTENTS

vii

Chemoradiation Therapy: The Evolving Role in Head and Neck Cancer and
Its Application to Oral Cavity Tumors
Barbara A. Murphy and Anthony Cmelak

605

The role of chemotherapy and radiation in the treatment of head and neck cancer has changed dramatically over the past decade. Concurrent chemotherapy with radiation has become a standard treatment option for head and neck cancer patients with resectable
tumors who desire a function preservation approach, patients with unresectable disease,
and high-risk postoperative patients. Although the data is conflicting, concurrent chemotherapy with radiation appears to be less effective in oral cavity tumors than in other
sites. Surgical resection remains the treatment of choice for most oral cavity lesions. Concurrent chemotherapy with radiation is recommended in unresectable patients and in
high-risk postoperative patients with involved margins, multiple positive lymph nodes,
and extracapsular extension. Improved outcomes are gained at the expense of increased
toxicity.

Management of Locoregional Recurrence in Oral Squamous Cell Carcinoma

Deepak Kademani and Eric Dierks

615

The best opportunity for a cure in the patient with oral cancer is at the time of initial diagnosis. The result of treatment of recurrent disease after prior definitive treatment is
generally poor. Although there is a paucity of reliable data regarding the management
of recurrent oral cancer, there is a general consensus that surgical resection of recurrent
disease in patients willing and able to tolerate extirpative procedures provides the most
durable long-term salvage rates. Surgery for recurrent disease is often an extensive undertaking having significant functional, esthetic, social, and financial implications for
these patients. As surgeons, we have to consider whether the efforts of extirpation are
likely to provide the patient with a meaningful likelihood of salvage or significant increase in lifespan without considerable functional compromise.

Interventional Approaches to Management of Pain of Oral Cancer

Sukdeb Datta and Umeshraya T. Pai

627

The management of pain of oral cancer is often difficult. Causes of pain are complex. A
proper management protocol includes a multidisciplinary approach including surgery, radiation therapy or chemotherapy directed at neoplasms and pain relief, systemic analgesic
therapy or therapy with hormones or steroids. Another important approach to the management of pain of oral cancer is the selection of appropriate interventional pain management techniques. Such techniques include the use of peripheral nerve blocks, blockade of
appropriate autonomic ganglia in the head and neck (trigeminal, sphenopalatine or stellate) and the use of central neuraxial (examples include intraventricular, intracisternal
or intrathecal administration of opioids with or without local anesthetics) techniques.

End of Life Issues in Oral Cancer

James Bridges and John Mulder

643

The head and neck surgeons role is important throughout the entire illness trajectory of
a patient who has cancer. Even if patients with advanced head and neck cancer are referred to hospice or palliative care, they still present with major problems, such as pharyngocutaneous fistulas, dysphagia, and compromised airways that head and neck
surgeons are uniquely qualified to palliate. Continuing involvement in patient care is rewarding for patients, families, and their surgeons. Successfully alleviating a terminal patients suffering also can bring an enormous sense of professional accomplishment.
Patients and families are just as grateful for a well-managed death as they are for treatments that result in a cure.

viii

CONTENTS

FORTHCOMING ISSUES
February 2007
Management of Impacted Teeth
Vincent J. Perciaccante, DDS, Guest Editor
May 2007
Treatment of the Female OMS Patient
Leslie R. Halpern, DDS, MD and
Meredith August, DMD, MD, Guest Editors
August 2007
Orthognathic
Joseph E. Van Sickels, DDS, Guest Editor

PREVIOUS ISSUES
August 2006
Modern Surgical Management of the
Temporomandibular Joint
A. Thomas Indresano, DMD, Guest Editor
May 2006
Perioperative Management of the OMS Patient, Part II
Harry Dym, DDS, Guest Editor
February 2006
Perioperative Management of the OMS Patient, Part I
Harry Dym, DDS, Guest Editor

THE CLINICS ARE NOW AVAILABLE ONLINE!

Access your subscription at:
http://www.theclinics.com

Oral Maxillofacial Surg Clin N Am 18 (2006) xi

Preface

Sanjay P. Reddi, BDS, MD

Guest Editor

Oral cancer is an insidious disease that aicts

over 30,000 new patients annually in the United
States and many more in other countries. Despite
advances in diagnostic methods, imaging modalities, surgical techniques, and adjuvant therapies,
the ve-year survival rate for oral cancer patients
unfortunately still remains at about fty percent.
This issue is devoted to improving this outcome
and improving our understanding of this disease
from diagnostic, molecular, perioperative/surgical, and post-treatment standpoints. The articles
cover all of these issues as well as the vital areas
of pain management and palliative care.
The authors represent diverse disciplines and
are recognized and respected for their contributions. I am very grateful to them and sincerely
appreciate their eorts. I hope that all of the
readers will derive new knowledge and recognize
the authors eorts. I want to thank John Vassallo
from Elsevier/Saunders for keeping me on track

and Dr. Richard Haug for selecting me for this

project.
Finally, I want to express my deep gratitude
and appreciation to all my teachers at the University of Miami/Jackson Memorial Hospital,
most notably, Drs. Stuart Kline, Robert Marx,
and Eric Carlson. Dr. Kline emphasized three
simple and eective rules: Pay attention to detail,
do it right the rst time, do it by the book. I have
realized that these rules are extremely useful in the
day-to-day practice of oral and maxillofacial
surgery and, undoubtedly, will be just as useful
for surgery related to oral cancer.
Sanjay P. Reddi, BDS, MD
Permian Basin Oral & Maxillofacial Surgery
2453 East 11th Street
Odessa, TX 79761, USA
E-mail address: orif98mmf@hotmail.com

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.08.001

oralmaxsurgery.theclinics.com

Oral Maxillofacial Surg Clin N Am 18 (2006) 425433

Oral Premalignant Lesions: Management

Considerations
Sanjay P. Reddi, BDS, MDa,*, Adam T. Shafer, DMDb
a
Private Practice, Permian Basin Oral and Maxillofacial Surgery, Odessa, TX, USA
Department of Oral and Maxillofacial Surgery, Vanderbilt University Medical Center, Nashville, TN, USA

Introduction
Oral cancer represents about 3% of all malignancies in men and 2% of all malignancies in
women in the United States [1]. The American
Cancer Society estimates that 30,000 new cases
of oral cancer will be diagnosed in the United
States population, of which 90% will be squamous cell carcinomas. Most oral cancers are squamous cell carcinomas because the mutagens in
tobacco, alcohol, and viruses have prolonged exposure to the supercial layers of the oral mucosa,
creating an opportunity for genetic mutation
within the supercial mucosal layers that leads
to a phenotypical mucosal premalignant lesion.
Oral squamous cell carcinoma (OSCCA) generally
arises in middle-aged and older people, with
a male to female ratio of greater than 2:1 [2].
This ratio is changing because of increased incidence of tobacco and alcohol use among women.
Despite recent advances in surgery, radiation, and
chemotherapy, the 5-year survival rate remains
between 50% and 55% [3,4].
Because the 5-year survival rate is directly
related to the stage of malignancy at the time of
diagnosis, prevention and early detection are vital
to decrease the incidence and improve the survival
odds of individuals who develop the disease. Oral
premalignant lesions and early stage malignancies
often arise as subtle lesions and require an alert

* Corresponding author. Permian Basin Oral and

Maxillofacial Surgery, 2453 East 11th Street, Odessa,
TX 79765.
E-mail address: orif98mmf@hotmail.com
(S.P. Reddi).

clinician with a high index of suspicion, especially

if any of the risk factors are present. Invasive
OSCCA usually is preceded by the presence of
a clinically detectable premalignant lesion of the
oral mucosa. A premalignant lesion is dened as
a morphologically altered tissue in which cancer
is more likely to occur than in its apparently
normal counterpart [5]. Frequently these lesions
can manifest as white lesions, referred to as leukoplakia, or red lesions, referred to as erythroplakia.
As these lesions progress, patients or their doctors
may notice an ulcer or a mass, which usually is
asymptomatic. Occasionally, OSCCA arises from
other preexisting mucosal lesions, such as lichen
planus, reverse smokers palate, and tobacco
pouch keratosis. This article focuses on relevant
aspects of the more common premalignant lesions, leukoplakia, erythroplakia, and lichen planus, with regard to their malignant potential.
Leukoplakia
Leukoplakia is the most common oral premalignant lesion. The World Health Organization
(WHO) denes leukoplakia as a white patch or
plaque that cannot be characterized clinically or
pathologically as any other disease [6]. It is therefore a diagnosis of exclusion and only a clinical
descriptive term and has no diagnostic or prognostic implication. When clinicians inform patients about identifying leukoplakia in the oral
cavity, it is imperative that other entities, such as
frictional keratosis, candidiasis, leukoedema, and
lichen planus, have been considered and eliminated from the dierential diagnosis. The concept
of a two-step process of malignancy (ie, initial

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.08.002

oralmaxsurgery.theclinics.com

426

REDDI & SHAFER

presence of a precursor [premalignant] lesion subsequently developing into OSCCA) is well established and leukoplakia seems to be the best
known precursor for OSCCA [7]. One study reveals that between 16% and 62% of OSCCA are
associated with leukoplakia at the time of diagnosis. Leukoplakia is identied most frequently in
middle-aged and older men, with increasing prevalence with age [8]. It is prevalent in 8% of men
over the age of 70 and 2% of women over the
age of 70 [8]. The most common sites include
buccal mucosa, alveolar mucosa, and lower lip.
Lesions arising on the oor-of-mouth, lateral
tongue, and lower lip are most likely to show dysplastic or malignant changes, and are considered
high-risk sites [9].
Leukoplakia can be classied based on clinical
appearance into: (1) early/thin, (2) thick/homogenous, (3) granular/nodular, (4) proliferative verrucous, and (5) speckled leukoplakia types [10].
The early/thin variant appears as a thin, minimally elevated gray-white plaque with either
well-dened or poorly dened borders [10]. It
gradually progresses to a thick, homogenous lesion with a leathery white ssured surface. Some
lesions progress from the early type to the granular/nodular type with pebbly surface irregularities.
Some early lesions progress to a widespread multifocal lesion with a papillary surface. This uncommon variant, called papillary verrucous
leukoplakia [11,12], will be discussed in a subsequent article.
The malignant transformation rate of oral
leukoplakia ranges between 0.5% and 20% in
some studies and between 15.6% and 39% by
other accounts [6,13]. In one particularly large series, Waldron and Shafer studied 3256 cases of
oral leukoplakia and reported that 19.9% had
some degree of epithelial dysplasia [9]. Within
this subgroup, 3.1% of cases had squamous cell
carcinoma, 4.6% had severe dysplasia or carcinoma in situ, and 12% were mild to moderate
dysplasia [9]. This study also revealed that the location of leukoplakia had signicant correlation
with the rate of dysplastic or malignant changes.
Floor-of-mouth was the highest risk site with
42.9% of lesions demonstrating either dysplasia,
carcinoma in situ, or squamous cell carcinoma
[9]. Tongue and lip were also high-risk sites with
24% of lesions demonstrating either dysplasia or
carcinoma [9] (Fig. 1).
Clinical appearance of the leukoplakic lesion
also indicates a possible relationship with the
probability of dysplastic or malignant ndings

Fig. 1. Tongue leukoplakia with central area of

dysplasia.

within the lesion. As the lesion increases in

thickness, the probability of dysplasia or malignancy increases. Papillary verrucous leukoplakia
has the greatest chance of dysplasia/malignancy
[11,12] compared with thick leukoplakia, which in
turn has a higher dysplastic/malignant potential
than the thin lesion [10].
Speckled leukoplakia (Fig. 2) is a mixed redwhite lesion and has the highest malignant potential among all subtypes with a rate of about 44%
and a dysplasia rate of 51% according to Pindborg and coworkers study [13]. Regardless of
the subtype or location, all leukoplakias should
be considered at risk for malignant transformation and biopsy should be obtained after diagnosis
and elimination of other white lesions. Histologic
examination of the specimen denitively excludes
other white lesions in most cases and establishes
the degree of epithelial dysplasia if present. Patients should be educated about avoiding risk factors and the importance of periodic long-term
follow-up visits to monitor the lesion should be

Fig. 2. Speckled leukoplakia of oor-of-mouth.

MANAGEMENT OF ORAL PREMALIGNANT LESIONS

stressed. The risk factors for malignant transformations of oral leukoplakia include: (1) female
patients, (2) nonsmokers, (3) longstanding lesions,
(4) oor-of-mouth and tongue subsites (Fig. 3),
(5) nonhomogeneous leukoplakia, (6) presence
of Candida albicans within the lesion, and (7)
presence of epithelial dysplasia in the lesion
[14,15].
Management of leukoplakia
After establishing a diagnosis of leukoplakia
by eliminating other white lesions and obtaining
tissue for histopathologic examination, the immediate concern is to identify and quantify presence
of dysplasia or squamous cell carcinoma. In cases
of no dysplasia to mild dysplasia, the decision to
observe versus denitively treat the lesion may be
inuenced by the site of origin and clinical
subtype of leukoplakia. Lesions that exhibit
carcinoma in situ or early invasive OSCCA
warrant excision with margins and there is little
dispute in the literature. For lesions that demonstrate moderate to severe dysplasia, and mild
dysplasia in high-risk sites, treatment options are
variable and there seems to be a lack of consensus
regarding the denitive treatment modality.
Treatment options include surgical excision, cryosurgery, CO2 laser surgery, and topical and systemic application of carotenoids.
Surgical excision
This traditional method of treatment is done
with a cold steel scalpel and may or may not
involve removing a margin of clinically uninvolved tissue in select cases. Recurrence rates are
variable, ranging from 15% to 35% and recurrences are noted adjacent to the excised lesions,
especially in high-risk sites. Other problems

Fig. 3. Extensive lesion of palate.

427

include inability to excise widespread or diuse

lesions (Fig. 3) without causing signicant morbidity and esthetic compromise, and lack of willingness on the part of the patient to undergo
widespread excisions because these lesions are essentially asymptomatic and have low malignant
potential. Reasons for the high recurrent rates include diculty in identifying the margins of the lesions and involvement of adjacent structures, such
as salivary ducts. Other disadvantages include
scarring of the residual tissue bed with eventual
distortion, excessive bleeding, especially from
tongue and oor-of-mouth, edema, and the possibility of requiring a local ap for coverage of
larger excisions. It is a useful method of treatment
of smaller, localized lesions. Studies indicate,
however, that surgically excised lesions developed
cancer at a higher rate than non-excised lesions,
and there are no randomized trials comparing surgical excision versus no treatment to indicate that
surgery is the preferred method of control [16,17].
These studies also conrmed that nonhomogenous leukoplakia had a signicantly higher
malignant transformation rate than homogenous
lesions.
Cryosurgery
This modality essentially ablates soft tissue by
therapeutic freezing and has some disadvantages,
including lack of depth control in the freezing
process, lack of specimen availability because of
the ablative process, and pain and swelling. The
main advantage is ease of use. This procedure is
now replaced by the CO2 laser.
CO2 laser
The CO2 laser can be used either to evaporate
the entire lesion without obtaining tissue for biopsy or to excise a lesion and provide a tissue
sample and create hemostasis of the residual
base. When used for excision biopsy, the advantages include obtaining a specimen with decreased
morbidity and adequate hemostasis, especially
when larger lesions are excised. Healing of the residual bed occurs by secondary intention and less
tissue distortion is noted when compared with
steel scalpel excision because of characteristics of
the CO2 laser. When mild to moderate dysplasia
is identied in a low-risk site lesion, laser ablation
may be performed provided a biopsy was done before and the patient understands the importance
of long-term follow-up. For higher grade lesions,
laser excision may be performed to remove the
clinically identiable lesion for biopsy and obtain

428

REDDI & SHAFER

hemostasis. The main drawback to laser excision

is that the specimen margins may be cauterized
and therefore lateral spread of dysplasia or malignancy cannot be assessed. Several retrospective
studies demonstrating the ecacy of the CO2 laser
for prophylactic and therapeutic removal of oral
leukoplakia (OL) have been published that show
recurrence rates ranging from 7% to 38% and
malignant transformation rates ranging from
1% to 2% [1820]. Proponents of this modality
advocate excision for lesions on nonkeratinized
epithelia and evaporation for lesions arising on
keratinized oral tissues, before which incision biopsy must be obtained from the lesion. The primary disadvantage of evaporation is lack of
tissue for histopathologic examination. Primary
advantages cited include hemostatic eld, minimal
patient discomfort, minimal surgical bed distortion, no need for ap surgery, and comparable local recurrence and malignant transformation rates
as with scalpel excision techniques.
Nonsurgical therapies
Vitamin A, retinoids, beta-carotene, vitamin E,
bleomycin, and alpha tocopherol are some of the
agents used topically or systemically with variable
results. Vitamin A and beta carotene have been
shown to cause regression of oral lesions, but the
studies lack long-term follow-up and adequate
number of patients. Detailed discussion regarding
this topic can be found in a subsequent article.
Erythroplakia
Oral erythroplakia (OE) generally is considered a premalignant lesion of the oral mucosa and
is uncommon compared with leukoplakia. It is
identied predominantly in middle-aged and elderly individuals and has a predilection for soft
palate, buccal mucosa, and oor-of-mouth [21].
Literature review regarding this lesion is fairly
scarce because of decreased incidence and somewhat confusing classication system. The classication problem is inherent because of OE
sometimes being associated with its counterpart
OL and the varying degrees of red and white
that are identied clinically.
Denition and classication
The term erythroplakia (erythroplasia) was
used to describe and dierentiate red lesions of
the oral mucosa from the white lesions of oral
leukoplakia. Although the terms early history
was originated from Queyrat in 1911 after

describing a bright-red precancerous lesion of

the glans penis, the term erythroplasie eventually became translated into erythroplakia after
its analogy to leukoplakia. Much later authors
drew attention to the fact that these lesions do not
form plaques at all, in contrast to their white
counterpart, and are actually depressed below the
surrounding mucosa [22]. In addition, confusion
arose when lesions showed clinical overlap between OE and OL, which introduced more terms,
such as erythroleukoplakia, erosive leukoplakia,
leukoerythroplakia, and speckled erythroplakia.
Diculty in classication exists because of this
and most reports have focused only on the
true or homogeneous, velvety, bright-red OE.
Over the years, the denition of OE has changed.
The WHO dened OE in 1978; the term was reconrmed during an international seminar on
OL in 1983 and later changed in 1994 at another
symposium [23] to its most widely accepted denition: the term erythroplakia is used analogously
to leukoplakia to designate lesions of the oral mucosa that present as red areas and cannot be diagnosed as any other denable lesion [23]. Similar
to OL, principles of provisional diagnosis and
denitive diagnosis were introduced for OE.
Provisional diagnosis was stated as follows:
A provisional diagnosis of OE is made when a
lesion at clinical examination cannot be clearly
diagnosed as any other disease of the oral mucosa
with red appearance [23]. Denitive diagnosis
was dened as follows: A denitive diagnosis of
OE is made as a result of identication, and if possible, elimination of suspected etiologic factors
and, in the case of persistent lesions, histopathological examination [23]. Many more denitions
were proposed; however, Pindborg and colleagues
[24] dened the most widely accepted term of OE
as a ery red patch that cannot be characterized
clinically or pathologically as any other denable
lesion. Almost all true erythroplakias demonstrate signicant epithelial dysplasia, carcinoma
in situ, or invasive squamous cell carcinoma at
diagnosis.
Incidence and prevalence
OE occurs less commonly than OL [25]; however, the gures on prevalence, incidence, and
relative frequency are lacking. Reports published
exclude OE in large-scale epidemiologic studies
of oral mucosal lesions [2527] and those that
did, did not include OE in their follow-up study.
Of those reports published, Mehta and coworkers

MANAGEMENT OF ORAL PREMALIGNANT LESIONS

[26] found in their study from 1966 to 1969 9 cases

of OE in a survey of 50,915 Indian individuals
(0.02%). Shafer and Waldron [27] described 58
cases of OE among 64,345 biopsies, indicating
a prevalence of 0.09%. Other groups showed in
their studies from Malaysia, [28,29], Burma [30],
and South Africa [31] a prevalence of 0.02%,
0.08%, and 5.7%, respectively. A range of prevalence between 0.02% and 0.83% from dierent
geographic areas has been documented. The causes
are believed to be the same as those associated
with invasive squamous cell carcinoma (SCC)
of the mouth (ie, multifactorial external factors,
such as tobacco, alcohol, and syphilis, and intrinsic
factors, such as systemic or generalized states).
Clinical aspects
In general, oral squamous mucosal lesions
appear red when the overlying mucosa is atrophic
and subepithelial capillaries dilate (Box 1). OE
might appear as smooth, granular, or nodular
red mucosa with well-dened margins (Fig. 4)

Box 1. Differential diagnosis of red

lesions of the oral mucosa
Mycotic infections
Oral candidiasis
Erythematous candidiasis
Generalized candidal erythema
Denture-induced stomatitis
Histoplasmosis
Bacterial infections
Tuberculosis
Mucosal diseases
Atrophic oral lichen planus
Systemic lupus erythematosis
Pemphigus
Pemphigoid
Others
Amelanotic melanoma
Hemangioma
Telangiectasias, lingual varices
Kaposi sarcoma
Oral purpura
Adapted from Reichart PA, Philipsen HP.
Oral erythroplakia: a review. Oral Oncol
2005;41(6):55161.

429

Fig. 4. Erythroplakia.

[24]. Shear [32] identied dierent morphologic

characteristics, such as the surface being irregular
or granular with white or yellow foci and presence
of leukoplakia within the lesion, giving it a granular or speckled appearance. He also described OE
being associated with or adjacent to OL. OE is
typically soft to palpation and does not become
indurated until it develops into an invasive
carcinoma [33]. Buccal mucosa, soft palate, and
oor-of-mouth are most commonly aected by
OE, although dierences in location have been
observed in dierent genders. In men the most
commonly aected site was the oor-of-mouth
followed by the retromolar trigone; however, in
women mandibular alveolar mucosa, mandibular
gingiva, and the mandibular vestibule exhibited
the most occurrences. It is uncommon for the
tongue to be aected. In India, reverse smoking
has been reported to be associated with a higher
incidence on the palate [34].
OE occasionally may be associated with oral
lichen planus, along with OL. Eight cases of OE
were found in a cohort of 740 patients who had
oral lichen planus, according to one study [12].
The clinical appearance of these lesions has been
described as having denitive borders and being
situated 0.1 to 0.2 mm below the level of the surrounding oral mucosa, showing the dierence of
the OE lesions from the atrophic type of oral lichen planus, which has no denitive border and
occurs at the same level as the surrounding mucosa. Holmstrup and Pindborg [12] reported in
patients who had OE and lichen planus that the
buccal mucosa was always aected (n 8). Bouquot and Ephros [33] have described the typical
OE lesion as being less than 1.5 cm in diameter,
and half are less than 1 cm; lesions larger than
4 cm have been observed infrequently.

430

REDDI & SHAFER

Rate of malignant transformation

OE has the highest risk for malignant transformation compared with all other premalignant
and potentially malignant oral mucosal lesions,
including OL, oral lichen planus, and oral submucosal brosis. Malignant transformation rates
for oral carcinoma in situ or severe epithelial
dysplasia (including red, white, and combined redwhite macules) were shown to average 26.3%,
ranging from 14.3% to 50.0%. Histopathologic
reports conclude that among OE of the homogenous type, 51% showed invasive carcinoma, 40%
carcinoma in situ, and 9% mild or moderate
dysplasia [27].
Management
Because of the high incidence of signicant
epithelial dysplasia, carcinoma in situ, or early
invasive squamous cell carcinoma at diagnosis,
surgical intervention is necessary. Complete excision of the lesion with clear margins down to the
submucosal level provides a specimen that can be
assessed adequately for margin control and may
reduce the risk for local recurrence signicantly.
Lichen planus
Lichen planus is a mucocutaneous disease than
can aect the skin or mucosa. Oral lichen planus
(OLP) is a chronic disease that has predilection
for the buccal mucosa, gingiva, and lateral
tongue, and seems to aect women more often
than men. The cause of OLP is not fully understood; however, it has been suggested that an
immunologically induced degeneration of the
basal cell layer of the mucosa is the cause [35].
The mechanism behind the basal cell layer destruction is a cell-mediated immune process involving antigen presenting cells (APC), such as
Langerhans cells, T lymphocytes, and macrophages. Regezi and colleagues [36] showed
through histochemical assays that cells involved
in OLP were of T-cell origin, specically CD4
helper and CD8 cytotoxic subtypes. The role of
the APC is to acquire and present the antigenic
fragments, or epitopes, together with class II major histocompatibility complex, to CD4 T-cells.
These cells also release interleukin-1, which attracts the T lymphocytes to the area and causes
them to produce interleukin-2, which causes Tcell activation and proliferation. Studies have
shown [36] that once activated these CD8 cells
have an anity for the basal cells and they

produce gamma interferon, which induces keratinocytes to present the class II histocompatibility
antigens HLA-DR and also to increase their rate
of dierentiation. Clinically, this results in a white
lesion with a thickened surface. HLA-DR is mutually expressed between APCs and lymphocytes,
and this expression may carry over to epithelial
cells from direct contact causing self antigens to
be recognized as foreign and to be a cause for
autoimmunity [37].
Classication
Clinically, there are various OLP presentations. A study in 1968 [38] divided OLP into six
types: reticular, papular, plaque-like, atrophic,
bullous, and erosive OLP.
Reticular: Most common type of OLP that
usually arises in the buccal mucosa bilaterally
and characteristically has ne, radiating white
striae known as Wickham striae, which may be
surrounded by a denite erythematous border
[39]. It tends to be asymptomatic.
Papular: Characteristically present as ne,
small, white pinpoint papules about 0.5 mm in
size and are asymptomatic also.
Plaque-like: Clinically may resemble oral leukoplakia with asymptomatic, homogenous white
patches that are slightly elevated and smooth or
irregular and multifocal. Most common sites are
buccal mucosa and dorsum of the tongue. More
common in smokers [40].
Atrophic: Characteristically presents as diuse
red patch with a rim of white striae surrounding
a painful lesion. Commonly involves the attached
gingiva and causes a burning sensation. A potentially malignant lesion.
Bullous: Typically present as small, fragile bullae or vesicles. After rupture, an ulcerated surface
remains that can be painful. Common sites of
involvement are the buccal mucosa and lateral
margins of the tongue.
Erosive: The second most common type of
OLP. Irregularly shaped and may be covered
with a brinous plaque or pseudomembrane over
the erosion. Very painful; has a greater potential
to undergo malignant change (Fig. 5) [41,42].
Management
Currently there is no denitive cure for OLP,
and this is evident in the numerous modalities
used for treatment. Because of low risk for
malignant transformation, patients ideally need
to be followed long term, similar to OL [4345].

MANAGEMENT OF ORAL PREMALIGNANT LESIONS

431

from 0.4% to 5.6%, although the actual overall

frequency of malignant transformation is low, between 0.3% and 3% [43]. This statistic is complicated because of some studies wherein lichenoid
contact reactions were reported as OLP. The
two OLP subtypes with an increased potential to
undergo malignant transformation are the erosive
and atrophic forms [43,44]. The malignant transformation of OLP in patients who had dysplasias
was not associated with exposure to carcinogens.
Summary
Fig. 5. Erosive lichen planus with carcinoma in situ.

Topical or systemic corticosteroid therapy seems

to be the mainstay in management of symptomatic OLP, recognizing that prolonged steroid use
may cause adrenal suppression and oropharyngeal candidiasis. Such patients require periodic
follow-up visits to monitor ecacy of therapy
and status of the lesion. Reports on ecacy of vitamin A or retinoids are controversial, with one
report citing that etretinate, a vitamin A analog,
should be the drug of choice for erosive OLP
[46], whereas other reports state that the lesions
remained unchanged after similar protocols [47].
Cyclosporine also has been used for management
because it inhibits the proliferation of T lymphocytes by reducing the production of lymphokines
[48]. Serum level of cyclosporine needs to be monitored regularly to avoid side eects. Griseofulvin
has been proposed to help with erosive OLP if
corticosteroid treatment is contraindicated [49].
Dapsone, photochemotherapy, phenytoin, azathioprine, and levamisole all have been used with
limited success [50]. Surgical management has
been considered for plaque-like lesions, although
its application is limited because it is not suitable
for the erosive and atrophic types in which the
surface epithelium is eroded.
Malignant potential
Malignant potential of OLP remains controversial [51,52]. Studies range from reporting
insucient evidence to consider OLP as a premalignant condition to reports of a relatively high incidence of malignant transformation. There seems
to be a slightly higher incidence of oral SCC in patients who have OLP than in the general population [45]. Malignant transformation rates range

No single method of treatment is proven to

prevent malignant transformation of leukoplakia
[53]. Despite this, smoking and alcohol cessation
should be emphasized to patients who have leukoplakia to minimize local recurrence and therefore
decrease the incidence of malignancy. Histopathologic examination of the lesion is imperative
before surgical treatment, regardless of the modality selected. CO2 laser-assisted excision certainly
has a role in the denitive management of oral
leukoplakia. Erythroplakia requires excision with
margin control because of the high incidence
of dysplasia or malignancy. Patients who have lichen planus require an incision biopsy followed
by periodic follow-up visits to monitor changes
that might suggest malignant transformation.
References
[1] Neville BW, Day TA. Oral cancer and precancerous
lesions. CA Cancer J Clin 2002;52:195215.
[2] Neville BW, Damm DD, Allen CM, et al. Oral &
maxillofacial pathology. 2nd edition. Philadelphia:
Saunders; 2002. p. 33769.
[3] Silverman S Jr. Demographics and occurrence of
oral and pharyngeal cancers. The outcomes, the
trends, the challenge. J Am Dent Assoc 2001;132:
7S11S.
[4] Ries LAG, Hankey BF, Miller BA, et al. Cancer statistics review 19731988. 1991. National Cancer Institute, Publication #NIH 912789.
[5] World Health Organization. Report from meeting
on histological denition of precancerous lesions.
1973, Geneva, Switzerland.
[6] Shafer WB, Waldron CA. A clinical and histopathologic study of oral leukoplakia. Surg Gynecol Obstet
1961;112:41120.
[7] Reibel J. Prognosis of oral pre-malignant lesions.
Crit Rev Oral Biol Med 2003;14(1):4762.
[8] Bouquot JE, Gorlin RJ. Leukoplakia, lichen planus,
and other oral keratoses in 23,616 white Americans
over the age of 35 years. Oral Surg Oral Med Oral
Pathol 1986;61:37381.

432

REDDI & SHAFER

[9] Waldron CA, Shafer WG. Leukoplakia revisited.

Cancer 1975;36:138692.
[10] Bouquot JE, Whitaker SB. Oral leukoplakia: rationale for diagnosis and prognosis of its clinical subtypes or phases. Quintessence Int 1994;25:13340.
[11] Hansen LS, Olson JA, Silverman S Jr. Proliferative
verrucous leukoplakia. A long-term study of thirty
patients. Oral Surg Oral Med Oral Pathol 1985;60:
28598.
[12] Silverman S Jr, Gorsky M. Proliferative verrucous
leukoplakia. A follow-up study of 54 cases. Oral
Surg Oral Med Oral Pathol Oral Radiol Endod
1997;84:1547.
[13] Pindborg JJ, Renstrup G, Poulsen HE, et al. Studies
in oral leukoplakia.V. Clinical and histologic signs
of malignancy. Acta Odont Scand 1963;21:40714.
[14] Silverman S Jr. Observations on the clinical characteristics and natural history of oral leukoplakia.
J Am Dent Assoc 1968;76:7727.
[15] van der Waal I, Schepman KP, van der Meij EH,
et al. Oral leukoplakia: a clinicopathologic review.
Oral Oncol 1997;33(5):291301.
[16] Holmstrup P, Vedtofte P, Reibel J, et al. Long-term
treatment outcome of oral premalignant lesions.
Oral Oncol 2006;42(5):46174.
[17] Pandey M, Thomas G, Somanathan T, et al. Evaluation of surgical excision of non-homogenous oral
leukoplakia in a screening intervention trial, Kerala,
India. Oral Oncol 2001;37:1039.
[18] van der Hem PS, Nauta JM, van der Waal I, et al.
The results of CO2 laser surgery in patients with
oral leukoplakia: a 25 year follow up. Oral Oncol
2005;41:317.
[19] Ishii J, Fujita T, Komori T. Laser surgery as a treatment for oral leukoplakia. Oral Onc 2003;39:75969.
[20] Chandu A, Smith ACH. The use of CO2 laser in the
treatment of oral white patches: outcomes and factors aecting recurrence. Int J Orl Max Surg 2005;
34(4):396400.
[21] Reichart PA, Philipsen HP. Oral erythroplakia: a review. Oral Oncol 2005;41(6):55161.
[22] Cawson RA, Langdon JD, Eveson JW. Ertyroplasia
(ertyroplakia). In: Cawson RA, editor. Surgical
pathology of the mouth and jaws. Oxford: Wright;
1996. p. 180.
[23] Axell T, Pindborg JJ, Smith CJ, et al. International
Collaborative Group on Oral White Lesions. Oral
white lesions with special reference to precancerous
and tobacco related lesions: conclusions of an international symposium held in Uppsala, Sweden, May
1821, 1994. J Oral Pathol Med 1996;25:4954.
[24] Pindborg JJ, Reichart PA, Smith CJ, et al. Histological typing of cancer and precancer of the oral mucosa. 2nd edition. Berlin: Springer Verlag; 1997.
[25] Scully C. Oral and maxillofacial medicine The basis
of diagnosis and treatment. Edinburgh, London,
New York: Wright. Elsevier Science Ltd.; 2004.
p. 28990.

[26] Mehta FS, Pindborg JJ, Hamner JE III. Report on

oral cancer and precancerous conditions in Indian
rural populations, 19661969. Basic Dental Research Unit. Tata Institute of Fundamental Research, Bombay. Copenhagen: Munksgaard; 1971.
[27] Shafer WG, Waldron CA. Erththroplakia of the oral
cavity. Cancer 1975;36:10218.
[28] Dental Division Ministry of Health. Malaysia. Dental epidemiological survey of adults in Peninsular
Malaysia 1974/75. Dental Division, Ministry of
Health Malaysia 1978;3637:989.
[29] Zain RB, Ikeda N, Razak IA, et al. A national epidemiological survey of oral mucosal lesions in Malaysia. Commun Dent Oral Epidemiol 1997;25:37783.
[30] Lay K, Sein M, Myint A, et al. Epidemiologic study
of 6000 villagers of oral precancerous lesions in Bilugyun: preliminary report. Commun Dent Oral Epdidemiol 1982;10:1525.
[31] Feller L, Altini M, Slabbert H. Pre-malignant lesions
of the oral mucosa in a South Africa sample: a clinicopathological study. J Dent Assoc S Afr 1991;46:
2625.
[32] Shear M. Ertyhroplakia of the mouth. Int Dent J
1972;22(4):46073.
[33] Bouquet JE, Ephros H. Erthroplakia: the dangerous
red mucosa. Pract Periodontics Aesthet Dent 1995;7:
5967.
[34] Daftary DK, Murti PR, Shah HT. Reverse chutta
smoking and palatal lesions. Control of tobaccorelated cancers and oral diseases. In: Gupta PC,
Hamner JE III, Murti PR, editors. International
symposium 1990. Bombay: Oxford University Press;
1992.
[35] Boisnic S, Frances C, Branchet MC, et al. Immunohistochemical study of oral lesions of lichen planus:
diagnostic and pathophysiologic aspects. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1990;70:
4625.
[36] Regezi JA, Deegan MJ, Hayward JR. Lichen planus:
immunologic and morphologic identication of the
sub mucosal inltrate. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 1978;46:4452.
[37] Mollaoglu N. Oral lichen planus: a review. Br J Oral
Maxillofac Surg 2000;38:3707.
[38] Andreasen JO. Oral lichen planus. 1. A clinical evaluation of 115 cases. Oral Surg Oral Med Oral Pathol
Oral Radiol Endod 1968;25:3142.
[39] Scully C, el-Kom M. Lichen planus: review and update on pathogenesis. J Oral Pathol Med 1985;14:
43158.
[40] Thorn JJ, Holmstrup P, Rindum J, et al. Course of
various clinical forms of oral lichen planus. A prospective follow-up study of 611 patients. J Oral
Pathol Med 1988;17:2138.
[41] Marder MZ, Deesen KC. Transformation of oral lichen planus to squamous cell carcinoma: a literature
review and report of case. J Am Dent Assoc 1982;
105:5560.

MANAGEMENT OF ORAL PREMALIGNANT LESIONS

[42] Silverman S Jr, Grith M. Studies on oral lichen

planus. II. Follow-up on 200 patients, clinical characteristics, and associated malignancy. Oral Surg
Oral Med Oral Pathol 1974;37:70510.
[43] Murti PR, Daftary KD, Bhonsle RB, et al. Malignant potential of oral lichen planus: observations
in 722 patients from India. J Oral Pathol Med
1986;15:717.
[44] Barnard NA, Scully C, Eveson JW, et al. Oral cancer
development in patients with oral lichen planus.
J Oral Pathol Med 1993;22:4214.
[45] Silverman S Jr. Lichen planus. Curr Opin Dent 1991;
1:76972.
[46] Gorsky M, Raviv M. Ecacy of etretinate (Tigason) in symptomatic oral lichen planus. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1992;
73:525.
[47] Ferguson MM, Simpson NB, Hammersley N. The
treatment of erosive lichen planus with a retinoid

[48]
[49]

[50]

[51]

[52]
[53]

433

etretinate. Oral Surg Oral Med Oral Pathol Oral

Radiol Endod 1984;58:2837.
Eisen D. The therapy of oral lichen planus. Crit Rev
Oral Biol Med 1993;4:14158.
Bagan JV, Silvestre FJ, Mestre S, et al. Treatment of
lichen planus with griseofulvin. Report of seven
cases. Oral Surg Oral Med Oral Pathol Oral Radiol
Endod 1985;60:60810.
McCreary CE, McCartan BE. Clinical management
of oral lichen planus. Br J. Oral Max Surg 1999;37:
33843.
Krutchkof DJ, Cutler L, Laskowski S. Oral lichen
planus: the evidence regarding potential malignant
transformation. J Oral Pathol 1978;7:17.
Larson A, Warfvinge G. Malignant transformation
of oral lichen planus. Oral Oncol 2003;39:6301.
Lodi G, Sardella A, Bez C, et al. Systematic review of
randomized trials for treatment of oral leukoplakia.
J Dent Educ 2002;66(8):896902.

Oral Maxillofacial Surg Clin N Am 18 (2006) 435444

Evaluation and Staging of Oral Cancer

Vishtasb Broumand, DMD, MDa,b,*, Teresa E. Lozano, MD, DDSc,
Jesus A. Gomez, DDSa,d
a

Division of Oral and Maxillofacial Surgery, University of Miami School of Medicine,

9380 SW 150th Street, Suite 170, Miami, FL 33157, USA
b
Private Practice, Florida Oral and Facial Surgical Associates, 549 Health Boulevard,
Daytona Beach, FL 32114, USA
c
Private Practice, Oral and Maxillofacial Surgery, 7755 SW 87th Avenue, Suite 100, Miami, FL 33173-2534, USA
d
Department of Surgery, Jackson Memorial Hospital, ACC East, 161 NW 12th Avenue, Miami, FL 33136, USA

Oral cancer is an increasingly prevalent disease

in the United States that accounts for the demise
and deformity of many patients. Most oral cancers are epithelial squamous cell carcinomas
(OSCCA) because most of the risk factors injure
the most supercial layers of the mucosa and gingiva. These areas experience a constant turnover
and prolonged exposure to risk factors, including
tobacco, alcohol, viruses, drugs, and betel nut
(alone or in combination), which cause genetic
mutations that ultimately result in an increase in
incidence and prevalence of the disease. Non
squamous cell, nonsalivary gland oral mucosal
malignancies are generally rare. These lesions
arise from mesodermal and endodermal tissues
and include melanoma, which usually metastasizes via the hematogenous route. This article reviews the evaluation and staging of patients who
are diagnosed with OSCCA. It is important to
perform a thorough evaluation using all pertinent
diagnostic modalities to provide optimum treatment options and bring it to review by a multidisciplinary team before initiation of therapy.
In 2005, an estimated 29,370 new cases of
OSCCA were diagnosed in the United States,
and unfortunately, approximately 7320 people
may succumb to the disease [1]. OSCCA accounts
for 3% of all malignancies in men and 2% in
* Corresponding author. Florida Oral and Facial
Surgical Associates, 549 Health Boulevard, Daytona
Beach, FL 32114.
E-mail address: vbrouma1@newssun.med.miami.edu
(V. Broumand).

women [2] and occurs more frequently in African

American and Hispanic groups than in whites [3].
More than 90% of oral cancers occur in patients
over the age of 45 years. The incidence increases
steadily with age until 65 and then levels o. Prior
reports had noted a substantial increase in the incidence of OSCCA in the United States during
1973 to 1997 among adults younger than 40 years,
particularly tongue cancer [4]. Over the last two decades, however, there has been a slight decrease in
incidence and mortality rates in the United States.
The primary risk factors for OSCCA in American men and women are tobacco (cigarettes and
smokeless) and alcohol use. Infection with human
papillomavirus 16 has been associated with an
increased risk of developing squamous cell carcinoma of the oropharynx [5].
Primary OSCCA lesions are usually noticed by
patients, dentists, and physicians as an ulcer or
a mass, and community screening examination
can be made more ecient by focusing attention
on high-risk sites where 90% of such lesions arise:
oor of mouth, ventrolateral aspect of tongue,
and soft palate [6]. An inspection of the oral cavity is often part of the physical examination by
a physician, and it has been noted that high-risk
individuals visit their medical physicians more frequently than they visit their dentists. Although
physicians are more likely to provide risk factor
counseling (eg, tobacco or alcohol cessation),
they are less likely than dentists to perform
a proper oral cancer examination [7]. Overall,
only a fraction (approximately 20%) of Americans receives a focused oral cancer examination.

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.07.001

oralmaxsurgery.theclinics.com

436

BROUMAND

African Americans, Hispanics, and individuals

who have a lower level of education are less likely
to receive such an examination because of a general lack of access to medical care [7]. A proper
oral examination should place emphasis on identifying any mucosal abnormality, especially white
lesions, red lesions, ulcers, or masses, which can
progress to OSCCA [810]. Although easily detected and often cured in early stages, most oral
cancers are moderately advanced (regional stage)
at the time of diagnosis, and unfortunately, this
pattern has not changed over time [11].
The TNM system
The TNM system for the classication of
malignant tumors was developed by Pierre Denoix in France between 1943 and 1952 and
eventually was adapted by the International
Union against Cancer [12]. TNM classication
has proved useful for the staging of most cancers.
It has withstood the test of time and evolved to
meet new expectations, and it was revised to include more specic delineations. The classication
system is recognized worldwide, and the newer
sixth version was published in 2002. It is an anatomic/clinically based staging system that includes
surface dimension of the primary tumor, involvement of regional cervical lymph nodes, and distant
metastases. It does not take into account the depth
of primary tumor or histopathologic grade.
Principles
Early observation by Denoix and others revealed the fact that cancer survival rate was higher
for patients with localized disease than for patients with disease beyond the site of origin. This
nding called for a classication system in which
patients were grouped depending on the primary
site of disease. The objectives of such classication
are as follows:
To aid the clinician in treatment planning
To provide prognostic value
To evaluate the results of treatment
To facilitate exchange of information between
surgical teams
To contribute to the continuing investigation
of human cancer
To meet these objectives, a classication system is needed whose basic principles are applicable to all sites regardless of treatment and that
can be supplemented in the future by information
that becomes available from histopathology or

et al

surgery. The TNM meets these requirements

eectively and is invaluable in the evaluation of
patients who have OSSCA.
General rules of the TNM system
TNM system is based on the assessment of
three main components: (1) T, surface diameter of
the primary tumor, (2) N, absence or presence and
extent of regional/cervical lymph node metastasis,
and (3) M, absence or presence of distant metastasis. These three components are further subdivided based on the surface diameter of the
primary tumor, presence or absence of cervical
lymph node involvement, and distant metastases
into T0, T1, T2, T3, T4; N0, N1, N2, N3 and M0,
M1 subclasses (Boxes 13). Physical examination
and imaging can be used to delineate each of these
components. This system can be considered
a shorthand notation for describing the

Box 1. T staging for tumors of the lip

and oral cavity
TX Primary tumor cannot be assessed
T0 No evidence of primary tumor
Tis Carcinoma in situ
T1 Tumor %2 cm in greatest dimension
T2 Tumor >2 cm but not >4 cm in
greatest dimension
T3 Tumor >4 cm in greatest dimension
T4a
Lip Tumor invades through cortical
bone, inferior alveolar nerve, floor of
mouth, or skin of face (ie, chin or
nose)*
Oral Tumor invades through cortical
bone, into deep [extrinsic]
Cavity Muscle of tongue (genioglossus,
hyoglossus, palatoglossus, and
styloglossus), maxillary sinus, or skin
of face
T4b Tumor involves masticator space,
pterygoid plates, or skull base and/or
encases internal carotid artery
* Superficial erosion alone of bone/tooth
socket by gingival primary is not sufficient to
classify as T4.
From Patel SG, Shah JP. TNM staging of
cancers of the head and neck: striving for uniformity among diversity. CA Cancer J Clin 2005;
55(4):24258; with permission.

437

EVALUATION AND STAGING OF ORAL CANCER

Box 2. N staging for all head and neck

sites except the nasopharynx and larynx

Box 3. M staging for head and neck

tumors

Nx Regional lymph nodes cannot be

assessed
N0 No regional lymph node metastasis
N1 Metastasis in a single ipsilateral
lymph node, %3 cm in greatest
dimension
N2 Metastasis in a single ipsilateral
lymph node, >3 cm but not >6 cm in
greatest dimension; or in multiple
ipsilateral lymph nodes, none >6 cm in
greatest dimension; or in bilateral or
contralateral lymph nodes, none >6 cm
in greatest dimension
N2a Metastasis in a single ipsilateral
lymph node >3 cm but not >6 cm in
greatest dimension
N2b Metastasis in multiple ipsilateral
lymph nodes, none >6 cm in greatest
dimension
N2c Metastasis in bilateral or
contralateral lymph nodes, none >6 cm
in greatest dimension
N3 Metastasis in a lymph >6 cm in
greatest dimension

Mx Distant metastasis cannot be

assessed
M0 No distant metastasis
M1 Distant metastasis
Mx Distant metastasis cannot be
assessed
M0 No distant metastasis
M1 Distant metastasis

From Patel SG, Shah JP. TNM staging of

cancers of the head and neck: striving for uniformity among diversity. CA Cancer J Clin 2005;
55(4):24258; with permission.

location, extent, and metastasis of a particular tumor at a given point in time [13]. Carcinomas that
arise in minor salivary glands of the oral cavity
are classied according to the rules of this system
in addition to oral mucosal tumors.
Each primary OSCCA lesion can be described
using the following details from the TNM system
and histopathologic grade:
Anatomic site and subsites when appropriate
Involvement of regional lymph nodes
TNM clinical classication
pTNM: pathologic classication
G: histopathologic grading
Stage grouping

From Patel SG, Shah JP. TNM staging of

cancers of the head and neck: striving for uniformity among diversity. CA Cancer J Clin 2005;
55(4):24258; with permission.

cavity, including those of minor salivary glands.

There should be histologic conrmation of the
disease [13].
Anatomic sites and subsites
Lip (C00)
1. External upper lip
(C00.0)
2. External lower lip
(C00.1)
3. Commissures (C00.6)

(vermillion

border)

(vermillion

border)

Oral cavity (C02-C06)

1. Buccal mucosa
(i) Mucosa of upper lip (C00.3, 4)
(ii) Cheek mucosa (C06.0)
(iii) Retromolar area (C06.2)
(iv) Bucco-alveolar sulci, upper and lower
(vestibule of mouth) (C06.1)
2. Upper alveolus and gingival (C03.0)
3. Lower alveolus and gingival (C03.1)
4. Hard palate (C05.0)
5. Tongue
(i) Dorsal surface and lateral borders anterior to vallate papillae (anterior two
thirds) (C02.0, 1)
(ii) Inferior (ventral) surface (C02.2)
6. Floor of the mouth (C04)

Regional/cervical lymph node involvement

Lip and oral cavity (ICD-0 C00, C02-C06)
The classication applies only to OSCCA of
the vermillion surface of the lips and the oral

The denition of N stage for all head and

neck sites, except nasopharynx and thyroid, are
the same. Midline nodes are considered ipsilateral

438

BROUMAND

et al

nodes except in the thyroid. Details of N

staging can be found in Box 2.

Table 1
Stage grouping for all head and neck tumors except nasopharynx and thyroid

Distant metastases

Stage group

T stage

N stage

M stage

M stage refers to distant metastases beyond

the connes of the cervical nodes (Box 3), and
substages M1 and pM1 may be specied further
according to the following system:

0
I
II
III

Tis
T1
T2
T3
T1
T2
T3
T4a
T4a
T1
T2
T3
T4a
T4b
Any T
Any T

N0
N0
N0
N0
N1
N1
N1
N0
N1
N2
N2
N2
N2
Any N
N3
Any N

M0
M0
M0
M0
M0
M0
M0
M0
M0
M0
M0
M0
M0
M0
M0
M1

Pulmonary (PUL) (C34)

Bone marrow (MAR) (C42.1)
Osseous (OSS) (C40, 41)
Pleura (PLE) (C38.4)
Hepatic (HEP) (C22)
Peritoneum (PER) (C48.1, 2)
Brain (BRA) (C71)
Adrenals (ADR) (C74)
Lymph nodes (LYM) (C77)
Skin (SKI) (C44)
Others (OTH)
Histopathologic grading
This category in not included in the TNM
system, and the details of the G categories
apply to all head and neck sites except thyroid.
Gx:
G1:
G2:
G3:
G4:

Grade of dierentiation cannot be assessed

Well dierentiated
Moderately dierentiated
Poorly dierentiated
Undierentiated

R classication
The absence or presence of residual tumor
after treatment is denoted by the symbol R. The
denitions of the R classication apply to all head
and neck sites.
Rx: Presence of residual tumor cannot be
assessed
R0: No residual tumor
R1: Microscopic residual tumor
R2: Macroscopic residual tumor [13]

IVA

IVB
IVC

From Patel SG, Shah JP. TNM staging of cancers of

the head and neck: striving for uniformity among diversity.
CA Cancer J Clin 2005;55(4):24258; with permission.

 Stage II
T2, N0, M0
 Stage III
T3, N0, M0
T1, N1, M0
T2, N1, M0
T3, N1, M0
 Stage IVA
T4a, N0, M0
T4a, N1, M0
T1, N2, M0
T2, N2, M0
T3, N2, M0
T4a, N2, M0
 Stage IVB
Any T, N3, M0
T4b, any N, M0
 Stage IVC
Any T, any N, M1 [14]
Evaluation of patients with oral cancer

AJCC stage groupings

The stage grouping for all head and neck tumors except nasopharyngeal and thyroid tumors
can be found in Table 1.
 Stage 0
Tis, N0, M0
 Stage I
T1, N0, M0

Initial evaluation
The evaluation for a patient with OSCCA
must include history and general physical examination, detailed head and neck examination,
histopathologic tissue diagnosis, radiographic
evaluation, and psychosocial assessment [14,15].
Patients may present with a presumed diagnosis,
and complete evaluation and review of the

EVALUATION AND STAGING OF ORAL CANCER

histopathology can yield a denitive diagnosis and

optimize the treatment plan with respect to adjuvant therapies, such as radiation and chemotherapy. Duration and location of symptoms related
to the tumor may suggest involvement of adjacent
structures and spaces. For example, trismus may
suggest masticator space involvement, and inferior alveolar nerve involvement may manifest as
hypoesthesia or paresthesia [16]. A common presenting symptom in oral cancer is the presence
of a painful lesion in the mouth. Other common
symptoms are bleeding, nonhealing ulcers, and
presence of a mass in the oral cavity. Late-stage
symptoms include cranial nerve decits, loose
teeth, ill-tting dentures, trismus, dysphagia,
weight loss, and hoarseness.
History
The presence of etiologic risk factors for
development of OSCCA, such as tobacco use
(including smokeless and smoked variants) and
alcohol dependence and abuse, increases the
likelihood of primary cancer and recurrence. The
pack year history of smoking is less valuable as
a predictor in the development of oral cancer; the
key feature seems to be the age at which patients
began smoking. Laboratory oral cancer induction
studies have demonstrated that the rate of cancer
formation has a linear relationship to the frequency of the carcinogen contact on a mucosal
tissue surface and by a square function to the dose
of carcinogen [17]. A review of the patient medical
and surgical history and identication of the risk
factors and identication of psychosocial needs
of the cancer patient reect on the overall health
of the patient. This information allows identication of comorbidities and operative risk data that
aect the nal treatment plan and diagnosis. Meticulous history taking allows consideration of all
diagnostic and therapeutic possibilities. Generally, abuse of tobacco and ethanol is underreported by patients, and occupational use of and
exposure to recreational drugs may not be divulged unless prompted by the clinician. Occupational exposure to heavy metals (eg, nickel),
history of human papillomavirus infections, previous head and neck radiation, and use of oral
chews (eg, betel nuts and slake lime) also must
be considered if indicated [18].
Physical examination
Physical examination begins with a complete
evaluation of the head and neck, including

439

a thorough examination of the oral cavity. Noteworthy characteristics of typical primary OSCCA
tumors, including size, precise location, appearance, texture, color, xation to bone/adjacent
structures, and presence of metachronous lesion,
should be noted. Inspection and palpation of all
mucosal surfaces, skin, scalp, tongue, hard and
soft palate, dentition, cervical nodes, and the
cranial nerves (specically cranial nerves V, VII,
X, XI, and XII) should be performed [19].
Complete examination of the cranial nerves is
performed, exploring tongue mobility for hypoglossal nerve function. The facial and spinal accessory nerve also must be evaluated because these
structures can be involved by OSCCA. Areas of
leukoplakia or erythroplakia with ulceration and
masses that are rm or xed to neighboring structures always raise the index of suspicion for malignancy [14,20,21].
Palpation of the neck is critical because the
presence of cervical nodal metastasis is the single
most reliable prognostic factor in patients who
have OSCCA [22]. The number, site, and size of
lymph nodes should be noted. The nodal groups
at risk for metastatic disease in early stage oral
cancer are levels I, II, and III [23]. In patients
with cervical metastasis at the time of diagnosis,
5-year survival rate is reduced by approximately
50% [22,24]. Cervical nodal metastasis is noted
on initial examination in approximately 30% of
patients who have OSCCA. Studies have shown
that the sensitivity, specicity, and accuracy of detection of cervical neck metastasis by clinical examination are 70%, 65%, and 68%, respectively
[25].
Certain positive or negative ndings on physical examination may yield operative risk data and
help formulate an appropriate treatment plan,
especially in consideration of possible sacrice of
the involved vital structures. The presence of
trismus or decreased mobility of the tongue is
a sign of invasion of the pterygomaxillary space or
deep tongue muscles. Perineural invasion can be
assessed via evaluation of sensation of the cheeks,
lips, chin, palate, and alveolar gingiva. The incidence of perineural invasion as a result of
squamous cell carcinoma of the oral cavity
(SCCOC) has been reported to be 27%, although
other authors have reported this incidence to be as
high as 52% [26,27].
Some patients require examination under anesthesia, direct laryngoscopy, and esophagoscopy
for biopsy, better visualization, and accurate
evaluation and staging, especially when symptoms

440

BROUMAND

referable to the oropharynx, hypopharynx, or

larynx or esophagus exist or when the clinician
is unable to assess the extent of the disease with
history, physical examination, and imaging techniques [2831]. With many clinicians introducing
exible beroptic technology into their oces, endoscopic assessment can be used not only for
routine screening and to identify possible synchronous tumors but also for assessment of extent of
disease [32].
Radiographic evaluation
Evaluation of deep tissue involvement of oral
cancer, presence of cervical lymphadenopathy,
and further evaluation of primary tumor often
requires use of several imaging modalities. Plain
lm radiographs, such as Panorex, occlusal view,
and anteroposterior and oblique views of the
mandible and maxilla may only demonstrate gross
bone involvement (Fig. 1). Plain lms are not useful for routine screening because they fail to show
early cortical invasion. Because of technical reasons, the midline lingual cortex of the mandible
cannot be evaluated adequately by a Panorex.
Periapical and occlusal lms may augment a Panorex in demonstrating invasion by tumor arising
in proximity to the maxillary or mandibular
arch [33].
CT scans with intravenous contrast are the
most common imaging modality used in the
assessment of deep tissue extension of tumors of
the oral cavity [34]. CT scans may not be optimal
for routine evaluation of all oral cavity tumors, especially if they are obscured by scatter artifact.
Soft tissue and bone windows that demonstrate
contrast enhancement can show tumor invasion
and bone destruction. CT scans clearly demonstrate bone changes, such as cortical destruction
of skull base and mandible, and tumor invasion
in the mandibular canal. The ability of CT to distinguish tumor from fat makes it useful in

Fig. 1. Panorex of patient with gingival carcinoma of

anterior mandible demonstrates gross bone involvement.

et al

evaluating tumors in the oral cavity and the

head and neck. Without the use of intravenous
contrast it is dicult to delineate between tumor
and muscle or vessels because they have similar
densities.
Advantages of CT include availability at all
centers, progressively lower cost, and good soft
tissue discrimination. With the advent of software
such as the DentaScan, one can obtain better
images of the mandible, maxilla, and teeth. CT
scans also are better tolerated than MRI scans by
most patients [35].
The decision to use MRI versus CT should be
based on the required information needed for
proper surgical planning. MRI is superior in
dening soft tissue details, has multiplanar imaging capability, and can better demonstrate intracranial extension of tumor [36]. Bony details
are not clear on MRI; however, MRI easily discerns subtle variations in soft tissue and distinguishes inammatory changes from brosis or
recurrent tumor [35]. MRI uses radio waves and
magnets, and a key advantage of MRI is the ability to produce images in any orientationdaxial,
coronal, sagittaldwithout the need for patient repositioning. Imaging times are still longer with
MRI; however, resolution comparable to that of
CT can be achieved with MRI. If any bone resection is anticipated in the treatment plan, bone involvement must be assessed properly by CT or
MRI. MRI can be used in conjunction with or
in place of CT and can be superior in certain
cases, especially in patients with numerous dental
amalgam restorations.
Ultrasound has been used with limited success
in evaluation of oral cancer, although neck nodes
can be evaluated easily. Ultrasound technology
remains inexpensive and readily accessible and has
gained some popularity as an initial diagnostic
tool for evaluation of the neck before obtaining
further imaging of the neck [37]. Margins of nodes
along with the size can be determined, which
makes ultrasound a useful guide in performing
ne needle aspirations for obtaining cytology
[38]. With smaller ultrasound probes one can determine whether a lesion is cystic or solid, as
long as there is no bone involvement. Bone does
not transmit sound, therefore the mandible or
maxilla cannot be evaluated.
Positron emission tomography (PET), which is
a form of nuclear medicine study, has been used to
dierentiate malignant disease from recurrent
tumor and is currently being used in some institutions to identify nodal metastasis and

EVALUATION AND STAGING OF ORAL CANCER

recurrent disease (Fig. 2) [3941]. Fluoro-deoxyglucose (FDG) is an imaging isotope used in

FDG-PET using the dierence in metabolism of
radiolabeled glucose molecules between normal
and malignant tissue [42]. The clinical use of this
modality is limited by cost and availability. PET
is proving to be ecacious, with even superior
sensitivity and specicity to CT or MRI in detection of recurrent tumors and distinguishing tumors from postradiation therapy aects and
scarring as a result of oncologic surgery [42,43].
The detection of tumor by PET depends on
anity of tumor for FDG, ability to dierentiate
normal from abnormal FDG uptake, and tumor
size. Malignant lesions of at least 1 cm in size are
generally detected by current PET scanners if the
tumor takes up FDG. Tumors are variable in
their FDG uptake; tumors of salivary gland
origin make PET an unreliable imaging modality.
Other confounding factors are that muscle can
have intense physiologic uptake of FDG, and
tumors with a decline in number of viable cells
have variable FDG uptake. The ability to distinguish abnormal from physiologic FDG uptake
along with tumor detection has been enhanced
greatly with the development of combined PET/
CT scanners. These combined anatomic CT
images, along with functional PET images, have
increased the ecacy of PET in accurate tumor
detection [44]. Current trends are targeting

Fig. 2. PET scan of patient with history of right

mandibular gingival carcinoma with metastatic 1.5 
1.0 cm lesion at right skull base.

441

investigational use of PET/CT for directing external beam radiation, brachytherapy, and interstitial radiation and monitoring therapeutic eects
of therapy [45,46]. PET use is currently limited
by higher cost and is being used for select cases
in which CT or MRI cannot produce enough information alone.
Metastatic evaluation
Oral cancer is generally considered a regional
disease; however, the possibility of systemic metastasis should not be overlooked [21]. Even with
metastatic disease, SCCOC tends to remain localized above the clavicle [47,48]. Thirty percent of
patients who have SCCOC present with cervical
metastasis on initial evaluation (Fig. 3) [49].
The most common sites of distant metastasis
include the lungs (66%), bone (22%), and the liver
(9.5%) [50]. Advanced disease stage and lymphatic or vascular invasion by the primary tumor
are associated with an increased rate of distant
metastasis. [50,51]. Anteroposterior and lateral
view chest radiography generally has been considered to be adequate for routine screening; however, pulmonary metastasis does occur in 15%
to 20% of patients who eventually succumb to
their disease [52]. More than 90% of patients diagnosed with distant metastasis die within 2 years.
The appropriate metastatic evaluation of patients
with low risk of metastasis is a yearly chest radiograph and serum liver function tests when indicated. To assess for cervical metastases in
patients with a clinically negative neck node, one
must consider obtaining CT scans, including oral
cavity and neck, when clinically indicated.

Fig. 3. CT scan with intravenous contrast of neck in

patient with SCCOC with left neck cervical metastases.
Note the hypodense center and the hyperdense capsule
of the lymph node.

442

BROUMAND

In patients with advanced stage cancer or

individuals with locoregional failure, the risk of
distant metastasis is still relatively low at 10%,
and PET or CT scan of the lungs is recommended
for assessment of distant metastasis in patients
with an abnormal chest radiograph [50,53,54].
CT scan of the abdomen can be used to rule
out metastasis to the liver in patients with
abnormal liver enzymes. Bone scans are not
recommended in asymptomatic patients and are
of little value in ruling out subclinical bone
metastasis in cases of SCCOC [55,56]. Several
studies compared the usefulness of PET with ultrasound, CT, and MRI in detection of lymph
node metastasis in SCCOC. Ultrasound had the
highest sensitivity (84%), and PET had the highest specicity (82%) [25].
Other studies have shown that the detection rate
of cervical lymphadenopathy increases from 75%
with physical examination alone to 91% when
physical examination is combined with CT [57].
In patients with known oral cancer and cervical metastasis, ne needle aspiration biopsy has
limited diagnostic value in the metastatic evaluation. Fine needle aspiration biopsy is useful for
patients with neck masses of unknown cause
without obvious evidence of oral cancer, because
it may spare patients from an open biopsy [58,59].
Several authors have examined the accuracy of
a staging supraomohyoid neck dissection as an
invasive diagnostic modality to assess lymph node
metastasis in SCCOC. Currently the staging
supraomohyoid neck dissection is the best modality for detecting cervical metastasis, with an
estimated accuracy in detection of regional metastasis of 98%, a sensitivity of 95%, and specicity of 100% [60,61].
The following list is a guideline for evaluation:
History and physical
Biopsy
Chest radiograph or chest CT
As indicated, Panorex or head and neck CT
Examination under anesthesia (if indicated)
Paresthesia studies
Dental evaluation [62]
Other imaging studies as indicated
Multidisciplinary consult as necessary [62]
Summary
Oral cancer is a prevalent disease in the United
States and accounts for the death and deformity
of many patients every year. As time progresses

et al

there are more new cases. Adequate evaluation

and staging are paramount in providing patients
with the best chance of cure and survival. During
diagnosis and treatment, it is recommended that
a multidisciplinary team discusses each particular
case to integrate the opinion and expertise of each
specialty involved. These opinions help to nd the
best possible solution regarding a patients needs
given what is available at the time. Modern
medical technology gives us valuable means for
this purpose and will continue to evolve and
provide even more possibilities to help our patients. It is important to continue to stress prevention, screening, and adequate diagnosis and
treatment modalities for oral cancer.

References
[1] American Cancer Society. Cancer facts and gures
2005. Atlanta (GA): American Cancer Society; 2005.
[2] American Cancer Society. Cancer facts and gures
2004. Atlanta (GA): American Cancer Society; 2004.
[3] Ries LA, Kosary CL, Hankey BF, et al, editors.
SEER cancer statistics review 19731995. Bethesda
(MD): National Cancer Institute; 1998.
[4] Schantz SP, Yu GP. Head and neck cancer incidence
trends in young Americans, 19731997, with a special analysis for tongue cancer. Arch Otolaryngol
Head Neck Surg 2002;128(3):26874.
[5] Mork J, Lie AK, Glattre E, et al. Human papillomavirus infection as a risk factor for squamous-cell carcinoma of the head and neck. N Engl J Med 2001;
344(15):112531.
[6] Mashberg A, Barsa P. Screening for oral and oropharyngeal squamous carcinomas. CA Cancer
J Clin 1984;34(5):2628.
[7] Kerr AR, Changrani JG, Gany FM, et al. An academic dental center grapples with oral cancer disparities: current collaboration and future opportunities.
J Dent Educ 2004;68(5):53141.
[8] Chiodo GT, Eigner T, Rosenstein DI. Oral cancer
detection: the importance of routine screening for
prolongation of survival. Postgrad Med 1986;80(2):
2316.
[9] Sudb J, Kildal W, Risberg B, et al. DNA content as
a prognostic marker in patients with oral leukoplakia. N Engl J Med 2001;344(17):12708.
[10] Poh CF, Zhang L, Lam WL, et al. A high frequency
of allelic loss in oral verrucous lesions may explain
malignant risk. Lab Invest 2001;81(4):62934.
[11] Miller BA, Kolonel LN, Bernstein L, et al. Racial/
ethnic patterns of cancer in the United States
19881992. Bethesda (MD): National Cancer Institute. NIH Pub. No. 96-4104, 1996.
[12] Marx RE, Stern D. Oral and maxillofacial pathology: a rationale for diagnosis and treatment.

EVALUATION AND STAGING OF ORAL CANCER

[13]
[14]

[15]

[16]

[17]

[18]

[19]

[20]
[21]
[22]

[23]

[24]
[25]

[26]

[27]

[28]

[29]

Hanover Park (IL): Quintessence Publishing; 2003.

p. 2849.
Denoix PF. Nomenclature des Cancer. Bull Inst Nat
Hyg (Paris) 1944;6973.
TNM classication of malignant tumours. In:
Sobin LH, Wittekind C, editors. 6th edition. New
York: Wiley-Liss; 2002.
Patel SG, Shah JP. TNM staging of cancers of the
head and neck: striving for uniformity among diversity. CA Cancer J Clin 2005;55(4):24258 [quiz:
2612, 264].
American Joint Committee on Cancer. AJCC cancer
staging manual. 6th edition. New York: SpringerVerlag; 2002.
Pelczar BT, Weed HG, Schuller DE, et al. Identifying high-risk patients before head and neck oncologic surgery. Arch Otolaryngol Head Neck Surg
1993;119:861.
Chen YK, Huang HC, Lin LM, et al. Primary
oral squamous cell carcinoma: an analysis of
703 cases in southern Taiwan. Oral Oncol 1999;
35:1739.
Mashberg A, Meyers H. Anatomical site and size of
222 early asymptomatic oral squamous cell carcinomas: a continuing prospective study of oral cancer.
II. Cancer 1976;37:214957.
Craig RM. Speckled leukoplakia of the oor of the
mouth. J Am Dent Assoc 1981;102:6902.
Pindborg JJ. Oral cancer and precancer. Bristol:
John Wright and Sons, Ltd; 1980.
Johnson JT, Barnes L, Myers EN, et al. The extracapsular spread of tumors in cervical node metastasis.
Arch Otolaryngol 1981;107:7259.
Byers RM, Newman R, Russell N, et al. Results of
treatment of squamous carcinoma of the lower
gum. Cancer 1981;47:22368.
Landis SH, Murray T, Bolden S, et al. Cancer statistics. CA Cancer J Clin 1999;1999(49):831.
Stuckensen T, Kovacs AF, Adams S, et al. Staging of
the neck in patients with oral cavity squamous cell
carcinomas: a prospective comparison of PET, ultrasound, CT and MRI. J Craniomaxillofac Surg 2000;
28:31924.
Soo K, Carter RL, OBrien CJ, et al. Prognostic implications of perineural spread in squamous carcinoma of the head and neck. Laryngoscope 1986;
96:1145.
Fagan JJ, Collins B, Barnes L, et al. Perineural invasion in squamous cell carcinoma of the head and
neck. Arch Otolaryngol Head Neck Surg 1998;124:
637.
Benninger MS, Enrique RR, Nichols RD. Symptom-directed selective endoscopy and cost containment for evaluation of head and neck cancer. Head
Neck 1993;15:5326.
Shaha AR, Hoover EL, Mitrani M, et al. Synchronicity, multicentricity, and metachronicity of
head and neck cancer. Head Neck Surg 1988;10:
2258.

443

[30] Hordijk GJ, Bruggink T, Ravasz LA. Panendoscopy: a valuable procedure? Otolaryngol Head
Neck Surg 1989;101:4268.
[31] Henle G, Henle W. Epstein-Barr virus-specic IgA
serum antibodies as an outstanding feature of nasopharyngeal carcinoma. Int J Cancer 1976;17:117.
[32] Lin DT, Cohen SM, Coppit GL, et al. Squamous cell
carcinoma of the oropharynx and hypopharynx.
Otolaryngol Clin North Am 2005;38(1):5974.
[33] King JM, Caldarelli DD, Petasnick JP. Dentascan:
a new diagnostic method for evaluating mandibular
and maxillary pathology. Laryngoscope 1992;102:
37987.
[34] Dillon WP, Harnsberger HR. The impact of radiologic imaging on staging of cancer of the head and
neck. Semin Oncol 1991;18:64.
[35] Piollet H, Lufkin RB, Hanafee W. Magnetic resonance imaging of tumors of the upper aerodigestive
tract. Oncology 1989;3:93.
[36] Di Martino E, Nowak B, Hassan HA, et al. Diagnosis and staging of head and neck cancer: a comparison of modern imaging modalities (positron
emission tomography, computed tomography,
color-coded duplex sonography) with panendoscopic and histopathologic ndings. Arch Otolaryngol Head Neck Surg 2000;126:1457.
[37] Koischwitz D, Gritzmann N. Ultrasound of the
neck. Radiol Clin North Am 2000;38:102945.
[38] Van Den Brekel MW, Stel HV, Castelijno JA, et al.
Lymph node staging in patients with clinically negative neck examination by ultrasound and ultrasound
guided aspiration cytology. Am J Surg 1991;162:362.
[39] Ichiya Y, Kuwabara Y, Otsuka M, et al. Assessment
of response to cancer therapy using uorine18-uorodeoxyglucose and positron emission tomography. J Nucl Med 1991;32:1655.
[40] Farber LA, Bernard F, Machtay M, et al. Detection
of recurrent head and neck squamous cell carcinomas after radiation therapy with 218F-uoro2-deoxy-D-glucose positron emission tomography.
Laryngoscope 1999;109:9705.
[41] Anzai Y, Carroll WR, Quint DJ, et al. Recurrence of
head and neck cancer after surgery or irradiation:
prospective comparison of 2-deoxy-2-[F-18] uoroD-glucose PET and MR imaging diagnoses. Radiology 1996;200:13541.
[42] Hanasono MM, Kunda LD, Segall GM, et al. Uses
and limitations of FDG positron emission tomography in patients with head and neck cancer. Laryngoscope 1999;109:8805.
[43] Greven KM, Williams DW III, Keyes JW Jr, et al.
Can positron emission tomography distinguish tumor recurrence from irradiation sequelae in patients
treated for larynx cancer? Cancer J Sci Am 1997;3:
333.
[44] Minn H, Paul R, Ahonen A. Evaluation of treatment response to radio-therapy in head and neck
cancer with uorine-18 uorodeoxyglucose. J Nucl
Med 1988;29:1521.

444

BROUMAND

[45] Bruschini P, Giorgetti A, Bruschini L, et al. Positron

emission tomography (PET) in the staging of head
neck cancer: comparison between PET and CT.
Acta Otorhinolaryngol Ital 2003;23(6):44653.
[46] Langlois D, Hostetter S, Malissard L, et al. Salvage
irradiation of oropharynx and mobile tongue with
192-iridium brachytherapy in Centre Alexis Vautrin.
Int J Radiat Oncol Biol Phys 1988;14:84953.
[47] Snow GB, Balm AJ, Arendse SW, et al. Prognostic
factors in neck node metastasis. In: Larson DL,
Ballantyne AJ, Guillamondegui OM, editors. Cancer in the neck: evaluation and treatment. New
York: MacMillan; 1986. p. 53.
[48] Calhoun KH, Fulmer P, Weiss R, et al. Distant metastases from head and neck squamous cell carcinomas. Laryngoscope 1994;104:199.
[49] Mackay EN, Sellers AH. A statistical review of
carcinoma of the lip. Can Med Assoc J 1964;90:
6705.
[50] Ferlito A, Shaha AR, Silver C, et al. Incidence and
sites of distant metastases from head and neck cancer. ORL J Otorhinolaryngol Relat Spec 2001;63:
2027.
[51] Holsinger FC, Myers JN, Roberts DB, et al. Clinicopathologic predictors of distant metastases from
head and neck squamous cell carcinoma [abstract
200]. Presented at the 5th International Conference
on Head and Neck Cancer. San Francisco, July
2000.
[52] Betka J. Distant metastasis from lip and oral cavity
cancer. ORL J Otorhinolaryngol Relat Spec 2001;
63:21721.
[53] Shaha AR, Spiro RH, Shah JP, et al. Squamous carcinoma of the oor of the mouth. Am J Surg 1984;
148:1004.

et al
[54] Callery CO, Spiro RH, Strong EW. Changing trends
in the management of squamous carcinoma of the
tongue. Am J Surg 1984;148:44954.
[55] Sham JS, Tong CM, Choy D, et al. Role of bone
scanning in detection of subclinical bone metastasis
in nasopharyngeal carcinoma. Clin Nucl Med
1991;16:27.
[56] van Veen SA, Balm AJ, Valdes Olmos RA, et al. Occult primary tumors of the head and neck: accuracy
of thallium 201 single-photon emission computed tomography and computed tomography and/or magnetic resonance imaging. Arch Otolaryngol Head
Neck Surg 2001;127:406.
[57] Merritt RM, Williams MF, James TH, et al.
Detection of cervical metastasis: a meta-analysis
comparing computed tomography with physical examination. Arch Otolaryngol Head Neck Surg 1997;
123:14952.
[58] Shaha A, Webber C, Marti J. Fine-needle aspiration
in the diagnosis of cervical lymphadenopathy. Am J
Surg 1986;152:420.
[59] Karayianis SL, Francisco GJ, Schumann GB. Clinical utility of head and neck aspiration cytology.
Diagn Cytopathol 1988;4:187.
[60] Davidson J, Biem J, Detsky A. The clinically negative
neck in patients with early oral cavity carcinoma:
a decision analysis approach to management.
J Otolaryngol 1995;24:3239.
[61] Byers RM, Wolf PF, Ballantyne AJ. Rationale for
elective modied neck dissection. Head Neck 1988;
10:1607.
[62] National Comprehensive Cancer Network. Clinical
practice guidelines in oncology: head and neck
cancers. Available at: http://www.nccn.org/
professionals/physician_gls/PDF/head-and-neck.pdf.

Oral Maxillofacial Surg Clin N Am 18 (2006) 445463

Imaging of Oral Cavity Squamous Cell Carcinoma

Joseph M. Aulino, MDa,*, Megan K. Strother, MDa,
Jason L. Shipman, MDb
a

Neuroradiology Section, Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center
North, 1161 21st Avenue South, Nashville, TN 372322675, USA
b
Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center North,
1161 21st Avenue South, Nashville, TN 372322675, USA

Ninety percent of oral cavity cancers are

squamous cell carcinoma (SCC) [1]. Most are obvious on clinical examination, so the focus of imaging is on staging and surveillance. This article
attempts to illuminate our successes and setbacks
in imaging SCC of the oral cavity. Imaging options include panoramic radiography, CT, MRI,
and positron emission tomography (PET). Using
these tools, radiologists contribute information regarding the characteristics of the primary tumor
(T stage, as dened by the American Joint Committee on Cancer Staging) (Table 1) [2], including
tumor dimensions, submucosal extent, muscular
inltration, bony involvement, and neurovascular
invasion, and evaluate for regional lymphatic
spread. Malignancies in dierent locations within
the oral cavity vary in behavior depending on the
site, so that familiarity with the expected pathways of extension is required for accurate
interpretation.
Imaging choices
Imaging strategies vary from patient to patient
because of technical considerations as well as the
location and extent of the primary tumor. Imaging choices should be guided by knowledge of the
potential benets and pitfalls for each modality.
Some general technical considerations for each
modality apply universally and are discussed
briey before turning to patient-specic criteria.
* Corresponding author.
E-mail address: joseph.aulino@vanderbilt.edu
(J.M. Aulino).

CT of the neck should be performed from the

top of the frontal sinuses through the manubrium,
after the administration of intravenous iodinated
contrast, to evaluate for regional lymphatic
spread and to characterize the extent of local
oral cavity disease. Multidetector-row scanning
with faster gantry rotation speeds allows the study
to be completed in less than 10 seconds, signicantly reducing motion artifacts. The helically
acquired volume of image data is processed to
form the standard axial images, which are usually
generated at a 2- to 5-mm slice thickness. Because
of the helical acquisition, the data can generally be
reformatted in the coronal or sagittal plane
without compromising resolution. If bone
invasion is suspected, additional images can be
generated using a lter to improve bone detail
(bone algorithm lter).
The most signicant drawback to CT, when
imaging the oral cavity, is beam-hardening artifact attributable to dental amalgam. Special
techniques to diminish this artifact include
open-mouth imaging [3], which requires an additional acquisition through the oral cavity. An
alternative technique to reduce beam-hardening
artifact uses a postprocessing algorithm, but this
is only available on scanners from certain manufacturers at the present time. Older single-slice
CT scanners possessed the capability of gantry
tilting, allowing angulation of the CT gantry to
perform a buttery technique to scan though
the oral cavity and oropharynx at dierent angles,
changing the location of the metallic streak artifact on each acquisition. Newer multidetector
scanners do not allow for signicant angulation

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.011

oralmaxsurgery.theclinics.com

446

AULINO

Table 1
Oral cancer primary tumor classication system (T)
Stage

Description

TX
T0
Tis
T1
T2
T3
T4 (lip)

Cannot be assessed
No evidence of primary tumor
Carcinoma in situ
2 cm or less in greatest dimension
24 cm
More than 4 cm
Invades through cortical bone,
involves inferior alveolar nerve,
extends to the oor of mouth, or
involves skin of the face
Invades through cortical bone, into
extrinsic muscles of the tongue,
into maxillary sinus, or involves
skin of the face
Invades masticator space, pterygoid
plates, or skull base, or encases
internal carotid artery

T4a (oral cavity)

T4b (oral cavity)

Data from Greene FL, Page DL, Fleming ID, et al,

editors. AJCC Cancer Staging Manual. 6th edition.
New York: Springer; 2002.

of the gantry; thus, this maneuver is not an option. Because of the preponderance of beam-hardening artifact in the oral cavity, MRI is generally
preferred for evaluation of an oral cavity primary
tumor. Although dental amalgam may cause some
distortion on MRI scans, it is generally less than
would be seen on CT.
A standard MRI examination of the neck
includes T1-weighted axial and coronal images
before and after contrast. The fat signal is suppressed on postcontrast images to improve the
distinction between enhancing tumor and surrounding tissue. Axial T2-weighted images are
also acquired through the oral cavity and neck,
ideally with a fat-suppression technique. The
examination requires 20 to 45 minutes, depending
on patient compliance, scanner hardware, and
level of detail required. Although MRI of the neck
includes coverage of the oral cavity, dedicated
higher resolution MRI through the face may
provide additional details about the primary
tumor that may be missed with routine neck
MRI. MRI may be limited by motion artifact
from patient swallowing, lip smacking, talking,
and tongue motion. Faster imaging hardware (eg,
3-T magnets and parallel imaging), faster MRI
sequences, and patient education may help to
diminish patient motion.
In patients who do not have extensive dental
hardware or amalgam, the choice between CT and

et al

MRI is more nuanced. MRI is generally preferred

to CT for delineating tumor margins because of its
superior soft tissue contrast. Tumor thickness is
most accurately assessed on T1-weighted sequences before and after contrast. T2-weighted
sequences may overestimate tumor size, because
peritumoral inammation and edema can have
the same bright T2 signal as the tumor [4]. This information is important for treatment planning.
CT is faster, less expensive, and more userindependent than MRI. CT also better demonstrates cortical bone destruction in advanced
tumors and is more sensitive for lymphadenopathy than MRI. Some patients can be stratied by
their clinical histories. MRI should be used for
patients who cannot receive CT contrast (eg,
because of renal insuciency or contrast allergy).
CT should be used in claustrophobic patients or in
patients who cannot easily lie still. CT and MRI
are comparable for T staging of oral cavity
tumors, but MRI provides more information for
surgical planning purposes.
Squamous cell carcinoma of oral cavity by region
Lips
The lips are the most common site of oral
cavity SCC, but disease is usually limited at
presentation. Imaging is not typically indicated
with early SCC of the lips, because the extent of
disease is readily assessed clinically. In more
advanced cases, tumor margins may not be delineated on clinical examination. CT or MRI
should focus on the mental and inferior alveolar
canals for bony destruction and perineural tumor
(PNT) spread, which upstages these tumors to T4.
Oral tongue
The oral tongue includes the anterior two thirds
of the tongue, extending posterior to the circumvallate papillae. Roughly two thirds of oral tongue
cancers are moderately to far advanced at the time
of initial diagnosis. Common (and ominous) patterns of spread include mandibular involvement
and bilateral lymphadenopathy. Tumor extension
across the midline lingual septum may require
a total glossectomy or relegate the patient to
nonsurgical treatment (Fig. 1). On imaging, note
should be made of involvement of the ipsilateral
and contralateral neurovascular bundles (Fig. 2),
which enter the tongue at the level of the tongue
base and extend anteriorly at the level of the sublingual space to supply the tongue. Oral cavity lesions

IMAGING OF ORAL CAVITY SCC

447

Fig. 1. Oral tongue cancer with oor of mouth extension and cervical lymphadenopathy. Axial contrast CT images. (A)
Enhancing tumor in the oral tongue extends to but does not cross the midline. (B) Dilatation of Whartons duct (long
arrow) conrms extension to the oor of mouth on the right. Foci of necrosis are seen within the involved right level IIa
node (short arrows). (C) Intraglandular submandibular duct dilatation (long arrow). Ipsilateral bulky level II lymphadenopathy with area of necrosis (short arrow).

are upgraded to a T4 tumor stage when they

invade bone (usually the mandible but also the
maxillary sinus and skull base) (Fig. 3) or involve
the extrinsic tongue musculature (Fig. 4). The
extent of mandibular involvement determines
the type of mandibulectomy.
Nodal drainage of the oral tongue is to levels I,
II, and III; occasionally, metastatic disease may
bypass these groups and spread directly to level
IV. Cross-drainage of lymphatics in the tongue is
common, increasing the likelihood of contralateral nodal involvement [5].
MRI is generally superior to CT to evaluate
cancers of the oral tongue. T1-weighted pre- and
postcontrast MRI scans in the axial, coronal,
and sagittal planes are particularly well-suited
for the purpose. Additionally, fat-suppressed T2weighted images may be benecial and should
routinely be used to evaluate the oral cavity and
neck. CT may be helpful to conrm the extent of

mandible involvement. Studies have demonstrated

that in the oral tongue, tumor thickness is an
independent prognostic factor to predict nodal
metastases, local recurrence, and patient survival.
Particularly in patients with N0 disease, tumor
thickness signicantly inuences patient care.
Lam and colleagues [4] found that only 8% of patients with a tumor thickness less than or equal to
3 mm had subclinical nodal metastasis. When the
thickness exceeded 9 mm, however, 53% of their
patients had subclinical nodal metastases. Tumor
thickness is well evaluated with coronal and sagittal planes on MRI.
Floor of mouth
The oor of mouth is the most common site of
oral cavity cancer after the lips [6]. The space is divided into the sublingual and submandibular
spaces by the mylohyoid muscle. The sublingual

448

AULINO

et al

Fig. 2. Large oral tongue cancer involves the neurovascular bundle and invades the oor of mouth with cervical and
mediastinal nodal disease. (A) Enhancing tumor is easily identied on the axial images of the contrast CT because it
obliterates normal fat in the anterior tongue. The lingual septum has been breached, and both genioglossus muscles
are involved. The right neurovascular bundle (arrowheads) is encased by tumor. (B) More caudal image from the
same patient conrms tumor extension to the left sublingual space, obstructing Whartons duct (arrow). Bulky submandibular and level IIa nodes are also present. Coronal maximum intensity projection PET image (C) and fused axial PETCT image (D) conrm oral tumor with bilateral cervical lymphadenopathy and identify metastatic involvement of an
anterior mediastinal node in the anteroposterior window (arrows).

space lies above the mylohyoid muscle, and the

submandibular space lies below the mylohyoid
muscle. Within the sublingual space are the sublingual glands, the Wharton submandibular ducts
(Fig. 5), the hypoglossal nerves, and the lingual
arteries and nerves. Because of its fat content,
the sublingual space should demonstrate low attenuation on CT and be T1-hyperintense on
MRI. The submandibular space contains the
main bodies of the submandibular glands. MRI
has been shown to dierentiate accurately between direct invasion of the submandibular gland
and sialectasis [7]. Tumors that arise in the oor of
mouth can spread posteriorly along the mylohyoid muscle to the tongue base and superior to
the oral tongue. Floor of mouth tumors that
spread laterally can involve the mandible
(Fig. 6). Mandibular invasion may be limited to

cortical involvement (obvious destruction or

more subtle saucerization) or may extend to involve cancellous bone. The type of resection is
based on the amount of erosion. Disease limited
to the mandibular cortex may be treated with
a marginal mandibulectomy, but tumor invasion
of marrow necessitates segmental resection [8].
Mandibular involvement may be assessed with
panoramic tomographic radiography, CT, MRI,
and bone scans with single photon emission
computed tomography (SPECT). Panoramic tomographic radiography is insensitive for the detection of marrow invasion. The superior and
inferior cortices of the mandible are imaged
tangentially by this technique, and subtle cortical
erosion in these regions, as by alveolar ridge
carcinoma, may be readily detected. This method
of imaging does not visualize the lingual cortex of

IMAGING OF ORAL CAVITY SCC

449

Fig. 3. Floor of mouth cancer invades the mandible with perineural spread. (A) Contrast axial CT image demonstrates
enhancing tumor in the oor of mouth on the right. The ipsilateral neurovascular bundle is encased by tumor; compare
its appearance with the normal neurovascular bundle on the left (arrowheads), which is surrounded by darker fat. There
is bony destruction of the mandibular symphysis. (B) Bone algorithm axial CT image demonstrates the destruction in the
mandibular symphysis. Prominent enlargement of the left inferior alveolar canal (arrow) reects perineural extension.
Bone and perineural involvement place this tumor at the T4 stage. The patient was treated medically. (C) Contrast axial
CT performed 6 months later demonstrates marked interval worsening of left mandibular involvement, with breakthrough of the buccal cortex (arrowheads).

the mandible in tangent, and lingual (as well as

buccal) cortical bone involvement may be overlooked. CT has the advantage of more accurately
depicting the pattern of bone invasion than
panoramic radiography (Fig. 7), whether it is erosive, invasive, or mixed [9]. Helical CT scanning
allows the imaging data to be reconstructed in
any plane at identical resolution. CT scanner
manufacturers oer postprocessing options to reconstruct cross-section images that are perpendicular to the mandible as well as panoramic images
of the mandible with dental planning software
packages. These may be useful to determine the
extent of mandibular invasion with oral SCC
[10,11].
Brown and Lewis-Jones [12] reviewed the published evidence to support the use of the dierent
modalities to assess mandibular invasion in the

setting of oral cancer. They found nuclear medicine bone scan imaging using SPECT to be quite
sensitive (97%) but not as specic (76%) for
SCC, because noncarcinomatous pathologic ndings within the mandible may cause abnormal uptake. It is essential that the bone scan (with or
without SPECT) be interpreted in conjunction
with other imaging studies that can more accurately localize the tumor. CT is more specic
(86%) than radiography and serves as a useful adjunct to bone scanning. The meta-analysis by
Brown and Lewis-Jones [12] showed that MRI
provides a sensitivity of 85% and specicity of
72%. Bolzoni and coworkers [13] found more
encouraging data to support the use of MRI. In
a recent 2004 study, they found MRI to have
a sensitivity of 93% and specicity of 93%. The
older literature should be viewed cautiously,

450

AULINO

et al

Fig. 4. Oral tongue cancer with sublingual space extension. (A) Axial, fat-suppressed, T2-weighted MRI scan shows high
signal within the tumor, which comprises most of the oral tongue at this level. The left sublingual space fat (dark on this
fat-suppressed image) is invaded on the left; note the preserved, dark, right-sided sublingual fat (arrowheads). The left
mylohyoid muscle is involved (arrow), but there is no bony extension; the sharp T2-hypointense cortical bone appears
preserved. The fatty marrow of the mandible is relatively bright because of incomplete fat suppression. (B) Coronal,
contrast, T1-weighted image with fat saturation demonstrates the thickness of the tumor (asterisks) as well as its clear
demarcation from normal tongue. The diameter of the tumor, which was greater than 4 cm, made this a stage T4 tumor.

because MRI and CT technologies continue to

improve rapidly and older investigations into their
applicability may not be accurate. For example,
CT scans are now routinely obtained at submillimeter resolution, and images of the mandible
can be generated in any plane on the radiology
imaging workstation. This is in contrast to older
published studies that evaluated the use of 3- to

5-mm axial slices to determine bone invasion,

without the ability to reconstruct and adjust the
display window and level settings. If involvement
of the mandibular inferior alveolar canal is identied, the extent of PNT spread should be evaluated
with MRI.
The extent of spread determines treatment of
cancers of the oor of mouth. Tumors that extend
inferiorly along the mylohyoid or hyoglossus
muscle may require a combined transoral and
cervical approach for complete surgical resection.
Oral cavity tumors that spread via the tongue base
to the pre-epiglottic space, vallecula, or tonsillar
region may require a partial pharyngectomy if
treated surgically. The primary lymph drainage
from the oor of mouth is to level I and level II
nodes; occasionally, these nodal groups are bypassed with direct spread to level III nodes [5].
Buccal and alveolar ridge

Fig. 5. Floor of mouth cancer with obstructed submandibular duct. Contrast axial CT image demonstrates
subtle replacement of normal fat in the left sublingual
space by tumor (white parallel lines), which extends to
the mylohyoid muscle (black arrow). The tumor has
obstructed the left submandibular gland (asterisk),
which is mildly enlarged and enhances avidly compared
with the normal right submandibular gland. The
intraglandular ducts are dilated.

The epithelial surface of the cheek and lip is the

buccal mucosa, dened superiorly and inferiorly
by the gingivobuccal sulci and posteriorly by the
pterygomandibular raphe region. Cancer of the
buccal mucosa may extend to deeper structures to
involve the buccinator muscle and buccal space,
including the parotid duct and body of the buccal
fat pad, as well as to the dermis and mandible
(Fig. 8). Extension posteriorly to involve the retromolar trigone (RMT) is also a concern. Small
buccal tumors may not be visible by standard

IMAGING OF ORAL CAVITY SCC

451

Fig. 6. Floor of mouth cancer with mandibular invasion. (A) Axial T1-weighted MRI scan without contrast through the
oor of mouth demonstrates a soft tissue mass on the left (asterisk). The sublingual space T1-hyperintense fat appears
normal (arrowheads). There is bony extension into the left mandible, however, with replacement of normal T1-hyperintense marrow fat (white arrow). Compare with contralateral normal marrow (double black arrows). (B) Coronal T1weighted MRI scan without contrast or fat suppression again demonstrates left mandible marrow replacement (arrow).
(C) With contrast and fat suppression, the enhancing tumor (star) is well demarcated from the normal tongue. The image
contrast between normal and abnormal mandible marrow is lost, however, because both enhance (curved arrow). Note
loss of the normal T1-hypointense bony cortex. Noncontrast T1-weighted images are best suited for evaluation of marrow replacement; dark cortical bone should be inspected on all sequences.

CT or MRI, and the pued cheek dynamic CT

maneuver may help the radiologist to visualize the
lesion [3,14].
The upper alveolar ridge refers to the alveolar
process of the maxilla, and the lower alveolar
ridge refers to the alveolar process of the mandible. The lingual extent of the lower alveolar ridge
mucosa is marked by the free mucosa of the oor
of mouth. In regard to SCC of the gingiva,
invasion of the adjacent buccal space occurs in
42% of the lower gingival tumors and 47% of the
upper gingival tumors [15]. Spread into the masticator space is seen in 14% of all lower gingival tumors (via the RMT and buccal space) but rarely
occurs in SCC arising from the upper alveolar
ridge [15]. Floor of the mouth (sublingual space)
extension may occur with lower gingival tumors
[15,16] and is well depicted by MRI. Eighty-ve

percent of imaged lower alveolar ridge mucosal

cancers invade the mandible (Fig. 9), and 88%
of imaged upper gingival cancers invade the
bony maxillary alveolus [15].
PNT spread from gingival and other oral
cavity tumors may occur with direct mandibular
involvement, aecting the inferior alveolar nerve,
or secondarily from masticator space extension
directly to the third division of the trigeminal
nerve (V3). With inferior alveolar nerve tumor
involvement, CT may reveal asymmetry of the
inferior alveolar canals within the mandible or
loss of the normal fat pad located at the insertion
of the inferior alveolar nerve into the mandibular
foramen. The inferior alveolar nerve within the
mandible displays normal mild enhancement by
MRI, but asymmetric enhancement should raise
the concern for PNT spread.

452

AULINO

et al

Fig. 7. SCC of the right inferior alveolar ridge. (A) Detail of panoramic radiograph of the mandible shows erosion of the
superior cortex of the anterior mandibular body (arrowheads) and ill-dened lucency within the adjacent marrow space
(star). The radiologist interpreting the examination was provided with the history of trauma, and the study was
interpreted as no fracture. The patient underwent an incisional biopsy without curettage. (B) Enhanced CT coronal
reconstruction image displayed in a soft tissue window setting obtained 19 days later shows rim-enhancing tumor
superior to the right mandibular body (black asterisk). (C) Same image with a bone window setting. There is erosion
of the superior right mandibular body involving the inferior alveolar canal. Note the normal left inferior alveolar canal
(arrow).

Evaluation of V3 involvement can be tricky,

and attention should be paid to normal enhancement in V3 as well as to normal variability in the
size of the foramen ovale, as appreciated on CT.
False-positive results can be avoided on MRI by
searching for asymmetric enhancement compared
with the contralateral foramen ovale. Middle
cranial fossa dural and cavernous sinus involvement may be subtle as well and is best visualized
on postcontrast, coronal, fat-suppressed, T1weighted images. As a general rule, MRI of the
face and skull base is more sensitive than CT for
evaluation of PNT spread, which can be clinically
silent. Thus, dedicated MRI of the face and skull
base may be indicated even when there is no
clinical evidence for PNT spread.
Retromolar trigone
The RMT begins behind the last mandibular
molar tooth and extends superior to the maxillary
tuberosity. It is, basically, the mucosal surface of
the posterior mandibular body superior extent
and inferior ramus anterior inferior extent. The

submucosal pterygomandibular raphe runs along

the medial aspect of the RMT, extending from the
mylohyoid ridge of the mandible to the inferior
aspect of the medial pterygoid process. The posterior buccinator muscle and the superior anterior
extent of the superior pharyngeal constrictor muscles attach to the pterygomandibular raphe.
The RMT is a relatively small region, and
carcinomas arising in this location often present
with invasion of adjacent structures, which include the buccal space, the alveolar ridge, the
mandible, the oropharynx (eg, soft palate, palatine tonsil), the pterygoid process of the buccal fat
pad (the bulk of which is posterior to the
maxillary sinus and orbit), the oor of mouth,
and the muscles of mastication (eg, pterygoid,
masseter muscles) (Fig. 10). Infratemporal fossa
and pterygopalatine fossa (PPF) involvement,
from PNT spread or direct invasion, may occur
secondarily (Fig. 11). Bone erosion, PNT spread,
or muscle invasion elevates the T stage of all
oral cancers to T4 [8,17]. Direct tumor spread
may not be visible by physical examination but
is readily identied with MRI [18] and less well

IMAGING OF ORAL CAVITY SCC

453

Fig. 8. Advanced buccal carcinoma. (A) Patient photograph shows a large left buccal mucosal lesion. Reconstructed
coronal enhanced CT images in soft tissue (B) and bone (C) window settings show destruction of the left maxillary
alveolar ridge and lateral hard palate. Invasion of the left maxillary sinus and inferior nasal cavity is evident. Sinus
secretions and mucosal thickening cannot reliably be distinguished from tumor. T2-weighted MRI (not shown) revealed
the secretions as bright signal, separate from the tumor.

with CT [19]. The diagnostic accuracy of MRI to

depict the T stage is 86% [18]. MRI is the preferred modality to evaluate RMT SCC because
of the increased contrast between tumor and the
surrounding soft tissue [17]. Denervation inammatory changes from V3 PNT spread may cause
diuse enhancement of the muscles of mastication, and PET may be required to dene the extent of the primary tumor (Fig. 12).
Hard palate
Although a rare location for primary SCC,
hard palate disease deserves special mention. SCC
involvement of the hard palate most commonly
occurs secondarily to extension from the adjacent
maxillary alveolar gingiva. The size of the primary
tumor is readily assessed with a combination
of clinical examination, CT, or MRI. Subtle
spread of tumor along the palatine nerves can be

overlooked, however. CT (displayed using a bone

window setting) may show asymmetry of the
palatine foramina, which is located roughly at
the junction of the hard and soft palates, and
MRI may show asymmetric enhancement of the
palatine nerves. With advanced PNT spread, there
is loss of the normal fat hypodensity (CT) and T1
hyperintense signal (MRI) within the involved
PPF. Tumor may then spread posteriorly along the
second division of the trigeminal nerve through the
foramen rotundum to involve the cavernous sinus
or superiorly into the inferior and superior orbital
ssures, with further extension possible from these
locations. Perineural extension from the PPF
posteriorly via the vidian nerve is also possible,
and the sharp cortical margins of the vidian canal
should be evaluated carefully to exclude tumor
involvement on CT scans of advanced tumors.
Hard palate SCC may extend superiorly to
invade the maxillary sinus and inferior nasal

454

AULINO

et al

Fig. 9. SCC of the left inferior alveolar ridge invades the mandible. (A) Axial enhanced CT image displayed in a soft
tissue window setting shows destruction of the left mandible with gas within the defect, which is related to ulceration.
(B) Same axial slice displayed with a bone window image setting shows destruction of the lingual cortex, cancellous
bone invasion, and thinning of the buccal cortex.

cavity. Maxillary sinus mucosal disease may be

indistinguishable from invasive tumor by CT [20].
MRI can usually distinguish tumor from retained
paranasal sinus secretions with T2-weighted images, however. Secretions are typically brighter
than tumor on T2-weighted images. T1-weighted
images, before and after contrast, occasionally
contribute information to make the distinction
[21,22].
Lymphadenopathy
Regional nodal disease from oral cavity SCC
usually involves nodal groups I (submandibular,
submental) and II (superior internal jugular
chain) (Table 2). The presence or absence of nodal
metastases greatly aects survival of patients with
head and neck cancer [23,24], and criteria that inuence N staging include laterality of nodes and
nodal dimensions (Table 3). Patients rarely present with distant disease unless there is extensive
tumor. CT is more sensitive than physical examination for detecting involved nodes [2529] and
should complement the clinical evaluation routinely. A recent study by Malard and colleagues
[30] showed CT evidence of nodal involvement
in 30% of patients with clinically N0 necks.
On MRI and CT, lymph nodes are judged by
their morphology and location. CT has been
shown to perform slightly better than MRI for
the detection of nodes involved by metastatic SCC
[31,32], although a recent study has shown MRI
to be comparable to CT when the MRI slice thickness is reduced and the interslice gap is removed

[33]. Tumor necrosis within lymph nodes is the

most specic marker of nodal involvement
(Fig. 13) [17], and routine neck CT is performed
at a higher resolution than routine neck MRI,
allowing for improved detection of this necrosis.
Of note, necrosis can only be appreciated on contrast images (necrosis is missed on unenhanced
studies).
Necrotic nodes are always considered pathologic in the setting of oral SCC, although
specicity may be diminished when a normal fatty
hilum is mistaken for central necrosis because of
volume averaging. Another clue for metastatic
nodal involvement is irregular spiculated nodal
margins on CT or MRI. This sign is highly
suggestive of extracapsular tumor spread in the
absence of prior surgery, radiation, and inammation. CT has generally been more accurate than
MRI in detecting extracapsular extension [31,34],
although MRI may be comparable to CT when
the axial MRI slice thickness is reduced [35].
Various imaging size criteria have been proposed to determine if a cervical node is involved
by tumor. If a larger diameter cuto is used (eg,
15 mm), small involved nodes are not detected
and sensitivity drops. If a smaller cuto is used
(eg, 8 mm), specicity suers. Additionally, differently sized criteria may be applied to the
dierent nodal levels. Castelijns and van den
Brekel [36] have recommended using a maximum
short axis dimension, as measured by ultrasound,
of 7 mm for level II nodes and 6 mm for other cervical nodal groups, based on previous work [37].
To reduce the number of false-positive nodes

IMAGING OF ORAL CAVITY SCC

455

Fig. 10. Large left RMT tumor. (A) Axial enhanced CT image through the level of the maxillary alveolus shows an
enhancing mass extending to the midline, involving the oropharynx. There is destruction of the posterior left maxillary
alveolus. The left muscles of mastication appear diusely involved. (B) Axial, fat-suppressed, T1-weighted MRI scan
reveals irregular enhancement within the mass, the left masseter muscle (star), and the left parotid gland (asterisk).
Parotid gland enhancement is related to Stenson duct obstruction. (C) Axial CT image in a bone window setting shows
subtle periosteal reaction along the lateral margin of the left mandibular ramus (arrowheads). The left pterygoid plates
are destroyed. Note the normal right pterygoid plates (arrows).

that these criteria would produce, these investigators sample the nodes using ultrasound guidance.
When ultrasound, combined with ne-needle aspiration cytology, conrms an N0 neck, the elective
neck dissection may be deferred [36,38,39]. This
approach has gained wider acceptance in Europe
than in North America, where more liberal size
criteria are used and ultrasound-guided ne-needle aspiration is not utilized. Generally, a 10-mm
long-axis dimension (on axial images) is used for
all cervical nodal groups except for level II, where
15 mm is generally used. Others, however, recommend using 15 mm as the maximum dimension in
level I [34]. Cervical node size criteria have not
been validated and serve as a general guideline
for determining regional spread by imaging. Involved subcentimeter nodes are missed, and enlarged, reactive, benign nodes are described as
involved by tumor.

Size should be factored in with nodal shape

when predicting malignancy. Most normal lymph
nodes possess an ovoid shape, and a node that is
more rounded should be viewed with concern.
Rounded nodes may be benign, however, reecting the sequelae of chronic infection and poor
dental hygiene; thus, roundness may not correlate
with malignancy [40]. Because of these drawbacks,
nodal necrosis remains the most specic imaging
nding on CT or MRI.
Another imaging modality that can help to
identify nodal disease is nuclear medicine lymphoscintigraphy. A radiotracer is injected around
the tumor, and gamma camera imaging or a handheld probe is used to localize the node or nodes
draining that region. Surgical treatment can then
be focused on these sentinel nodes. Preliminary
results show that the test is technically feasible
and promising in oral cancer [41].

456

AULINO

et al

Fig. 11. PNT spread from SCC of the left RMT in the same patient as Fig. 10. (A) Axial enhanced CT through the face
shows tumor replacement of the left PPF fat (arrow). Compare with the normal right PPF (double arrows). (B) Axial T1weighted MRI scan at a similar level shows loss of the normal left PPF fat signal (arrow). Compare with the normal right
PPF T1-hyperintense fat (double arrows). (C) Coronal, fat-suppressed, T1-weighted, postcontrast, coronal MRI scan
through the foramen ovale in the same patient. Prominent enhancement is seen throughout the left muscles of
mastication extending superior to the left foramen ovale (arrow). The muscles of mastication may enhance secondarily
from subacute denervation injury from V3 PNT spread or may be involved by direct tumor extension. If management is
aected, PET may be useful to dene the extent of the primary tumor (not shown).

Positron emission tomography

Introduction to positron emission tomography
In recent years, PET has emerged as an
exciting oncologic imaging modality. It has
proven eective in identifying highly metabolically active tumors and is now approved for
diagnosis, staging, and restaging in such tumors
as adenocarcinoma of the breast, non-small cell
carcinoma of the lung, lymphoma, and SCC of the
head and neck. Although its role is still evolving,
it promises to be a powerful tool in the care of
patients with cancer.
PET is based on positron generation from a
positron heavy nucleus. Although many molecules
have been investigated and used on an experimental basis, 18F-uorodeoxyglucose (FDG) remains

the vanguard of PET technology. FDG, a glucose

analogue, is taken up and trapped inside metabolically active cells, where it emits the positron that
ultimately disintegrates into two incident photons
of equal energy (511 keV). These photons travel in
opposite directions and are received by ring detectors surrounding the patient.
The capture of FDG inside cells aords the
observer contrast between metabolically active
and inactive cells. This represents a great advantage in identifying a suspicious mass or node. Yet,
the physics of PET inherently couple power with
limitation. The distance traveled by the positron,
the attenuation by surrounding tissues, and the
intrinsic ineciency of high-energy detectors limit
spatial resolution to 5 mm. Although much
progress has been made in crystal eciency and

IMAGING OF ORAL CAVITY SCC

457

Fig. 12. Right mandibular gingiva SCC with PNT spread along the V3 with denervation inammatory changes of all the
muscles of mastication mimicking tumor invasion. (A) Axial, fat-suppressed, postcontrast, T1-weighted image through
the level of the mandibular condyles shows abnormal enhancement throughout the right masticator space, including the
masseter (asterisk) and lateral pterygoid (star) muscles, suggesting diuse tumor invasion. Axial 18F-uorodeoxyglucose
(FDG)-PET image (B) and corresponding unfused CT transmission image (C) through the same level as (A) show only
a single focus of increased radiotracer uptake (arrow in B) in the expected location of the V3. (D) Coronal PET image
shows marked FDG uptake within the mandibular gingiva primary tumor (black diamond). PNT spread is seen
extending superiorly (arrows). (E) Axial, fat-suppressed, T2-weighted image shows edema within the temporalis muscle
(arrow), supplied by V3, which is an unusual (but not impossible) location for tumor spread. The diuse involvement of
the muscles of mastication suggests denervation as the cause rather than tumor extension; PET conrms the suspicion.

computer-generated reconstruction algorithms to

bring the technology this far, there is always going
to be a gap in the spatial resolution of transmission imaging (CT, MRI) and emission imaging
(PET).
In addition, attempts have been made to
standardize the activity observed in a lesion according to the dose administered and the patients
body weight, such that higher activity might be
a predictor of metabolic activity (ie, prognosis).
The standard uptake value (SUV) is dened as
tissue activity divided by the injected dose activity
per unit of body weight [42]. Although it is easy to
calculate, it should be used clinically only after judicious calibrating of the scanner to a known dose
as well as correcting for dose inltration, residual
dose in the syringe and tubing, body mass index,
glucose metabolism, and uptake time [4345], all

of which vary from patient to patient. Although

many studies have sought to use this SUV to predict malignancy versus benignancy, its use alone
should be viewed with healthy suspicion.
Recently, PET has been used in concert with
CT. PET-CT combines the molecular contrast of
PET with the anatomic precision of CT [4648].
It also allows attenuation correction, which decreases perceived photon scatter from overlying
tissue. Many of the initial studies involving PET
compared it alone with MRI or CT. In this scenario, PET oered mixed results, depending on
the disease process [4952]. More recent studies
involve dedicated PET-CT scanners that combine
the PET detector ring and the CT gantry in
one patient table. The common patient table
allows the images to be integrated for molecular
contrast on CT. Although there is a paucity

458

AULINO

et al

Table 2
Cervical nodal levels

Table 3
Oral cancer nodal classication system (N)

Level

Description

Level

Description

Ia
Ib
IIa

Submental
Submandibular
Superior internal jugular; skull base to
bottom of hyoid, anterior to internal
jugular vein posterior margins
Superior internal jugular; posterior to
internal jugular veins
Middle internal jugular; bottom of hyoid
to bottom of cricoid
Inferior internal jugular; bottom of cricoid
to level of clavicles
Posterior triangle, spinal accessory, and
transverse cervical chains; behind
posterior border of sternocleidomastoid
muscles, superior to clavicles
Visceral compartment; bottom of hyoid to
suprasternal notch, bounded laterally
by the carotid sheaths
Superior mediastinum; inferior to
suprasternal notch

NX
N0
N1

Cannot be assessed
No known nodal metastases
Single, ipsilateral node, %3 cm in greatest
dimension; a midline node is considered
ipsilateral
Single, ipsilateral node O3 cm and %6 cma
Multiple, ipsilateral nodes all %6 cma
Bilateral or contralateral nodes, all %6 cma
O6 cm node

IIb
III
IV
V

VI

VII

Data from Greene FL, Page DL, Fleming ID, et al,

editors. AJCC Cancer Staging Manual. 6th edition.
New York: Springer; 2002.

of investigation, initial studies comparing PETCT with MR, CT, and PET alone have been
promising [53,54].
To evaluate abnormal FDG uptake, normal
uptake patterns must be recognized. FDG is
normally taken up by the brain, liver, myocardium, bowel, and urinary tract. These areas of
uptake do not interfere with oral SCC imaging.
There is also normal uptake in the lymphatic tissue
in the tonsils and base of the tongue (Waldeyers
ring), however [55,56]. A V-shaped pattern of
activity in the oor of mouth usually represents
the sublingual glands. The remaining salivary
glands are variably active. Varying activity levels
between glands is thought to be attributable to
the primarily serous function of the parotid and
submandibular glands, as opposed to the mucin
production by the sublingual glands [5759].
Manifold causes of false-positive ndings during PET exist, and careful attention is necessary to
increase imaging interpretation specicity. Muscle
activity is commonly seen and is usually attributed
to activity just before or during the uptake phase
of the FDG. If the patient does not rest with the
eyes closed but, instead, reads or talks after FDG
is administered, ocular muscles or laryngeal activity is seen. Anxiety may manifest as activity
within paraspinous muscles and may be prevented

N2a
N2b
N2c
N3
a

It is recognized that most nodal masses greater

than 3 cm are not single nodes but are conuent nodes.
Data from Greene FL, Page DL, Fleming ID, et al,
editors. AJCC Cancer Staging Manual. 6th edition.
New York: Springer; 2002.

with administration of low-dose diazepam. Brown

fat uptake is seen occasionally when the patient is
cold during the examination. Especially in the
setting of head and neck cancer, brown fat can be
confounding. Cells that are inamed from infection or recent perturbation from surgery or
radiation demonstrate increased activity [58,60].
Although the activity associated with inammation is not as focal or as strong as that seen in
recurrence, it can be confusing. Asymmetry in activity can be used to dierentiate normal from abnormal tissue. After treatment, however, normal
uptake may appear asymmetric because of surgical and radiation changes. Finally, macrophages
within arterial plaques, particularly in the carotid
arteries, take up FDG.
Poor glucose control or a prior insulin bolus
may alter glucose entry into cells, causing falsenegative results [61]. High blood glucose competes
for FDG and is associated with abnormally high
blood pool activity (circulating blood volume
within the heart and great vessels). Exogenous insulin given to control serum glucose or endogenous
insulin from a nonfasting state causes FDG to be
preferentially directed into the heart and muscle
groups. Recommendations for diabetics include
taking half the normal insulin dose and fasting before the examination. Dose inltration also results
in false-negative ndings but is recognized by
prominent uptake near the intravenous site.
PET scans may also demonstrate false-positive
ndings when the attenuation correction algorithm
miscorrects for dental metalwork [62] and other
metallic surgical hardware. Radiologists avoid
this pitfall by inspecting the noncorrected images
in concert with the corrected images. In addition,

IMAGING OF ORAL CAVITY SCC

459

Fig. 13. Subtle nodal necrosis within a subcentimeter node involved by tumor from a left inferior alveolar ridge primary
SCC tumor. (A) Axial enhanced CT image shows subtle low attenuation within a left submandibular level I
subcentimeter node (arrow). (B) Follow-up study reveals interval enlargement of this involved node.

patient motion may misregister the PET images

onto the CT images, causing false-positive results.
PET-CT fusion imaging avoids this pitfall [60].
Positron emission tomography of oral cancer
When used alone, PET has shown little advantage over CT or MRI in classifying primary
tumors because it lacks the spatial resolution
necessary to detail structural integrity and tumor
invasion [63,64]. Clearly, the role of PET in initial
staging must be in concert with CT or MRI. Studies have clearly indicated the superiority of PET
and PET-CT compared with other imaging modalities in nodal staging [6567]. With PET, small
malignant nodes (which do not meet CT or MRI
pathologic criteria) are identied. The controversy
lies, however, in the clinical signicance of superior accuracy in nodal staging. Some studies
found that despite increased sensitivity and specicity during staging, there was no dierence in
outcome between patients staged by PET and
those staged by other means. In addition, some
studies found PET to be inferior in accuracy
[49]. These studies, however, compared PET alone
without the benet of fused CT scans.
PET may be useful in those patients who could
avoid therapy with low nodal staging. Seventy
percent of patients who are staged at N0 clinically
undergo negative nodal dissection. A negative PET
scan may change this treatment arm, although, at
present, the results are mixed. One study used PET
to dierentiate those who would receive neck
dissection from sentinel node biopsy alone. Kovacs
and colleagues [68] reported saving 12 of 124 neck
sides that would normally have undergone neck

dissection without the benet of PET. Although

there is still some debate regarding this protocol
[6971], this is an intriguing use of PET for staging.
In our experience, fusion PET-CT greatly increases
the diagnostic accuracy of interpretation compared
with PET alone, and the PET examination should
never be reported without a corresponding CT or
MRI examination of the neck.
At present, the most valuable contribution of
PET in head and neck SCC imaging is in
advanced tumors and in surveillance. PET truly
has an advantage over CT and MRI for nding
distant metastases and second primary tumors.
Ten percent of patients with oral cavity and
oropharyngeal SCC have a second primary tumor
at presentation, and 22% develop a second primary tumor within 5 years. Finding these lesions
signicantly alters therapy [72,73]. For this reason, PET should be used routinely to evaluate patients presenting with large primary tumors or
extensive regional disease (Fig. 14).
A possible future benet of PET may concern
the prognostic information that comes from the
relative activity. Although the pitfalls of the SUV
alone have been previously discussed, some studies indicate that the FDG avidity may predict
tumor response to therapy [74,75].
Studies demonstrate the superiority of PET to
anatomic imaging in tumor surveillance. Local
recurrence can be dicult to assess on CT or MRI
because of the distortion of normal tissue planes
by surgery and radiation. PET has superior
sensitivity and specicity for recurrence, with
reported negative predictive values up to 97% to
100%. Positive predictive values in the same
studies range from 55% to 71% [76,77].

460

AULINO

et al

Fig. 14. Benet of whole-body PET to identify distant disease in the patient with a large primary tumor. Axial PET image (A) and corresponding axial noncontrast transmission CT image (B) obtained on a CT-PET scanner show increased
radiotracer uptake within the large anterior right and midline tongue mass and an involved 10-mm right level II node
(arrow). (C) Coronal maximum intensity projection PET image shows additional involved nodes within the right neck
and mediastinum (arrows).

The timing of surveillance PET is still being

debated. Current recommendations are to wait 3
to 4 months after treatment (eg, surgery, radiation) before imaging with PET. This diminishes
false-positive results related to expected inammatory activity in the posttreatment neck. Positive PET scans should be followed with a guided
biopsy. Some investigators have sought to shorten
the time to PET scanning after radiation and
chemotherapy. By scanning early, investigators
hope to change therapy in a tumor that is not
responding. Early scans also may lead to surgical
treatment before radiation brosis develops.
These studies remain preliminary [78,79].
Finally, PET has some use in the diagnosis of an
unknown primary tumor. Two percent to 5% of
patients with head and neck cancer present with
nodal disease and an unknown primary. Hidden
tumors usually reside in the base of the tongue,
tonsils, and nasopharynx and have been treated

with wide radiotherapy in the past. This approach

can be debilitating to underlying structures. PET
has demonstrated good results (20%50% detection) in nding unknown tumors after traditional
evaluations were negative [8082]. This information can be used to avoid wide eld radiation.
PET, and especially PET-CT, is an emerging
force in the diagnosis and treatment of oral cavity
SCC. Although its complete role in oral SCC
imaging is still evolving, PET seems to have great
utility in staging advanced disease for nodal and
distant metastases as well as in monitoring response to therapy and surveillance imaging.
Summary
Familiarity with the patterns of oral cavity
tumor spread is essential to appropriate interpretation of imaging studies. All relevant imaging
studies should be viewed simultaneously, in

IMAGING OF ORAL CAVITY SCC

conjunction with the appropriate clinical history

provided by the clinician surgeon. CT and MRI of
the oral cavity may be complementary to evaluate
SCC, although MRI is generally better suited for
the purpose. Currently, PET-CT may be useful in
the patient with a large primary tumor to identify
distant disease, to evaluate the patient with an
unknown primary tumor, and to identify residual
or recurrent tumor in the posttreatment neck. The
routine use of PET for nodal staging is controversial, although the benet of fusion PET-CT
has not been evaluated. MRI and CT technologies
are rapidly evolving, and limitations of each of
these modalities that have been described in
older published reports should be taken into
consideration.

References
[1] Weber AL, Romo L, Hashmi S. Malignant tumors
of the oral cavity and oropharynx: clinical, pathologic, and radiologic evaluation. Neuroimaging
Clin N Am 2003;13:44364.
[2] Greene FL, Page DL, Fleming ID, et al, editors.
American Joint Committee on Cancer staging manual. 6th edition. New York: Springer; 2002.
[3] Henrot P, Blum A, Toussaint B, et al. Dynamic
maneuvers in local staging of head and neck malignancies with current imaging techniques: principles
and clinical applications. Radiographics 2003;23:
120113.
[4] Lam P, Au-Yeung K, Cheng P, et al. Correlating
MRI and histologic tumor thickness in the assessment of oral tongue cancer. AJR Am J Roentgenol
2004;182:8038.
[5] Mukherji SK, Armao D, Joshi VM. Cervical nodal
metastases in squamous cell carcinoma of the head
and neck: what to expect. Head Neck 2001;23:
9951005.
[6] Yousem DM, Chalian AA. Oral cavity and pharynx.
Radiol Clin North Am 1998;36:96781.
[7] Campbell RSD, Baker E, Chippindale AJ, et al.
MRI T staging of squamous cell carcinoma of the
oral cavity: radiological-pathological correlation.
Clin Radiol 1995;50:53340.
[8] Mukherji SK, Castelijns J, Castillo M. Squamous
cell carcinoma of the oropharynx and oral cavity:
how imaging makes a dierence. Semin Ultrasound
CT MR 1998;19:46375.
[9] Nakayama E, Yoshiura K, Yuasa K, et al. A study
of the association between the prognosis of
carcinoma of the mandibular gingiva and the pattern
of bone destruction on computed tomography. Dentomaxillofac Radiol 2000;29:1639.
[10] Au-Yeung KM, Ahuja AT, Ching ASC, et al. DentaScan in oral imaging. Clin Radiol 2001;56:70013.

461

[11] Brockenbrough JM, Petruzzelli GJ, Lomasney L.

DentaScan as an accurate method of predicting
mandibular invasion in patients with squamous cell
carcinoma of the oral cavity. Arch Otolaryngol
Head Neck Surg 2003;129:1137.
[12] Brown JS, Lewis-Jones H. Evidence for imaging the
mandible in the management of oral squamous cell
carcinoma: a review. Br J Oral Maxillofac Surg
2001;39:4118.
[13] Bolzoni A, Cappiello J, Piazza C, et al. Diagnostic accuracy of magnetic resonance imaging in
the assessment of mandibular involvement in oraloropharyngeal squamous cell carcinoma. Arch
Otolaryngol Head Neck Surg 2004;130:83743.
[14] Weissman JL, Carrau RL. Pued-cheek CT improves evaluation of the oral cavity. AJNR Am J
Neuroradiol 2001;22:7414.
[15] Kimura Y, Sumi M, Sumi T, et al. Deep extension
from carcinoma arising from the gingiva: CT and
MR imaging features. AJNR Am J Neuroradiol
2002;23:46872.
[16] Sumi M, Izumi M, Yonetsu K, et al. Sublingual
gland: MR features of normal and diseased states.
AJR Am J Roentgenol 1999;172:71722.
[17] Castelijns JA, van den Brekel MWM. Magnetic
resonance imaging evaluation of extracranial head
and neck tumors. Magn Reson Q 1993;9:11328.
[18] Crecco M, Vidiri A, Angelone ML, et al. Retromolar
trigone tumors: evaluation by magnetic resonance
imaging and correlation with pathological data.
Eur J Radiol 1999;32:1828.
[19] Lane AP, Buckmire RA, Mukherji SK, et al. Use of
computed tomography in the assessment of mandibular invasion in carcinoma of the retromolar trigone.
Otolaryngol Head Neck Surg 2000;122:6737.
[20] Araki K, Ariji E, Shimizu M, et al. Computed
tomography of carcinoma of the upper gingiva
and hard palate: correlation with the surgical and
histopathological ndings. Dentomaxillofac Radiol
1997;26:17782.
[21] Som PM, Shapiro MD, Biller HF, et al. Sinonasal
tumors and inammatory tissues: dierentiation
with MR imaging. Radiology 1988;167:8039.
[22] Lanzieri CF, Shah M, Krauss D, et al. Use of gadolinium-enhanced MR imaging for dierentiating
mucoceles from neoplasms in the paranasal sinuses.
Radiology 1991;178:4258.
[23] Candela FC, Kothari K, Shah JP. Patterns of
cervical node metastases from squamous carcinoma
of the oropharynx and hypopharynx. Head Neck
1990;12:197203.
[24] Shah JP. Cervical lymph node metastasesddiagnostic, therapeutic, and prognostic implications. Oncology (Huntingt) 1990;4:619.
[25] Ali S, Tiwari RM, Snow GB. False-positive and
false-negative neck nodes. Head Neck Surg 1985;8:
7882.
[26] Shingaki S, Suzuki I, Nakajima T, et al. Computed
tomographic evaluation of lymph node metastases

462

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

AULINO

in head and neck carcinomas. J Craniomaxillofac

Surg 1995;23:2337.
Merritt RM, Williams MF, James TH, et al. Detection of cervical metastasis: a meta-analysis comparing computed tomography with physical exam.
Arch Otolaryngol Head Neck Surg 1997;123:
14952.
Atula TS, Varpula MJ, Kurki TJI, et al. Assessment
of cervical lymph node status in head and neck cancer patients: palpation, computed tomography and
low eld magnetic resonance imaging compared
with ultrasound-guided ne-needle aspiration cytology. Eur J Radiol 1997;25:15261.
Som PM, Curtin HD, Mancuso AA. Imaging-based
nodal classication for evaluation of neck metastatic
adenopathy. AJR Am J Roentgenol 2000;174:
83744.
Malard O, Toquet C, Jegoux F, et al. Computed tomography in TN stage evaluation of oral cavity and
oropharyngeal cancers. Clin Imaging 2004;28:3607.
Yousem DM, Som PM, Hackney DB, et al. Central
nodal necrosis and extracapsular neoplastic spread
in cervical lymph nodes: MR imaging versus CT.
Radiology 1992;182:7539.
Curtin HD, Ishwaran H, Mancuso AA, et al. Comparison of CT and MR imaging in staging of neck
metastases. Radiology 1998;207:12330.
King AD, Tse GMK, Ahuja AT, et al. Necrosis in
metastatic neck nodes: diagnostic accuracy of CT,
MR imaging, and US. Radiology 2004;230:7206.
Gor DM, Langer JE, Loevner LA. Imaging of cervical lymph nodes in head and neck cancer: the basics.
Radiol Clin North Am 2006;44:10110.
King AD, Tse GMK, Yuen EHY, et al. Comparison
of CT and MR imaging for the detection of extranodal neoplastic spread in metastatic neck nodes. Eur J
Radiol 2004;52:26470.
Castelijns JA, van den Brekel MWM. Imaging of
lymphadenopathy in the neck. Eur Radiol 2002;12:
72738.
van den Brekel MWM, Castelijns JA, Snow GB. The
size of lymph nodes in the neck on sonograms as a radiologic criterion for metastasis: how reliable is it?
AJNR Am J Neuroradiol 1998;19:695700.
van Den Brekel MW, Castelijns JA, Stel HV, et al.
Modern imaging techniques and ultrasound-guided
aspiration cytology for the assessment of neck
node metastases: a prospective comparative study.
Eur Arch Otorhinolaryngol 1993;250:117.
Hodder SC, Evans RM, Patton DW, et al. Ultrasound and ne needle aspiration cytology in the staging of neck lymph nodes in oral squamous cell
carcinoma. Br J Oral Maxillofac Surg 2000;38:
4306.
Leslie A, Fyfe E, Guest P, et al. Staging of squamous
cell carcinoma of the oral cavity and oropharynx:
a comparison of MRI and CT in T- and N-staging.
J Comput Assist Tomogr 1999;23:439.

et al
[41] Chikamatsu K, Kamada H, Ninomiya H, et al. A
preliminary study on sentinel lymph node biopsy:
feasibility and predictive ability in oral cavity cancer.
Ann Nucl Med 2004;18:25762.
[42] Hamberg LM, Hunter GJ, Alpert NM, et al. The
dose uptake ratio as an index of glucose metabolism:
useful parameter or oversimplication? J Nucl Med
1994;35:130812.
[43] Lindholm P, Minn H, Leskinen-Kallio S, et al. Inuence of the blood glucose concentration on FDG uptake in cancerda PET study. J Nucl Med 1993;34:
16.
[44] Kim CK, Gupta NC, Chandramouli B, et al. Standardized uptake values of FDG: body surface area
correction is preferable to body weight correction.
J Nucl Med 1994;35:1647.
[45] Kim EE, Lee M, Inoue T, et al. Clinical PET, principles and applications. New York: Springer-Verlag;
2004.
[46] Beyer T, Townsend DW, Brun T, et al. A combined
PET/CT scanner for clinical oncology. J Nucl Med
2000;41:136979.
[47] Goerres GW, von Schulthess GK, Steinert HC. Why
most PET of lung and head-and-neck cancer will be
PET/CT. J Nucl Med 2004;45(Suppl 1):66S71S.
[48] Schoder H, Yeung HWD, Gonen M, et al. Head and
neck cancer: clinical usefulness and accuracy of
PET/CT image fusion. Radiology 2004;31:6572.
[49] Dammann F, Horger M, Mueller-Berg M, et al. Rational diagnosis of squamous cell carcinoma of the
head and neck region: comparative evaluation of
CT, MRI, and 18FDG PET. AJR Am J Roentgenol
2005;184:132631.
[50] Ng SH, Yen TC, Liao CT, et al. 18F-FDG PET and
CT/MRI in oral cavity squamous cell carcinoma:
a prospective study of 124 patients with histologic
correlation. J Nucl Med 2005;46:113643.
[51] Yen TC, Chang JT, Ng SH, et al. Staging of untreated squamous cell carcinoma of buccal mucosa
with 18F-FDG PET: comparison with head and
neck CT/MRI and histopathology. J Nucl Med
2005;46:77581.
[52] Goerres GW, Schmid DT, Schuknecht B, et al. Bone
invasion in patients with oral cavity cancer: comparison of conventional CT with PET/CT and SPECT/
CT. Radiology 2005;237:2817.
[53] Syed R, Bomanji JB, Nagabhushan N, et al. Impact
of combined 18F-FDG PET/CT in head and neck
tumours. Br J Cancer 2005;92:104650.
[54] Branstetter BF IV, Blodgett TM, Zimmer LA, et al.
Head and neck malignancy: is PET/CT more accurate than PET or CT alone? Radiology 2005;235:
5806.
[55] Goerres GW, von Schulthess GK, Hany TF. Positron emission tomography and PET/ CT of the
head and neck: FDG uptake in normal anatomy,
in benign lesions, and in changes resulting from
treatment. AJR Am J Roentgenol 2002;19:133743.

IMAGING OF ORAL CAVITY SCC

[56] Nakamoto Y, Tatsumi M, Hammoud D, et al. Normal FDG distribution patterns in the head and neck:
PET/CT evaluation. Radiology 2005;234:87985.
[57] Graham MM, Menda Y. Positron emission tomography/computed tomography imaging of head and
neck tumors: an atlas. Semin Nucl Med 2005;35:
22052.
[58] Kapoor V, Fukui M, McCook BM. Role of 18FFDG PET/CT in the treatment of head and neck
cancers: principles, technique, normal distribution,
and initial staging. AJR Am J Roentgenol 2005;
184:57987.
[59] Lowe V, Stack B, Bogsrud T. PET and PET/CT
imaging in head and neck cancer. In: Valke PE,
Delbeke D, Bailey DL, et al, editors. Positron
emission tomography: clinical practice. London:
Springer-Verlag, in press.
[60] Kapoor V, Fukui M, McCook BM. Role of 18FFDG PET/CT in the treatment of head and neck
cancers: posttherapy evaluation and pitfalls. AJR
Am J Roentgenol 2005;184:58997.
[61] Goerres GW, Schmid DT, Eyrich GK. Do hardware
artifacts inuence the performance of head and neck
PET scans in patients with oral cavity squamous cell
cancer? Dentomaxillofac Radiol 2003;32:36571.
[62] Torizuka T, Zasadny KR, Wahl RL. Diabetes decreases FDG accumulation in primary lung cancer.
Clin Positron Imaging 1999;2:2817.
[63] Laubenbacher C, Saumweber D, Wagner-Manslau
C, et al. Comparison of uorine-18-uorodeoxyglucose PET, MRI and endoscopy for staging head and
neck squamous-cell carcinomas. J Nucl Med 1995;
36:174757.
[64] Wong WL, Chevretton EB, McGurk M, et al. A prospective study of PET-FDG imaging for the assessment of head and neck squamous cell carcinoma.
Clin Otolaryngol Allied Sci 1997;22:20914.
[65] Braams JW, Pruim J, Freling NJ, et al. Detection of
lymph node metastases of squamous cell cancer of
the head and neck with FDG-PET. MRI. J Nucl
Med 1995;36:2116.
[66] Benchaou M, Lehmann W, Slosman DO, et al. The
role of FDG-PET in the pre-operative assessment of
N staging in head and neck cancer. Acta Otolaryngol
1996;116:3325.
[67] Adams S, Baum RP, Stuckensen T, et al. Prospective
comparison of 18F-FDG PET with conventional
imaging modalities (CT, MRI, US) in lymph node
staging of head and neck cancer. Eur J Nucl Med
1998;25:125560.
[68] Kovacs AF, Dobert N, Gaa J, et al. Positron emission tomography in combination with sentinel
node biopsy reduces the rate of elective neck dissections in the treatment of oral and oropharyngeal
cancer. J Clin Oncol 2004;22:397380.
[69] Stoeckli SJ, Steinert H, Pfaltz M, et al. Is there a role
for positron emission tomography with 18F-uorodeoxyglucose in the initial staging of nodal negative

[70]
[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78]

[79]

[80]

[81]

[82]

463

oral and oropharyngeal squamous cell carcinoma.

Head Neck 2002;24:3459.
Pohar S. Positron emission tomography paradigm
fuzzier than reported. J Clin Oncol 2004;23:48034.
Smith BD, Lally BE, Haty BG, et al. Do PET and
SNB reduce the rate of elective neck dissection? A
hypothesis still in need of validation. J Clin Oncol
2005;23:28745 [author reply: 28756].
Teknos TN, Rosenthal EL, Lee D, et al. Positron
emission tomography in the evaluation of stage III
and IV head and neck cancer. Head Neck 2001;23:
105660.
Schwartz DL, Rajendran J, Yueh B, et al. Staging of
head and neck squamous cell cancer with extendedeld FDG-PET. Arch Otolaryngol Head Neck Surg
2003;129:11738.
Kitagawa Y, Sano K, Nishizawa S, et al. FDG-PET
for prediction of tumour aggressiveness and response to intra-arterial chemotherapy and radiotherapy in head and neck cancer. Eur J Nucl Med
Mol Imaging 2003;30:6371.
Allal AS, Slosman DO, Kebdani T, et al. Prediction
of outcome in head-and-neck cancers using the standard uptake value of 2-[18F]uoro-2-deoxy-Dglucose. Int J Radiat Oncol Biol Phys 2004;59:
1295300.
Yao M, Graham MM, Smith RB, et al. Value of
FDG PET in assessment of treatment response and
surveillance in head-and-neck cancer patients after
intensity modulated radiation treatment: a preliminary report. Int J Radiat Oncol Phys 2004;60:
14108.
Porceddu SV, Jarmolowski E, Hicks RJ, et al. Utility
of positron emission tomography for the detection
of disease in residual neck nodes after (chemo)radiotherapy in head and neck cancer. Head Neck 2005;
27:17581.
Goerres GW, Schmid DT, Bandhauer F, et al. Positron emission tomography in the early follow-up
of advanced head and neck cancer. Arch Otolaryngol
Head Neck Surg 2004;130:1059.
Nam SY, Lee SW, Im KC, et al. Early evaluation of
the response to radiotherapy of patients with
squamous cell carcinoma of the head and neck using
18FDG-PET. Oral Oncol 2005;41:3905.
Bocca E, Calearo C, de Vincentiis I, et al. Occult metastases in cancer of the larynx and their relationship
to clinical and histological aspects of the primary tumor: a four-year multicentric research. Laryngoscope 1984;94:108690.
Braams JW, Pruim J, Kole AC, et al. Detection of
unknown primary head and neck tumors by positron
emission tomography. Int J Oral Maxillofac Surg
1997;26:1125.
Assar OS, Fischbein NJ, Caputo GR, et al. Metastatic head and neck cancer: role and usefulness of
FDG PET in locating occult primary tumors.
Radiology 1999;210:17781.

Oral Maxillofacial Surg Clin N Am 18 (2006) 465482

Advances in the Detection and Diagnosis of Oral

Precancerous and Cancerous Lesions
John R. Kalmar, DMD, PhD
Section of Oral and Maxillofacial Surgery, Pathology, and Anesthesiology, The Ohio State University College
of Dentistry, 305 West 12th Avenue, Columbus, OH 43210, USA

In the United States, an estimated 29,370 new

cases of oral and pharyngeal cancer were diagnosed in 2005, with more than 7320 tumor-related
deaths [1]. Oral cancer represents roughly 3% of
total cancer cases in the United States and is the
ninth most common form of malignancy among
American men. Although the concept of early diagnosis leads to improved prognosis applies to
oral cancer, most patients present with regional
or distant (stage III or IV) disease, which is a problem especially notable among African Americans.
The tendency for delayed or late diagnosis is
reected in an overall 5-year survival rate of approximately 59% for data pooled from 1995
through 2001. Although this gure represents
a signicant improvement for oral cancer survival
for the rst time in decades (up from 54% in
19741976), survival within the African-American
population has remained comparatively lower
(36% in 19741976, 40% in 19952001) [2]. Increased mortality from oral cancer is especially
marked in African-American men, whose 5-year
survival rate (34%) is substantially lower than
that of their female counterparts (52%).
The most common form of oral cancer is primary mucosal squamous cell carcinoma (O90%
of cases), although malignancies of salivary gland
origin, sarcomas, lymphomas, melanoma, and
metastatic disease also contribute to the total
cancer burden. Because squamous cell carcinoma
and its variants represent most oral cancer cases,
this article focuses on the diagnosis and detection
of this condition and its precursors. The ability to
diagnosis precursor (precancerous) lesions is

E-mail address: kalmar.7@osu.edu

critical to the battle against oral cancer. With

early detection, diagnosis, and treatment, noninvasive intraepithelial lesions (grades of epithelial
dysplasia or carcinoma in situ [CIS]) can be
conservatively managed with minimal surgical
morbidity and 100% survival. In addition, advances in molecular diagnosis suggest that genetic
or protein markers of precancerous change are
likely detectable before clinically apparent mucosal lesions can be identied. If the promise of such
prediagnosis can be realized, early detection of
patients at increased risk for initial or recurrent
disease would be possible and would hopefully
lead to reduced patient morbidity and mortality.
Clinical features of oral precancerous
and cancerous lesions
The signs and symptoms of precancerous
lesions and even some early squamous cancers
are often so subtle that they probably go unnoticed or ignored by patients and practitioners
alike. Distinguishing lesional tissue from the surrounding mucosa, especially in the presence of
complicating factors, such as local trauma or
superimposed infection, can be dicult for even
well-trained health care professionals. Together
with estimates that only approximately half of the
US adult population sees a dentist even once
a year, it should probably not be surprising that
most patients with oral cancer (60%) are diagnosed with stage III or IV disease.
Given that notable symptoms are typically
a late-stage feature of oral cancer, early detection
and diagnosis of oral precancerous and cancerous
lesions clearly depend on patient participation in
periodic (annual) oral examinations and the

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.013

oralmaxsurgery.theclinics.com

466

KALMAR

sensitivity and specicity of the oral examiner or

examination procedure. Detecting the mucosal
alterations that often precede the development
of squamous cell carcinoma requires a sound
knowledge of oral anatomy and anatomic variations as well as a thorough understanding of local
and systemic factors or conditions that can mimic
or obfuscate underlying precancerous change.
Leukoplakia
The term leukoplakia is dened as a white plaque or patch of oral mucosa that cannot be
rubbed o or cannot be diagnosed as any other
condition clinically, or subsequently, by microscopic evaluation. Leukoplakia is not a diagnosis;
it is a descriptive term that encompasses a surprising variety of localized whitish areas of mucosal
change that cannot be readily explained at the
clinical level. The term has no diagnostic and,
thus, no prognostic value. Written or imagingbased documentation of clinical features, including site, size, border, surface character, and
presence of ulceration is a medicolegally sound practice that should always be performed as a standard
part of patient examination. Inspection of the lesion border is of particular importance, because
a well-dened sharply demarcated margin is suggestive of clonal (preneoplastic or neoplastic)
growth (Fig. 1). Depending on the precise clinical
setting, dierential considerations, such as traumatic, reactive, or infectious conditions, can usually be addressed through local conservative
measures and follow-up re-evaluation. Any leukoplakia that persists or progresses after 10 to 14
days despite appropriate conservative treatment

Fig. 1. Large leukokeratotic plaque (leukoplakia) of the

left posterior and inferior buccal mucosa with a sharply
dened border and surface irregularity, including plexiform ssuring. (Courtesy of C.M. Allen, DDS, MS, Columbus, OH.)

should be considered a potentially premalignant

condition.
Leukoplakia is most commonly seen in older
adult men, and more than 80% of patients have
a history of smoking [3,4]. Although the buccal
mucosa and gingiva are the most frequently affected sites (see Fig. 1), lesions that occur on the
ventral tongue, oor of the mouth, and tonsillar
pillars are more likely to demonstrate histologic
evidence of dysplasia or carcinoma. These latter
areas have been recognized for years as some of
the oral anatomic regions at greatest risk for the
development of squamous cell carcinoma (Fig. 2)
[5]. For this reason, persistent leukoplakia in these
areas should be considered as suspicious for carcinoma. Scalpel biopsy is warranted for any suspicious lesion and should be scheduled or performed
as soon as conveniently possible. Use of diagnostic adjuncts, such as toluidine blue staining, may
be helpful in guiding the biopsy procedure; however, heavily keratinized lesions are often negative
with this vital stain. Cytologic methods, including
brush cytology, are not advised for clinically suspicious lesions, because these tests can delay scalpel biopsy, denitive diagnosis, and appropriate
therapy. Therefore, brush cytology would not be
indicated for any persistent leukoplakia in the highrisk zone for oral cancer (see Fig. 2B). Even with
small lesions (Fig. 3), excisional biopsy in these
areas would be preferable for two reasons. First,
complete removal of lesional tissue is more easily
accomplished by scalpel biopsy and typically restores a normal background appearance to the
mucosa at this site. Against this mucosal equivalent of a clean slate, the clinicians ability to detect
signs of local recurrence is improved. Second,
scalpel biopsy leads more directly to a nal tissue
diagnosis, decreasing the interval to appropriate
treatment if needed. Finally, regardless of location, any leukoplakia that exhibits intralesional
areas of reddish or erythematous change (eg,
speckled leukoplakia, erythroleukoplakia) should
also be viewed as a high-risk presentation (see
the section on erythroplakia) that demands scalpel
biopsy. With speckled leukoplakias, toluidine blue
has been shown to stain the less heavily keratinized (reddish) areas suspicious for dysplasia or
carcinoma and may be helpful in directing the biopsy procedure.
Proliferative verrucous leukoplakia (PVL) is
a more aggressive and often multifocal form of
leukoplakia that frequently occurs in the absence
of a signicant smoking history [6,7]. Although it
can aect any area of the oral cavity, the buccal

DETECTION AND DIAGNOSIS OF ORAL LESIONS

467

Fig. 2. (A) Site of origin of 209 consecutive cases of mouth cancer from the Memorial Hospital Head and Neck service
between 1962 and 1965. (B) Cancer-prone crescent from which 75% of cancerous lesions originate. (From Moore C, Catlin D. Anatomic origins and locations of oral cancer. Am J Surg 1967;114(4):511; with permission.)

mucosa is a favored site among female patients,

whereas the tongue is involved most often in
male patients. In addition, female patients tend
to be older (mean age of 6570 years) than male
patients (mean age of 49 years) at the time of diagnosis. Progression of PVL lesions to involve signicant portions of the oral mucosa is often seen
despite surgical treatment, and relatively rapid
transformation to squamous cell carcinoma is
a recognized complication.
Erythroplakia
As with its whitish counterpart, the term erythroplakia is used to describe a red macule or plaque
that cannot be rubbed o or diagnosed clinically
as any other condition. Although not a diagnosis,
this presentation should always arouse clinical

concern, because nearly 100% of true erythroplakias have been found on biopsy to represent severe dysplasia, CIS, or squamous cell carcinoma
[8,9]. Not surprisingly, most erythroplakias arise
in oral sites at the highest risk for squamous cell
carcinoma: the oor of the mouth, ventrolateral
surfaces of the tongue, tonsillar pillars, and soft
palate (see Fig. 2B). Admixed areas of keratinization (speckled erythroplakia) may be seen.
Depending on the precise clinical presentation,
immediate scalpel biopsy of erythroplakia may
be warranted even without conservative treatment
or follow-up evaluation. Toluidine blue staining
may be useful in biopsy site selection for cases
of erythroplakia. As previously noted, the high
index of suspicion for signicant dysplasia or carcinoma in cases of erythroplakia would be a contraindication for cytologic methods.
Squamous cell carcinoma

Fig. 3. Close-up view of small (0.8 cm
0.3 cm), welldemarcated, asymptomatic leukoplakia of the right ventral tongue.

Most cases of oral squamous cell carcinoma

present initially with clinical features of leukoplakia, erythroplakia, or both. Although any site can
be aected, anatomic areas of increased risk for
this disease have been recognized for years. In
1967, Moore and Catlin [5] presented scattergrams of oral cancer cases that provided a visual
depiction of their distribution (see Fig. 2A). These
plots were used to outline a cancer-prone crescent (see Fig. 2B), where more than 75% of the
cancer cases were found, despite the fact that
this region represented only 20% of the entire
oral mucosa. Subsequently, the area of elevated

468

KALMAR

cancer risk has been extended by other authors to

include the tonsillar pillar and soft palate complex
[10]. As mentioned previously, the nding of any
persistent mucosal alteration in this cancer risk
zone should raise the clinicians index of suspicion and serve as a trigger for surgical biopsy.
The risk for oral cancer increases with age, and
most patients are diagnosed after the age of
40 years. Men are more commonly aected than
women, and, as mentioned previously, the risk is
particularly high for African-American men. The
major risk factor for oral squamous cell carcinoma is cigarette smoking, and roughly 80% of
aected patients have a positive smoking history
[3]. Alcohol consumption has a less well-dened
association and may serve more as a cofactor, together with smoking. Smokeless forms of tobacco
have also been considered as risk factors for oral
cancer. Recent evidence, however, suggests that
this historical view may need to be revised as several epidemiologic studies published during the
past 10 years have failed to detect a signicant association between the use of smokeless tobacco
and the development of oral squamous cell carcinoma [1120]. The only form of oral cancer not
directly associated with smoking is cancer of the
lip. This is related to sun exposure, and roughly
90% of such cases arise on the lower lip vermilion.
It is also well recognized that patients can develop
squamous cell carcinoma in the absence of any
known risk factors. In patients less than 40 years
of age, the most common site for this to occur is
the ventrolateral aspect of the tongue. In older female patients, the gingiva is frequently aected.
Spread of oral squamous cell carcinoma is
usually by local extension into and destruction of
underlying tissues, including alveolar bone. Metastatic spread is commonly through the lymphatics to involve the ipsilateral cervical or
submandibular lymph nodes.

Cytology
Oral exfoliative cytology has been an adjunct
to oral diagnosis for many years; however, until
recently, it has been primarily used to provide
rapid and inexpensive identication of supercial
infectious agents, such as fungi (using periodic
acidSchi or KOH staining), or viruses (using
Papanicolaou staining to permit visualization of
the viral cytopathic eect in infected epithelial
cells), such as herpes simplex virus (HSV; human
herpesvirus [HHV]-1,2) and varicella zoster virus
(VZV; HHV-3).
Use of oral cytology to test potentially precancerous epithelial lesions lost popularity for
several decades after studies from the late 1960s
through early 1970s had false-negative rates as
high as 31% [2123]. Given the signicant margin
of error, most practitioners abandoned this technique in the mid-1970s in favor of surgical biopsy
analysis for potentially precancerous or cancerous
lesions.
Brush cytology (brush biopsy)
Brush cytology (brush biopsy; OralCDx; CDx
Laboratories, Suern, New York) was introduced
in 1999 as an alternative to conventional exfoliative cytology for investigating persistent oral
epithelial lesions not considered suspicious for
carcinoma [24]. Using materials provided in
a commercially available kit (Fig. 4), the technique diers from conventional exfoliative cytology in two signicant ways. First, a small
circular brush instrument is provided for use in
a rotary fashion to collect a transepithelial specimen. The brush is continually rotated against
lesional tissue until pinpoint bleeding is detected
clinically, indicating penetration of the basement

Diagnostic adjuncts
A variety of aids or adjuncts to the diagnosis of
oral precancerous and cancerous lesions have
been developed over the years, several within the
past decade. Although primarily developed for
use by the general dental practitioner, data have
been published to suggest possible utility in the
hands of specialists as well. As with any test,
proper case selection and correct performance of
the test itself are critical to the sensitivity and
specicity of its result.

Fig. 4. Fixative and brush instrument of the OralCDx

brush biopsy system.

DETECTION AND DIAGNOSIS OF ORAL LESIONS

membrane and ensuring the likelihood of a fullthickness (transepithelial) sample. The instrument
is then unloaded by rotating the brush against
a glass slide to deposit and disperse the disaggregated epithelial cells. The sample is xed with a solution provided by the company (see Fig. 4) and
returned for interpretation. Automated computer-assisted specimen analysis initially determines specimen adequacy, and then identies
and stores cytologic abnormalities found within
the specimen. These abnormal ndings are subsequently reviewed by a pathologist trained in oral
cytology, who provides a test result.
Results of brush cytology specimens are classied into one of four categories:
1. Inadequate:
incomplete
transepithelial
specimen
2. Negative: no epithelial abnormality
3. Atypical: abnormal epithelial changes of uncertain diagnostic signicance
4. Positive: denitive cellular evidence of epithelial dysplasia or carcinoma
For atypical or positive results, the company
recommends that patients receive follow-up scalpel biopsy. This recommendation reects the fact
that the brush result is limited to reporting
evidence of cellular abnormalities or atypia; it
does not provide a nal diagnosis. In the case of
a negative result, clinical follow-up of persistent
oral lesions is recommended.
Several studies have shown encouraging data
with oral brush cytology for evaluation of oral
precancerous and cancerous lesions. Sciubba [24]
reported 100% sensitivity with 100% specicity
for positive results and 92.9% specicity for atypical results in 945 patients. Unfortunately, biopsy
conrmation of the brush result was not obtained
for all atypical or negative cases, and the lack of
such information has raised concerns that falsenegative or false-positive results may have been
left undetected [25]. In another study of 298 patients, the positive predictive value of an abnormal brush cytology nding resulting in a scalpel
biopsy report of dysplasia or carcinoma was
38.3% [26]. A comparative study of brush cytology and scalpel biopsy in 80 patients reported
the brush technique to have 92% sensitivity and
94% specicity for both positive and atypical results in detecting dysplasia and oral cancer [27].
For positive results alone, sensitivity was 62%
and specicity was 97%. A positive likelihood
ratio [sensitivity/(1  specicity)] of 16.2 was
also recorded for the brush technique, meaning

469

that a positive or atypical result was 16.2 times

more likely in a mucosal lesion with dysplasia or
carcinoma than in a lesion without precancerous
or cancerous change.
In contrast, results from a study of 112 patients
reported a sensitivity of 71%, specicity of 32%,
and positive predictive value of 44.1% with the
oral brush system [28]. The authors were
concerned that 6 of 15 lesions with a negative
OralCDx result were found to have dysplasia or
carcinoma on subsequent scalpel biopsy. Such
a nding validates previous concerns with earlier
studies for failing to provide follow-up scalpel biopsy ndings on all cases, possibly resulting in an
overestimation of sensitivity and specicity with
the brush technique [25]. Finally, in a series of
four cases of oral squamous cell carcinoma, the
diagnosis of carcinoma was determined by scalpel
biopsy despite negative brush biopsy results [29].
The time delay from the initial brush sampling
to nal diagnosis varied from 5 to 292 days (average 117 days).
The brush system is easy to use, although its
cost is not negligible. In addition to its application
for innocuous-appearing but persistent mucosal
lesions, it could be a useful alternative for assessing lesions in patients who refuse a scalpel biopsy.
Brush cytology, especially in combination with
vital staining, may also be useful for sampling
multiple areas of a large lesion, cases of PVL, or in
the follow-up of patients previously treated for
dysplasia or squamous cell carcinoma.
Tissue uorescence
Recently, a technique previously used as an
adjunct to the examination of cervical mucosa
(speculoscopy) has been adapted and approved
for use in the oral cavity. Several dierent
commercial products designed for this technique
have been marketed, including: ViziLite (Zila,
Phoenix, Arizona; now available as ViziLite Plus
or ViziLite with TBlue marking system), MicroLux DL (AdDent Inc., Danbury, CT), and VELscope (LED Dental Inc., Vancouver, British
Columbia, Canada) (visually enhanced lesion
scope). With the ViziLite system (Fig. 5) and the
MicroLux DL, the oral mucosa is rst rinsed
with mild acetic acid and then illuminated by an
activated chemiluminescent (ViziLite) or batteryoperated portable light source (MicroLux DL)
with output in the blue-white spectrum. The acetic
acid wash helps to remove surface debris and reportedly causes the epithelial cells to dehydrate

470

KALMAR

Fig. 5. ViziLite system components, including a disposable light source, acetic acid solution, and light holder.
ViziLite Plus (ViziLite with TBlue marking system)
kits also provide a toluidine blue (tolonium chloride)
solution.

slightly, increasing the relative prominence of

their nuclei. Under blue-white illumination, normal epithelium appears lightly bluish in color,
whereas abnormal epithelium appears distinctly
white. ViziLite Plus consists of the same device
packaged together with a tolonium chloride solution (see section on toluidine blue). The tolonium
chloride is intended for use as a marking dye to
help highlight lesions identied with the light
source. VELscope (Fig. 6) is an alternating current
(AC)powered, portable, reusable light source
that provides a blue emission spectrum unique
from the ViziLite or MicroLux DL system. With
this device, areas of reduced autouorescence are
considered suspicious for abnormality or a positive
nding.
In a survey study of 150 patients, the ViziLite
system was visually shown to amplify areas of the
mucosa where hyperkeratinization or chronic inammation was identied [30]. Conditions like

Fig. 6. VELscope light source unit with viewing handpiece and ber optic light guide.

leukoedema, nonspecic ulcer, and broma were

shown to be chemiluminescent-positive, together
with two leukoplakias that were subsequently
characterized as atypical by brush cytology or as
hyperkeratosis and epithelial atypia by scalpel
biopsy. No attempt was made by the authors to
assess the sensitivity or specicity of the system.
In a study of 40 Malaysian subjects, the sensitivity
of the ViziLite test with follow-up scalpel biopsy
was reportedly 100%, with a specicity of 14%
[31]. The authors raised several concerns about
the technique, including its cost and a high falsepositive rate (19%). Finally, a published abstract
has reported that the ViziLite test result was positive in 78% of all clinically suspicious lesions,
including 66% of suspicious leukoplakias (61
of 92 cases) and 60% of erythroleukoplakias (6
of 10 cases) but only 25% of clinically suspicious
erythroplakias (5 of 20 cases) [32]. In addition,
19% (12 of 58 cases) of the keratoses judged to
be clinically innocuous were positive; however,
additional histologic or diagnostic information
was not provided.
Recently, investigators using an electrically
powered uorescent light source similar to the
VELscope unit presented results from a pilot
study involving 44 patients with a history of
biopsy-conrmed dysplasia or squamous cell carcinoma [33]. The patients rst received routine
oral examinations under white light, followed by
re-examination in a darkened room using the uorescent unit. Compared with the uniform autouorescence of normal mucosa, areas of reduced
uorescence (as compared with adjacent mucosa
and mucosa from the contralateral anatomic
site) were considered positive or suspicious.
Next, the uorescent results were correlated with
microscopic features in 50 oral biopsies from the
patient cohort. Of 7 biopsies from sites with normal autouorescence, 6 exhibited normal surface
epithelium, although 1 was diagnosed as severe
dysplasia or CIS. Of the remaining 43 specimens
obtained from sites with reduced autouorescence, 10 showed severe dysplasia or CIS and
33 were diagnosed as squamous cell carcinoma.
These data corresponded to a reported sensitivity
of 98% and a specicity of 100%. The authors
noted that the decision to perform a biopsy was
not based on tissue autouorescence but on standard clinical features (patient history, clinical appearance, and toluidine blue staining results).
Unfortunately, the authors failed to correlate
these features with tissue uorescence, making it
impossible to assess the added diagnostic value

DETECTION AND DIAGNOSIS OF ORAL LESIONS

of the uorescent examination. A published abstract from the same group reported that a signicantly higher proportion of oral premalignant
lesions (n 69) with reduced uorescence were
dysplastic (n 42 [81%]) compared with lesions
with normal uorescence (n 17 [41%]) [34]. In
another abstract, 8 patients undergoing surgery
for recently diagnosed T0 to T2 oral cancer were
studied. In each case, the clinical lesions, areas
of reduced tissue uorescence (uorescent-positive), and surgical margins were delineated, and
punch biopsies (n 18) were obtained from uorescent-positive areas that extended beyond the
margin of visibly abnormal tissue. Of these biopsies, 6 were diagnosed as carcinoma (33%), 4 as
severe dysplasia (22%), 4 as mild to moderate dysplasia (22%), and 4 as hyperplasia or normal
(22%). These results suggest that uorescent examination may permit detection of precancerous
lesions even when the oral mucosa appears clinically normal [35].
The ViziLite Plus test is simple to use; however, its cost is not negligible, and the light stick
can only be activated once. Although the MicroLux DL provides a multiple-use light source, there
is currently little evidence to suggest that either
system improves detection of oral precancerous or
cancerous lesions beyond visual inspection alone.
The VELscope unit is a portable, multiuse,
uorescent device that is also simple to operate,
but the unit is expensive and its durability has not
been proven. Additional prospective studies are
needed to evaluate the potential diagnostic benet
of tissue uorescence for oral cavity examination.
Toluidine blue (tolonium chloride)
In 1964, Niebel and Chomet [36] rst reported
on the use of toluidine blue as a vital tissue stain
to aid in the early detection of oral precancerous
and malignant lesions. Also known by its chemical name of tolonium chloride, toluidine blue is
a basic metachromatic stain that binds to DNA.
Although not cancer specic, it has been reported
to stain mitochondrial DNA, altered DNA in premalignant and malignant epithelial lesions, and
cells with relatively increased amounts of DNA
[37]. From 1964 to 1992, a number of studies
showed toluidine blue to exhibit sensitivity that
ranged from 86% to 100%, with a specicity
ranging from 63% to 100%. A meta-analysis published in 1989 reported toluidine blue sensitivity as
ranging from 93.5% to 97.8%, with a specicity
ranging from 73.3% to 92.9% [38].

471

In 1996, Warnakulasuriya and Johnson [39] reported that all oral cancers (18 of 18 cases) tested
were toluidine blue-positive; however, lower sensitivity (79.5%) and specicity (62%) were found
with precancerous lesions, and a false-negative
rate of 20.5% was observed. Problems with toluidine blue sensitivity, specicity, or both were
noted in other studies of precancerous lesions in
the middle to late 1990s and early 2000s [40,41].
In addition, false-positive rates as high as 35%
were reported [41]. Variable study results over several decades probably explain why toluidine blue
currently lacks widespread acceptance among generalists or specialists.
A series of recent reports may revive professional interest in this technique, however.
Toluidine blue positivity was higher in oral premalignant lesions that showed loss of heterozygosity (LOH) at chromosome regions associated
with the development of squamous cell carcinoma
(3p, P .13; 17p, P .049) and was more likely
seen with lesions that showed LOH in greater than
two regions [42]. Importantly, the authors suggested that lesions with weak toluidine blue staining should be viewed suspiciously, because their
molecular proles were essentially identical to lesions that stained strongly. Similar molecular
ndings were reported in a study of 100 oral premalignant lesions that also examined clinical outcome, with an average follow-up time of 44
months [43]. Although only 5% (3 of 64 cases)
of toluidine blue-negative lesions progressed to
cancer, carcinomatous transformation was observed in 33% (12 of 36 cases) of the toluidine
blue-positive lesions. This corresponded to
a greater than sixfold elevation in cancer risk (relative risk 6.67, 95% condence interval [CI]:
1.8723.70). Toluidine blue staining was associated with multiple LOH, especially including
LOH at 3p or 9p, and this, in turn, was associated
with a marked increased risk of carcinomatous
transformation (P .0002 or P ! .00001). Of
particular interest in this study, toluidine bluepositive lesions with minimal or no identiable
dysplasia on initial biopsy were almost fourfold
more likely to transform to carcinoma than lesions found to be toluidine blue-negative (relative
risk 3.92, 95% CI: 0.9216.80).
Use of tolonium chloride has also been of
reported benet in the follow-up of patients with
previously treated upper aerodigestive cancer. In
an examination of 96 biopsies performed in 81
patients, the sensitivity for detecting recurrent or
secondary disease by clinical examination alone

472

KALMAR

was 40% compared with 97% with vital staining

(P .0002) [44]. Because the positive predictive
values were similar for both arms of the study,
the authors noted that the increased sensitivity
with tolonium chloride did not come at the
expense of unnecessary biopsies (false-positive
results). In a separate report of 46 patients previously treated for oropharyngeal cancer, toluidine
blue was used to direct subsequent follow-up
punch biopsies of the stained tissue in these patients, together with nonstaining adjacent mucosa
in 34 cases [37]. Evidence of equivalent LOH was
noted in 25 of the 34 sample pair cases regardless
of staining status, with discordant LOH in the remaining cases. Of these, 8 of the 9 cases showed
a greater degree of LOH in the toluidine bluepositive sample compared with the unstained
sample. In addition, the authors found that 59%
of morphologically innocuous lesions initially
thought to be false-positive results contained
LOH, consistent with the hypothesis that toluidine blue staining may permit clinical detection
of altered DNA even if the tissue appears microscopically benign. Most recently, a smaller study
of 18 patients suggested that only dark toluidine
blue staining should be viewed as a positive result
[45]. The study was hampered by a high false-positive rate (31%) and the fact that all dark-stained
lesions in their series were clinically ulcerated.
Because this report stands in contrast to the
earlier molecular-associated ndings (similarly abnormal LOH patterns with dark- and light-stained
lesions), conrmatory studies are needed.

Diagnostic methods
Despite the growing number of adjuncts available to assist in the clinical evaluation of lesions
with uncertain biologic potential, surgical biopsy
remains by far the most popular means of obtaining a nal tissue diagnosis. Once a diagnosis is
established, additional studies (including imaging
modalities) may be needed to determine the stage
of disease and to guide treatment plan development. A variety of approaches have been used to
obtain diagnostic tissue samples of suspicious oral
lesions, and several are discussed here.
Punch biopsy
A punch biopsy is a soft tissue sampling instrument having a circular cutting edge of varying
diameter. It is most frequently used by dermatologists to sample skin lesions but can be used on

mucosal surfaces as well. Deep biopsies in areas

like the palate can be relatively simple to obtain
with a punch biopsy instrument; however, controlling the sample depth may be dicult, and
subsequent use of scissors or a scalpel is often
needed to free the specimen base from underlying
tissues. For study purposes, an advantage of the
punch instrument is its ability to provide reproducibly sized epithelial samples of lesion or
control tissues.
Scalpel biopsy
The simplest form of surgical sampling may be
the shave biopsy, where a shallow saucer-shaped
or elliptically shaped specimen (including a thin
layer of connective tissue) is removed using a
scalpel or curved razor blade. As with the use of
a punch biopsy, a shave biopsy is favored by
dermatologists for the diagnosis of supercial
lesions, such as actinic keratosis or early basal
cell carcinoma, in which evaluation of deep
margins is not considered essential. Because a determination of tissue invasion is critical to the
distinction between intraepithelial neoplasia (dysplasia or CIS) and oral squamous cell carcinoma,
use of a shave technique is typically not recommended for the diagnosis of suspicious intraoral
lesions.
The nal diagnosis for suspicious lesions of the
oral cavity is usually made on the basis of an
incisional or excisional scalpel biopsy. Excisional
biopsy is most often reserved for clinically benign
or, at worst, precancerous mucosal lesions that
are less than 2 cm in diameter. In cases in which
carcinoma is strongly expected, excision of lesional tissue should only be performed by the
surgeon who is to be directly involved with
denitive patient management. Otherwise, healing
of the surface mucosa may obscure the precise
location of the original lesion and hinder denitive treatment planning.
Most suspicious lesions of the oral cavity are
diagnosed through an incisional biopsy, where
a portion of the abnormal surface tissue is removed for histopathologic interpretation. As a
rule, the tissue sample should include the most
clinically suspicious portion of the lesion, including areas of erythroplakia, speckled leukoplakia,
surface granularity, or ulceration. Careful application of toluidine blue staining may be useful in
this setting by highlighting suspicious areas. For
lesions greater than 3 cm in diameter, the use of
multiple incisional biopsies and vital staining may

473

DETECTION AND DIAGNOSIS OF ORAL LESIONS

be warranted to help identify or exclude focal

carcinomatous transformation.
With oral precancerous or dysplastic lesions,
little correlation has been identied between grade
of dysplasia (mild, moderate, or severe) and the
risk of progression to cancer [4648]. In the absence of reliable prognostic information associated with morphology, molecular approaches
have been used to help identify genetic features
that might better dene the risk of progression
for a given lesion. These are discussed in more detail in the section on cytochemical and molecular
studies in this article.
In the case of squamous cell carcinoma,
predicting tumor behavior based on its microscopic features has also been an ongoing challenge
for the pathologist. Tumor grade, or degree of
dierentiation, has not been a satisfactory predictor of local recurrence or patient survival,
especially compared with tumor stage (tumor
extent). Although the thickness of early (T1)
squamous cell carcinoma of the tongue has been
strongly associated with the risk for regional node
metastasis and survival, it does not predict the risk
of local recurrence [4951]. A multiparameter
analysis of squamous cell carcinoma, incorporating variables like degree of keratinization, pattern
of invasion, nuclear pleomorphism, mitotic rate,
and lymphocytic response, has been advocated
by a number of authors to help predict local

recurrence and overall survival [5257]. The Martinez-Gimeno scoring system, an analysis of six
histologic criteria plus primary tumor size (T classication), was shown in a prospective study to
have a sensitivity of 100% (95% CI: 98%
100%) and a specicity of 55% (95% CI: 44%
66%) with a positive predictive value of 59%
(95% CI: 48%70%) and a negative predictive
value of 100% (95% CI: 98%100%) for the
risk of locoregional metastatic disease in cases of
oral squamous cell carcinoma [58].
Recently, the concept of multiparameter analysis was examined and modied by BrandweinGensler and colleagues [59] to produce a histologic
risk assessment system based on (1) perineural invasion greater than 1 mm involving nerves, (2)
lymphocytic response, and (3) worst pattern of invasion (WPOI) (Table 1). In a study of 292 patients
with cancer, the authors demonstrated that their
three-tiered system of risk assignment was strongly
predictive of local recurrence and overall survival
(log rank: P .0004 and P ! .0001, respectively)
across uniformly treated patients (Fig. 7). Margin
status, however, was not signicantly related to
disease recurrence or survival. This system provides a logical basis for the recommendation of adjuvant radiotherapy or chemotherapy for patients
with oral cancer, including the newly dened group
with T1/T2 N0/N1 tumors and negative resection margins but a risk score of greater than 3

Table 1
Proposed histopathologic risk assessment system for oral squamous cell carcinoma
Point assignment for risk scoring
Histologic variable
Perineural invasion
Lymphocytic inltrate
at interface
WPOI at interface

0
None
Continuous band

1
Small nerves
Large patches

3
Large nerves
Little or none

#1 or #2 or #3

#4

#5

Risk score
(sum of all point
assignments)

Risk for local

recurrence

Overall survival
probability

Low

Good

1 or 2

Intermediate

Intermediate

39

High

Poor

Adjuvant
treatment
recommendations
No local
disease-free benet
seen for adjuvant RT
No local
disease-free benet
seen for adjuvant RT
RT regardless of 5 mm
margins

Abbreviations: RT, radiotherapy; WPOI, worst patternal invasion.

From Brandwein-Gensler L, Teixeira MS, Lewis CM, et al. Oral squamous cell carcinoma: histologic risk assessment,
but not margin status, is strongly predictive of local disease-free and overall survival. Am J Surg Pathol 2005;29(2):175;
with permission.

474

KALMAR

retrospective series of salivary gland lesions diagnosed by FNA, where a sensitivity of 73% and
a specicity of 91% were recorded [62].
Sentinel node biopsy and cytology

Fig. 7. Kaplan-Meier overall survival curves classied

by risk assessment scoring system. (From BrandweinGensler M, Teixeira MS, Lewis CM, et al. Oral squamous cell carcinoma: histologic risk assessment, but
not margin status, is strongly predictive of local
disease-free and overall survival. Am J Surg Pathol
2005;29(2):175; with permission.)

(high-risk histologic features). Although prospective studies are needed to corroborate and extend
these ndings, the potential benets of this simple
yet elegant scoring system should be obvious to
clinicians and pathologists alike.
Fine-needle aspiration cytology
Fine-needle aspiration (FNA) cytology is
a valuable tool in the diagnosis of supercial
masses of the head and neck region. Although
most of these masses represent benign conditions,
testing for cancerous lesions can include cervical
or submandibular masses suspicious for metastatic nodal disease or conditions like primary
salivary gland malignancy or lymphoma. A good
discussion of this technique has been provided in
a previous issue of Oral and Maxillofacial Surgery
Clinics of North America [60]. More recently,
FNA has been applied to the concept of sentinel
node examination. Expertise in aspiration technique and cytologic interpretation of FNA specimens is essential for reliable results with this
procedure. Although tumor sampling has been
aided through guidance technology (ultrasound
or CT), sampling errors or diagnostic challenges
are reported with this technique and may necessitate subsequent open biopsy [61]. These limitations
have been documented in a large (n 6249)

Taken from its initial application with melanoma, the technique of investigating sentinel
node tissue has recently been applied to oropharyngeal malignancies, such as squamous cell
carcinoma. This procedure is intended to identify
micrometastatic disease within a sentinel node
considered most likely to drain the tumor bed and
receive initial metastatic deposits from the primary malignancy. Sentinel node biopsy thus
represents a less invasive means of providing
staging information for the patient with oral
cancer with an N0 neck.
The sentinel node technique uses lymphoscintigraphy, where the primary cancer site is initially
injected with a radioactive tracer material, such as
Tc 99m sulfur colloid. Dierent molecular weights
of this material can be selected depending on the
desired transit time for the study. Conventional
radiography is then used to locate the sentinel
node, and the patient is taken to the operating
room. For open biopsy, the surgeon may inject
a blue dye into the tumor bed to assist with visual
identication of the node, supplemented by an
intraoperative gamma detector. Use of a dye is
not always recommended for head and neck
tumors, because some authors claim that it can
interfere with node identication or even tumor
resection [63]. The node is then removed and examined histopathologically for micrometastatic
disease, often aided by serial sections and the
use of immunohistochemistry (IHC). Because
only 6 of 10 occult metastases from primary squamous cell carcinoma of the oral cavity primary
were reportedly detected using frozen sections, intraoperative evaluation of sentinel nodes does not
seem to be suciently reliable for routine use [64].
A recent meta-analysis of this approach for
squamous cell carcinoma of the oral cavity and
oral pharynx reported a pooled sensitivity of 92.6%
(95% CI: 0.8520.964) [65]. In a study of 50 patients
with oral, pharyngeal, or laryngeal cancer, 46 had
identiable sentinel nodes that were harvested by
open biopsy [63]. All patients subsequently underwent neck dissection (39 unilateral and 21 bilateral). Occult metastases were found by open
biopsy in 12 patients, and the authors noted that tumor detection required additional sectioning or
IHC in three cases. For 9 of the patients, the sentinel

DETECTION AND DIAGNOSIS OF ORAL LESIONS

node was the only one to show micrometastatic disease, whereas multiple positive nodes were found in
3 patients. In addition, no patient with a negative
sentinel node result was found to have tumor in
other nonsentinel lymph nodes. Ultrasound-guided
FNA cytology has also been used in an eort to
provide an even more conservative approach to
sentinel node assessment. Unfortunately, a lower
sensitivity rate of 42% to 73% has been reported
with ultrasound-guided specimens [66]. Some authors suggest an adjunctive role for FNA cytology
in the evaluation of the patient with N0 neck cancer
and have recommended that negative FNA results
be followed by open biopsy of the sentinel node
[65,66].
Another recent technology that has been used
together with sentinel lymph node biopsy is positron emission tomography (PET) using 18F-uoro2-deoxy-D-glucose (18FDG). This imaging study is
based on the increased metabolic activity of most
cancer cells that results in preferential uptake of radiolabeled glucose by tumors, such as squamous
cell carcinoma. In a prospective study involving resectable T1 to T3 lesions of oral and oropharyngeal
squamous cell carcinoma, PET and CT were obtained in 62 patients [67]. A total of 38 patients
with PET-negative ndings were subsequently
tested by sentinel node biopsy, including step-serial
sections and IHC analysis. Five of these patients
were found to have metastatic disease (PET falsenegative results) and were treated with neck dissection. Although no signicant dierences were
noted between PET and CT, negative neck sides
were better predicted by PET. Only 41 (33%) of
a possible 124 neck sides were treated after PET
staging, positive sentinel node biopsy, or intraoperative evaluation of tumor extension. In contrast, standard treatment guidelines and CT
examination would reportedly have resulted in
100 neck side procedures (81%). Importantly,
none of the 41 patients diagnosed as PET-negative
had evidence of clinical relapse, with a median follow-up of 33 months (range: 1052 months). The
authors proposed a staging ladder for clinically
N0 patients based on the high specicity of prerequisite PET examination followed by the high sensitivity of sentinel node biopsy, which may result in
fewer unnecessary neck dissection procedures. Finally, some authors suggest that use of combined
(fused) PET and CT imaging could provide an additional advantage in patient staging over either
modality alone or MRI [6870].
Complicating factors with sentinel node biopsy
and cytology include the rich lymphatic system of

475

the head and neck that can produce bilateral

drainage patterns by lymphoscintigraphy as well
as the complex anatomy that can make precise
localization and identication of suspicious nodal
tissue quite dicult. The nding of multiple
radioactive nodes can hinder determination of
the true sentinel or rst echelon node, with
some authors favoring harvest or sampling of the
three most strongly radioactive nodes [63,66,71].

Cytochemical and molecular studies

The diagnosis of oral precancerous and cancerous lesions continues to be made almost
exclusively on the basis of routine morphologic
evaluation of formalin-xed paran-embedded
tissue sections of scalpel biopsy specimens. Wellrecognized cytologic and architectural changes
associated with premalignant oral epithelial lesions are used to determine the presence and
degree of epithelial dysplasia. This time-honored
system represents the gold standard for identication of oral premalignancies and is used, at
least broadly, to predict biologic behavior or risk
of malignant transformation for a given precancerous lesion. Unfortunately, the earliest morphologic signs of dysplasia can be mimicked by a host
of reactive conditions, and numerous studies have
documented signicant variability (interobserver
and intraobserver) in the diagnosis of oral epithelial dysplasia. The predictive value of increasing
degrees of dysplasia for the risk of malignant
transformation is also unreliable. At the same
time, although the histologic diagnosis of squamous cell carcinoma is less susceptible to variability among pathologists, studies relating its
morphologic features to biologic behavior and
prognosis have only recently been reported.
Although most experts agree that cellular
alterations at the DNA level almost certainly
precede microscopic morphologic changes that
can be recognized by even the most experienced
pathologist, a consensus has yet to be reached as
to what parameter(s) might be most useful in the
diagnosis and management of oral lesions. In this
section, some of the chemical, IHC, and molecular
markers that have been used in the early characterization of oral epithelial dysplasias and squamous cell carcinomas are presented.
Nucleolar organizing regions
Nuclear organizing regions (NORs) are loops
of ribosomal DNA loops located on the short

476

KALMAR

arms of chromosomes 13, 14, 15, 21, and 22 and

are associated with acidic nonhistonic proteins
that can be visualized by silver-staining techniques
(argyrophilic) [72]. Because the number or size of
argyrophilic NORs (AgNORs) correlates positively with cellular proliferation, they have been
used to study a variety of neoplastic conditions,
including dysplastic and malignant oral epithelial
lesions, as has been previously discussed in an earlier issue of Oral and Maxillofacial Surgery Clinics
of North America [60].
Mean AgNOR counts were shown in a recent
report to be dier signicantly between nondysplastic (2.14) and dysplastic (2.65) clinical leukoplakias (95% CI: 0.6700.936), with a sensitivity of
75% and specicity of 83%, with a cuto mean
AgNOR value of 2.37 [73]. Using this cuto value,
a subsequent report compared AgNOR counts
with the gold standard of histopathologic diagnosis in 52 archival biopsy specimens. The test sensitivity was 67%, and the specicity was 59%,
whereas the false-positive and false-negative
rates were 41% and 33%, respectively [74]. The
authors noted that mean AgNOR count had
little correlation to the diagnosis of dysplasia and
suggested lowering the cuto value to reduce the
high false-negative rate. Even more recently,
mean AgNOR number, size, and percentage of total nuclear area were signicantly increased in 12
cases of squamous cell carcinoma compared with
corresponding normal patient tissue (P ! .01), although signicant case-to-case variability was
noted [75]. For example, although AgNOR size
was signicantly larger in 11 of the 12 carcinoma
cases, AgNOR number diered in only 8 of 12
cases, and the percentage area of the nucleus occupied by AgNORs varied in only 6 of 12 cases.
In two separate studies, one research group has
combined brush cytology sampling with AgNOR
counts. Using image analysis technology, a combined sensitivity of 98.2%, specicity of 100%,
positive predictive value of 100%, and negative
predictive value of 99.5% were reported for the
detection of cancer cells in 251 samples from 181
patients [72]. As discussed previously, however,
these ndings are confounded by the fact that
only 63% of the brush cytologyAgNOR results
were conrmed by scalpel biopsy, including only
57% (47of 83 lesions) described as leukoplakia.
In a follow-up paper using manual AgNOR counts,
a sensitivity of 92.5%, specicity of 100%, positive
predictive value of 100%, and negative predictive
value of 84.6% for the detection of squamous cell
carcinoma were reported in 337 samples from 75

patients [76]. Once again, corresponding biopsy results were obtained in only 64 (19%) of 337 cases.
Limitations with the technique include the time
and eort required to perform the study manually
as well as staining variability and counting subjectivity. Although computer-assisted image analysis has the potential to overcome some of these
problems, such hardware adds to the overall cost
and maintenance requirements.
Abnormal DNA segregation (DNA aneuploidy)
It has been recognized for many years that
abnormal chromosomal segregation resulting in
aneuploidy can be a marker of neoplastic transformation. In the view of some authors, errors in
DNA segregation may be one of the many causes
of cancer and not merely a result [77]. Abnormal
chromosomal content is thought to be the most
common characteristic of solid tumors in human
beings. Discussion of abnormal DNA content
was provided in a previous issue of Oral and Maxillofacial Surgery Clinics of North America in an
article on ow cytometric analysis [60]. Since
then, studies using ow and image cytometry
have provided additional information regarding
the diagnostic usefulness of this technique
[72,7884]. In a series of 25 resected oral squamous cell carcinomas, all tumors were found by
image analysis to be aneuploid and multiploidy
was seen in 15 cases (60%); however, aneuploidy
alone did not seem to be related to clinical progression [78]. A total of 29 mucosal lesions in 21
patients that progressed to carcinoma were compared with 29 control lesions that did not progress
(mean follow-up of 112 months). The lesions were
matched for location and level of dysplasia [79].
Using quantitative image analysis, the progressive
lesions exhibited signicantly greater levels of
DNA aberration than controls (P .0096). Of
clinical importance, 3 lesions initially judged to
be reactive by histopathologic examination but
found to be nondiploid or aneuploid by DNA
analysis all progressed to carcinoma. Comparison
of ow cytometric and image cytometric results
was performed in another study involving 32 cases
of oral squamous cell carcinoma [80]. Image cytometry of stained sections was found to be
more sensitive in detecting abnormal cell DNA
content, and the presence of aneuploidy had prognostic implications. Nine of the 32 patients died
within 5 years of initial treatment, and the tumors
had greater than 10% abnormal cellular DNA in
all 9 cases. To test the prognostic value of this

DETECTION AND DIAGNOSIS OF ORAL LESIONS

nding, a multivariate analysis was performed on

195 patients with oral cancer. In descending order
of importance, the three independent variables
found to have a statistical association with survival were (1) abnormal DNA content, (2) clinical
stage, and (3) growth pattern (endophytic versus
exophytic).
Combining brush cytology with DNA cytometry has been used to provide a less invasive means
of patient sample analysis [72,76,82]. This combination was compared with incisional biopsy in
a report of 98 patients, with a sensitivity of
100% and specicity of 97.4% [82]. Aneuploidy
was found in 1 of 21 leukoplakia cases, 3 of 3
erythroplakia cases, and 15 of 15 squamous cell
carcinoma cases. The ndings of Remmerbach
and colleagues [76] indicated that DNA aneuploidy
was detectable in brush cytology specimens from 1
to 15 months before histologic evidence could
conrm the presence of malignancy. In contrast,
ow cytometry results were recently reported in
67 cases of oral squamous cell carcinoma [83].
Although 27% of the tumors were aneuploid, the
authors found no relation between ploidy status and local recurrence, distant metastases, or
survival.
Despite promising results, there has been recent public acknowledgment of signicant scientic fraud by one of the most active proponents of
aneuploidy analysis in the study of oral cancer
[84,85]. This disclosure makes it impossible to
summarize our current knowledge with any certainty until the evidence presented in several reports is re-examined and the conclusions are
revised as necessary. Because of the ongoing controversy, the future of aneuploidy in the diagnosis
and management of oral precancerous and cancerous lesions is unclear. Readers are advised to
stay abreast of the related scientic literature as
the story unfolds.
DNA alteration (loss of heterozygosity (LOH))
As mentioned previously, the progression of
oral epithelium from a benign to malignant process begins at the genetic (DNA) level and is
ultimately expressed at the cellular and clinical
levels. For some lesions, such as a true leukoplakia with no observable dysplasia on biopsy,
a clinical lesion may even precede the pathologists ability to detect histomorphologic evidence
of premalignancy. It is further recognized that
carcinogenesis does not result from a single area
of DNA damage but is a multistep process that

477

requires an accumulation of several DNA alterations collectively resulting in uncontrolled neoplastic growth.
Recent work has supported the concept that
premalignant oral lesions may have an identiable
genetic prole associated with the risk for malignant transformation. Using microsatellite analysis
to detect areas of loss of allelic balance or LOH,
early DNA changes in precancerous oral lesions
have been found in the 3p and 9p chromosomal
regions [86]. For a given lesion, the risk for progression to cancer was low if no LOH was found,
intermediate if LOH at 3p and 9p was found, and
high if LOH at 3p and 9p was seen together with
additional areas of genetic damage [86,87]. Overall, lesions with LOH at 3p and 9p plus other dened chromosomal areas had a 33-fold increased
risk for progression to squamous cell carcinoma
compared with lesions without LOH [86,88].
Others have reported that only 2% of low-risk lesions by LOH analysis are likely to progress to
cancer over a 5-year period compared with 50%
of high-risk lesions [16,87,89]. In examination of
the known high-risk zones for oral cancer (see
Fig. 2B), genetic analysis showed that leukoplakia
from sites of high risk (71 cases) possessed
a greater degree of LOH than similar lesions in
low-risk sites (56 cases) [90].
In addition to permitting insight into the risk
of progression for a given lesion, the discovery
that clinically and microscopically normal margins can harbor genetic damage signicantly alters
the concept of a clear or negative excisional
margin with oral precancerous conditions. In
a study of 66 dysplastic lesions designed to assess
the treatment impact on patient outcome, such
clinical features as sex (male versus female),
history of smoking (nonsmoker versus ever
smoker), location (high-risk versus low-risk site),
and appearance (homogeneous versus nonhomogeneous) were not associated with lesion progression or recurrence [87]. Likewise, the
histologic grade of dysplasia (mild or moderate)
was not related to progression. Using LOH analysis to assign low-, intermediate-, and high-risk
patterns, the authors found that although lesion
treatment (surgical removal of clinically abnormal
tissue) reduced progression to cancer for lesions of
all LOH patterns, the reduction was not statistically signicant. To further examine this nding,
repeat biopsy was performed on 19 patients at
the site of the original excision. In 17 patients,
LOH patterns observed in the repeat biopsy indicated incomplete removal of the initial lesion.

478

KALMAR

Importantly, in 8 of the 17 cases, there was no

clinical evidence of mucosal abnormality at the
time of the second biopsy. When the treatment
impact was reassessed by combining molecular
and clinical criteria (evidence of residual clones,
completeness of surgical removal, or clinical evidence of recurrence), the risk of progression risk
was signicantly reduced in cases with intermediate- and high-risk LOH patterns.
LOH is also seen in virtually all cases of
squamous cell carcinoma and may be useful in
predicting biologic behavior. Allelic imbalance
has been reported at several chromosome arms,
including 3p, 4q, 5q, 7q, 8p, 10q, 11q, 13q, 18q,
20q, and 22q [91]. In addition, accumulation of
multiple LOH seems to be related to the risk of tumor recurrence. In a study of 68 patients with previously treated oral cancer, biopsies of subsequent
leukoplakic lesions that did or did not progress to
a second oral malignancy were performed [92].
Progressing lesions were 26 times more likely to
exhibit LOH at 3p or 9p than nonprogressive lesions. In contrast, histopathologic evidence of
dysplasia was not associated with increased risk
of a second malignancy. Such data arm the
idea that molecular evaluation of lesional tissue
and margin status may be more informative
than routine histopathologic evaluation in the
management of oral squamous cell carcinoma
and precancerous disease.
The future
Despite the clinical promise shown by molecular techniques in areas ranging from early diagnosis
to the follow-up of previously treated patients with
oral cancer, such technology is primarily used at
a handful of major clinical research centers in
North America and is not routinely available to
most oral and maxillofacial surgeons. Limiting
factors include the added costs and requirement
for additional equipment, the complexity of the
tests themselves, and the need to calibrate for those
who interpret test results. The natural reluctance
of practitioners to adopt new techniques or protocols, even in the face of compelling evidence, also
makes the standard of care for precancerous and
cancerous lesions slow to change.
Ideally, diagnostic adjuncts or tests should be
relatively aordable, should be simple to perform,
and should use easily obtainable patient samples or
specimens. As has been done with AgNOR counting, the combination of brush cytology sampling
with thin-lm slide preparation and molecular

nucleic acid analysis may provide optimal sample

material for studies of LOH or other genetic
abnormalities using specimens obtained directly
from clinically normal or abnormal mucosa.
It has also been shown that tumor markers or
associated biomarkers can be detected in the
serum and, more recently, the saliva of patients
with cancer [9395]. Saliva, especially, has many
ideal characteristics for future diagnostic applications that range from routine screening to posttreatment follow-up of the patient with oral
cancer. Saliva is easily accessible, is collectible by
noninvasive means, and has previously been
shown to contain identiable DNA abnormalities
in patients with oral squamous cell carcinoma
[96,97]. Recent evidence using the reverse transcriptase polymerase chain reaction (RT-PCR)
has also revealed quantiable levels of mRNA in
saliva [95]. Using microarray analysis, Li and colleagues [95] examined the saliva from 10 patients
with recently diagnosed oral cancer in comparison
to age and sex-matched controls with comparable
smoking histories. From a total of 10,316 transcripts, 1679 were shown to dier signicantly
(up- and downregulated) between patient and
control samples. Using more stringent selection
criteria, the authors presented a total of seven salivary mRNAs as candidate cancer-related biomarkers, including interleukin (IL)-8, IL-1B,
DUSP1, HA3, OAZ1, S100P, and SAT. Together,
these biomarkers gave a sensitivity of 91% and
specicity of 91% for distinguishing patients
with oral cancer from controls. Of additional interest, these researchers have recently reported
that salivary IL-8 mRNA and protein levels, as
measured by quantitative RT-PCR and ELISA,
respectively, were signicantly higher in patients
with oral cancer compared with matched controls
[98]. Serum IL-6 mRNA and protein levels were
also elevated in these same patients with cancer.
The combination of salivary IL-8 and serum IL6 gave a sensitivity of 99% and a specicity of
90% for detecting oral squamous cell carcinoma.
Finally, serum analysis by matrix-assisted laser
desorption and ionization time-of-ight mass
spectrometry was recently used to evaluate specimens from 57 patients with oral cancer and 29
control patients [99]. In this technique, chip-based
arrays are used to bind various proteins through
a number of interactions, including hydrophobic
and/or hydrophilic, anionic and/or cationic, and
metal-binding properties. The technology permits
large numbers of specimens to be screened simultaneously, but it is expensive and requires

DETECTION AND DIAGNOSIS OF ORAL LESIONS

additional characterization work for proteins of

interest. Of several proteins initially identied,
the authors reported that a C-terminal fragment
of the brinogen a-chain was the most highly predictive marker for cancer, with a sensitivity of
100% and specicity of 96.6%. Elevated tissue brinogen levels have previously been reported in
association with breast cancer, small cell carcinoma of the lungs, and melanoma.
Technology is poised to play an increasingly
active role in the diagnosis and management of
patients with oral precancerous and cancerous
lesions. Until more laboratories, including those
based in smaller community hospitals, acquire the
needed molecular technologies to perform new
more diagnostically robust sample analyses, routine histopathologic examination is likely to remain the standard of care for most patients with
oral precancerous and cancerous disease. Careful
prospective examination of the recently proposed
scoring system for oral squamous cell carcinoma
(see Table 1) may represent an important step toward bridging the gap between current clinical
and laboratory practice and the molecular diagnostics of the future.

References
[1] Jemal A, Murray T, Ward E, et al. Cancer statistics,
2005. CA Cancer J Clin 2005;55(1):1030.
[2] Morse DE, Kerr AR. Disparities in oral and pharyngeal cancer incidence, mortality and survival among
black and white Americans. J Am Dent Assoc 2006;
137(2):20312.
[3] Neville BW, Damm DD, Allen CM, et al. Epithelial pathology. In: Oral and maxillofacial pathology. 2nd edition. Philadelphia: WB Saunders;
2002. p. 33745.
[4] Waldron CA, Shafer WG. Leukoplakia revisited. A
clinicopathologic study of 3256 oral leukoplakias.
Cancer 1975;36(4):138692.
[5] Moore C, Catlin D. Anatomic origins and locations
of oral cancer. Am J Surg 1967;114(4):5103.
[6] Hansen LS, Olson JA, Silverman S Jr. Proliferative
verrucous leukoplakia. A long-term study of thirty
patients. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 1985;60(3):28598.
[7] Silverman S Jr, Gorsky M. Proliferative verrucous
leukoplakia: a follow-up study of 54 cases. Oral
Surg Oral Med Oral Pathol Oral Radiol Endod
1997;84(2):1547.
[8] Shafer WG, Waldron CA. Erythroplakia of the oral
cavity. Cancer 1975;36(3):10218.
[9] Reichart PA, Philipsen HP. Oral erythroplakiad
a review. Oral Oncol 2005;41(6):55161.

479

[10] Mashberg A, Samit A. Early diagnosis of asymptomatic oral oropharyngeal squamous cancers. CA
Cancer J Clin 1995;45(6):32851.
[11] Bundgaard T, Wildt J, Frydenberg M, et al. Casecontrol study of squamous cell cancer of the oral
cavity in Denmark. Cancer Causes Control 1995;
6(1):5767.
[12] Muscat JE, Richie JP Jr, Thompson S, et al. Gender
dierences in smoking and risk for oral cancer. Cancer Res 1996;56(22):51927.
[13] Schildt EB, Eriksson M, Hardell L, et al. Oral snu,
smoking habits and alcohol consumption in relation
to oral cancer in a Swedish case-control study. Int J
Cancer 1998;77(3):3416.
[14] Lewin F, Norell SE, Johansson H, et al. Smoking tobacco, oral snu, and alcohol in the etiology of squamous cell carcinoma of the head and neck:
a population-based case-referent study in Sweden.
Cancer 1998;82(7):136775.
[15] Schwartz SM, Daling JR, Doody DR, et al. Oral
cancer risk in relation to sexual history and evidence
of human papillomavirus infection. J Natl Cancer
Inst 1998;90(21):162636.
[16] Lee JJ, Hong WK, Hittelman WN, et al. Predicting
cancer development in oral leukoplakia: ten years of
translational research. Clin Cancer Res 2000;6(5):
170210.
[17] Boetta P, Aagnes B, Weiderpass E, et al. Smokeless
tobacco use and risk of cancer of the pancreas and
other organs. Int J Cancer 2005;114(6):9925.
[18] Rosenquist K, Wennerberg J, Schildt EB, et al. Use
of Swedish moist snu, smoking and alcohol
consumption in the aetiology of oral and oropharyngeal squamous cell carcinoma. A population-based
case-control study in southern Sweden. Acta Otolaryngol 2005;125(9):9918.
[19] Accortt NA, Waterbor JW, Beall C, et al. Cancer incidence among a cohort of smokeless tobacco users
(United States). Cancer Causes Control 2005;16(9):
110715.
[20] Henley SJ, Thun MJ, Connell C, et al. Two large
prospective studies of mortality among men who
use snu or chewing tobacco (United States). Cancer
Causes Control 2005;16(4):34758.
[21] Folsom TC, White CP, Bromer L, et al. Oral exfoliative study. Review of the literature and report of
a three-year study. Oral Surg Oral Med Oral Pathol
Oral Radiol Endod 1972;33(1):6174.
[22] Shklar B, Cataldo E, Meyer I. Reliability of cytologic smear in diagnosis of oral cancer. A controlled
study. Arch Otolaryngol 1970;91(2):15860.
[23] Rovin S. An assessment of the negative oral cytologic
diagnosis. J Am Dent Assoc 1967;74(4):75962.
[24] Sciubba JJ. Improving detection of precancerous
and cancerous oral lesions. J Am Dent Assoc 1999;
130(10):144557.
[25] Rick GM, Slater L. Oral brush biopsy: the problem
of false positives. Oral Surg Oral Med Oral Pathol
Oral Radiol Endod 2003;96(3):252.

480

KALMAR

[26] Svirsky JA, Burns JC, Carpenter WM, et al. Comparison of computer-assisted brush biopsy results
with follow up scalpel biopsy and histology. Gen
Dent 2002;50(6):5003.
[27] Scheifele C, Schmidt-Westhausen AM, Dietrich T,
et al. The sensitivity and specicity of the OralCDx
technique: evaluation of 103 cases. Oral Oncol
2004;40(8):8248.
[28] Poate TW, Buchanan JA, Hodgson TA, et al. An audit of the ecacy of the oral brush biopsy technique
in a specialist oral medicine unit. Oral Oncol 2004;
40(8):82934.
[29] Potter TJ, Summerlin DJ, Campbell JH. Oral malignancies associated with negative transepithelial
brush biopsy. J Oral Maxillofac Surg 2003;61(6):
6747.
[30] Huber MA, Bsoul SA, Terezhalmy GT. Acetic acid
wash and chemiluminescent illumination as an adjunct to conventional oral soft tissue examination
for the detection of dysplasia: a pilot study. Quintessence Int 2004;35(5):37884.
[31] Ram S, Siar CH. Chemiluminescence as a diagnostic
aid in the detection of oral cancer and potentially
malignant epithelial lesions. Int J Oral Maxillofac
Surg 2005;34(5):5217.
[32] Kerr AR, Sirois DA. Clinical evaluation of a new adjunctive technique for oral mucosal examinations.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
2004;97(4):451.
[33] Lane PM, Gilhuly T, Whitehead P, et al. Simple device for the direct visualization of oral-cavity tissue
uorescence. J Biomed Opt 2006;11(2):024006.
[34] Zhang L, Ng SP, Williams M, et al. Detection of
high-risk oral premalignant lesions by multi-spectral
uorescent visualization. J Dent Res 2004;83(Spec
Issue A):1242.
[35] Poh CF, Anderson DW, Williams PM, et al. Identication of clinically occult disease in oral cancer patients in the operating room (OR): a real-time
application of uorescence visualization. Proc Am
Assoc Cancer Res 2005;46:1592.
[36] Niebel HH, Chomet B. In vivo staining test for delineation of oral intraepithelial neoplastic change: preliminary report. J Am Dent Assoc 1964;68:8016.
[37] Guo Z, Yamaguchi K, Sanchez-Cespedes M, et al.
Allelic losses in OraTest-directed biopsy of patients
with prior upper aerodigestive tract malignancy.
Clin Cancer Res 2001;7(7):19638.
[38] Rosenberg D, Cretin S. Use of meta-analysis to evaluate tolonium chloride in oral cancer screening. Oral
Surg Oral Med Oral Pathol Oral Radiol Endod
1989;76(5):6217.
[39] Warnakulasuriya KA, Johnson NW. Sensitivity and
specicity of OraScan toluidine blue mouthrinse in
the detection of oral cancer and precancer. J Oral
Pathol Med 1996;25(3):97103.
[40] Epstein JB, Oakley C, Millner A, et al. The utility of
toluidine blue application as a diagnostic aid in patients previously treated for upper oropharyngeal

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

carcinoma. Oral Surg Oral Med Oral Pathol Oral

Radiol Endod 1997;83(5):53747.
Onofre MA, Sposto MR, Navarro CM. Reliability
of toluidine blue application in the detection of
oral epithelial dysplasia and in situ and invasive
squamous cell carcinomas. Oral Surg Oral Med
Oral Pathol Oral Radiol Endod 2001;91(5):
53540.
Epstein JB, Zhang L, Poh C, et al. Increased allelic
loss in toluidine blue-positive oral premalignant lesions. Oral Surg Oral Med Oral Pathol Oral Radiol
Endod 2003;95(1):4550.
Zhang L, Williams M, Poh CF, et al. Toluidine blue
staining identies high-risk primary oral premalignant lesions with poor outcome. Cancer Res 2005;
65(17):801721.
Epstein JB, Feldman R, Dolor RJ, et al. The utility
of tolonium chloride rinse in the diagnosis of recurrent or second primary cancers in patients with prior
upper aerodigestive tract cancer. Head Neck 2003;
25(11):91121.
Gandolfo S, Pentenero M, Broccoletti R, et al. Toluidine blue uptake in potentially malignant oral lesions in vivo: clinical and histological assessment.
Oral Oncol 2006;42(1):8995.
Silverman S Jr, Gorsky M, Lozada F. Oral leukoplakia and malignant transformation. A follow-up
study of 257 patients. Cancer 1984;53(3):5638.
Mincer HH, Coleman SA, Hopkins KP. Observations on the clinical characteristics of oral lesions
showing histologic epithelial dysplasia. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1972;
33(3):38999.
Cruz IB, Snijders PJ, Meijer CJ, et al. p53 expression
above the basal cell layer in oral mucosa is an early
event of malignant transformation and has predictive value for developing oral squamous cell carcinoma. J Pathol 1998;184(4):3608.
Kim HC, Kusukawa J, Kameyama T. Clinicopathologic parameters in predicting cervical nodal metastasis in early squamous cell carcinoma of the oral
cavity. Kurume Med J 1993;40(4):18392.
Martinez-Gimeno C, Rodriguez EM, Vila CN, et al.
Squamous cell carcinoma of the oral cavity: a clinicopathologic scoring system for evaluating risk of cervical lymph node metastasis. Laryngoscope 1995;
105(7 Pt 1):72833.
Po Wing Yuen A, Lam KY, Lam LK, et al. Prognostic factors of clinically stage I and II oral tongue carcinoma. A comparative study of stage, thickness,
shape, growth pattern, invasive front malignancy
grading, Martinez-Gimeno score, and pathologic
features. Head Neck 2002;24(6):51320.
Bryne M, Koppang HS, Lilleng R, et al. New malignancy grading is a better prognostic indicator than
Broders grading in oral squamous cell carcinomas.
J Oral Pathol Med 1989;18(8):4327.
Bryne M, Koppang HS, Lilleng R, et al. Malignancy
grading of the deep invasive margins of oral

DETECTION AND DIAGNOSIS OF ORAL LESIONS

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

squamous cell carcinomas has high prognostic value.

J Pathol 1992;166(4):37581.
Odell EW, Jani P, Sherri M, et al. The prognostic
value of individual histologic grading parameters
in small lingual squamous cell carcinomas. The importance of the pattern of invasion. Cancer 1994;
74(3):78994.
Bryne M, Jenssen N, Boysen M. Histological grading in the deep invasive front of T1 and T2 glottic
squamous cell carcinomas has high prognostic value.
Virchows Arch 1995;427(3):27781.
Spiro RH, Guillamondegui O Jr, Paulino AF, et al.
Pattern of invasion and margin assessment in patients with oral tongue cancer. Head Neck 1999;
21(5):40813.
Bundgaard T, Rossen K, Henriksen SD, et al. Histopathologic parameters in the evaluation of T1 squamous cell carcinomas of the oral cavity. Head Neck
2002;24(7):65660.
Martinez-Gimeno C, Molinero AP, Castro V, et al.
Prospective validation of the Martinez-Gimeno clinicopathologic scoring system (MGSS) for evaluating
risk of cervical lymph node metastases of squamous
cell carcinoma of the oral cavity. Head Neck 2005;
27(4):3205.
Brandwein-Gensler M, Teixeira MS, Lewis CM,
et al. Oral squamous cell carcinoma: histologic risk
assessment, but not margin status, is strongly predictive of local disease-free and overall survival. Am J
Surg Pathol 2005;29(2):16778.
Kerpel SM, Fornatora M, Freedman PD. Advances
and new concepts in oral and maxillofacial pathology. Oral Maxillofac Surg Clin North Am 1997;
9(3):299315.
Stanley MW. Selected problems in ne needle aspiration of head and neck masses. Mod Pathol 2002;
15(3):34250.
Hughes JH, Volk EE, Wilbur DC. Pitfalls in salivary
gland ne-needle aspiration cytology: lessons from
the College of American Pathologists Interlaboratory Comparison Program in Nongynecologic
Cytology. Arch Pathol Lab Med 2005;129(1):2631.
Hoft S, Maune S, Muhle C, et al. Sentinel lymphnode biopsy in head and neck cancer. Br J Cancer
2004;91(1):1248.
Civantos FJ, Gomez C, Duque C, et al. Sentinel
node biopsy in oral cavity cancer: correlation with
PET scan and immunohistochemistry. Head Neck
2003;25(1):19.
Paleri V, Rees G, Arullendran P, et al. Sentinel node
biopsy in squamous cell cancer of the oral cavity and
oral pharynx: a diagnostic meta-analysis. Head
Neck 2005;27(9):73947.
Nieuwenhuis EJ, van der Waal I, Leemans CR, et al.
Histopathologic validation of the sentinel node concept in oral and oropharyngeal squamous cell carcinoma. Head Neck 2005;27(2):1508.
Kovacs AF, Dobert N, Gaa J, et al. Positron emission tomography in combination with sentinel

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]
[78]

[79]

[80]

[81]

481

biopsy reduces the rate of elective neck dissections

in the treatment of oral and oropharyngeal cancer.
J Clin Oncol 2004;22(19):397380.
Syed R, Bomanji JB, Nagabhushan N, et al. Impact
of combined (18)F-FDG PET/CT in head and neck
tumours. Br J Cancer 2005;92(6):104650.
Menda Y, Graham MM. Update on 18F-uorodeoxyglucose/positron emission tomography and positron emission tomography/computed tomography
imaging of squamous head and neck cancers. Semin
Nucl Med 2005;35(4):2149.
Muylle K, Castaigne C, Flamen P. 18F-uoro-2deoxy-D-glucose positron emission tomographic
imaging recent developments in head and neck cancer. Curr Opin Oncol 2005;17(3):24953.
Werner JA, Dunne AA, Ramaswamy A, et al. Sentinel node detection in N0 cancer of the pharynx and
larynx. Br J Cancer 2002;87(7):7115.
Remmerbach TW, Weidenbach H, Pomjanski N,
et al. Cytologic and DNA-cytometric early diagnosis
of oral cancer. Anal Cell Pathol 2001;22(4):21121.
Chattopadhyay A, Ray JG, Caplan DJ. AgNOR
count as objective marker for dysplastic features in
oral leukoplakia. J Oral Pathol Med 2002;31(9):
5127.
Ray JG, Chattopadhyay A, Caplan DJ. Usefulness
of AgNOR counts in diagnosing epithelial dysplasia.
J Oral Pathol Med 2003;32(1):716.
Guler N, Uckan S, Celik I, et al. Expression of
Fas and Fas-ligand and analysis of argyrophilic
nucleolar organizer regions in squamous cell carcinoma: relationships with tumor stage and grade,
and apoptosis. Int J Oral Maxillofac Surg 2005;34:
9006.
Remmerbach TW, Weidenbach H, Muller C, et al.
Diagnostic value of nucleolar organizer regions
(AgNORs) in brush biopsies of suspicious lesions
of the oral cavity. Anal Cell Pathol 2003;25(3):
13946.
Sen S. Aneuploidy and cancer. Curr Opin Oncol
2000;12(1):828.
Staibano S, Mignogna MD, Lo Muzio L, et al. Overexpression of cyclin-D1, bcl-2, and bax proteins,
proliferating cell nuclear antigen (PCNA), and
DNA-ploidy in squamous cell carcinoma of the
oral cavity. Hum Pathol 1998;29(11):118994.
Hogmo A, Lindskog S, Lindholm J, et al. Preneoplastic oral lesions: the clinical value of image cytometry DNA analysis, p53 and p21/WAF1 expression.
Anticancer Res 1998;18(5B):364550.
Kim J, Shin DM, El-Naggar A, et al. Chromosome
polysomy and histological characteristic in oral premalignant lesions. Cancer Epidemiol Biomarkers
Prev 2001;10(4):31925.
Noguchi M, Kinjyo H, Kohama GI, et al. Invasive
front in oral squamous cell carcinoma: image and
ow cytometric analysis with clinicopathologic correlation. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 2002;93(6):6827.

482

KALMAR

[82] Maraki D, Becker J, Boecking A. Cytologic and

DNA-cytometric very early diagnosis of oral cancer.
J Oral Pathol Med 2004;33(7):398404.
[83] Saiz-Bustillo R, Corchero-Martin G, GarciaMontesinos-Perea B, et al. Oral squamous cell
carcinoma. Cytometric parameters of prognostic interest. Med Oral Patol Oral Cir Bucal 2005;10(5):
4627.
[84] Couzin J, Schirber M. Scientic misconduct. Fraud
upends oral cancer eld, casting doubt on prevention trial. Science 2006;311(5760):4489.
[85] Curfman GD, Morrissey S, Drazen JM. Expression
of concern [letter to the editor]. N Engl J Med 2006;
354(6):638.
[86] Mao L. Can molecular assessment improve classication of head and neck premalignancy? Clin Cancer
Res 2000;6(2):3212.
[87] Zhang L, Poh CF, Lam WL, et al. Impact of localized treatment in reducing risk of progression of
low-grade oral dysplasia: molecular evidence of incomplete resection. Oral Oncol 2001;37(6):50512.
[88] Rosin MP, Cheng X, Poh C, et al. Use of allelic loss
to predict malignant risk for low-grade oral epithelial dysplasia. Clin Cancer Res 2000;6(2):35762.
[89] Epstein JB, Zhang L, Rosin M. Advances in the diagnosis of oral premalignant and malignant lesions.
J Can Dent Assoc 2002;68(10):61721.
[90] Zhang L, Cheung KJ Jr, Lam WL, et al. Increased
genetic damage in oral leukoplakia from high risk
sites: potential impact on staging and clinical management. Cancer 2001;91(11):4855.
[91] Nagpal JK, Das BR. Oral cancer: reviewing the present understanding of its molecular mechanism and

[92]

[93]

[94]
[95]

[96]

[97]

[98]

[99]

exploring the future directions for its eective management. Oral Oncol 2003;39(3):21321.
Rosin MP, Lam WL, Poh C, et al. 3p14 and 9p21
loss is a simple tool for predicting second oral malignancy at previously treated oral cancer sites. Cancer
Res 2002;62(22):644750.
Hamana K, Uzawa K, Ogawara K, et al. Monitoring of circulating tumour-associated DNA as a prognostic tool for oral squamous cell carcinoma. Br J
Cancer 2005;92(12):21814.
Sidransky D. Nucleic acid-based methods for the detection of cancer. Science 1997;278(5340):10549.
Li Y, St. John MA, Zhou X, et al. Salivary transcriptome diagnostics for oral cancer detection. Clin Cancer Res 2004;10(24):844250.
El-Naggar AK, Mao L, Staerkel G, et al. Genetic
heterogeneity in saliva from patients with oral
squamous carcinomas: implications in molecular
diagnosis and screening. J Mol Diagn 2001;3(4):
16470.
Liao PH, Chang YC, Huang MF, et al. Mutation of
p53 gene codon 63 in saliva as a molecular marker
for oral squamous cell carcinomas. Oral Oncol
2000;36(3):2726.
St. John MA, Li Y, Zhou X, et al. Interleukin 6
and interleukin 8 as potential biomarkers for oral
cavity and oropharyngeal squamous cell carcinoma.
Arch Otolaryngol Head Neck Surg 2004;130(8):
92935.
Cheng AJ, Chen LC, Chien KY, et al. Oral cancer
plasma tumor marker identied with bead-based afnity-fractioned proteomic technology. Clin Chem
2005;51(12):223644.

Oral Maxillofacial Surg Clin N Am 18 (2006) 483491

Molecular Biology and Clinical Behavior

of Oral Cancer
Brian L. Schmidt, DDS, MD, PhD
Department of Oral and Maxillofacial Surgery, University of California,
San Francisco, 521 Parnassus Avenue, C-522, PO Box 0440,
San Francisco, CA 941430440, USA

Our understanding of the molecular biology

of oral squamous cell carcinoma (SCC) has
progressed signicantly over the past decade.
The profound increase in basic science knowledge, however, has not aected our ability as
clinicians to control oral SCC or provided us
with tools to improve patient outcome. The
challenges we face as oral and maxillofacial
surgeons providing comprehensive surgical management for patients with oral cancer have
remained the same: diculty in predicting the
capricious clinical behavior of oral cancer, recurrence at the primary site after resection,
cervical and distant metastasis, and the development of second primary oral cancers. The causes
and the solutions of these clinical challenges have
a molecular basis. As molecular technologies advance, genetic and proteomic approaches are
likely to be integrated into clinical practice. Molecular approaches are clearly going to be used to
predict clinical behavior, determine prognosis,
guide surgical treatment, and assist with tumor
surveillance. In this article, I rst review the traditional histopathologic features that have been
used to treat patients with oral cancer. I then
present some of the more recent molecular studies and technologies that we, as surgeons, might
be using in the future to tip the balance in our
patients favor.

E-mail address: brian.schmidt@ucsf.edu

Traditional histopathologic features used

to manage oral cancer
Histopathologic evaluation of margins
In discussing the impact of histologic margins
on outcome, one of the primary diculties is the
lack of a clear denition among clinicians, pathologists, and investigators regarding what is
meant by clear and close margins as well as
the distinction between mucosal and deep margins
[15]. Although surgeons typically try to resect
oral cancer with a 1-cm margin of clinically normal tissue, pathologic evaluation of the specimen
almost always demonstrates signicantly less normal tissue surrounding the cancer. This reduced
margin can partially be explained by the tissue
shrinkage that occurs with specimen processing.
A study in dogs demonstrated that 30% to 50%
tissue shrinkage (from the clinical to histologic
margin) occurs with specimen processing [6]. Another issue that can compound the interpretation
of margins and studies evaluating margins in
oral cancer is the site of the primary cancer. Woolgar and Triantafyllou [5] have demonstrated that
the oral cancer subsite signicantly inuences the
status of the margins. Using the denition of
a 1-mm margin as an involved margin, these authors showed that the percentage of involved margins was highest in maxillary alveolar (45%),
retromolar (38%), buccal mucosa (33%), mandibular alveolus (17%), and tongue (11%) SCC [5].
Not surprisingly, the incidence of involved margins increased with the tumor T stage in this study
[5]. Even when histologic margins are clear, recurrences occur and are most likely secondary to retained histologically normal but genetically

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.010

oralmaxsurgery.theclinics.com

484

SCHMIDT

altered mucosa [7]. As molecular technology advances in terms of eciency and sensitivity, it is
anticipated that molecular margin analysis should
signicantly improve surgeons ability to obtain
histologically and genetically clear margins.
Tumor grading
Although Broders tumor grading system [8] of
well, moderately, and poorly dierentiated carcinomas is commonly used in pathology reports,
the value of tumor grading in the management
of oral SCC remains equivocal. Tumor grade
has been investigated to determine whether higher
grade correlates with an increased cervical metastasis rate. Byers and colleagues [9] looked at oral
tongue SCCs and demonstrated a signicant association between tumor grade and metastasis. It is
important to note in this report, however, that the
clinically negative nodes in the patients with T1
and T2 lesions were not analyzed. A more recent
study demonstrated no association between tumor
grade and subclinical nodal metastasis [10]. Tumor grade is currently not used in clinical practice
to predict metastasis or to guide treatment of oral
SCC.
Tumor size
Tumor size (T in the tumor node metastasis
[TNM] staging system) [11] has not been particularly eective in predicting cervical metastasis.
Multiple studies have demonstrated that even T1
tongue SCCs are associated with a signicant
rate (greater than 20%) of cervical metastasis
(Table 1). It is generally agreed that if the risk
of cervical metastasis is greater than 20%, treatment of the neck is indicated. Weiss and coworkers [12] used a computer model and
decision analysis to determine the optimal strategy for the treatment of the N0 neck as a function
of the probability of occult cervical metastasis.
The data analyzed included studies with large
Table 1
Studies on occult metastases in T1 tongue squamous cell
carcinomas: occult neck metastases (%)
Author

Date

Site

T1

Spiro and Strong

Lee and Litton
Whitehurst and Droulias
Johnson et al
Ho et al
Yuen et al

19711974
1972
1977
1980
1992
1999

Tongue
Tongue
Tongue
Tongue
Tongue
Tongue

29.4
24
24
36
46
21

numbers of patients and contained a minimum

2-year follow-up, with results analyzed in terms
of outcome as a function of stage of neck disease.
A sensitivity analysis was performed to determine
the optimal threshold for treatment of the neck.
The authors concluded that for patients with
head and neck SCC and stage N0 neck status,
treatment of the neck is indicated if the probability of occult cervical metastasis is greater than
20%. Based on this prediction model [12] and
the available studies (see Table 1), surgical treatment of the neck is often indicated for patients
with T1N0 oral SCC lesions. A possible exception
would be in lesions demonstrating only supercial
invasion (ie, less than 2 mm of invasion). If this
decision is based on an incisional biopsy, however,
sampling error must be considered.
Tumor thickness
Tumor thickness has been proposed to provide
more predictive information regarding metastasis
than tumor size. Similar to Breslows work [13]
showing that malignant melanoma thickness is
a primary predictive variable, work in the area
of oral SCC has also demonstrated the importance of tumor thickness in predicting cervical metastasis. In oral SCC, the critical thickness has not
been determined and studies have varied considerably regarding the thickness that suggests the
cervical metastasis rate is high. In 1986,
Mohit-Tabatabai and colleagues [14] and Spiro
and coworkers [15] looked at the predictive value
of tumor thickness with SSC of the oor of the
mouth. Mohit-Tabatabai and colleagues [14] retrospectively reviewed 84 cases of patients with
early oor of the mouth SCC and found that
when patients had a tumor thickness less than
1.5 mm, the incidence of cervical metastasis was
1.8%; however, a tumor thickness greater than
1.5 mm carried with it a cervical metastasis rate
of 48%. Spiro and coworkers [15] retrospectively
evaluated 105 patients with tongue or oor of
the mouth SCC and found that a tumor thickness
of 2 mm or less carried a metastasis rate of 7.5%
and a tumor thickness greater than 2 mm carried
a metastasis rate of 38%. Byers and colleagues [9]
evaluated 91 patients with tongue SCC and attempted to correlate lymph node metastasis with
multiple preoperative and intraoperative factors,
including thickness of the specimen; depth of
muscle invasion; frozen margin status; perineural,
vascular, or lymphatic invasion; histologic
dierentiation; and DNA ploidy. In this study,

MOLECULAR BIOLOGY AND CLINICAL BEHAVIOR

the best predictors for nodal metastasis were

greater than 4 mm of muscle invasion, double
DNA ploidy, and histologic dierentiation. The
DNA ploidy results agree with those of other
studies demonstrating a weak correlation between
DNA ploidy status and cervical metastasis in patients with SCC of the tongue [16]. Currently,
DNA ploidy must be used with other pathologic
tumor characteristics to predict cervical
metastasis.
Invasive pattern
These ndings on tumor thickness suggest that
oral SCCs are composed of a subset of oral SCCs
that have increased invasive potential or that as
an oral SCC grows in thickness, it approaches and
invades blood and lymphatic vessels. The dierent
patterns of tumor invasion have been shown to
have varying metastasis rates [17,18]. SCCs that
inltrate along a broad front are much less likely
to metastasize than SCCs that are composed of
separate packets of invading cells. Woolgar and
Triantafyllou [5] have concluded that the pattern
of invasion aects the interpretation of margins
and that even a clear (ie, 5 mm) margin might
be inadequate in a case with widely separated clusters of invading tumor cells. Despite these studies,
an evaluation of invasive pattern is generally not
used in clinical practice to guide treatment.
Perineural invasion
Some oral SCCs are neurotrophic, and expression of nerve cell adhesion molecule (NCAM) on
the SCC is associated with perineural invasion
[19,20]. Although most studies have demonstrated
that perineural invasion is a histopathologic feature that correlates with increased local recurrence
and cervical metastasis [2124], perineural invasion has also been shown not to correlate with metastasis. The use of perineural invasion as an
independent factor to predict prognosis and recommend treatment is unclear at this time.

Molecular alterations in oral squamous

cell carcinoma
A genetic model similar to the genetic model
for colon cancer provided by Fearon and Vogelstein [25] can likely be applied to oral SCC. Oral
cancer is a rare cancer in which biopsies can be
obtained at multiple dierent times as the lesion
progresses through premalignant phases to cancer. Therefore, a genetic model of progression,

485

similar to the work of Califano and colleagues

[26], might allow for the creation of a genetic progression model for oral cancer. The available genetic models suggest that cells are altered by
inactivation of tumor suppressor gene or activation of oncogenes [25,27]. These genetic changes
produce a cellular growth advantage to the altered
cells, which expand to produce a malignancy. The
timing and accumulation of genetic events leading
to carcinoma are likely critical. Because genetic
changes precede phenotypic changes, molecular
and genetic analysis might aord the practitioner
the possibility to intervene.
Molecular analysis of surgical margins
Because genetic alterations precede phenotypic
changes, histologic assessment is likely inadequate
to analyze the adequacy of resection and to
predict recurrence [28]. Intraoperative histologic
analysis of surgical margins after resection cannot
detect epithelial cells that appear phenotypically
normal but have undergone transformation at
the molecular level. Therefore, occult tumor cells
present in normal-appearing tissue can lead to local recurrence [29].
There is clear molecular evidence of genetic
alterations in histologically normal margins surrounding oral SCC. Tumor suppressor genes,
such as TP53, provide a negative growth signal
through cell cycle arrest or apoptosis. A tumor
suppressor gene can be inactivated through a mutation or deletion or by being bound to a viral or
cellular protein. Such inactivation can lead to uncontrolled SCC growth. The p53 mutation is the
most common genetic event in human cancers
[30]. The incidence of p53 mutations has varied
signicantly between studies. Fifteen percent to
60% of oral SCCs have been shown to contain
p53 mutations [31,32]. Brennan and coworkers
[7] have demonstrated that histologically normal
surgical margins that contain p53 mutations are
associated with much higher local recurrence
rates. Using an assay based on the polymerase
chain reaction (PCR), which has the capacity to
detect 1 mutant cancer cell among 10,000 normal
cells, the authors analyzed the surgical specimens
to determine whether microscopically occult neoplastic cells could be identied in surgical margins
and lymph nodes obtained after resection for 30
head and neck cancer cases. The presence of
a p53 mutation in the margin was associated
with a signicantly increased incidence of local recurrence [7]. Local recurrence occurred in 5 (38%)

486

SCHMIDT

of 13 patients with positive margins assessed by

this method compared with no recurrence in patients with negative margins. Molecular analysis
identied neoplastic cells in 6 (21%) of 28 lymph
nodes that initially tested negative by traditional
histopathologic analysis. This study highlights
the potential utility of molecular margin analysis.
After resection, margins could be obtained and
analyzed during surgery to conrm genetically
clear margins. Currently, there is not a rapid
and accurate comprehensive technique available
to identify and map genetic aberrations in samples
to guide surgical management of oral SCC.
TP16 is a tumor suppressor gene located on
chromosomal region 9p21. This gene, which is
also known as CDKN2A, encodes a protein,
p16, that binds to cyclin-dependent kinases 4
and 6 and prevents their association with cyclin
D1. The cyclin-dependent kinase-cyclin D1 complex inhibition prevents retinoblastoma phosphorylation, leading to the inhibition of the cell cycle
G1/S transition. Deletions, point mutations, or
promoter hypermethylation inactivates the p16
gene and allows potentially mutagenic DNA damage to escape repair before cell division [33]. The
CDKN2A gene locus encodes two dierent proteins derived from alternative splicing: p16 acts
as a G1 cell cycle regulator, and p14ARF regulates cell growth through independent eects on
the p53 pathway. Therefore, p16 and p14ARF
both function as negative regulators of cell cycle
progression. Methylation of CDKN2A is a
key mechanism leading to inactivation of the
gene [33]. O6methylguanine-DNA-methyltransferase
(MGMT) is a DNA repair gene that is critical for
the rapid reversal of methylation of guanine [34].
Inactivation is rarely attributable to deletion or
mutation, but methylation of a discrete region of
the MGMT promoter has been associated with
the silencing of this gene [35]. Goldenberg and coworkers [36] explored the intraoperative evaluation of promoter methylation as a potential
method to analyze the molecular status of surgical
margins. The study was designed to evaluate the
possibility of performing the assay in a real-time intraoperative manner. The authors used a novel
quantitative methylation-specic PCR protocol
that required approximately 5 hours to evaluate
for methylation of the gene. Thirteen patients
with head and neck SCC were initially characterized for molecular alterations in their tumor at
the time of biopsy. Six primary tumors were found
to harbor promoter hypermethylation for p16 and
MGMT genes. Rapid quantitative methylation-

specic PCR was then used at the time of the operation to identify promoter hypermethylation of
these genes in the surgical margins. The authors
found that 3 patients had methylation-positive
margins. Tumor margins from 2 patients were
methylated for p16 alone, and margins from 1 patient were methylated for p16 and MGMT simultaneously. In this preliminary study, the authors
demonstrated the feasibility of the promoter hypermethylation assay for intraoperative use. Although no clinical follow-up data were provided,
future studies could examine the clinical implications of surgical margin hypermethylation. Ultimately, this method of molecular margin analysis
could facilitate more precise and complete surgical
resection of oral SCC and improve the locoregional control rate.
Field cancerization
Clearly, if a patient with oral cancer has a eld
eect, continued resection to obtain histologically
and genetically clear margins might not be indicated or possible. After histopathologic evaluation of specimens from 783 patients with head
and neck cancer, Slaughter and colleagues [37]
demonstrated that the epithelium surrounding
head and neck cancer was abnormal, and 11%
of patients were found to have an independent
area of malignancy. The authors concluded that
the epithelial changes were likely extensive and
likely contributed to the development of a malignancy. The hypothesis proposed by Slaughter and
colleagues [37] now has molecular support [38].
Molecular evidence exists for multiple foci of
genetically altered mucosa as well as for lateral
spread of clones. As the resolution of molecular
techniques improves, the clinician might be able
to determine whether patients with extensive or
multiple sites of genetically altered mucosa can
be managed surgically. Also, as chemopreventive
approaches are improved, patients with extensive
elds would be appropriate candidates for this
treatment strategy.

Genomics
Applications of microarrays
The results of the Human Genome Project in
combination with microarray technology have
provided investigators with a high-throughput
method to study the roles played by specic genes
in the development and progression of cancer.

MOLECULAR BIOLOGY AND CLINICAL BEHAVIOR

The strength of microarray technology is the

ability to evaluate the expression of tens of
thousands of genes within a specimen. In the
future, genomic technology is likely to be used by
clinicians to understand and predict the clinical
behavior of oral cancer. Gene expression has been
used to predict metastasis and outcome in patients
with breast cancer [39]. A large number of novel
genes within previously unknown sequences have
already been identied in oral SCC specimens using microarray technology. The current challenge
is to nd a way to organize and catalog this vast
amount of information into a form that can be analyzed to understand better the roles of particular
genes and gene families in cancer. DNA microarray technology facilitates the identication and
classication of DNA genome information and
is allowing for an understanding of the function
of newly identied genes. As more information accumulates, we should be able to use microarrays
to ask increasingly complex questions, such as
why, despite similar histologic ndings, oral
SCC behaves clinically as a heterogeneous group.
The function of new genes should be understood
based on similarities in expression patterns with
those of known genes. Given that the product of
any one gene is likely to interact with the products
of other genes, coordinated gene expression and
an understanding of gene interrelations should
be gained.
DNA microarrays consist of solid supports (ie,
glass microscope slides or silicone chips) onto
which thousands of gene sequences are xed at
specic locations. The DNA can be printed or
spotted onto the support. Microarray technology
depends on the ability of an mRNA molecule to
hybridize to one of the DNA templates on the
support. Hybridization occurs when two complementary sequences nd each other, such as the
target DNA xed to the slide or chip and
the mobile probe DNA, cDNA, or mRNA. The
mobile probe is uorescently labeled. After hybridization, the slide or chip is scanned with a laser
that excites the uorescent tag on the mobile
probe. The chip is read with a specialized microscope and camera. A red-to-green uorescence
ratio is determined signifying the presence of
expressed genes (ie, the amount of uorescently
labeled cDNA or mRNA bound to the spots on
the microarray). By using a microarray containing
many DNA samples, the expression levels of
thousands of genes within a cell can be determined, generating a prole of genes expressed in
the cell.

487

Microarray technology can be used in a number of ways, including measuring changes in gene
expression levels, identifying genomic gains and
losses, and determining mutations in DNA. In the
rst case, determining the level at which a certain
gene is expressed is called gene expression analysis
[40]. The immobilized DNA on the chip is cDNA
derived from the mRNA of known genes. The
control DNA and sample DNA consist of
cDNA derived from the mRNA of normal and
diseased tissue, respectively. The control DNA
and sample DNA are dierentially labeled. If
a particular gene is overexpressed in disease,
more sample cDNA than control cDNA hybridizes to the spot, representing that expressed
gene. In the case of an expression array, this particular spot representing the gene of interest uoresces red with greater intensity than it uoresces
green. As we gain knowledge of the gene expression patterns of particular disease states, mRNA
can be extracted from diseased tissue from any individual and tested to determine whether the gene
expression pattern in that tissue matches the expression pattern of the disease state. In addition
to diagnosis, it is hoped that gene expression chips
can be used to develop new therapeutic approaches. For example, if a certain gene is overexpressed in oral SCC, an expression chip could be
used to test whether a particular drug reduces expression of the gene or gene family of interest and
subsequently improves cancer control.
Microarray technology has been used to investigate the expression of up to 40,000 genes in
oral cancer specimens and to compare those
expression levels with normal tissue. These studies
have conrmed the heterogeneity of oral SCC
compared with other normal tissues. These ndings point to the fact that oral SCC needs to be
studied separately from other head and neck sites.
Mendez and colleagues [41] demonstrated the
heterogeneity of oral SCC by evaluating the expression proles of 26 oral SCCs relative to 18
normal oral tissues using oligonucleotide arrays
evaluating 7000 genes. This study demonstrated
that many of the upregulated genes present in
oral SCCs are also present in the premalignant
tissues. Supervised regression analysis demonstrated 239 genes that were upregulated and 75
genes that were downregulated. Many genes involved in invasion and metastasis were upregulated. There was no dierence in expression of
these genes between early- and late-stage tumors,
suggesting that the genetic changes leading to invasion and metastasis are present early on and

488

SCHMIDT

could be identied and treated appropriately.

Microarray technology holds great promise for
being able to predict the clinical behavior of
oral SCC. Warner and coworkers [42] analyzed
the gene expression proles of 20 oral SCCs using
cDNA microarrays containing 19,200 sequences.
The authors identied two sample clusters that
correlated with higher T category and cervical
node metastasis. This report demonstrates that
despite the clinical heterogeneity of oral SCC,
molecular subtyping by cDNA microarray is
able to identify distinct patterns of gene expression that can be correlated with relevant clinical
parameters. Such an approach represents an advance in the genetic classication of oral SCCs
and might provide the clinician with useful information to tailor therapy.
Genetic instability seems to be a consequence
of transformation. Multiple chromosomal sites
are altered in oral SCC, including 3p, 4q, 5q21 to
5q22, 8p21 to 8p23, 9p21 to 9p22, 11q13, 11q23,
13q, 14q, 17p, 18q, and 22q [4354]. Microarray
technology can also be used to identify gains
and losses of chromosomal regions across the entire genome. This technique is called array comparative genomic hybridization (CGH) and can
be used to look for genomic gains and losses involved in a disease state. To date, array CGH
technology has been used primarily to study cancer [55]. DNA repair genes play a major role in
cancer, and mutations of the DNA repair genes
can produce broken chromosomes, leading to certain chromosomal gains and losses. Changes in
chromosomes are related to cancer progression,
and the patterns of these changes are relevant to
clinical prognosis. Array CGH allows for accurate
quantication of DNA copy number variations
over the entire genome and provides reliable detection of single copy changes from diploid [55].
Array CGH shows great promise for classifying
disease states, assessing risk, and predicting
prognosis. In solid tumors, the chromosomal
aberrations are often extensive, which makes identication of candidate genes involved in carcinogenesis dicult. We performed array CGH on
89 oral SCCs and focused on narrow regions of
gene amplication to facilitate identication of genetic pathways important in oral SCC development [55]. We found genes involved in integrin
signaling, survival, adhesion, and migration as
well as members of the hedgehog and notch
pathways to be amplied and overexpressed.
This study points to the potential utility of
novel, high-throughput, automated technologies

to identify genes and gene products that could

provide therapeutic, prognostic, or diagnostic
targets.

Proteomics
The proteome is dened as all the proteins
encoded by the genome. Proteomics is the study
and mapping of the human proteome. Proteomics
represents a link between the information from
the Human Genome Project and the use of that
information to understand and treat cancer.
Proteomics encompasses identication and quantication of proteins as well as modications,
interactions, and activities of those proteins.
Cancer is a consequence of genetic and epigenetic
alterations that lead to protein dysregulation
aecting cell division, dierentiation, immune
recognition, tissue invasion, and metastasis.
Once genes are expressed, the resulting mRNA
contains introns and exons. The introns are
removed, and the exons are spliced together,
producing an uninterrupted series of coding
sequences for protein translation. Because the
initial transcript can be spliced in dierent ways,
a number of dierent proteins can be produced by
a single gene. Proteins can also be altered after
translation. Posttranslational modications include alternative protein splicing, chemical modication of certain amino acids, and the
attachment of small molecules (eg, carbohydrates). Proteomic approaches can be used successfully to identify disease markers. Recent
developments in proteomics suggest that the
technologies available are already suciently
advanced to approach many biologic questions
relevant to the management of oral cancer.
Current proteomic technologies include twodimensional gel electrophoresis for proling complex mixtures, followed by mass spectrometry for
protein identication; mass spectrometry comparative analysis of isotopically labeled reagents from
diseased and normal individuals or tissues and
immunoprecipitation coupled to mass spectrometry for studying protein-protein interactions; and
computational methods for predicting protein
structure-function relations.
One potentially fruitful area of proteomics is
early diagnosis by proteomic patterns. Proteomic
analysis has been applied to biouids for the
possible diagnosis of cancer. This approach has
been used to diagnose cancer using serum. Using
a complex combination of mass spectrometry and

MOLECULAR BIOLOGY AND CLINICAL BEHAVIOR

bioinformatics, serum analysis was able to diagnose ovarian cancer with a sensitivity of 100%,
specicity of 95%, and positive predictive value of
94% [56]. Using a similar approach, prostate cancer could be diagnosed after proteomic serum
analysis with a sensitivity of 83% and specicity
of 97% [57]. Clearly, for oral cancer, the identication of a salivary protein or proteomic pattern
would signicantly aid in the screening of patients
at risk for oral cancer. We have demonstrated that
certain salivary peptides are increased in patients
with oral cancer relative to normal controls [58].
Salivary analysis is theoretically ideal for the diagnosis of oral or systemic disease, given that saliva
is readily available. The identication and measurement of a salivary biomarker would be particularly benecial for screening patients at risk for
oral SCC. Screening patients with a history of
oral cancer, those at risk because of alcohol and
tobacco use, or those who have a potentially malignant epithelial lesion could be signicantly improved with noninvasive salivary analysis.

References
[1] Looser KG, Shah JP, Strong EW. The signicance
of positive margins in surgically resected epidermoid carcinomas. Head Neck Surg 1978;1(2):
10711.
[2] Loree TR, Strong EW. Signicance of positive margins in oral cavity squamous carcinoma. Am J Surg
1990;160(4):4104.
[3] Spiro RH, Guillamondegui O Jr, Paulino AF, et al.
Pattern of invasion and margin assessment in patients with oral tongue cancer. Head Neck 1999;
21(5):40813.
[4] van Es RJ, van Nieuw Amerongen N, Slootweg PJ,
et al. Resection margin as a predictor of recurrence
at the primary site for T1 and T2 oral cancers. Evaluation of histopathologic variables. Arch Otolaryngol Head Neck Surg 1996;122(5):5215.
[5] Woolgar JA, Triantafyllou A. A histopathological
appraisal of surgical margins in oral and oropharyngeal cancer resection specimens. Oral Oncol 2005;
41(10):103443.
[6] Johnson RE, Sigman JD, Funk GF, et al. Quantication of surgical margin shrinkage in the oral cavity. Head Neck 1997;19(4):2816.
[7] Brennan JA, Mao L, Hruban RH, et al. Molecular
assessment of histopathological staging in squamous-cell carcinoma of the head and neck. N Engl
J Med 1995;332(7):42935.
[8] Broders AC. Squamous cell epithelioma of the lip;
a study of ve hundred and thirty seven cases.
JAMA 1920;74:65664.

489

[9] Byers RM, El-Naggar AK, Lee YY, et al. Can we detect or predict the presence of occult nodal metastases in patients with squamous carcinoma of the oral
tongue? Head Neck 1998;20(2):13844.
[10] Po Wing Yuen A, Lam KY, Lam LK, et al. Prognostic factors of clinically stage I and II oral tongue carcinomada comparative study of stage, thickness,
shape, growth pattern, invasive front malignancy
grading, Martinez-Gimeno score, and pathologic
features. Head Neck 2002;24(6):51320.
[11] Greene FL. American Joint Committee on Cancer,
American Cancer SocietyAJCC cancer staging manual. 6th edition. New York: Springer-Verlag; 2002.
[12] Weiss MH, Harrison LB, Isaacs RS. Use of decision
analysis in planning a management strategy for the
stage N0 neck. Arch Otolaryngol Head Neck Surg
1994;120(7):699702.
[13] Breslow A. Thickness, cross-sectional areas and
depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 1970;172(5):9028.
[14] Mohit-Tabatabai MA, Sobel HJ, Rush BF, et al. Relation of thickness of oor of mouth stage I and II
cancers to regional metastasis. Am J Surg 1986;
152(4):3513.
[15] Spiro RH, Huvos AG, Wong GY, et al. Predictive
value of tumor thickness in squamous carcinoma
conned to the tongue and oor of the mouth. Am
J Surg 1986;152(4):34550.
[16] Monasebian DM, Ruskin JD. Flow cytometric analysis of squamous cell carcinoma of the tongue. J Oral
Maxillofac Surg 1994;52(6):5748.
[17] Crissman JD, Liu WY, Gluckman JL, et al. Prognostic value of histopathologic parameters in squamous cell carcinoma of the oropharynx. Cancer
1984;54(12):29953001.
[18] Yamamoto E, Miyakawa A, Kohama G. Mode of
invasion and lymph node metastasis in squamous
cell carcinoma of the oral cavity. Head Neck Surg
1984;6(5):93847.
[19] McLaughlin RB Jr, Montone KT, Wall SJ, et al.
Nerve cell adhesion molecule expression in squamous cell carcinoma of the head and neck: a predictor of propensity toward perineural spread.
Laryngoscope 1999;109(5):8216.
[20] Vural E, Hutcheson J, Korourian S, et al. Correlation of neural cell adhesion molecules with perineural spread of squamous cell carcinoma of the head
and neck. Otolaryngol Head Neck Surg 2000;
122(5):71720.
[21] Cottel WI. Perineural invasion by squamous-cell
carcinoma. J Dermatol Surg Oncol 1982;8(7):
589600.
[22] Fagan JJ, Collins B, Barnes L, et al. Perineural invasion in squamous cell carcinoma of the head and
neck. Arch Otolaryngol Head Neck Surg 1998;
124(6):63740.
[23] Goepfert H, Dichtel WJ, Medina JE, et al. Perineural invasion in squamous cell skin carcinoma of the
head and neck. Am J Surg 1984;148(4):5427.

490

SCHMIDT

[24] Soo KC, Carter RL, OBrien CJ, et al. Prognostic

implications of perineural spread in squamous carcinomas of the head and neck. Laryngoscope 1986;
96(10):11458.
[25] Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990;61(5):75967.
[26] Califano J, van der Riet P, Westra W, et al. Genetic
progression model for head and neck cancer: implications for eld cancerization. Cancer Res 1996;
56(11):248892.
[27] Kim MM, Califano JA. Molecular pathology of
head-and-neck cancer. Int J Cancer 2004;112(4):
54553.
[28] Nathan CA, Amirghahri N, Rice C, et al. Molecular
analysis of surgical margins in head and neck squamous cell carcinoma patients. Laryngoscope 2002;
112(12):212940.
[29] Gath HJ, Brakenho RH. Minimal residual disease
in head and neck cancer. Cancer Metastasis Rev
1999;18(1):10926.
[30] Harris CC, Hollstein M. Clinical implications of the
p53 tumor-suppressor gene. N Engl J Med 1993;
329(18):131827.
[31] Brennan JA, Boyle JO, Koch WM, et al. Association
between cigarette smoking and mutation of the p53
gene in squamous-cell carcinoma of the head and
neck. N Engl J Med 1995;332(11):7127.
[32] Ralhan R, Nath N, Agarwal S, et al. Circulating p53
antibodies as early markers of oral cancer: correlation with p53 alterations. Clin Cancer Res 1998;
4(9):214752.
[33] Merlo A, Herman JG, Mao L, et al. 50 CpG island
methylation is associated with transcriptional silencing of the tumour suppressor p16/CDKN2/MTS1 in
human cancers. Nat Med 1995;1(7):68692.
[34] Esteller M, Hamilton SR, Burger PC, et al.
Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human
neoplasia. Cancer Res 1999;59(4):7937.
[35] Costello JF, Futscher BW, Tano K, et al. Graded
methylation in the promoter and body of the O6
methylguanine DNA methyltransferase (MGMT)
gene correlates with MGMT expression in human
glioma cells. J Biol Chem 1994;269(25):1722837.
[36] Goldenberg D, Harden S, Masayesva BG, et al.
Intraoperative molecular margin analysis in head
and neck cancer. Arch Otolaryngol Head Neck
Surg 2004;130(1):3944.
[37] Slaughter DP, Southwick HW, Smejkal W. Field
cancerization in oral stratied squamous epithelium;
clinical implications of multicentric origin. Cancer
1953;6(5):9638.
[38] Ha PK, Califano JA. The molecular biology of mucosal eld cancerization of the head and neck. Crit
Rev Oral Biol Med 2003;14(5):3639.
[39] vant Veer LJ, Dai H, van de Vijver MJ, et al. Gene
expression proling predicts clinical outcome of
breast cancer. Nature 2002;415(6871):5306.

[40] DeRisi J, Penland L, Brown PO, et al. Use of

a cDNA microarray to analyse gene expression patterns in human cancer. Nat Genet 1996;14(4):
45760.
[41] Mendez E, Cheng C, Farwell DG, et al. Transcriptional expression proles of oral squamous cell carcinomas. Cancer 2002;95(7):148294.
[42] Warner GC, Reis PP, Jurisica I, et al. Molecular
classication of oral cancer by cDNA microarrays
identies overexpressed genes correlated with nodal
metastasis. Int J Cancer 2004;110(6):85768.
[43] Adamson R, Jones AS, Field JK. Loss of heterozygosity studies on chromosome 17 in head and neck
cancer using microsatellite markers. Oncogene
1994;9(7):207782.
[44] El-Naggar AK, Coombes MM, Batsakis JG, et al.
Localization of chromosome 8p regions involved in
early tumorigenesis of oral and laryngeal squamous
carcinoma. Oncogene 1998;16(23):29837.
[45] el-Naggar AK, Hurr K, Batsakis JG, et al. Sequential loss of heterozygosity at microsatellite motifs
in preinvasive and invasive head and neck squamous
carcinoma. Cancer Res 1995;55(12):26569.
[46] Lazar AD, Winter MR, Nogueira CP, et al. Loss of
heterozygosity at 11q23 in squamous cell carcinoma
of the head and neck is associated with recurrent disease. Clin Cancer Res 1998;4(11):278793.
[47] Lee DJ, Koch WM, Yoo G, et al. Impact of chromosome 14q loss on survival in primary head and neck
squamous cell carcinoma. Clin Cancer Res 1997;
3(4):5015.
[48] Mao EJ, Schwartz SM, Daling JR, et al. Loss of heterozygosity at 5q21-22 (adenomatous polyposis coli
gene region) in oral squamous cell carcinoma is common and correlated with advanced disease. J Oral
Pathol Med 1998;27(7):297302.
[49] Mao L, Lee JS, Fan YH, et al. Frequent microsatellite alterations at chromosomes 9p21 and 3p14 in
oral premalignant lesions and their value in cancer
risk assessment. Nat Med 1996;2(6):6825.
[50] Nawroz H, van der Riet P, Hruban RH, et al. Allelotype of head and neck squamous cell carcinoma.
Cancer Res 1994;54(5):11525.
[51] Ogawara K, Miyakawa A, Shiba M, et al. Allelic loss
of chromosome 13q14.3 in human oral cancer: correlation with lymph node metastasis. Int J Cancer
1998;79(4):3127.
[52] Ransom DT, Barnett TC, Bot J, et al. Loss of heterozygosity on chromosome 2q: possibly a poor prognostic factor in head and neck cancer. Head Neck
1998;20(5):40410.
[53] Reis PP, Rogatto SR, Kowalski LP, et al. Quantitative real-time PCR identies a critical region of deletion on 22q13 related to prognosis in oral cancer.
Oncogene 2002;21(42):64807.
[54] Waber P, Dlugosz S, Cheng QC, et al. Genetic alterations of chromosome band 9p21-22 in head and
neck cancer are not restricted to p16INK4a. Oncogene 1997;15(14):1699704.

MOLECULAR BIOLOGY AND CLINICAL BEHAVIOR

[55] Snijders AM, Schmidt BL, Fridlyand J, et al.

Rare amplicons implicate frequent deregulation
of cell fate specication pathways in oral squamous cell carcinoma. Oncogene 2005;24(26):
423242.
[56] Petricoin EF, Ardekani AM, Hitt BA, et al. Use of
proteomic patterns in serum to identify ovarian cancer. Lancet 2002;359(9306):5727.

491

[57] Adam BL, Qu Y, Davis JW, et al. Serum protein ngerprinting coupled with a pattern-matching algorithm distinguishes prostate cancer from benign
prostate hyperplasia and healthy men. Cancer Res
2002;62(13):360914.
[58] Pickering V, Jordan RCK, Schmidt BL. Elevated salivary endothelin levels in oral cancer patientsda pilot
study. Oral Oncol 2006 Jun 4; [Epub ahead of print].

Oral Maxillofacial Surg Clin N Am 18 (2006) 493511

Oral Cancer Prevention

M. Abraham Kuriakose, MD, BDS, FDSRCS,
FFDRCS, FRCS Ed, FRCSa,b,*, Rajeev Sharan, MS, MDa
a

Head and Neck Institute, AIMS Hospital, Amrita Lane, Elamakkara, Kochi 682 026, India
Department of Oral and Maxillofacial Surgery, New York University College of Dentistry,
New York, New York, USA

Oral cancer is largely a preventable disease.

Improved understanding of risk factors, molecular
basis of predilection for oral cancer and carcinogenesis process, and advances in chemoprevention
have made oral cancer prevention a clinical reality.
Oral cancer prevention is not only feasible but also
most desirable. Despite advances in modern multimodality treatment, the survival rate of oral cancer
has only modestly improved to approximately 60%
in the recent past. The treatment of oral cancer can
produce signicant functional impairment in
speech, mastication and swallowing, dental health,
and even the ability to interact socially. It must be
considered among the most debilitating and disguring of all cancers [1]. The morbidity of oral
cancer treatment is proportional to the stage of
disease at diagnosis. Measures to downstage the
disease or prevent it will have a signicant benet
to society. This article reviews the current status
of early detection and chemoprevention measures
in oral cancer.
Oral carcinogenesis
Oral cancer prevention essentially involves detouring of the carcinogenesis process. Understanding the mechanism of oral carcinogenesis is
essential for developing various strategies to reverse this process. Oral cancer in general develops
as multistep, multifocal disease as a result of
various carcinogenic insults on oral mucosa [2,3].
The lesions often start as a premalignant condition
that may be a clinically obvious lesion, such as

* Corresponding author.
E-mail address: mak12@nyu.edu (M.A. Kuriakose).

leukoplakia, or a subclinical condition with morphologically normal mucosa. Over the years it accumulates genetic and epigenetic alterations and
progresses through well-dened pathologic stages
to invasive squamous cell carcinoma. Oral carcinogenesis is both a multistep and multifocal process.
Multistep process
The genetic mechanism of oral carcinogenesis
is well understood [4]. This includes mutation of
tumor-suppressor genes and activation of various
oncogenes at various stages of carcinogenesis
(Fig. 1). Loss of heterozygosity (LOH) at chromosomal region 3p14, 9p21 and 17p leading to p16
promoter hypermethylation and p53 inactivation;
and gene amplication at 11q13 with overexpression cyclin D1 and telomerase activation are involved in various stages of oral carcinogenesis
[2,3]. Tobacco may cause oral cancer, partly
through eects on p53 and mutation at the chromosomal region 3p [5,6]. Altered p53 expression
is associated with increased genomic instability
(eg, aneuploidy) in oral premalignant lesion,
which may drive the acceleration in the rate of genetic alterations during oral tumorigenesis [7].
Overexpression of cyclo-oxygenase-2 (COX-2),
phospho-epidermal growth factor receptor
(EGFR), RAR, and PCNA are important events
in oral carcinogenesis. This forms the basis for
targeted prevention strategies in oral cancer.
Multifocal process
Prolonged carcinogenic exposure is required
for the transformation of a normal cell to a malignant cell. Because oral epithelium is continuously being renewed, the only possible cells that

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.003

oralmaxsurgery.theclinics.com

494

KURIAKOSE & SHARAN

Histologic
Steps

Chromosomal
Alteration

Gene expression
Changes

Normal

Hyperplasia

9p

Mild
dysplasia

3p
17p

Moderate
dysplasia

Severe
dysplasia

11q
13q
14q

Carcinoma
in situ

Invasive
carcinoma

Distant
Metastasis

6p
8p
4q

RAR, EGFR, p16, p53, PCNA/Ki67, COX-2, NF-B, H-RAS, TGF-

Fig. 1. Key histologic, chromosomal, and gene expression alterations in multistep carcinogenesis.

can be subjected to this long-term insult should be

those at the basal layers, in particular the stem cell
population. It is hypothesized that the carcinogenic process starts at these epithelial stem cells.
These transformed stem cells parent clonal units
of malignant cells with similar genetic alterations
to form a patch. These transformed cells acquire
further genetic alterations and gain growth
advantage over the adjacent normal mucosal
cells. These cells spread by lateral migration or
inoculation through saliva to a distant site or by
forming a dierent clone to form a eld (Fig. 2).

This eld cancerization, often observed in the

oral cavity, may be responsible for the development of second primary tumors (SPTs). This concept of eld cancerization introduced by Slaughter
and colleagues [8], which was based on the exposure of wide elds of epithelial surface within the
aerodigestive tract to carcinogens such as tobacco
and alcohol, may be responsible for the development of multiple oral cancers that arise from similar or separate clones. Genetic analyses indicate
that subsequent cancers develop in a eld from
the spread of the original clone [9]. This entails

Fig. 2. Multifocal oral carcinogenesis and mechanism of expansion of carcinogenic eld. (From Braakhuis BJ, Tabor
MP, Kummer JA, et al. A genetic explanation of Slaughters concept of eld cancerization: evidence and clinical implications. Cancer Res 2003;63:172730; with permission.)

495

ORAL CANCER PREVENTION

the clinical implication that transformed elds

often remain after treatment of the primary premalignant or malignant lesion. This may lead to
the development new cancers, currently designated
as SPTs or local recurrences [2,10]. Preventive
strategy should attempt to detour this carcinogenic
process to correct these genetic alterations.

evolution to cancer. Tertiary prevention, which

intends to lower the morbidity of treatment, can
be undertaken through down-staging the tumor
by organized oral cancer screening program of
a population.

Risk factor reduction

Risk factors for oral cancer predilection
Epidemiologic and molecular biology research
has helped to dene the high-risk group with
a predilection to develop oral cancer (Fig. 3). Epidemiologic studies have identied certain populations with high risk of developing oral cancer,
including people from the Indian subcontinent,
Polynesia, and Mexico, and Black Americans.
Within the high-risk population, individuals with
susceptibility for oral cancer can be identied
through their risk habits of tobacco use and alcohol abuse. Single nucleotide polymorphisms
(SNPs) of genes that code for carcinogen metabolism enzymes or DNA repair genes may explain
individual dierences in the susceptibility to carcinogens. The carcinogen-metabolizing genes
include CYP1-A1, GSTM1, and GSTM1T1.
Within susceptible individuals, there are premalignant lesions with varying risk of malignant transformation. The malignant transformation rate
varies from 0.13% to 36.4% [1115]. The biomarkers, which predict malignant transformation,
are degree of dysplasia, loss of heterozygosity of
3p14 and 9p21 region and genomic instability.
Oral cancer prevention strategies
Various prevention strategies can be implemented at various stages of oral carcinogenesis to
reduce the burden of oral cancer. Primary prevention, in which the goal is to prevent the onset
of the disease, can be attempted by interrupting
high-risk lifestyles, particularly those that involve
the use of tobacco. Secondary prevention is aimed
at treating an individual with premalignant condition or an in situ neoplasia and blocking its

There is sucient scientic evidence to conclude that oral cancer is largely related to lifestyle.
The major etiologic factors include the use of
tobacco, alcohol abuse, and, in the case of lip
cancer, exposure to the sun. The risk for development of oral cancer is three- to ninefold
greater in individuals who smoke or drink and as
much as 100 times greater in individuals who
smoke and drink heavily versus people who
neither smoke nor drink [16]. Tobacco may be
consumed in a smoked or smokeless form. For
smoking, previous studies showed that 70% to
90% of patients with leukoplakia were tobacco
users and that 78% of leukoplakia lesions disappeared or regressed within 12 months after
smoking cessation [17]. In a case-control study,
Kulasegaram and colleagues [18] found a signicant dose-response relationship between leukoplakia and tobacco use. Similar ndings also
were observed in other studies [1922].
For the association between alcohol consumption and the development of leukoplakia, Kulasegaram and colleagues [18] found that alcohol
consumption was not related to leukoplakia.
The synergetic eect between alcohol and smoking, as found in oral cancer, was not observed
for leukoplakia [19,20,22]. Mashberg and colleagues [23] also indicated that there was no substantial dierence in the risk of leukoplakia
caused by dierences in the type of alcohol consumed. Although the malignant transformation
from leukoplakia to oral cancer has been reported
to range from 0.13% to 36.4% [1115], these rates
pertained to the malignant transformation after
medical regimen rather than the disease natural
history.

High Risk
Population

High Risk
Individual

High Risk
Lesions

Indian Sub-continent
Polynesia
South America
US Black

H/o HNSCC
Risk habits
Polymorphisms of
CYP1A1,GSTM1. GSTM1T1

LOH
of 3p14, 9p21
Dysplasia

Fig. 3. Factors that determine oral cancer predilection.

496

KURIAKOSE & SHARAN

Several studies have addressed the association

between betel quid chewing and leukoplakia
because this habit prevailed in areas such as India
and Taiwan. One study in Taiwan showed a signicant association between betel quid chewing
and leukoplakia [24]. In addition to smoking, the
use of smokeless tobacco, whether in the form of
snu (ground and nely cut tobacco) or chewing
tobacco (loose leaf tobacco), has been linked
strongly to oral cancer.
Martin and colleagues [25] evaluated the prevalence and risk of developing oral leukoplakia in
smokeless tobacco users and the response of the
leukoplakic lesions after six weeks of involuntary
tobacco cessation in US Air Force basic military
trainees. Of the 3051 male trainees examined
(mean age, 19.5 years), 9.9% (302/3051) were
identied as current smokeless tobacco users.
Among current smokeless tobacco users, 39.4%
(119/302) had leukoplakia versus 1.5% (42/2749)
of nonusers of smokeless tobacco (odds ratio,
41.9, 95% condence interval, 28.162.6). At the
end of the involuntary cessation of tobacco use,
97.5% of the leukoplakic lesions had complete
clinical resolution. The type of smokeless tobacco
used (snu versus chewing tobacco), amount used
(cans or pouches per day), length of use (months),
number of days since last use, and brand of snu
used were signicantly associated with the risk of
developing leukoplakic lesions among smokeless
tobacco users.
Schinke and colleagues [26] studied tobacco
use trends, perceptions, and prevention eects
for 1281 fth and sixth graders enrolled in 12 randomly selected Washington State elementary
schools. During the 2-year study, smoked and
smokeless tobacco use rates increased in all
groups, and youths in the skills intervention group
consistently showed the lowest rates relative to the
other groups. These ndings demonstrated the potential of skills intervention methods for lowering
tobacco use rates among adolescents.
In a meta-analysis of randomized clinical trials, Sankaranarayanan and colleagues [27] reconrmed the eectiveness of smoking cessation
counseling. School-based programs to prevent
smokeless tobacco use have had mixed results
[2830]. Shiu and Chen [31] studied the eects of
three risk factors (ie, chewing, smoking, and
drinking) on occurrence of oral leukoplakia and
malignant transformation to oral cancer in a hospital-based case-control study design. A total of
74 oral cancer patients, 164 patients with oral leukoplakia, and 187 controls were interviewed to

collect information on their betel chewing, smoking, and drinking habits. The eects of the three
risk factors on the progression rates of the threestage disease process were estimated. Subjects
who chewed betel quid were at greater risk of leukoplakia (adjusted odds ratio [OR] 17.7 [9.03
34.5]) but there was no signicant eect on malignant transformation (OR 1.04 [0.611.76]). Smoking played a major role in the onset of leukoplakia
(OR 4.26 [2.218.23]) but a minor role in malignant transformation (OR 1.36 [0.692.68]). Alcohol was positively associated with malignant
transformation (OR 2.37 [1.473.82]) but unrelated to occurrence of leukoplakia (OR 0.76
[0.041.43]). They have concluded that smoking
and betel quid were two signicant risk factors
for the occurrence of leukoplakia, whereas alcohol
was responsible for malignant transformation.
The elucidation of three risk factors associated
with the natural history of leukoplakia and oral
cancer provides a way of designing an appropriate
prevention program in relation to oral cancer, including primary prevention in elimination risk
factors or secondary prevention in reducing malignant transformation. The probabilities of developing leukoplakia and oral cancer for a person
who chews betel quid or smokes over a 10-year
period were 17.8% and 7.65%, respectively. The
corresponding gures were 0.31% and 0.075%
for a person without one of the three risk factors.
The probabilities of malignant transformation
over the 10-year period were 75.1% for a leukoplakia case in the presence of three risk factors and
36.1% in the absence of any risk factors.
Economic data suggest reduction in consumption of tobacco in several populations. Ageadjusted rate of oral cancer over the years has
shown an overall reduction in Indian and SEER
(United States) cancer registries. The reduction is
much steeper in the Indian population with high
prevalence of the disease (Fig. 4). Similar data
from Eastern Europe, where the tobacco consumption rate remains high, show a steady increase in incidence of oral cancer (Fig. 5).
Overall reduction in the incidence of oral cancer
may be attributed to low tobacco consumption.
Lowering tobacco consumption will have a more
pronounced eect in high prevalent regions.
Primary prevention, including giving advice on
smoking and tobacco cessation, alcohol moderation, and sun protection, should be an important
goal of every health program targeted against oral
cancer [3234]. Health care providers involved in
the care of oral disease should be well aware of

497

ORAL CANCER PREVENTION

50
45
40
35
30
25
20
15
10
5
0

Mumbai-Female
Mumbai Male
SEER-Female
SEER-Male

19731977

19781982

19831987

19881992

19931997

Fig. 4. Change in trends of age-adjusted rate of oral cancer in Mumbai (India) and SEER, USA cancer registries.

and involved in implementing the concepts of

avoidable cancer risk lifestyle, which will have
considerable benet over several decades in reducing the burden of oral cancer in the community.
Screening and early detection and treatment
The oral cavity is one of the few regions in the
body that lends itself to routine screening of
asymptomatic population for early cancer detection. Attempts to downstage the disease by early
detection programs may improve the functional
outcome and the cost-eectiveness of the treatment [35,36]. Patients who have oral cancer are at
risk of developing synchronous or metachronous
SPTs at a rate of approximately 7% a year to
an aggregate risk of 25%. Individuals with tobacco and alcohol habits also are at signicantly
high risk for developing oral cancer. The potential

strategic benets of prevention of oral cancer can

be achieved best by screening individuals at high
risk for oral cancer.
A signicant number of oral cancer developments are preceded by visible mucosal changes,
which oer opportunity for intervention and
eective treatment with low morbidity. However,
most cases of oral cancer are diagnosed at an
advanced stage, when the treatment is complex
and costly, with associated functional compromise
and disgurement. This situation may be related
to lack of awareness of oral cancer among the
public, dental, and medical communities. The lack
of awareness of oral cancer signs and symptoms
and risk factors are believed to greatly inuence
diagnostic delay [1]. It has also been reported that
general practitioners rarely conduct routine oral
examinations or enquire about their patients tobacco and alcohol use [1,37,38].

Fig. 5. Change in trends of age-adjusted rate of oral cancer in Europe. (From La Vecchi C, Lucchini F, Negri E. Trends
in oral cancer mortality in Europe. Oral Oncol 2004;40:4339; with permission.)

498

KURIAKOSE & SHARAN

Screening for oral cancer is a simple, noninvasive procedure that requires only a 5-minute
visual inspection of the oral mucosa with adequate lighting, gauze, and gloves, whereas the
detection of most other malignancies in their early
asymptomatic stages almost always requires special, costly, and often invasive techniques [39].
Down-staging oral cancer by systematic or opportunistic screening will have a major impact on reducing the disease burden and morbidity of the
disease in the community [40].
The oral cavity is an easily accessible site for
screening by dentists, physicians, nurses, and
health care workers or by self-examination. Visual
screening has been shown to detect early oral
neoplasia if provided as part of routine health care
by health care workers [4148]. The sensitivity of
oral visual inspection to detect lesions varied
from 57.7% to 61.4% in previous studies, and
the specicity ranged from 98.6 to 98.8% [41
48]. Early cases of oral cancer have a better
prognosis than cases with advanced disease. A
nationwide oral cancer screening program has
been ongoing in Cuba since 1984. A signicant reduction in the risk of advanced oral cancer was
seen in a case-control study of oral screening in
Cuba [47]. Moles and colleagues [49] reported
a meta-analysis of oral cancer screening that reiterated its eectiveness for early detection of these
tumors (Fig. 6).
Sankaranarayanan and colleagues [50] reported for the rst time in a randomized clinical
trial of more than 190,000 people in Kerala, India
that oral cancer screening by visual examination
by trained health care workers can reduce mortality from oral cancer in high-risk individuals. This
was a cluster randomized trial designed to have
80% power to detect a 35% reduction in oral
cancer mortality within 12 years of enrollment
between the intervention and control group
through screening every three years (Fig. 7). The

investigators reported that nine years after the

start of screening, there was a 32% reduction in
mortality in high-risk individuals in the intervention group (42% when only male tobacco/alcohol
users are considered). Overall, these data suggest
that oral visual screening in high-risk patients
could prevent approximately 40,000 deaths from
oral cancer worldwide.
Early recognition of potential malignant oral
disease with appropriate referral of patients can
signicantly inuence patient outcome [37,51].
Often the referral of patients with potential malignant disease by dentally or medically qualied primary health care providers is neither appropriate
nor timely [37,52,53].
Leao and colleagues [54] showed that only
34% of a group of 129 dentists in Brazil could correctly identify common clinical characteristics of
oral squamous cell carcinoma, and only 15
(11%) could identify the etiologic agents responsible for the disease. This poor knowledge is in accordance with studies of other groups of dental
care workers in other parts of the world [55,56],
which demonstrates a need for continuing education courses to increase knowledge on most aspects of oral malignancy. The situation is
expected to be worse among medical practitioners.
The oral cancer diagnosis continues to rely on
patients symptoms and physical examination
with biopsy conrmation, which may result in
delay in diagnosis accounting for the fact that
more than 75% of all patients are diagnosed at
stage III and IV. Although the cure rate of earlystage disease is more than 80%, with regional
metastasis the cure rate become less than 40%. To
detect these cancers at an early curable stage,
dental care workers must identify patients who are
at increased risk because of unhealthy lifestyles
and target advice and counseling to them and
perform systematic examination of the oral cavity.
Recently, the American Cancer Society identied

Fig. 6. Outcome of oral cancer screening trials. (From Moles DR, Downer MC, Speight PM. Meta-analysis of measures
of performance reported in oral cancer and precancer screening studies. Br Dent J 2002;192:3404; with permission.)

499

ORAL CANCER PREVENTION

R
A
N
D
O
M
I
Z
E

Control
(n=95,356)

Intervention
(n=96,517)
Arm A: Intervention group
N= 96,517
Compliance 90.8%
3 rounds of visual examination
at three year intervals
Arm B:Control group
N= 95,356
Compliance 83.9%
Standard care

Study Period
1996 to 2004

Precancer

2,252

Screen detected
cancer

131

Interval cancer

74

Total cancer

205

158

Incidence rate 43.7 per 100,000

37.6 per 100,000

Oral cancer
Mortality

87

Mortality rate

77

16.4 per 100,000

20.7 per 100,000

Fig. 7. Eect of screening on oral cancer mortality. Oral cancer screening by visual examination can lower mortality of
oral cancer, particularly in patients with high-risk habits. (From Sankaranarayanan R, Ramadas K, Thomas G, et al, for
the Trivandrum Oral Cancer Screening Study Group. Eect of screening on oral cancer mortality in Kerala, India: a cluster-randomised controlled trial. Lancet 2005;365:192733; with permission.)

cancers of the oral cavity as a potentially preventable cause of major morbidity and mortality, and
a comprehensive oral cancer examination is recommended every 3 years for persons aged 20 to 39
years and annually for individuals aged 40 years
and older [57].
Nicotera and colleagues [58] explored knowledge of risk factors and diagnostic procedures
for oral cancer and attitudes and behavior among
dental hygienists in Italy. A random sample of 500
dental hygienists received a questionnaire by mail
that focused on demographics and practice characteristics, knowledge, and attitudes and behaviors regarding oral cancer assessment practices.
Almost all dental hygienists correctly indicated tobacco usage and having a prior oral cancer lesion
as risk factors. Less than half (42.8%) recognized
that an early oral lesion usually is a small, painless, red area, and only 4.2% knew the examination procedures of the tongue. Further
educational interventions to early detect and prevent oral cancer are strongly needed.

an ideal model to investigate various chemopreventive strategies for the following reasons:

Chemoprevention

Retinoids

Chemoprevention is dened as the administration of agent(s) to block or reverse carcinogenesis

[59]. Oral cancer is one of the solid tumors in
which the proof of principle of this strategy was
rst established. Chemoprevention in oral cancer
has been directed toward reversing premalignant
lesions and preventing SPT. Interfering with the
carcinogenetic process early in the pathway before
malignant transformation and preventing second
primary lesions represents an attractive approach
for reducing the incidence and related morbidity
and mortality of oral cancer. Oral cancer also is

Retinoid, either in the natural or synthetic

form, is the most widely investigated chemopreventive agent in oral cancer.
Hong WK and colleagues [60] reported the
rst randomized chemopreventive study for oral
cancer (Fig. 8). They studied the role of high-dose
retinoid as a chemopreventive agent. The patients
were randomized into two arms: arm A involved
isotretinoin (13-cRA), 1 to 2 mg/kg for 3 months,
and arm B was the placebo arm. All patients were
followed up for six months. They recruited 24 patients in arm A and 20 patients in arm B. Clinical

1. It has known etiologic factors, namely, tobacco, alcohol, betel nut chewing, and viruses
2. Dierences in the risk factors inuences only
the subsite prediction and not the biologic
behavior or clinical response
3. Strong association with established premalignant lesions, such as leukoplakia, erythroplakia, and oral submucous brosis
4. It has a well-dened tumor progression
model in which cancer progresses from normal epithelium to mild, moderate, and severe
dysplasia to carcinoma in situ and frank invasive cancer
5. The lesions can be examined readily and repeatedly and subjected to biopsy
Various agents have been studied for chemoprevention of oral cancer, including retinoids,
beta-carotene, vitamin E, selenium, and COX-2
inhibitors.

500

KURIAKOSE & SHARAN

R
A
N
D
O
M
I
Z
E

Arm A:
13 cRA x 3 months
Isotretinoin1-2mg/kg

Follow-up
x 6 months

13-cRA
(n=24)

Placebo
(n=20)

Clinical response

16 (67%)

2 (10%)

Histologic response

13 (54%)

2 (10%)

Adverse effects

Arm B:
Placebo x 3 months

Cutaneous 19 (79%) 4 (20%)

Conjunctivitis 13 (54%) 2 (10%)
Hyperglyceridemia 27(70%) 6 (25%)

Fig. 8. Retinoid as chemoprevention agent. Landmark study provides proof of principle that chemoprevention may be
eective in oral cancer. High-dose retinoid had unacceptable toxicity and 50% relapsed within 3 months of stopping
therapy. (Data from Hong WK, Endicott J, Itri LM, et al. 13-cis-retinoic acid in the treatment of oral leukoplakia.
N Engl J Med 1986;315:15015.)

response was found in 16 patients (67%) in arm A

and only two patients (10%) in arm B. Thirteen
patients (54%) of arm A showed histologic response, whereas only two patients (10%) in arm
B showed histologic response. Isotretinoin was associated with signicant adverse eects, however,
such as dermatitis (79%), conjunctivitis (54%),
and hyperglyceridemia (70%). The authors concluded that isotretinoin is eective but associated
with unacceptable toxicity, and 50% of patients
experienced relapse within three months of stopping treatment.
Two years after this landmark trial, Stich and
colleagues [61] published the result of a large international collaborative, randomized clinical trial
among smokeless tobacco users with leukoplakia
in Kerala, India (Fig. 9). The result demonstrated
that vitamin A is as eective in patients using
smokeless tobacco (57.1%) as that reported by
Hong and colleagues (67%). Cessation of the therapeutic intervention in these studies showed

Tobacco
Chewers
From Kerala, India

R
A
N
D
O
M
I
Z
E

Arm A:
2000 IU Vit A / wk
x 6 months
Arm B:
Placebo x 6 months

relapse of the lesion in three months. Signicant

toxicities associated with the treatment precluded
its long-term use.
The issues of toxicity and relapse were subsequently addressed by Lippman and colleagues
[62], who studied the eect of low-dose 13-cRA
(0.5 mg/kg/d) versus beta-carotene (30 mg/kg/d)
for short-term maintenance after induction with
high-dose 13-cRA (Fig. 10). Consistent with the
results of Hongs study, the clinical response rate
after induction therapy was 55%, with reversal
of dysplasia in 43% of these responders. Fiftythree patients were fully assessable after maintenance. Further lesion response was observed in
33% and 10% of patients in the 13-cRA and the
beta-carotene groups, respectively. The rate of
disease progression during or after low-dose 13cRA maintenance therapy was only 8% versus
55% in the beta-carotene arm (P ! 0.001). This
study established that induction with 13-cRA followed by low-dose maintenance is eective in

Vitamin A
(n=21)

Placebo
(n=33)

Complete
remission

12 (57%)

1 (3%)

Prevention of
new lesions

21 (100%)

7 (21%)

Disease Progression

Fig. 9. Vitamin A chemoprevention is eective in tobacco chewers as in smokers. (Data from Stich HF, Homby AP,
Mathew B, et al. Response of oral leukoplakias to the administration of vitamin A. Cancer Lett 1988;40:93101.)

501

ORAL CANCER PREVENTION

Isotretinoin
1.5 mg/kg
x 3 months

CR or PR 36 (55%)
Stable 23 (35%)

Progression 7

R
A
N
D
O
M
I
Z
E

Arm A:
Isotretinoin (0.5 mg/kg)
x 9 months
Arm B:
Beta-carotine 30mg
x 9 months

Isotretinoin
(n=26)
Response 22 (92%)
/stable

Beta-carotine
(n=33)
13 (45%)

Fig. 10. Long-term maintenance low dose isotretinoin is eective in oral premalignant lesions with minimal adverse
eects. (Data from Lippman SM, Batsakis JG, Toth B, et al. Comparison of low-dose isotretinoin with beta-carotene
to prevent oral carcinogenesis. N Engl J Med 1993;328:1520.)

preventing the recurrence associated with treatment cessation, with acceptable toxicity.
In a placebo-controlled study, Han and colleagues [63] revealed the signicant activity of synthetic retinoid fenritinide (4-HPR) in reversing
oral premalignant lesion. Chisea and colleagues
[64] from Italy showed that 4-HPR is eective as
an adjuvant to prevent relapse after excision of
leukoplakia with minimal toxicity (Fig. 11).
Retinoids to prevent second primary tumor
Patients who have undergone successful management of primary oral cancer are at increased
risk of SPT in the upper aerodigestive tract, which
is a leading cause of mortality. In light of the trials
of retinoids in oral leukoplakia, Hong and colleagues [65] prospectively studied 103 patients
with stage I-IV (M0) HNSCC in a phase III adjuvant chemoprevention trial. After denitive therapy, patients were randomized to receive either

Leukoplakia
surgical excision

R
A
N
D
O
M
I
Z
E

high-dose 13-cRA (100 mg/m2/d) or placebo. Although there were no signicant dierences between the two groups in the number of local,
regional, or distant metastases of the primary cancers, the treatment group had signicantly fewer
SPTs than the placebo group (7% versus 33% placebo; P 0.008). SPTs in the retinoid-treated patients also took signicantly longer time to
develop. In a study to investigate the eectiveness
of retinoid as an adjuvant to oral cancer treatment, Benner and colleagues [66] observed that although it did not improve the loco-regional
disease control rate, it signicantly lowered the
episodes of SPT development (Fig. 12).
Khuri et al [67,68] reported data from an intergroup, placebo-controlled, double-blinded study
that evaluated the ecacy of low-dose isotretinoin
(30 mg) in the prevention of SPTs in patients with
stage I or II HNSCC and the impact of smoking
status on SPT development (Fig. 13). They
randomized 1190 patients to receive 30 mg of

Arm A:
4-HRP x 12 months
Arm B:
Placebo x 12 months

Local relapse/
new lesions

4-HRP
(n=39)

Placebo
(n=41)

3 (8%)

12 (29%)

Fig. 11. Fen-retinoid (4-HRP) is eective as adjuvant treatment after excision of oral leukoplakia with minimal toxicity.
(Data from Chiesa F, Tradati N, Marazza M, et al. Prevention of local relapses and new localisations of oral leukoplakias with the synthetic retinoid fenretinide (4-HPR). Preliminary results. Eur J Cancer B Oral Oncol 1992;28B;97102.)

502

KURIAKOSE & SHARAN

R
A
N
D
O
M
I
Z
E

Arm A:
13 cRA x 12 months
50-100 mg/m2
Arm B:
Placebo x 12 months

13-cRA
(n=49)

Placebo
(n=51)

15 (31%)
16 (33%)

17 (33%)
17 (33%)

2 (4%)
7 (14%)

12 (24%)
16 (31%)

Disease Progression
Follow-up
(Median 54.5
Months)

32 months
54 months
Second Primary Tumors
32 months
54 months

Fig. 12. Retinoid as adjuvant to cancer treatment is ineective. Retinoid therapy lowers SPTs. (Data from Benner SE,
Winn RJ, Lippman SM, et al. Regression of oral leukoplakia with alpha-tocopherol: a community clinical oncology program chemoprevention study. J Natl Cancer Inst 1993;85:447.)

isotretinoin daily or placebo for three years. The

annual SPT rate was 4.7% in both arms, and there
were no signicant dierences in overall survival
or recurrence-free survival. There was a transient,
statistically nonsignicant pattern of fewer recurrences with isotretinoin, which was lost after treatment cessation. Continued smoking was shown to
have a signicant adverse eect on disease-free
survival [67,68].
Similarly, a large European study (EUROSCAN), which investigated the ecacy of retinoids in lowering SPTs, reported a negative result
[69]. The EUROSCAN study consisted of two parallel clinical trials of two high-risk populations: patients with completely resected nonsmall-cell lung
cancer and patients with head and neck cancer after curative treatment. The agents used were N-acetylcysteine and retinyl palmitate. The study found
no benet in either study in terms of survival,
event-free survival, or SPT rate.

Stage I / II
HNSCC
(n=1,190)

R
A
N
D
O
M
I
Z
E

Biochemoprevention
Biochemoprevention therapy combines highdose isotretinoin, alpha-tocopherol, and interferon-a Alpha-tocopherol was chosen because of
its synergistic eects with retinoids, its ability to
decrease the toxicity of isotretinoin, and its
minimal side eect prole. Although it showed
response in laryngeal dysplasia, a prospective,
nonrandomized phase II trial by Papdimitrakopoulou and colleagues [70] was found to be ineffective in patients with oral leukoplakia.
Shin and colleagues [71,72] conducted a bioadjuvant phase II trial with isotretinoin, alphatocopherol, and interferon-a for 12 months in patients with locally advanced head and neck cancer
and demonstrated its eectiveness as an adjuvant
to prevent recurrences after denitive treatment of
head and neck cancer. Its eect was long lasting
(Fig. 14). Combination therapy also was tried in

13-cRA
Arm A:
13 cRA x 36 months

Disease Progression

Arm B:
Placebo x 36 months

Second Primary Tumors

Placebo

no difference

4.7%

4.7%

Fig. 13. Retinoids is ineective in reducing SPTs. Retinoid does not improve disease control rate. (Data from Khuri FR,
Kim ES, Lee JJ, et al. The impact of smoking status, disease stage, and index tumor site on second primary tumor incidence and tumor recurrence in the head and neck retinoid chemoprevention trial. Cancer Epidemiol Biomarker Prev
2001;10(8):8239.)

503

ORAL CANCER PREVENTION

retinoid-resistant tumors, including advanced premalignant lesions of the oral cavity and larynx.
Lippman and colleagues [73] showed p53 overexpression, high levels of polysomy, and lack of
RAR-beta upregulation as predictors of retinoid
resistance. In terms of locoregional control and
incidence of SPT, Toma and colleagues [74] reported a clinical trial with negative results using
biochemoprevention as adjuvant treatment for
stage III/IV head and neck cancer (Fig. 15).

Other agents
Other agents, such as selenium, have been
investigated in nonrandomized clinical trials and
have showed response in oral leukoplakia [75]. A
trial of alpha-tocopherol produced a response
rate of 46% in some retinoid-resistant lesions [64].
Bowman-Birk inhibitor
Bowman-Birk inhibitor (BBI) is a protease inhibitor found in soybeans that has demonstrated
a broad spectrum of anticarcinogenesis. The BBIs
mechanisms of action remain unknown; however,
they likely function by acting like one or more of
the endogenous tumor suppressor proteins that
possess protease inhibitory activity. Alternatively,
BBI also maintains anti-inammatory properties
while inhibiting free radical production [76]. BBI
concentrate prevents development of malignancy
in a large number of animal models [77]. Phase I
and phase II trials showed no evidence of

Locally
Advanced
HNSCC

R
A
N
D
O
M
I
Z
E

Arm A:
13 cRA
IFN-
x 12 months
-TP
Arm B:
No Rx x 12 months

signicant (grade 2 or higher) toxicity [78,79].

The latter trial demonstrated an overall clinical
response to BBI administration in 31% of subjects
(10/32). Two intermediate markersdprotease
activity and neu expressiondare being evaluated
in oral cancer chemoprevention trials of BBI.
Sulindac
Sulindac is a pan-COX inhibitor. It has been
shown to have anti-neoplastic eects against
human oral squamous cell carcinoma. Sulindac
also is reported to be eective in preventing colon
and esophageal tumors in several animal models.
Recently it was shown to be safe and eective in
humans for the prevention of polyps in familial
adenomatous polyposis. It is currently being tested
for ecacy in the management of oral leukoplakia
in a collaborative clinical trial at Memorial SloanKettering Cancer Institute, New York, and Amrita Institute of Medical Sciences, India. This study
was initiated because of the known dramatic
overexpression of COX-2 in head and neck cancers
and leukoplakia compared with true normal tissue
and the known high levels of prostaglandins that
may contribute to carcinogenesis in these patients.
Ketorolac
In the 2003 American Society of Clinical
Oncology proceedings, the National Cancer
Institute reported a negative result of a clinical
trial of the topical nonsteroidal anti-inammatory
drug ketorolac for oral leukoplakia.

B
I
O
P
S
Y

Follow-up
x 6 months

Biomarker Studies
p53 mutations
p53 expression
11q13 amplification
9p21 loss
cyclin D1 expression
p16 expression

Fig. 14. Bioadjuvant phase II trial of isotretinoin, interferon-a, alpha-tocopherol. (Data from Shin DM, Richards TJ,
Seixas-Silva JA. Phase II trial of bioadjuvant therapy with interferon-alpha2a, 13-cis-retinoic acid, and alpha-tocopherol
for locally advanced squamous cell carcinoma of the head and neck: long term follow up. Proc Am Soc Clin Oncol 2003;
22:1995a.)

504

KURIAKOSE & SHARAN

Stage III/IV
HNSCC
(n=260)

R
A
N
D
O
M
I
Z
E

Arm A:
13 cRA 0.5mg/kg/d
IFN2a 3,000,000 UI x3/week

Arm B:
Observation

13-cRA+IFN2a Control
(n=126)
(n= 126)
Mean
Overall survival
F/U
39 months Disease free survival
Second Primary Tumors

NS
NS
NS

Fig. 15. Biochemoprevention well tolerated but ineective to reduce SPT. (Data from Toma S, Bonelli L, Sartoris A,
et al. 13-cis retinoic acid in head and neck cancer chemoprevention: results of a randomized trial from the Italian
Head and Neck Chemoprevention Study Group. Oncol Rep 2004;11(6):1297305.)

Curcumin
Curcumin is a polyphenol derived from the
plant Curcuma longa, commonly called turmeric.
Turmeric is a member of the ginger family; its rhizomes produce a brilliant yellow dye. The primary
bioactive constituents in turmeric have been
found to be the phenolic curcuminoids, the most
important of which is curcumin (diferuloylmethane). Curcumin is one of the three curcuminoids
in turmeric, which is known to possess varying degrees of antioxidants [80,81] and anti-inammatory properties [8284]. Curcumin protects cells
against free radicals that promote cancer by damaging DNA and activating genes. Analysis of its
structure revealed the presence of beta-diketone
moiety and phenolic hydroxy groups that are believed to contribute to antioxidation.
Curcumin is a powerful inhibitor of nuclear
factor kappa beta (NF-kB) [85]. NF-kB is implicated in the oncogenesis of several malignancies,
which indicates that NF-kB activation may be
a common pathway of broad importance in cancer. Curcumin can suppress tumor initiation, promotion, and metastasis of cancer [86]. It has been
shown to inhibit angiogenesis by chelating metals
used by enzymes metalloproteinases that promote
angiogenesis [87]. Importantly, dietary administration of curcumin during initiation or postinitiation periods signicantly suppresses development
of chemically induced tumors in mice [88]. It
also reduces formation of focal areas of dysplasia
and aberrant crypt foci in the colon that are early
preneoplastic lesions in rodents. Pereira and
colleagues [89] reported that administration of
curcumin continuously during the initiation
and postinitiation phases signicantly inhibited

development of colonic adenomas in a rat model.

Curcumin treatment can be delayed after the carcinogen administration in experimental carcinogenesis and still be eective so as to provide
baseline knowledge for possible clinical use of
this agent in secondary prevention of colon cancer
in high-risk individuals, such as patients with colonic polyps.
A phase II/III double-blind, randomized, placebo-controlled trial of chemoprevention of oral
premalignant lesions using curcumin is currently
underway at Amrita Institute of Medical Science
in Cochin, India. The primary objective of this
study is to evaluate the ecacy of curcumin in
subjects with oral premalignant lesions by clinical
and histologic response. The secondary objective
is to evaluate the mechanisms of in vivo modulation of NF-kB by curcumin and evaluate the
treatment eects in modulating the expression of
biomarkers I kappa B alpha kinase activation and
p65 nuclear translocation after 4 months of
treatment in comparison with the placebo.

Other dietary agents

Chemoprevention using dietary agents, a costeective approach, has received growing consideration as a potential means to control the incidence of cancer. Recently, much attention has
been focused on fruits and vegetables with potent
antimutagenic and anticarcinogenic properties
[90]. In particular, tomatoes and garlic are recognized to possess a wide range of benecial eects.
Tomatoes and tomato-based products are rich
sources of lycopene, an antioxidant carotenoid reported to be a more stable and potent singlet

ORAL CANCER PREVENTION

oxygen-quenching agent compared with other carotenoids [91].

Garlic (Allium sativum)
Garlic, which is used for avoring and has
been used as a medicinal herb for centuries,
exhibits antioxidant, antimutagenic, and anticarcinogenic eects [92]. Epidemiologic studies have
provided evidence that increased intake of tomato
and garlic is associated with decreased cancer risk
[93,94]. Although lycopene has been found to exhibit only marginal inhibitory eects on phase I
enzymes, it strongly induces phase II enzymes
[95,96]. The organosulfur constituents of garlic,
such as diallyl-di-sulde and diallyl-tri-sulde,
are potent inhibitors of phase I enzymes and inducers of phase II enzymes [97]. Bhuvaneswari
and colleagues [98] evaluated the combinatorial
chemopreventive eects of tomato and garlic on
hamster buccal pouch carcinogenesis induced by
7,12-dimethylbenz[a]anthracene (DMBA). Combined administration of tomato and garlic eectively suppressed the incidence and mean tumor
burden of hamster buccal pouch carcinomas
and the frequency of bone marrow micronuclei.
This study clearly indicated that a broad spectrum
of chemopreventive eects could be achieved
through altered expression of xenobiotic-metabolizing enzymes. The combinatorial chemopreventive eect of tomato and garlic must be
established in dierent animal tumor models because of tissue-specic variations in the expression
of phase I and II enzyme systems. It also must be
investigated in systematic clinical trials.
Potential new targets for chemoprevention
Progress in molecular biology has made it
possible to identify the genotypic alterations that
lead to the development of malignant clones and
molecular markers of specic stages in multistep
carcinogenesis. Potential new targets for chemotherapy, which are under consideration in oral
cancer, include the following:
H-ras gene
EGFR
p53 gene
COX-2 inhibitors
NF-kB
H-ras gene
The ras proto-oncogene products are membrane-associated, guanine nucleotide-binding

505

proteins that serve as molecular switches for signal transduction pathways. Saranath and colleagues [99] observed frequent mutations of Hras in an Indian population that had betel-nut
chewing habits and oral cancer. This nding was
conrmed in other studies [100].
Epidermal growth factor receptor
EGFR, which is highly expressed by many
solid tumor types [101], is one target for novel
therapies. EGFR is a member of the erbB family
of receptorsdtransmembrane glycoproteins, that
play an important role in cell growth and dierentiation using tyrosine kinase activity as the signal
transduction mechanism. EGFR signaling has
been shown to be critical for many aspects of tumor biology: cell proliferation, angiogenesis, metastasis, and inhibition of apoptosis [101]. Most
head and neck tumors express high levels of
EGFR relative to normal tissues [102104]. The
level of EGFR expression has prognostic importance in head and neck cancer, with higher levels
indicating a poorer outcome in terms of progression-free
survival
and
overall
survival
[102,103,105,106]. There are at least two EGFR ligandsdepidermal growth factor and transforming growth factor-alpha. The phosphorylated
form of the receptor is biologically active. Tobacco smoke extract induces EGFR phosphorylation (which seems to be critical to tobaccoinduced COX-2 expression) in oral carcinogenesis
[107].
In human head and neck cancer cell lines, the
EGFR thyrosine kinase inhibitors and EGFR
monoclonal antibodies (centuximab) inhibit cell
proliferation in a dose-dependent manner [108
111]. Currently, ongoing phase III clinical trials
are investigating their therapeutic potential in
head and neck cancer. An ongoing phase II clinical trial is investigating the chemopreventive eectiveness of EGFR thyrosine kinase inhibitor in
oral cancer.
Nuclear factor kappa beta
NF-kB, a nuclear transcription factor, is an
inducible and ubiquitously expressed transcription
factor for genes involved in cell survival, cell
adhesion, inammation, dierentiation, and
growth [112,113]. A heterotrimeric complex that
consists of p50, p65, and IkBa [114,115], NF-kB
is present in its inactive state in the cytoplasm.
When activated, IkBa is degraded and p50-p65
heterodimer is translocated to the nucleus, binds

506

KURIAKOSE & SHARAN

the DNA (at the promoter region), and activates

genes. Several genes that mediate tumorigenesis
and metastasis are regulated by NF-kB.
NF-kB is activated by carcinogens, tumor
promoters, inammatory cytokines, and chemotherapeutic agents [116119]. Cancer cells often
overexpress NF-kB and use it as a means to proliferate. It is constitutively activated in human and
murine squamous cell carcinomas [120]. The activation of NF-kB can suppress apoptosis, thus
promoting chemoresistance and tumorigenesis.
Cumulative evidence in head and neck, lymphoid,
breast, gastric, colorectal, and prostate cancers is
consistent with the hypothesis that NF-kB is constitutively activated and is a major factor in the
pathogenesis of cellular and host alterations in
these cancers. It also plays an important role in
oral carcinogenesis. Oral cancer constitutively expresses NF-kB and IKK (IkBf kinase). Evidence
suggests that NF-kB is an ideal target for chemoprevention and chemosensitization. Treatment
with curcumin-inhibited NF-kB was monitored
by DNA binding, IKK activation, and p65 nuclear translocation [85]. An ongoing clinical trial
is investigating the eect of curcumin as a chemopreventive agent in oral cancer.
p53 gene
The p53 gene encodes a nuclear protein with
transcription factor activity, an essential component of the G1-S transition checkpoint, thus contributing to cell cycle control, DNA repair and
synthesis, and genomic integrity [121,122]. The alterations of p53 impair the ability of cells to repair
and undergo apoptosis in response to DNA damage. The most common p53 alteration is a point
mutation conned primarily to exons 5 to 8. The
dysregulation of p53 in the mucosal epithelium
correlates with increased proliferative activity
[123]. p53 status has been shown to be a predictor
of progression of premalignant oral dysplasias to
invasive cancers [124]. The identication of distinct p53 mutations in tumor-distant epithelia suggests that p53 is a useful molecular marker of the
eld wide and multifocal nature of the head and
neck carcinogenic process [123]. The use of p53
as a molecular marker has suggested that multiple
primary head and neck tumors originate independently and that distinct p53 alterations may help
to dierentiate between an SPT and primary tumor recurrence or metastasis [125127]. Brennan
and colleagues [128] found that 52% of patients
(13/25) who seemed to have had complete tumor

resection on the basis of a negative histopathologic assessment were positive for a p53 mutation
in at least one tumor margin. In 5 of 13 patients
with positive margins by this method (38%), the
carcinoma recurred locally compared with 0 of
12 patients with negative margins. Molecular
analysis identied neoplastic cells in 6 of 28 lymph
nodes (21%) that were initially negative by histopathologic assessment. An ongoing clinical trial
using Onxyx virus, a replication-decient virus
that specically infects cells with p53 mutated
cells, will attempt to validate the chemopreventive
eectiveness in oral leukoplakia.
Cyclo-oxygenase-2 inhibitors
There are two isoforms of COX. The COX-1
isoform is related to normal cell activity, but
COX-2 plays a key role in inammation. Celecoxib is one of newly developed inhibitors that
specically target COX-2 but not COX-1. The
study by Wang and colleagues [129] provided the
rst evidence that celecoxib is highly eective and
safe for the inhibition of oral carcinomas cells in
a nude mouse model when administered early. It
suggests its preventive ecacy in oral cancer.
Antiangiogenic activity of celecoxib is thought
to be a major cause of its ecacy in the early
treatment of carcinoma. Sood and colleagues
[130] performed a long-term chemopreventive experiment using celecoxib (3% or 6%) through
topical application in a DMBA-induced hamster
oral cancer model. Both 3% and 6% reduced
the incidence of squamous cell carcinoma at the
postinitiation stage. Such a cancer preventive effect was correlated with inhibition of prostaglandin E2 biosynthesis [131].
These ndings warrant human clinical trials to
determine the ecacy of celecoxib or other COX
inhibitors as chemoprevention, especially in preventing secondary lesions in individuals after
primary treatment of oral cancer.

Summary
Along with attempts to improve the cure rate
of oral cancer; concerted eorts to prevent the
disease in the community should be undertaken.
This is particularly true for the high-risk population and high-risk individuals. A targeted prevention in high-risk individuals with high-risk
lesions using agents targeted to key molecules in
the oral carcinogenesis process should have an
impact in lowering the disease morbidity and

ORAL CANCER PREVENTION

mortality. Advances in molecular biology have

helped to dene these high-risk individuals with
high-risk lesions and novel chemopreventive targets. This targeted approach should overcome
some of the earlier setbacks observed in oral
cancer prevention research.
References
[1] Mignogna MD, Fedele S, Lo Russo L, et al. Oral
and pharyngeal cancer: lack of prevention and
early detection by health care providers. Eur J Cancer Prev 2001;10:3813.
[2] Lippman SM, Hong WK. Molecular markers of
the risk of oral cancer. N Engl J Med 2001;344:
13236.
[3] Mao L, Hong WK, Papadimitrakopoulou VA. Focus on head and neck cancer. Cancer Cell 2004;5:
3116.
[4] Abbruzzese JL, Lippman SM. The convergence of
cancer prevention and therapy in early phase clinical drug development. Cancer Cell 2004;6:3216.
[5] Brennan JA, Boylo JO, Koch WM, et al. Association between cigarette smoking and mutation of
the p53 gene in squamous cell carcinoma of the
head and neck. N Engl J Med 1995;332:7127.
[6] Wu X, Lippman SM, Lee JJ, et al. Chromosome instability in lymphocytes: a potential indicator of
predisposition to oral premalignant lesions. Cancer
Res 2002;62:28138.
[7] Shin DM, Charuruks N, Lippman SM, et al. p53
protein accumulation and genomic instability in
head and neck multistep tumorigenesis. Cancer
Epidemiol Biomarkers Prev 2001;10:6039.
[8] Slaughter DP, Southwick HW, Smejkal W. Field
cancerization in oral stratied squamous epithelium: clinical implications of multicentric origin.
Cancer 1953;6:9638.
[9] Braakhuis BJ, Tabor MP, Kummer JA, et al. A genetic explanation of Slaughters concept of eld
cancerization: evidence and clinical implications.
Cancer Res 2003;63:172730.
[10] Jang SJ, Chiba I, Hirai A, et al. Multiple oral squamous epithelial lesions: are they genetically related?
Oncogene 2001;20:223542.
[11] Weisberger D. Precancerous lesions. J Am Dent
Assoc 1957;54:5078.
[12] Einhorn J, Wersall J. Incidence of oral carcinoma
in patients with leukoplakia of the oral mucosa.
Cancer 1967;20:218993.
[13] Silverman SJ. Observations on the clinical characteristics and natural history of oral leukoplakia.
J Am Dent Assoc 1968;76:7727.
[14] Silverman SJ, Bhargava K, Mani NJ, et al. Malignant transformation and natural history of oral leukoplakia in 57,518 industrial workers in Gujarat,
India. Cancer 1976;38:17905.

507

[15] Silverman SJ, Gorsky M, Lozada F. Oral leukoplakia and malignant transformation: a follow-up
study of 257 patients. Cancer 1984;53:5638.
[16] Neville BW, Day TA. Oral cancer and precancerous lesions. CA Cancer J Clin 2002;52:195215.
[17] Roed-Petersen B. Cancer development in oral leukoplakia: follow-up of 331 patients. J Dent Res
1971;50:711.
[18] Kulasegaram R, Downer MC, Jullien JA, et al.
Case-control study of oral dysplasia and risk habits
among patients of a dental hospital. Eur J Cancer
1995;31:22731.
[19] Wynder EL, Bross JJ, Feldman RM. A study of the
etiologic factors in cancer of the mouth. Cancer
1957;10:130023.
[20] Rothamn KJ, Keller AZ. The eect of joint exposure to alcohol and tobacco on the risk of cancer
of the mouth and pharynx. J Chronic Dis 1972;
25:7116.
[21] Graham S, Dayal H, Rohrer T. Dentition, diet, tobacco and alcohol in the epidemiology of oral cancer. J Natl Cancer Inst 1977;59:16118.
[22] Brugere J, Guenel P, Leclerc A, et al. Dierential
eects of tobacco and alcohol in cancer of the
larynx, pharynx and mouth. Cancer 1986;57:
3915.
[23] Mashberg A, Boetta P, Winkelman R, et al. Tobacco smoking, alcohol drinking, and cancer of
the oral cavity and oropharynx among US veterans. Cancer 1993;72:136975.
[24] Shiu MN, Chen THH, Chang SH, et al. Risk factors for leukoplakia and malignant transformation
to oral carcinoma: a leukoplakia cohort in Taiwan.
Br J Cancer 2000;82:18714.
[25] Martin GC, Brown JP, Eier CW, et al. Oral leukoplakia status six weeks after cessation of smokeless
tobacco use. J Am Dent Assoc 1999;130(7):94554.
[26] Schinke SP, Gilchrist LD, Schilling RF II, et al.
Smoking and smokeless tobacco use among
adolescents: trends and intervention results. Public
Health Rep 1986;101(4):3738.
[27] Sankaranarayanan R, Nair MK, Mathew B, et al.
Recent results of oral cancer research in Kerala, India. Head Neck 1992;14:10712.
[28] Sussman S, Dent CW, Stacy AW, et al. Project towards no tobacco use: 1-year behaviour outcomes.
Am J Public Health 1993;83:124550.
[29] Elder JP, Wildey M, de Moor C, et al. The
long term prevention of tobacco use among junior high school students: classroom and telephone interventions. Am J Public Health 1993;
83(9):123944.
[30] Stevens MM, Freeman DH Jr, Mott LA, et al.
Smokeless tobacco use among children: the New
Hampshire study. Am J Prev Med 1993;9:1607.
[31] Shiu M-N, Chen TH-H. Impact of betel quid, tobacco and alcohol on three-stage disease natural
history of oral leukoplakia and cancer: implication

508

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

KURIAKOSE & SHARAN

for prevention of oral cancer. Eur J Cancer Prev

2004;13:3945.
American Cancer Society. Update January 1992:
the American Cancer Society guidelines for the cancer -related checkup. CA Cancer J Clin 1992;42:
445.
British Dental Association. Oral cancer: guidelines
for early detection. Occasional paper no. 5; London: British Dental Association; 1998.
Workgroup of the UICC Epidemiology and Prevention Programme. Progress report from the
workgroup on the success of prevention and detection programmes. Geneva (Switzerland): International Union Against Cancer; 2001.
Burzynski N, Firriolo FJ, Butters JM, et al. Evaluation of oral cancer screening. J Cancer Educ 1997;
12:959.
Mignogna MD, Fedele S, Lo Russo L, et al. Costs
and eectiveness in the care of patients with oral
and pharyngeal cancer: analysis of a paradox. Eur
J Cancer Prev 2002;11:2058.
Allison P, Locker D, Feine JS. The role of diagnostic delays in the prognosis of oral cancer:
a review of the literature. Oral Oncol 1998;34:
16170.
Mashberg A. Diagnosis of early oral and oropharyngeal squamous carcinoma: obstacles and their
amelioration. Oral Oncol 2000;36:2535.
Mignogna MD, Fedele S, Lo Russo L. The world
cancer report and the burden of oral cancer. Eur
J Cancer Prev 2004;13:13942.
UICC. Consensus/consultation session on oral
cancer. Presented at the 15th Asia Pacic Cancer
Conference. Madras, India, December 15, 1999.
Warnakulasuriya KAAS, Ekanayake ANI, Sivayoham S, et al. Utilisation of primary care workers for
early detection of oral cancer and precancer cases in
Sri Lanka. Bull World Health Organ 1984;62:
24350.
Warnakulasuriya KAAS, Nanayakkara BG. Reproducibility of an oral cancer and precancer detection program using a primary health care
model in Sri Lanka. Cancer Detect Prev 1991;
15:3314.
Mehta FS, Gupta PC, Bhonsle RB, et al. Detection
of oral cancer using basic health workers in an area
of high oral cancer incidence in India. Cancer Detect Prev 1986;9:21925.
Mathew B, Sankaranarayanan R, Sunilkumar KB,
et al. Reproducibility and validity of oral visual inspection by trained health workers in the detection
of oral precancer and cancer. Br J Cancer 1997;76:
3904.
Mathew B, Wesley R, Dutt SC, et al. Cancer
screening by local volunteers. World Health Forum
1996;17:3778.
Mashberg A, Barsa P. Screening for oral and oropharyngeal squamous cell carcinomas. CA Cancer
J Clin 1984;34:2628.

[47] Fernandez Garrote L, Sankaranarayanan R, Lence

Anta JJ, et al. An evaluation of the oral cancer control programme in Cuba. Epidemiology 1995;6:
42831.
[48] Mathew B, Sankaranarayanan R, Wesley R, et al.
Evaluation of mouth self-examination in the control of oral cancer. Br J Cancer 1995;71:3979.
[49] Moles DR, Downer MC, Speight PM, et al. Metaanalysis of measures of performance reported in
oral cancer and precancer screening studies. Br
Dent J 2002;192(6):3404.
[50] Sankaranarayanan R, Ramadas K, Thomas G,
et al. for the Trivandrum Oral Cancer Screening
Study Group. Eect of screening on oral cancer
mortality in Kerala, India: a cluster-randomised
controlled trial. Lancet 2005;365:192733.
[51] Allison P, Franco E, Black M, et al. The role of professional diagnostic delays in the prognosis of upper aerodigestive tract carcinoma. Oral Oncol
1998;34(2):14753.
[52] Allison P, Franco E, Feine J. Predictors of professional diagnostic delays for upper aerodigestive
tract carcinoma. Oral Oncol 1998;34(2):12732.
[53] Kerdpon D, Sriplung H. Factors related to advanced stage oral squamous cell carcinoma in
southern Thailand. Oral Oncol 2001;37(3):21621.
[54] Leao JC, Goes P, Sobrinho CB, et al. Knowledge
and clinical expertise regarding oral cancer among
Brazilian dentists. Int J Oral Maxillofac Surg
2005;34:4369.
[55] Clovis JB, Horowitz AM, Poel DH. Oral and pharyngeal cancer: knowledge and opinions of dentists
in British Columbia and Nova Scotia. J Can Dent
Assoc 2002;68(7):41520.
[56] Macpherson LM, McCann MF, Gibson J, et al.
The role of primary healthcare professionals in
oral cancer prevention and detection. Br Dent J
2003;195(5):27781.
[57] Smith RA, Cokkinides V, Eyre HJ. American Cancer Society guidelines for the early detection of cancer. CA Cancer J Clin 2003;53:2743.
[58] Nicotera G, Gnisci F, Bianco A, et al. Dental hygienists and oral cancer prevention: knowledge, attitudes and behaviors in Italy. Oral Oncol 2004;40:
63844.
[59] Lippman SM, Benner SE, Hong WK. Cancer chemoprevention. J Clin Oncol 1994;12:85173.
[60] Hong WK, et al. 13-cis retinoic acid in the treatment of oral leukoplakia. N Engl J Med 1986;
315:1301.
[61] Stich HF, Homby AP, Mathew B, et al. Response
of oral leukoplakias to the administration of vitamin A. Cancer Lett 1988;40:93101.
[62] Lippman SM, Batsakis JG, Toth BB, et al. Comparison of low-dose isotretinoin with beta-carotene
to prevent oral carcinogenesis. N Engl J Med 1993;
328:15.
[63] Han J, Jiao L, Lu Y, et al. Evaluation of N-4(hydroxycarbophanyl) retinamide as a cancer

ORAL CANCER PREVENTION

[64]

[65]

[66]

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

prevention agent and as a cancer chemotherepeutic

agent. In Vivo 1990;4:15360.
Chiesa F, Tradati N, Marazza M, et al. Prevention
of local relapses and new localisatons of oral leukoplakias with the synthetic retinoid fenretinide (4HPR): preliminary results. Eur J Cancer B Oral
Oncol 1992;28B:97102.
Hong WK, Lippman SM, Itri LM, et al. Prevention
of second primary tumors with 13cRA in squamous-cell carcinoma of the head and neck.
N Engl J Med 1990;323:795.
Benner SE, Winn RJ, Lippman SM, et al. Regression of oral leukoplakia with alpha-tocopherol:
a community clinical oncology program chemoprevention study. J Natl Cancer Inst 1993;85:447.
Khuri FR, Kim ES, Lee JJ, et al. The impact of
smoking status, disease stage, and index tumor
site on second primary tumor incidence and tumor
recurrence in the head and neck retinoid chemoprevention trial. Cancer Epidemiol Biomarker Prev
2001;10(8):8239.
Khuri FR, Lee JJ, Lippman SM, et al. Isotretinoin
eects on head and neck cancer recurrence and second primary tumors [abstract 359] Proc Am Soc
Clin Oncol 2003;22:90.
van Zandwijk N, Dalesio O, Pastorino U, et al.
EUROSCAN: a randomized trial of vitamin A
and N-Acetylcysteine in patients with head and
neck cancer or lung cancer. J Natl Cancer Inst
2000;92:97786.
Papadimitrakopoulou VA, Clayman GL, Shin
DM, et al. Biochemoprevention for dysplastic lesions of the upper aerodigestive tract. Arch Otolaryngol Head Neck Surg 1999;125(10):10839.
Shin DM, Khuri FR, Murphy B, et al. combined
interferon-alpha, 13-cis-retinoic acid, and alpha-tocopherol in locally advanced head and neck squamous cell carcinoma: novel bioadjuvant phase II
trial. J Clin Oncol 2001;19:30107.
Shin DM, Richards TJ, Seixas-Silva JA. Phase II
trial of bioadjuvant therapy with interferonalpha2a, 13-cis-retinoic acid, and alpha-tocopherol
for locally advanced squamous cell carcinoma of
the head and neck: long term follow up. Proc Am
Soc Clin Oncol 1995;2003:22.
Lippman S, Shin D, Lee J, et al. p53 and retinoid
chemoprevention of oral carcinogenesis. Clin Cancer Res 1995;55:169.
Toma S, Bonelli L, Sartoris A, et al. 13-cis retinoic
acid in head and neck cancer chemoprevention: results of a randomized trial from the Italian Head
and Neck Chemoprevention Study Group. Oncol
Rep 2004;11(6):1297305.
Toma S, Coialbu T, Collecchi P, et al. Biological aspects and perspectives applicable to the chemoprevention of the cancer of the upper respiratory
digestive tract. Acta Otorhinolaryngol Ital 1990;
10(Suppl 27):4154.

509

[76] Armstrong WB, Wan XS, Kennedy AR, et al.

Development of the Bowman-Birk inhibitor for
oral cancer chemoprevention and analysis of
neu immunohistochemical staining intensity with
Bowman-Birk inhibitor concentrate treatment.
Laryngoscope 2003;113(10):1687702.
[77] Messadi DV, Billings P, Schklar G, et al. Inhibition
of oral carcinogenesis by a protease inhibitor.
J Natl Cancer Inst 1986;76:44752.
[78] Armstrong WB, Kennedy AR, Wan XS, et al.
Single-dose administration of Bowman-Birk inhibitor concentrate in patients with oral leukoplakia.
Cancer Epidemiol Biomarkers Prev 2000;9:437.
[79] Armstrong WB, Kennedy AR, Wan XS, et al. Clinical modulation of oral leukoplakia and protease
activity by Bowman-Birk inhibitor concentrate in
a phase IIa chemoprevention trial. Clin Cancer
Res 2000;6:468491.
[80] Sharma OP. Antioxidant activity of curcumin and
related compounds. Biochem Pharmacol 1976;25:
18112.
[81] Toda S, Arichi H, Tanizawa H, et al. Natural antioxidant III: antioxidative components isolated
from rhizome of Curcuma longa L. Chem Pharm
Bull (Tokyo) 1985;33:17258.
[82] Ruby AJ, Babu KD, Rajasekharan KN, et al. Antitumour and antioxidant activity of natural curcuminoids. Cancer Lett 1995;94(1):7983.
[83] Tonnesen HHC, Lee CY, Haung MT, editors. Phenolic compounds in food and their eect on health:
analysis, occurrence and chemistry. ACS Symposium Series No. 506; 1992. p. 14353.
[84] Srimal RC. Pharmacology of diferuloylmethane
(curcumin), a non-steroidal anti-inammatory
agent. J Pharm Pharmacol 1973;25:44752.
[85] Han S-S, Chung S-T, Robertson DA, et al. Curcumin causes the growth arrest and apoptosis of B cell
lymphoma by downregulation of egr-1, C-myc,
Bcl-XL, NF-B, and p53. Clin Immunol 1999;
93(2):15261.
[86] Kawamori T, Steele VE, Kello GJ, et al. Chemopreventive eect of curcumin, a naturally occurring
anti-inammatory agent, during the promotion/
progression stages of colon cancer. Cancer Res
1999;59:597601.
[87] Arbiser JL, Klauber N, Rohan R, et al. Curcumin
is an in vivo inhibitor of angiogenesis. Mol Med
1998;4(6):37683.
[88] Rao CV, Simi B, Reddy B. Chemoprevention of colon carcinogenesis by dietary curcumin, a naturally
occurring plant phenolic compound. Cancer Res
1995;55:25966.
[89] Pereira MA, Grubbs CJ, Barnes LH, et al. Eects
of the phytochemicals, curcumin and quercetin,
upon azoxymethane-induced colon cancer and
7,12-dimethylbenz[a]anthracene-induced
mammary cancer in rats. Carcinogenesis 1996;17(6):
130511.

510

KURIAKOSE & SHARAN

[90] La Vecchia C, Tavani A. Fruits and vegetables and

human cancer. Eur J Cancer Prev 1998;7:38.
[91] Heber D, Lu Q-Y. Overview of mechanisms of action of lycopene. Exp Biol Med 2002;227:9203.
[92] Kik C, Kahane R, Gebhardt R. Garlic and health.
Nutr Metab Cardiovasc Dis 2001;11:5765.
[93] Giovannucci E. Tomatoes, tomato-based products,
lycopene, and cancer: review of the epidemiologic
literature. J Natl Cancer Inst 1999;91:31731.
[94] Thomson M, Ali M. Garlic (Allium sativum): a review of its potential use as an anti-cancer agent.
Curr Cancer Drug Targets 2003;3:6781.
[95] Breinholt V, Lauridsen ST, Daneshvar B, et al.
Dose-response eects of lycopene on selected
drug-metabolizing and antioxidant enzymes in the
rat. Cancer Lett 2000;154:20110.
[96] Gradelet S, Astorg P, Leclerc J, et al. Eects of canthaxanthin, astaxanthin, lycopene and lutein on
liver xenobioticmetabolising enzymes in the rat.
Xenobiotica 1996;26:4963.
[97] Wu CC, Sheen LY, Chen HW, et al. Dierential effects of garlic oil and its three major organosulfur
components in the hepatic detoxication system
in rats. J Agric Food Chem 2002;50:37883.
[98] Bhuvaneswari V, Abraham SK, Nagini S. Combinatorial antigenotoxic and anticarcinogenic eects
of tomato and garlic through modulation of xenobiotic-metabolizing enzymes during hamster buccal
pouch carcinogenesis. Nutrition 2005;21:72631.
[99] Saranath D, Chang SE, Bhoite LT, et al. High frequency mutation in codons 12 and 61 of H-ras oncogene in chewing tobacco-related human oral
carcinoma in India. Br J Cancer 1991;63:5738.
[100] Kuo MY, Chang HH, Hahn LJ, et al. Elevated ras
p21 expression in oral premalignant lesions and
squamous cell carcinoma in Taiwan. J Oral Pathol
Med 1995;24:25560.
[101] Baselga J. New technologies in epidermal growth
factor receptor-targeted cancer therapy. Signal
2000;1:1221.
[102] Rubin Grandis J, Melhem MF, Gooding WE, et al.
Levels of TGF-a and EGFR protein in head and
neck squamous cell carcinoma and patient survival.
J Natl Cancer Inst 1998;90:82432.
[103] Poon TCW, Chan ATC, To KF. Expression and
prognostic signicance of epidermal growth factor
receptor and HER2 protein in nasopharyngeal carcinoma [abstract 913]. Proc Am Soc Clin Oncol
2001;20:229.
[104] Santini J, Formento JL, Francoual M, et al. Characterization, quantication, and potential clinical
value of the epidermal growth factor receptor in
head and neck squamous cell carcinomas. Head
Neck 1991;13:1329.
[105] Dassonville O, Formento JL, Francoual M, et al.
Expression of epidermal growth factor receptor
and survival in upper aerodigestive tract cancer.
J Clin Oncol 1993;11:18738.

[106] Numico G, Colantonio I, Comino A, et al. EGFR

and survival of patients with locally advanced head
and neck cancer (HNC) and treated with alternating chemotherapy radiation (CT-RT): results of
a retrospective analysis [abstract 2018]. Proc Am
Soc Clin Oncol 2003;22:501.
[107] Dannenberg AJ, Lippman SM, Mann JR, et al.
Cyclooxygenase-2 and epidermal growth factor receptor: pharmacologic targets for chemoprevention. J Clin Oncol 2005;23:25466.
[108] Huang SM, Bock JM, Harari PM. Epidermal
growth factor receptor blockade with C225 modulates proliferation, apoptosis, and radiosensitivity
in squamous cell carcinomas of the head and
neck. Cancer Res 1999;59:193540.
[109] Di Gennaro E, Barbarino M, Bruzzese F, et al.
Critical role of both p27KIP1 and p21CIP1/
WAF1 in the antiproliferative eect of ZD1839
(Iressa), an epidermal growth factor receptor tyrosine kinase inhibitor, in head and neck squamous
carcinoma cells. J Cell Physiol 2003;195:13950.
[110] Huang SM, Li J, Harari PM. Molecular inhibition
of angiogenesis and metastatic potential in human
squamous cell carcinomas after epidermal growth
factor receptor blockade. Mol Cancer Ther 2002;
1:50714.
[111] Matheny KE, Barbieri CE, Sniezek JC, et al. Inhibition of epidermal growth factor receptor signaling decreases p53 expression in head and neck
squamous carcinoma cells. Laryngoscope 2003;
113:9369.
[112] Kabrun N, Enrietto PJ. The Rel family of proteins
in oncogenesis and dierentiation. Semin Cancer
Biol 1994;5(2):10312.
[113] Baldwin AS. Control of oncogenesis and cancer
therapy resistance by the transcription factor NFkappa B. J Clin Invest 2001;107(3):2416.
[114] Karin M, Delhase M. The IkappaB kinase (IKK)
and NF-kappa B: key elements of proinammatory
signaling. Semin Immunol 2000;12:8598.
[115] Bharti AC, Aggarwal BB. Nuclear factor-kappa B
and cancer: its role in prevention and therapy. Biochem Pharmacol 2002;64:8838.
[116] Pahl HL. Activators and target genes of Rel/NFkappaB transcription factors. Oncogene 1999;
18(49):685366.
[117] Yan Z, Subbaramaiah K, Camilli T, et al. Benzo
(a)pyrene induces the transcription of cyclooxygenase-2 in vascular smooth cells: evidence for the
involvement of extracellular signal-regulated kinase and NF-kappaB. J Biol Chem 2000;275(7):
494955.
[118] Rioux N, Castonguay A. The induction of cyclooxygenase-1 by a tobacco carcinogen in U937 human
macrophages is correlated to the activation of NFkappaB. Carcinogenesis 2000;21(9):174551.
[119] Li N, Karin M. Ionising radiation and short wavelength UV activate NF-kappaB through two

ORAL CANCER PREVENTION

[120]

[121]

[122]
[123]

[124]

[125]

distinct mechanisms. Proc Natl Acad Sci U S A

1998;95(22):130127.
Loercher A, Lee TL, Ricker JL, et al. Nuclear
factor-kB is an important modulator of the altered
gene expression prole and malignant phenotype in
squamous cell carcinoma. Cancer Res 2004;64:
651123.
Harris CC, Hollstein M. Clinical implications of
the p53 tumor-suppressor gene. N Engl J Med
1993;329:131827.
Hartwell LH, Kastan MB. Cell cycle control and
cancer. Science 1994;266:18218.
Nees M, Homann N, Discher H, et al. Expression
of mutated p53 occurs in tumor-distant epithelial
head and neck cancer patients: a possible molecular
basis for the development of multiple tumors. Cancer Res 1993;53:418996.
Lee JJ, Hong WK, Hittelman WN, et al. Predicting
cancer development in oral leukoplakia: ten years
of translational research. Clin Cancer Res 2000;6:
170210.
Chung KY, Mukhopadhyay T, Kim J, et al. Discordant p53 gene mutations in primary head and
neck cancers and corresponding second primary
cancers of the upper aerodigestive tract. Clin Cancer Res 1993;53:167683.

511

[126] Zariwala M, Schmid S, Pfaltz M, et al. p53 gene

mutations in oropharyngeal carcinomas: a comparison of solitary and multiple primary tumors and
lymph-node metastases. Int J Cancer 1994;56:
80711.
[127] Gasparotto D, Maestro R, Brazan L, et al. Recurrences and second primary tumours in the head and
neck region: dierentiation by p53 mutation analysis. Ann Oncol 1995;6:9339.
[128] Brennan JA, Mao L, Hruban RH, et al. Molecular
assessment of histopathological staging in squamous-cell carcinoma of the head and neck.
N Engl J Med 1995;332(7):42935.
[129] Wang Z, Fuentes CF, Shapshay SM. Antiangiogenic and chemopreventive activities of celecoxib
in oral carcinoma cell. Laryngoscope 2002;112:
83943.
[130] Sood S, Shi SJ, Yang CS, et al. Selection of topically applied non-steroidal anti-inammatory
drugs for oral cancer chemoprevention. Oral Oncology 2005;41:5627.
[131] Li N, Sood S, Wang S, et al. Overexpression of
cyclooxygenase 2 and 5-lipoxygenase in hamster
and human oral cancer and chemopreventive effects of celecoxib and zileuton. Clin Cancer Res
2005;11(5):20838.

Oral Maxillofacial Surg Clin N Am 18 (2006) 513519

Verrucous Carcinoma
Rocco R. Addante, DMD, MDa,*, Samuel J. McKenna, DDS, MDb
a

Section of Maxillofacial Surgery, Dartmouth-Hitchcock Medical Center, One Medical Center Drive,
Lebanon, NH 037560001, USA
b
Oral and Maxillofacial Surgery, Vanderbilt University School of Medicine, 1623 The Vanderbilt Clinic,
Nashville, TN 372325225, USA

Verrucous carcinoma (VC) is a distinct lowgrade variant of squamous cell carcinoma of

the oral cavity. It was rst identied as a clinical
and histologic entity by Ackerman [1] in 1948 in
a study of 31 patients. He postulated that this tumor should be considered separately from other
epidermoid tumors, because patients with these
tumors, even when extensive, had a better prognosis. Most of the patients he studied were elderly
men, many of whom gave a history of chewing tobacco use and many of whom were noted to have
poorly tting dentures, poor oral hygiene, or carious and jagged teeth. Consequently, many of
these associations have been cited in the literature
in follow-up case studies and reports, although
not entirely substantiated by controlled epidemiologic investigations. For example, although most
studies demonstrate a high incidence of mucosal
VC with the use of inhaled and smokeless tobacco
[2], 16% to 51% of oral VCs are found in persons
without tobacco habits [3].
The similar clinical and histologic appearance
of VC aecting the upper aerodigestive tract,
genitalia (condyloma acuminatum), and extremity
skin (carcinoma cuniculatum) raises the possibility of a common cause [4]. Specically, human
papilloma virus (HPV) has been implicated as
a cause for verrucoid-appearing lesions at each
of these sites. For example, HPV has been detected in 45% to 85% of laryngeal VCs by polymerase chain reaction [5]. In contrast, ndings
from studies of HPV and oral squamous cell cancer have varied widely, with frequencies of HPV
* Corresponding author.
E-mail address: rocco.r.addante@hitchcock.org
(R.R. Addante).

identication ranging from 0% to 78% [6]. A related nding in some cases of verrucous lesions
is alteration of p53 tumor suppression gene expression and alteration of the keratinocyte cell cycle [7,8]. The role of HPV infection and p53 gene
expression in the pathogenesis of verrucous oral
lesions and oral carcinoma has yet to be dened.
VC is a rare tumor, representing only 3% to
4% of all oral carcinomas, with an annual incidence of one to three cases for every 1 million
persons [9]. Among the 411,534 cases of head and
neck carcinoma recorded in the National Cancer
Data Bank (NCDB) between 1985 and 1996,
2350 (0.6%) were identied as VC. Ninety-one
percent of all cases arose in the oral cavity
(56%) and larynx (35%).
VC is a disease of the elderly, with most cases
being observed in patients in their sixth decade or
older. The median age at diagnosis is 69.0 years.
Within the oral cavity, the most commonly involved locations are the buccal mucosa and
gingiva. Men have a higher proportion of laryngeal tumors, whereas women have a higher
proportion of oral cavity tumors [10]. This is consistent with the ndings of McCoy and Waldron
[11] in a study of 49 cases. When occurring in the
larynx, the vocal cords are the preferred location.
VC has also been described in genitalia, nasal passages, and esophagus [12].
Clinical features
VC is characterized by a slow-growing, painless, broad-based verrucous or wart-like papillary
lesion. The lesion lacks ulceration and has a propensity for local invasion rather than metastatic
spread. Clinically, the lesion is often well

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.007

oralmaxsurgery.theclinics.com

514

ADDANTE & MCKENNA

Fig. 1. (A) VC arising from the right mandibular ridge in a 50-year-old man with a history of moderate smoking and
ethanol abuse. (B) Occlusal view showing well-demarcated margins. (C) Radiograph demonstrating underlying bony resorption indicative of osseous invasion.

demarcated (Fig. 1) and can exhibit a white, red

and white, or red appearance, depending on the
amount of surface keratinization (Fig. 2) [13].
With progression, it can assume an exophytic
and fungating quality (Figs. 3 and 4). It bluntly
invades underlying soft tissues and may become
xed to the periosteum with subsequent invasion
of bone (see Fig. 1). Despite a deceptively benign
microscopic appearance, VC can extensively
inltrate and destroy adjacent tissue, including

Fig. 2. VC originating from the right buccal mucosa in

an 84-year-old woman demonstrating a mixed red and
white presentation.

muscle, cartilage, and bone (Fig. 5). VC may present in a eld of generalized leukoplakia (Figs. 6
and 7). Although a patient may present with enlarged cervical lymph nodes, such lymphadenopathy is usually inammatory or reactive in nature.
In contrast to oral squamous cell carcinoma, VC
does not metastasize regionally or distantly.
The macroscopic papillary and wart-like surface features of VC are similar to those seen in
verrucous hyperplasia (VH) (Fig. 8) and proliferative verrucous leukoplakia (PVL) (see Fig. 7).
Although current terminology related to the

Fig. 3. VC arising from the gingiva on the lingual aspect

of the anterior mandible in a 69-year-old woman.

VERRUCOUS CARCINOMA

515

Fig. 4. VC in a 75-year-old woman originating in the buccal mucosa with extension to the oral commissure (A), maxilla
(B), lip (C), and tongue. Over 7 years, this woman underwent multiple surgical excisions, including a maxillectomy and
(D) extensive nasolabial and cervical ap perioral reconstruction. Tongue involvement progressed to squamous cell
carcinoma with pulmonary metastases and death.

histologic dierentiation of these lesions is confusing and controversial, microscopically, these lesions all share the feature of atypical epithelial
hyperplasia, with varying potential for progression to dysplasia and carcinoma. In particular,
VH may represent a stage in a continuum, if not
a precursor of VC, and thus must be thoroughly

evaluated. Indeed, VC often arises in an area or

background of VH.
Described in 1985 by Hansen and colleagues
[14], PVL is considered to be a progressive oral
leukoplakia. In contrast to VC, PVL tends not
to be associated with tobacco use. PVL, like VC,
is a chronic indolent process typically beginning

Fig. 5. (A) VC of the maxilla in a 74-year-old woman with extension into the vestibule. (B) CT image demonstrating a
2-cm  4-cm right maxillary mass with destruction of the maxillary alveolus, although sparing the sinus and the nasal
cavity.

516

ADDANTE & MCKENNA

squamous cell carcinoma after 11 years. The diffuse nature of this lesion makes complete surgical
excision challenging, if not impossible, accounting
for frequent recurrence (see Fig. 7). In fact, PVL is
typically resistant to all treatment modalities, including surgery, radiation, chemotherapy, and laser excision [17]. VC arising in the setting of PVL
is most commonly associated with congruent invasive squamous cell carcinoma [18].
Microscopic features
Fig. 6. VC arising in the right oor of the mouth in a
61-year-old woman with extensive generalized oral
leukoplakia.

as a nondysplastic keratosis, which, through lateral spread, develops into large diuse keratoses.
PVL is an evolving dysplasia with a high frequency of progression to squamous cell carcinoma [15]. Silverman and Gorsky [16] noted
that 70% of 54 patients with PVL developed

Histologically, VC is broadly based and invasive, with plump papillary invaginations of

thickened and infolding epithelium that lack the
usual cytologic criteria of malignancy [12]. The tumor cells are enlarged but do not show pleomorphism. There is no signicant cellular atypia and
minimal individual cell dysplasia. Mitoses are
rare; when observed, they are located in a suprabasal position immediately above the basal cell
layer, where they normally occur. Throughout
the lesion, the basement membrane remains

Fig. 7. VC of the right hard palate in a 49-year-old nonsmoking woman with a history of proliferative verrucous
leukoplakia (A), involving the buccal mucosa (B), gingiva (C), oor of the mouth (D), and tongue (E).

VERRUCOUS CARCINOMA

517

features of the border between the lesion and

normal tissue, the biopsy should include some
adjacent normal-appearing tissue.
Treatment

Fig. 8. Asymptomatic 1.5-cm VH of the palate in 71year-old woman with a smoking history of 25 packs
per year.

intact; despite the invasive nature of the tumor,

there is an absence of cells violating the basement
membrane as seen in other types of squamous cell
carcinoma. The inferior margin is often described
as pushing rather than invading, and this blunt
margin characteristically invokes a mononuclear
inammatory reaction in the stroma immediately
adjacent to the advancing margin. The inammatory cells are typically plasma cells and lymphocytes. Perineural invasion and vascular invasion
are not features of VC. There is usually an abrupt
transition from VC to adjacent normal tissue. The
features of the border zone have been described in
detail by Jacobson and Shear [19], who noted that
inwardly projecting epithelial folds often cause
a margin of normal epithelium to retract down
into the underlying connective tissue. This feature
may be helpful in distinguishing VC from reactive
inammatory epithelial hyperplasia.
Occasionally, lesions with the general morphology of VC may contain foci of invasive
squamous cell carcinoma of varying grade. These
lesions have been designated hybrid tumors [20].
Although there is no universally agreed on
amount of squamous cell carcinoma within VC
to qualify it as a hybrid tumor, the best course
of action is to treat any VC with a component
of squamous cell carcinoma as invasive squamous
cell carcinoma [21].
Because of the scarcity of cytologic features of
dysplasia, the presence of background inammation, clinical overlap with PVL, and risk of
progression to squamous cell carcinoma, it is
important that adequate tissue be obtained at
the time of biopsy of suspected VC. A generous
portion of the lesion, including the submucosa,
should be submitted. Because of the characteristic

Since VC was rst reported by Ackerman [1],

there has been ongoing discussion regarding the
preferred treatment. Treatment modalities have
included surgery, radiation therapy, chemotherapy, cryotherapy, laser therapy, photodynamic
therapy, and treatment with recombinant a-interferon. Surgical excision remains the preferred treatment for the primary lesion [22].
Transformation of VC to anaplastic or poorly
dierentiated squamous cell carcinoma has been
described after radiation therapy. This phenomenon was rst reported by Perez and coworkers [23]
in 1966 when they noted its occurrence in 8 of 17 patients with VC. Anaplastic transformation resulting in a more aggressive behavior has been shown
to occur as a result of chemotherapy, cryosurgery,
laser surgery, and even after multiple conventional
operations [24]. In an eort to examine the ecacy
of radiation therapy as a primary modality in the
treatment of VC, Ferlito and colleagues [25] critically evaluated 148 patients with histologically conrmed VC managed with radiation therapy. Most
of these patients experienced treatment failure
with persistence or recurrence and with a local control rate of only 43.2% (64 of 148 patients). These
ndings corroborate the observation that radiation
therapy is less eective than surgery, because VC,
although not radioresistant, is less radiosensitive
than squamous cell carcinoma. Of the 148 cases
treated with primary radiotherapy, 10 cases
(6.7%) of true anaplastic transformation were identied. The prognosis of anaplastic transformation
after treatment with radiation therapy is dismal.
A review of NCDB data with specic attention
to localized cases of oral VC demonstrated 85%
and 42% 5-year survival rates after surgery and
radiation therapy, respectively [10]. Further, improved outcomes were not seen with the addition
of radiation therapy as an adjuvant to surgery. Although these data clearly suggest better outcomes
with surgical intervention, denitive statements
cannot be made without the benet of randomized
trials, which are dicult to conduct in these rare
tumors. A degree of treatment selection bias also
has to be factored into the analysis of treatment outcomes, because irradiation alone is often chosen for
patients with extensive disease that would be dicult to resect primarily or in patients with signicant

518

ADDANTE & MCKENNA

comorbidities who would tend to have lower 5-year

survival data [10]. Irrespective of treatment, there is
a 26% to 57% incidence of recurrence. Therefore,
close long-term follow-up is necessary.
There is no consensus with respect to the size of
surgical margins necessary to decrease the risk of
recurrence. Certainly with extensive gingival involvement, consideration should be given to segmental or marginal osseous resection. Resection of
VC involving tooth-bearing mucosa must include
adjacent alveolus to ensure that any periodontal
extension, including alveolar involvement, is adequately treated. A review of eight cases of gingival
VC demonstrated histologic evidence of bony invasion in three patients (37.5%) [26].
Because VH has many histopathologic and
clinical features of VC and probably represents
a precursor to VC, there should be consideration
for similar surgical management. Procrastination
in hopes of involution or spontaneous reversal of
the process can lead to loss of control and
development of squamous cell carcinoma [27].
Similarly, conservative excision misguided by the
indolent and chronic nature of VC, especially
when tooth-bearing mucosa is involved, predisposed to treatment failure and recurrence.
Summary
VC is an indolent low-grade carcinoma with
a deceptively benign histologic appearance. Notwithstanding the potential for slow growth and
lack of metastases, the pattern of local invasion
and the risk of progression to squamous cell
carcinoma dictate an aggressive approach with
complete surgical excision. VC involving toothbearing mucosa should especially be treated with
excision of underlying alveolus. Unfortunately,
recurrence is common, and long-term vigilant
surveillance is necessary.
References
[1] Ackerman L. Verrucous carcinoma of the oral
cavity. Surgery 1948;23:6708.
[2] Wray A, McGuirt WF. Smokeless tobacco usage associated with oral carcinoma: incidence, treatment,
outcome. Arch Otolaryngol Head Neck Surg 1993;
119:92933.
[3] Neville BW, Damm DD, Allen CM, et al. Oral and
maxillofacial pathology. Philadelphia: WB Saunders; 1995. p. 3046.
[4] Prioleau PG, Santa Cruz DJ, Meyer JS, et al. Verrucous carcinoma: a light and electron microscopic,
autoradiographic and immunouorescence study.
Cancer 1980;45:284957.

[5] Fliss DM, Noble-Topam SE, McLachlin CM, et al.

Laryngeal verrucous carcinoma: a clinicopathologic
study and detection of human papillomavirus using
polymerase chain reaction. Laryngoscope 1994;104:
14652.
[6] Palefsky JM, Silverman S Jr, Abdel-Salaam M, et al.
Association between proliferative verrucous
leukoplakia with human papilloma virus type 16.
J Oral Pathol Med 1995;24:1937.
[7] Gopalakrishnan R, Weghorst CM, Lehman TA,
et al. Mutated and wild-type p53 expression and
HPV integration in proliferative verrucous leukoplakia and oral squamous cell carcinoma. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1997;
83(4):4717.
[8] Fettig A, Pogrel A, Silverman S, et al. Proliferative
verrucous leukoplakia of the gingiva. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 2000;
90(6):72330.
[9] Bouquot JE. Oral verrucous carcinoma: incidence
in two US populations. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 1998;86:31824.
[10] Koch BD, Trask DK, Homan HT, et al. National
survey of head and neck verrucous carcinoma. Patterns of presentation, care and outcome. Cancer
2001;92:11020.
[11] McCoy JM, Waldron CA. Verrucous carcinoma of
the oral cavity. Oral Surg 1981;52:6239.
[12] Batsakis JG, Hybels R, Crissman JD, et al. The pathology of head and neck tumors: verrucous
carcinoma. Part 15. Head Neck Surg 1982;5:2938.
[13] Ferlito A, Recher G. Ackermans tumor (verrucous
carcinoma) of the larynx. A clinicopathologic study
of 77 cases. Cancer 1980;46:161730.
[14] Hansen LS, Olson JA, Silverman S. Proliferative
verrucous leukoplakia. A long-term study of thirty
patients. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 1985;60:28598.
[15] Murrah VA, Batsakis JG. Proliferative verrucous
leukoplakia and verrucous hyperplasia. Ann Otol
Rhinol Laryngol 1994;103:6603.
[16] Silverman S, Gorsky M. Proliferative verrucous leukoplakia: a follow-up study of 54 cases. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1997;
84(2):1547.
[17] Gopalakrishnan R, Weghorst CM, Lehman RA,
et al. Mutated and wild-type p53 expression and
HPV integration in proliferative leukoplakia and
oral squamous cell carcinoma. Oral Surg Oral Med
Oral Pathol Oral Radiol Endod 1997;83:4717.
[18] Vigilante CE, Quinn PD, Alawi R. Proliferative verrucous leukoplakia: report of a case with characteristic long-term progression. J Oral Maxillofac Surg
2003;61:62631.
[19] Jacobson S, Shear M. Verrucous carcinoma of the
mouth. J Oral Pathol 1972;1:6675.
[20] Medina JE, Dichtel MAJW, Luna MA. Verrucous
squamous carcinomas of the oral cavity. Arch Otolaryngol 1984;110:43740.

VERRUCOUS CARCINOMA

[21] Slootweg PJ, Richardson M. Squamous cell carcinoma

of the upper aerodigestive system. In: Gnepp DR, editor. Diagnostic surgical pathology of the head and
neck. Philadelphia: WB Saunders; 2001. p. 1978.
[22] Spiro RH. Verrucous carcinoma, then and now. Am
J Surg 1998;175:3937.
[23] Perez CA, Kraus FT, Evans JC, et al. Anaplastic
transformation in verrucous carcinoma of the oral
cavity after radiation therapy. Radiology 1966;86:
10815.
[24] Yoshimura Y, Mishima K, Obara S. Treatment
modalities for oral verrucous carcinomas and their

519

outcomes: contribution of radiation therapy and

chemotherapy. Int J Clin Oncol 2001;6:192200.
[25] Ferlito A, Rinaldo A, Mannara GM. Is primary radiotherapy an appropriate option for the treatment
of verrucous carcinoma of the head and neck? J Laryngol Otol 1998;112(2):1329.
[26] Ogawa A, Fukuta Y, Nakajima T, et al. Treatment
results of oral verrucous carcinoma and its biological
behavior. Oral Oncol 2004;40:7937.
[27] Murray VA, Batsakis JG. Proliferative verrucous
leukoplakia and verrucous hyperplasia. Ann Otol
Rhinol Laryngol 1994;103:6603.

Oral Maxillofacial Surg Clin N Am 18 (2006) 521531

Special Considerations with Floor of Mouth

and Tongue Cancer
Oneida A. Arosarena, MDa,*, Matthew Madsen, BSb,
Richard Haug, DDSc
a

Department of Otolaryngology, Temple University, Philadelphia, PA, USA

b
College of Dentistry, University of Kentucky, Lexington, KY, USA
c
Department of Oral and Maxillofacial Surgery, College of Dentistry, University of Kentucky, Lexington, KY, USA

Nearly 24,000 new cases of oral cavity cancer

are diagnosed yearly, and oral cavity cancer
accounts for 14% to 24% of all head and neck
malignancies in the United States [1,2]. The incidence of oral cavity cancer has decreased considerably in the past three decades (from 3.6 to 2.7
per 100,000 population), and these malignancies
are being diagnosed at earlier stages because of
more widespread dental professional screening
[1]. Five-year survival rates have not improved
meaningfully from the early 1970s, however, despite a shift toward the use of combined treatment
(surgery and postoperative radiotherapy) as opposed to single modality therapy [1,3]. In fact, 5year survival rates for patients with regional stage
oral cavity cancer have decreased from 49.2% to
43.8% [1]. Floor of mouth and tongue malignancies comprise most oral cavity malignancies. Floor
of mouth cancers comprise 9.3% to 16.8% and
mobile tongue cancers comprise 20% to 35.6%
of oral cavity cancers [47]. Treatment of malignancies of the anterior oral cavity can result in signicant functional impairment [8]. Patients with
epidermoid malignancies of the oral tongue and
oor of mouth also are at increased risk for eld
cancerization and development of second or multiple primary cancers in the upper aerodigestive
tract [911]. Squamous cell carcinoma is the
most common malignancy of the oral cavity, accounting for 90% of all oral cavity tumors [12].

* Corresponding author.
E-mail address: oneida.arosarena@temple.edu
(O.A. Arosarena).

Further references to oral cavity cancer in this article imply epidermoid carcinoma unless otherwise
specied.
Treatment modalities
Surgery and radiotherapy
Although there is evidence in the literature that
surgery and denitive radiotherapy are equally
eective for local control in early oral cavity
cancer, it is widely accepted that surgical resection
is the treatment of choice for T1 and T2 oral cavity
malignancies [1315]. In a four-decade study of
more than 1500 patients with oral cavity cancer,
Carvalho and colleagues [14] found that patients
with T1 and T2 oral cavity cancer had the highest
rate of local recurrence when treated by radiotherapy alone (32.8%) as compared with surgery alone (13.9%), chemoradiation therapy
(14.3%), and combined surgery and radiotherapy
(15.8%). In a study of 161 patients with head and
neck cancer treated with preoperative radiotherapy, oral cavity cancer was found to be an independent risk factor for locoregional recurrence
[16]. Similarly, Yao and colleagues [17] evaluated
the outcomes of intensity-modulated radiotherapy
given for denitive treatment and postoperatively
in the management of head and neck cancer.
Patients with oropharyngeal cancer were found
to have better responses than patients with oral
cavity and laryngeal cancer. Given the relative locoregional aggressiveness of oral cavity epidermoid malignancies, ipsilateral neck dissection
should accompany resection of the primary tumor

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.005

oralmaxsurgery.theclinics.com

522

AROSARENA

for T2 and greater disease. If the tumor is within

1 cm of the midline, however, bilateral neck dissections should be performed [18].
For T3 and T4 oral cavity cancer, combined modality therapy (surgery and postoperative
radiotherapy) is the standard of care [14]. However,
Carvalho and colleagues [14] found nearly equal
local recurrence rates with surgery, combined surgery and radiotherapy, and chemoradiation therapy
(26.3%, 28.0%, and 29.7%, respectively). Radiotherapy alone had the highest local recurrence rate
(41.5%) for T3 and T4 disease. Radiotherapy and
chemoradiation therapy had unacceptably high
rates of locoregional recurrence for advanced disease (22.0% and 31.7%, respectively) compared
with surgery alone and combined modalities involving surgery (5.7% to 7.9%).

Role of chemotherapy
Chemotherapy has been used with radiation
therapy, surgery, or combined modality therapy to
attempt improved survival in head and neck cancer
patients. Induction or neoadjuvant treatment
before denitive therapy, concurrent chemoradiation therapy, and adjuvant chemotherapy after
denitive surgery or radiotherapy have been studied [1921]. None of the trials testing induction
chemotherapy have demonstrated a survival
advantage, and two have shown decreased survival
in patients receiving induction chemotherapy
[19,2123].
Approximately 70% of patients have some
response to induction chemotherapy, and 20%
can be expected to have a complete response [21].
A randomized, controlled trial that compared induction chemotherapy followed by surgery with
surgery alone for resectable oral cavity cancer
concluded that chemotherapy did not improve
survival but could limit the extent of surgical resection needed for locoregional disease control
[24]. Similarly, in a series of 33 patients with stage
III and IV oral cavity cancer treated with neoadjuvant cisplatin-based chemotherapy followed by
surgical resection and radiotherapy, Ruggeri and
colleagues [25] found a clinically signicant increase in disease-free survival only in patients
who had complete clinical and pathologic responses to induction chemotherapy. In another
study that compared induction chemotherapy
with standard surgery and postoperative radiotherapy for oor of mouth and tongue epidermoid
carcinomas, Nemeth and colleagues [26] found

et al

a survival advantage in patients treated with cytostatic chemotherapy.

Unlike malignancies in other head and neck
sites, such as the larynx and hypopharynx, combined chemoradiation therapy for oral cavity
malignancies has not been shown to aect disease-free survival signicantly in most clinical
trials and in meta-analyses of these trials. Despite
encouraging pilot data that chemotherapy could
act as a sensitizer for radiotherapy, most trials
have not shown a survival advantage [21].
The role of postoperative chemotherapy for
oral cavity cancer has not been dened clearly
because few trials have tested this treatment
method. The rationale for postoperative chemotherapy is based on the hypothesis of rst order
kill (i.e., that chemotherapy is more eective in
killing the relatively small number of residual
tumor cells that remain after surgery in comparison to the attempted eradication of large tumor
bulk by neoadjuvant chemotherapy). Kovacs and
colleagues [27] demonstrated a signicant 5-year
survival advantage in patients with oral cavity
cancer treated with radical surgery and postoperative chemotherapy in comparison to patients
treated with radical surgery alone.
Several reasons have been proposed for the
relative ineectiveness of chemotherapy for head
and neck cancer, including low growth fractions
in squamous cell carcinomas and insensitivity of
squamous cells to chemotherapy. Trials testing
adjuvant chemotherapy may be limited by decreased vascularity to the tumor bed caused by
surgery and radiotherapy, which compromise
chemotherapy delivery. Researchers also have
postulated that induction chemotherapy may delay denitive treatment or induce biochemical
resistance, which limits the eectiveness of adjuvant chemotherapy. Probably most importantly,
in almost every head and neck chemotherapy trial,
investigators report poor compliance or tolerance
because of the toxicity of chemotherapy, which
limits the number of cycles given. Although
animal models and some clinical trials support
adjuvant chemotherapy for head and neck cancer,
much needs to be learned about optimal patient
selection, and less toxic but eective treatment
protocols must be developed [21,28].
Floor of mouth cancer
Cancers of the oor of mouth are more likely
to be multifocal than other oral cavity malignancies [29]. Because the loose connective and

FLOOR OF MOUTH AND TONGUE CANCER

muscular tissues of the submental and submandibular spaces provide no barrier to spread, advanced cancers often present with extension into
the soft tissues of the neck [13,29]. One third to
one half of patients with oor of mouth cancer
present with regional metastatic disease (10% of
T1, 30% of T2, and 65% of T3 tumors)
[13,29,30]. The incidence of occult metastasis is
21% with T1 tumors, 40% with T2 tumors, and
62% to 70% with T3-4 tumors [29,31].
Assessing tumor thickness is essential in the
management of oor of mouth cancer, although it
can be dicult to determine clinically or by CT
because of reactive inammatory induration in
the soft tissues of the submental and submandibular triangles [13]. Intraoral ultrasonography has
been shown to be superior to CT and MRI for
measuring tumor thickness [32]. Tumor thickness
determines the extent of primary tumor resection
and management of occult neck disease in early
oor of mouth cancer; its estimation is critical
for locoregional control.
Supercial T1 and T2 tumors without extension
into the submandibular and submental spaces can
be managed by resection, including the submandibular and sublingual glands with their associated lymphatics [33]. Reconstruction of these
small defects can be accomplished with split-thickness skin grafting or pedicled nasolabial aps
[34,35]. Floor of mouth tumors frequently extend
onto the ventral tongue, however, and skin grafting of defects that involve the ventral tongue can
result in tethering and limitation of tongue mobility. The use of local cervical myocutaneous aps,
including the platysma and sternocleidomastoid
aps, has been advocated for small defects in selected patients, with reported success rates between 89% and 92%. Necrosis of the skin
paddle is a frequent complication with these aps
and has limited their use, however [3639]. The
use of a central or paramedian island tongue
ap also has been described for repair of defects
in T1 and T2 oor of mouth malignancies, without
compromise of speech or deglutition [40].
More extensive (T3 and T4) oor of mouth tumors are best managed by a pull-through approach, in which the contents of the
submandibular and submental spaces are delivered through the neck in continuity with the
neck dissection specimen. These larger defects
are best reconstructed with microvascular free tissue transfer, although a regional ap, such as the
pectoralis major myocutaneous ap, is a good option for patients whose medical conditions do not

523

permit microvascular reconstruction [41]. The

type of microvascular free tissue transfer needed
depends greatly on whether there is mandibular
involvement with tumor.
Floor of mouth tumors that extend to the
mandibular gingiva have associated mandibular
invasion in 7% of T1, 55% of T2, and 63% of T3
lesions [29]. Harrison and Fass [13] pointed out
that early anterior oral cavity carcinomas can
cause pressure saucerization without actual bone
invasion, complicating the assessment of bony involvement. In a study of 34 patients, most of
whom had alveolar ridge squamous cell carcinoma (2 had retromolar trigone malignancies),
Goerres and colleagues [42] found that coregistration of CT with single photon emission CT with
technetium 99m (99mTc) had the least accuracy
(88%) for predicting bony invasion as compared
with coregistration of 2-[F-18]-uoro-2-deoxy-Dglucose positron emission tomography [PET]
(97%), and contrast-enhanced CT scanning alone
(94%). PET/CT had the highest sensitivity
(100%), and contrast enhanced CT alone had
the highest specicity (100%) for mandibular involvement. This is in contrast to a study by Imola
and colleagues [43] that found that 99mTc single
photon emission CT had greater accuracy and
sensitivity but less specicity than CT scanning
and Panorex radiographs for detecting mandibular invasion by oral squamous cell carcinoma.
Floor of mouth tumors with frank bony invasion, and T3 and T4 tumors are appropriately
managed with surgery and postoperative radiotherapy because bone invasion renders cure by radiotherapy unlikely [13]. Tumors with periosteal
involvement can be resected adequately with marginal mandibulectomy, whereas frank cortical erosion
requires
segmental
mandibulectomy.
Anterior segmental mandibular defects are best
reconstructed with bone-containing free aps,
such as the bular or scapular aps, which provide thin, pliable soft tissue paddles for reconstruction of the soft tissues of the oor of mouth
[4446]. The iliac crest free ap also can be used
for segmental mandibular defects, but the soft tissue paddle associated with this ap can be excessively bulky, limiting speech and deglutition. The
vascular geometry of inferior epigastric pedicle
limits the arc of rotation of the iliac crest free
ap skin paddle with respect to the iliac crest
bone [47]. Use of the bone-containing radial forearm free ap is typically discouraged because of
the signicant risk of radial fracture that results
in substantial upper extremity disability. The

524

AROSARENA

radial forearm fasciocutaneous ap is commonly

used for reconstruction of large oor of mouth defects with mandibular continuity. In patients with
lateral mandibular defects who are poor candidates for reconstruction with bone-containing
free aps, use of a reconstruction plate with a local
or free myocutaneous ap (such as the pectoralis
major ap) remains a good option [48,49].
Floor of mouth tumors that extend into the
root of the tongue require partial or complete
glossectomy for control depending on extent.
Defects that result from resection of these tumors
are best reconstructed with myocutaneous free
aps, such as the rectus abdominus or latissimus
dorsi free aps to provide the bulk needed for
swallowing and speech [50]. Local myocutaneous
aps also can be used to reconstruct these defects
in patients whose comorbidities do not allow free
tissue transfer. Skin grafting the oor of mouth
component of these bulky muscle-containing aps
improves tongue mobility [51].
Denitive radiotherapy is a treatment option
for patients who refuse surgery or are poor
surgical candidates. Although equal local control
rates are reported with surgery and radiotherapy
for oor of mouth cancers and improved functional outcome has been reported with radiotherapy for tumors of the oor of mouth and oral
tongue, lesions that abut or are tethered to the
periosteum are best managed with surgery so as to
prevent osteoradionecrosis of the mandible
[13,52,53]. When brachytherapy, external beam
radiotherapy, or a combination of both is used
to treat oor of mouth cancers, local control rates
of 88% to 98% for T1 tumors, 47% to 93% for T2
tumors, and 51% to 86% for T3 tumors have been
reported [13,29,5457].

Oral tongue cancer

Tongue malignancies usually arise from the
lateral aspect of the tongue (47%), whereas 20%
occur on the anterior third and 4% occur on the
dorsum [13,29]. Cervical metastases occur more
commonly with oral tongue cancer than with cancer at other sites in the oral cavity [29]. Thirty-ve
percent of oral tongue cancers present with cervical metastases [13,58]. Oral tongue malignancies
can become extensive before becoming symptomatic because the tongue musculature does not provide a barrier to tumor growth. Pain, which
indicates lingual or glossopharyngeal nerve involvement, and dysphagia, which is related to

et al

hypoglossal nerve invasion or tumor bulk, are relatively late signs [29].
Early oral tongue cancers can be managed
successfully with brachytherapy, external beam
radiotherapy, or a combination of the two.
Signicant lingual edema can occur with brachytherapy, necessitating tracheotomy [29]. Control
rates with radiotherapy reported are 79% to
100% for T1 lesions, 61% to 91% for T2 lesions,
and 68% for T3 lesions [57,5961]. Similar control
rates (85% for T1 tumors, 77% for T2 tumors, and
50% for T3 tumors) have been reported with glossectomy alone [62]. Given the signicant local failure rates with unimodality treatment for advanced
disease, it is recommended that T3 and T4 tumors
be treated with surgery followed by radiotherapy
[13]. Only 10% to 15% of patients with local recurrence are amenable to surgical salvage, and
of those patients, disease is controlled in only
30% [29,63]. The risk of regional failure is significant for T2 and greater disease. In these patients,
radiotherapy or selective neck dissection is recommended for the N0 neck [13,29]. In a study that
compared outcomes in patients with oral tongue
and oor of mouth tumors treated with surgery
and postoperative radiotherapy, the local failure
rate for patients with oral tongue cancer was
38% compared with 11% for oor of mouth cancers even after controlling for clinical stage. Similarly, the median survival after recurrence was
9 months for tongue malignancies and 40 months
for oor of mouth malignancies, which demonstrates the relative local aggressiveness of mobile
tongue malignancies. Patients with oor of mouth
tumors were more likely to have distant metastases (50%) than patients with tongue lesions
(21%), which may reect the longer survival in patients with oor of mouth malignancies [64].
Large tongue malignances that extend into the
oor of mouth and abut or invade the mandible
are managed as described for oor of mouth cancers with mandibular involvement.
Reconstruction of oral tongue defects is challenging because of the importance of maintaining
mobility of the tongue. Fortunately, patients are
often able to compensate with postoperative
speech and swallowing therapy and can have
acceptable speech and swallowing function, with
defects extending up to half of the anterior
tongue. Defects that involve up to half of the
mobile tongue can be reconstructed successfully
with a split-thickness skin graft or a pedicled
nasolabial ap, which provides more bulk than
the split-thickness skin graft [35]. Skin grafts that

FLOOR OF MOUTH AND TONGUE CANCER

extend from the oor of mouth onto the tongue can

result in tethering of the tongue, however. Larger
lingual defects are best reconstructed with bulky,
muscle-containing free aps or regional aps, although folded fasciocutaneous aps can be used
successfully [41,65,66]. If total glossectomy is
needed for local control in patients with poor pulmonary function, total laryngectomy should be
considered to separate the digestive and respiratory tracts permanently and prevent aspiration.
Assessment of risk for recurrence and metastasis
Survival can vary signicantly with early stage
oral tongue cancers from 60% to more than 90%
[6770]. As with other head and neck cancers, the
status of regional lymph nodes is the most important prognostic factor in oral cavity cancer. In
general, the presence of nodal metastases decreases cure and survival rates by half
[14,29,53,7173]. Several studies have demonstrated that the locoregional recurrence rate of
oral cavity cancer is worse than that of head and
neck cancer at other sites [16]. For these reasons,
interest has grown in methods of detection of micrometastatic disease, or prediction of metastatic
potential for early stage oral cavity cancer.
Characteristics of the tumor at the primary site
Indicators of local tumor aggressiveness have
not been shown clearly to have an inuence on
metastatic risk in oral tongue and oor of mouth
cancer. Although many studies have indicated
that histologic dierentiation, anatomic location,
status of surgical margins, and vascular invasion
correlate with increased risk for cervical metastases [31,53,68,71,7380], Klotch and colleagues [72]
found that vascular invasion, bone invasion, and
even positive margins did not aect survival
from oor of mouth carcinoma. Perineural invasion at the primary site has been shown to correlate with cervical metastases by several
investigators [68,70,74,75,79,81] but not by others
[82]. Similarly, tumor stage has been shown to
have independent prognostic value by some
authors [15,83] and not by others [84].
Several studies have demonstrated tumor
thickness to be predictive of local control, cervical
lymph node involvement, and survival in patients
with T1 and T2 squamous cell carcinomas of the
tongue and oor of mouth. Results of several
studies have varied signicantly in determining
the thickness at which prognosis changes,

525

however. Change in prognosis has been reported

at cut-o thicknesses varying from 1.5 mm to 10
mm [70,76,77,8487]. Woolgar and Scott [75] reported mean thickness of cancers with metastases
varying by anatomic site (7.6 mm in the oor of
mouth and 15.1 mm in the tongue). Several studies have demonstrated changes in prognosis with
tumor thickness between 4 mm and 5 mm
[68,83,88,89]. Close and colleagues [82] found no
correlation between tumor thickness and subsequent development of nodal metastases. Large,
multi-institutional studies are needed to assess
the value of tumor thickness and other clinicopathologic characteristics as indicators of metastatic potential.
Imaging studies
Imaging studies, including CT, MRI, and
ultrasonography, have been used to identify nonpalpable cervical metastatic disease but have
signicant false-positive and false-negative result
rates because of their dependence on lymph node
size and shape for detection of metastases [68].
Palpable lymphadenopathy and central necrosis
of lymph nodes are relatively late indications of
metastatic disease [68,90]. Myers and Wax [71]
found PET to have superior sensitivity, specicity,
positive predictive value, and negative predictive
value in comparison to CT for detection of nodal
metastases from oral cavity cancers with clinically
negative necks. PET had less specicity than CT,
however, and the study had only 23 subjects. It
is known that foci of cancer smaller than 1 cm
are below the limits of resolution of PET [90].
Lymphatic mapping
The management of the clinically negative
neck in stage I and II oral cavity cancer is
controversial. Most surgeons advocate supraomohyoid selective neck dissection or radiotherapy for
treatment of the clinically negative neck rather
than watchful waiting, because the risk of subsequent regional failure is approximately 22%
[14,91,92]. In a study that compared neck dissection with observation in patients with T1N0M0
squamous cell carcinoma of the oral tongue and
oor of mouth, 24% of patients in the observation
group developed neck disease as compared with
a 4% neck recurrence rate in the patients who
had undergone neck dissection. Patients who underwent neck dissection had a 23% greater disease-free survival advantage compared with the

526

AROSARENA

patients whose necks were observed [93]. Micrometastases are found in only 20% to 34% of patients with clinically negative necks, however,
and routine supraomohyoid neck dissection may
expose patients to unnecessary risk and radiotherapy may result in excessive morbidity. Other surgeons view supraomohyoid neck dissection as
purely a staging procedure, citing the need for
postoperative radiotherapy or more extensive
neck dissection if micrometastases are found
[94,95]. Still others cite the possibility of metastases to level IV of the neck with no evidence of disease at levels I-III. These skip metastases have
been reported to occur in 2% to 15.8% of patients
who have oral cavity cancer, and in a study of 253
patients with oral cavity and oropharyngeal malignancies, Woolgar found that these erratic metastases only occurred from primary sites in the
oral tongue [18,29,94,96,97].
Because of the successful use of sentinel lymph
node biopsy for melanoma, interest has grown in
lymphatic mapping of stage I and II oral cavity
cancers to minimize the use of surgery and
treatment costs. Most reports of sentinel lymph
node mapping for oral cavity cancers include few
patients with limited follow-up [94]. Balkissoon
and colleagues [94] recommend lymphoscintigraphy and intraoperative gamma probe use for identication of high-risk nodal basins that are not
encompassed by supraomohyoid neck dissection.
In a study of 48 patients with oral cavity cancer,
Gallegos-Hernandez and colleagues [83] were
able to identify patients with anatomically unexpected lymphatic drainage using preoperative lymphogammagraphy, and intraoperative blue dye
and gamma probe use. In this study, 10.4% of patients demonstrated lymphatic drainage outside
the region typically encompassed by supraomohyoid neck dissection. Similarly, Civantos and colleagues [90] also recommend intraoperative use of
the gamma probe for identication of the sentinel
lymph node, but in their series of 18 patients
with oral cavity cancer, they also identied situations that call into question the basic premise of
lymphoscintigraphy and sentinel lymph node
biopsy.
In four cases, microscopic foci of cancer were
missed at the time of frozen section and subsequently were identied with permanent histology or immunohistochemistry. In two cases,
a sentinel node was identied deep to the omohyoid muscle, whereas more proximal, indurated
nodes had no radioactive uptake. In both of these
cases, pathologic review revealed that the

et al

proximal, indurated, cold lymph nodes were

completely replaced by carcinoma without evidence of normal lymphatic architecture. Although
the authors believed that these proximal pathologic lymph nodes were not enlarged by CT or
preoperative physical examination criteria, it is
obvious that obstruction of lymphatic ow resulted in a new pathway that bypassed the indurated nodes. In yet another case, the sentinel
node did not have metastatic disease, whereas an
adjacent cold node did. Although these cases
challenge the capacity to identify rst echelon
lymph nodes with lymphoscintigraphy, the reported sensitivity of lymphoscintigraphy with
intraoperative gamma probe use was 69% [90].
Gallegos-Hernandez and colleagues [83] reported that the specicity and sensitivity of combined intraoperative gamma probe use and blue
dye were related to the T stage of the tumor.
For T1 tumors, the sensitivity of sentinel node biopsy was 100% and the specicity was 93%. For
T2 lesions, the sensitivity was 73% and the specicity was 100%. Several patients in this study had
more than one sentinel node, and the authors indicated that the presence of two or more sentinel
nodes decreased the likelihood of occult metastases in nonsentinel nodes. In a study of 134 patients
with oral cavity and oropharyngeal cancer, the
sensitivity of sentinel node biopsy for oor of
mouth cancer was 80% compared with 100%
for malignancies at other sites. In this study, the
sentinel node was identied in 86% of oor of
mouth tumors compared with a 97% identication rate for cancers at other sites [98]. This diculty in identifying the sentinel node in oor of
mouth cancers may relate to the fact that the rst
echelon nodes for malignancies in this area may
be within the submandibular or submental spaces
and are dicult to distinguish from the primary
tumor using lymphoscintigraphy or the gamma
probe intraoperatively because of shine artifact
[33,94]. The feasibility of sentinel node biopsy
for oral cavity malignancies has been proved,
but large, multi-institutional studies are needed
to standardize and rene the technique of lymphatic mapping [94].
Extracapsular spread
In patients with metastatic disease to the
regional lymph nodes, extracapsular spread has
been shown to reduce survival signicantly
[78,99,100]. Five-year survival rates for patients
with regional disease limited to the lymph nodes

FLOOR OF MOUTH AND TONGUE CANCER

are 50% to 70%, whereas the presence of extracapsular spread lowers the 5-year survival rate to
25% to 30% [101]. Although extracapsular spread
is correlated with the size of the lymph node, it
frequently occurs in small, N1 lymph nodes and
even in clinically negative necks [29,71,101]. Alvi
and Johnson [101] found extracapsular spread in
49% of head and neck cancers with occult regional metastatic disease.
Distant metastases
Although locoregional failures from oral cavity and oropharyngeal cancer have decreased from
26% to 16% overall in the past 20 years, the
incidence of distant failure has increased from 3%
to 8% [102]. Distant metastases are thought to occur in 8% to 17% of oral cavity cancers. The
probability of distant metastasis correlates with
extent of the primary tumor, presence of regional
lymphatic disease, presence of extracapsular
spread of nodal disease, and locoregional disease
persistence or relapse [12,103]. Patients who present with stage IV oral cavity cancer have been reported to have a distant metastatic rate at the time
of presentation as high as 10% to 40%. In contrast, patients who present with early stage disease
can be expected to have a distant metastatic rate
at initial evaluation of 2% to 3% [12]. Distant metastases most commonly aect the lung (54% to
71%), liver (10% to 36%), bones (15% to 22%),
and mediastinum (3.4% to 23%), with reports of
sporadic metastases to the adrenal gland, kidney,
axilla, heart, and brain [12,104]. For patients who
present with advanced disease, Betka [12] recommends PET, or if PET is not available, CT of
the lungs as a screening tool for distant metastatic
disease before initiation of denitive oncologic
treatment. For other patients with oral cavity cancer, a chest roentgenogram and liver function tests
serve as adequate screening and surveillance tools
for distant metastatic disease.

Summary
Although the incidence of oral cavity cancer
has decreased in the past 30 years, and although
these malignancies are being detected earlier
because of improved screening, the survival rates
from oral cavity malignancies have not improved
despite the benets of combined modality therapy
on locoregional control. Epidermoid carcinomas
of the oral tongue and oor of mouth are the
most common oral cavity malignancies, and they

527

demonstrate locally aggressive behaviors related

to lack of anatomic barriers to spread and a propensity for early cervical metastases. The surgical
management of oral tongue and oor of mouth
cancers can result in signicant functional impairment, and large defects that result from
advanced disease are best reconstructed with
microvascular free tissue transfer. The status of
cervical lymph nodes remains the most important
prognostic factor in cancer survival. Several
methods for predicting metastatic risk and detecting micrometastases show promise. Large multiinstitutional studies are needed to assess the value
of primary tumor characteristics, such as T stage,
tumor dierentiation, vascular embolization,
perineural invasion, and tumor thickness as indicators of metastatic potential. Renements in
lymphoscintigraphy also will permit standardization of this technique for predicting the status of
the oral cavity nodal drainage basin. Similarly,
ne-tuning of PET-CT coregistration eventually
will permit accurate detection of mandibular
invasion and regional and distant metastatic
disease. Current guidelines for oral tongue and
oor of mouth cancer management include surgery for stage I and II disease with elective
treatment of the nodal drainage basin and combined surgery and postoperative radiotherapy for
stage III and IV disease. As less toxic chemotherapeutic protocols become available and patient
selection is optimized, chemotherapy may assume
a more prominent role in the treatment of oral
cavity cancer.

References
[1] Carvalho AL, Nishimoto IN, Califano JA, et al.
Trends in incidence and prognosis for head and
neck cancer in the United States: a site-specic
analysis of the SEER database. Int J Cancer 2005;
114:80616.
[2] Funk GF, Karnell LH, Robinson RA, et al. Presentation, treatment, and outcome of oral cavity cancer: a National Cancer Data Base report. Head
Neck 2002;24:16580.
[3] Shiboski CH, Shiboski SC, Silverman S. Trends in
oral cancer rates in the United States, 19731996.
Community Dent Oral Epidemiol 2000;28(4):
24956.
[4] Hogan LC, Hall GM, Chambers IG. Oral cancer at
a Tasmanian tertiary referral clinic, the Royal
Hobart Hospital 19962002. Aust Dent J 2005;
50(1):316.
[5] Howell RE, Wright BA, Dewar R. Trends in incidence of oral cancer in Nova Scotia from 1983 to

528

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

AROSARENA

1997. Oral Surg Oral Med Oral Pathol Oral Radiol

Endod 2003;95(2):20512.
Otoh E, Johnson N, Olasoji H, et al. Intra-oral carcinomas in Maiduguri, north-eastern Nigeria. Oral
Dis 2005;11(6):37985.
Tsuhako K, Nakazato I, Miyagi J, et al. Comparative study of oral squamous cell carcinoma in Okinawa, Southern Japan and Sapporo in Hikkaido,
Northern Japan; with special reference to human
papillomavirus and Epstein-Barr virus infection.
J Oral Pathol Med 2000;29(2):709.
Pauloski BR, Logemann JA, Colangelo LA, et al.
Surgical variables aecting speech in treated patients with oral and oropharyngeal cancer. Laryngoscope 1998;108(6):90816.
Larson JT, Adams GL, Fattah HA. Survival statistics for multiple primaries in head and neck cancer.
Otolaryngol Head Neck Surg 1990;103(1):1424.
Jovanovic A, van der Tol IG, Kostense PJ, et al.
Second respiratory and upper digestive tract cancer
following oral squamous cell carcinoma. Eur J
Cancer B Oral Oncol 1994;30B(4):2259.
Thomson PJ. Field change and oral cancer: new evidence for widespread carcinogenesis? Int J Oral
Maxillofac Surg 2002;31(3):2626.
Betka J. Distant metastases from lip and oral cavity
cancer. ORL J Otorhinolaryngol Relat Spec 2001;
63(4):21721.
Harrison LB, Fass DE. Radiation therapy for oral
cavity cancer. Dent Clin North Am 1990;34(2):
20522.
Carvalho AL, Magrin J, Kowalski LP. Sites of recurrence in oral and oropharyngeal cancers according to the treatment approach. Oral Dis 2003;9:
1128.
Brennan CT, Sessions DG, Spitznagel EL, et al.
Surgical pathology of cancer of the oral cavity
and oropharynx. Laryngoscope 1991;101:117597.
Shikama N, Sasaki S, Nishikawa A, et al. Risk factors for local-regional recurrence following preoperative radiation therapy and surgery for head
and neck cancer (stage IIIVB). Radiology 2003;
228(3):78994.
Yao M, Dornfeld KJ, Buatti JM, et al. Intensitymodulated radiation treatment for head-and-neck
squamous cell carcinoma: the University of Iowa
experience. Int J Radiat Oncol Biol Phys 2005;
63(2):41021.
Woolgar JA. Histological distribution of cervical
lymph node metastases from intraoral/oropharyngeal squamous cell carcinomas. Br J Oral Maxillofac Surg 1999;37(3):17580.
Chang TM. Induction chemotherapy for advanced
head and neck cancers: a literature review. Head
Neck Surg 1988;10:1509.
Choski AJ, Dimery IW, Hong WK. Adjuvant chemotherapy of head and neck cancer: the past, the
present, and the future. Semin Oncol 1988;
15(Suppl 3):4559.

et al
[21] Jacobs C, Makuch R. Ecacy of adjuvant chemotherapy for patients with respectable head and neck
cancer: a subset analysis of the head and neck contracts program. J Clin Oncol 1990;8:83847.
[22] Stell PM, Dalby JE, Strickland P, et al. Sequential
chemotherapy and radiotherapy in advanced head
and neck cancer. Clin Radiol 1983;34(4):4637.
[23] Toohill RJ, Duncavage JA, Grossman TW, et al.
The eects of delay in standardized treatment due
to induction chemotherapy in two randomized prospective studies. Laryngoscope 1987;97(4):40712.
[24] Licitra L, Grandi C, Guzzo M, et al. Primary chemotherapy in respectable oral cavity squamous
cell cancer: a randomized controlled trial. J Clin
Oncol 2003;21(2):32733.
[25] Ruggeri EM, Carlini P, Pollera CF, et al. Longterm survival in locally advanced oral cavity
cancer: an analysis of patients treated with neoadjuvant cisplatin-based chemotherapy followed by
surgery. Head Neck 2005;27:4528.
[26] Nemeth Z, Szigeti K, Mathe M, et al. Eect of induction chemotherapy on changes of laminin and
syndecan expression in oral squamous cell carcinomas: a prospective, randomized, clinicopathologic
and immunohistochemical study. J Craniofac
Surg 2005;16(2):20512.
[27] Kovacs AF, Ghahremani MT, Stefenelli U, et al.
Postoperative chemotherapy with cisplatin and
5-uorouracil in cancer of the oral cavity and the
oropharynx: long-term results. J Chemother 2003;
15(5):495502.
[28] Shikani AH, Domb AJ. Polymer chemotherapy for
head and neck cancer. Laryngoscope 2000;110(6):
90717.
[29] Sharma PK, Schuller DE, Baker SR. Malignant
neoplasms of the oral cavity. In: Cummings CW,
Fredrickson JM, Harker LA, et al, editors. Otolaryngology head and neck surgery. 3rd edition.
St. Louis (MO): Mosby; 1998. p. 141862.
[30] Fu K, Lichter A, Galante M. Carcinoma of the
oor of mouth: an analysis of treatment results
and the sites and causes of failures. Int J Radiat
Oncol Biol Phys 1976;1:82937.
[31] Hicks WL, Loree TR, Garcia RI, et al. Squamous
cell carcinoma of the oor of mouth: a 20-year review. Head Neck 1997;19(5):4005.
[32] Shintani S, Yoshihama Y, Ueyama Y, et al. The
usefulness of intraoral ultrasonography in the evaluation of oral cancer. Int J Oral Maxillofac Surg
2001;30(2):13943.
[33] Dutton JM, Graham SM, Homan HT. Metastatic
cancer to the oor of mouth: the lingual lymph
nodes. Head Neck 2002;24(4):4015.
[34] Birt BD, Gruss JS. Intra-oral reconstruction using
the nasolabial ap. J Otolaryngol 1985;14(4):
2336.
[35] Varghese BT, Sebastian P, Cherian T, et al. Nasolabial aps in oral reconstruction: an analysis of
224 cases. Br J Plast Surg 2001;54(6):499503.

FLOOR OF MOUTH AND TONGUE CANCER

[36] Golabek W, Kondratowicz J. Sternocleidomastoid

myocutaneous ap for intraoral reconstruction.
Eur Arch Otorhinolaryngol 1990;247(6):37981.
[37] Peng LW, Zhang WF, Zhao JH, et al. Two designs
of platysma myocutaneous ap for reconstruction
of oral and facial defects following cancer surgery.
Int J Oral Maxillofac Surg 2005;34(5):50713.
[38] Tashiro H, Ozeki S, Ohishi M, et al. Cervical island
skin ap for intraoral repair following cancer surgery. Br J Oral Maxillofac Surg 1992;30(1):1823.
[39] Zhao YF, Zhang WF, Zhao JH. Reconstruction of
intraoral defects after cancer surgery using cervical
pedicle aps. J Oral Maxillofac Surg 2001;59(10):
11426.
[40] Fischinger J, Zargi M. Repair of anterior oor of
mouth defects by a central or paramedian island
tongue ap. J Laryngol Otol 2003;117(5):3915.
[41] Haughey BH, Taylor SM, Fuller D. Fasciocutaneous ap reconstruction of the tongue and oor of
mouth: outcomes and techniques. Arch Otolaryngol Head Neck Surg 2002;128(12):138895.
[42] Goerres GW, Schmid DT, Schuknecht B, et al.
Bone invasion in patients with oral cavity cancer:
comparison of conventional CT with PET/CT
and SPECT/CT. Radiology 2005;237:2817.
[43] Imola MJ, Gapany M, Grund F, et al. Technetium
99m single positron emission computed tomography scanning for assessing mandible invasion in
oral cavity cancer. Laryngoscope 2001;111(3):
37381.
[44] Dierks EJ, Karakourtis MH. Segmental resection
of the anterior mandibular arch with bular microvascular reconstruction. Atlas Oral Maxillofac
Surg Clin North Am 1997;5(2):5573.
[45] Urken ML, Sullivan MJ. Fibular osteocutaneous.
In: Urken ML, Cheney ML, Sullivan MJ, et al, editors. Atlas of regional and free aps for head and
neck reconstruction. New York: Raven Press;
1995. p. 291306.
[46] Urken ML, Sullivan MJ. Scapular and parascapular fasciocutaneous and osteofasciocutaneous. In:
Urken ML, Cheney ML, Sullivan MJ, et al, editors.
Atlas of regional and free aps for head and neck
reconstruction. New York: Raven Press; 1995.
p. 21736.
[47] Urken ML. Iliac crest osteocutaneous and osteomusculocutaneous. In: Urken ML, Cheney ML,
Sullivan MJ, et al, editors. Atlas of regional and
free aps for head and neck reconstruction. New
York: Raven Press; 1995. p. 26190.
[48] Ghali GE, Sikes JW. Lateral mandibulectomy and
partial glossectomy with plate application. Atlas
Oral Maxillofac Surg Clin North Am 1997;5(2):
114.
[49] Urken ML, Biller HF. Pectoralis major. In:
Urken ML, Cheney ML, Sullivan MJ, et al, editors.
Atlas of regional and free aps for head and neck
reconstruction. New York: Raven Press; 1995.
p. 328.

529

[50] Urken ML. Rectus abdominis. In: Urken ML,

Cheney ML, Sullivan MJ, et al, editors. Atlas of regional and free aps for head and neck reconstruction. New York: Raven Press; 1995. p. 11938.
[51] Butler CE. Skin grafts used in combination with
free aps for intraoral oncological reconstruction.
Ann Plast Surg 2001;47(3):2938.
[52] Teichgraeber J, Bowman J, Goepfert H. Functional
analysis of treatment of oral cavity cancer. Arch
Otolaryngol Head Neck Surg 1986;112:95965.
[53] Sessions DG, Spector GJ, Lenox J, et al. Analysis
of treatment results for oor-of-mouth cancer. Laryngoscope 2000;110(10 Part 1):176472.
[54] Chu A, Fletcher G. Incidence and causes of failure
to control by irradiation the primary lesions in
squamous cell carcinomas of the anterior twothirds of the tongue and oor of mouth. Am J
Roentgenol Radium Ther Nucl Med 1973;117(3):
5028.
[55] Pernot M, Hostetter S, Peiert D, et al. Epidermoid carcinomas of the oor of mouth treated by
exclusive irradiation: statistical study of a series of
207 cases. Radiother Oncol 1995;35(3):17785.
[56] Marsiglia H, Haie-Meder C, Sasso G, et al. Brachytherapy for T1T2 oor-of-the-mouth cancers: the
Gustave-Roussy Institute experience. Int J Radiat
Oncol Biol Phys 2002;52(5):125763.
[57] Wadsley JC, Patel M, Tomlins CD, et al. Iridium192 implantation for T1 and T2a carcinoma of
the tongue and oor of mouth: retrospective study
of the results of treatment at the Royal Berkshire
Hospital. Br J Radiol 2003;76(906):4147.
[58] Lindberg R. Distribution of cervical lymph node
metastases from squamous cell carcinoma of the
upper respiratory and digestive tracts. Cancer
1972;29(6):14469.
[59] Decroix Y, Ghossein N. Experience of the Curie
Institute in treatment of cancer of the mobile
tongue: I. Treatment policies and results. Cancer
1981;47(3):496502.
[60] Mendenhall WM, van Cise WS, Bova FJ, et al.
Analysis of time-dose factors in squamous cell carcinoma of the oral tongue and oor of mouth
treated with radiation therapy alone. Int J Radiat
Oncol Biol Phys 1981;7(8):100511.
[61] Yamazaki H, Inoue T, Koizumi M, et al. Comparison of long-term results of brachytherapy for
T12N0 oral tongue cancer treated with Ir-192
and Ra-226. Anticancer Res 1997;17(4A):281922.
[62] Spiro R, Strong E. Epidermoid carcinoma of the
mobile tongue: treatment by partial glossectomy
alone. Am J Surg 1971;122(6):70710.
[63] Franceschi D, Gupta R, Spiro RH, et al. Improved
survival in the treatment of squamous carcinoma of
the oral tongue. Am J Surg 1993;166(4):3605.
[64] Zelefsky MJ, Harrison LB, Fass DE, et al. Postoperative radiotherapy for oral cavity cancers: impact
of anatomic subsite on treatment outcome. Head
Neck 1990;12(6):4705.

530

AROSARENA

[65] Urken ML, Biller HF. A new bilobed design for the
sensate radial forearm ap to preserve tongue mobility following signicant glossectomy. Arch Otolaryngol Head Neck Surg 1994;120(1):2631.
[66] Yokoo S, Komori T, Umeda M, et al. Functional
reconstruction of mobile tongue and suprahyoid
muscles after resection of cancer of the tongue. Br
J Oral Maxillofac Surg 2001;39(4):2525.
[67] Beenken SW, Krontiras H, Maddox WA, et al. T1
and T2 squamous cell carcinomas of the oral
tongue: prognostic factors and the role of elective
lymph node dissection. Head Neck 1999;21:12430.
[68] OBrien CJ, Lauer CS, Fredricks S, et al. Tumor
thickness inuences prognosis of T1 and T2 oral
cavity cancer: but what thickness? Head Neck
2003;25:93745.
[69] Shibuya H, Hoshina M, Takeda M, et al. Brachytherapy for stage I and II oral tongue cancer: an
analysis of past cases focusing on control and complications. Int J Radiat Oncol Biol Phys 1993;26:
518.
[70] Yuen APW, Lam KY, Lam LK, et al. Prognostic
factors of clinically stage I and II oral tongue carcinoma: a comparative study of stage, thickness,
shape, growth pattern, invasive front malignancy
grading, Martinez-Gimeno score, and pathologic
features. Head Neck 2002;24(6):51320.
[71] Myers LL, Wax MK. Positron emission tomography in the evaluation of the negative neck in patients with oral cavity cancer. J Otolaryngol 1998;
27(6):3427.
[72] Klotch DW, Muro-Cacho C, Gal TJ. Factors affecting survival for oor of mouth carcinoma. Otolaryngol Head Neck Surg 2000;122(4):4958.
[73] Sessions DG, Spector GJ, Lenox J, et al. Analysis
of treatment results for oral tongue cancer. Laryngoscope 2002;112(4):61625.
[74] Spiro RH, Guillamondegui O, Paulino AF, et al.
Pattern of invasion and margin assessment in patients with oral tongue cancer. Head Neck 1999;
21:40813.
[75] Woolgar JA, Scott J. Prediction of cervical lymph
node metastasis in squamous cell carcinoma of
the tongue/oor of mouth. Head Neck 1995;17:
46372.
[76] Al-Rajhi N, Khafaga Y, El-Husseiny J, et al. Early
stage carcinoma of the oral tongue: prognostic factors for local control and survival. Oral Oncol 2000;
36(6):50314.
[77] Brown B, Barnes L, Mazariegos J, et al. Prognostic
factors in mobile tongue and oor of mouth carcinoma. Cancer 1989;64(6):1195202.
[78] Jones AS. Prognosis in mouth cancer: tumour factors. Eur J Cancer B Oral Oncol 1994;30B(1):815.
[79] Kowalski LP, Medina JE. Nodal metastases: predictive factors. Otolaryngol Clin North Am 1998;
31(4):62137.
[80] Sparano A, Weinstein G, Chalian A, et al. Multivariate predictors of occult neck metastasis in early

et al

[81]

[82]

[83]

[84]

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92]

[93]

[94]

oral tongue cancer. Otolaryngol Head Neck Surg

2004;131(4):4726.
OBrien CJ, Lahr CJ, Soong SJ, et al. Surgical treatment of early stage carcinoma of the oral tongue:
would adjuvant treatment be benecial? Head
Neck Surg 1986;8(6):4018.
Close CG, Burns DK, Reisch J, et al. Microvascular invasion in cancer of the oral cavity and oropharynx. Arch Otolaryngol Head Neck Surg
1987;113(11):11915.
Gallegos-Hernandez J, Hernandez- Hernandez D,
Flores-D az R, et al. The number of sentinel nodes
identied as prognostic factor in epidermoid cancer. Oral Oncol 2005;41:94752.
Mohit-Tabatabai MA, Sobel HJ, Rush BF, et al.
Relation of thickness of oor of mouth stage I
and II cancers to regional metastasis. Am J Surg
1986;152:3513.
Spiro RH, Huvos AG, Wong GY, et al. Predictive
value of tumour thickness in squamous carcinoma
conned to the tongue and oor of mouth. Am J
Surg 1986;152:34550.
Matsuura K, Hirokawa Y, Fujita M, et al. Treatment results of stage I and II oral tongue cancer
with interstitial brachytherapy: maximum tumor
thickness is prognostic of nodal metastasis. Int J
Radiat Oncol Biol Phys 1998;40(3):5359.
Nathanson A, Agren K, Biorklund A, et al. Evaluation of some prognostic factors in small squamous cell carcinoma of the mobile tongue:
a multicenter study in Sweden. Head Neck
1989;11(5):38792.
Kurokawa H, Yamashita Y, Takeda S, et al. Risk
factors for late cervical lymph node metastases in
patients with stage I or II carcinoma of the tongue.
Head Neck 2002;24:7316.
Morton RP, Ferguson CM, Lambie NK, et al. Tumour thickness in early tongue cancer. Arch Otolaryngol Head Neck Surg 1994;120:71720.
Civantos FJ, Gomez C, Duque C, et al. Sentinel
node biopsy in oral cavity cancer: correlation with
PET scan and immunohistochemistry. Head Neck
2003;25:19.
Byers RM, El-Naggar AK, Lee YY, et al. Can we
detect or predict the presence of occult nodal metastases in patients with squamous carcinoma of
the oral tongue? Head Neck 1998;20(2):13844.
McGuirt WF, Johnson JT, Myers EN, et al. Floor
of the mouth carcinoma: the management of the
clinically negative neck. Arch Otolaryngol Head
Neck Surg 1995;121(3):27882.
Dias FL, Kligerman J, Matos de Sa G, et al. Elective neck dissection versus observation in stage I
squamous cell carcinomas of the tongue and oor
of the mouth. Otolaryngol Head Neck Surg 2001;
125(1):239.
Balkissoon J, Rasgon BM, Schweitzer L. Lymphatic mapping for staging of head and neck cancer. Semin Oncol 2004;31(3):38293.

FLOOR OF MOUTH AND TONGUE CANCER

[95] Yuen AP, Lam KY, Chan AC, et al. Clinicopathological analysis of elective neck dissection for N0
neck of early oral tongue carcinoma. Am J Surg
1999;177(1):902.
[96] Byers RM, Weber RS, Andrews T, et al. Frequency
and therapeutic implications of skip metastases
in the neck from squamous carcinoma of the oral
tongue. Head Neck 1997;19(1):149.
[97] Khaf A, Lopez-Garza JR, Medina JE. Is dissection of level IV necessary in patients with T1T3
N0 tongue cancer? Laryngoscope 2001;111:
108890.
[98] Ross GL, Soutar DS, Gordon-MacDonald D, et al.
Sentinel node biopsy in head and neck cancer: preliminary results of a multicenter trial. Ann Surg
Oncol 2004;11(7):6906.
[99] Greenberg JS, Fowler R, Gomez J, et al. Extent of
extracapsular spread: a critical prognosticator in
oral tongue cancer. Cancer 2003;97(6):146470.

531

[100] Myers JN, Greenberg JS, Mo V, et al. Extracapsular spread: a signicant predictor of treatment failure in patients with squamous cell carcinoma of the
tongue. Cancer 2001;92(12):30306.
[101] Alvi A, Johnson JT. Extracapsular spread in the
clinically negative neck (N0): implications and outcome. Otolaryngol Head Neck Surg 1996;114(1):
6570.
[102] Taneja C, Allen H, Koness RJ, et al. Changing patterns of failure of head and neck cancer. Arch Otolaryngol Head Neck Surg 2002;128(3):3247.
[103] Shintani S, Matsuura H, Hasegawa Y, et al.
Regional lymph node involvement aects the
incidence of distant metastasis in tongue squamous cell carcinomas. Anticancer Res 1995;15:
15736.
[104] Rayatt SS, Dancey AL, Fagan J, et al. Axillary metastases from recurrent oral carcinoma. Br J Oral
Maxillofac Surg 2004;42(3):2646.

Oral Maxillofacial Surg Clin N Am 18 (2006) 533546

Surgical Management of the Neck in Oral Cancer

Eric R. Carlson, DMD, MD, FACS*, Ivo Miller, DDS
Department of Oral and Maxillofacial Surgery, University of Tennessee Graduate School of Medicine, University
of Tennessee Cancer Institute, 1930 Alcoa Highway, Suite 335, Knoxville, TN 37920, USA

History of neck dissections for oral cancer

Surgical removal of the cervical lymph nodes
plays an important role in the comprehensive
management of squamous cell carcinoma of the
oral cavity. Two of the most important aspects of
the assessment of patients with these cancers
include the clinical evaluation of the lymph nodes
of the neck and the prediction of occult neck
disease in the case of a clinically negative neck.
Occult neck disease is dened as disease that is
present microscopically but cannot be palpated
clinically and may defy identication by special
imaging studies. As such, patients deemed to have
occult neck disease were staged as N0 clinically
and underwent elective surgical dissection of their
cervical lymph nodes with the histopathologic
identication of positive nodes (pN). It is the
philosophy of these authors, and that of most,
that elective surgical removal of the cervical
lymph nodes that are clinically negative (N0)
should be executed liberally with curative intent.
Oral cancer is most commonly treated surgically,
so it is most appropriate that the neck also be
treated surgically, with radiotherapy being added
postoperatively when pathologic evidence so dictates [1]. Observing the N0 neck, only to operate
on the neck in the case of future, clinically apparent nodal disease is oncologically unsafe in most
cases of oral squamous cell cancer. This statement
is based on the realization that salvage rates for
these patients are unfavorable.
To this end, Warren in 1839 recommended
removal of lymph nodes in the submandibular
triangle associated with tongue cancer in the best
* Corresponding author.
E-mail address: Ecarlson@mc.utmck.edu
(E.R. Carlson).

interests of improving the curability of cancer at

that site [2]. One of the rst systematic descriptions of the importance of cervical lymph nodes
in head and neck cancer was reported by Maximilian von Chelius in 1847 [2]. In 1906, a since frequently quoted paper was published in the
Journal of the American Medical Association by
Dr. George Crile of the Cleveland Clinic in Ohio
[3]. It was entitled Excision of cancer of the
head and neckdWith special reference to the
plan of dissection based on one hundred and
thirty-two operations. Although the 1906 paper
is the most popular of its kind and often is
referred to as the rst of Criles great works on
this subject, it is actually his paper published in
1905, entitled On the surgical treatment of cancer of the head and neck in which Crile initially
described an en bloc dissection of the neck [4,5].
In the 1905 paper, Crile created an analogy between breast cancer, in which regional lymph nodes are routinely excised, and head and neck
cancer, in which a similar approach should be applied. He stated that such a dissection of lymph
nodes of the neck is indicated regardless of
whether the glands are palpable. According to
Crile, palpable glands may be inammatory
and impalpable glands may be carcinomatous.
A strict rule of excision should therefore be followed. Early in the introduction of his 1906
paper, Crile astutely identied that the immediate
extension from the primary malignant focus principally occurred by permeation and metastasis in
the regional lymphatics. As such, Crile summarized his recommendations for surgical management of the neck by stating that an incomplete
operation would lead to dissemination of disease,
stimulate the growth of the cancer, shorten the
patients life, and diminish comfort. He emphasized his previously stated philosophy that

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.002

oralmaxsurgery.theclinics.com

534

CARLSON & MILLER

isolated excision of the primary focus of the cancer was as unsurgical as excision of a breast in
the case in which the regional lymph nodes remained unaddressed. He oered support of en
bloc removal of cervical lymph nodes because
excision of individual lymphatic glands would
not result in cure of the patient but would rather
be followed by greater dissemination and more
rapid growth. He emphasized that a block dissection of the regional lymphatics and the primary
malignancy was necessary for eective treatment
of these patients.
This block dissection included lymph nodes in
levels I-V of the neck (Table 1), the sternocleidomastoid muscle, the internal jugular vein, and
the spinal accessory nerve. Crile performed this
treatment in the management of patients in
whom lymph nodes were enlarged (N neck)
and in patients whose lymph nodes were not
clinically enlarged (N0 neck). Although Criles
comments were directed to head and neck cancer
as a whole, oral cavity cancers represented only
a minority, including four cases of oor of mouth
cancer, one soft palate cancer, two alveolar ridge
cancers, and 12 cancers of the tongue. This
notwithstanding, this paper served as a model
for treatment of the neck in patients with oral
cancer.
Interestingly, the most common cancer treated
by Crile in his report of 132 cancers was that of
the lips, accounting for 31 of these cases. By 2006
standards, most of these lip cancers could have
been managed without neck treatment. Although
the often-quoted theme of Criles paper was
radical neck dissection (RND), only 36 patients
underwent such treatment in his report. In a 1923
paper, Crile elaborated on his recommendations
for excision of the cervical lymph nodes by stating
that early cancer of the gingiva or cheek that
metastasizes late does not demand excision of the
lymph nodes, whereas cancer of the lip, however
early, demands the complete excision of all lymph
nodes that drain the involved area, and cancer of
the tongue or lip calls for complete removal of the
lymph nodes of the neck on both sides. In these
cases, he indicated that a tracheostomy was
doubly indicated, because aside from the short
circuiting of respiration and xing the trachea, it
produced a wall of protective granulations across
the top of the dangerous mediastinal area. Criles
1923 paper reiterated many of the comments
made in the 1905 and 1906 papers, including
other oncologically safe principles, such as not
handling the specimen [6].

Dr. Criles three papers represented the landmark articles regarding neck dissections for head
and neck cancer until Dr. Hayes Martin published
his paper entitled Neck dissection in 1951 [7].
This extensive review commented on an experience of 1450 neck dissections performed from
1928 to 1950, although statistics were derived
from 665 operations performed in 599 patients.
Dr. Martin did not believe that a routine prophylactic RND was practical in managing patients
with cancer of the tongue and lip and presented
data from a survey sent to 75 of his colleagues,
the consensus of which supported his contentions.
His conclusion with regard to RND was that routine prophylactic neck dissection was considered
illogical and unacceptable for cancer of the
oral cavity. He made these comments because of
his thoughts about oncologic safety and not about
functional consequences, stating that no one
could carry out prophylactic neck dissection to
a degree sucient to eect signicant improvements in cure rates. He believed that the RND
was too radical a technique for elective use. In
other words, the RND should not be used for
the N0 neck, a philosophy that is largely observed
currently. With regard to the elective neck dissection, Martin reported that this concept was not
practiced on the Head and Neck Service of Memorial Hospital at the time. Rather, he believed
that denite clinical evidence that cancer was present in the lymph nodes represented one criterion
for neck dissection.
Although RND has proved to be a reliable
method of treating patients with oral/head and
neck cancer, it is associated with substantial
morbidity. Nahum and colleagues [8] described
a syndrome of pain and decreased range of abduction in the shoulder after RND, which has been
referred to as shoulder syndrome and relates to
the sacrice of the spinal accessory nerve. Research has shown that preservation of the spinal
accessory nerve during neck dissection ameliorates the syndrome [9]. The morbidity of the
RND gave way to the development of the numerous modications of the RND that maintain oncologic safety while also reducing morbidity of
the RND. These modications of the RND were
designed to preserve the sternocleidomastoid muscle, spinal accessory nerve, and internal jugular
vein and have been realized in the form of the
modied RND (MRND) proper, functional
neck dissection, and the selective neck dissections
represented by the supraomohyoid neck dissection. By 2006 standards, radical and MRNDs

SURGICAL MANAGEMENT OF THE NECK IN ORAL CANCER

535

Table 1
Oncologic lymph node levels of the neck
Lymph node group

Description

IA (submental)

Lymph nodes within the triangular boundary of the anterior

belly of the digastric muscles and the hyoid bone
Lymph nodes within the boundaries of the anterior belly of the
digastric muscle and the stylohyoid muscle and the inferior
border of the mandible
Lymph nodes located around the upper third of the internal
jugular vein and the adjacent spinal accessory nerve; level
IIA lymph nodes are located anterior (medial) to the spinal
accessory nerve; level IIB lymph nodes are located posterior
(lateral) to the spinal accessory nerve
Lymph nodes located around the middle third of the internal
jugular vein; nodes are located between the inferior border
of the hyoid bone and the inferior border of the cricoid cartilage
Lymph nodes located around the lower third of the internal
jugular vein; nodes extend from the inferior border of the
cricoid cartilage to the clavicle
Lymph nodes located along the lower half of the spinal
accessory nerve and the transverse cervical artery; supraclavicular
nodes are located in this group of lymph nodes
Lymph nodes in the prelaryngeal, pretracheal, paratracheal, and
tracheoesophageal groove; boundaries are the hyoid bone to the
suprasternal notch and between the medial borders of the
carotid sheaths; lymph nodes are generally not dissected in
oral cancer patients
Lymph nodes in the anterior superior mediastinum and
tracheoesophageal grooves, extending from the suprasternal
notch to the innominate artery; lymph nodes are generally
not dissected in oral cancer patients

IB (submandibular)

IIA and IIB (upper jugular)

III (middle jugular)

IV (lower jugular)

V (posterior triangle)

VI (central compartment)

VII (superior mediastinal)

are most commonly performed as therapeutic

neck dissections for clinically N disease, whereas
selective neck dissections are most commonly performed as elective neck dissections for clinically
N0 disease.

Anatomy of cervical lymph nodes in relation

to oral cancer
A rational approach to the surgical management of the neck in patients with oral squamous
cell carcinoma requires a thorough understanding
of the lymphatic anatomy of the neck and the
patterns of nodal metastasis from these cancers.
To this end, ve oncologic levels have been
described in patients with oral cavity cancers
and form the basis of the nomenclature for the
classication of neck dissections (Table 1, Fig. 1).
Cervical lymph nodes are categorized into seven
levels by numerous experts at the Memorial Sloan
Kettering Cancer Center in New York [5]. To
make matters more confusing, classications for

Fig. 1. The oncologic lymph node levels of the neck as

applied to oral cavity squamous cell carcinoma. (From
Regezi J, Sciubba J, Jordan RCK, editors. Oral pathology: clinical pathologic correlations. 4th edition. Philadelphia: WB Saunders Co; 2003.)

536

CARLSON & MILLER

neck dissections by the American Head and Neck

Society reviewed six levels for dening the boundaries of neck dissection [10,11]. Despite this confusion, in general terms, lymph nodes in levels I-III
are considered sentinel or rst echelon lymph nodes for oral cavity cancers. In this sense, these are
the rst lymph nodes that are expected to contain
metastatic squamous cell carcinoma when the
neck in fact contains cancer. This well-accepted
concept forms the basis for elective neck dissections in which the likelihood of occult neck disease
exceeds 20%.

Classication of neck dissections

So as to develop uniformity regarding nomenclature, Robbins and colleagues [10] developed
standardized neck dissection terminology in 1991
and updated the classication in 2002 (Table 2)
[11]. Their original classication was based on
the following concepts: (1) the RND is the fundamental procedure with which all other neck dissections are compared, (2) MRND denotes
preservation of one or more nonlymphatic structures, (3) selective neck dissections denote preservation of one or more groups of lymph nodes, and
(4) extended RND denotes removal of one or
more additional lymphatic or nonlymphatic structures. An MRND refers to the excision of all
lymph nodes routinely removed by RND with
preservation of one or more nonlymphatic structures, such as the spinal accessory nerve, internal
jugular vein, and sternocleidomastoid muscle. As
such, lymph node levels I-V are removed in this
neck dissection (Table 1). Typically, a type I
MRND involves preservation of the spinal
Table 2
Classication of neck dissections
1991 Classication

2002 Classication

1. Radical
neck dissection
2. Modied
radical neck dissection
3. Selective
neck dissection
a. Supraomohyoid
b. Lateral
c. Posterolateral
d. Anterior

1. Radical
neck dissection
2. Modied
radical neck dissection
3. Selective
neck dissection: each
variation is depicted
by SND and the use
of parentheses
to denote the levels
or sublevels removed

4. Extended
neck dissection

4. Extended
neck dissection

accessory nerve, a type II MRND involves preservation of the spinal accessory nerve and the internal jugular vein, and a type III MRND involves
preservation of the spinal accessory nerve, internal jugular vein, and sternocleidomastoid muscle
[12]. It seems that most authors favor the type I
MRND for the N neck in oral cavity cancers
(Fig. 2) [13], and this modication of the traditional RND does not compromise oncologic
safety [14]. Neck dissections may be further classied as comprehensive or selective. Comprehensive
neck dissections are those in which cervical lymph
nodes are removed in levels I-V. Such neck dissections are represented by the radical and MRNDs
for N disease and commonly remove nonlymphatic tissue. Selective neck dissections are those
in which cervical lymph nodes are selectively removed, most commonly for N0 disease. The
most commonly performed selective neck dissection for oral cavity cancer is the supraomohyoid
neck dissection, which removes lymph nodes in
levels I, II, and III. The anterolateral neck dissection removes lymph nodes in levels II, III and IV,
and the posterolateral neck dissection removes
lymph nodes in levels II, III, IV, and V. The functional neck dissection is a poorly understood and
often misquoted neck dissection insofar as the dened sacrice of lymph node levels.
Modied radical neck dissections
The surgical concepts of MRNDs are based on
the understanding that the aponeurotic system of
the neck encases the internal structures that are
usually removed during RND. The MRND works
within these planes of dissection and still results
in an en bloc lymphadenectomy while preserving
structures such as the spinal accessory nerve,
sternocleidomastoid muscle, and internal jugular
vein. Many modications of the RND have been
proposed, including MRND proper, functional
neck dissection, and selective neck dissections as
represented by the commonly performed supraomohyoid neck dissection.
Functional neck dissection
In 1967, Bocca and Pignataro [15] published
their work on a more conservative neck dissection
that has been referred to as the functional neck
dissection. The rst author to describe the functional neck dissection was Osvaldo Suarez from
the University of Cordoba Medical School in Argentina. He published the rst original, systematic

SURGICAL MANAGEMENT OF THE NECK IN ORAL CANCER

537

Fig. 2. (AC) A 52-year-old man with obvious right

neck swelling. Examination revealed an indurated lymph
node that measured approximately 3.5 cm in level II. Examination of the oral cavity showed a 3-cm ulcerated
mass (D) that was diagnosed as squamous cell carcinoma of the right oor of mouth by incisional biopsy.
Staging was T2N2aM0 squamous cell carcinoma of the
right oor of mouth. Computerized tomogram (E) demonstrated a large lymph node of level II of the right neck
that was compressing the right internal jugular vein. Surgery consisted of a composite resection that included
a right type I modied radical neck dissection and
a wide excision of the oor of mouth primary cancer
in concert with a lingual cortical resection of the right
mandible. The neck incision incorporated an upper
member McFee incision with an oblique release to the
clavicle (F). The deep aspect of this neck dissection is
the carotid artery, vagus nerve, phrenic nerve, and other
deep neck structures (G). The specimen (H) consists of
an en bloc resection of the neck, lingual cortex of mandible, and oor of mouth. The resultant defect (I) shows
the carotid system, vagus and phrenic nerves, and the remaining mandible stabilized with a reconstruction bone
plate so as to avoid pathologic fracture in the postoperative course. Four of 41 lymph nodes in the neck dissection specimen were positive for metastatic squamous cell
carcinoma. The patient underwent postoperative radiation therapy and displayed acceptable facial form at 6
months postoperatively (J, K).

538

CARLSON & MILLER

Fig. 2 (continued)

approach to this neck dissection in 1963 [5]. History indicates that Bocca learned the technique
from Suarez and later published numerous observations on this technique as an elective neck dissection. Before that time, the elective neck
dissection was the RND. In their report the authors described absence of lymphatic recurrences

in approximately 100 neck dissections in which

only the lymphatic tissue of the neck was sacriced and the sternocleidomastoid muscle, internal
jugular vein, and spinal accessory nerve were preserved. Of interest are the authors comments regarding the ability to not include level I lymph
nodes and structures in this conservative neck

Fig. 2 (continued)

540

CARLSON & MILLER

dissection [12,15,16]. Even when the dissection required the removal of level I lymph nodes, the authors stated that the submandibular gland could
be preserved. In either case, preservation of some
or all structures in the sentinel level I of the neck
results in questionable use of this neck dissection
when managing an oral cavity cancer. In 1984,
these authors published their ndings of 1500
functional neck dissections in 843 patients operated on between 1961 and 1982 [17]. Cancer of
the larynx made up 87% of the patients in this series. Twelve hundred of these neck dissections were
elective (N0), whereas 300 were therapeutic (N).
Neck recurrences occurred in 68 cases (8.1%). Of
these, 16 occurred in the elective (N0) functional
neck dissection patients (2.38%) and 52 cases of
recurrence occurred in the 171 curative (N) functional neck dissection patients (30.4%). Calearo
and Teatini [18] reviewed 476 functional neck dissections that were performed in 211 patients with
only nine recurrences (3.5%) during a 3-year follow-up period. Other authors have expressed similar satisfaction with this neck dissection [19,20].
Supraomohyoid neck dissection
The supraomohyoid neck dissection is the ideal
solution to the dilemma for some surgeons as to
how to manage the N0 neck properly (Fig. 3) [21].
A sucient body of literature supports the performance of elective neck dissections for T1N0 and
T2N0 squamous cell carcinomas of the oral cavity,
identifying the incidence of 36% to 42% of occult
neck disease in these cases [2224]. As such, numerous authors have recommended the supraomohyoid neck dissection enthusiastically as a staging
procedure in the management of the N0 neck associated with oral cavity squamous cell carcinoma
[2529]. By denition, the supraomohyoid neck
dissection removes lymph node levels I-III while
preserving the spinal accessory nerve, internal jugular vein, and sternocleidomastoid muscle. The senior author shares the opinion of many that
delaying elective surgery of the N0 neck and adopting the wait-and-watch approach to the N0 neck is
foolhardy in most cases. It is particularly true for
cases of tongue cancer, in which survival in the observation group has been noted to be 33% compared with 55% in the neck dissection group,
and locoregional control increased from 50% to
91% when neck dissection was performed [30].
The supraomohyoid neck dissection is
a straightforward surgical procedure that requires
little time to perform and oers prognostically

signicant information to the patient and the

radiation oncologist, who might otherwise be
unable to render treatment based on objective
information. Most importantly, it addresses neck
disease in an occult stage whereby survival is
improved. Its scientic basis is the observation
that lymph nodes in levels I-III are sentinel nodes.
In a classic study by Shah and colleagues [31] that
examined the specimens of 501 patients undergoing RND, only 9% of patients showed pathologically proved metastases in level IV when the
neck dissection was elective in nature. The incidence of positive nodes in level V was only 2%.
These data indicate that levels IV and V probably
do not require removal in the management of the
N0 neck. Shah and Singh [32] recently recommended the excision of level IV lymph nodes along
with levels I, II, and III when operating on the
N0 neck related to primary cancers of the lateral
border of the oral tongue. Crean and colleagues
[33] found occult metastases in level IV in 5 of
49 cases of oral cavity cancers. Their conclusion
was that extending the traditionally performed
supraomohyoid neck dissection to include the easily accessible level IV should be adopted as standard treatment in the management of the N0
neck. Byers and colleagues [34] found an overall
frequency of skip metastases in squamous cell carcinoma of the oral tongue in 15.8% of their patients studied. These patients had either level IV
metastases as the only manifestation of disease
in the neck or the level III node was the only positive node present without disease in level I and II.
The authors conclusions were that the usual
supraomohyoid neck dissection is inadequate for
a complete pathologic evaluation of all the nodes
at risk for patients with squamous carcinoma of
the oral tongue.
Some authors have recommended performing
the supraomohyoid neck dissection with frozen
section analysis of the lymph nodes to permit
intraoperative extension of the neck dissection in
the form of an MRND [35,36]. Other authors
have stressed the importance of including level
IV beyond the traditional supraomohyoid neck
dissection when managing the N0 neck where the
primary tumor is located in the tongue [34,37]. Although many surgeons do not advocate the use of
the supraomohyoid neck dissection in the management of the N neck, some recent evidence
suggests the ecacy of this type of neck dissection
with postoperative radiation therapy in the management of head and neck cancer patients with
an N neck [38].

SURGICAL MANAGEMENT OF THE NECK IN ORAL CANCER

541

Fig. 3. (A, B) A 49-year-old gentleman with a 3.5-cm biopsy proven squamous cell carcinoma of the left tongue.
Staging was consistent with a T2N0M0 squamous cell
carcinoma of the tongue. Surgery consisted of a left partial glossectomy and an ipsilateral supraomohyoid neck
dissection. Access to the neck was accomplished with
a modied apron incision from mastoid to the submental region (C). The specimen is elevated o the carotid
sheath contents and the suprahyoid and strap muscles
(D). The neck specimen (E) contained 28 lymph nodes,
with one level II node showing metastatic squamous
cell carcinoma. Note the neck defect (F). The tongue excision (G, H) observed 1-cm linear margins. A primary
closure was accomplished. Extensive perineural invasion
was noted in the tongue specimen, thereby requiring
postoperative radiation therapy. One-year postoperative
evaluation of the patient (I, J) and the tongue (K)
showed no evidence of disease.

542

CARLSON & MILLER

Fig. 3 (continued)

Decision making regarding what type of neck

dissection to perform
The senior author performs surgical treatment
of the neck primarily using two main neck
dissections. In patients with a clinically N0 neck,

the sentinel or rst echelon nodes are present in

levels I, II, and III. As such, a supraomohyoid
neck dissection as described would address the
neck adequately in these patients. Stage I and II
squamous cell carcinoma of the tongue is a disease
with a high likelihood of occult neck disease,

SURGICAL MANAGEMENT OF THE NECK IN ORAL CANCER

543

Fig. 3 (continued)

which warrants elective treatment of the neck. The

senior author does not adopt the wait-and-watch
approach in such circumstances. Consideration
should be given to the dissection of level IV in
many of these cases. The use of the supraomohyoid neck dissection in patients who have pathologically negative nodes has resulted in excellent
control of disease in the neck, with failure rates
of less than 10% [39]. In the 30% of patients
who are found to have pathologically proved occult neck disease, the failure rate with the supraomohyoid neck dissection alone ranges from 10%
to 24% depending on the number of positive

nodes and the presence of extracapsular spread

[39]. When postoperative radiation therapy is
added to this scenario, the failure rates drop to
0 to 15%, again depending on the extent of nodal
metastases [40].
In patients with palpable cervical metastasis
(N), the levels at highest risk are levels I-IV. In
the opinion of the senior author, the most prudent
operation to perform is the type I MRND, which
spares the spinal accessory nerve as long as it is
not involved with tumor. This approach is advocated because of the fact that even palpable nodes
smaller than 3 cm in diameter have a substantial

544

CARLSON & MILLER

incidence of extracapsular spread of disease [41].

The presence of extracapsular spread may breech
the aponeurotic planes relied on when the sternocleidomastoid muscle and internal jugular vein
are otherwise preserved. As such, a functional
neck dissection is probably contraindicated in
managing the N neck related to oral cavity cancer
[39]. In patients with clinically palpable neck disease, the results of type I MRND followed by postoperative radiation therapy depend on the bulk of
disease in the neck, the presence of extracapsular
spread in the lymph node, and the radiosensitivity
of microscopic disease remaining in the neck.
When patients receive postoperative radiation
therapy after a type I MRND for an N neck,
neck failure rates in N1 patients are 7% to 10%
and approximately 12% in N2 patients [42].
These results compare favorably to rates obtained
with RND in similar patients [39]. It is oncologically safe and appropriate to preserve cranial nerve
XI.
Although a great deal of thought must be
exerted in determining which neck dissection is
best suited for patients with oral cancer, perhaps
the more controversial aspect of oral cancer care
surrounds the small cohort of patients who do not
require a neck dissection for eective management
of their oral cancer. Because of the relative lack of
morbidity and high yield of the supraomohyoid
neck dissection in managing the N0 neck, it seems
reasonable to perform this neck dissection in
most, if not all, patients with squamous cell carcinoma of the oral cavity. For some patients with
early oral cancer, elective neck dissection is not
necessary. In general terms, the authors believe
that the risk of occult neck disease is less than
20% in some patients with primary T1N0 and
T2N0 cancers of the lip and anterior maxillary
and mandibular gingiva squamous cell carcinoma.
As such, elective neck dissections may not be
required in these patients in select clinical circumstances. Robbins and Samant [43] indicated that
a supraomohyoid neck dissection should be
performed for all T1-T4 tongue cancers, T2-T4 at
all other sites, and where there is identied perineural or lymphatic invasion in the primary specimen.
Considerations for future management of the neck
in oral cancer
The next era of neck dissection might be best
referred to as superselective neck dissection using
sentinel node biopsies [44]. This approach
was originally developed for patients who have

melanoma [45], and it specically analyzes lymphoscintigraphy-guided biopsies of sentinel nodes

in the neck in determining if a neck dissection is
indicated. Careful examination of a specic lymph
node, as occurs in the evaluation of sentinel lymph
node biopsies but not elective neck dissections,
may increase the use of sentinel node mapping.
Ambrosch and colleagues [46] analyzed a series
of 76 neck dissection specimens from patients
originally staged as pathologically negative. The
authors used serial section in 10-mm intervals
and hematoxylin-eosin staining and cytokeratin
staining to reassess the lymph nodes. Eight previously undiagnosed micrometastases were identied in six specimens from six patients, which
resulted in upstaging. Another study revealed
that 36 of 96 (37%) pathologically negative elective neck dissection specimens contained micrometastases upon serial resectioning [47]. Similar
ndings have been noted by other authors
[48,49]. The implication is that elective neck dissections might be therapeutic in more necks than
previously shown. In the case of examination of
a sentinel lymph node, serial examination is essential [44]. In a study investigating the accuracy of
sentinel lymph node biopsies in patients with
squamous cell carcinoma, Shoaib and colleagues
[50] were able to identify sentinel lymph nodes in
36 of 40 necks (90%). In four necks, nonsentinel
lymph nodes contained tumor in the presence of
pathologically positive sentinel lymph nodes. In
only one case was a nonsentinel lymph node
found to contain tumor, which gave the impression of a false-negative neck based on sentinel
lymph node biopsies. Werner and colleagues [51]
reported that sentinel lymph node biopsies correctly identied metastatic disease in 97% of their
90 patients with head and neck squamous cell carcinoma. They indicated that if only the lymph
node with the highest tracer activity had been excised, 39% of cancer-positive necks would have
been missed. Payoux and colleagues [52] examined
30 patients with 37 neck dissections for N0 necks.
Preoperative lymphoscintigraphy, was performed
and sentinel nodes were identied. In 29 necks,
the sentinel node and neck dissection were negative for metastatic disease. Lymph node mapping
allowed for identication of 6 of 7 positive necks
(86%). The authors concluded that lymphoscintigraphic sentinel node detection might play a role
in the management of squamous cell carcinoma
of the head and neck. They recommended that
randomized clinical trials be performed before
the technique is used widely.

SURGICAL MANAGEMENT OF THE NECK IN ORAL CANCER

Summary
Signicant renement has occurred since 1905
regarding surgical management of the neck in
patients with oral cancer. These renements have
been described and implemented in the best
interests of enhancing the potential cure of these
patients while also maintaining patients function
and quality of life. It is anticipated that further
developments will occur in this exciting surgical
discipline. In the meantime, surgeons must plan
treatment strategically for the clinically negative
neck with the same degree of enthusiasm as that
performed for the clinically positive neck, although with dierent techniques.

[12]

References

[17]

[1] Smith GI, OBrien CJ, Clark J, et al. Management

of the neck in patients with T1 and T2 cancer in
the mouth. Br J Oral Maxillofac Surg 2004;42:
494500.
[2] Shah JP. Head and neck surgery in crisis: preparing
for the future [commentary]. Arch Otolaryngol
Head Neck Surg 2005;113:55660.
[3] Crile G. Excision of cancer of the head and neck,
with special reference to the plan of dissection based
on one hundred and thirty-two operations. JAMA
1906;47:17806.
[4] Crile GW. On the surgical treatment of cancer of the
head and neck, with a summary of one hundred and
twenty-one operations performed upon one hundred
and ve patients. Transactions Southern Surgical
and Gynecology Association 1905;17:10827.
[5] Ferlito A, Rinaldo A. Errare humanum est, in errore
perseverare stultum: this is true also for neck dissection. Oral Oncol 2005;41:1324.
[6] Crile GW. Carcinoma of the jaws, tongue, cheek,
and lips. Surg Gynecol Obstet 1923;36:15962.
[7] Martin H, Del Valle B, Ehrlich H, et al. Neck dissection. Cancer 1951;4:44199.
[8] Nahum AM, Mullally W, Marmor L. A syndrome
resulting from radical neck dissection. Arch Otolaryngol 1961;74:826.
[9] Short SO, Kaplan JN, Laramore GE, et al. Shoulder
pain and function after neck dissection with or without preservation of the spinal accessory nerve. Am J
Surg 1984;148:47884.
[10] Robbins KT, Medina JE, Wolfe GT, et al. Standardizing neck dissection terminology: ocial report of
the Academys Committee for Head and Neck Surgery and Oncology. Arch Otolaryngol Head Neck
Surg 1991;117:6015.
[11] Robbins KT, Clayman G, Levine PA, et al. Neck
dissection classication update: revisions proposed
by the American Head and Neck Society and the
American Academy of Otolaryngology, Head and

[13]

[14]

[15]

[16]

[18]

[19]
[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

545

Neck Surgery. Arch Otolaryngol Head Neck Surg

2002;128:7518.
Medina JE. A rational classication of neck dissections. Otolaryngol Head Neck Surg 1989;100:
16976.
Myers EN, Fagan JJ. Treatment of the N neck
in squamous cell carcinoma of the upper aerodigestive tract. Otolaryngol Clin North Am 1998;31:
67186.
Khaf RA, Gelbsh GA, Asase DK. Modied radical neck dissection in cancer of the mouth, pharynx,
and larynx. Head Neck 1990;12:47682.
Bocca E, Pignataro O. A conservation technique in
radical neck dissection. Ann Otol Rhinol Laryngol
1967;76:97587.
Ferlito A, Rinaldo A, Robbins KT, et al. Changing
concepts in the surgical management of the cervical
node metastasis. Oral Oncol 2003;39:42935.
Bocca E, Pignataro O, Oldini C, et al. Functional
neck dissection: an evaluation and review of 843
cases. Laryngoscope 1984;94:9425.
Calearo CV, Teatini G. Functional neck dissection:
anatomical grounds, surgical technique, clinical observations. Ann Otol Rhinol Laryngol 1983;92:
21522.
Ariyan S. Functional radical neck dissection. Plast
Reconstr Surg 1980;65:76876.
Gavilan J, Gavilan C, Herranz J. Functional neck
dissection: three decades of controversy. Ann Otol
Rhinol Laryngol 1992;101:33941.
Persky MS, Lagmay VMP. Treatment of the clinically negative neck in oral squamous cell carcinoma.
Laryngoscope 1999;109:11604.
Yuen APW, Lam KY, Chan CL, et al. Clinicopathological analysis of elective neck dissection for N0
neck of early oral tongue carcinoma. Am J Surg
1999;177:902.
Ho CM, Lam KH, Wei WI. Occult lymph node metastasis in small oral tongue cancers. Head Neck
1992;14:35963.
Beenken SW, Krontiras H, Maddox WA. T1 and T2
squamous cell carcinoma of the oral tongue: prognostic factors and the role of elective lymph node dissection. Head Neck 1999;21:12430.
Henick DH, Silver CE, Heller KS, et al. Supraomohyoid neck dissection as a staging procedure for
squamous cell carcinomas of the oral cavity and oropharynx. Head Neck 1995;17:11923.
Medina JE, Byers RM. Supraomohyoid neck dissection: rationale, indications, and surgical technique.
Head Neck 1989;11:11122.
Kligerman J, Lima RA, Soares JR. Supraomohyoid
neck dissection in the treatment of T1/T2 squamous
cell carcinoma of oral cavity. Am J Surg 1994;168:
3914.
Kowalski LP, Magrin J, Waksman G, et al. Supraomohyoid neck dissection in the treatment of head
and neck tumors. Arch Otolaryngol Head Neck
Surg 1993;119:95863.

546

CARLSON & MILLER

[29] Spiro JD, Spiro RH, Shah JP, et al. Critical assessment of supraomohyoid neck dissection. Am J
Surg 1988;156:2869.
[30] Jalisi S. Management of the clinically negative neck
in early squamous cell carcinoma of the oral cavity.
Otolaryngol Clin North Am 2005;38:3746.
[31] Shah JP, Candela FC, Poddar AK. The patterns of
cervical lymph node metastasis from squamous carcinoma of the oral cavity. Cancer 1990;66:10913.
[32] Shah JP, Singh SG. Cervical lymph nodes. In: Head
and neck surgery and oncology. 3rd edition. Edinburgh: Elsevier; 2003. p. 35394.
[33] Crean SJ, Joman A, Potts J, et al. Reduction of
occult metastatic disease by extension of the supraomohyoid neck dissection to include level IV. Head
Neck 2003;25:75862.
[34] Byers RM, Weber RS, Andrews T, et al. Frequency
and therapeutic implications of skip metastases in
the neck from squamous carcinoma of the oral
tongue. Head Neck 1997;19:149.
[35] Manni JJ, van den Hoogen FJA. Supraomohyoid
neck dissection with frozen section biopsy as a staging procedure in the clinically node-negative neck in
carcinoma of the oral cavity. Am J Surg 1991;162:
3736.
[36] Van den Hoogen FJA, Manni JJ. Value of the supraomohyoid neck dissection with frozen section analysis as a staging procedure in the clinically negative
neck in squamous cell carcinoma of the oral cavity.
Eur Arch Otorhinolaryngol 1992;249:1448.
[37] Woolgar JA. Pathology of the N0 neck. Br J Oral
Maxillofac Surg 1999;37:2059.
[38] Muzaar K. Therapeutic selective neck dissection:
a 25-year review. Laryngoscope 2003;113:14605.
[39] Shah JP, Andersen PE. Evolving role of modications in neck dissection for oral squamous carcinoma. Br J Oral Maxillofac Surg 1995;33:38.
[40] Byers RM. Modied neck dissection: a study of 967
cases from 19701980. Am J Surg 1985;150:41421.
[41] Snow GB, Annyas AA, Van Slooten EA, et al. Prognostic factors of neck node metastasis. Clin Otolaryngol 1982;7:18592.
[42] Anderson PE, Spiro RH, Cambronero E, et al. The
role of comprehensive neck dissection with

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

preservation of the spinal accessory nerve in the clinically positive neck. Am J Surg 1994;168:499502.
Robbins KT, Samant S. Neck dissection. In:
Cummings CW, editor. Cummings otolaryngology
head and neck surgery. 4th edition. Philadelphia:
Mosby; 2005. p. 261445.
Myers EN, Gastman BR. Neck dissection: an operation in evolution. Arch Otolaryngol Head Neck
Surg 2003;129:1425.
Morton DL, Wen D, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early
stage melanoma. Arch Surg 1992;127:3929.
Ambrosch P, Kron M, Pradler O, et al. Ecacy of
selective neck dissection: a review of 503 cases of
elective and therapeutic treatment of the neck in
squamous cell carcinoma of the upper aerodigestive
tract. Otolaryngol Head Neck Surg 2001;124:
1807.
Van den Brekel MWM, van der Wall I, Meijer
CJLM, et al. The incidence of micrometastases
in neck dissection specimens obtained from elective neck dissections. Laryngoscope 1996;106:
98791.
Enepekides DJ, Sultanem K, Hguyen C. Occult cervical metastases: immunoperoxidase analysis of the
pathologically negative neck. Otolaryngol Head
Neck Surg 1999;120:7137.
Woolgar JA. Micrometastasis in oral/oropharyngeal squamous cell carcinoma: incidence, histopathological features and clinical implications. Br J Oral
Maxillofac Surg 1999;37:1816.
Shoaib T, Soutar DS, MacDonald DG, et al. The accuracy of head and neck carcinoma sentinel lymph
node biopsy in the clinically N0 neck. Cancer 2001;
91:207783.
Werner JA, Dunne AA, Ramaswamy A, et al. The
sentinel node concept in head and neck cancer: solution for the controversies in the N0 neck? Head Neck
2004;26:60311.
Payoux P, Dekeister C, Lopez R, et al. Eectiveness
of lymphoscintigraphic sentinel node detection for
cervical staging of patients with squamous cell carcinoma of the head and neck. J Oral Maxillofac Surg
2005;63:10915.

Oral Maxillofacial Surg Clin N Am 18 (2006) 547563

Sentinel Lymph Node Biopsy

in the Staging of Oral Cancer
Thomas D. Shellenberger, DMD, MDa,b
a

Head and Neck Surgical Oncology, M. D. Anderson Cancer Center Orlando,

1400 South Orange Avenue, MP 760, Orlando, FL 32806, USA
b
Head and Neck Surgical Oncology, The University of Texas M. D. Anderson Cancer Center,
1515 Holcombe Boulevard, Houston, TX 770304009, USA

Sentinel lymph node biopsy (SNB) is a technique under investigation for the staging of the
regional lymph nodes in patients with squamous
cell carcinoma of the oral cavity. SNB oers the
potential of improving the accuracy of regional
staging and reducing the morbidity of elective
node dissection in patients without metastases.
Thus, the use of SNB might signicantly advance
the treatment of oral cancer. SNB is generally
accepted for the staging of melanoma and breast
cancer and has shown encouraging preliminary
data for the staging of oral cancer. Acceptance of
SNB in the staging of oral cancer requires careful
consideration of the technique and a critical review of the studies that tested the accuracy of
SNB for primary tumors of the oral cavity.
Therefore, the rationale, diagnostic ecacy, and
remaining obstacles for SNB in the staging of oral
cancer are discussed.
Rationale for sentinel lymph node biopsy
in oral cancer
Eective management of patients with squamous cell carcinoma of the oral cavity depends on
accurate staging to determine the prognosis and to
select appropriate therapeutic strategies. The stage
of disease depends highly on the status of the
regional cervical lymph nodes at risk for metastasis from the primary tumor. Cervical metastasis
is the single most important prognostic factor in
The author is supported by an American Society of
Clinical Oncology Young Investigator Award.
E-mail address: tdshelle@mdanderson.org

patients with oral cancer, with the presence of

nodal spread decreasing the 5-year disease-free
survival rate by approximately 50% [13]. Moreover, the recent advances in multimodal therapy
have been reached by appropriately selecting patients with regional disease for radiation therapy
[4,5] and chemoradiation [6,7]. Further advances
in adjuvant and neoadjuvant therapy can be expected to rely even more on identifying the patients who are most likely to benet from new
strategies. Thus, achieving the goals in management of oral cancer requires accurately determining the progression of disease by staging the
regional lymphatics.
Although clinically evident cervical metastases
are found at presentation in approximately 30%
of patients, occult metastasis occurs in 20% to
30% of patients with oral cancer who have no
clinical evidence of regional disease [8]. This disparity between clinical and pathologic lymph
node staging was underscored by a study of 266
patients with squamous cell carcinoma of the
oral tongue who underwent primary surgery at
the University of Texas M. D. Anderson Cancer
Center [9]. Pathologic evidence of regional metastasis was found in 34% of clinically N0 patients;
moreover, extracapsular spread was present in
nearly 20% of those upstaged. Selecting the best
management options for these patients requires
accurately identifying subclinical cervical metastases. The accuracy of the currently available means
of evaluating the neck before treatment is limited,
however. In the detection of nodal disease, imaging techniques, including ultrasonography, CT,
MRI, and positron emission tomography, lack

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.012

oralmaxsurgery.theclinics.com

548

SHELLENBERGER

the sensitivity and specicity necessary to guide

treatment decisions in patients staged clinically
as N0. As a result of imprecision in the staging
of patients before treatment, inappropriately selected management of the clinically N0 neck can
result in morbidity and mortality.
The management options for the clinically N0
neck include (1) selective neck dissection with the
rationale of regional staging and elective treatment, (2) irradiation of the neck as elective
treatment, and (3) clinical follow-up with therapeutic neck dissection or irradiation reserved for
patients who develop detectable disease. Elective
neck dissection is warranted in patients with
a 15% to 20% or greater risk of occult metastasis
based on features of the primary tumor (Fig. 1).
At The University of Texas M. D. Anderson Cancer Center, patients with oral cancer undergo selective dissection of the clinically N0 neck if they
have primary tumors greater than 2 cm in diameter or a depth of invasion of more than 4 mm or if
their primary tumor has pathologic features associated with metastasis, such as perineural or

lymphovascular invasion [10]. The benets of elective neck dissection are realized when disease is
found on pathologic evaluation of the specimen
and include the early removal of metastatic disease and more accurate staging information on
which to base adjuvant radiation therapy. In up
to 65% to 75% of patients, however, no disease
is found, and the morbidity associated with neck
dissection is unjustied. Likewise, elective irradiation of the neck overtreats most patients, and even
worse, it provides no staging information on
which to estimate prognosis or guide further management. Moreover, few treatment alternatives
are left in those who develop second primary tumors. Watchful waiting may allow disease to present at a more advanced stage in the neck, making
salvage treatment more dicult [11,12]. Thus,
more precise staging before treatment is requisite
to prevent the consequences of inappropriately selected management strategies for the clinically N0
neck in oral cancer.
The need for optimal management of regional
nodes at risk for metastasis from other cancers led
to a search for more accurate and less invasive
approaches of staging. As a result, SNB has
become an accepted staging tool in the management of melanoma and breast cancer. The procedure accurately stages the regional lymphatics
based on the status of the rst-echelon nodes in
the lymphatic basin draining the primary tumor
site while limiting morbidity caused by unnecessary lymph node dissection. The potential of SNB
to stage head and neck malignancies, including
oral cancer, is currently under investigation.
Successful application of the procedure would
bring a major advance to the staging and management of oral cancer.
Premise for staging oral cancer with sentinel
lymph node biopsy

Fig. 1. The usual extent of selective neck dissection

(supraomohyoid neck dissection) for oral cancer, including levels I, II, and III (shaded). For primary tumors of
the oral tongue, selective neck dissection is extended to
level IV to include additional nodes at risk. For primary
tumors involving the oor of the mouth, ventral surface
of the tongue, or midline of the tongue, bilateral selective
neck dissection is performed to include lymph nodes at
risk on both sides of the neck. (From Byers RM, Weber
RS, Andrews T, et al. Frequency and therapeutic implications of skip metastases in the neck from squamous
carcinoma of the oral tongue. Head Neck 1997;19:15;
with permission.)

The potential use of SNB for the staging of

oral cancer developed from the success of the
procedure in malignant melanoma and breast
cancer. SNB has improved the accuracy of staging
while reducing the morbidity caused by unnecessary lymphadenectomy in these cancers. As a result, SNB has been investigated in the staging of
other cancers, including upper gastrointestinal,
lung, urogenital, and colorectal carcinomas. In
parallel, the procedure has been studied in the
staging of squamous cell carcinoma of various
sites in the head and neck, including the oral
cavity.

SENTINEL LYMPH NODE BIOPSY IN STAGING

Sentinel lymph node concept

The staging of cancer by sentinel lymph node
identication and biopsy is based on the concept
that metastasis from a primary tumor occurs by
predictable orderly spread to the rst-echelon
nodes before reaching nodes in the remainder of
the lymphatic basin. Therefore, histopathologic
examination of the sentinel node can dene the
disease status of the entire regional lymphatic
nodal basin. A detailed pathologic examination of
the sentinel lymph node can increase the accuracy
of staging. A sentinel node demonstrating metastasis can identify patients most likely to benet
from therapeutic lymph node dissection. Furthermore, the decision to proceed with dissection is
based on documented evidence of metastasis
rather than on the probability of metastasis
implied by less certain clinical or radiographic
data. Moreover, a true-negative sentinel node can
spare patients from dissection, with minimal risk
of disease recurrence in the neck at an advanced
stage.
In 1955, Seaman and Powers [13] laid the
groundwork for lymphoscintigraphy and lymphatic mapping. They rst described the peritumoral injection of radioactive colloidal gold and
showed its uptake and distribution in regional
lymph nodes of breast cancer patients. Soon after,
Gould and colleagues [14] independently observed
that the routine excision of a sentinel node
found at the origin of the common facial vein at
the time of parotidectomy oered diagnostic signicance. The report suggested that the sentinel
node accurately predicts the disease status of a regional nodal basin and proposed that intraoperative frozen section analysis of the node could
guide the decision for radical neck dissection for
parotid malignancies.
The concepts of these early ndings were
drawn together in 1977, when Mortons group
[1517] introduced cutaneous lymphoscintigraphy
as a means of identifying the regional lymphatic
basin of primary drainage at risk for metastasis
from melanomas. In 1992, the same investigators
developed and rened the technique of selectively
identifying the sentinel node by intraoperative
mapping with blue dye and radiotracer injected
around the primary tumor [18]. Further, they reported on 223 patients with clinical stage I melanoma who underwent intraoperative lymphatic
mapping and en bloc lymphadenectomy. A sentinel node was successfully identied in 82% of
lymph node basins, and metastasis was

549

demonstrated in 21% of these specimens. Moreover, the remainder of the lymph node basin
was the sole site of metastasis in only two specimens with negative sentinel lymph nodes, establishing a false-negative rate of less than 1% for
SNB. Thus, SNB was established as accurately
identifying patients with early-stage melanoma
who have nodal metastasis and are likely to benet from radical lymphadenectomy.
Evolution of the sentinel lymph node biopsy
Since the seminal report of Morton and
colleagues [18], the technique of SNB for regional
staging of cutaneous melanoma has undergone
further evolution and renement and is reaching
worldwide acceptance. The currently available
technique allows identication of sentinel lymph
nodes in 97% of patients with a sensitivity of
more than 95% and a false-negative rate of less
than 2% [19]. This high level of accuracy allows
the appropriate selection of patients for therapeutic lymph node dissection so as to improve regional disease control and for consideration in
trials of adjuvant therapy. Furthermore, patients
with pathologically negative sentinel lymph nodes
who would not benet from elective lymphadenectomy can be spared the morbidity of surgery
and can condently undergo close clinical observation. Thus, the staging of regional lymphatics
by SNB has revolutionized the management of
patients with early-stage cutaneous malignant
melanoma.
The technique has been standardized for the
staging of cutaneous melanoma, allowing for the
comparison of data between multiple institutions
participating in trials. The statistical power
achieved by that approach has hastened research
advances in the management of melanoma. Moreover, accumulation of experience and long-term
follow-up data has shown that sentinel node
status is a strong predictor of survival. For
example, Jansen and coworkers [20] showed an
overall survival rate of 93% at 3 years when the
sentinel node was negative and a 67% survival
rate when the sentinel node was positive.
A more detailed examination of sentinel lymph
nodes by serial sectioning and immunohistochemical analysis more precisely identies the subset of
patients with micrometastatic regional disease and
improves the accuracy of staging. In addition,
a new classication for lymph node disease has
emerged to dierentiate microscopic from macroscopic disease. One study has shown that patients

550

SHELLENBERGER

with clinically negative but microscopically occult

metastatic disease who undergo lymph node
dissection have a better prognosis than those
who undergo delayed resection of clinically evident macroscopic disease [21]. A multi-institutional study of patients with stage I and II
melanoma showed that the status of the sentinel
lymph node was the strongest predictor of disease-free survival [22]. These ndings prompted
major revisions in the American Joint Committee
on Cancer staging system for melanoma in 2002
[23].
Although SNB has improved the accuracy of
staging for melanoma, the benets extend beyond
those of quality in patient care. Establishing SNB
as a standard of care in certain groups of patients
has resulted in cost savings for patients, providers,
payers, employers, and industry [24]. Cost savings
result from SNB performed as outpatient surgery
and from SNB results obviating elective lymphadenectomy in a substantial number of patients.
Furthermore, the costs of adjuvant therapy are reduced by targeting a highly selected subset of patients based on SNB.
The technique of SNB was developed in the
staging of breast cancer with a rationale similar to
that of melanoma: that axillary lymph node
dissection may constitute excessive treatment in
early-stage disease. That supposition is underscored by the availability of modern screening
methods that have improved the detection of
breast cancer in early stages. Moreover, the safety
and ecacy of SNB in reducing the need for
complete axillary dissection have been demonstrated in breast cancer patients with small
primary tumors. The results of a randomized
study comparing SNB with axillary lymph node
dissection in 616 patients showed that SNB had
an overall accuracy of 96.9% and a sensitivity of
91.2% [25]. Clinical follow-up showed that patients with negative SNB results who did not undergo axillary dissection had a high regional
control rate and no axillary metastases. In addition, less pain and better arm mobility were found
in the patients who underwent SNB alone. Thus,
the use of SNB to stage breast cancer is rapidly
gaining acceptance at cancer centers worldwide.
Sentinel lymph node biopsy in cancers
of the head and neck
Although studies examining the use of SNB to
stage cutaneous malignant melanoma have focused mainly on primary tumors of the trunk and

extremities, the studies have also shown the use of

SNB to be feasible in tumors of the head and
neck. Moreover, several cancer centers have recently reported sentinel node identication rates
of 96% to 99% from SNB of head and neck
primary melanomas [2628]. At M. D. Anderson
Cancer Center, a prospective study was conducted
for melanoma of the head and neck with intraoperative lymphatic mapping and SNB in the context of elective neck dissection [29]. In 43 patients,
the sentinel node identication rate was 98%, and
no patients with negative sentinel nodes had disease in the neck dissection specimen.
Although this study demonstrated the accuracy of SNB in head and neck melanoma, it also
described some of the diculties of the procedure
that are unique to primary tumors of the head and
neck. The study observed that tumors of the head
and neck had a multiplicity of sentinel nodes
(a mean 3.6 nodes per patient), widespread
distribution of sentinel nodes (2.2 basins per
patient), and frequent localization of sentinel
nodes to the parotid gland (44% of patients).
The Sunbelt Melanoma Trial demonstrated similar ndings and reported a signicantly higher
false-negative rate with SNB for tumors at head
and neck sites compared with that of SNB for
tumors of the trunk and extremities [30]. Thus,
these studies suggest that the value of SNB may
be limited in many patients with primary melanoma of the head and neck, for whom parotidectomy and selective neck dissection are indicated.
Therefore, the role of SNB for melanoma of the
head and neck has not yet been fully dened [31].
For squamous cell carcinoma of the head and
neck, the rst successful identication of a sentinel
lymph node was performed in 1996 by Alex and
Krag [32] in a patient with a laryngeal primary tumor. Shortly thereafter, the procedure was introduced for oral cancer by Koch and colleagues
[33] and Pitman and coworkers [34]. Although
nding the technique feasible in selected cases,
Koch and colleagues [33] noted that radiocolloid
could not identify sentinel nodes with radiocolloid
in the presence of metastases. Furthermore, Pitman and coworkers [34] observed that isosulfan
blue dye could not be seen in the lymphatics and
cervical lymph nodes. By combining blue dye
and radiocolloid, however, Shoaib and colleagues
[35] successfully identied sentinel lymph nodes in
16 of 17 patients with oral and oropharyngeal primary tumors. Moreover, the SNB correctly identied the pathologic status of the neck in 9 of 9
patients with oral cavity primary tumors and

SENTINEL LYMPH NODE BIOPSY IN STAGING

clinically N0 necks, thereby demonstrating the accuracy of SNB. The discovery of the potential for
SNB in the staging of head and neck cancer has
provided the foundation for further study to rene
the technique.
In subsequent investigations, the indications
for SNB in oral cancer have emerged to focus
study on the patients most likely to benet from
the procedure. The sentinel lymph node concept in
carcinoma of the oral cavity is based on the
drainage of the primary tumor by collecting
vessels that reach rst-echelon nodes of the regional basin (Fig. 2). For patients with small primary tumors of the oral cavity, the injection of
tracer is technically easier and may be accomplished under topical or local anesthesia. Therefore, preoperative lymphoscintigraphy can be
performed in the nuclear medicine suite before
the patient is brought to the operating room.
Moreover, unlike patients with bulky invasive primary tumors, those with small primary tumors
pose the greatest uncertainty regarding the regional nodal status. Likewise, patients without
clinical evidence of regional disease are more
likely to undergo successful lymphoscintigraphy
and sentinel node mapping. In contrast, clinically
positive nodes and even grossly positive nodes
that are missed on clinical evaluation are dicult
to identify by sentinel node mapping. Such diseased nodes often uptake tracer poorly or divert

551

lymphatic ow altogether, resulting in the labeling

of downstream nodes [33]. Therefore, in patients
with T1 or T2 oral cavity carcinoma, SNB is
most applicable for staging the following: (1) the
ipsilateral clinically N0 neck drained by a unilateral primary tumor, (2) the bilateral clinically
N0 neck drained by a midline tumor or tumor
crossing the midline, and (3) the contralateral clinically N0 neck drained by a midline tumor or tumor crossing the midline in the presence of
a clinically positive ipsilateral neck (Box 1).
In oral cancer, staging of the clinically N0 neck
by SNB initially progressed along a single methodologic approach in which patients undergo
SNB-assisted elective neck dissection, which combines sentinel lymph node identication and
excision with a standard selective neck dissection
[36]. This approach underlies most published pilot
studies designed to validate the accuracy of SNB
in reference to the pathologic ndings of the elective neck dissection specimen. Further, SNB-assisted elective neck dissection may result in a higher
stage than would have occurred with elective neck
dissection because of sensitive techniques of pathologic evaluation of the sentinel node. Indeed, the
additional pathologic methods of serial sectioning
and immunohistochemical analysis of sentinel
nodes have upstaged up to 8% of neck dissection specimens in patients with head and neck

Box 1. Application of sentinel lymph

node biopsy for the staging of oral
cancer
T1 or T2 tumors

Tongue

Retromolar trigone

Buccal mucosa

Lip

Alveolar ridge

Hard palate

Floor of the mouth

Fig. 2. The collecting lymphatic vessels of the tongue

and regional nodal groups. The collecting vessels receive
lymphatic uid from the mucosal and muscular networks and drain to the nodal groups of the submental
and submandibular triangles as well as the internal jugular chain. (From Droulias C, Whitehurst JO. The lymphatics of the tongue in relation to cancer. Am Surg
1976;42(9):671; with permission.)

Clinically N0 neck

Ipsilateral neck for a unilateral primary
tumor

Bilateral neck for a midline tumor or
tumor crossing the midline

Contralateral neck for a midline tumor
or tumor crossing the midline in the
presence of a clinically positive
ipsilateral neck

552

SHELLENBERGER

squamous cell carcinoma [37,38]. The clinical

signicance of upstaging by this methodology
remains to be determined by studies with longterm follow-up, however. Thus, the role of SNBassisted elective neck dissection in the management of oral cancer remains unclear.
More recently, a second methodologic approach in the staging of the clinically N0 neck in
oral cancer has emerged with the use of SNB as
a stand-alone procedure, potentially replacing
elective neck dissection. SNB alone involves
sentinel lymph node identication, excision of
only the sentinel nodes, and additional pathologic
evaluation of the specimen most likely to contain
metastasis [36]. The advantage of this approach is
that it is minimally invasive and can potentially be
more sensitive in pathologic staging. On the basis
of the sentinel node status, patients then undergo
a therapeutic neck dissection or clinical follow-up.
Despite early reports of technical diculties
with SNB, the studies of SNB in mucosal cancers
of the upper aerodigestive tract have spread
widely. By 2001, an international conference on
SNB in head and neck cancer was convened by
investigators with experience in the eld from 21
cancer centers around the world [39]. The pilot
studies of these centers culminated in a European
multicenter trial using a standardized technique to
allow meaningful comparisons. Moreover, the
College of Surgeons Oncology Group initiated
a multicenter prospective trial (Protocol Z0360)
to compare SNB with elective neck dissection in
the staging of patients with oral cavity carcinoma
[40]. Furthermore, at the end of 2005, a search of
the medical literature found 153 articles that described the use of SNB in head and neck squamous cell carcinoma [41]. Thus, sucient
experience with SNB is available to evaluate the
technical considerations, results, and future promise of the procedure in the staging of oral cancer.

Technique of sentinel lymph node biopsy

in oral cancer
In patients with oral cancer, the procedure for
SNB is usually performed in coordination with
excision of the primary tumor and elective neck
dissection. At our institution, on the day of
surgery, the patient undergoes peritumoral injection of radioactive tracer, followed by scintigraphic imaging in the nuclear medicine suite to
provide a map of the regional lymphatics. During surgery, sentinel nodes are identied by

intraoperative detection with a handheld gamma

probe alone or with the injection of vital blue dye.
Radioactive or blue sentinel lymph nodes are then
excised and submitted for histopathologic evaluation. Successful sentinel node identication,
excision, and interpretation require close collaboration between the surgeon, the nuclear radiologist, and the pathologist, with each member of the
team playing a critical role (Box 2) [42].
Preoperative lymphoscintigraphy
Under topical or local anesthesia, radiocolloid
tracer is injected into the submucosa around the
periphery of the entire primary tumor (Fig. 3).
The tracer is injected by using a tuberculin syringe
with a 25-gauge needle. Small-particlesized technetium-99mlabeled sulfur colloid (99mTc-SC)
travels suciently through the interstitial space
and undergoes uptake by the submucosal lymphatic vessels to reach the regional lymphatic
channels. The tracer travels quickly to the rstechelon sentinel lymph nodes, where it becomes
trapped and ltered through the nodes before
reaching the second- and third-echelon nodes at
a lower radioactivity level.
Tracers of varying pharmacokinetic and pharmacodynamic properties have been reported in
studies of lymphoscintigraphy for oral cancer.
Small-particle colloids like 99mTc-SC are more
likely to pass down the lymphatic channels to penetrate the sentinel node in the presence of disease
or blocked lymphatics. Further advantages of
99m
Tc-SC include low overall exposure to radiation, a short half-life, and an energy peak within
the range of detection of most gamma cameras
and handheld gamma probes [43,44]. Although
a range of doses and volumes of tracer have
been reported, a 400-mCi dose [45] is generally sufcient and can be delivered in a volume of approximately 0.2 to 0.4 mL.
Lymphoscintigraphic imaging serves as a road
map for planning the surgical procedure (Fig. 4).

Box 2. Technique of sentinel lymph

node biopsy in oral cancer
1.
2.
3.
4.

Preoperative lymphoscintigraphy
Intraoperative lymphatic mapping
Excision of sentinel lymph nodes
Pathologic evaluation of sentinel
lymph nodes

SENTINEL LYMPH NODE BIOPSY IN STAGING

Fig. 3. A T2 squamous cell carcinoma primary tumor of

the lateral border of the tongue appropriate for the peritumoral injection of radiocolloid tracer under topical or
local anesthesia.

The resulting images identify the nodal basins at

risk for metastasis and the location of sentinel
lymph nodes within the basins. Dynamic and
static scans are obtained after injection of the
tracer. The areas of focal uptake in the cervical
lymphatics corresponding to the sentinel lymph
nodes are marked on the patients skin. Because
the location of sentinel nodes may vary within
a basin, marking the position of the sentinel
node in reference to other nodes facilitates the design of an appropriate incision.
Intraoperative lymphatic mapping
After lymphoscintigraphy is performed, the
head and neck region is prepared and draped for
surgery in the operating room with the patient
under general anesthesia. Whenever possible,
the primary tumor is excised rst to reduce

553

background radioactivity. The handheld gamma

probe is calibrated for the detection of 99mTc-SC.
The neck is then probed transcutaneously to conrm the location of the radioactive lymph nodes
that were identied by lymphoscintigraphy and
marked on the skin (Fig. 5A). The radioactivity
of the hot spot relative to the background levels
(in vivo activity ratio) is recorded. A skin incision
is then designed for adequate access in harvesting
the sentinel lymph nodes and for incorporation
into an incision for elective neck dissection.
If vital blue dye is used to identify sentinel
lymph nodes, 0.5 to 1.0 mL is injected into the
submucosa around the periphery of the primary
tumor before excision. The most investigated dye
in oral cancer is isosulfan blue in a 1% aqueous
solution (Lymphazurin; United States Surgical
Corporation, North Haven, Connecticut). This
dye enters the lymphatics rapidly, with minimal
diusion into the surrounding tissues. When the
dye enters a sentinel lymph node, it stains part
of the node pale blue, thus clearly distinguishing
the sentinel node from the surrounding nonsentinel nodes.
Vital blue dye and radiocolloid are complementary techniques in lymphatic mapping; when
they are used simultaneously, the probability of
success in identifying the sentinel node increases.
Moreover, blue dye may permit sentinel node
identication when radiocolloid fails, such as
when the sentinel nodes are located in lymphatic
basins lying close to the primary tumor. The
resultant shine-through of radioactivity increases
background levels that interfere with gamma

Fig. 4. Static lymphoscintigraphic images obtained after the injection of 99mTc-SC at a dose of 500 mCi around the periphery of a T2 squamous cell carcinoma of the left lateral border of the tongue. (A) There is evidence of drainage to
middle and lower left cervical nodes. (B) A large star artifact around the injection site might obscure more proximal
lymph node drainage close to the primary tumor site, however.

554

SHELLENBERGER

Fig. 5. Intraoperative lymphatic mapping and excision of sentinel lymph nodes. (A) Neck is probed transcutaneously to
conrm the location of radioactive lymph node(s) previously identied by lymphoscintigraphy. (B) Surgical dissection is
aided by the handheld gamma probe to identify a sentinel lymph node dened by probe counts with an activity ratio of
3:1 or higher in vivo. (C) After excision, the radioactivity of the node is measured with the probe to determine an ex vivo
activity ratio and conrm correct identication of the sentinel node. (D) Basin is reprobed for residual radioactivity to
verify that all sentinel nodes have been removed.

probe identication of the sentinel node. Conversely, the mapping of lymphatic basins with
blue dye alone has not resulted in adequate rates
of sentinel node identication. The use of isosulfan blue for lymphatic mapping in oral cancers
has thus not been widespread [41]. In addition,
isosulfan blue has been associated with a signicant incidence of allergic reactions (range: 1%
3%), including life-threatening anaphylaxis [46].
Excision of sentinel lymph nodes
The skin is incised, and the subcutaneous
tissues are divided to the level of the platysma.
The muscle is then incised, and subplatysmal aps
are elevated to provide sucient access. Surgical
dissection is aided by the handheld gamma probe,
which directs the dissection down to the sentinel
node (see Fig. 5B). Blue-stained aerent lymphatic vessels that lead to a blue node may aid
in identication further. A sentinel node is dened
as a node with an activity ratio of 3:1 or higher
in vivo (counts at least three times the background levels). Once identied, the entire node is
removed by sharp dissection or electrocautery.

Surrounding nonsentinel lymph nodes can be removed to provide an internal control. The radioactivity level of all excised nodes is measured
with the probe and recorded (see Fig. 5C). A ratio
of radioactivity in an excised sentinel node relative
to a nonsentinel lymph node (ex vivo activity ratio) of 10:1 or higher further conrms correct
identication of the sentinel node. Additional nodes with in vivo counts greater than 10% of the
hottest node are removed. The basin is reprobed
for residual radioactivity to verify that all sentinel
nodes were removed (see Fig. 5D). If radioactivity
has not fallen to background levels, further probedirected dissection is warranted.
Pathologic examination of sentinel lymph nodes
After the sentinel nodes are harvested, specimens with a high likelihood of harboring metastasis are submitted for detailed pathologic
evaluation. The limited material from an SNB
allows a focused analysis with a far more detailed
search for metastasis than that possible from
a neck dissection specimen. The Second International Conference on Sentinel Node Biopsy in

SENTINEL LYMPH NODE BIOPSY IN STAGING

Mucosal Head and Neck Cancer [47] resulted in

the following consensus recommendations. To begin pathologic evaluation, the xed node is bisected through the long axis and divided into
slices of 2.5 mm. The entire node is then serially
step-sectioned at 150-mm intervals; four sections
are cut and mounted from each step. The rst section is stained with hematoxylin and eosin (H&E)
for conventional analysis. The nding of metastasis signies pathologic nodal positivity. In the absence of metastasis, the second section is stained
with an antibody to pancytokeratin for immunohistochemical analysis. In the absence of cytokeratin staining, the node is deemed pathologically
nodal-negative. If cytokeratin staining is found,
the third section is stained with H&E to conrm
the presence of viable tumor cells. The fourth section is reserved for evaluation that may be necessary later. This technique is expected to identify
all metastases and micrometastases.
Intraoperative frozen section analysis of sentinel lymph nodes oers the potential to decide
immediately whether to perform neck dissection
during the same procedure. This benet must be
weighed against the potential of compromising
the denitive pathologic evaluation by losing
material from the sentinel node because of frozen
section processing. Although the feasibility and
accuracy of frozen section analysis of the sentinel
node have been suggested in oral cancer [48], further evaluation is required to reach a consensus.

555

the procedure in determining the status of the

neck (Box 3).

Sentinel lymph node identication

The technical feasibility of SNB is determined
by the likelihood of identifying sentinel lymph
nodes with preoperative lymphoscintigraphy and
intraoperative detection using a handheld gamma
probe. Data from institutions worldwide demonstrate high rates of sentinel lymph node identication for head and neck squamous cell
carcinoma, which suggests that the procedure is
feasible.
Pilot studies from 20 centers contributing to
the Second International Conference on Sentinel
Node Biopsy in Mucosal Head and Neck Cancer
were collectively analyzed [47]. In these studies,
sentinel lymph nodes were identied by the peritumoral injection of radioactive tracer, followed by
preoperative lymphoscintigraphy and intraoperative detection with a handheld gamma probe
alone or with the aid of blue dye. Among 379 patients with clinically N0 disease, sentinel lymph
nodes were identied in 366 patients, for an
identication rate of 97%. Some variations in
the technique of lymphatic mapping were attributable to dierences in available radiotracers and
scanning equipment. Nevertheless, the published
rates were comparable between the institutions.
Stoeckli and coworkers [47] attributed the successful identication of sentinel nodes to overcoming
a learning curve and noted the paramount

Diagnostic ecacy of sentinel lymph node biopsy

in oral cancer
The lower incidence of head and neck squamous carcinoma compared with melanoma and
breast cancer poses a signicant obstacle in the
study of regional staging with SNB. Although
a number of centers around the world have
performed pilot studies of SNB in oral cancer,
no large single-institution studies are available.
Nonetheless, conclusions on the diagnostic ecacy of sentinel node procedures can be drawn
from reports of the cumulative experience of
multiple international centers. Moreover, new
insight comes from a recent meta-analysis of the
world literature and from preliminary results of
a European multicenter prospective trial. The
results of SNB must be evaluated to determine
the feasibility of SNB to identify sentinel lymph
nodes, the extent to which metastasis can be
detected within those nodes, and the accuracy of

Box 3. Reported diagnostic efficacy

of sentinel lymph node biopsy in oral
cancer
Sentinel lymph node identification rate

96%98% for SNB-assisted elective
neck dissection

91% for SNB alone in the Canniesburn
trial
Detection of occult metastasis

34%60% of patients upstaged
Accuracy

Sensitivity: 86%100%

Negative predictive value: 83%99%

Likelihood ratio: 0.09 in the
Canniesburn trial

556

SHELLENBERGER

importance of obtaining comparable results for

any center performing the procedure.
A systematic review and diagnostic metaanalysis of all published literature on SNB in
head and neck cancer through 2003 were performed by Paleri and colleagues [41]. The analysis
included 301 patients with oral cavity primary tumors from 19 studies [33,45,4965] since 1999. Despite a relatively low number of patients and the
methodologic limitations inherent in diagnostic
meta-analysis, the overview provided valid results.
The authors addressed a focused question and selected articles by appropriate inclusion criteria
and quality assessment. The results revealed an
overall sentinel lymph node identication rate of
97.7%. The mean number of nodes removed per
patient was 1.6. As in the report of Stoeckli and
coworkers [47], a high rate of sentinel node identication was found despite variations in technique,
including a wide range in the dose of radioactive
tracer.
In 2004, preliminary results were published
from a multicenter trial (the Canniesburn trial)
begun in 1998 that included the largest cohort of
patients studied thus far [66]. The study described
134 SNB procedures using a standardized technique at six centers in the United Kingdom, Denmark, Spain, Italy, and Germany. Staging was
performed on patients with clinically N0 necks
who had T1 or T2 tumors of the oral cavity and
oropharynx. Preoperative lymphoscintigraphy
was combined with intraoperative detection,
which used a handheld gamma probe and injection of blue dye. Overall, a sentinel lymph node
was successfully identied in 93% of patients.
Sentinel lymph nodes were removed in SNB-assisted elective neck dissection and as a stand-alone
procedure with success rates of 96% and 91%,
respectively.
Success in sentinel node identication has been
shown to correlate with a surgeons level of
experience with SNB. The concept of a learning
curve was rst described in Morton and coworkers seminal 1992 paper [18] in which the
rate of identication as well as the sensitivity increased with the increasing number of procedures
performed. The First International Conference of
Head and Neck Sentinel Node Biopsy also found
a signicantly higher rate of identication at centers that had performed at least 10 procedures
[39]. Clearly, the standardization of SNB in melanoma has set a level that must be met if the technique is to be applied in oral cancer. Moreover,
any center embarking on an investigation of the

technique must validate the procedure before its

application in patients.
Detecting occult metastasis by sentinel
lymph node biopsy
If SNB is feasible in detecting metastatic oral
cancer, the next step in evaluating SNB lies in
determining the extent to which occult metastases
are identied. The nding of occult metastasis
within the sentinel node on pathologic evaluation
results in upstaging of the clinically negative N0
neck. The rate of upstaging by SNB must be
compared with that by elective neck dissection,
which is currently the best available method of
detecting occult metastasis. With traditional pathologic evaluation of specimens stained with H&E,
elective neck dissection upstages approximately
30% of patients.
Many of the early validation studies evaluated
sentinel lymph nodes by traditional methods with
H&E staining. Recently published studies have
used a variety of pathologic methods to search for
metastatic disease in sentinel nodes, however. In
the Canniesburn trial [66], sentinel lymph nodes
were examined by a protocol that included routine
H&E staining, serial sectioning, and immunohistochemical analysis for cytokeratin. In that trial,
upstaging of disease occurred in 34% of cases.
Nodal metastasis was identied by H&E staining
alone in 26% of cases and by additional pathologic means in 11% of cases. In a subgroup of patients undergoing SNB in the context of an
elective neck dissection, metastatic disease was
identied in 42% of cases. Another study reported
upstaging with H&E staining in up to 60% of
cases in which nodes were sectioned at 2 to
3 mm [67]. Therefore, although the extent to
which SNB upstages the neck by traditional pathologic methods seems similar to that with elective
neck dissection, the use of additional pathologic
methods results in perhaps an even greater level
of detection of disease.
Although SNB increases the rate of upstaging
in the clinically N0 neck, additional pathologic
methods, such as serial sectioning and immunohistochemical analysis, may increase the identication of micrometastasis (Fig. 6), dened as
single or multiple tiny deposits of tumor within
the lymph node sinuses with minimal replacement
of the node architecture and measuring no more
that 3 mm at any level of sectioning [38]. As stated
by Ross and Shoaib [68], the identication of micrometastases creates a dilemma for the current

SENTINEL LYMPH NODE BIOPSY IN STAGING

557

Fig. 6. Squamous cell carcinoma micrometastases in a cervical lymph node stained with H&E. (A) Multiple tiny deposits
of tumor are demonstrated within the lymph node sinus with minimal replacement of the lymph node architecture
(arrows) (original magnication  2). (B) Single deposit of tumor measures 1.5 mm at the level of section (original magnication  5). (Courtesy of A.K. El-Naggar, MD, Houston, TX.)

staging system for oral cancer. Patients with micrometastatic disease found by additional pathologic methods and with no additional disease in
the neck specimen are upstaged from clinically
N0 to a new classication of pNmi, or pathologic
nodal micrometastasis, for which the prognosis is
unknown. Although data for the current regional
disease classications indicate prognosis and
guide the use of adjuvant treatment with radiation, no such data are available for pNmi. The signicance of micrometastases in head and neck
squamous cell carcinoma must await the results
of well-designed prospective studies with uniform
pathologic evaluation.
Accuracy of sentinel lymph node biopsy
SNB must also be evaluated for its accuracy in
determining the status of the neck. Staging the
clinically N0 neck with SNB has dichotomous
outcomes: the presence or absence of occult
metastasis. Thus, before it can be considered
applicable in the staging of patients, SNB must
be critically appraised as a diagnostic test. First,
the evidence about the accuracy of SNB must
be proved valid [69]. Second, if this evidence is
valid, it must demonstrate an ability to distinguish
accurately who does and does not have occult
metastases [70].
The validity of a diagnostic test is based
foremost on independent blind comparison with
a reference standard [69]. In evaluating SNB, most
studies compared the histopathologic assessment
of sentinel nodes with that of an elective neck dissection specimen. Indeed, each of the 19 studies in
the meta-analysis of Paleri and colleagues [41] met
this criterion. Elective neck dissection serves as an

appropriate reference standard because of its accuracy in detecting occult metastasis and its high
rates of regional disease control. Moreover, elective neck dissection is the most ethical control
that can be incorporated into clinical trials with
appropriate equipoise. Even elective neck dissection can fail to detect occult metastases, however,
resulting in regional disease recurrence in a small
percentage of patients. Thus, clinical follow-up
for disease recurrence in the neck after SNB is
the reference standard with the greatest validity.
The accuracy of a test to determine the
presence or absence of disease is described by
the clinical performance parameters of the test.
Most studies of SNB in oral cancer have determined the accuracy of SNB by evaluating its
sensitivity, which is dened as the proportion of
patients with occult metastasis who are found to
have positive sentinel nodes. The sensitivity of
SNB ranges from 86% to 100% in published
reports. In the Canniesburn trial [66], SNB revealed metastatic disease in 22 of 53 patients
who underwent SNB and elective neck dissection.
In 1 patient, disease was not identied in the sentinel node but was found in the neck dissection
specimen. Thus, the sensitivity was 96%. In the
72 patients who underwent SNB alone, the sensitivity determined by clinical follow-up was 90%.
In the meta-analysis of Paleri and colleagues
[41], a pooled sensitivity of 92.6% (95% condence interval [CI]: 85.2%96.4%) was calculated
by a random eects model in 301 patients. These
studies suggest that SNB has a low false-negative
rate and is thus suciently sensitive to rule out occult metastasis when the result is negative.
The performance parameter complementary to
sensitivity is specicity, which is dened as the

558

SHELLENBERGER

proportion of patients without occult metastasis

who are found to have negative sentinel nodes.
Because a positive pathologic evaluation denes
the presence of occult metastasis in the neck
regardless of the pathologic ndings in the neck
dissection specimen, the concept of a false-positive
test result does not apply to SNB. Thus, the
specicity of SNB is dened as 100%.
A tests sensitivity is considered when deciding
whether to perform it. Once the results of the test
are available, however, the sensitivity of the test is
no longer relevant, because sensitivity values
apply only to persons known to have the disease.
In contrast, the dilemma for SNB is in determining the probability of disease in a patient with
negative sentinel nodes. That probability is described by the negative predictive value and is
determined by the sensitivity and specicity of the
test and the prevalence of occult metastasis in the
population undergoing the test. In the Canniesburn trial, the negative predictive value for SNB
was 97% in the 53 patients who underwent SNB
and elective neck dissection and 96% in the 72
patients who underwent SNB alone (95% CI:
83%99% and 95% CI: 87%99%, respectively)
[66]. Those results, although limited by wide CIs,
suggest a low probability of occult metastasis
when sentinel nodes are negative. The accuracy
rates (the proportion of all test results, positive
and negative, that were correct) of the two approaches were 98% and 97%, respectively.
An alternative way of describing the performance of a diagnostic test is the likelihood ratio,
which is dened as the probability of a particular
test result in patients with disease relative to the
likelihood of that test result in patients without
disease. In evaluating the performance of SNB,
the likelihood ratio for a negative test result
predicts the probability of metastatic disease in
a patient found to have negative sentinel nodes.
The likelihood ratio for a negative SNB in oral
cancer can be calculated from the data of 72
patients who underwent SNB alone in the Canniesburn trial ([1  sensitivity]/specicity). The likelihood ratio reveals that a negative SNB is 0.09
times more likely in a patient with occult metastasis than in a patient without occult metastasis.
More importantly, the likelihood ratio indicates
the degree to which the probability of occult metastasis before SNB is changed by the result of
the pathologic evaluation. If a patient with oral
cancer has a 20% probability of occult metastasis
before SNB and a likelihood ratio of 0.09 for
a negative test result, the probability of occult

metastasis after a negative pathologic evaluation

is reduced to less than 2% (by the likelihood ratio
nomogram). Therefore, a low likelihood of regional disease recurrence after a negative SNB
could justify clinical follow-up.

Remaining challenges
For SNB to become accepted for widespread
use in the staging of oral cancer, the procedure
must oer advantages over elective neck dissection, the current standard in management. Although elective neck dissection is the reference
standard for accuracy, the management of patients staged by SNB has not been compared with
management by elective neck dissection in prospective studies. Currently, several factors must be
investigated before SNB can reach clinical applicability (Box 4).
Is sentinel lymph node biopsy practical
in oral cancer?
The currently available data support the use of
SNB to stage only a limited number of patients
with oral cancer. The results of the Canniesburn
trial [66] gave important insights into the diculties of successfully identifying sentinel nodes
draining from primary tumors with certain characteristics. The trial, limited to patients with
early-stage oral cancer, further dened the indications for which SNB techniques are most applicable. Bulky or deeply inltrative primary tumors
that invade adjacent anatomic subsites clearly
pose technical diculties for peritumoral injection. Furthermore, the interpretation of lymphatic
mapping is complicated by altered lymphatic
drainage pathways and the involvement of subsites and tissues with diering drainage patterns.
Therefore, the current SNB techniques are limited
to the staging of the N0 neck of patients with
early-stage primary tumors.

Box 4. Remaining challenges to the

acceptance of sentinel lymph node
biopsy in oral cancer

Making the technique for SNB practical

Validating the sentinel lymph node
concept

Supporting the use of SNB with longterm follow-up data

SENTINEL LYMPH NODE BIOPSY IN STAGING

In addition to the diculties posed by tumors

advanced beyond an early stage, tumors at certain
subsites within the oral cavity present pitfalls for
SNB. The Canniesburn trial showed that although
sentinel lymph nodes were successfully identied
for tumors of the tongue, retromolar trigone,
buccal mucosa, lip, alveolar ridge, and hard
palate, oor of the mouth tumors posed a dilemma
[66,71]. The sentinel node identication rate for
oor of the mouth tumors (86%) was signicantly
lower than that for tumors at other subsites
(97%). In that trial, among the nine patients in
whom sentinel nodes could not be identied, six
had oor of the mouth primary tumors. The diculty of identifying sentinel nodes in patients with
these tumors may be explained by the proximity
of the primary tumor to the anterior oral cavity,
creating radioactivity shine-through to the draining basin of level I. High background counts
from the nearby primary tumor may thus overshadow the counts from the sentinel node, complicating identication with the gamma probe.
Additionally, surgical access may be limited for
deep nodes located close to primary tumors of
the anterior oral cavity. This problem has not
been overcome by techniques used thus far, including the use of lead shields to separate the primary tumor from the neck, removing the primary
tumor before sentinel node identication, and
elective dissection of level I nodes [72,73].
The nature of the regional lymphatics also
poses a challenge in SNB for oral cancers. The
excision of sentinel lymph nodes from oral cancers
by a minimally invasive approach may not be
possible. In single-institution pilot studies, sentinel lymph nodes were identied in the context of
elective neck dissection (SNB-assisted elective
neck dissection). Thus, the nodes were accessed
through standard neck dissection incisions after
ap elevation. This allowed for easy identication
of nodes by the gamma probe alone or with the
aid of blue dye. In addition, multiple and deep
nodes can be excised easily with the access
aorded by standard neck dissection. SNB performed through a limited access incision poses
technical diculties, however. Removing nodes
that are mapped to multiple levels requires extension of the incision or access by additional incisions. In addition, the excision of nodes that are
mapped to the deep jugular chain through a limited incision carries higher risks of neurovascular
injury. Only in the Canniesburn trial did a separate
cohort of patients undergo excision of sentinel
lymph nodes by a minimally invasive approach

559

(SNB alone) [66]. In that trial, the sentinel lymph

node identication rate was similar between the
group that underwent SNB-assisted elective neck
dissection and the group that underwent SNB
alone. In any case, further evidence of successful
identication and excision of sentinel lymph nodes
by minimally invasive access is required to support the use of SNB.
Even if proven feasible, additional limitations
remain for the use of SNB alone. The procedure,
even by a limited access approach, must be
performed along with the excision of the primary
tumor under general anesthesia. Because intraoperative frozen section analysis has not yet been
proven accurate, a second procedure to perform
therapeutic neck dissection is required in the event
of upstaging on nal pathologic evaluation. Furthermore, a subsequent therapeutic neck dissection would be impossible in the substantial
number of patients who undergo immediate freeap reconstruction of the primary site with
microvascular anastomosis in the neck.
Is the sentinel node concept valid in oral cancer?
Although the sentinel node concept has been
proven valid in cutaneous malignant melanoma of
the trunk and extremities, the biologic behavior of
oral squamous cell carcinoma combined with the
anatomic complexity of the head and neck poses
unique obstacles. As Pitman and coworkers [74]
noted, SNB may be an imperfect means of predicting the lymph node status of cancer in some
instances.
Although cervical metastasis from oral cancer
usually occurs in an orderly progression from the
rst-echelon nodes of level I to subsequent levels,
the disease may bypass level I or II and spread
directly to level III or IV as the only manifestation
of disease (Fig. 7). This phenomenon, known as
skip metastasis, is anatomically supported by
Rouvieres description of direct drainage pathways from the oral tongue to lower levels of the
neck [75]. Skip metastasis from carcinoma of the
oral tongue occurred in 15.8% of patients in
a study by Byers and colleagues [76]. Thus, tumor
cells escaping the rst draining nodes and establishing metastases at lower levels in the basin
may go undetected in some patients undergoing
SNB.
Soft tissue deposits of squamous cell carcinoma, representing total eacement of a lymph
node or extralymphatic spread of disease, may be
found in neck dissection specimens. Jose and

560

SHELLENBERGER

signicance of new staging classications that

are dened by the pathologic evaluation of sentinel nodes must be determined.
Will long-term follow-up data support
sentinel lymph node biopsy in the staging
of oral cancer?

Fig. 7. Composite drawing of the lymphatic drainage of

the tongue shows the supercial network (A) and collecting vessels (B). Dashed lines represent drainage from the
anterior third, dotted lines that from the middle third,
and continuous lines that from the posterior third.
Note direct pathways of drainage to bilateral nodes,
contralateral nodes, and nodes below the rst echelon.
(From Droulias C, Whitehurst JO. The lymphatics of
the tongue in relation to cancer. Am Surg 1976;
42(9):671; with permission.)

coworkers [77] reported soft tissue deposits of carcinoma in nearly one quarter of patients undergoing neck dissection for head and neck squamous
cell carcinoma and found an association with an
aggressive clinical course. More important, soft
tissue deposits of carcinoma were found in the absence of additional disease in 10% of patients, including those with clinically N0 disease. These
manifestations of regional disease with signicant
prognostic implications would go undetected by
SNB.
Although the routine evaluation of sentinel
node specimens may underestimate the presence
of metastasis, additional pathologic methods, as
stated previously, upstage an additional 5% to
8% of patients [37,38]. The analysis of markers by
immunohistochemical staining and molecular assays, such as reverse-transcriptase polymerase
chain reaction, as pioneered by Ferris and colleagues [78] are powerful means of detecting disease with new levels of precision. To validate
SNB for oral cancer, however, the prognostic

Ongoing studies of SNB in head and neck

squamous cell carcinoma are expected to dene
the role of the technique further in staging oral
cancer. The American College of Surgeons Oncology Group trial, which is comparing SNB with
elective neck dissection in the staging of patients
with oral cancer [40], was nearing the completion
of accrual with 160 patients at the end of 2005
(J.N. Myers, MD, PhD, personal communication,
2005). Patients with T1 or T2 primary tumors and
without clinical or radiologic evidence of regional
disease undergo preoperative lymphoscintigraphy,
followed by resection of the primary tumor with
SNB and elective neck dissection of levels I through
IV. Pathologic evaluation of the sentinel lymph
node by multiple sections and anticytokeratin
staining is to be compared with evaluation of the
largest lymph node from each level of the neck dissection specimen. A similar trial sponsored by the
European Organization for Research and Treatment of Cancer is underway [79]. Both prospective
trials are expected to provide results that should determine the validity of SNB in oral cancer.
These two trials are designed to study the
accuracy of SNB in the staging of oral cancer in
comparison with elective neck dissection. Nevertheless, the widespread acceptance of SNB ultimately depends on prospective trials designed
with end points determined during clinical follow-up. To date, only the Canniesburn trial has
reported data on patients followed for regional
recurrence after undergoing SNB. Although its
results suggest promise for the technique, only
a small number of patients (n 52) have been
followed for a short period (mean of 24 months)
[66]. Therefore, determining the merit of SNB in
oral cancer requires future trials designed to study
regional control and overall survival. Moreover,
those outcomes must be evaluated in patients
with oral cancer who are undergoing management
based on the results of SNB compared with those
managed by elective neck dissection.
Summary
The studies on SNB in oral cancer suggest that
its use is feasible in the identication of sentinel

SENTINEL LYMPH NODE BIOPSY IN STAGING

lymph nodes, that it is capable of detecting occult

metastasis, and that it accurately predicts the
status of the lymphatic basins draining the primary tumor. Although most currently available
data are from single-institution pilot studies,
ongoing multi-institution prospective trials are
expected to evaluate the results of SNB better in
the staging of oral cancer. Nonetheless, several
obstacles remain if the procedure is to gain
acceptance and widespread use. If SNB is to oer
signicant advantages over elective neck dissection, it must be proven to be practical for primary
tumors at a wider spectrum of subsites within the
oral cavity and must be performed as a minimally
invasive procedure. Moreover, the sentinel node
concept remains invalidated in oral cancer because of the biologic behavior of squamous cell
carcinoma and the anatomic complexity of the
head and neck. Finally, regional control and
overall survival rates of patients undergoing
SNB in the staging and management of oral
cancer must be determined to be equivalent or
superior to those achieved with elective neck
dissection if the potential of the procedure is to
be realized.

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

Acknowledgments
The author is indebted to Drs. Helmuth
Goepfert, Randal Weber, and Gary Clayman for
their inspiration and is grateful to Martha V.
Morrison for the scientic editing of the
manuscript.

[15]

[16]

[17]

References
[1] Johnson JT, Barnes EL, Myers EN, et al. The extracapsular spread of tumors in cervical node metastasis.
Arch Otolaryngol 1981;107:7259.
[2] Snow GB, Annyas AA, van Slooten EA, et al. Prognostic factors of neck node metastasis. Clin Otolaryngol 1982;7:18592.
[3] Schuller DE, McGuirt WF, McCabe BF, et al. The
prognostic signicance of metastatic cervical lymph
nodes. Laryngoscope 1980;90:55770.
[4] Byers RM, Wolf PF, Ballantyne AJ. Rationale for
elective modied neck dissection. Head Neck Surg
1988;10:1607.
[5] Amdur RJ, Parsons JT, Mendenhall WM, et al.
Postoperative irradiation for squamous cell carcinoma of the head and neck: an analysis of treatment
results and complications. Int J Radiat Oncol Biol
Phys 1989;16:2536.
[6] Cooper JS, Pajak TF, Forastiere AA, et al. Postoperative concurrent radiotherapy and chemotherapy

[18]

[19]

[20]

[21]

561

for high-risk squamous-cell carcinoma of the head

and neck. N Engl J Med 2004;350:193744.
Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer.
N Engl J Med 2004;350:194552.
Boyle JO, Strong EW. Oral cavity cancer. In:
Shah JP, editor. Cancer of the head and neck. Hamilton (Ontario, Canada): BC Decker; 2001. p. 10026.
Greenberg JS, El Naggar AK, Mo V, et al. Disparity
in pathologic and clinical lymph node staging in oral
tongue carcinoma. Implication for therapeutic decision making. Cancer 2003;98:50815.
Byers RM, El Naggar AK, Lee YY, et al. Can we detect or predict the presence of occult nodal metastases in patients with squamous carcinoma of the oral
tongue? Head Neck 1998;20:13844.
Lydiatt DD, Robbins KT, Byers RM, et al. Treatment of stage I and II oral tongue cancer. Head
Neck 1993;15:30812.
Yuen AP, Wei WI, Wong YM, et al. Elective neck
dissection versus observation in the treatment of
early oral tongue carcinoma. Head Neck 1997;19:
5838.
Seaman WB, Powers WE. Studies on the distribution of radioactive colloidal gold in regional lymph
nodes containing cancer. Cancer 1955;8:10446.
Gould EA, Winship T, Philbin PH, et al. Observations on a sentinel node in cancer of the parotid.
Cancer 1960;13:778.
Fee HJ, Robinson DS, Sample WF, et al. The determination of lymph shed by colloidal gold scanning in
patients with malignant melanoma: a preliminary
study. Surgery 1978;84:62632.
Holmes EC, Moseley HS, Morton DL, et al. A rational approach to the surgical management of melanoma. Ann Surg 1977;186:48190.
Robinson DS, Sample WF, Fee HJ, et al. Regional
lymphatic drainage in primary malignant melanoma
of the trunk determined by colloidal gold scanning.
Surg Forum 1977;28:1478.
Morton DL, Wen DR, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early
stage melanoma. Arch Surg 1992;127:3929.
Morton DL, Thompson JF, Essner R, et al. Validation of the accuracy of intraoperative lymphatic
mapping and sentinel lymphadenectomy for earlystage melanoma: a multicenter trial. Multicenter Selective Lymphadenectomy Trial Group. Ann Surg
1999;230:45363.
Jansen L, Nieweg OE, Kapteijn AE, et al. Reliability
of lymphoscintigraphy in indicating the number of
sentinel nodes in melanoma patients. Ann Surg
Oncol 2000;7:62430.
Cascinelli N, Morabito A, Santinami M, et al. Immediate or delayed dissection of regional nodes in
patients with melanoma of the trunk: a randomised
trial. WHO Melanoma Programme. Lancet 1998;
351:7936.

562

SHELLENBERGER

[22] Gershenwald JE, Thompson W, Manseld PF, et al.

Multi-institutional melanoma lymphatic mapping
experience: the prognostic value of sentinel lymph
node status in 612 stage I or II melanoma patients.
J Clin Oncol 1999;17:97683.
[23] Kim CJ, Reintgen DS, Balch CM, for the AJCC
Melanoma Staging Committee. The new melanoma
staging system. Cancer Control 2002;9:915.
[24] Brobeil A, Cruse CW, Messina JL, et al. Cost analysis of sentinel lymph node biopsy as an alternative
to elective lymph node dissection in patients with
malignant melanoma. Surg Oncol Clin N Am
1999;8:43545.
[25] Veronesi U, Paganelli G, Viale G, et al. A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer. N Engl J
Med 2003;349:54653.
[26] Wells KE, Rapaport DP, Cruse CW, et al. Sentinel
lymph node biopsy in melanoma of the head and
neck. Plast Reconstr Surg 1997;100:5914.
[27] Bostick P, Essner R, Sarantou T, et al. Intraoperative lymphatic mapping for early-stage melanoma
of the head and neck. Am J Surg 1997;174:5369.
[28] Wagner JD, Park HM, Coleman JJ III, et al. Cervical sentinel lymph node biopsy for melanomas of the
head and neck and upper thorax. Arch Otolaryngol
2000;126:31321.
[29] Eicher SA, Clayman GL, Myers JN, et al. A prospective study of intraoperative lymphatic mapping
for head and neck cutaneous melanoma. Arch Otolaryngol 2002;128:2416.
[30] Chao C, Wong SL, Edwards MJ, et al. Sentinel
lymph node biopsy for head and neck melanomas.
Ann Surg Oncol 2003;10:216.
[31] Lentsch EJ, McMasters KM. Sentinel lymph node
biopsy for melanoma of the head and neck. Exp
Rev Anticancer Ther 2003;3:67383.
[32] Alex JC, Krag DN. The gamma-probe-guided resection of radiolabeled primary lymph nodes. Surg
Oncol Clin N Am 1996;5:3341.
[33] Koch WM, Choti MA, Civelek AC, et al. Gamma
probe-directed biopsy of the sentinel node in oral
squamous cell carcinoma. Arch Otolaryngol 1998;
124:4559.
[34] Pitman KT, Johnson JT, Edington H, et al. Lymphatic mapping with isosulfan blue dye in squamous
cell carcinoma of the head and neck. Arch Otolaryngol 1998;124:7903.
[35] Shoaib T, Soutar DS, Prosser JE, et al. A suggested
method for sentinel node biopsy in squamous cell
carcinoma of the head and neck. Head Neck 1999;
21:72833.
[36] Ross G. Rationale for sentinel node biopsy to stage
N0 head and neck squamous-cell carcinoma. Cancer
Biother Radiopharm 2004;19:27384.
[37] Ambrosch P, Brinck U. Detection of nodal micrometastases in head and neck cancer by serial sectioning and immunostaining. Oncology 1996;10:
12216.

[38] Woolgar JA. Micrometastasis in oral/oropharyngeal squamous cell carcinoma: incidence, histopathological features and clinical implications. Br J Oral
Maxillofac Surg 1999;37:1816.
[39] Ross GL, Shoaib T, Soutar DS, et al. The First International Conference on Sentinel Node Biopsy in
Mucosal Head and Neck Cancer and adoption of
a multicenter trial protocol. Ann Surg Oncol 2002;
9:40610.
[40] Civantos F, Zitsch RP, Myers J. A trial of lymphatic
mapping and sentinel node lymphadenectomy for
patients with T1 or T2 clinically N0 oral cavity squamous cell carcinoma. Chicago: American College of
Surgeons, 2002.
[41] Paleri V, Rees G, Arullendran P, et al. Sentinel node
biopsy in squamous cell carcinoma of the oral cavity:
a diagnostic meta-analysis. Head Neck 2005;27:
73947.
[42] Reintgen D, Haddad FF, Pendas S, et al. Lymphatic
mapping and sentinel lymph node biopsy. Surgical
techniques. Sci Am 1998;117.
[43] Sri-Pathmanathan R, Railton R. Lymphoscintigraphy in the detection of cervical metastases from
oral carcinoma: a pilot study. Ann R Coll Surg
Engl 1989;71:2814.
[44] Kaplan WD, Davis MA, Rose CM. A comparison
of two technetium-99m-labeled radiopharmaceuticals for lymphoscintigraphy: concise communication. J Nucl Med 1979;20:9337.
[45] Civantos FJ, Gomez C, Duque C, et al. Sentinel
node biopsy in oral cavity cancer: correlation with
PET scan and immunohistochemistry. Head Neck
2003;25:19.
[46] Thevarajah S, Huston TL, Simmons RM. A comparison of the adverse reactions associated with isosulfan blue versus methylene blue dye in sentinel
lymph node biopsy for breast cancer. Am J Surg
2005;189:2369.
[47] Stoeckli SJ, Pfaltz M, Ross GL, et al. The Second International Conference on Sentinel Node Biopsy in
Mucosal Head and Neck Cancer. Ann Surg Oncol
2005;12:91924.
[48] Tschopp L, Nuyens M, Stauer E, et al. The value of
frozen section analysis of the sentinel lymph node in
clinically N0 squamous cell carcinoma of the oral
cavity and oropharynx. Otolaryngol Head Neck
Surg 2005;132:99102.
[49] Alex JC, Sasaki CT, Krag DN, et al. Sentinel lymph
node radiolocalization in head and neck squamous
cell carcinoma. Laryngoscope 2000;110:198203.
[50] Lopez MC, Amoros Sebastia LI, Ferrer Ramirez
MJ, et al. Preliminary results of the relevance the
sentinel node in head and neck tumors. Acta Otorrinolaringol Esp 2003;54:18590.
[51] Pitman KT, Johnson JT, Brown ML, et al. Sentinel
lymph node biopsy in head and neck squamous cell
carcinoma. Laryngoscope 2002;112:210113.
[52] Altinyollar H, Berberoglu U, Celen O. Lymphatic
mapping and sentinel lymph node biopsy in

SENTINEL LYMPH NODE BIOPSY IN STAGING

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]
[61]

[62]

[63]

[64]

[65]

squamous cell carcinoma of the lower lip. Eur J Surg

Oncol 2002;28:724.
Barzan L, Sulfaro S, Alberti F, et al. Gamma probe
accuracy in detecting the sentinel lymph node in clinically N0 squamous cell carcinoma of the head and
neck. Ann Otol Rhinol Laryngol 2002;111:7948.
Hyde NC, Prvulovich E, Newman L, et al. A new approach to pre-treatment assessment of the N0 neck
in oral squamous cell carcinoma: the role of sentinel
node biopsy and positron emission tomography.
Oral Oncol 2003;39:35060.
Kohno N, Ohno Y, Kihara K, et al. Feasibility of
sentinel lymph node radiolocalization in neck
node-negative oral squamous cell carcinoma patients. ORL J Otorhinolaryngol Relat Spec 2003;
65:6670.
Kontio R, Leivo I, Leppanen E, et al. Sentinel lymph
node biopsy in oral cavity squamous cell carcinoma
without clinically evident metastasis. Head Neck
2004;26:1621.
Zitsch RP III, Todd DW, Renner GJ, et al. Intraoperative radiolymphoscintigraphy for detection of
occult nodal metastasis in patients with head and
neck squamous cell carcinoma. Otolaryngol Head
Neck Surg 2000;122:6626.
Ionna F, Chiesa F, Longo F, et al. Prognostic value
of sentinel node in oral cancer. Tumori 2002;88:
S189.
Taylor RJ, Wahl RL, Sharma PK, et al. Sentinel
node localization in oral cavity and oropharynx
squamous cell cancer. Arch Otolaryngol 2001;127:
9704.
Pastore A. Treatment of the N0 neck in carcinoma of
the larynx. Suppl Tumori 2002;1:S5760.
Wang G, Fan W, Zhang W, et al. Sentinel lymph
node detection by combined dye-isotope technique
and its predictive value for cervical lymph node metastasis in patients with lingual carcinoma. Chin
Med J (Engl) 2003;116:12135.
Werner JA, Dunne AA, Ramaswamy A, et al. Sentinel node detection in N0 cancer of the pharynx and
larynx. Br J Cancer 2002;87:7115.
Stoeckli SJ, Steinert H, Pfaltz M, et al. Sentinel
lymph node evaluation in squamous cell carcinoma
of the head and neck. Otolaryngol Head Neck
Surg 2001;125:2216.
Asthana S, Deo SV, Shukla NK, et al. Intraoperative neck staging using sentinel node biopsy and imprint cytology in oral cancer. Head Neck 2003;25:
36872.
Shoaib T, Soutar DS, MacDonald DG, et al. The accuracy of head and neck carcinoma sentinel lymph
node biopsy in the clinically N0 neck. Cancer 2001;
91:207783.

563

[66] Ross GL, Soutar DS, Gordon MD, et al. Sentinel

node biopsy in head and neck cancer: preliminary results of a multicenter trial. Ann Surg Oncol 2004;11:
6906.
[67] Rigual NR, Douglas W, Lamonica D, et al. Sentinel lymph node biopsy: a rational approach for
staging T2N0 oral cancer. Laryngoscope 2005;
115:221720.
[68] Ross GL, Shoaib T. Role of sentinel node biopsy in
the management and staging of the N0 neck. Odontology 2005;93:16.
[69] Jaeschke R, Guyatt G, Sackett DL. Users guides to
the medical literature. III. How to use an article
about a diagnostic test. A. Are the results of the
study valid? Evidence-Based Medicine Working
Group. JAMA 1994;271:38991.
[70] Jaeschke R, Guyatt GH, Sackett DL. Users guides
to the medical literature. III. How to use an article
about a diagnostic test. B. What are the results and
will they help me in caring for my patients? Evidence-Based Medicine Working Group. JAMA
1994;271:7037.
[71] Loree TR. Sentinel lymph node biopsy for early
stage clinical N0 squamous cell carcinoma of the
oral cavity. Ann Surg Oncol 2004;11:7256.
[72] Ross G, Shoaib T, Soutar DS, et al. The use of sentinel node biopsy to upstage the clinically N0 neck in
head and neck cancer. Arch Otolaryngol 2002;128:
128791.
[73] Ross GL, Soutar DS, Shoaib T, et al. The ability of
lymphoscintigraphy to direct sentinel node biopsy in
the clinically N0 neck for patients with head and
neck squamous cell carcinoma. Br J Radiol 2002;
75:9508.
[74] Pitman KT, Ferlito A, Devaney KO, et al. Sentinel
lymph node biopsy in head and neck cancer. Oral
Oncol 2003;39:3439.
[75] Rouviere H. Anatomy of the human lymphatic system. Ann Arbor (MI): Edward Brothers; 1938.
[76] Byers RM, Weber RS, Andrews T, et al. Frequency
and therapeutic implications of skip metastases in
the neck from squamous carcinoma of the oral
tongue. Head Neck 1997;19:149.
[77] Jose J, Moor JW, Coatesworth AP, et al. Soft tissue
deposits in neck dissections of patients with head
and neck squamous cell carcinoma: prospective
analysis of prevalence, survival, and its implications.
Arch Otolaryngol 2004;130:15760.
[78] Ferris RL, Xi L, Raja S, et al. Molecular staging of
cervical lymph nodes in squamous cell carcinoma of
the head and neck. Cancer Res 2005;65:214756.
[79] Shah JP. Extent of surgical intervention in case of
N0 neck in head and neck cancer patients. Eur
Arch Otorhinolaryngol 2004;261:2934.

Oral Maxillofacial Surg Clin N Am 18 (2006) 565571

Access Surgery for Oral Cancer

Rui Fernandes, DMD, MDa,*, Robert Ord, DDS, MD, FACSb
a

Division of Oral and Maxillofacial Surgery, Department of Surgery, University of Florida College of Medicine,
Jacksonville, 653-1 West Eight Street, Jacksonville, FL 32209, USA
b
Department of Oral and Maxillofacial Surgery, University of Maryland School of Dentistry,
Greenbaum Cancer Center, University of Maryland Medical Center, 419 West Redwood Street, Suite 410,
Baltimore, MD 21201, USA

The oral cavity is the surgical home for the oral

and maxillofacial surgeon. Familiarity with this
area gives the surgeon the ability to carry out
demanding resections without the need to make
facial skin incisions. This ability continues to be
one of the major strengths of our specialty, in fact,
separating us from most of our counterparts.
Although most of our resections can be carried
via the transoral route, at times, the position of
the tumor and the disease process require a more
exacting resection. The management of malignant
processes requires that the resection be performed
with negative margins. To this end, the surgeon is
sometimes required to increase the access to the
oral cavity to resect with clear margins.
This article highlights some of the most common surgical accesses to the oral cavity. There are
a multitude of surgical accesses for the facial
skeleton based on the concept of modular osteotomies. These are primarily used for tumors in the
nasopharynx or the skull base. The goal of this
article is to review the surgical access used to aid
in the resection of tumors of the oral cavity. To
this end, the article is divided into two parts:
midface access for maxillary pathologic ndings
and transmandibular access for tumors in the oral
cavity and the base of tongue. To provide the
ideal approach to the tumor, in selected cases, the
ablative surgeon needs to incorporate the midface
approach as well as one of the transmandibular
accesses in a combined fashion. The method used

* Corresponding author.
E-mail address: rui.fernandes@jax.u.edu
(R. Fernandes).

is predicated on a thorough understanding of the

head and neck anatomy and careful presurgical
planning. This oers the surgeon an understanding of the location of the tumor and how it
encroaches on its surroundings. An appreciation
of the tumor location as well as the anatomic
constraints of the region should reveal to the
experienced surgeon the area of greatest diculty
for resection. The use of the correct surgical access
is predicated on this understanding.
Maxillary approach
Weber-Fergusson maxillectomy incision
The term transfacial is used to describe any procedure that mobilizes the midfacial skeleton
through a facial (skin) incision, irrespective of
the extent of midface disassembly used [1]. One
of the most commonly used transfacial approaches
to the midface for the resection of maxillary tumors is the Weber-Fergusson (WF) maxillectomy
incision. This approach was rst described by Weber [2] in the German literature and was later modied by Fergusson [3] in the English literature.
The benet of this incision is that it allows for
improved access to the maxilla and for unimpeded
resection of tumors aecting the anterior and
superior aspects of the maxilla. It may also be
used to aid in swinging the maxilla laterally while
pedicled to the cheek ap. The lateral maxillary
swing is used routinely in accessing the nasopharynx for the resection of nasopharyngeal carcinomas [4]. An important caveat to using the WF
incision is that when the tumor involves the posterior aspect of the maxilla and/or the pterygoid
plates, the approach does not improve the access.

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.008

oralmaxsurgery.theclinics.com

566

FERNANDES & ORD

Since the rst description of the WF approach,

it has undergone multiple minor modications;
yet, the essence of the incision is the same. One of
the most common modications of this incision is
Altemirs incision [5], whereby the straight line incision in the lip is moved from the midline and
placed on the philtrum.
The authors approach is to mark the patient
by carefully placing the incision in an area in
which the healed incision leaves the least noticeable scar. This is accomplished by placing the
incision along the philtrum, carrying the lip
extension with a slight step away and with an
option of placing a chevron in the vermillion
portion of the incision. The superior extent of the
lip incision should be performed into the nasal sill
and then extend out along the base of the ala and
in a cephalad direction. The lateral nasal incision
should be raised and placed in the nasal side wall
at the junction of the nasal subunit (Fig. 1). Once
the medial canthal region is approached, the incision may be extended laterally inferior to the
lower eyelid in one of the creases (as in the Dieffenbach modication) or extended superiorly
into a Lynch incision. A full-thickness ap is
then elevated while maintaining sound oncologic
margins. The bony osteotomies are made as the
tumor dictates. At the conclusion of the resection
and reconstruction, the ap is reapproximated,
taking care to have correct anatomic closure so
as to minimize the postoperative scar.

approach requires familiarity with closed rhinoplasty techniques by the ablative surgeon [6]. The
technique incorporates the use of vestibular and
intranasal incisions to lift or deglove the facial
skin from the facial skeleton, improving the access
to the maxillary tumor.

Midfacial degloving approach

Transmandibular approach

An alternative approach to maxillary tumors is

by use of the midface degloving technique. This

Lip-split mandibulotomy approach

Fig. 1. WF incision marking with a chevron placed on

the vermilion of the lip and extension of the incision
into the nasal sill and onto the lateral nasal wall subunit.

Posterior maxillary approach

When the surgeon is confronted with a large
tumor that extends anteriorly, superiorly, and
posteriorly with extension toward the pterygoid
plates or the infratemporal fossa, it may be
advantageous to use two approaches. The WF
approach may be combined with a lip-splitting
mandibulotomy approach so that the posterior
aspect of the tumor can be resected in a more
direct fashion. The combination of the two techniques was rst reported by Dingman and Conley
[7] using a lateral mandibular osteotomy, and it
was later modied by Lawson and colleagues [8]
by moving the osteotomy to a midline mandibulotomy. The approach to the posterior maxilla
and the retromaxillary and/or infratemporal space
would not be complete without mentioning Obwegesers work on the subject [9]. The detailed surgical technique may be found in his article on the
temporal approach to the temporomandibular
joint (TMJ), the orbit, and the retromaxillaryinfracranial region. The major impetus behind
this approach is the lack of facial skin incisions.

The mandibulotomy approach is typically used

in combination with a lip split. The placement of
the incision for the lip split varies signicantly
among surgeons. The rst description of a lip-split
technique was by Roux [10] in the middle of the
nineteenth century. Since then, there have been
a multitude of variations to the lip-splitting incision. The three classic incisions most often used
are (1) the Roux incision [2], whereby a straightline incision is made from the midline of the lower
lip, extending through the chin and connecting to
the neck incision (Fig. 2); (2) the McGregor incision [11], which extends in a vertical fashion
from the midline of the lip toward the chin, circumventing the chin along the chin and/or labiomental groove; and (3) the Robson incision [12],
which is placed on a relaxed skin tension line medial to the commissure and descends down toward
the neck to meet with the neck incision. Hayter

ACCESS SURGERY

Fig. 2. Roux incision for the lip-splitting mandibulotomy used in this patient, who needed a bilateral neck
dissection at the time of resection of the primary tumor.

and colleagues [13] recently published a modication of the McGregor incision that incorporates
a chevron into the vermillion margin and the midline lip incisions to improve closure, thereby improving the nal esthetics. The authors routinely
use the McGregor incision in association with
a mandibular osteotomy for oral access. The incision is marked in the midline of the lower lip, with
or without a chevron, and extended to the labiomental crease. Once the crease is reached, the incision is extended to the ipsilateral side of the
neck along the mental crease toward the midline
of the submental region; from there, the incision
is connected to the neck incision (Fig. 3). The

Fig. 3. Skin markings for a lip-splitting mandibulotomy

in conjunction with a neck dissection.

567

placement of the circum-mental incision on the ipsilateral side rather than the contralateral side of
the neck dissection is to decrease the possibility
of ischemic necrosis of the chin [14]. Before making the incision, the senior author prefers to place
a large hemostatic suture on each side of the lip.
A 1-cm section of angiocatheter is threaded over
a 0-silk suture to protect the lip vermillion. Before
tying the stitch, the skin incision is completed with
a number 11 blade to prevent distortion from the
hemostatic stitch (Fig. 4). The mucosal cuts are
placed, leaving a cu for closure at the end of
the procedure. Once the ap is elevated, the osteotomy is then marked anterior to the mental foramen. Two 2-0 plates are then contoured and
placed at the marked osteotomy site, with one superior and one at the inferior border. The inferior
plate used is a locking 2-0 plate with bicortical
screws. This is our preferred method based on
the work performed at our institution by Engro
and coworkers [15]. The plates are then removed
and placed on the back table to be used at the
time of mandibular xation. The osteotomy is
completed rst, taking care to elevate the lingual
mucosa and protect it. The osteotomy is completed, and the mucosa incision is performed on
the oor of the mouth along the lingual mucosa.
A cu is left for later closure if possible, or a gingivosulcular incision is placed to reect a lingual
mucoperiosteal ap. The mucosal incisions
through the buccal and lingual gingival mucosa
should be stepped away from the bone cuts so
that the soft tissue closure does not lie directly
over the osteotomy site.

Fig. 4. Hemostatic sutures in the lip also used for ap

retraction. Note the angiocatheter protecting the vermillion (inset).

568

FERNANDES & ORD

With the osteotomy and mucosal incision

completed, the mandible is then able to be
reected laterally, providing a clear view of the
tumor and allowing for uncompromised resection
(Fig. 5). This lateral swinging of the mandible is
why this technique is at times referred to as the
mandibular swing approach [16]. Once the resection and reconstruction are completed, the xation is reapplied and the incisions are closed
(Fig. 6). When the incisions are closed with attention to the careful alignment of tissues, the postoperative scar is imperceptible (Fig. 7).

swings the segment in a superior direction. As described by Attia and coworkers [17], this procedure involves a lip-splitting incision. We have
used a similar double osteotomy with a modication of the proximal osteotomy to a vertical subsigmoid osteotomy. This modication allows for
the use of the double osteotomy without the
need for a lip-split incision [18]. Whenever this
technique is used for the oral cavity, one needs
to pay meticulous attention to the overlying attached mucosa so as to maintain perfusion to
the mandible.

Nonlip-splitting mandibulotomy approach

Pull-through technique

The use of a nonlip-splitting mandibulotomy

is most commonly employed to improve access to
the parapharyngeal space and the infratemporal
fossa. In some circumstances, however, it may be
used to improve access to the oral cavity. This
approach incorporates two osteotomies: one at
the parasymphyseal area and another at the
ramus region. The approach of Attia and coworkers [17] uses a parasymphyseal and horizontal ramus osteotomy above the lingula and

The concept of in-continuity resection has been

around for many years. The rationale for this
approach lies in the possibility of a tumor focus in
the lymphatics between the primary and the
cervical nodes. To that end, proponents of this
theory advocate that the resection should remove
the primary, the intervening lymphatics, and the
cervical nodes in one block specimen. The technique was rst described by Scheunemann [19] in
1975 and was later modied by Stanley [20]. The

Fig. 5. (A) Lip-splitting mandibular swing approach demonstrates increased access to the oral cavity for tumor
extirpation. (B) Lip-splitting mandibular swing approach demonstrates increased access to the oral cavity and resected
posterior oor of the mouth and tongue tumor. (C) Lip-splitting mandibular swing approach demonstrates increased
access to the posterior oral cavity and resected posterior oor of the mouth and tongue tumor.

ACCESS SURGERY

569

Visor ap

Fig. 6. Closure of ap with careful reapproximation of

the lip-vermillion and chin ap.

pull-through operation is performed by making

a releasing incision on the lingual aspect of the
gingival and the elevation of a mucoperiosteal
ap. The dissection is performed until the mylohyoid muscle is encountered, taking care to leave
a small cu of muscle attached to the mandible
to aid in closure. Similarly, the genioglossus and
geniohyoid muscles are detached, and the contents of the oral cavity are then pulled into the
neck, allowing for a more direct resection through
unimpeded visual and tactile feel of the tumor
(Fig. 8). Devine and colleagues [21] compared
their experience between the lip-split mandibulotomy and the mandibular lingual releasing access
and found that patients with lingual release reported more problems with speech, swallowing,
and chewing.

An alternative to the pull-through technique is

the visor ap. This ap is mostly used in the
resection of tumors involving the anterior oral
cavity; however, at times, it may also be used in
the resection of lateral tongue and oor of the
mouth tumors. The incision for the ap is
performed from the mastoid process on one side
and is extended toward the contralateral mastoid
or to the mandibular symphysis on the unaected
side. The ap is elevated as usual, taking care to
identify and preserve the marginal mandibular
branch of the facial nerve. Once the inferior
border of the mandible is encountered, a periosteal
incision is made along the exposed mandible. An
intraoral incision is made along the gingivobuccal
sulcus, and a mucoperiosteal ap is elevated and
reected toward the inferior border of the mandible. The two dissections are then connected. The
ap is tethered to the mandible only at the mental
foramen region (Fig. 9). The mental nerve is incised, and the ap is then mobilized in a cephalad
direction with the aid of two rubber drains. With
the drains secured, the surgeon has an unimpeded
view of the oral cavity to proceed with the tumor
extirpation.
One of the obvious drawbacks of this technique is the sacrice of the mental nerve, leading
to anesthesia of the lip and chin. The benet is
that the surgeon is able to remove the tumor
without the need for a lip-splitting incision,
however.

Fig. 7. (A) Preoperative view of patient compared with postoperative view. (B) McGregor lip-splitting incision
mandibulotomy. Note the excellent healing of the incision.

570

FERNANDES & ORD

Fig. 8. (AC) Pull-through approach for resection of intraoral tumors. Note that the mylohyoid, genioglossus, and
geniohyoid muscles have been disinserted to all the contents of the oral cavity to drop or be pulled into the neck.

Summary
The most important task in managing oral
cancer is the resection of tumors without positive
margins. To accomplish this goal, the surgeon is
often faced with placing facial skin incisions to
improve access to the oral cavity. This article has
reviewed some of the most commonly used
approaches and highlighted techniques used by
the authors to resect tumors with minimal postoperative scaring. Although esthetics are

Fig. 9. Development of a visor ap. Note the elevation

of the periosteum from the mandible with only the
mental nerve tethering the ap to the mandible before
incising the oral mucosa.

important to the patient, they remain a distant

third consideration to the resection of the tumor
and functional reconstruction. The surgeon and
the patient should not lose sight of this fact.
References
[1] Evans BT. Access surgery. In: Landon J, Patel M,
editors. Operative maxillofacial surgery. Chapman
Hall Medical; 1998. p. 238.
[2] Weber O. Vorstellung einer kranken mit Resection
des Unterkiefers Verhandlungen des naturhist-med
Vereins z Heidelberg. 1845;4:802 [in German].
[3] Fergusson W. In operation of the upper jaw. A system of practical surgery. Edinburgh (UK): John
Churchill; 1842. p. 484.
[4] Wei WI, Lam KH, Sham JS. New approach to the
nasopharynx: the maxillary swing approach. Head
Neck 1991;13(3):2007.
[5] Altemir FH. Transfacial access to the retromaxillary
area. J Maxillofac Surg 1986;14:16570.
[6] Browne JD. The midfacial degloving procedure for
nasal, sinus, and nasopharyngeal tumors. Otolaryngol Clin North Am 2001;34(6):1095104.
[7] Dingman DL, Conley J. Lateral approach to the
pterygomaxillary region. Ann Otol Rhinol Laryngol
1970;79:9679.
[8] Lawson W, Naidu RK, Benger JL, et al. Combined
median mandibulotomy and Weber-Fergusson maxillectomy. Arch Otolaryngol Head Neck Surg 1990;
116:5969.
[9] Obwegeser HJ. Temporal approach to the TMJ, the
orbit, and the retromaxillary-infracranial region.
Head Neck Surg 1985;7(3):18599.
[10] Roux PJ. In: Butlin HT, Spencer GJ, editors. Diseases of the tongue. London: Cassell; 1900. p. 359.

ACCESS SURGERY

[11] McGregor IA, McDonald DG. Mandibular osteotomy in the approach to the oral cavity. Head Neck
Surg 1983;5:45762.
[12] Robson MC. An easy access incision for the removal
of some intraoral malignant tumours. Plast Reconstr
Surg 1979;64:8345.
[13] Hayter JP, Vaughan ED, Brown JS. Aesthetic lip
splits. Br J Oral Maxillofac Surg 1996;34:4325.
[14] Becker GD. The extended single transverse neck
incision for composite resections. Surg Gynecol
Obstet 1979;148:902.
[15] Engro SL, Blanchaert RH Jr, von Fraunhofer JA.
Mandibulotomy xation: a laboratory analysis. J
Oral Maxillofac Surg 2003;61(11):1297301.
[16] Spiro RH, Gerold FP, Strong EW. Mandibular
swing approach for oral and oropharyngeal
tumors. Head Neck Surg 1981;3:3718.

571

[17] Attia EL, Bentley KC, Head T, et al. A new external

approach to the pterygomaxillary fossa and the parapharyngeal space. Head Neck Surg 1984;6:8849.
[18] Kolokythas A, Fernandes R, Ord R. A non-lip splitting double mandibular osteotomy for access to parapharyngeal tumors. J Oral Maxillofac Surg, in press.
[19] Scheunemann H. Pull-through surgery in mouth
oor-tongue neoplasms. Acta Stomatol Belg 1975;
72(2):22930.
[20] Stanley RB. Mandibular lingual releasing approach
to oral and oropharyngeal carcinomas. Laryngoscope 1984;94(5 Pt 1):596600.
[21] Devine JC, Rogers SN, McNally D, et al. A comparison of aesthetic, functional and patient subjective
outcomes following lip-splitting mandibulotomy
and mandibular lingual releasing access procedures.
Int J Oral Maxillofac Surg 2001;30(3):199204.

Oral Maxillofacial Surg Clin N Am 18 (2006) 573604

Postablative Reconstruction Techniques

for Oral Cancer
D. David Kim, DMD, MD*, G.E. Ghali, DDS, MD, FACS
Department of Oral and Maxillofacial Surgery, Louisiana State University Health Sciences Center,
1501 Kings Highway, Shreveport, LA 71103, USA

The estimated 20,000 new cases of oral cavity

carcinoma that occur annually in the United
States provide many patients who may require
reconstructive surgery for their surgical defects [1].
The functional and anatomic complexity of the
head and neck makes reconstruction of these defects an intricate task. Although not all patients
with oral cancer have surgery as their primary
mode of therapy, surgery is generally considered
the preferred method of treatment for all but the
most advanced cases of disease in medically t patients. With greater emphasis on early detection,
more patients may be identied in earlier stages
of disease in which complex reconstruction may
not be necessary. Currently, however, more than
50% of oral cavity cancers are diagnosed in the
advanced stages of disease (stage III or IV), in
which surgery can leave large, functionally crippling defects [2].
The resultant defects from surgical ablation of
early stage oral cancer are often amenable to
relatively simple reconstruction techniques, including primary closure, healing by secondary
intention, or skin grafting. These techniques are
described in more detail elsewhere in the literature
and are not covered in this article [3,4]. The use of
regional and distant tissue for reconstruction
of larger, more complex oral defects is the focus
of this article. The appropriate use of these techniques should be determined based on the variables that may be present for each individual
patient. These variables include comorbid conditions, patient desires and expectations, quality of

* Corresponding author.
E-mail address: dkim1@lsuhsc.edu (D.D. Kim).

prospective donor sites, and the size and character

of the ablative defect.
Classication of aps
Flaps can be described based on the proximity
of the donor tissue to the ablative defect. The use
of the terms local, regional, and distant aps result in clinically useful and descriptive terminology. With the exception of most local aps,
these aps are all based on a vascular pedicle
to provide blood ow to the transferred tissue.
The classication of aps provided by Mathes
and Nahai [5] describe the anatomic basis of these
vascular pedicles. These authors proposed classication system is described in Table 1. When used
with Taylor and Palmers [6] anatomic study of
the human bodys angiosomes, these classications allow reconstructive surgeons to systematically classify aps according to the type of
vascular supply and predict the angiosome distribution of the pedicle being used.
Regional aps
The use of regional tissue for reconstruction of
oral cavity defects has been considered the mainstay of head and neck reconstruction for many
years. These aps have proved over time to be
relatively simple and reliable for use in the head
and neck. All such regional or rotational, axial
pattern aps are limited by their arc of rotation
and their proximity to the defect in question.
Depending on ones denition of regional,
numerous aps have been described that fall
into this category. Most of these aps are based
on a musculocutaneous conguration in which

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.004

oralmaxsurgery.theclinics.com

574

KIM & GHALI

Table 1
The classication of muscle aps
Type

Pedicle

Anatomic location

Type 1 One vascular

pedicle
Type 2

Type 3

Type 4

Type 5

Gastrocnemius, rectus
femoris, tensor fascia
lata
One dominant
Abductor digiti
vascular pedicle
minimi, abductor
plus minor
hallucis, biceps femoris,
pedicles (most
exor digitorum brevis,
common
gracilis, peroneus
longus, peroneus brevis,
pattern)
platysma,
semitendinosis, soleus,
sternocleidomastoid,
temporalis, trapezius,
vastus lateralis
Two dominant
Gluteus maximus,
pedicles
rectus abdominus,
serratus anterior,
semimembranosus
Segmental vascular Extensor digitorum
pedicles
longus, extensor
hallucis longus, exor
digitorum longus, exor
hallucis longus,
sartorius, tibialis
anterior
Pectoralis major,
One dominant
latissimus dorsi
vascular pedicle
and secondary
segmental
vascular pedicles

Data from Mathes SJ, Nahai F. Classication of the

vascular anatomy of muscles: experimental and clinical
correlation. Plast Reconstr Surg 1981;67:17787.

a skin paddle may be transported based on

perforators of a main vascular pedicle that travel
through the underlying muscle. Notable exceptions to this are the deltopectoral and forehead
aps, which are transported without a muscle.
Pectoralis major myocutaneous ap
First described for use in head and neck defects
by Ariyan in 1979 [7], the pectoralis major myocutaneous ap (PMMF) has enjoyed a great deal of
popularity in the head and neck and has been
called the workhorse ap for head and neck reconstruction. Part of this popularity is because of
the aps relative ease of harvest, hearty vascular
supply, low morbidity, and good versatility. Although microvascular soft tissue aps recently
have overshadowed the PMMF, it remains a mainstay in head and neck reconstruction and a rescue

ap for use in microvascular reconstruction

failures.
The pectoralis major is a broad, fan-shaped
muscle that covers the anterior chest wall. Its
origin is the sternum, clavicle, and external
oblique aponeurosis, and it forms a thick tendon
that inserts into the greater tubercle of the
humerus. The action of this muscle is to adduct
and medially rotate the arm with some action in
exing (superior bers) and extending (inferior
bers) the arm. The loss of the pectoralis major
muscle after ap harvest seems to be tolerated well
by patients, with most of the actions of the muscle
compensated by the latissimus dorsi muscle.
The pectoralis major is a class V muscle
according to Mathis and Nahai [5] with one dominant pedicle, the thoracoacromial artery, and
a secondary segmental blood supply from the
internal mammary parasternal perforators. A
branch of the second part of the axillary artery,
the thoracoacromial arterys pectoral branch is
the main arterial supply to the PMMF. The lateral
thoracic and the superior thoracic arteries also
provide some vascularity to the PMMF but are
usually sacriced to obtain a greater arc of rotation. The lateral thoracic artery has been shown
to provide substantial vascular supply to the pectoralis major muscle and have a caliber as large as,
if not larger than, the pectoral branch of the thoracoacromial artery [8].
These descriptions of the vascular territories of
the pectoralis major muscle apply, only in part, to
the overlying chest wall skin. Although the pectoralis major muscle derives its main blood supply
from the thoracoacromial artery, the segmental
branches of the internal thoracic artery also provide substantial vascular supply to the muscle,
especially in the portion of the muscle medially and
inferior to the fourth intercostal space. The overlying skin in this region receives its blood supply
from the corresponding parasternal vascular pedicles through direct musculocutaneous perforators.
Harvesting a PMMF based on the pectoral branch
of the thoracoacromial artery with the skin paddle
oriented inferior to the fourth intercostal space
requires use of the network of choke vessels that
link adjacent angiosomes to provide vascular
supply to the skin paddle. It is this anatomic reason
that an inferior-medially placed skin paddle of the
PMMF may undergo necrosis even when the blood
supply to the muscle remains intact [9].
The orientation of the skin paddle for the
PMMF can be designed in many variations
according to the defects characteristics. The

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

most commonly used orientation is along the

medial surface of the muscle in an inferior
location to maximize the aps arc of rotation
(Fig. 1). This location of the skin paddle requires
patent anastomotic connections between the thoracoacromial and the internal mammary systems.
Extending this skin paddle o the pectoralis major
muscle and onto the abdominal skin, by including
the underlying rectus fascia, has been described.
This orientation is tenuous, because this random skin is an angiosome once removed
from the main arterial pedicle of the PMMF
[10]. A somewhat more reliable skin paddle orientation would be designed around the free lateral
border of the muscle in the area of the inframammary fold (Fig. 2). According to ink injection studies by Reid and Taylor [11], the skin in
this region derives its blood supply through
direct fasciocutaneous perforators from the
thoracoacromial axis.

Flap harvest technique

The patient should be prepared and draped
widely to include the upper abdomen past the
midline of the chest, the shoulder, and the axilla,
and the head and neck should be prepared as
necessary for ablative surgery. Some surgeons
may prefer to have the ipsilateral arm extended
on an armboard to facilitate severing the pectoralis majors humeral attachments. Both surgical
sites are contiguous, so little eort should be
placed in keeping the operative site for the PMMF
under clean conditions because the entire head,
neck, and chest eld is clean-contaminated with
oral secretions. Whenever possible, however,

Fig. 1. PMMF design with an inferomedially placed

skin paddle and curvilinear anterior chest incision.

575

Fig. 2. PMMF design with a laterally placed skin paddle

and inframammary crease incision design.

scalpels, electrocautery tips, and scissors should

be changed if they were used during tumor
resection.
Depending on the skin paddle design, the
approach to the PMMF can vary in many ways.
Most commonly, with an inferior-medially placed
skin paddle, the simplest incision design involves
a curvilinear incision from the ipsilateral shoulder
to the superior margin of the skin paddle (see
Fig. 1). This incision allows direct exposure of the
pectoralis major with little tunneling necessary,
with the obvious disadvantage of a cosmetically
objectionable scar. Alternatively, the incision
may be placed in the inframammary fold to conceal the scar (Fig. 2). This incision design aords
adequate access but requires signicant cutaneous
undermining to visualize the entire muscle.
The skin is incised from the shoulder to the
superior margin of the skin paddle through subcutaneous fat and fascia down to the muscle. The
fascia overlying the muscle should be incised and
undermined laterally with the overlying skin until
the lateral and inferior margins of the pectoralis
major muscle are identied. Only the lateral
border of the skin paddle should be incised at
this point down to muscle. This approach allows
the surgeon to evaluate the location of the skin
paddle on the muscle and whether adjustments to
the skin paddle design should be made to minimize its random portion. Once this location is
veried, the circumference of the skin paddle is
incised to the same level (Fig. 3). The superior
portion of the skin and fascial ap should be

576

KIM & GHALI

Fig. 4. A branch of the pectoral nerve (arrow) can be

seen lying next to the main thoracoacromial vascular
pedicle (dashed arrow).
Fig. 3. PMMF with lateral skin ap elevated and skin
paddle oriented over the inferior edge of the muscle.

undermined to connect with the neck dissection

incisions over the clavicle. Caution should be exercised not to disturb the second and third intercostal perforators to the deltopectoral ap during this
superior dissection to preserve their integrity in
case this ap is needed at a later date.
Once the skin/fascia aps have been elevated,
the pectoralis major muscle can be separated from
the chest wall. This separation is started laterally
at the lateral border of the muscle using blunt and
sharp dissection through a well-dened, thin
fascia between the pectoralis major and the
anterior chest wall and superiorly between the
pectoralis major and minor muscles. This dissection can be performed mostly with gentle nger
dissection except for sharp incision of the medial
pectoralis major origins. Care should be taken to
identify and ligate the intercostal perforators,
especially when undermining the portion of the
muscle that underlies the skin paddle. Monopolar
cautery should be avoided during this portion of
the dissection.
As the muscle is being raised from the chest
wall, the sternal origin of the muscle should be
incised sharply to ligate any perforators. The
second and third intercostal perforators should
be left intact during this dissection by cutting the
pectoralis major attachments lateral to their
emergence through the muscle. As this dissection
continues up the chest wall, the pectoral branch of
the thoracoacromial artery comes into view on the
undersurface of the pectoralis major along the
medial aspect of the pectoralis minor muscle. In

the same vicinity, one or two branches of the

pectoral nerve are encountered and should be
severed to provide greater arc of rotation (Fig. 4).
This maneuver eectively de-innervates the
muscle.
With the thoracoacromial pedicle in clear view,
the humeral attachments can be transected, with
careful attention paid to hemostasis in this area to
prevent postoperative hematoma formation. The
cephalic vein may be encountered during this
process in the deltopectoral groove and should
be preserved if possible. Further skeletonization
of the vascular pedicle can be achieved with
careful blunt dissection through the areolar tissue
surrounding the vessels.
With the dissection complete, the ap can be
passed through the tunnel created to the neck
incisions to reach the head and neck site being
reconstructed. Careful hemostasis should be
achieved, and primary closure usually can be
obtained with wide undermining over suction
drains (Fig. 5).

Fig. 5. Primary closure over suction drains at the donor

site usually can be obtained with wide undermining.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

577

Latissimus dorsi myocutaneous ap

Available as a pedicled or free ap, the
latissimus dorsi myocutaneous ap can be used
to resurface nearly any area of the head and neck.
Recorded as the rst musculocutaneous ap to be
reported in the medical literature in 1896 by
Tansini [12], it was not used in the head neck until
1978 [13]. The latissimus dorsi is a at, broad
muscle that arises from T6T12, the thoracolumbar fascia, lower four ribs, and the iliac crest. It
converges around the teres major to insert into
the intertubular groove of the humerus [14]. Because of its large surface area, the latissimus dorsi
can be used to cover huge defects in the head and
neck.
The blood supply to the latissimus dorsi myocutaneous ap is analogous to that of the PMMF.
Like the pectoralis major, the latissimus dorsi is
a class 5 muscle according to Mathes and Nahai [5],
with its one dominant pedicle being the thoracodorsal artery and segmental secondary pedicles
consisting of the paraspinal perforators. The
angiosomes over the latissimus dorsi can be separated into thirds. The rst lies over the superior
third of the muscle supplied by perforators of the
thoracodorsal artery. The second corresponds to
the middle one third of the muscle, which is supplied by the paraspinal perforators. Finally, over
the caudal portion of the muscle, the skin is fed by
the lumbar arteries (Fig. 6). In this third portion,
placement of a skin paddle that relies on the thoracodorsal vessels for blood supply may be prone to
necrosis because of its once removed angiosome
relationship to the aps pedicle. There is a particular abundance of perforators located at the aps
anterior, free margin (middle zone) where a skin
paddle is most ideally designed.
The thoracodorsal artery is a branch of the
subscapular system, which is itself a branch of the
third part of the axillary artery. During its course
through the axilla, the thoracodorsal artery gives
o branches to the teres major, serratus anterior,
subscapularis, and an angular branch to the
inferior tip of the scapula.
Flap harvest technique
The patient should be placed in a prone or,
preferably, a lateral decubitus position with the
ipsilateral arm extended anteriorly and superiorly.
Great care should be taken to protect the patients
pressure points and airway during positioning.
A line is drawn from the midpoint of the axilla to
a point midway between the anterior and

Fig. 6. The angiosomes over the latissimus dorsi muscle.

I, Proximal zone with direct perforators from the thoracodorsal. II, Intermediate zone with contributions from
thoracodorsal and paraspinal perforators. III, Distal
zone from the lumbar arteries.

posterior superior iliac spines. This line represents

the anterior free margin of the latissimus dorsi
muscle (Fig. 7). If a skin paddle is being used, it
can be designed parallel to this line with care
not to extend it too far caudally, whereby entering
the tertiary angiosome.
The initial incision is made along this line and
the anterior border of the skin paddle through
skin and subcutaneous fat and fascia until the
muscles free margin is identied. From here, the
dissection can proceed in two ways. First,
the muscle can be dissected bluntly from the chest
wall under its free margin to expose branches of
the thoracodorsal vessels. Specically, the branch
to the serratus anterior can be identied and
followed proximally to the main thoracodorsal
artery. These branches can be ligated and the
vascular pedicle dissected proximally. Once sucient dissection of the pedicle is obtained, the
circumference of the skin paddle can be incised
and the medial skin aps elevated in a subfascial
plane to the midline, just superior to the iliac crest

578

KIM & GHALI

Fig. 7. Patient marking for latissimus dorsi ap. The

anterior line represents the free anterior margin of the
muscle. A skin paddle can be designed along this line
to take advantage of abundant perforators in the area.

and over the scapula. The origins of the muscle are

sharply incised and the muscle elevated from the
chest wall with careful ligation of perforators
encountered during this dissection. Alternatively,
this skin ap development can be performed before
identication of the vascular pedicle, and the
thoracodorsal vessels come into view on the undersurface of the muscle as it is lifted o the chest wall.
Once the dissection is complete, if a pedicled
ap is planned, a tunnel through the axilla must
be made between the pectoralis major and minor
muscles to communicate with the neck dissection
incisions. This tunnel can be achieved with blunt
dissection and should be generous enough so
compression of the vascular pedicle is avoided.
Care should be used so as not to injure the
thoracoacromial vessels during this maneuver to
preserve the blood supply to the PMMF in the
event this ap is necessary at another time. The
ap can be passed carefully through under direct
vision of the pedicle to the head and neck site.
Wide undermining of the skin of the back
usually allows primary closure over suction drains
to prevent hematoma and seroma formation. Skin
grafts may be used if necessary.

fasciocutaneous ap can be used for various resurfacing indications in the head and neck. With the
introduction of myocutaneous and free aps, the
deltopectoral ap generally has been relegated to
cutaneous reconstruction of neck defects.
The aps notorious distal skin reliability has
spawned numerous articles to extend the aps
transferable skin surface area. The area of skin
available for the deltopectoral ap lies just below
the clavicle to approximately the fourth intercostal space and from the parasternal region to
a variable extent laterally. This lateral skin is
prone to necrosis when transferred with this ap.
Once again, the angiosomes of this area of the
superior thoracic wall can explain this well-known
problem.
The primary angiosome of the internal mammary perforators extend from the parasternal
region to the deltopectoral groove. The region of
the deltopectoral groove is its own angiosome
served by a direct perforator from the thoracoacromial axis. Lateral to the deltopectoral groove
and overlying the deltoid muscle is the territory of
the deltoid branch of the thoracoacromial system
and the circumex humeral artery (Fig. 8). This

Deltopectoral ap
Known as the Bakamjian ap after the surgeon
who popularized the ap in 1965 [15], the deltopectoral ap was the primary method for resurfacing head and neck defects before the introduction
of myocutaneous aps, such as the pectoralis major. Based on the rst through third perforators of
the internal mammary artery, the ap is usually
transferred using only the second and third. This

Fig. 8. The angiosomes of the deltopectoral region. I,

Direct perforators of the rst through third intercostal
arteries. II, Direct perforator of the thoracoacromial
artery. III, Perforators from the deltoid branch of the
thoracoacromial artery and the circumex humeral
artery.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

tertiary angiosome is unreliable based on the internal mammary pedicle of the deltopectoral ap
[16]. To improve the viability of this lateral skin,
the blood ow through the choke vessels between angiosomes must be reversed to rely wholly
on the ow from the internal mammary vessels.
This reverse is accomplished in a delay procedure in which the ap is elevated, and the distal
vessels (direct cutaneous branch and the deltoid
branch of the thoracoacromial artery) are ligated
and sutured back to the donor site for a variable
length of time to allow the new vascular pattern
to establish itself [17].
Flap harvest technique
The superior limb of the ap parallels the clavicle
from the midline, and the inferior limb parallels this
line at approximately the fourth intercostal space.
The length of the ap is determined by the location
of the head and neck defect in question, but if the
ap must be carried onto the deltoid muscle, a delay procedure should be performed rst. The skin
is incised through subcutaneous fat and fascia down
to the pectoralis major muscle. The ap is then
elevated from a lateral to medial direction until
approximately 2 cm lateral to the midline. The
source vessels do not have to be visualized during
the dissection. The direct cutaneous perforator from
the thoracoacromial vessels is encountered in the
deltopectoral groove. It must be ligated and transected as the dissection proceeds medially. This vessel
is also one that must be ligated during the delay
procedure for an extended deltopectoral ap.
The ap can be rotated into the recipient site
with the exteriorized pedicle tubed or the intervening skin can be depithelialized and tunneled
under the neck skin. In the case of an exteriorized
pedicle, a secondary procedure (or a tertiary
procedure in the case of a previous delay
procedure) is necessary to sever the pedicle after
approximately 3 to 5 weeks. Alternatively, the
intervening neck skin between the defect and the
donor pedicle can be removed to accommodate
the skin of the entire ap.
An attempt at primary closure can be performed by wide undermining, but it results in
great tension in this area. A skin graft should be
used if primary closure cannot be achieved.
Microvascular reconstructive techniques
Although some debate exists, Alexis Carrel [18]
is credited for the standardization of end-to-end
vascular anastomoses in 1902. He was awarded

579

the Nobel Prize in Medicine and Physiology in

1912 for this work. The use of these techniques
in human small vessels was limited because of
the lack of adequate instrumentation and magnication, however. Although the operating microscope was introduced to clinical surgery in 1921
by Nylen [19], it was not until 1960 that it was
used in microvascular anastomoses by Jacobson
and Suarez [20] in animal vessels as small as
1.4 mm in diameter. The threshold of the 1-mm
vessel anastomosis was breached by Buncke and
Schulz [21] in 1964, when they successfully replanted a rabbit ear with vessels of 1 mm in diameter. For this accomplishment, Buncke is
considered by many as the father of microvascular surgery. Conict exists regarding the rst
microvascular tissue transfer to the head and
neck. As early as 1957, a free jejunal ap for an
esophageal defect was reported by Seidenberg
and colleagues [22]. Others credit McLean and
Buncke [23] with the transplant of omentum to
ll a large scalp defect in 1969. In 1974, Panje
and colleagues [24] were recognized as the rst
to use free tissue transfer for oral cavity
reconstruction.
Although the term microvascular surgery
was coined by Jacobson in his work with the
operating microscope in vascular anastomosis, the
advancement of microvascular surgery truly
parallels the improvement of microscopes and
microinstrumentation, and the manufacturing of
ultra-ne suture materials and needles. Many of
the pioneers of microvascular surgery developed
their own suture materials and instrumentation
that were modied from jewelers instruments to
facilitate microvascular anastomoses.

Advantages and disadvantages of free tissue

transfer
The concept of microvascular free tissue transfer (MFTT) has not been embraced universally as
the standard in head and neck reconstruction after
tumor extirpation. Many researchers espouse the
merits of the use of regional aps or delayed
nonvascularized bone grafts for use in these
reconstructions [25].
Although there are numerous advantages to
free tissue transfer, to some practitioners they do
not oset the disadvantages that may be present.
It is true, however, that not all patients are
suitable or agree to have free ap reconstruction
of their oral cavity defects.

580

KIM & GHALI

One of the foremost advantages of MFTT is

the ability to tailor the donor ap to the specic
needs of the ablative site. For example, a thin and
pliable radial forearm ap would be ideal for
a partial defect of the oor of mouth or tongue,
whereas the muscular bulk and large skin paddle
of a rectus abdominus ap may be better suited
for total glossectomy or maxillectomy defects.
Pedicled aps are less ideal for defects that need
either extreme bulk of tissue or thin, pliable tissue.
MFTT is not restricted by the arc of rotation of
the pedicle. Flaps can be placed in areas that may
not be reached easily by rotational aps as long as
recipient vessels are nearby. In many cases of
MFTT, if a longer vascular pedicle is needed,
more of the donor vessels can be dissected to
provide this length, or vein grafts may be used if
necessary.
Another advantage of MFTT is that it is
a more ecient use of harvested tissue in which
nearly all is used directly for the reconstruction. In
contrast, most rotational aps require mobilization of large amounts of tissue with a relatively
small area used for the actual reconstruction.
Because microvascular ap reconstructions bring
their own blood supply, they are well suited to
placement in such inhospitable areas as the oral
cavity. Their excellent perfusion can improve
wound healing and protect against wound breakdown. MFTT of bone-containing aps practically
eliminates the resorption of bone seen in nonvascularized bone grafts. The use of primary
microvascular reconstruction has allowed for
more aggressive ablative surgery with clear margins, even in large tumors [26].
Because of the added length of operating time
and need for technical expertise, MFTT is perceived as being a less reliable and more costly
procedure than rotational ap reconstruction.
Several studies that compared patients who have
received MFTT reconstruction of head and neck
compared with rotational ap have suggested that
this may not be true. In an often quoted article,
Brown and colleagues [27] showed that the risk of
postoperative complications in patients who underwent free ap reconstruction was not signicantly greater than those matched patients who
received pedicled ap reconstruction. Their study
showed a trend toward shorter intensive care
unit stays and overall hospital stays, although
not statistically signicant. Several retrospective
reviews identied the PMMF as having a higher
risk of major postoperative complications, such
as stula formation, ap loss, infection, and

hematoma compared with radial forearm [28] or

rectus abdominus aps [29].
Other possible advantages of MFTT include
(1) the possibility of neural anastomoses for the
theoretical fabrication of a sensate ap [30], (2)
possible superior functional outcomes, and (3)
a possible improved quality of life. There are disadvantages to MFTT, however, not the least of
which are the technical demands on the surgical
team, operating room sta, anesthesiology team,
and the intensive care unit. The initial anesthetic
time for MFTT is longer than for nonmicrovascular reconstructions, but studies have shown that
the overall cost, length of hospital and intensive
care unit stay, and incidence of complications
are not signicantly increased with MFTT
[27,31]. The potential for ap loss is also a distinct
disadvantage of MFTT. Even the most accomplished microsurgeons have a 5% to 10% failure
rate that is usually attributable to thrombosis.
Comparisons to the pectoralis myocutaneous
ap have shown greater reliability in terms of
lower incidence of ap loss with radial forearm
and rectus abdominus aps [28,29], and some series show major complications with pectoralis myocutaneous aps to be as high as 58% [32]. This
gure can be compared with reports of large series
of MFTT failure and re-exploration rates of 2%
to 8% and 6% to 14%, respectively [33,34]. Although this low failure rate was not always true
of MFTT, the current state of microvascular reconstructive surgery has evolved from a radical,
last resort procedure to the rst and most reliable
choice in all types of reconstruction, including
head and neck defects.
Donor site morbidity is of great concern to
reconstructive surgeons. A hallmark of the
PMMF is its acceptable morbidity prole. Dierent free ap donor sites have dierent morbidity
proles and are discussed later in this text. Many
of the most commonly used aps, such as the
radial forearm, bula, rectus abdominus, and,
recently, the anterolateral thigh aps, have acceptable levels of postoperative morbidity with few
long-term decits.
Patient management
Preoperative
Preoperative patient evaluation for MFTT for
oral cavity defects is of paramount importance for
multiple reasons. The initial evaluation allows
reconstructive surgeons to (1) identify any contraindications to MFTT, (2) estimate the expected

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

defect and the need for reconstruction, (3) identify

the ideal donor site or sites, (4) perform appropriate preoperative testing, (5) verify the patients
medical and emotional tness to undergo the
procedure, and (6) formulate alternative reconstructive options.
As with any presurgical assessment, evaluation
of the microvascular reconstruction candidate
should begin with a thorough history, including
any comorbid disease states, medications, and
allergies. A patients social history also should
be elicited at this time. Although absolute contraindications for microvascular surgery are rare,
several common disease states may serve as
relative contraindications and direct the treatment
plan toward more conventional reconstructive
techniques. Many of these conditions are controversial as to their impact on microvascular reconstructive surgery.
Hypercoagulable states are the only true contraindications to MFTT, including diseases such
as polycythemia, thrombocytosis, and possibly
sickle cell anemia (Table 2) [35]. The risk of
thrombosis in these conditions is too high to justify MFTT. Patients taking the antiestrogen medication Tamoxifen for prevention or treatment of
breast cancer should be taken o this medication
before surgery because of its known thrombogenic
activity [36]. For similar reasons of hypercoaguability and decreased ap perfusion and impaired

Table 2
Hypercoagulable states
Inherited

Acquired

Antithrombin III
deciency
Protein C deciency
Protein S deciency
Activated protein C
resistance
Factor V Leiden
Dysbrinogenemia
Plasminogen activator
deciency
Plasminogen deciency
Heparin cofactor II
deciency
Factor XII deciency
Polycythemia
Thrombocytosis
Sickle cell anemia

Prolonged immobilization
Pregnancy
Surgery/trauma
Oral contraceptives/
antiestrogens
Homocystinuria
Vitamin K deciency
Malignancy
Smoking
Nephrotic syndrome
L-asparaginase
Diabetes mellitus
Hyperlipidemia
Disseminated intravascular
coagulation
Lupus anticoagulant
Anticardiolipin antibody

581

wound healing, smokers should be encouraged to

stop at least 1 week before surgery.
Age as a risk factor for postoperative complications with MFTT has not been shown to be
a signicant problem. Although studies have
implied that advanced age poses a greater risk to
prolonged hospital stay, medical complication,
and in-hospital death [37,38], others have shown
no signicant dierence in complications after
major head and neck reconstructive surgery [39].
Advanced age is associated with atherosclerotic
disease and increased vessel fragility, however,
which may pose a specic, although not absolute,
problem in MFTT. Conversely, pediatric patients
may have small vascular caliber, which makes
anastomosis dicult and less reliable, especially
when vessels are less than 1 mm in diameter.
Diabetes can predispose patients to microvascular disease and atherosclerosis and negatively
inuence wound healing, which may result in
delayed wound healing, infection, and possible
ap loss. Atherosclerosis is most problematic
when evaluating the patient for harvest of the
bula ap, which is discussed later. The direct
inuence of atherosclerosis on microvascular
anastomoses is not known, however.
Once a thorough medical history is obtained,
evaluation of the lesion should be performed to
note location, size, and involvement of adjacent
tissues. The extent of the resection is estimated and
the ideal donor site is chosen according to this
prediction. A detailed explanation of the proposed
procedure should be discussed with the patient and
the family to delineate risks, benets, capabilities,
and limitations of the reconstruction. This dialog
helps the reconstructive surgeon determine if the
patient and family are emotionally suitable
candidates for microvascular reconstruction.
Any preoperative testing should be ordered at
this time, and specic tests for the proposed donor
sites are performed. Noteworthy are the performance of Allens tests for radial forearm ap
candidates and evaluation of distal pulses and
extremity angiography for bula ap harvest.
Postoperative
The intraoperative details of MFTT are
discussed in the section covering microvascular
techniques. Of at least equal importance to free
ap success is the postoperative management of
patients who receive MFTT. Protocols vary by
institution, but certain standards are followed by
most and should be applied to all free ap

582

KIM & GHALI

reconstructions regardless of site. Most postoperative management procedures are not based on
scientic evidence of ecacy, however, but rather
surgeon preference, personal experience, and,
possibly, superstition.
It stands to reason that pressure on the
microvascular pedicle should be avoided at all
times. In head and neck reconstruction, the vessels
in the neck can be compressed easily by circumferential ties or straps, such as those for tracheotomies or oxygen tents. Although nursing orders
may be written to avoid the use of such devices,
a sign over a patients bed also should be used to
prevent other personnel from making an innocent
mistake. Similarly, kinking or undue tension on
the vessels must be avoided, especially during the
immediate postoperative period. The position of
the neck that optimizes vessel geometry that was
determined intraoperatively should be maintained
as adamantly as possible. This position can be
sustained by the use of postoperative sedation
with ventilatory support. A paralytic agent also
can be added to this pharmacologic management
if desired. Surgeons who prefer not to sedate
patients in the postoperative period rely on
patient compliance for maintenance of neck
position. The use of a rigid cervical collar is an
extreme but sometimes useful tool in some cases.
Hemodynamics should be maintained as close
to normal limits as possible. Avoiding extreme
hyper- or hypotension is crucial to the limitation
of hematoma formation while maintaining perfusion of the ap. A balance in the oxygen-carrying
capacity of blood and blood viscosity must be
achieved. Although little scientic evidence exists,
a hematocrit of 28% to 30% is generally accepted
as the desired goal for the immediate postoperative period.
Flap-monitoring protocols vary drastically between institutions. Although many modalities of
monitoring aps are currently available, the
modern standard is clinical examination of the
aps skin paddle. The parameters evaluated
include color, capillary rell, ap turgor, and
warmth (Fig. 9). The frequency of serial evaluations of aps postoperatively also varies by institution from every hour to every 4 hours. The
frequency is based on the fact that the time between the onset of a thrombotic event and its recognition may be critical to the aps salvage [40].
In pig skin aps, this critical time is 7 hours according to one study [41]. The no-reow phenomenon may occur if the aps circulation
cannot be re-established within 8 to 12 hours [42].

Fig. 9. Early postoperative appearance of a healthy

bula aps skin paddle. Notice the skins natural color
with no ecchymosis or edema.

Possibly the most commonly used adjunct to

the clinical examination is the pin-prick test. A 25gauge needle is used at the center of the skin
paddle, and the rapidity, color, and amount of
blood return are evaluated. A healthy ap bleeds
bright red blood within 1 to 3 seconds. Rapid
return of dark colored blood combined with an
ecchymotic skin paddle suggests venous insuciency (Fig. 10), whereas no blood return with
a ap that is cool to the touch suggests arterial
thrombosis. This examination, although widely
used, is subject to much interobserver error but
may prove a valuable aid to the clinical
examination.
The next modality most commonly used in
postoperative ap monitoring is the hand-held
surface Doppler probe. Although the ability to
conrm venous outow has been proposed, the
practical use of the Doppler probe is mainly for

Fig. 10. Ecchymotic and edematous bula ap skin

paddle. This ap was found to have venous insuciency
caused by a neck hematoma compressing the venous
anastomosis.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

conrmation of arterial ow [43]. Because thrombotic complications are usually venous in nature,
however, the use of this modality is severely limited. Observance of a change in the character of
the phases of the Doppler signal may help a clinician identify the impending failure of a ap. One
must be aware that Doppler signals obtained at
the anastomotic site may be unreliable because
of the proximity of the carotid arteries.
Finally, a few of the adjunctive modalities that
have been studied most frequently with microvascular tissue transfer should be mentioned,
including tissue PO2 monitoring [44], tissue pH
[45], pulse oximetry [46], photoplethysmography
[47], and laser Doppler owmeters [48,49]. Each
of these instruments has its own advantages and
disadvantages, but none has made its way into
mainstream microsurgical monitoring. Discussion
of these modalities is beyond the intent of this article. Although rarely indicated, technetium scanning within 5 to 7 days of surgery also remains
the ideal method of evaluating the perfusion of
vascularized bone-containing aps [50].
Antiplatelet and antithrombotic pharmacotherapy
There is little agreement on the ideal pharmacologic agent for the prevention of thrombosis in
microvascular surgery. Like many areas of microvascular surgery, many of the institutional variations are caused by individual practitioner
experience or retrospective case control studies.
Heparin. Although the thrombus in microvascular anastomosis depends highly on platelets, the
glue that holds the large clumps of platelets
together is brin. The use of heparin to inhibit the
formation of brin has logical merit in microvascular surgery. The heparin family of molecules
acts by binding to and inducing a conformational
change in antithrombin III, which alters antithrombin III from a slow inhibitor of coagulation
to a more rapid one. Its inhibition of thrombin
(factor IIa) and other factors in the coagulation
cascade (XIIa, XIa, IXa, and Xa) results in the
anticoagulant eect seen with heparin [51].
The genesis of arterial thrombosis usually
occurs in areas of high or disrupted ow or sites
of atherosclerotic plaque rupture. These thrombi
consist mainly of platelet aggregates bound together by thin brin strands, and they depend on
the action of platelets and the coagulation cascade. Venous thrombi may rely more on the
coagulation cascade, however, because they are
formed in areas of blood stasis and consist mostly

583

of red blood cells and brin rather than platelets

[52]. In a study by Khouri and colleagues [53], inhibition of the platelet aggregation pathway in
rats did not improve vessel patency because brin
alone could form a thrombus without platelet aggregation. The authors were able to show that
heparin at therapeutic levels more eectively prevented arterial and venous occlusion than antiplatelet agents.
Regardless of these data, the systemic postoperative use of heparin in microvascular surgery
is generally limited to a few institutions or for the
salvage of a threatened ap, probably because of
fear of hemorrhagic complications, such as hematoma formation and donor site bleeding. Somewhat more commonly used is an intraoperative
bolus of heparin (1000 U) given just before removal of the microvascular clamps, which has
been found to be benecial in a rabbit model [54].
The topical use of heparin as a local irrigant in varying concentrations (between 50 and 250 U/mL)
is common. Although low doses generally have no
systemic eects, concentrations more than 250
U/mL have been shown to alter the partial thromboplastin time. Although commonly used and
advocated in some reviews [55], little research supports its role as an eective topical antithrombotic
agent [56].
Aspirin. Aspirin (acetylsalicylic acid) has been
available for more than 100 years and has proved
to be an eective preventive medication in myocardial infarction, stroke, and other occlusive
vascular events. Aspirin acetylates cyclo-oxygenase and inhibits the production of arachidonic
acid metabolites, including thromboxane and
prostacyclin. Thromboxane A2 is a potent vasoconstrictor and platelet-activating agent produced
by platelet cyclo-oxygenase. Conversely, prostacyclin (prostaglandin I2) is derived from endothelial cyclo-oxygenase and is a potent vasodilator
and inhibitor of platelet aggregation [52]. Low
doses of aspirin (15 mg/kg) are theorized to preferentially inhibit thromboxane and not prostacyclin production [57]. Although its usefulness in
microvascular surgery is somewhat suspect, it remains a popular pharmacologic agent for prevention of thrombus formation. One study in an
animal model showed low-dose aspirin to inhibit
venous thrombosis and improve microcirculatory
perfusion [58]. It is generally given orally or
crushed via nasogastric tube immediately postoperatively. Although side eects of low-dose therapy are rare, bleeding, gastritis, and renal failure

584

KIM & GHALI

are possible sequelae and should be considered

carefully, especially in elderly patients.
Dextran. A recently maligned agent that nonetheless remains popular among microvascular surgeons is dextran [59]. Usually used in its low
molecular weight form (40,000 MW, dextran
40), dextran is a polysaccharide synthesized from
sucrose by Leuconostoc mesenteroides. Although
its mechanism of action is not completely understood, ve proposed actions of dextran are as follows: (1) increased electronegativity of platelets
and endothelium, which results in decreased platelet aggregation, (2) brin structure modication,
which increases its degradation, (3) inhibition of
alpha-2 antiplasmin with subsequent plasminogen
activation, (4) decreased factor VIII and von Willebrand factor, and (5) altered rheologic properties of blood by acting as a volume expander [52].
Dextran has potential antigenicity, and a test
dose of 20 mL is recommended by some authors
before continuous infusion [60]. The infusion is
generally given at a rate of 25 to 50 cc per hour
for 5 days postoperatively. It is usually initiated
just before release of microvascular clamps, immediately preoperatively or immediately postoperatively. Dextran is discontinued after 5 days
without tapering, and its eect can be expected
to persist for several hours or days after discontinuation [56].
Some conict exists regarding the benet of
dextran to microvascular anastomoses. Retrospective reviews and anecdotal evidence suggest
a favorable microvascular patency rate with
dextran, however. This nding, combined with
a generally low side-eect prole, allows dextran
to remain popular as an adjunct to microvascular
surgery. An association with dextran and noncardiogenic pulmonary edema has been made but
rarely reported [61], thus vigilance of observations
of signs and symptoms of this complication
should be practiced. A loose association with the

use of dextran and systemic postoperative complications (myocardial infarction, congestive heart
failure, pulmonary edema, pleural eusion, and
pneumonia) also was suggested in a recent article
[59]. Whether these associations have a direct
causal relationship remains to be elucidated.
Streptokinase, urokinase, and tissue plasminogen
activator. These three brinolytic agents may be
eective in the salvage of a thrombosed ap when
infused locally; however, evidence suggests a high
rate of rethrombosis when the brinolytic agent is
stopped. Heparin can be added to the salvage
protocol of a thrombosed ap to prevent reocclusion. Other pharmacotherapeutic agents that may
have a place in microvascular surgery in the future
include hirudin, which is secreted by leeches in
a recombinant form [62], tissue factor pathway inhibitor [63], clopidogrel, and glycoprotein IIb-IIIa
inhibitors [64], and some preliminary reports have
surfaced for each of these agents. These agents require further clinical and experimental experience
before their use can be advocated.
Microvascular techniques
Reconstructive surgery in the head and neck is
dissimilar to that in other regions of the body. A
complex three-dimensional anatomic recipient site
provides a true challenge to the reconstructive
imagination when designing aps for this area. As
dierent as the reconstructive surgery may be,
however, the microvascular portion of these
operations should be much the same. Besides
some minor variables of patient positioning, recipient vessel selection, and vessel diameter, when
under the microscope every anastomosis should
be nearly identical. The surgical instrumentation
involves specialized instruments that at minimum
include paired jewelers forceps, nonlocking
microneedle drivers, adventitia scissors (Fig. 11),
and microvascular clamps. Nylon suture designed

Fig. 11. Microvascular instruments from left to right: Adventitia scissors, long and short straight microscissors, two
pairs of jewelers forceps, curved jewlers forcep, and nonlocking microneedle driver.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

for microvascular surgery is necessary. Usually 90 or 10-0 sutures are used with a taper-cutting needle, depending on surgeons preference. Spatulashaped, tapered, or reverse cutting needles are
also available. An operating microscope designed
for assisted microsurgery is also necessary, or
high-powered, wide-eld loupe magnication
may be used in some larger caliber anastomoses.
The following section deals with specic microvascular techniques from vessel selection to ap
salvage.
Recipient vessel selection
Because of the abundance of vessels in the head
and neck, the availability of recipient vessels for
microvascular anastomosis is usually not an issue.
Choosing which vessels to use depends on many
factors, including vessel diameter, pedicle length
or geometry, and presence of atherosclerotic
plaques and evidence of traumatic dissection. If
a neck dissection is being performed before
MFTT, the type of dissection determines the
presence of remaining vascular structures. The
communication between reconstructive and
ablative surgical teams is crucial to maximize the
preservation of neck vasculature while maintaining an oncologically sound neck dissection.
Although some reconstructive surgeons insist on
evaluating the completed neck dissection for
available recipient vessels, this process can be
greatly minimized or eliminated with clear
communication and a comfortable working
relationship between the two teams. It is a rare
instance in which at least some option for
microvascular anastomosis is not available in
even the most aggressive ablative procedure.
Probably the most common recipient neck
artery for MFTT is the facial artery. This branch
of the external carotid artery is commonly
transected during neck dissection as it traverses
the inferior border of the mandible and as it enters
the submandibular triangle. A length of facial
artery often can be preserved if the submandibular
triangle is not grossly involved with tumor and the
submandibular gland can be separated from
the facial artery without excessive trauma to the
artery. Alternatively, the stump of the artery can
be traced back to its origin from the external
carotid and brought out from behind the posterior
belly of the digastric and stylohyoid muscles to
create length.
Another branch of the external carotid that is
commonly used is the superior thyroid artery. The

585

caliber of this vessel is often adequate for only

a short distance from its origin at the external
carotid. Alternatively, end-to-side anastomosis
(ESA) or even end-to-end anastomosis (EEA)
directly to the external carotid can be performed
if no other adequate recipient artery is available.
The transverse cervical artery o the thyrocervical trunk can be preserved in most neck dissections, and substantial lengths are obtainable by
tracing the arterys path across the neck to the
undersurface of the trapezius muscle. These arteries may be less prone to atherosclerosis than the
branches of the external carotid but do have
a greater susceptibility to vasospasm. Up to
20% of transverse cervical arteries may arise
directly from the subclavian artery and run
a tortuous course through the brachial plexus,
which would make its use as a recipient vessel
dicult [65].
Recipient veins in the neck most commonly
include the internal and external jugular veins,
a stump of the facial or common facial veins, and
the transverse cervical vein. Similar to arterial
preservation, the type of neck dissection, the
proximity and stage of nodal disease, and the
skill and communication with the ablative surgeon often determine which vessels can be preserved. Venous dissection should be performed
with great care because even intact veins can
sustain intimal damage caused by careless dissection and promote thrombus formation.
The anterior jugular veins may be an option
for venous anastomosis only in cases in which
a tracheotomy is not used. Because many head
and neck cancer patients have a tracheotomy at
least temporarily, however, the caudal portion of
these veins runs in proximity to the stoma, which
increases risk of thrombosis caused by tracheal
contamination.
Finally, if none of the previously mentioned
veins is available, vein grafts to extend the pedicle
to the contralateral neck can be used. Common
sites for vein grafting are the greater and lesser
saphenous veins of the lower extremities. These
veins are not dicult to harvest and have an
adequate and constant caliber throughout most of
their length. Higher rates of ap failure have been
shown with aps that require vein grafts, however.
Vascular anastomosis
Before vascular anastomosis, the ap should
be inset to the defect and any necessary tunneling
for the passage of the pedicle should be

586

KIM & GHALI

performed. For oral cavity reconstruction, a watertight closure of the ap to the oral mucosa must
be obtained to prevent the catastrophic complication of salivary leak and stula formation. Salivary
contamination of the vascular anastomosis results
in thrombosis. For this purpose, the authors prefer
vicryl sutures passed in an interrupted horizontal
mattress fashion. This technique allows wound
edge eversion and a tight seal. Once the inset is
complete or nearly complete, attention can be
turned to preparing recipient and donor vessels.
Regardless of vessel type or the technique of
anastomosis being performed, the initial preparation of all donor and recipient vessels is essentially
the same. Extreme care should be exercised when
dissecting the vessels. At no time should the vessel
wall be grasped directly with any instrument. The
only part of the vessel that should be handled is
the adventitia.
The recipient vessels are prepared by 360 degree dissection under magnication for an ample
distance to allow manipulation of the vessel during suturing of front and back walls. A length of
3 to 4 cm of vessel should be adequate for this purpose in most cases. All branches should be clipped
meticulously, and a suitable site for anastomosis
that is relatively devoid of clips that may interfere
with suturing is identied. Small, noncrushing
microvascular clamps are used to occlude blood
ow proximally in recipient arteries and veins.
End-to-end anastomosis
The type of microvascular anastomosis applied
to any reconstruction depends on many factors,
including pedicle length, vessel size match, availability of recipient vessels, and surgeons preference. The EEA is probably the simplest
anastomotic concept in that the sharply cut end
of the donor vessel is sutured to the cut end of the
recipient vessel. The major variations of the EEA
technique involve the pattern of suture placement
at the anastomosis. Whichever scheme one chooses, the principle of placing six to ten sutures
around the circumference of the vessel perpendicular to the cut end of the vessel should be followed
strictly.
The original EEA for macrovascular anastomosis described by Alexis Carrel [18] involved
placing two sutures between 120 and 150 apart
and a third suture in a position to complete the
isosceles triangle (Fig. 12). The remaining sutures
are placed between the initial three sutures until
the anastomosis is complete. This technique was

Fig. 12. The Carrel EEA describes placing sutures at

120 to 150 apart and a third in a position to complete
an isosceles triangle in an attempt to minimize the risk of
back-wall suturing.

suggested to decrease the chance of back-wall

suturing the vessel, in which a suture is placed
through the intended vessel wall and the opposite
wall on the same vessel, which eectively diminishes the lumen of the vessel and most denitely
leads to an occlusion of the anastomosis.
An alternative technique used more commonly
by both authors is the halfway estimation
technique. The suturing of the vessel ends is begun
by placing two sutures approximately 180 or
more from each other (Fig. 13). The third suture is

Fig. 13. The halfway estimation technique involves

placing sutures 180 from each other and a third at their
midpoint. The subsequent sutures are placed by bisecting the remaining spaces.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

587

Fig. 14. (A) Tenting the lumen of the vessel by gently allowing the jewelers forceps provides tension to aid in placing
initial sutures. (B) Grasping only the adventitia, the opposite vessel ends suture is placed.

placed at the halfway point between the rst two.

The remaining vessel circumference between the
rst two sutures is then halved by each additional
suture until the side is complete. Once completed,
the clamps are ipped and a similar procedure is
performed on the back side. The most dicult sutures in this and in most techniques are the initial
anchoring sutures. To aid in placing slight tension on the vessel wall to facilitate suture passage,
the tips of the jewelers forceps may be placed
gently into the lumen of the vessel and slightly
opened to tent the wall (Fig. 14A). The opposite
vessel end is grasped by the adventitia, and the
needle can be passed at the corresponding position as the initial needle passage (Fig. 14B). The
sutures are tied with a surgeons knot with at least
two additional throws (Fig. 15 AC). The second
suture is placed in a similar fashion approximately
180 from the rst (Fig. 16). It is important to remember that the vessel should never be grasped by
any instrument and only the adventitia should be
used to manipulate the vessel.
Similar techniques for EEA can be used for
venous anastomosis; however, the vessels walls
are much thinner and more susceptible to damage
from aggressive traction, dissection, or other

manipulations. Smaller, more controlled movements are necessary to avoid damaging veins
when these techniques are used.
Another potential pitfall of venous anastomosis is a higher likelihood of back-wall suturing.
The technique of placing the jewelers forceps into
the lumen of the vessel to facilitate needle passage
helps to avoid this error. Alternatively, a skilled
assistant can help prevent back-wall suturing
by applying a gentle stream of heparinized saline
irrigation to the vessel lumen, which allows it to
balloon while placing the sutures (Fig. 17). This
technique, although useful, does create a change
in the optics of the procedure because the surgeon
must visualize the vessel through a pool of uid,
which decreases depth perception and may aect
the accuracy of needle placement.
End-to-side anastomosis
The ESA is applied in situations in which
proper size-matched recipient vessels are unavailable or the selected recipient vessels native ow is
desired to be maintained and not sacriced for the
sole purpose of the ap. Although somewhat
more technically dicult, the ESA is a reliable

Fig. 15. (AC) Tying sutures in square knots. The initial throw is a surgeons knot with at least two additional throws.

588

KIM & GHALI

Fig. 16. Placement of the second suture is approximately 180 from the rst.

microvascular technique and, because of the

elasticity of the vascular walls, has a tendency to
tent the anastomosis open. The most common
ESA performed clinically in head and neck reconstruction is for venous outow of donor vein
to the internal jugular vein. It is the preferred
method of venous anastomosis of some authors
[6668]. One study showed that the incidence of
internal jugular vein thrombosis on the rst postoperative day is 24.7% after modied neck dissections, however, which potentially leads to obvious
consequences to venous outow [69].
The donor vessel is clamped with a single
microclamp some distance proximal to the cut
end. The recipient vessel should be dissected

Fig. 17. The assistant may provide opening of the vessel

lumen by providing a gentle stream of heparinized
irrigation solution.

Fig. 18. Use of vessel loops to occlude blood ow for an

ESA.

circumferentially for an adequate length to allow

manipulation. When the recipient is the internal
jugular vein, care must be taken to avoid damage
to the vein itself, vagus nerve, and carotid artery.
Any branches should be clipped meticulously
before transection. For a large vessel such as the
internal jugular, the use of microclamps may not
be possible. Vessel loops may be used at either end
of the proposed ESA to occlude blood ow
temporarily (Fig. 18). Alternatively, a baby Satinsky vena cava clamp (Fig. 19) can be used atraumatically for this purpose.
A venotomy appropriate to the size of the
lumen of the donor vessel is created either with
scissors or a microscalpel. The venotomy should
be made slightly larger than the cut end of the
donor vessel and in an elliptical or diamond
shape. Two ends of the venotomy are sutured to
the donor vessel, which leaves the needles attached to each suture (Fig. 20A). The remaining
sutures are placed in a running fashion starting
with the back wall. These sutures can be placed
from inside the lumen (Fig. 20B). When completed, the end can be tied to the tail of the opposite anchoring suture. The same process is

Fig. 19. The baby Satinsky vena cava clamp also may be
used as a noncrushing clamp for ESA.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

589

Fig. 20. (A) The authors preferred method of ESA. The rst two sutures are placed at either end of the venotomy with
the ends of the suture left long. (B) The sutures are then placed in the back wall rst from inside the lumen in a running
fashion. This can be tied to the tail of the anchoring suture on the opposite side. The procedure same is performed on the
front wall.

performed on the front wall to complete the anastomosis. Otherwise, an interrupted suture technique may be performed (Fig. 21).
Attempts to compare the performance of EEA
and ESA have been made without success. One
study retrospectively reviewed more than 2000
microvascular anastomoses comparing EEA and
ESA. No signicant dierence in failure rate of
arterial or venous anastomoses was identied [70].
There may be some argument that ESA may be
more appropriate for some aps because they
may draw as much blood supply as they need
rather than become engorged with the terminal
blood ow from EEA [71]. In terms of venous

Fig. 21. Alternatively, an interrupted suture technique

may be used.

anastomoses, a recent study that compared anastomosis to the external jugular (EEA) versus directly to the internal jugular (ESA) or to its
branches (EEA) showed a signicantly higher
rate of failure with use of the external jugular
vein. These authors concluded that the internal
jugular system should be used whenever possible
[72].
Flap salvage
Although not a pleasant topic, the threatened
ap must be identied as early as possible to
maximize the possibility of salvaging the reconstruction. Depending on the timing of the thrombotic event, return to the operating room may be
necessary to surgically revise one or more of the
microvascular anastomoses. Alternatively, nonsurgical methods of salvage may be indicated if
the event occurs in the late postoperative period.
The techniques used during ap salvage procedures are not well documented in the literature.
The actual procedures depend on the clinical
situation and the appearance of the anastomosed
vessels. For example, a relatively common situation may occur in which an expanding hematoma
in the neck causes compression on the venous
outow of a ap, which causes it to appear purple
and edematous. In this circumstance, the neck
should be explored and the only salvage procedure that may be necessary is evacuation of the

590

KIM & GHALI

hematoma and bipolar coagulation or clipping of

any bleeding vessels so that venous outow can be
restored. There are other instances in which
venous and arterial anastomoses must be taken
down and revised, however. Most venous thromboses occurred within 24 hours of anastomosis
of rat femoral veins in a recent article [73]. In
a clinical situation, this would equate to an early
thrombosis of the venous anastomosis, which
warrants immediate surgical exploration.
Once a thrombosed artery or vein is identied,
an attempt should be made to determine the
extent of the clot by palpating of the length of
the vessel. The rmness of the thrombosis should
be noted at the site of the clot, whereas the vessel
is compressible proximally and distally. If identication of a thrombosis is uncertain, a strip test
(Fig. 22) can be performed in which the lumen of
the vessel is occluded downstream and a stripping
motion is applied upstream and released. Rapid
rell of the vessel usually indicates adequate ow.
The anastomosis should be opened by removing some or all of the sutures. The area of the

Fig. 22. The strip test. The downstream end of a segment

of vessel suspected to be thrombosed is gently occluded
with jewelers forceps. A second pair of forceps is used to
strip the vessel lightly in a smooth motion upstream and
released. Rapid rell of the vessel indicates adequate
ow.

thrombosis can be resected and a new anastomosis attempted at the new site or the thrombus can
be removed using Fogarty catheters. Which technique to use depends on the clinical circumstances, length and caliber of vessel, and quality
of the existing vessel opening. For example, if the
thrombosis occurred because of endothelial damage at the original site of anastomosis, that section
of artery or vein should be removed and revised at
an undamaged site.
Other situations that may necessitate revision
of microvascular anastomoses are kinking or
compression of the vascular pedicle. Compression
may occur because of a hematoma or external
sources, such as tracheal ties. These issues are best
avoided but their remedy is relatively straightforward. A kinked vessel can be handled by
reorienting and suspending the pedicle with carefully placed sutures or by taking down the
anastomosis, trimming the vessel, and revising,
or an alternative recipient vessel may be chosen to
improve the pedicles geometry.
The topic of exsanguination therapies must be
discussed as an option for aps whose venous
anastomoses are not able to be revised or in
which a revision has failed. The use of these
modalities allows for continued viability of tissues by maintaining perfusion while neovascularization is allowed to occur. Although shown to
be useful in nger replantations, the use of these
therapies with head and neck reconstruction
requires the exsanguination of a much larger
surface area of tissue, which may lead to considerable blood loss and the need for transfusion.
One paper recorded that an average of 13 U of
packed red blood cells per patient was necessary
to maintain adequate hemoglobin concentrations
during therapy [74]. The use of medicinal leech
therapy in head and neck MFTT should proceed
with much care so the leeches do not migrate
down the esophagus or larynx. Antibiotic prophylaxis with anti-pseudomonal penicillin or a
uoroquinolone should be instituted before leech
treatment to prevent infection by Aeromonas
hydrophila [75].
Finally, the use of thrombolytic agents, such as
streptokinase, urokinase, and tissue-type plasminogen activator, has been studied. Although some
promising laboratory and clinical data have been
published, their use is still limited [52]. This limitation may be caused by the risk of bleeding with
systemic exposure of these substances or possibly
a fear that the thrombosis will return upon discontinuation of the drug.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

Selected aps
Soft tissue aps
Radial forearm fasciocutaneous ap
Similar to the pectoralis major ap, the radial
forearm fasciocutaneous ap has become the
workhorse microvascular ap for the head
and neck. Since Soutars popularization of the
ap in the 1980s [76], the radial forearm fasciocutaneous ap has become one of the most commonly used and reliable soft-tissue free aps for
head and neck reconstruction. Based on the radial
artery, venae comitantes, or cephalic vein, the ap
can be transferred with bone (partial-thickness radius), tendon (palmaris longus), muscle (brachioradialis), nerve (lateral antebrachial cutaneous),
and skin and fascia.
If necessary, the skin of nearly the entire
forearm from the exor crease of the wrist to
the antecubital fossa and nearly the entire circumference except a small strip on the ulnar aspect of
the forearm can be raised with the ap. The ap
should be centered over the radial vessels and
cephalic vein, but the shape and overall dimensions can be tailored appropriately to the clinical
situation. Bilobed, dual cutaneous paddles, fascia
only, folded, and tubed aps have been designed
and used successfully for dierent reconstructive
needs. The skin of the volar forearm is generally
thin and pliable, with men having somewhat
thinner subcutaneous tissue in this region than
women, with varying densities of hair follicles.
This characteristic makes the radial forearm
a nearly ideal donor site for reconstruction of
areas that require thin, mobile tissue, including
the oor of mouth, tongue, and soft palate.
The blood supply to the lower arm and hand is
delivered by the radial and ulnar arteries, which
are branches of the brachial artery. The palmar
arches are the terminal ends of these arteries. The
radial artery terminates in the deep palmar arch,
and the ulnar artery leads to the supercial palmar
arch. After harvesting the radial forearm ap, the
blood supply to the hand and digits relies completely on the ulnar artery. Whereas the blood
supply to the third, fourth, and fth digits is
almost never in question, the perfusion to the
thumb and index nger is at greatest risk when
harvesting this ap. The existence of two anatomic
variations must be present for thumb or index
nger ischemia to occur: (1) an incomplete supercial palmar arch that does not send branches to
the thumb and index nger and (2) a complete lack

591

of communicating branches between the supercial and deep palmar arches. A cadaveric study
showed this combination of anomalies to be
present in approximately 12% of specimens [77].
The performance of an accurate Allens test is
essential in avoiding potential ischemic complications after ap harvest. In cases in which the
Allens test may be equivocal or dicult to interpret, radial and ulnar artery duplex may be performed that can determine objectively the pattern
of ow and reversal of ow upon occlusion of
the radial artery.
Flap harvest technique. Flap design and orientation should be performed with care because of the
complex, three-dimensional nature of oral cavity
reconstructions. Especially in areas such as the
soft palate, in which the skin paddle may need to
be folded for adequate reconstruction, the position and orientation of the pedicle is of paramount
importance. Once the design is conrmed, the arm
is exsanguinated with an Esmark bandage and
a tourniquet inated to 250 mm Hg. The time the
tourniquet has been in use should be recorded, as
should the time the harvested ap spends without
blood ow during inset and microvascular anastomosis. Although no absolute time limit has been
discovered for this ischemia time, every attempt
should be made to keep it to a minimum. Notably,
aps that have experienced longer than 6 hours of
ischemia are prone to experiencing a no-reow
phenomenon that is irreversible, in experimental
models, after 8 to 12 hours [42].
Flap elevation begins at the distal-most aspect
of the skin paddle. Incision through skin and
subcutaneous fat and fascia allows for identication of the distal radial artery and cephalic vein,
which are ligated and transected (Fig. 23). The

Fig. 23. Dissection of the radial forearm ap begins at

the distal end of the skin paddle. The distal radial artery
and venae comitantes are identied and ligated.

592

KIM & GHALI

supercial branches of the radial nerve also are

encountered at that time and may be preserved
or sacriced depending on surgeon preference.
The longitudinal portions of the skin paddle can
be incised through skin and subcutaneous tissue
down through the fascia of the brachioradialis
muscle radially and the exor muscles on the ulnar side. This dissection continues in the subfascial plane to the lateral intermuscular septum
between the brachioradialis and the exor carpi
radialis. Care must be taken to leave a thin lm
of paratenon on the tendons of the wrist exors
or wound complications caused by skin graft failure will be a problem.
At that point, the proximal portion of the skin
paddle can be incised. Care must be taken to avoid
injury to the cephalic vein as it emerges from the
ap in the subcutaneous fat. A linear incision is
made from the proximal portion of the ap to the
antecubital fossa to facilitate dissection of the
vascular pedicle. This incision is undermined in
the subcutaneous plane laterally and medially.
Completion of the dissection of the cephalic vein
can be performed at this stage to the appropriate
length and in a circumferential fashion. Finally,
the ap can be elevated from distal to proximal by
carefully dissecting the radial artery and its venae
comitantes. Vascular clips are necessary to control
the numerous branches to the surrounding musculature and radial bone. The fascia of the
brachioradialis must be incised along its length
to facilitate the proximal dissection of the radial
artery. The proximal extent of this dissection is
limited by the radial recurrent artery branch near
the antecubital fossa (Fig. 24).

Fig. 24. The completed forearm ap dissection. The radial artery (solid arrow) is separated from its venae comitantes by gentle dissection under loupe magnication.
The cephalic vein (dashed arrow) is also shown with
excellent caliber.

Once dissection is complete, the tourniquet is

deated and the ap allowed to reperfuse for
a period of at least 20 minutes. During this time,
hemostasis can be achieved carefully with cautery
and hemoclips and the nal preparations to the
recipient site can be performed. Once an adequate
reperfusion time has been allowed, the ap can be
harvested by using hemoclips at the desired
pedicle length. A suction drain is placed and
wound closure occurs in layers, with absorbable
sutures in the deep tissue to reapproximate the
divided fascia and the subdermal layer. Primary
closure of the proximal linear incision is easily
achieved and may be possible for the area of the
skin paddle. This area usually requires a skin graft
to avoid excessive tension, however. A split- or
full-thickness graft may be used and inset with
resorbable sutures. The graft should be perforated
to allow for seepage of uid during healing. This
graft should be compressed to the recipient bed by
petroleum gauze and gauze that is supported by
a plaster splint fabricated to the volar aspect of
the lower arm and hand. This splint should be left
in place for 7 days to prevent sheer forces on the
skin graft. Before placing the nal dressing, the
perfusion to the index nger and thumb should be
veried by checking capillary rell.
The ap should be inset and the microvascular
anastomosis performed with alacrity to minimize
ischemia time. The radial artery is usually a good
size match to many recipient arteries in the head
and neck, including the facial, superior thyroid,
and transverse cervical. The cephalic vein is also
of excellent caliber and can be matched to the
external jugular or facial veins (Fig. 25). This ap
has been described as a macro-microvascular
ap. The decision to use the cephalic vein and
the venae comitantes should be made on a caseby-case basis. Although the ap is known to

Fig. 25. A completed vascular anastomosis to the facial

artery and common facial vein.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

survive with drainage from either the supercial

or deep venous systems, many authors perform
more than one venous anastomosis when adequate recipient veins are present. The supercial
veins are usually chosen if only one venous anastomosis is to be performed because of their larger
caliber, thicker vessel walls and greater exibility
of recipient vessel selection and pedicle geometry.
Rectus abdominus myocutaneous ap
Based on the deep inferior epigastric artery and
vein (DIEA and DIEV), the rectus abdominus
myocutaneous ap has been a reliable and popular ap for many areas of reconstruction. The ap
oers the ability to transfer large amounts of skin,
subcutaneous tissue, muscle, and fascia along with
a fairly long vascular pedicle of good caliber. This
ap is well suited for defects that require large
volumes of tissue, such as total maxillectomy,
total glossectomy, skull base, or large scalp defects. Pedicled and free aps of the rectus muscle
based on the superior and inferior epigastric
arteries have been well described for breast
reconstruction.

593

Advantages for its use in head and neck

reconstruction include the ability for two teams
to operate at the same time. This advantage is
somewhat overstated in many cases in which two
teams would be crowded around a shorter patient,
however. The volume of tissue available is generally substantial with the ability to ll large areas
of dead space. A patients body habitus greatly
aects the amount of tissue available, however.
This volume decreases over time because of denervation of the muscle. The length of the pedicle
is adequate with a good caliber for microvascular
anastomosis. The skin paddle can be designed in
several dierent orientations to the rectus muscle
for various applications (Fig. 26AE). Finally, the
DIEA perforator ap, which allows for a much
thinner skin paddle without harvesting the muscle, has been described [63]. In this ap, the cutaneous perforators to the paraumbilical skin are
dissected through the rectus muscle to the DIEA
and DIEV. This perforator ap has enjoyed somewhat of a burst in popularity along with other perforator aps in recent times [78].

Fig. 26. (AE) Various skin paddle designs for the rectus abdominus myocutaneous ap. (Adapted from Urken ML.
Free aps: muscle and musculocutaneous aps. Rectus abdominis. In: Atlas of regional and free aps for head and
neck reconstruction. New York: Raven Press; 1995. p. 122; with permission.)

594

KIM & GHALI

Although the rectus muscle shares its blood

supply from the deep superior epigastric artery
and vein and the DIEA and DIEV, the DIEA and
DIEV have approximately twice the diameter of
the deep superior epigastric artery and vein. The
musculocutaneous perforators are directly associated with the DIEA and DIEV and only anastomose to the deep superior epigastric artery and
vein through a series of choke vessels in which
ow can be reversed. These dominant perforators
to the anterior abdominal skin are concentrated
around the umbilicus [79]. The DIEA branches
from the external iliac artery just superior to the
inguinal ligament. It travels superiorly and medially to penetrate the transversalis fascia 3 to 4
cm caudal to the arcuate line on the underside
of the rectus muscle.
The anatomy of the anterior abdominal wall
should be reviewed, because the preservation of
fascial sheaths helps prevent the formation of
abdominal hernias postoperatively. From the
pubis to the xiphoid, the fascial sheaths formed
by the brous aponeuroses of the abdominal
muscles change in their overall composition.
Most importantly, at the level of the anterior
superior iliac spine is the arcuate line, which
marks the transition between these dierent compositions. Above the line, the posterior rectus
sheath is formed by extensions of the transversalis
fascia and part of the internal oblique aponeurosis. This double layered sheath is adequate to
prevent hernias. Below the arcuate line, however,
the posterior rectus sheath is formed by the
transversalis fascia alone (Fig. 27). Bulging or hernia complications may occur if the fascia is not reinforced by preserving the anterior sheath below
the arcuate line and closing it as a separate layer.

Fig. 27. (A,B) Schematic representation of the layers of

the anterior abdominal wall above and below the arcuate line.

Because the skin paddle of the rectus abdominus myocutaneous ap can be oriented in so
many dierent congurations, ap design and
pedicle orientationdespecially in maxillary reconstructiondshould be performed with great
care. An oblique skin paddle that crosses the
abdominal wall obliquely from costal margin to
just inferior and lateral to the umbilicus, as
described by Urken [80], is useful in head and
neck reconstruction. Alternatively, vertical ord
less commonly for head and neck defectsdtransverse skin paddles may be used. The oblique
skin paddle design is described for ap harvesting.
Flap harvest technique. Flap elevation begins by
incising the superior margin of the skin paddle
through skin, subcutaneous tissue, fat, and fascia
to expose the rectus muscle from the linea alba to
the linea semilunaris. The inferior margin is then
incised to the same level. Once the linea semilunaris is identied superiorly and inferiorly, the
lateral extent of the rectus muscle is dened and
the remainder of the skin paddle laterally can be
incised down to subcutaneous fat. The skin can be
undermined at the level of Scarpas fascia until the
linea semilunaris is reached. The rectus sheath is
then incised at the linea semilunaris to preserve
the musculocutaneous perforators present over
the muscle. Similarly, the medial portion of the
skin paddle is incised and undermined deep to
Scarpas fascia until the linea alba is recognized
and the rectus sheath incised at this medial margin
(Fig. 28).
A vertical incision from the inferior portion of
the skin paddle oriented toward the femoral
vessels is made down to rectus sheath. The rectus
sheath is incised at the midpoint of the muscle to
expose the rectus muscle (Fig. 29). Once the entire

Fig. 28. Completion of the circumferential skin paddle

incision after identication of the rectus sheath both medially at the linea alba and laterally at the linea
semilunaris.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

595

needles or other sharp instruments to prevent

visceral injury. Primary skin closure generally can
be achieved with undermining.

Fig. 29. The exposure of the inferior portion of the rectus abdominus muscle after completing the vertical incision toward the femoral vessels. The inferior portion of
the rectus sheath (arrows) has been divided vertically to
expose the muscle. This portion of the sheath should be
approximated during closure of the donor site to prevent
hernia formation.

caudal portion of the muscle is in view, the dissection continues by elevating the muscle o the posterior rectus sheath using blunt dissection
(Fig. 30). The nerve supply, which includes mixed
motor/sensory nerves from the intercostal nerves,
is encountered at this point. Also on the undersurface of the muscle, the DIEA and DIEV come into
view. The nerves may be transected and the vascular pedicle followed inferiorly until the desired
length of pedicle is achieved. The DIEV joins to
form a single vein just before its take-o from
the external iliac vein.
Closure begins by approximating the inferior
portion of the cut anterior rectus sheath. The
superior portion that was harvested with the ap
also may be closed with relatively large, slowly
resorbable suture. Care should be taken not to
puncture the posterior rectus sheath with suture

Fig. 30. Completion of the dissection of the ap o the

abdominal wall with the pedicle still attached.

Anterolateral thigh ap
Recently, the reconstructive arena has been
inundated by a dramatic increase in the knowledge of the path of the cutaneous blood supply
from its underlying source vessel. This information is known as the angiosome theory, and the
concept indicates that the human body is composed of approximately 40 blocks of composite
tissue with unique cutaneous territories all supplied by a discrete source vessel. In a simplied
characterization, all arteries to the skin either
travel directly to their destination (direct perforators) or rst travel through some other tissue that
is usually muscle (indirect perforators) [81]. The
emergence of the anterolateral thigh ap, which
has been pioneered by colleagues in Asia, has
proved to be the spark in the recent explosion of
perforator ap popularity.
One of the foremost concerns of free ap
planning is donor site morbidity. The borrowing
of composite tissue from one site to rebuild
another is the basis for reconstructive surgery.
Although many of the aforementioned free aps
have reasonable morbidity proles, a new standard may be forming for acceptable donor site
morbidity in free ap reconstruction. With perforator aps, the skin territory to be harvested is
dissected directly with the skin perforator through
intervening tissue to the source vessel. As with the
deep inferior epigastric perforator ap versus the
rectus abdominus myocutaneous ap, the morbidity of harvesting the rectus muscle with skin is
avoided along with its attendant sequelae.
The anterolateral thigh ap is based on the
cutaneous perforators of the descending branch of
the lateral circumex femoral artery, a branch of
the profunda femoris system (Fig. 31). The pedicle
travels between the rectus femoris and the vastus
lateralis muscles along with the motor supply to
the vastus lateralis. The perforator may exist in
the intermuscular septum between these muscles
or travel through part of the vastus lateralis [82].
Some variations may exist, as noted in an article
by Koshima and colleagues [83], in which they
found the cutaneous perforators to the lateral
thigh existing as branches from the transverse
branch of the lateral circumex femoral artery
or directly from the deep femoral artery. This variation in vascular anatomy and a somewhat inconsistent location of cutaneous perforators have

596

KIM & GHALI

Fig. 32. Line drawn from anterior superior iliac spine to

the superior lateral pole of the patella represents the axis
of the descending branch of the lateral circumex femoral artery. At this lines midpoint, a 3-cm radius circle is
drawn and Doppler probing of the area usually identies
a perforator in the superior lateral quadrant of the circle.

Fig. 31. Schematic of the blood supply of the anterolateral thigh ap. (Adapted from Cipriani R, Contedini F,
et al. Three-dimensional reconstruction of the oral cavity
using the free anterolateral thigh ap. Plast Reconstr
Surg 2002;109:54; with permission.)

impeded this aps overall acceptance into head

and neck reconstruction.
No special preoperative evaluations need to be
performed before ap harvesting. Doppler probe
localization of the cutaneous perforators may be
performed but may not be accurate. The total area
that this ap can support is reported up to 800
cm2 from the level of the greater trochanter of the
femur to a line 3 cm above the patella. The dominant perforator to the lateral thigh skin is identied by drawing a line from the anterosuperior
iliac spine to the superolateral corner of the patella. The midpoint of this line is marked and
a 3-cm radius circle is drawn. Most of the cutaneous perforators are located in the inferolateral
quadrant of this circle. The aps skin paddle
should be designed to center on this area (Fig. 32).
Flap harvest technique. Flap harvesting begins by
incising skin and subcutaneous tissue on the
medial surface of the skin paddle. This dissection
continues down to the tensor fascia lata. Depending on the thickness of tissue needed, the ap may
be elevated to include or not include the tensor
fascia lata. A great advantage to this ap is the
possibility of harvesting a thin apdas little as
5 mm of thickness is reported to maintain the

suprafascial vascular plexus based on the perforating vessels. In either case, the skin paddle is
gently dissected laterally until the cutaneous
perforators are identied. These perforators are
followed either through the intermuscular septum
between the rectus femoris and vastus lateralis (in
the case of septocutaneous perforators) or through
the vastus lateralis (in the case of musculocutaneous perforators) (Fig. 33). If dissection through
the vastus lateralis is necessary, careful ligation of
muscular branches must be performed. This procedure is continued until the source vessel is identied (descending branch of the lateral circumex

Fig. 33. With musculocutaneous perforators, the vascular pedicle must be followed through the vastus lateralis
to the main source vessel. Here, the dissection through
the vastus lateralis muscle is complete, and the vascular
pedicle in clear view. The rectus femoris muscle is seen
under the retractors.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

femoral), and it may be dissected for the desired

length up to 16 cm [84].
Care must be taken when dissecting the source
artery and vein to avoid injury to the nerve to
the vastus lateralis. The muscle should be approximated if intramuscular dissection was necessary
after adequate hemostasis is achieved. The wound
edges often can be closed primarily with wide
undermining when the skin paddle dimensions
remain smaller than 6 to 9 cm. Larger defects may
require skin grafting.
A possible disadvantage to this ap is the
inconsistent location and size of the cutaneous
perforators. The rare incidence of a total lack of
perforators has been reported [85]. Celik and
colleagues [86] regarded many of these reports to
be a lack of recognition of small perforators or
a surgeons inability to trace musculocutaneous
perforators through the vastus lateralis. Because
of these possible variations in anatomy, some authors recommend that the reconstructive surgeon
be prepared to convert from an anterolateral thigh
ap to an anteromedial thigh ap or a tensor fascia lata ap.
Composite bone-containing aps
Fibula osteocutaneous ap
Although the bula ap was rst described by
Taylor and colleagues [87] in 1975 for lower extremity reconstruction, Hidalgo introduced the
bula osteocutaneous ap for oromandibular reconstruction in 1989 [88]. It is currently the most
commonly used microvascular bone-containing
ap for mandibular reconstruction. Factors that
aected its widespread use include relative ease
of harvest, low morbidity prole, distant site to facilitate two-team approaches, and ability to contour the bone to the required shape.
The bula is a nonweight-bearing bone with
an average length of 22 to 25 cm of bone available
for harvest. Based on the peroneal artery and
venae comitantes, the bula ap can be transferred as a bone-only ap or an osteocutaneous
ap with the skin of the lateral calf perfused by
cutaneous perforators traveling through or along
the posterior crural septum. The length of bula
that can be obtained is the entire bone except for 6
to 7 cm at the proximal and distal ends of the
bone to preserve the integrity of the knee and
ankle joints. The length of the vascular pedicle
that can be harvested is limited by the bifurcation
of the posterior tibial artery. Proximal bular
bone not necessary for reconstruction can be

597

discarded after careful subperiosteal dissection to

increase the eective pedicle length.
The tubular bular bone must be contoured
for most applications for mandibular or maxillary
reconstruction. These contours may be created by
careful subperiosteal dissection circumferentially
around where the osteotomy is to be performed
while carefully protecting the vascular pedicle
during bone cuts. These closing osteotomies
allow for a rough approximation of the bula to
the normal contour of the mandible (Fig. 34).
Although a controversial topic, before harvesting of the bula osteocutaneous ap the perfusion
of the foot should be evaluated in some manner.
Because many patients with head and neck cancer
also have risk factors for peripheral vascular
disease, a patient who has compromised perfusion
to the foot after ap harvesting is a possibility.
The presence of a peroneal artery magna variation
or two-vessel ow to the distal lower extremity
also may place the foot at risk for postoperative
ischemia. Angiography has been the standard
imaging modality for the preoperative assessment
of the lower extremities before bula osteocutaneous ap harvest, but magnetic resonance angiography [89] and recently CT angiography [90] in
some centers have been shown to provide adequate information, although they are technique
and equipment sensitive. Some surgeons do not
believe that preoperative assessment is necessary
and rely on clinically evaluating the anterior and
posterior tibial vessels during ap elevation to imply adequate ow to the distal extremity [91].
Other surgeons, however, have shown that a sucient percentage of bula ap candidates have signicant nding that may alter the reconstructive
plan when imaged preoperatively (Fig. 35) [92].
The presence of a septocutaneous or musculocutaneous perforator to supply the attached

Fig. 34. Closing osteotomies allow for contouring of the

tubular bula bone.

598

KIM & GHALI

Fig. 35. Standard lower extremity angiography shows

only two vessel ow to bilateral distal lower extremities
with early loss of ow to the anterior tibial arteries bilaterally. This patient is not a good candidate for a bula
osteocutaneous ap because ischemic injury to the foot
is highly likely.

skin paddle in a bula osteocutaneous ap should

be identied during the initial stages of the
dissection. If no perforator is seen, a dierent
soft tissue ap option should be pursued. The
overall reliability of the bular skin paddle is
reported as 90% to 100%. Skin paddles should be
centered on the junction of the middle and distal
thirds of the bula with a relatively long axial
dimension to increase the likelihood of capturing
these perforators.
Damage to the common peroneal nerve can be
avoided by preserving the proximal 6 to 7 cm of
bula and avoiding excessive traction during
dissection of the proximal pedicle. An injury to
this nerve can lead to an equinovarus deformity of
the foot along with anesthesia to the anterior and
lateral lower leg and dorsum of the foot. Damage
to the deep peroneal nerve or scarring of the exor
hallicus longus may result in weakness in dorsiexion of the foot. Diculty with ambulation
usually resolves within a few months when no
neural injury is present.
Flap harvest technique. Landmarks for the harvesting of the bula osteocutaneous ap include
the lateral epicondyle of the ankle and the bular
head. A line between these points estimates the
posterior crural septum. The segments of bula
not harvested to preserve the joints are marked

and an appropriately sized skin paddle is drawn

centered on the posterior crural septum and the
middle and distal thirds of the bula bone
(Fig. 36). The patients lower extremity is prepared circumferentially from toes to hip, and
a large pneumatic tourniquet is placed on the
thigh and inated to 350 mm Hg.
As described by Fleming and colleagues [93],
the dissection begins by incising the skin of the anterior portion of the skin paddle through subcutaneous tissue and fascia overlying the peroneus
longus muscle. A subfascial plane of dissection
proceeds posteriorly until the posterior crural septum is encountered. Septocutaneous or musculocutaneous perforators should be visible. The
posterior crural septum is followed to the bula,
dissection along the anterior portion of the bula
is performed, and the anterior tibial vessels are
identied. Staying close to the bula, the medial
dissection is continued to identify the interosseous
membrane (Fig. 37). The initial incision is extended proximally and distally to facilitate this
dissection.
The bula can be osteotomized to allow the
completion of the dissection. Subperiosteal dissection occurs at proximal and distal ends circumferentially around the bula to protect the
vascular pedicle. Oscillating, reciprocating, or
Gigli saws may be used depending on surgeon
preference. Once the osteotomies are complete,
the distal peroneal artery and vein can be identied, ligated, and divided. At that time, the

Fig. 36. Markings for a bula osteocutaneous ap. The

lateral epicondyle of the ankle and bular head at the
knee are the main landmarks, and a line between the two
points represents the posterior crural septum. The skin
paddle should be centered over this line and lie between
the middle and distal thirds of the bone to capture cutaneous perforators in that area. Here, the entire palpable
bula is marked for demonstration.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

Fig. 37. Dissection along the posterior crural septum

transitions over the anterior and medial surface of the
bula. The interosseous septum is visualized and the anterior tibial vessels can be visualized or palpated in the
tissues retracted anteriorly.

posterior portion of the skin paddle is incised

down through the fascia overlying the soleus and
gastrocnemius muscles. Depending on the type of
perforators present, a cu of soleus may or may
not need to be included with posterior crural
septum to ensure adequate vascularity to the skin
paddle.
With the bula distracted, the medial dissection of the ap may proceed. The interosseous
septum is divided to reveal the chevron-shaped
tibialis posterior muscle. The peroneal artery and
venae comitantes are followed through this muscle, with the surgeon carefully ligating branches
with hemoclips. This process is continued until the
branching of the posterior tibial artery is identied (Fig. 38). The exor hallicus longus should be
transected at this time and the tourniquet

Fig. 38. Once the osteotomies are created, the bula can
be retracted laterally, thus facilitating medial dissection
of the vascular pedicle.

599

released. The leg should be allowed to reperfuse

for a period of at least 20 minutes before nal
ap harvest. During that time, capillary rell,
pulses, and warmth of the foot should be
evaluated.
Once harvested, the ap is transported to the
recipient site and contoured to its desired shape by
creating closing osteotomies and discarding unnecessary bone. The pedicle should be fashioned
on the lingual side to facilitate rigid xation, with
monocortical screws axing the ap to the reconstruction plate. The skin paddle is inset in the
standard fashion after bony contouring and
xation are complete. The skin paddle should be
transposed over the lateral surface of the bula
over the reconstruction plate to provide some
protection from exposure. The artery and venae
comitantes are a good size match to most of the
same vessels indicated previously, although their
close association limits recipient choices to an
artery and vein near one another.
The donor site is managed by obtaining adequate hemostasis, light reapproximation of the
muscles to decrease dead space, and placement of
two large caliber (19 French Blake, Ethicon, Inc.,
Somerville, NJ) suction drains at dierent tissue
levels. The cutaneous defect can be undermined and
primarily closed, but if closure is too tight, a compartment syndrome may occur. A split-thickness
skin graft takes nicely on the vascular bed of
muscle. A posterior splint is fabricated to support
the ankle in a dorsiexed position, and the patient
should remain in a nonweight-bearing position for
7 days. At that point, light physical rehabilitation
may begin.
Iliac crest osteocutaneous ap
The iliac crest osteocutaneous ap is based on
the deep circumex iliac artery and vein (DCIA
and DCIV) as described by Taylors dye injection
studies in 1979 [94]. Before then, vascularized iliac
crest bone was harvested with the supercial circumex iliac artery, ascending branch of the lateral circumex artery, and the superior branch
of the superior gluteal artery. Along with skin
and bone, the ap can include the internal oblique
muscle based on the ascending branch of the
DCIA, as described by Ramasastry and colleagues [95]. This allows the exibility of a thin,
mobile soft tissue component along with the relatively immobile cutaneous and osseous constituents of the ap.
Probably the greatest advantage of the DCIAbased aps is the quality and quantity of the bone

600

KIM & GHALI

that can be harvested. The ilium is ideally suited

for reconstruction of mandibular defects of nearly
all but the largest of mandibular resections. It is
probably most suited for somewhat smaller defects (up to 16 cm) of the mandibular ramus and
body, however, because of its natural curvature
and ease of designing patterns to reconstruct the
mandibular angle without osteotomies. Osteotomies may be performed to recreate the symphysis
of the mandible if the medial periosteum and
vascular pedicle are preserved.
There are several disadvantages to the DCIA
aps, however. First is the possibility of abdominal wall weakness and hernia. Although reported
as rare, it is not dicult to imagine the potential
eects of harvesting the internal oblique muscle.
Second, because the use of full-thickness ilium
makes stripping of the lateral thigh musculature
necessary, the occurrence of postoperative gait
disturbances may become a focus in patient
dissatisfaction.
The DCIA is a branch of the external iliac
artery that originates approximately 1 to 2 cm
above the inguinal ligament. The artery travels
laterally toward the anterior superior iliac spine
and travels along the medial surface of the ilium in
a groove formed by the junction of the transversalis fascia and the iliacus muscle approximately 0.4 to 2.2 cm inferior to the iliac crest
(Fig. 39). The ascending branch to the internal

Fig. 39. Schematic version of the vascular supply to the

iliac crest osteomyocutaneous ap.

oblique is given o variably along this course

and terminates within the muscle serving no cutaneous perforators. Perforators to the ilium and
skin perforators are given o throughout its
course. The artery terminates as a dominant skin
artery approximately 10 cm posterior to the anterior superior iliac spine [81]. Of note, the femoral
nerve is situated deep to the DCIA just lateral to
its take-o from the external iliac artery and the
lateral femoral cutaneous nerve travels medial to
the anterior superior iliac spine and may run supercial or deep to the DCIA.
Flap harvest technique. The femoral vessels are
palpated and identied at the inguinal ligament.
The take-o of the DCIA and DCIV is estimated
just above the inguinal ligament and medial to the
anterior superior iliac spine. The skin paddle
should be large enough to capture the cutaneous
perforators just medial to the ilium. A fusiform
skin paddle is marked along the axis of the
anterior superior iliac spine to the inferior border
of the scapula. The incision is begun at the
superior border of the skin paddle extending
medially toward the iliac vessels. This incision is
carried through skin, subcutaneous tissue to the
external oblique muscle. The external oblique
muscle is incised to reveal the internal oblique
muscle, leaving a cu of muscle attached to the
ilium of approximately 2 cm to preserve the
cutaneous perforators. The internal oblique is
also incised, leaving 2 cm attached to the inner
ilium. This incision reveals the ascending branch
of the DCIA, which can be ligated. An alternative
harvesting technique described in detail by Urken
[96] can be performed in which the internal oblique is harvested in its entirety with preservation
of the ascending branch.
At this stage, the cut ascending branch is
followed through the transversus abdominis until
the DCIA and DCIV are encountered. The
pedicle can be followed down to their origins
from the external iliacs. Laterally, the transversus
abdominis is transected, which leaves the same 2cm cu attached to the medial ilium as the other
layers of the abdominal wall. This procedure
allows exposure of the lateral femoral cutaneous
nerve and the iliacus muscle. The iliacus is then
incised to expose the ilium at the desired point.
One must remember, however, that the DCIA and
DCIV run in the groove between the iliacus and
the transversus abdominis, so a 2-cm cu of
iliacus should remain attached to the ilium to
help preserve the pedicle.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

At that stage, the inferior portion of the skin

paddle can be incised. The incision is brought
through skin, subcutaneous tissue, and the tensor
fascia lata and the tendon of the gluteus medius,
which provides exposure to the lateral ilium in
a subperiosteal plane for performance of osteotomies. These osteotomies are performed from the
lateral border while protecting the vascular pedicle medially. The vascular pedicle may be transected at its origin with double ligatures.
As with the rectus abdominus ap, wound
closure is an important part of this technique to
help prevent hernia formation. The transversus
abdominus and the internal oblique are approximated to the iliacus muscle with 2-0 vicryl sutures.
Care must be taken to avoid injury to the external
iliac vessels or the femoral nerve when placing deep
sutures. A second layer of muscular wall closure is
also performed by approximating the external
oblique and the tensor fascia lata and the gluteus
medius. Suction drains should be used between
layers, and nally the skin is closed with staples.
The lateral border of the bone is exposed in the
subperiosteal plane, and this side of the bone
should be xated to the reconstruction plate with
monocortical screws. The skin paddle can be used
for oral lining or inverted to provide external skin
coverage. If the internal oblique muscle is included
in the ap, multiple mucosal or skin surfaces can be
relined with this highly mobile portion of the ap.
Any osteotomies that must be performed should be
done with care by limited dissecting of the medial
periosteum and protecting the pedicle. Any large
gaps between osteotomies should be lled with free
corticocancellous bone chips.
Summary
The philosophy of reconstructive surgery has
undergone drastic changes within the last 30
years. With further understanding and excellent
anatomic studies on the angiosome patterns of the
human body, reconstructive surgeons have devised pedicled aps to supplant the random
pattern aps of the past. With renements of
microvascular techniques and improved instrumentation, free tissue transfer, a procedure once
thought of as high risk and radical, is currently
the reconstructive modality of choice for many
head and neck reconstructions. Although not fool
proof and certainly not indicated for every case of
head and neck reconstruction, often the consequences of aggressive ablative surgery of the head
and neck can be lessened by the appropriate

601

selection of donor tissue transplanted with microvascular techniques. In instances in which microvascular techniques cannot be used, harvesting of
pedicled aps continues to be a simple and reliable
technique.
References
[1] Jemal A, Thomas A, Murray T, et al. Cancer
statistics, 2002. CA Cancer J Clin 2002;52:2347.
[2] Wingo PA, Tong T, Bolden S. Cancer statistics,
1995. CA Cancer J Clin 1995;45(1):830.
[3] Shaw JP. Oral cavity and oropharynx. In: Shaw JP,
Patel SG, editors. Head and neck surgery and oncology. 3rd edition. London: Mosby; 2003. p. 173233.
[4] Wilk RM, Potter BE. Soft-tissue reconstruction of
tumor defects in the head and neck. Oral Maxillofac
Surg Clin North Am 1997;9:47788.
[5] Mathes SJ, Nahai F. Classication of the vascular
anatomy of muscles: experimental and clinical correlation. Plast Reconstr Surg 1981;67:17787.
[6] Taylor G, Palmer J. The vascular territories (angiosomes) of the body: experimental study and clinical
applications. Br J Plast Surg 1987;40:11341.
[7] Ariyan S. The pectoralis major myocutaneous ap:
a versatile ap for reconstruction in the head and
neck. Plast Reconstr Surg 1979;63:7381.
[8] Moloy P, Gonzales F. Vascular anatomy of the pectoralis major myocutaneous ap. Arch Otolaryngol
Head Neck Surg 1986;112:669.
[9] Rikimaru H, Kiyokawa K, Inoue Y, et al. Threedimensional anatomical vascular distribution in the
pectoralis major myocutaneous ap. Plast Reconstr
Surg 2005;115:134252.
[10] Magee W, McCraw J, Horton CE, et al. Pectoralis
paddle myocutaneous ap: the workhorse of
head and neck reconstruction. Am J Surg 1980;
140:50713.
[11] Reid CD, Taylor GI. The vascular territory of the
acromiothoracic axis. Br J Plast Surg 1984;37:
194212.
[12] Tansini I. Spora il mio muovo processo di amputazione della mammaella per cancre. Riforma Med
(Palermo, Napoli) 1896;12:3.
[13] Quillen C, Shearin J, Georgiade N. Use of the latissimus dorsi myocutaneous island ap for reconstruction in the head and neck area. Plast Reconstr Surg
1978;62:1137.
[14] Urken ML, Sullivan MJ. Latissimus dorsi. In:
Urken ML, Cheney ML, Sullivan MJ, et al, editors.
Atlas of regional and free aps for head and neck
reconstruction. New York: Raven Press; 1995.
p. 23758.
[15] Bakamjian VY. A two-stage method for pharyngoesophageal reconstruction with a primary pectoral
skin ap. Plast Reconstr Surg 1965;36:17384.
[16] Urken ML, Biller HF. Deltopectoral. In: Urken ML,
Cheney ML, Sullivan MJ, et al, editors. Atlas of re-

602

[17]
[18]

[19]

[20]
[21]

[22]

[23]

[24]

[25]

[26]
[27]

[28]

[29]

[30]

[31]

[32]

[33]

KIM & GHALI

gional and free aps for head and neck reconstruction. New York: Raven Press; 1995. p. 8596.
McGregor IA, Morgan G. Axial and random
pattern aps. Br J Plast Surg 1973;26:20213.
Carrel A. The operative technique of vascular anastomosis and transplantation of organs. Lyon Med
1902;98:85964.
Nylen CO. The microscope in aural surgery: its rst
use and later development. Acta Otolaryngol Suppl
1954;116:22640.
Jacobson JH, Suarez EL. Microsurgery in anastomosis of small vessels. Surg Forum 1960;11:2435.
Buncke HT, Schulz WP. Total ear reimplantation in
the rabbit utilizing microminiature vascular anastomoses. Br J Plast Surg 1966;19:1522.
Seidenberg B, Rosenak SS, Hurwitt ES, et al. Immediate reconstruction of the cervical esophagus by
a revascularized isolated jejunal segment. Ann Surg
1959;149(2):16271.
McLean DH, Buncke HJ Jr. Autotransplant of
omentum to a large scalp defect, with microsurgical
revascularization. Plast Reconstr Surg 1972;49(3):
26874.
Panje WR, Krause CJ, Bardach J, et al. Reconstruction of intraoral defects with the free groin ap. Arch
Otolaryngol 1977;103:7883.
Carlson ER, Marx RE. Part II. Mandibular
reconstruction using cancellous cellular bone grafts.
J Oral Maxillofac Surg 1996;54(7):88997.
Khouri RK. Free ap surgery: the second decade.
Clin Plast Surg 1992;19:75761.
Brown MR, McCullough TM, Funk GF, et al. Resource utilization and patient morbidity in head
and neck reconstruction. Laryngoscope 1997;107:
102831.
Shusterman MA, Kroll SS, Weber RS, et al. Intraoral soft tissue reconstruction after cancer ablation:
a comparison of the pectoralis major ap and
the free radial forearm ap. Am J Surg 1991;162:
3979.
Kroll SS, Reece GP, Miller MJ, et al. Comparison of
the rectus abdominus free ap with the pectoralis
major myocutaneous ap for reconstructions in the
head and neck. Am J Surg 1992;164:6158.
Urken ML, Vickery C, Weinberg H, et al. The
neurovascular radial forearm ap in head and neck
reconstruction: a preliminary report. Laryngoscope
1990;100:16173.
Huang RD, Silver SM, Hussain A, et al. Pectoralis
major myocutaneous ap: analysis of complications
in a VA population. Head Neck 1992;14:1026.
Kroll SS, Schusterman MA, Reece GP. Costs and
complications in mandibular reconstruction. Ann
Plast Surg 1992;29:3417.
Disa JJ, Pusic AL, Hidalgo DH, et al. Simplifying
microvascular head and neck reconstruction: a rational approach to donor site selection. Ann Plast Surg
2001;47:3859.

[34] Haughey BH, Wilson E, Kluwe L, et al. Free ap

reconstruction of the head and neck: analysis of
241 cases. Otolaryngol Head Neck Surg 2001;125:
107.
[35] Ayalya C, Blackwell KE. Protein C deciency in microvascular head and neck reconstruction. Laryngoscope 1999;109:25965.
[36] Peverill RE. Hormone therapy and venous thromboembolism. Best Pract Res Clin Endocrinol Metab
2003;17:14964.
[37] Polancyk CA, Marcantonio E, Goldman L, et al.
Impact of age on perioperative complications and
length of stay in patients undergoing noncardiac surgery. Ann Intern Med 2001;134:63743.
[38] Bhattacharyya N, Fried MP. Benchmarks for mortality, morbidity and length of stay for head and
neck surgical procedures. Arch Otolaryngol Head
Neck Surg 2001;127:12732.
[39] Shaari CM, Urken ML. Complications of head and
neck surgery in the elderly. Ear Nose Throat J 1999;
78:5102.
[40] Hidalgo DA, Jones CS. The role of emergent exploration in free tissue transfer: a review of 150
consecutive cases. Plast Reconstr Surg 1990;86:
4928.
[41] Kerrigan CL, Zelt RG, Daniel RK. Secondary critical ischemia time of experimental skin aps. Plast
Reconstr Surg 1984;74:5226.
[42] May JW, Chait LA, Obrient BM, et al. The noreow phenomenon in experimental free aps.
Plast Reconstr Surg 1978;61:25667.
[43] Jones NF. Intraoperative and postoperative monitoring of microsurgical free tissue transfers. Clin
Plast Surg 1992;19:78397.
[44] Achauer BM, Black KS. Transcutaneous oxygen
and aps. Plast Reconstr Surg 1984;74:7212.
[45] Dunn RM, Kaplan IB, Moncoll J, et al. Experimental and clinical use of pH monitoring of free tissue
transfers. Ann Plast Surg 1993;31:53945.
[46] Menick FJ. The pulse oximeter in free muscle ap
surgery: a microvascular surgeons sleep aid.
J Reconstr Microsurg 1998;4:3314.
[47] Harrison DH, Firling M, Mott G. Experience in
monitoring the circulation in free ap transfers. Plast
Reconstr Surg 1981;68:54353.
[48] Goldberg S, Sepka RS, Perona BP, et al. Laser
Doppler blood ow measurements of common
cutaneous donor sites for reconstructive surgery.
Plast Reconstr Surg 1990;85:5816.
[49] Yuen JC, Feng Z. Monitoring free aps using
laser Doppler owmeter: 5 year experience. Plast
Reconstr Surg 2000;105:5561.
[50] Zinberg EM, Wood MB, Brown ML. Vascularized
bone transfer: evaluation of viability by postoperative bone scan. J Reconstr Microsurg 1985;2:139.
[51] Rosenberg RD. Actions and interactions of antithrombin and heparin. N Engl J Med 1975;292:
14651.

POSTABLATIVE RECONSTRUCTION FOR ORAL CANCER

[52] Conrad MH, Adams WP. Pharmacologic optimization of microsurgery in the new millennium. Plast
Reconstr Surg 2001;108:200896.
[53] Khouri RK, Cooley BC, Kenna DM, et al. Thrombosis of microvascular anastomoses in traumatized
vessels: brin versus platelets. Plast Reconstr Surg
1990;86:1107.
[54] Vlastou C, Earle AS. Intraoperative heparin in replantation surgery: experimental study. Ann Plast
Surg 1983;10:1124.
[55] Das SK, Miller JH. Current status of topical antithrombotic agents in microvascular surgery. Microsurgery 1994;15:6302.
[56] Johnson PC, Barker JH. Thrombosis and antithrombotic therapy in microvascular surgery. Clin
Plast Surg 1992;19:799807.
[57] Peter FW, Steinau HU, Homann HH, et al. Aspirin
and microvascular surgery: an update. Plast
Reconstr Surg 2003;112:136870.
[58] Peter FW, Franken RJPM, Wang WZ, et al. Eect
of low dose aspirin on thrombus formation at
arterial and venous microanastomoses and on the
tissue microcirculation. Plast Reconstr Surg 1997;
99:11221.
[59] Disa JJ, Polvora VP, Pusic AL, et al. Dextranrelated complications in head and neck microsurgery: do the benets outweight the risk? A
prospective randomized analysis. Plast Reconstr
Surg 2003;112:15349.
[60] Renck H, Ljungstrom KG, Rosenberg B, et al. Prevention of dextran-induced anaphylactic reactions
by hapten inhibition: II. A comparison of the eects
of 20ml Dextran 1, 15% administered either
admixed or before dextran 70 or dextran 40. Acta
Chir Scand 1983;149:34953.
[61] Esclamado RM, Carroll WR. The pathogenesis of
vascular thrombosis and its impact in microvascular
surgery. Head Neck 1999;21:35562.
[62] Walenga JM, Pifarre R, Hoppensteadt DA, et al.
Development of recombinant hirudin as a therapeutic anticoagulant and antithrombotic agent: some
objective considerations. Thromb Haemost 1989;
15:31633.
[63] Khouri RK, Sherman R, Buncke HJ, et al. A phase
II trial of interluminal irrigation with recombinant
human tissue factor pathway inhibitor to prevent
thrombosis in free ap surgery. Plast Reconstr
Surg 2001;107:40815.
[64] Ching S, Thoma A, Monkman J, et al. Inhibition of
microsurgical thrombosis by the platelet glycoprotein IIb/IIIa antagonist SR121566A. Plast Reconstr
Surg 2003;112:17785.
[65] Urken ML. Recipient vessel selection in free tissue
transfer to the head and neck. In: Urken ML,
Cheney ML, Sullivan MJ, et al, editors. Atlas of regional and free aps for head and neck reconstruction. New York: Raven Press; 1995. p. 3317.
[66] Varvares MA, Cheney ML. Free aps for head and
neck reconstruction. In: Cheney ML, editor. Facial

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78]

[79]

[80]

[81]

603

surgery: plastic and reconstructive. Baltimore

(MD): Williams and Wilkins; 1997. p. 487507.
Chalian AA, Anderson TD, Weinstein GS, et al. Internal jugular vein versus external jugular vein anastomosis: implications for successful free tissue
transfer. Head Neck 2001;23:4758.
Yamamoto Y, Nohira K, Kuwahara H, et al. Superiority of end-to-side anastomosis with the internal
jugular vein: the experience of 80 cases in head and
neck microsurgical reconstruction. Br J Plast Surg
1999;52:8891.
Quraishi HA, Was MK, Granke K, et al. Internal
jugular vein thrombosis after functional and
selective neck dissection. Arch Otolaryngol Head
Neck Surg 1997;123:96973.
Samaha FJ, Oliva A, Buncke GM, et al. A clinical
study of end-to-end versus end-to-side techniques
for microvascular anastomosis. Plast Reconstr
Surg 1997;99:110911.
Sasmor MT, Reus WF, Straker DJ, et al. Vascular
resistance considerations in free-tissue transfers.
J Reconstr Microsurg 1992;8:195200.
Chalian AA, Anderson TD, Seinstein GS, et al. Internal jugular vein versus external jugular vein anastomosis: implications for successful free tissue
transfer. Head Neck 2001;23:4758.
Hui KC, Zhang F, Shaw WW, et al. Assessment of
the patency of microvascular venous anastomosis.
J Reconstr Microsurg 2002;18:1114.
Chepeha DB, Nussenbaum B, Bradford CR, et al.
Leech therapy for patients with surgically unsalvageable venous obstruction after revascularized free
tissue transfer. Arch Otolaryngol Head Neck Surg
2002;128:9605.
Kubo T, Yano K, Hosokawa K. Management of
aps with compromised venous outow in head
and neck microsurgical reconstruction. Microsurgery 2002;22:3915.
Soutar DS, Scheker CR, Tanner NSB, et al. The radial forearm ap: a versatile method for intraoral
reconstruction. Br J Plast Surg 1983;36:18.
Urken ML. Free aps: fascial and fasciocutaneous
aps. Radial forearm. In: Atlas of regional and
free aps for head and neck reconstruction. New
York: Raven Press; 1995. p. 14968.
Koshima I, Moriguchi T, Fukuda H, et al. Free
thinned paraumbilical perforator-based aps.
J Reconstr Microsurg 1991;7:3136.
Boyd JB, Taylor GI, Corlett R. The vascular
territories of the superior epigastric and the deep inferior epigastric systems. Plast Reconstr Surg 1984;
73:114.
Urken ML. Free aps: muscle and musculocutaneous aps. Rectus abdominis. In: Atlas of regional
and free aps for head and neck reconstruction.
New York: Raven Press; 1995. p. 11938.
Hallock GG. Direct and indirect perforator aps:
the history and the controversy. Plast Reconstr
Surg 2003;111:85565.

604

KIM & GHALI

[82] Song YG, Chen GZ, Song YL. The free thigh ap:
a new free ap concept based on the septocutaneous
artery. Br J Plast Surg 1984;37:14959.
[83] Koshima I, Fukuda H, Yamamoto H, et al. Free
anterolateral thigh aps for reconstruction of head
and neck defects. Plast Reconstr Surg 1993;92:
4218.
[84] Wei FC, Jain V, Celik N, et al. Have we found an
ideal soft-tissue ap? An experience with 672 anterolateral thigh aps. Plast Reconstr Surg 2002;109:
221926.
[85] Kimata Y, Uchiyama K, Ebihara S, et al. Anatomic
variations and technical problems of the anterolateral thigh ap: a report of 74 cases. Plast Reconstr
Surg 1998;102:151723.
[86] Celik N, Wei FC, Lin CH, et al. Technique and strategy in anterolateral thigh perforator ap surgery,
based on an analysis of 15 complete and partial failures in 439 cases. Plast Reconstr Surg 2002;109:
22116.
[87] Taylor GI, Miller DH, Ham FJ. The free vascularized bone graft: a clinical extension of microvascular techniques. Plast Reconstr Surg 1975;
55:53344.
[88] Hidalgo D. Fibular free ap: a new method of mandible reconstruction. Plast Reconstr Surg 1989;84:
719.
[89] Lorenz RR, Esclamado R. Preoperative magnetic
resonance angiography in bular free ap

[90]

[91]

[92]

[93]

[94]

[95]

[96]

reconstruction of head and neck defects. Head

Neck 2001;23:84450.
Karakas YL, Antony A, Rubin G, et al. Preoperative CT angiography for free bular transfer. Microsurgery 2004;24:1257.
Lutz BS, Wei FC, Ng SH, et al. Routing donor leg
angiography before vascularized free bula transplantation is not necessary: a prospective study in
120 cases. Plast Reconstr Surg 1999;103:1217.
Kessler P, Wiltfang J, Schultze-Mosqau S, et al.
The role of angiography in the lower extremity using free vascularized bular transplants for mandibular reconstruction. J Craniomaxillofac Surg
2001;29:3326.
Fleming A, Brough M, Evans N, et al. Mandibular
reconstruction using vascularized bula. Br J Plast
Surg 1990;43:4039.
Taylor GI, Townsend P, Corlett R. Superiority of
the deep circumex iliac vessels as the supply for
free groin aps: experimental work. Plast Reconstr
Surg 1979;64:595604.
Ramasastry SS, Granick MS, Futrell J. Clinical
anatomy of the internal oblique muscle. J Reconstr
Microsurg 1986;2:11722.
Urken ML. Free aps: composite free aps. Iliac
crest osteocutaneous and osteomusculocutaneous.
In: Atlas of regional and free aps for head and
neck reconstruction. New York: Raven Press;
1995. p. 26190.

Oral Maxillofacial Surg Clin N Am 18 (2006) 605614

Chemoradiation Therapy: The Evolving Role

in Head and Neck Cancer and Its Application
to Oral Cavity Tumors
Barbara A. Murphy, MDa,*, Anthony Cmelak, MDb
a

Division of Hematology and Oncology, Vanderbilt-Ingram Cancer Center, 2220 Pierce Avenue,
777 Preston Research Building, Nashville, TN 37232-6307, USA
b
Department of Radiation Oncology, Vanderbilt Center for Radiation Oncology, 22nd Avenue at Pierce,
B-1003 The Vanderbilt Clinic, Nashville, TN 37232-5671, USA

Overview of combined modality therapy

Induction chemotherapy
Early clinical trials demonstrated that treatment-na ve patients with locally advanced head
and neck cancer had a high response rate to
systemic chemotherapy. In 1982, investigators
at Wayne State University were the rst to report
on the results of a two-drug combination using
cisplatin and 5-uorouracil (FU) [2]. Of 26 evaluable patients, 19% had a complete response and
a 70% partial response rate (overall response
rate of 89%) after three cycles of induction therapy. Although similar results have been demonstrated by other investigators using other
combination regimens, cisplatin and 5-FU became
the most commonly used induction regimen for
the next two decades.
Because of the high response rates, there was
initial enthusiasm about the potential benet of
induction chemotherapy before surgical resection.
Unfortunately, the high response rates failed to
result in a statistically signicant survival. Similarly, adjuvant chemotherapy after surgical
resection has failed to demonstrate a survival
advantage. Although many adjuvant studies are
methodologically awed, a recent, well-conducted
Radiation Therapy Oncology Group (RTOG)

* Corresponding author.
E-mail address: Barbara.murphy@vanderbilt.edu
(B.A. Murphy).

trial conrmed the lack of a survival advantage

for adjuvant therapy. In this study, patients underwent surgical resection followed by standard
radiation or three cycles of cisplatin and 5-FU followed by standard radiation [3]. Results demonstrated no improvement in outcome with the
addition of systemic chemotherapy. It must be
noted that in a subset of patients with bulky
disease, patients who received systemic chemotherapy had improved outcome, which must be
seen as a hypothesis-generating observation that
warrants further investigation.
Induction chemotherapy also has been investigated as a part of combined modality therapy
before radiation therapy. These investigations
may be divided into three distinct settings: resectable patients who desire organ preservation,
patients who have unresectable squamous carcinomas, and patiemts who have locally advanced
nasopharynx cancers. In the resectable patient
population, a radiation-based organ preservation
approach was used most commonly in patients
with laryngeal, hypopharyngeal, and base of
tongue tumors. In this cohort of patients, surgical
resection could lead to signicant function loss.
Early phase II studies indicated that induction
chemotherapy followed by radiation has acceptable toxicity, comparable survival outcome to
historical surgical controls, and reasonable rates
of organ preservation.
Two sentinel phase III studies have compared
induction chemotherapy with radiation to primary surgery with postoperative radiation. The

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.001

oralmaxsurgery.theclinics.com

606

MURPHY & CMELAK

Veterans Aairs Laryngeal Cancer Study Group

randomized 332 patients with stage III-IV laryngeal cancer to total laryngectomy with postoperative radiation or induction chemotherapy with
three cycles of cisplatin and 5-FU followed by
denitive irradiation (6676 Gy). Local recurrences were increased in the induction chemotherapy/radiation arm (P 0.0005), although distant
metastases were fewer (P 0.016). The 2-year survival rate was 68% for both treatment arms. The
larynx preservation rate was 64% with induction
chemotherapy and radiation [4]. An analogous
result was shown by the European Organization
for Research and Treatment of Cancer in patients
with locally advanced hypopharyngeal cancer [5].
Unlike the Veterans Administration trial, however, a complete response was required after two
cycles to go on to the third cycle of chemotherapy
and denitive radiation. The median survival
obtained with induction chemotherapy and radiation was 44 months versus 25 months for immediate surgery (P ns). At 3 years, 42% of patients
who received induction chemotherapy and radiation retained a functional larynx. Treatment failures at local, regional, and second primary sites
occurred at the same frequency (12%, 19%, and
16%, respectively, for surgery and 17%, 23%,
and 13%, respectively, for induction chemotherapy radiation). These trials have been criticized
because they lack a third treatment arm with
radiotherapy alone.
Fewer data are available for the role of induction therapy in the unresectable patient population. Paccagnella conducted a randomized trial
in which patients were separated into two cohorts:
resectable and unresectable [6]. Within each cohort, patients were randomized to no induction
or four cycles of cisplatin and 5-FU. Induction
therapy did not improve survival in the surgical
cohort; however, survival was signicantly increased in patients who received induction chemotherapy followed by radiation as opposed to
induction therapy alone. These results have been
updated and remain statistically signicant after
10 years of follow-up. The 5- and 10-year overall
survival rates were 21% and 16%, respectively,
for chemoradiation and 8% and 6%, respectively,
for radiation alone (P 0.04) [7].
As new drugs are being developed they are
being incorporated into induction regimens in an
attempt to improve the ecacy with hopes of
improving survival. Two randomized trials have
investigated the use of aggressive three-drug
regimens as induction therapy before denitive

radiation. Hitt reported the results of a phase III

trial of cisplatin and 5-FU compared with cisplatin, 5-FU, and paclitaxel [7]. Patients who
received the three-drug regimen had an increase
in progression-free (21.7 versus 17.7 months;
P 0.024) and overall survival (median survival
not reached versus 37.7 months; P 0.038) [7].
Similarly, Vermorken and van Herpen [8] reported the results of a randomized phase III trial
of cisplatin and 5-FU versus cisplatin, 5-FU, and
docetaxel followed by radiation therapy. The
three-drug regimen demonstrated an improved response (67.8% versus 53.6%; P 0.007), progression-free survival (HR 0.72, 95% condence
interval 0.560.91; P 0.006), and overall survival (HR 0.73, 95% condence interval 0.57
0.94; P 0.016). Both studies provide strong
support for the further investigation of novel induction regimens in the treatment of locally advanced disease.
Concurrent chemoradiation
An alternative method for combining chemotherapy and radiation therapy is to give them
concurrently. There are several postulated mechanisms for radiosensitization: (1) alteration in
repair of sublethal cell damage, (2) alteration of
cell cycle kinetics, favoring G2/M arrest, and (3)
elimination of clonogens responsible for accelerated repopulation. Preclinical data indicate that
several commonly used chemotherapy agents can
enhance radiation ecacy, including cisplatin,
5-FU, mitomycin, hydroxyurea (Hydrea), bleomycin, actinomycin D, and doxorubicin (Adriamycin). Numerous phase I/II data demonstrate
that these agents can be administered concomitantly with radiation therapy; however, it is at
the expense of increased toxicity. Based on promising phase II data, investigators evaluated chemoradiation in comparison to radiation alone in
patients with locally advanced squamous carcinoma of the head and neck. The French Head
and Neck Oncology and Radiotherapy Group
conducted a randomized phase III trial using radiation alone compared with chemoradiation with
carboplatin and 5-FU in 226 patients with advanced oropharyngeal cancers [9]. Results showed
an improvement in 5-year survival (22% versus
16%; log rank P 0.05), disease-specic survival
(27% versus 15%; P 0.01), and local-regional
control (48% versus 25%; P 0.002) favoring
the combined therapy arm. The results of the intergroup trial comparing radiation alone versus

CHEMORADIATION THERAPY

concurrent radiation plus cisplatin versus a split

course concurrent chemoradiation regimen in
295 unresectable patients were similar [10]. The
3-year overall survival for patients enrolled on
the concurrent radiation plus cisplatin was superior to radiation alone (37% versus 23%; P
0.014). Several additional phase III trials also support the use of combined chemotherapy with
radiation over radiation alone in terms of progression-free, disease-free, relapse-free, and overall
survival [3,1124].
The question arose as to whether induction
chemotherapy followed by radiation or concurrent chemoradiation (CCR) provided superior
outcomes. For this reason, an intergroup phase
III trial was conducted to determine if induction
chemotherapy is an essential component of organ
preservation. Treatment arms included radiation
alone (70 Gy), induction cisplatin and 5-FU followed by radiation, and concomitant cisplatin
with radiation [25]. Function preservation rates
were 55% versus 65% versus 85%, respectively.
Concurrent cisplatin with radiation, although
providing no overall survival advantage, showed
statistically signicant improvement in organ
preservation rate and is considered a standard approach [25].
Three meta-analysesdtwo literature based and
one patient baseddthat evaluated the role of
chemotherapy in the primary treatment of squamous cell carcinoma of the head and neck have
been reported [2628]. A reported patient-based
meta-analysis with median follow-up of 6 years
conrmed these results in 10,717 patients enrolled
in 63 trials between 1965 and 1993 [28]. The addition of chemotherapy provided an overall 11%
risk reduction, with a 4% absolute survival benet
at 5 years (P 0.001). Adjuvant and neoadjuvant
chemotherapy provided a risk reduction of 2%
and 5%, respectively (absolute 5-year benet of
1% and 2%, respectively, P ns). In a subset
of patients who received induction therapy with
cisplatin and 5-FU, a signicant survival advantage favored induction therapy (HR 0.88, 95%
condence interval 0.790.97). Concomitant chemotherapy provided a risk reduction of 19%,
with an absolute survival benet of 8% at 5 years
(P 0.0001). It seems that the addition of concomitant chemotherapy confers a modest overall
survival advantage to radiation therapy alone
which comes at the cost of signicantly increased
toxicity. The benet of induction therapy was isolated to patients who received induction therapy
with cisplatin and 5-FU.

607

More recently, the use of CCR has been

expanded to use in postoperative patients at
high risk for recurrence [29]. Several small trials
reported improved survival in patients with locally advanced disease who underwent postoperative CCR when compared with patients who
received radiation alone [24,30,31]. Based on these
data, two large cooperative group trials, one in
Europe [32] and one in the United States [33],
were conducted to determine whether CCR improved survival in high-risk postoperative patients.
With the exception of eligibility requirements, the
studies had a similar design: postoperative patients
at high risk for recurrence were randomized to
postoperative radiation versus postoperative radiation with concurrent cisplatin 100 mg/m2 every 3
weeks during radiation therapy. In the European
study, local-regional control, progression-free,
and overall survival were higher for the combined
therapy arm (82%, 47%, and 53%, respectively)
when compared with the radiation-only arm
(69%, 36%, and 40%, respectively). All values
were statistically signicant in favor of postoperative CCR. In the US study, which was conducted
through the RTOG, CCR resulted in an improvement in local control (82% versus 72%) and disease-specic survival (HR of death 0.78,
condence interval 0.610.99; P 0.04). No survival advantage was found (HR of death 0.84,
condence interval 0.651.09; P 0.19). The authors note that at the last analysis the survival
curves were separating and may eventually
demonstrate a survival advantage for combined
therapy. Based on the results of these studies, patients at high risk for recurrence who have a good
performance status are considered for postoperative chemoradiation.
Future directions
For patients with locally advanced disease who
are being treated with a radiation-based regimen,
the current standard is to administer chemotherapy concurrently. The role of induction therapy is
unclear at this time. It is clear, that CCR has
achieved a toxicity ceiling. Further escalation or
intensication of regimens using current agents is
not feasible due to lack of patient tolerance. In
addition, as local disease becomes controlled with
CCR, patients are living long enough to develop
metastatic disease. This begs the question: in what
direction should we take investigative trials?
Obviously, one can test new radiation sensitizers
that have less toxicity or agents that ameliorate

608

MURPHY & CMELAK

CCR induced side eects. Bonner et al reported the

results of cetuximab, an anti-epidermal growth
factor receptor antibody, as a radiation sensitizer
in patients with squamous cancer of the head and
neck [34]. Patients with stage 3 or 4 squamous cancer of the head and neck were randomized to standard radiation or radiation with cetuximab.
Results demonstrated an increase in local-regional
control (P 0.02) and overall survival (P 0.02)
for patients treated with combined therapy versus
patients who were treated with radiation alone.
However, we still fail to cure a substantial percentage of patients with locally advanced disease.
Thus, an alternative approach, the use of induction
chemotherapy followed by CCR has garnered
a great of attention over the past several years.
The induction chemotherapy would allow systemically eective doses of chemotherapy to decrease
metastatic disease, whereas the concomitant chemoradiation would optimize local control.
Several phase II trials have been conducted to
investigate various regimens that combine induction and concurrent chemoradiation. These regimens vary in their intensity from highly aggressive
regimens that may only be used in good performance status patients to regimens designed to be
tolerable for patients with poor performance
status, comorbid disease, and suboptimal nutritional status. Based on the intriguing results of
phase II studies using induction therapy followed
by concurrent chemoradiation, numerous phase
III trials are underway to determine whether this
approach will enhance outcome and, if so, in
which patient populations. The only study reported to date was presented by Hitt and colleagues [7]. They reported on the rst 170 patients
accrued to a three-arm randomized trial that compared induction therapy with cisplatin and 5-FU
versus induction therapy with cisplatin, 5-FU
and taxotere, versus no induction therapy. All patients received concurrent cisplatin and radiation.
Preliminary results demonstrated a complete response rate of 36% for both induction therapy
regimens. The overall complete response rate after
completion of treatment was 88% for the twodrug induction regimen, 80% for the three-drug
induction regimen, and 47% for chemoradiation
alone. Survival data were not available, and accrual continues.
Chemoradiation in oral cavity cancers
Oral cavity tumors comprise approximately
25% of head and neck primaries [1]. They result

in between 2.3 and 3.6 deaths per 100,000 in the

United States depending on the geographic region
[35]. In a recent evaluation of 556 patients diagnosed with oral cavity tumors in Norway, the
subsite distribution was 39.9% tongue, 23.9%
gingival, 23.4% oor of mouth, and 12.8% other.
The stage distribution was 23.2% stage I, 21.2%
stage II, 11.2% stage III, 39.7% stage IVA, 4%
stage IVb, and 0.7% stage IVC. Survival was similar among subsites, with a 5-year survival rate of
40.6% [36]. Tumors of the oral cavity are heterogeneous with regard to risk factors, genetic abnormalities, and treatment considerations. For
example, oropharyngeal cancers have been associated with smoking and alcohol use [37]. The odds
ratio varies depending on the site within the oral
cavity [38]. Unfortunately, most epidemiologic
studies combine all subsites of oral cavity tumors
with oropharynx tumors, which makes it dicult
to dissect out incidence rates, stage at diagnosis,
response to treatment, and survival in patients
with various subsites.
Although radiation therapy is eective for
treating small tumors at various sites within the
oral cavity, most patients are treated with surgery
alone for small tumors and surgical resection
followed by radiation therapy for locally advanced disease [36]. The combination of chemotherapy and radiation therapy traditionally has
been reserved for patients with unresectable disease or patients who refuse surgical resection.
The data describing the use of chemoradiation
in oral cavity cancers are scant. Few studies or
reports have been dedicated solely to patients
with oral cavity primaries. The bulk of the data
is culled from trials that allow patients from various primary sites to be enrolled. The available
data do provide important guidance as to the
use of chemoradiation in this setting.
Induction therapy before surgery
Induction therapy before surgery results in
downsizing of tumors. There is no proven benet
with regard to local regional control or survival.
Licitra and colleagues [39] reported the results of
a randomized trial that evaluated the eect of
induction chemotherapy on local-regional recurrence and distant relapse. One hundred ninetyve patients with resectable oral cavity tumors
were randomized to surgery or three cycles of cisplatin and 5-FU followed by surgery. All high-risk
patients subsequently underwent postoperative
radiation therapy; 33% of patients had a complete

CHEMORADIATION THERAPY

response and 49% of patients had a partial response, for an overall response rate of 82%.
Thirty-three percent of patients in the induction
arm versus 46% of patients in the surgery alone
arm required postoperative radiation therapy. A
mandibular resection was required in 31% versus
52% of patients in the induction and surgery-only
arm, respectively. Although there was an improvement in the 5-year event-free survival rate
for patients who received induction therapy
(57% versus 46%), there was no statistically
signicant dierence in local-regional control or
distant relapse. The 5-year survival rate was
55% in both arms. There did seem to be a decrease
in second primary tumors. This study was
designed to look for an improvement in localregional control and distant relapse of R20%,
which may have been an overly ambitious goal
and the study may have been underpowered.
The authors concluded that induction chemotherapy did not improve long-term outcome; however,
it may allow less aggressive surgery or spare radiation therapy to the oral cavity in selected
patients. They advocated for further evaluation
of this approach.
The ability of induction therapy to downstage
oral cavity carcinomas was conrmed by Grau
and colleagues [40] in a prospective trial of induction chemotherapy for patients with resectable or
unresectable stage III or IV oral cavity cancer.
The primary outcome measures were response
to induction, local control, and survival. Of
1089 patients with oral cavity cancer who were
screened, 204 met entry criteria; 66% of patients
responded to induction chemotherapy (16% complete response and 50% partial response). Of
46 patients who were considered inoperable, 34
were able to undergo complete resection after induction therapy. Predictors of outcome included
initial stage, subsite, response to induction chemotherapy, and adjuvant radiation therapy. Diseasefree survival at 5 years was 26% for patients
undergoing resection and 22% for patients who
received chemoradiation.
Similar response rates were reported by Ruggeri and colleagues [41], who treated 33 patients
with stage III or IV oral cavity tumors using one
of three dierent induction regimens. The complete response rate was 48% (n 16) and the
pathologic complete response rate was 33%
(n 9). The overall 5- and 10-year survival rates
were 54.5% and 39.5%, respectively. Patients
who achieved a complete response to induction
therapy had a clinically signicant increase in

609

survival when compared with patients who failed

to achieve a complete response (P 0.05).
Concomitant chemoradiation
CCR can be used in patients with resectable
and unresectable cancer. In the resectable patient
population, substantial data indicate that early
stage squamous carcinomas of the oral cavity may
be treated eectively with radiation therapy alone.
Most patients with oral cavity cancers, regardless
of size, are treated with surgical resection to avoid
the acute and late eects of radiation to the oral
cavity, however (see the section on supportive care
issues). Investigators have evaluated the role of
CCR as a function preservation approach for oral
cavity primaries.
Fuchihata and colleagues [42] reported the results of CCR using bleomycin or peplomycin in
patients with resectable squamous cell carcinomas
of the lower gingiva. As expected, patients with
early stage disease (T1 or T2) had a higher complete response rate (67%, n 100) compared
with patients with advanced stage tumors (35%,
n 62). Patients who failed to achieve a complete
response went on to surgical salvage. Diseasespecic 5-year survival rate was 75% for stage I,
87% for stage II, 71% for stage III, and 51%
for stage IV. The complete response rate is substantially lower than would be expected at other
head and neck primary sites.
Investigators from the University of Michigan
undertook a clinical trial in patients with
advanced resectable oral cavity tumors using
induction chemotherapy as a marker of responsiveness. Patients were scheduled to receive one
cycle of cisplatin and 5-FU. Responders were to
proceed to denitive CCR with salvage surgery.
Patients who failed to respond to one cycle of
therapy were to proceed directly to surgical
salvage. Eighteen patients were entered on study;
2 patients died after cycle one of therapy before
re-evaluation. Nine of 16 patients who were
evaluable had a 50% response to therapy and
went on to CCR. Of the 9 patients who underwent
CCR, 6 had a complete response. The 3-year
survival rate was 47%, and the disease-specic
survival rate was 68.2%. The investigators concluded that an organ preservation approach was
not recommended for patients with locally advanced oral cavity tumors [43].
As with other sites of disease, CCR can result
in durable complete response in patients with
unresectable oral cavity tumors and should be

610

MURPHY & CMELAK

considered standard of care for patients with an

adequate performance status. The bulk of available data demonstrates a lower response rate and
survival for oral cavity tumors when compared
with other sites, however. Harrison and colleagues
[44] reported the results of a phase II trial of 82
patients with unresectable disease. Three-year survival rate for patients with oral cavity tumors was
worse than other sites (0 versus 47%; P 0.03). In
an RTOG trial that reported evaluating concurrent chemoradiation with cisplatin in 96 unresectable patients, the complete response rate after
therapy was 56% for patients with oral cavity
cancers versus 74% for oropharynx, 82% for
nasopharynx, 75% of larynx, and 37% for hypopharyngeal lesions [45]. Similarly, in a study by
Taylor and colleagues [46], six patients with oral
cavity tumors treated with CCR had a worse
prognosis (67% failure rate and 0% 8-year survival) compared with other sites.
Two randomized trials have compared postoperative radiation versus postoperative chemoradiation in high-risk patients. Both clinical trials
included patients with oral cavity primaries. The
RTOG trial stratied based on primary site; thus,
there are an equal number of patients with oral
cavity primaries in both arms, and oral cavity tumors compose a high percentage of the study population (30% arm one and 20% arm two). Neither
study reported the comparative results specically
for oral cavity tumors. Postoperative chemoradiation should be recommended for select patients at
high risk for recurrence.

Supportive care issues in patients with oral cavity

primaries
A host of supportive care issues must be taken
into account when patients are treated with CCR
[47]. Unfortunately, space does not permit an
exhaustive review of symptom control issues in
head and neck cancer. We do, however, mention
toxicities most pertinent to patients with oral cavity primaries.
Once patients begin therapy with concurrent
chemoradiation, several side eects are encountered. Perhaps the most critical side eect of
chemoradiation is mucositis. Radiation induces
tissue damage in the mucous membranes and soft
tissues within the radiated port. The tissue damage results in the activation of several biologic
pathways and is accompanied by an inltration of
inammatory cells [48]. This damage eventually

results in mucosal ulcerations that are painful

and usually require high doses of opioids. This tissue damage is also associated with edema, which
decreases function of vital structures, such as the
tongue and muscles of deglutition. After radiation
is completed, tissue damage begins to repair. Unfortunately, scar tissue with lymphedema may develop and result in long-term sequelae. The late
eects may include altered speech, swallowing,
and nutritional intake [49,50].
A great deal of eort has been expended trying
to identify eective preventive and treatment
measures. A plethora of traditional treatment
for oropharyngeal mucositis has been reported,
but rarely have they been substantiated or corroborated by rigorous well-controlled studies.
Various mouthwashes have been recommended:
hydrogen peroxide [51], chlorhexidine (alone or
with nystatin) [52], dilute sodium bicarbonate
and salt water [53], and Benzydamine, a nonsteroidal anti-inammatory agent with antimicrobial
and anesthetic properties [54]. Antiulcer medications, such as sucralfate [55] and prostoglandins,
have been studied, with mixed results [56,57],
and antibiotic lozenges that contain polymyxin,
tobramycin, and amphotericin B (PTA-lozenges)
have been developed after gram-negative bacteria
were shown to play a role in the pathogenesis and
exacerbation of radiation mucositis [58]. GM-CSF
[59], amifostine [60], and keratinocyte growth factor are newer agents currently undergoing testing
in randomized trials [61].
It has long been understood that patients who
have head and neck cancer are at risk for
malnutrition as a consequence of their tumor or
its treatment [62]. Pretreatment patients must be
assessed for nutritional status. Although most patients who present with newly diagnosed head and
neck cancers are able to maintain adequate oral
intake, patients who have malnutrition at presentation are at increased risk for morbidity from
treatment [63]. Weight loss also is associated
with decrease in quality of life [64]. Unfortunately,
radiation-induced mucositis and edema result in
poor oral intake and associated weight loss
[65,66]. The problem is exacerbated by the use
of concurrent chemoradiation, with weight loss
averaging 10% over the course of treatment.
The placement of a feeding tube does not guarantee adequate intake of calories, nutrients, or water
[67]. Metabolic alterations as a result of the tumor
and its treatment seem to play a role in the weight
loss and deconditioning experienced by patients
[68]. Nutritional support during therapy is vital.

CHEMORADIATION THERAPY

Preliminary data indicate that patients nutritional

inadequacies persist long-term [49]; thus, nutritional support should continue through the recovery phase as patients return toward a normal diet.
New tools are being developed to allow treating
physicians to screen quickly patients who are at
risk for nutritional inadequacies [69].
Another important eect of radiation to the
oral cavity is xerostomia. Even modest doses of
radiation to the salivary gland (25 Gy) begin to
result in marked decrease in salivary ow. Saliva
has many critical biologic functions, including
maintaining a healthy environment for teeth [70,
71]. Patients who have severe xerostomia may experience a wide array of symptoms or problems,
including rapid dental decay, diculty swallowing, altered taste, decreased quality of life, and
discomfort. Recent studies demonstrate that the
prophylactic use of amifostine and the use of intensity modulated radiation therapy can decrease
xerostomia [72,73].
Attention to dental status is imperative, and
a baseline dental evaluation is imperative, including full mouth radiographs, dental charting,
scaling and polishing, oral hygiene and home care
instructions. Maxillary and mandibular alginate
impressions and neutral pH preparations of uoride therapy (either sodium or stannous) are
mandatory. Preirradiation extractions should be
performed on any teeth with questionable ability
to tolerate R50 Gy radiation, and oral mouth
guards or changing amalgam lling/crowns to
composite material helps reduce radiation scatter
to adjacent mucosa and potentially lowers the risk
of osteoradionecrosis.

Summary
The use of the combination of chemotherapy
and radiation therapy has altered treatment dramatically for head and neck cancers. Oral cavity
tumors are unique with regards to response and
treatment-related sequelae, however, which makes
extrapolation of data from other head and neck
sites to cancers of the oral cavity problematic. It is
clear that patients who have unresectable disease
and high-risk postoperative patients should be
considered for concurrent chemoradiation. It
must be understood that improved outcome is at
the expense of increased toxicity. The treating
physician must see patients on a frequent basis
during and immediately after therapy. They also
must be familiar with support care measures and

611

have access to a team of consultants who can help

manage the acute and late eects of therapy.
References
[1] Jemal A, Ward E, Samuels A, et al. Cancer Statistics
2005. CA Cancer J Clin 2005;55(1):1030.
[2] Kish J, Jacos J. Clinical trial of cisplatin and 5-FU
infusion as initial treatment for advanced squamous
cell carcinoma of the head and neck. Cancer Treat
Rep 1982;66:4716.
[3] Al-Sarraf M, LeBlanc M, Giri PG, et al. Chemoradiotherapy versus radiotherapy in patients with
advanced nasopharyngeal cancer: phase III randomized Intergroup study 0099. J Clin Oncol 1998;16(4):
13107.
[4] Department of Veterans Aairs Laryngeal Cancer
Study Group. Induction chemotherapy plus radiation compared with surgery plus radiation in patients with advanced laryngeal cancer. N Engl
J Med 1991;324:168590.
[5] Lefebvre J, Chavalier D, Luboinski B. Larynx preservation in pyriform sinus cancer: preliminary results of a European Organization for Research and
Treatment of Cancer phase II trial. J Natl Cancer
Inst 1996;88:8909.
[6] Paccagnella A, Orlando A, Marchiori C, et al.
Phase III trial of initial chemotherapy in stage III
or IV head and neck cancers: a study by the Gruppo
di Studio sui Testa e del Collo. J Natl Cancer Inst
1994;86:26572.
[7] Hitt R, Lopez-Pousa A, Martinez-Trufero J, et al.
Phase III study comparing cisplatin plus uorouracil
to paclitaxel, cisplatin, and uorouracil induction
chemotherapy followed by chemoradiotherapy in locally advanced head and neck cancer. J Clin Oncol
2005;23(34):863645.
[8] Vermorken JB, van Herpen J. Standard cisplatin/
infusional 5-urouracil (PF) vs doctaxel (T) plus
PF (TPF) as neoadjuvant chemotherapy for nonresectable locally advanced squamous cell carcinoma
of the head and neck (LA-SCCHN): a phase III trial
of the EORTC Head and Neck Cancer Group
(EORTC #24971). Proc Am Soc Clin Oncol 2003;
23(Suppl):22.
[9] Denis F, Bardet E, Alfonsi M, et al. Final results
of the 9401 French Head and Neck Oncology and
Radiotherapy Group randomized trial comparing
radiotherapy alone with concomitant radiochemotherapy in advanced-stage oropharynx carcinoma.
J Clin Oncol 2004;22:6976.
[10] Adelstein DJ, Wagner H, Kish JA, et al. An intergroup phase III comparison of standard radiation
therapy and two schedules of concurrent chemoradiotherapy in patients with unresectable squamous
cell head and neck cancer. J Clin Oncol 2003;21:
928.
[11] Wendt T, Granbenbauer G, Rodel C. Simultaneous
radiochemotherapy versus radiotherapy alone in

612

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

MURPHY & CMELAK

advanced head and neck cancer: a randomized multicenter study. J Clin Oncol 1998;16(4):131824.
Lo T, Wiley A, Anseld F. Combined radiation therapy and 5-uorouracil for advanced squamous cell
carcinoma of the oral cavity and oropharynx: a randomized study. AJR Am J Roentgenol 1976;126:
22935.
Shanta V, Krishnamurthi S. Combined bleomycin
and radiotherapy in oral cancer. Clin Radiol 1980;
31:61720.
SECOG. A randomized trial of combined multidrug
chemotherapy and radiotherapy in advanced squamous cell carcinoma of the head and neck. Eur J
Surg Oncol 1986;12:28995.
Fu K, Philips T, Silverberg I. Combined RT and CT
with bleomycin in advanced inoperable head and
neck cancer: update of the Northern California Oncology Group randomized trial. J Clin Oncol 1987;
5(9):14108.
Weissberg J, Son Y, Papac R. Randomized clinical
trial of mitomycin C as an adjunct to radiotherapy
in head and neck cancer. Int J Radiat Oncol Biol
Phys 1989;17:39.
Adelstein D, Sharan V, Earle A. Simultaneous versus sequential combined technique for head and
neck cancer. Cancer 1990;65:168591.
Sanichiz R, Milla A, Torner J. Single fraction per
day versus two fractions per day versus radiochemotherapy in the treatment of head and neck cancer. Int
J Radiat Oncol Biol Phys 1990;19:134750.
Merlano M, Vitale V, Rosso R. Treatment of advanced squamous-cell carcinoma of the head and
neck. N Engl J Med 1992;327:111521.
Adelstein D, Saxton J, Lavertu T. A phase III randomized trial comparing concurrent radiotherapy
alone in advanced head and neck cancer. Head
Neck 1997;19(7):56775.
Brizel DM, Fisher SE, Scher RL, et al. Hyperfractionated irradiation with or without concurrent chemotherapy for locally advanced head and neck cancer.
N Engl J Med 1998;338:1798804.
Zakotnik B, Smid L, Budhina M. Concomitant radiotherapy with mitomycin C and bleomycin compared with radiotherapy alone in inoperable head
and neck cancer: nal report. Int J Radiat Oncol
Biol Phys 1998;41:11217.
Dobrowsky W, Naude J, Widder J, et al. Continuous
hyperfractionated accelerated radiotherapy with/
without mitomycin C in head and neck cancer. Int
J Radiat Oncol Biol Phys 1998;42:8036.
Bachaud JM, Altzieu C, David JM, et al. Combined postoperative radiotherapy and weekly cisplatin infusion for locally advanced head and
neck carcinoma: nal report of a randomized
trial. Int J Radiat Oncol Biol Phys 1996;36(5):
9991004.
Forastiere A, Berkey B, Maor M. Phase III trial to
preserve the larynx: induction chemotherapy and
radiotherapy versus concomitant chemotherapy

[26]

[27]
[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

versus radiotherapy alone: intergroup trial R9111

[abstract 4]. Proc Am Soc Clin Oncol 2001;20:2.
El-Sayed S, Nelson N. Adjuvant and adjunctive chemotherapy in the management of squamous cell carcinoma of the head and neck region: a meta-analysis
of prospective and randomized trials. Clin Oncol
1996;14:83847.
Munro A. Meta-analysis of chemotherapy in head
and neck cancer. Br J Cancer 1995;71:8391.
Pignon J, Bourhis J, Domenge C, et al. Chemotherapy added to locoregional treatment for head
and neck squamous-cell carcinoma: the meta-analyses of updated individual data. Lancet 2000;355:
94955.
Cooper JS, Forastiere A, Jacobs J, et al. Precisely dening high-risk operable head and neck tumors
based on RTOG #8503 and #8824: targets for
post-operative radiochemotherapy? Head Neck
1998;20:58894.
Lundahl RE, Bonner JA, Suman VJ, et al. Combined neck dissection and postoperative radiation
therapy in the management of the high-risk neck:
a matched-pair analysis. Int J Radiat Oncol Biol
Phys 1998;40(3):52934.
Hay B, Papac R, Saski CT, et al. Chemotherapy as
an adjuvant to radiation in the treatment of squamous cell carcinoma of the head and neck: results
of the Yale mitomycin randomized trials. J Clin
Oncol 1997;15(1):26876.
Bernier J, Ozaxhin M, Matuszewska K, et al. Postoperative irradiation with or without concomitant
chemotherapy for locally advanced head and neck
cancer. N Engl J Med 2004;350(19):194552.
Cooper JS, Forastiere AA, Jacobs J, et al. Postoperative concurrent radiotherapy and chemotherapy
for high-risk squamous-cell carcinoma of the head
and neck. N Engl J Med 2004;350(19):193744.
Bonner JA, Harari PM, Giralt J, et al. Radiotherapy
plus cetuximab for squamous-cell carcinoma of the
head and neck. N Engl J Med 2006;354(6):56778.
Goldberg HI, Wyatt SW, Crossett LS. Trends and
dierentials in mortality from cancers of the oral
cavity and pharynx in the United States, 1973
1987. Cancer 1994;74:56572.
Groome PA, Boysen M, Hall SF, et al. A comparison of published head and neck stage groupings in
carcinomas of the oral cavity. Head Neck 2001;23:
61324.
Blott W, McLaughlin J, Winn D. Smoking and
drinking in relation to oral and pharyngeal cancer.
Cancer Res 1988;48:32827.
Barasch A, Krutcho DJ, Eisenberg E. Smoking,
gender, and age as risk factors for site-specic
intraoral squamous cell carcinoma. Cancer 1994;
73:50913.
Licitra L, Guzzo M, Mariani L, et al. Primary chemotherapy in resectable oral cavity squamous cell
cancer: a randomized controlled trial. J Clin Oncol
2003;21(2):32733.

CHEMORADIATION THERAPY

[40] Grau JJ, Blanch JL, Verger E, et al. Multidisciplinary approach in advanced cancer of the oral cavity:
outcome with neoadjuvant chemotherapy according
to intention-to-treat local therapy. Oncology 2002;
63:33845.
[41] Ruggeri EM, Pollera CF, De Marco S, et al. Longterm survival in locally advanced oral cavity cancer:
an analysis of patients treated with neoadjuvant
cisplatin-based chemotherapy followed by surgery.
Head Neck 2005;27:4528.
[42] Fuchihata H, Murakami S, Kubo K, et al. Results of
combined external irradiation and chemotherapy of
bleomycin or peplomycin for squamous cell carcinomas of the lower gingiva. Int J Radiat Oncol Biol
Phys 1994;29:7059.
[43] Urba S, Carey T, Chepeha D, et al. One cycle of
induction chemotherapy to select for organ preservation for patients with advanced squamous carcinoma of the oral cavity. Proc Am Soc Clin Oncol
2005;23:[abstract 5555].
[44] Harrison LB, Pster DG, Zelefsky M, et al. A prospective phase II trial of concomitant chemotherapy
and radiotherapy with delayed accelerated fractionation in unresectable tumors of the head and neck.
Head Neck 1998;20:497503.
[45] Al-Sarraf M, Marical VA, Mowry P, et al. Concurrent radiotherapy and chemotherapy with cisplatin in inoperable squamous cell carcinoma of
the head and neck: an RTOG study. Cancer
1987;59:25965.
[46] Taylor SG, Griem K, Recine DC, et al. Concomitant
cisplatin/5-FU infusion and radiotherapy in advanced head and neck cancer: 8-year analysis of
results. Head Neck 1997;10:68491.
[47] Murphy BA, Bayles S, Dowling E, et al. Symptom
management: head and neck cancer. In: Oxford textbook of palliative medicine. Oxford: Oxford Press;
2003. p. 65873.
[48] Sonis S. The pathobiology of mucositis. Nature Reviews 2004;4:27784.
[49] Murphy BA, Dowling E, Cheatham R, et al. Dietary
intake and adaptations in head and neck cancer
patients treated with chemoradiation. Proc Am Soc
Clin Oncol 2002;21:[abstract 932].
[50] Murphy BA, Dowling E, Cmelak A, et al. Baseline
swallowing function for patients treated on 2399:
a phase II trial of function preservation with induction paclitaxel/carboplatin followed by radiation
plus weekly paclitaxel. Proc Am Soc Clin Oncol
2003;2:[abstract 2005].
[51] Feber T. Management of mucositis in oral irradiation. Clin Oncol 1996;8:10611.
[52] Barkvoll P. Eect of nystatin and chlorhexidine
digluconate on Candida albicans. Oral Surg Oral
Med Oral Pathol 1989;67:27981.
[53] Dewalt E, Haines A. The eects of specied stressors
on health of oral mucosa. Nurs Res 1969;18:227.
[54] Epstein J, Stevenson-Moore P, Jackson S. Prevention of oral mucositis in radiation therapy:

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64]

[65]

[66]
[67]

[68]

[69]

613

a controlled study with benzydamine hydrochloride

rinse. Int J Radiat Oncol Biol Phys 1989;16:15715.
Pheier P, Madsen EL, May O. A prospective pilot
study on the eect of sucralfate mouth-swishing in
reducing stomatitis during radiotherapy of the oral
cavity. Acta Oncol (Madr) 1989;27:4713.
Sinzinger H, Porteder H, Matejka M, et al. Prostaglandins in irradiation induced mucositis. Lancet
1989;1:556.
Kuhrer I, Kuzmits R, Linkesch W, et al. Topical
PGE2 enhances healing of chemotherapy associated
mucosal lesions. Lancet 1986;1:623.
Symonds R, McIlroy P, Khorrami J, et al. The reduction of radiation mucositis by selective decontamination antibiotic pastilles: a placebo controlled
double blind trial. Br J Cancer 1996;74:3127.
Nemunaitis J, Rosenfeld C, Ash R. Phase II
randomised double blind placebo controlled trial
of rh GM-CSF following allogeneic bone marrow
transplant. Bone Marrow Transplant 1995;15:
94954.
Sasse AD, Clark LG, Sasse EC, et al. Amifostine reduces side eects and improves complete response
rate during radiotherapy: results of a meta-analysis.
Int J Radiat Oncol Biol Phys 2006;64(3):78491.
Knerer B, Formanek M, Schickinger B. Dierent
KGF expression in squamous cell carcinomas of
the head and neck and in normal mucosa. Acta Otolaryngol 1998;18:43842.
Lopez MJ, et al. Nutritional support and prognosis
in patients with head and neck cancer. J Surg Oncol
1994;55:336.
Brookes GB. Nutritional status: a prognostic indicator in head and neck cancer. Otolaryngol Head Neck
Surg 1985;93:6974.
Hammerlid E, Wirblad B, Sandin C, et al. Malnutrition and food in take in relation to quality of life in
head and neck cancer patients. Head Neck 1998;20:
5408.
Collins M, Wright R, Partridge G. Nutritional consequences of radiotherapy in early laryngeal carcinoma. Ann R Coll Surg Engl 1999;81:37681.
Donaldson S. Nutritional consequences of radiotherapy. Cancer Res 1977;37:240713.
Silver HJ, Arnold DJ, Livingstone AS, et al. Older
adults on home enteral nutrition: enteral regimen,
provider involvement, and health care outcomes.
JPEN J Parenter Enteral Nutr 2004;28:928.
van Bokhorst-de van der Schuer MA, von Blomberg-van der Flier BM, Kuik DJ, et al. Survival of
malnourished head and neck cancer patients can be
predicted by human leukocyte antigen-DR expression and interleukin-6/tumor necrosis factor-alpha
response of the monocyte. JPEN J Parenter Enteral
Nutr 2000;24:32936.
Murphy BA, Wells N, Cmelat A, et al. Reliability
and validity for the Vanderbilt Head and Neck
Symptom Survey (VHNSS): a new tool to assess
symptom burden in patients undergoing

614

MURPHY & CMELAK

chemoradiation. Proc Am Soc Clin Oncol 2004;22:

[abstract 5569].
[70] Eisbruch A, et al. Radiation (RT) concurrent with
low dose gemcitabine (gem) for head and neck cancer: interim Results of a phase I study. Proc Am
Soc Clin Oncol 1998;17:[abstract 1563].
[71] Eisbruch A, Kin HM, March LH, et al. Dose,
volume, and function relationships in parotic
salivary glands following conformal and intensitymodulated irradiation of the head and neck cancer.
Int J Radiat Oncol Biol Phys 1999;45:57787.

[72] Wasserman T, Henke M, Eschwege F, et al. Inuence of intravenous amifostine on xerostomia, tumor control and survival after radiotherapy for
head and neck cancer: 2-year follow-up of a prospective, randomized phase III trial. Int J Radiat Oncol
Biol Phys 2005;63:98590.
[73] Braaksma M. Tools for optimal tissue sparing in
concomitant chemoradiation of advanced head
and neck cancer: subcutaneous amifostine and computed tomographic-based target delineation. Semin
Oncol 2002;6(Suppl 19):6370.

Oral Maxillofacial Surg Clin N Am 18 (2006) 615625

Management of Locoregional Recurrence

in Oral Squamous Cell Carcinoma
Deepak Kademani, DMD, MDa,*, Eric Dierks, DMD, MDb,c
a

Oral and Maxillofacial Surgery, Mayo Clinic College of Medicine, 200 First Street, SW, Rochester, MN 55902, USA
b
Oral and Maxillofacial Surgery, Oregon Health and Science University, Portland, OR, USA
c
Legacy Emanuel Hospital, Portland, OR, USA

Oral squamous cell carcinoma (OSCC) is the

sixth most common malignant tumor worldwide,
aecting more than 25% of 1 million people
annually [1,2]. It accounts for 4% of all cancers
and 2% of all cancer-related deaths worldwide
[3]. In Southeast Asia, particularly in India,
OSCC is the most common malignant tumor, accounting for 30% to 50% of all tumors [4,5]. In
the United States, approximately 30,000 new cases
of OSCC are diagnosed annually, and most are
seen in patients older than the age of 40 years,
with a mean age at diagnosis of 60 years [68]. Although a strong male predilection was seen at the
turn of the century, incidence rates between men
and women are now almost equal. The incidence
in young patients (!40 years old) ranges between
0.4% and 4% [9]. Unfortunately, the rate of second primary (synchronous or metachronous) tumors in the aerodigestive tract continues to be
high, at 3.7% per year [10,11]. This contributes
to the high death and registration index, which
is 0.45 for OSCC compared with 0.48 for carcinoma of the uterine cervix and greater than 0.38
for melanoma [3,12]. Early-stage tumors (T1 and
T2) are associated with a 60% to 80% 5-year survival rate. The status of the cervical lymph nodes
is the single most important prognostic factor in
OSCC, with the development of neck metastases
reducing the 5-year survival rate by 50% [13
15]. Despite advances in surgical and adjuvant
chemoradiotherapy, the diagnosis of OSCC continues to portend a poor prognosis, with an
* Corresponding author.
E-mail address: kademani.deepak@mayo.edu
(D. Kademani).

overall 5-year survival rate remaining essentially

unchanged at 50% for the past several decades
[13,16].
The mainstay of primary treatment of OSCC is
surgical resection with or without adjuvant radiotherapy. The goal of all primary resection aimed
at cure is complete eradication of all local disease
with adequate treatment of the cervical lymph
nodes. Despite adequate local control of OSCC,
recurrence rates of 25% to 48% have been
reported [1719]. There are several considerations
in dealing with patients with local or regional recurrence. Most recurrences from OSCC occur in
the rst 24 to 36 months; therefore, it is important
to maintain a strict follow-up schedule for tumor
surveillance [15]. To reduce the risk of recurrence,
it is important for patients to abstain from tobacco and alcohol consumption so as to minimize
the risk associated with second primary lesions.
Although the oral cavity is amenable to easy examination, the presence of a reconstructive ap
may make the diagnosis of deep recurrent disease
dicult. This is in contrast to supercial or marginal recurrent disease, which is more easily detected. The regular scheduled use of imaging
modalities, such as CT, MRI, and, more recently,
positron electron tomography (PET), is assuming
an increasing role in tumor surveillance. PET
scanning uses a radiolabeled glucose isotope (18uorodeoxyglucose [FDG]) to label areas of hypermetabolism. Although PET is proving to be
a valuable tool in oncologic surveillance, it should
be recognized that in the immediate postsurgical
setting, PET scanning may remain positive at
the primary or postoperative site for several
months. It is therefore recommended that for

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.009

oralmaxsurgery.theclinics.com

616

KADEMANI & DIERKS

purposes of tumor surveillance, PET scanning be

deferred until a minimum of 6 to 12 weeks after
surgery so as to avoid false-positive results. These
imaging modalities are highly useful not only to
identify early recurrence before it is clinically apparent but to delineate the spread of recurrent disease, which often lacks well-dened margins.
Although the use of combinations of chemotherapy, radiotherapy, and surgery has improved
locoregional control rates, the patient with recurrent disease usually falls into the hands of the
surgeon. These patients present a signicant surgical challenge in terms of increased rates of
complications attributable to brotic hypovascular tissue and the lack of donor vessels for microvascular free ap reconstruction. To improve
survival from recurrent OSCC, the complex interrelations between the patient, tumor, previous
treatment rendered, type of recurrence, salvage
options, functional outcome, quality of life, and
overall prognosis need to be considered and
individualized to optimize patient outcomes.
Local recurrence
Most recurrence from oral cancer occurs
locally. Recurrent cancer tends to occur at the
primary site and is seen within the rst 24 to 36
months after initial treatment. Several theories
have been advanced to explain the phenomenon
of local recurrence after tumor excision with
tumor-free margins. Recurrence may be attributable to the persistence of microscopic malignant
cells within local lymphatics and the ability of the
resulting recurrent tumor to inltrate along tissue
and fascial planes. The use of adjuvant radiotherapy may lead to tissue brosis, hypoxia, and
hypocellularity, which may encase persistent malignant cells, making detection dicult and eventually resulting in local failure.
The aim of all curative ablative surgery should
be to extirpate the tumor completely with wide
tumor-free margins (range: 1.01.5 cm). One of
the most signicant causes of local recurrence is
the persistence of tumor at the resection margins
[14,20,21]. Slootweg and colleagues [22] examined
the resection margins of 394 patients who underwent tumor resection and found that in patients
with negative histologic margins, the incidence
of local recurrence was 3.9% compared with
21.9% for patients with positive margins. Normal
p53 protein serves as a suppressor of cell growth,
and mutations of p53 are demonstrated in approximately 50% of head and neck cancers and

have also been seen in premalignant lesions [23].

The role of p53 tumor suppressor gene in local
recurrence has received much attention since
Brennan and coworkers [24] examined the causal
relation between p53 in resection margins and
the development of local recurrence. Fifty-three
percent of histologically negative margins contained p53 mutations, of which 38% developed local recurrence. The incidence and specic type of
p53 mutation may depend on the risk factor exposure pattern [25,26]. A retrospective study by
Brachman and colleagues [23] suggested that tumors with a p53 mutation had a shorter time to
treatment failure than tumors lacking a p53 mutation. Shindoh and coworkers [27] reported on tumor samples from 118 patients (tumor sites were
the oral cavity, oropharynx, larynx, and hypopharynx); median survival was signicantly shorter
in patients with a p53 mutation in their primary
tumor specimen than in those with no such mutation. There was no dierence in recurrence rates
of the primary tumor according to p53 status,
however, although patients with a p53 mutation
had a higher likelihood of developing a second
primary malignancy. Studies of the chromosome
9p21 to 9p22 indicated mutations in this region
in more than 70% of examined head and neck tumors, and a similar incidence of allelic loss was
found in preinvasive lesions [28,29]. These ndings suggest that loss of genetic information on
chromosome 9p is an early event in head and
neck squamous cell carcinogenesis. It also seems
that information on the myc-oncogene may be
useful for prognosis, but this remains controversial [17,30].
A dilemma that often confronts the surgeon is
the presence of epithelial dysplasia or carcinoma
in situ at the resection margins of OSCC. Approximately 5% to 18% of epithelial dysplasias
become malignant [31,32]. Although one might
expect a greater probability of malignant change
for those dysplastic lesions with a greater histologic degree of epithelial dysplasia, that relation
is hard to prove, because few cases of dysplasia
have been diagnosed and observed for malignant
change without treatment. A greater risk of malignant change in epithelial dysplasia has been associated with the following factors: (1) erythroplakia
within leukoplakia; (2) proliferative verrucous appearance; (3) location at a high-risk anatomic site,
such as the tongue or oor of the mouth; and (4)
presence of multiple lesions [31]. Silverman and
colleagues [31] monitored 257 patients with oral
leukoplakia: 22 had a diagnosis of epithelial

LOCOREGIONAL RECURRENCE OF OSCC

dysplasia, and the remaining 235 had a diagnosis

of hyperkeratosis. Eight (36.4%) of the 22 patients with epithelial dysplasia developed carcinoma. Of the 107 patients with a homogeneous
leukoplakic lesion and a diagnosis of hyperkeratosis, 7 (6.5%) developed carcinoma. Thirty (23.4%)
of the 128 patients with erythroplakic lesions and
a diagnosis of hyperkeratosis were eventually diagnosed with carcinoma. The time from initial diagnosis of epithelial dysplasia or hyperkeratosis to
carcinoma ranged from 6 months to 39 years. In
another study reported by Lumerman and Kerpel
[32], 7 (15.9%) of 44 patients with oral epithelial
dysplasia identied in a biopsy service developed
carcinoma; the mean time from biopsy to cancer
diagnosis was 33.6 months. Epithelial dysplasia
has been more extensively studied in association
with the uterine cervix than with the oral cavity.
Based on clinical reviews, approximately 12% of
cervical epithelial dysplasias progress to carcinoma in situ [33]. The estimated median time for
this progression depends on the histologic severity
of the epithelial dysplasia: 58 months for mild dysplasia, 38 months for moderate dysplasia, and 12
months for severe dysplasia [34]. Approximately
73% of carcinoma in situ cases evolve into fullblown carcinoma [34]. How important this information is for understanding the progression to
oral cancer is unclear, but it is consistent with observations that not all oral epithelial dysplasias
evolve into carcinoma in situ or full-blown carcinoma and that this transition, when it does occur,
takes months or years.
When local recurrence does occur, the time
interval between the completion of treatment and
recurrence is critical in determining the overall
prognosis and salvageability of the patient. This is
particularly true when surgical extirpation combined with neoadjuvant radiotherapy has been
performed previously. Because the treatment of
recurrent locoregional OSCC is individualized, it
is important to balance the potential for curing the
patient with the inevitable signicant decline in
quality of life that occurs after radical surgical
treatment of recurrent tumors (Fig. 1). Aggressive
surgical treatment is warranted when there is a signicant chance of cure or durable palliation
[34,35]. Therefore, if palliative surgery is to be undertaken, it is important to limit morbidity and
disability as much as possible. Schwartz and coworkers [36] noted a 28% recurrence rate among
350 patients with OSCC. Ninety-two percent of
recurrences were identied within 36 months. The
distribution of recurrences was predominantly

617

Fig. 1. A 29-year-old man was initially treated at a referring institution for OSCC of the tongue 6 years previously.
He was initially treated with surgical resection and
postoperative radiotherapy. Local recurrences at the
primary site were managed with excision. He eventually
presented with a massive unresectable locoregional recurrence. He was treated with intra-arterial chemotherapy
and high-dose fractionated radiotherapy for palliation.
(Courtesy of R. Foote, MD, Rochester, MN.)

local in 58% of patients, locoregional in 27%,

and regional in 16%. Recurrences that occurred
within 6 months of primary tumor treatment
showed a mean survival time of 20 months,
and no patients were cured. Recurrences occurring
later than 6 months after initial treatment had
a mean survival time of 58 months. Only 21% of
patients were salvaged with surgery. The time to recurrence was important, with no survivors in the
group in which recurrences were seen less than 6
months from the primary tumor resection.
It is important to note that in this study, the
stage of the recurrence had no predictive value for
survival; however, the stage of the primary tumor
was much more signicant in predicting salvageability. In conclusion, patients who are likely to
have the most survival benet from salvage are
patients (1) with stage I and II primary tumors (2)
who have a recurrence longer than 6 months after
initial treatment, and (3) who develop a recurrence
that is amenable to surgical resection [36,37].
Locoregional or regional recurrence
Currently, 50% of patients with OCSCC
continue to present with advanced-stage disease

618

KADEMANI & DIERKS

with cervical metastasis at the time of diagnosis

[15]. When patients fail regionally, there are several factors that need to be considered for salvageability (Figs. 2 and 3). Patients presenting with
N() neck disease typically undergo a standard
form of oncologic neck dissection. Using the description of the levels of the neck developed
at Memorial Sloan-Kettering Hospital, there are
several varieties of cervical lymphadenectomy
routinely used in the patients with N(). Supraomohyoid dissection (SOHND) removes levels I
through III and modied radical samples of
lymph nodes in levels I through V and typically
spares the sternocleidomastoid muscle (SCM), internal jugular vein (IJV), and spinal accessory
nerve. The traditional gold standard is radical
neck dissection, which involves cervical dissection
at levels I through V and sacrice of the SCM,
IJV, and accessory nerve.
Patients with N(
) neck disease are treated
with prophylactic staging dissections when the
risk of occult neck disease is greater than 15%.
This means that patients with T1 tongue tumors
with a depth of invasion greater than 4 mm and
virtually all T2 and greater tumors for all other
oral cavity sites are treated with staging

dissections. Should the neck dissection specimen

prove histologically negative, no further treatment
is usually initiated [14,3840]. In situations of the
clinically negative neck with positive histologic
ndings after neck dissection of one or more metastatic lymph nodes, multilevel disease, or the
presence of extracapsular spread (ECS), regional
radiotherapy with external beam or intensitymodulated radiation therapy (IMRT) is usually
initiated. Should recurrence later occur in the
neck, it should be identied whether this has occurred in a dissected or undissected eld, because
this has implications for patient survival as well as
salvage options. Patients with recurrence within
a neck dissection eld have a 32% 5-year survival
rate compared with an 18% rate for recurrences
within a previously operated neck. When regional
failure occurs in an unoperated neck, the 5-year
salvage rate is considerably better at 56%. It is interesting to note that in the same series, patients
who had recurrences after initial chemoradiotherapy alone were unsalvageable [41].
The ability of the tumor to metastasize to
cervical lymph nodes is a poor prognostic sign.
Furthermore, violation of the lymph node capsule
(ECS) has been associated with a worse prognosis.

Fig. 2. (A) CT scan shows an 80-year-old patient with recurrent OSCC of the right maxilla. He was initially treated with
infrastructure maxillectomy with tumor-free margins. (B) PET scan after 5 months shows local and ipsilateral cervical
metastasis without evidence of distant disease. He was managed with adjuvant chemoradiotherapy.

LOCOREGIONAL RECURRENCE OF OSCC

Fig. 3. Patient presented 11 months after resection for

tongue carcinoma with ipsilateral neck dissection and
chemoradiotherapy. Fungating ipsilateral regional recurrence is noted in a previously dissected eld. The patient was treated with palliative therapy.

The presence of ECS within the cervical lymph

nodes contributes to an increased rate of regional
failure. Patients with cervical metastasis without
the presence of ECS had a 52% survival rate
compared with 28% with ECS [42]. In 1999,
Woolgar and Rogers [43] examined the role of microscopic ECS (!3 mm) with prognosis and found
the presence of intranodal metastasis decreased
survival to 64% compared with 21% for micrometastatic ECS. Unidentied microscopic metastatic
disease may be present within the cervical lymph
nodes despite current pathologic techniques and
technology, and there is a patient subset in which
regional cervical lymph node metastasis is not
detected. There needs to be a tumor volume of approximately 1 million malignant cells within a 1-g
lymph node (1:1000 ratio) to facilitate detection
by routine hematoxylin and eosin staining. Reports have suggested that 2-mm serial lymph
node sectioning increases the rate of identication
of occult metastasis by up to 5.8%. The addition
of polymerase chain reaction (PCR) analysis for
p53 mutations increases the rate of detection of
cervical metastasis in previously N0 patients by
21% [25]. In a retrospective series from the
Mayo Clinic, histologic samples of T1 OSSC in
patients without a prior history of any oncologic
disease or regional or distant metastasis at

619

presentation were evaluated and reclassied according to Broders histologic grading system for
epidermoid carcinomas. Specimens were also evaluated with depth information from the level of the
adjacent mucosa to the most inltrative portion of
the tumor. All patients were treated with primary
tumor resection without neck dissection. The preliminary data suggest that patients with lesions
greater than 2 mm in depth have an increased
rate of local recurrence, patients with lesions
greater than 4 mm in depth have an increased
rate of regional failure, and patients with lesions
greater than 8 mm in depth tended to fail distantly
as the rst recurrence after surgical extirpation.
The time to recurrence was 2.7 and 1.3 years for
local and regional recurrence, respectively. Approximately two thirds of the study population
were successfully salvaged with multimodality
treatment; however, one third of the study population died, with an average time to death after
the rst recurrence of 3.4 years [40]. It should
also be noted that most patients who died early
in the course of their disease had tumors that
were positive for p53 tumor suppressor gene
mutations [40].
The role of histologic grade in the overall
prognosis and recurrence rates of patients with
oral cancer continues to be of interest. Patients
with poorly dierentiated tumors, regardless of
the site or size of the primary tumor, have a higher
rate of cervical metastasis on presentation
(48.5%) compared with 8.5% for patients with
well-dierentiated tumors. Poorly dierentiated
tumors also show a greater likelihood of close or
positive margins at the time of resection, with this
being seen in 4% of well-dierentiated lesions
versus 25% in poorly dierentiated tumors. The
risk of local or regional failure and patient
survival increases by approximately 44% per
grade based on a retrospective series [44].
Distant recurrence
The presence of distant metastasis from primary squamous cell carcinomas of the upper
aerodigestive tract, particularly the oral cavity, is
a relatively infrequent event, being observed in
2% to 9% of patients (Fig. 4) [10,29,33,4550].
The critical factor in evaluating patients with distant disease is to dierentiate whether the distant
tumor focus represents distant metastasis or a secondary primary tumor. This has tremendous implications for patient treatment and ultimate
survival. This distinction can be dicult in

620

KADEMANI & DIERKS

coworkers [52] evaluated the role of PET scanning

in the preoperative assessment and staging of patients with head and neck tumors and found a
4% incidence of second primary tumors and 4%
of patients with distant metastasis on presentation.
Distant failure in the patient with oral cancer is an
early event and is associated with a poor prognosis,
with a mean survival time of 4 to 12 months. This
raises the question of the utility of maintenance
chemotherapy after initial treatment in the patient
with advanced-stage oral cancer. Although this
may be benecial, most patients with oral cancer
die from locoregional failure and other comorbid
medical conditions. The development of distant
metastasis is still seen in only 5% of patients with
advanced-stage head and neck tumors [51].
Considerations for salvage

Fig. 4. PET scan shows massive widespread distant metastatic disease from a primary tongue OSCC. The patient refused surgical extirpation and went on to
receive chemoradiotherapy. On surveillance imaging 4
months after the conclusion of treatment, he was unfortunately noted to have widespread metastatic disease.
Note the bilateral pulmonary, adrenal, and long bone
metastases.

situations in which the primary and distant tumors are of the same histologic type. Traditionally, the Warren-Gates criteria have been used;
however, current approaches have used genetic
implications of allelic loss and p53 mutations to
relate tumors to each other, which have proven
to be accurate in 50% of cases [29]. The initial
clinical stage, status of the cervical lymph nodes,
depth, grade, and presence of ECS have been reported as the most important risk factors for distant metastasis. The most common site for distant
metastasis from OSCC is the lung, with a mean
survival time of 8.9 months. There are certain subgroups of patients presenting with pulmonary metastasis that may be surgically resected, and these
individuals tend to have a better prognosis [50].
The second most common site of distant metastasis is bone, which has an extremely poor prognosis, with average time from diagnosis to death of
1.8 months [51].
CT or whole-body PET scanning is playing an
increasingly important role in the postoperative
surveillance of these patients. Schmid and

Despite advances in multimodality treatment,

local and regional recurrences remain the most
frequent cause of death in patients with OSSC. Of
the patients who present with recurrence, 50% are
not considered salvageable because of the advanced stage of the tumor at presentation, involvement of local vital structures, and poor
surgical risk [38,5355]. There are several factors
that must be considered in the evaluation of the
patient with locoregional recurrence from oral
cancer. These include the previous treatment received, presenting symptoms, overall prognosis,
quality of life that can be expected, and patients
wishes for further aggressive therapy versus palliative care.
As previously noted, the treatment that the
patient has received before the development of
a local regional recurrence is an important factor
in determining the options that are available for
salvage. When recurrence occurs in the previously
reconstructed eld, this may be dicult to detect.
Marginal recurrences are often easily detectable
clinically; however, this is not the case with deep
recurrences, which are often dicult to evaluate
clinically and radiologically in the early stages and
may be mistaken for postoperative changes.
Because of the altered anatomy from tumor
extirpation and free tissue transfer, patterns of
tumor spread are often atypical. Patients in an
advanced stage often require adjuvant treatment
with chemoradiotherapy, which increases the degree of brosis and tissue hypoxia and leads to
a further lack of denition of recurrences and
tissue planes through which malignant cells may
spread.

LOCOREGIONAL RECURRENCE OF OSCC

Despite advances in multimodality treatment,

surgical resection of local recurrence, if feasible, is
still considered the gold standard of treatment as
compared with radiotherapy with or without
chemotherapy [36]. In 2004, Lin and colleagues
[56] examined the utility of primary surgical management and surgical salvage in patients with
recurrent OSSC. Five-year survival rates were
60% and 38%, respectively, with a higher rate
of overall complications in the salvage group. It
is important to note that most failures were seen
in the rst year, with those patients surviving for
more than 18 months being considered as salvaged. Surgical management of local recurrence
has also been shown to produce a 60% salvage
rate in patients primarily treated with radiotherapy [56,57]. This is in contrast to reports of
salvage with radiotherapy in recurrent tongue
cancer, with a mean survival of only 9 months
and a 2-year disease-free survival of 15% [58].
Eckardt and coworkers [59] compared the survival
of patients treated with surgery with that of patients treated with multimodality treatment over
a 20-year period and found a 31.0% rate of salvage with surgery as compared with a 15.4%
rate with surgery and radiotherapy, with a salvage
rate of 0% for patients treated with radiotherapy
or supportive therapy alone. When approaching
the treatment alternatives for patients with recurrent oral cancer, it is important that the stage of
the primary tumor rather than the stage of the recurrence be the most relevant indicator of overall
patient prognosis [36,37,56,59,60].
In patients who present with recurrent neck
disease, there are several factors to consider. The
importance of previous neck therapy and type of
neck dissection are important to consider in
overall patient survival. The type of neck dissection and whether the recurrence is seen in a previously dissected eld also have prognostic and
treatment implications. Wong [41] and others
[51,60] have evaluated the presence of regional recurrence from oral and oropharyngeal tumors and
found that a 56% surgical salvage rate could be
achieved with a 32% 5-year survival rate in patients with an ipsilateral neck recurrence in an unoperated neck. In patients with a history of prior
neck dissection, surgical salvage was reduced to
32% and, consistent with other reports, no durable salvage could be achieved with radiotherapy
or chemotherapy. In situations in which primary
treatment has been with radiotherapy, the ability
to salvage the patient in the face of a recurrence
is reduced. Mabanta and colleagues [61] analyzed

621

the recurrence patterns of 356 patients with N()

being treated with radiotherapy and found a 14%
rate of recurrence. Sixty-ve percent of patients
were unsalvageable by any means, with a 5-year
disease-free survival rate of 10%. The conclusion from this study was that successful salvage
treatment after neck recurrence after primary
radiotherapy is unlikely. Figs. 5 and 6 outline a
schematic approach to the management of recurrence in patients with primary early and advanced-stage tumors that present with local or
regional recurrence.

Summary
The best opportunity for a cure in the patient
with oral cancer is at the time of initial diagnosis.
The result of treatment of recurrent disease after
prior denitive treatment is generally poor. Local
recurrence at the primary site is relatively common as compared with regional and distant sites.
Local recurrence is seen in 20% to 30% of
patients and is the most common cause of death
in the patient with oral cancer. This is followed by
regional recurrence, which accounts for 10% to
15% of mortality. Death from distant disease is
relatively uncommon when local and regional
disease is controlled, occurring in 1% to 3.8%
of patients with tumors of the oral cavity [37,51].
Although there is a paucity of reliable data regarding the management of recurrent oral cancer,
there is a general consensus that surgical resection
of recurrent disease in patients willing and able to
tolerate extirpative procedures provides the most
durable long-term salvage rates. Even though surgical salvage is the best option for cure, this is not
without potential complications. Surgery for recurrent disease is often an extensive undertaking
having signicant functional, esthetic, social, and
nancial implications for these patients. As surgeons, we have to consider whether the eorts of
extirpation are likely to provide the patient with
a meaningful likelihood of salvage or signicant
increase in lifespan without considerable functional compromise. This was best summarized
by Goodwin [37] in a meta-analysis of 32 published reports on salvage surgery with recurrent
squamous cell carcinoma, where it was stated
that we must consider whether the end justies
the means in this patient population.
Salvage surgery is often the best option for
patients with disease that is amenable to resection
and who are able to tolerate surgery, particularly

622

KADEMANI & DIERKS

Patients with primary

tumor stage T1-2N0M0

Recurrence <6 months after

treatment of primary tumor

Recurrence >6 months after

treatment of primary tumor

Resectable Local
Recurrence

Unresectable Local
Recurrence

Regional
Recurrence

Salvage
surgery

Chemotherapy
radiation therapy

Salvage neck
dissection

Resectable Local/
Regional Recurrence

Unresecatble Local/
Regional Recurrence

Salvage surgery

Treatment
failure

Fig. 5. Algorithm for the treatment of recurrence in the early-stage I to II patient. (Adapted from Schwartz GJ, et al.
Salvage treatment for recurrent squamous cell carcinoma of the oral cavity. Head Neck 2000;22:37; with permission.)

among patients initially treated with irradiation.

For most patients, the original diagnosis of oral
cancer is often associated with tremendous anxiety, which is compounded when a recurrent tumor
is detected. It is imperative for the treating
physician to communicate the likelihood of salvage, overall prognosis, and expected quality of

life compassionately yet realistically. For patients

with a limited long-term prognosis, decisions
based on religious, cultural, and social value
systems play important roles in whether surgical
salvage should be contemplated. Head and neck
oncologists have made tremendous strides in
recent years with improved multimodality

Patients with primary

tumor stage T3-4N0M0

Resectable Local
Recurrence

Salvage
surgery

Unresectable Local
Recurrence

Regional Recurrence in
Previously Operated Neck

Chemotherapy
radiation therapy
Regional Recurrence in
Undissected Neck Field

Treatment
failure

Fig. 6. Algorithm for the treatment of recurrence in the advanced-stage III to IV patient. (Adapted from Schwartz GJ,
et al. Salvage treatment for recurrent squamous cell carcinoma of the oral cavity. Head Neck 2000;22:38; with
permission.)

LOCOREGIONAL RECURRENCE OF OSCC

treatments for oral cancer; however, there have

been no randomized prospective trials evaluating
the ecacy and most appropriate protocol for
treating patients with recurrent oral carcinoma.
From the retrospective data available, it is clear
that salvage surgery is appropriate in patients with
amenable recurrence longer than 6 months after
treatment who are medically able and willing to
tolerate tumor extirpation. This is particularly
ecacious in situations in which initial treatment
with radiotherapy has been administered or in
early-stage 1 and 2 primary tumors (see Figs. 5
and 6). In patients with recurrent advanced disease who are unable or unwilling to undergo surgery, neoadjuvant therapy has a role; however, the
salvageability of this group of patients in the long
term is likely to be dismal. It is imperative that
further controlled studies be performed to dene
the optimal treatment strategies for this challenging group of patients. Future improved outcome
data may enable patients and families to make
the best decisions on whether treatment of recurrent cancer is likely to be of benet.
Acknowledgments
The authors thank Robert Foote, MD, Professor of Radiation Oncology, Mayo Clinic, for
providing images used in this article.

References
[1] Parkin SM, Laara E, Muir CS. Estimates of the
worldwide frequency of sixteen major cancers. Int
J Cancer 1988;41:18497.
[2] Park BZ, Kohn WG, Malvitz DM. Preventing and
controlling oral and pharyngeal cancerdrecommendations from a national planning conference.
MMWR Morb Mortal Wkly Rep 1998;47:112.
[3] Johnson NW, Warnakulasuriya KAA. Oral cancer.
Is it more common than cervical? Br Dent J 1991;
170:1701.
[4] Nair UJ, Friesen M, Richard I, et al. Eect of lime
composition on the formation of reactive oxygen
species from the areca nut extract in vivo. Carcinogenesis 1990;11:21458.
[5] Hindle I, Downer MC, Speight PM. The epidemiology of oral cancer. Br J Oral Maxillofac Surg 1996;
34:4716.
[6] Devesa SS, Silverman DT, Young JL, et al. Cancer
incidence and mortality trends among whites in the
United States, 19471987. J Natl Cancer Inst 1987;
79:70170.
[7] Worrall SF. Oral cancer incidence between 1971 to
1989. Br J Oral Maxillofac Surg 1995;33:1959.

623

[8] Zakrzewska JM. Oral cancer and precancerdour responsibility. Br Dent J 1994;176:2867.
[9] Friedlander PL, Schantz SP, Shaha AR, et al. Squamous cell carcinoma of the tongue in young patients.
A matched paired analysis. Head Neck 1988;20:
3638.
[10] Day GL, Blot WJ. Second primary tumors in patients with oral cancer. Cancer 1992;70:149.
[11] Foulkes WD, Brunet JS, Sieh W, et al. Familial risks
of squamous cell carcinoma of the head and neck,
retrospective case controlled study. Cancer 1996;
313:71621.
[12] Johnson NW, Ranasingle AW, Warnakulasuriya
KAA. Epidemiology of oral cancer in the United
Kingdom. Community Dent Health 1993;10(Suppl):
1329.
[13] Moore S, Johnson N, Pierce A. The epidemiology of
mouth cancer. Oral Dis 2000;6:6574.
[14] Woolgar J, Scott J. Prediction of cervical metastasis
in squamous cell carcinoma of the tongue and oor
of mouth. Head Neck 1995;17:46372.
[15] Shah J. Cervical lymph node metastasis, its diagnostic, therapeutic and prognostic implications. Oncology 1990;4:619.
[16] Centers for Disease Control and Prevention. Current trends; deaths from oral cavity and pharyngeal
cancerdUnited States 1987. JAMA 1990;264:
67885.
[17] Jones KR, Lodge-Rigal RD, Reddick RL, et al.
Prognostic factors in the recurrence of stage 1 and
2 squamous cell carcinoma of the oral cavity. Arch
Otolaryngol Head Neck Surg 1992;118:4835.
[18] Nathanson A, Agren K, Biorklund A. Evaluation of
some prognostic factors in small squamous cell carcinoma of the mobile tongue: a multicenter study in
Sweden. Head Neck 1989;11:38792.
[19] Pearlman NW. Treatment outcome in recurrent
head and neck cancer. Arch Surg 1979;114:3942.
[20] Jones AS, Binhano Z, Nadapalan V, et al. Do positive resection margins after ablative surgery for
head and neck cancer adversely aect prognosis. A
study of 352 patients with recurrent carcinoma following radiotherapy treated by salvage surgery. Br
J Cancer 1996;74:12832.
[21] Ord RA, Aisner S. Accuracy of frozen sections in
assessing margins in oral cancer resection. J Oral
Maxillofac Surg 1997;55(7):66971.
[22] Slootweg PJ, Hordijk GL, Schade Y, et al. Treatment failure and margin status in head and neck cancer. A critical view on the potential value of
molecular pathology. Oral Oncol 2002;38(5):5003.
[23] Brachman DG, Graves D, Volkes E. Occurrence of
p53 gene deletions and human papillomavirus infection in human head and neck cancer. Cancer Res
1992;62:48326.
[24] Brennan JA, Boyle JO, Koch WF. Association between cigarette smoking and mutation of the p53
gene in squamous-cell carcinoma of the head and
neck. N Engl J Med 1995;332:7127.

624

KADEMANI & DIERKS

[25] Brennan JA, Mao L, Hruban RH. Molecular assessment of histopathological staging in squamous-cell
carcinoma of the head and neck. N Engl J Med
1995;332:42935.
[26] Shindoh M, Chiba I, Yasuda M. Detection of human papillomavirus DNA sequences in oral squamous cell carcinomas and their relation to p53 and
proliferating cell nuclear antigen expression. Cancer
1995;76:151321.
[27] Shindoh M, Kim J, Ro JY. Activation of p53
gene expression in premalignant lesions during
head and neck tumorigenesis. Cancer Res 1994;54:
3216.
[28] van der Riet P, Nawroz H, Hruban RH. Frequent
loss of chromosome 9p21-22 early in head and
neck cancer progression. Cancer Res 1994;54:
11568.
[29] Nawroz H, van der Riet P, Hruban RH, et al. Allelotype of head and neck squamous cell carcinoma.
Cancer Res 1994;54:11525.
[30] Brachman DG. Molecular biology of head and neck
cancer. Semin Oncol 1994;21:3209.
[31] Silverman S Jr, Gorsky M, Lozada MF. Oral leukoplakia and malignant transformation: a follow-up
study of 257 patients. Cancer 1984;53:5638.
[32] Lumerman HFP, Kerpel S. Oral epithelial dysplasia
and the development of invasive squamous cell carcinoma. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 1995;79:3219.
[33] Ostor AG. Natural history of cervical intraepithelial
neoplasia: a critical review. Int J Gynecol 1993;12:
8692.
[34] Richart RM, Barron AA. A follow-up study of patients with cervical dysplasia. Am J Obstet Gynecol
1993;105:38693.
[35] Ridge JA. Squamous cell carcinoma of the head and
neck; a surgical treatment of local and regional recurrence. Semin Oncol 1993;20:41929.
[36] Schwartz GJ, Mehra RM, Wening BL, et al. Salvage
treatment for recurrent squamous cell carcinoma of
the oral cavity. Head Neck 2000;22:3441.
[37] Goodwin WJ. Salvage surgery for patients with recurrent squamous cell carcinoma of the upper aerodigestive tract: When does the end justify the means.
Laryngoscope 2000;110:118.
[38] Shah J, et al. Carcinoma of the oral cavity. Factors
aecting treatment failure at the primary site and
neck. Am J Surg 1976;132:5047.
[39] Lindberg R. Distribution of cervical lymph node metastasis from squamous cell carcinoma of the upper
respiratory and digestive tracts. Cancer 1972;29:
14469.
[40] Kademani D, Baltali E, Keller EEK, et al. Analysis
of cause of death and patient survival in T1 oral
squamous cell carcinoma, in press.
[41] Wong L. Salvage of recurrent head and neck squamous cell carcinoma after primary curative surgery.
Head Neck 2003;25(11):9539.

[42] Johnson J, Barnes L, Myers E. The extracapsular

spread of tumors in cervical node metastasis. Arch
Otolaryngol 1981;107:72531.
[43] Woolgar J, Rogers S. Survival and patterns of recurrence in 200 oral cancer patients treated by radical
surgery and neck dissection. Oral Oncol 1999;35:
25765.
[44] Kademani D, Bell RB, Dierks EJ, et al. Prognostic
factors in intraoral squamous cell carcinoma: the inuence of histologic grade. J Oral Maxillofac Surg
2005;63(11):1599605.
[45] Al-Othman MOF, Morns GG, Hinerman RW, et al.
Distant metastases after denitive radiotherapy for
squamous cell carcinoma of the head and neck.
Head Neck 2003;25:62933.
[46] Belson TP, Lehman RH, Chobanian SL, et al. Bone
and liver scans in patients with head and neck carcinoma. Laryngoscope 1980;90:12916.
[47] de Bree R, Deurlow EG, Snow GB, et al. Screening
for distant metastases in patients with head and neck
cancer. Laryngoscope 2000;110:397401.
[48] Matsuo JMS, Matsuo YY. Clinical nodal stage is an
independently signicant predictor of distant failure
in patients with squamous cell carcinoma of the larynx. Ann Surg 2003;238:41222.
[49] Leon X, Quer M, Orus C, et al. Distant metastases in
head and neck cancer patients who achieved loco-regional control. Head Neck 2000;22:6806.
[50] Younes RN, Gross RL, Silva JF. Surgical treatment
of lung metastases of head and neck tumors. Am J
Surg 1997;174:499502.
[51] Kowalski LP, Carvalho AL, Prinate AVM, et al.
Predictive factors for distant metastasis from oral
and oropharyngeal squamous cell carcinoma. Oral
Oncol 2005;41:53441.
[52] Schmid DT, Stoecki JF, Bandhaem F, et al. Impact
of positron emission tomography on the initial staging and therapy in locoregional-advanced squamous
cell carcinoma of the head and neck. Laryngoscope
2003;113:88891.
[53] Leemans C, Tiwari R, Nauta JJ, et al. Recurrence at
the primary site in head and neck cancer and the signicance of neck lymph node metastasis as a prognostic factor. Cancer 1994;73:18790.
[54] Grandi C, Allossio M, Moglia DEA. Prognostic signicance of lymphatic spread in head and neck carcinomas: therapeutic implications. Head Neck Surg
1985;8:6773.
[55] Kowalski LP, Magrin J, Waksman G. Supraomohyoid neck dissection in the treatment of head and
neck tumors: survival results in 212 cases. Arch Otolaryngol Head Neck Surg 1993;119:95863.
[56] Lin YC, Hsiao JR, Tsai ST. Salvage surgery as the
primary treatment for recurrent oral squamous cell
carcinoma. Oral Oncol 2004;40:1839.
[57] Cherian T, Sebastian P, Ahamed MI, et al. Evaluation of salvage surgery in heavily irradiated cancer
of the buccal mucosa. Cancer 1991;68:2959.

LOCOREGIONAL RECURRENCE OF OSCC

[58] Sun LM, Levins W, Wang CJ, et al. The relapse

pattern and outcome of postoperative recurrent
tongue cancer. J Oral Maxillofac Surg 1994;55:
82731.
[59] Eckardt A, Barith EL, Kokemueller H, et al. Recurrent carcinoma of the head and neck: treatment
strategies and survival analysis in a 20-year period.
Oral Oncol 2004;40:42732.

625

[60] Gleich LL, et al. Recurrent advanced (T3, T4) head

and neck squamous cell carcinoma: is salvage possible.
Arch Otolaryngol Head Neck Surg 2005;130:358.
[61] Mabanta S, Mendenhall WM, Stringer SP, et al. Salvage treatment for neck recurrence after irradiation
alone for head and neck squamous cell carcinoma
with clinically positive neck nodes. Head Neck
1999;99:5914.

Oral Maxillofacial Surg Clin N Am 18 (2006) 627641

Interventional Approaches to Management

of Pain of Oral Cancer
Sukdeb Datta, MD, DABPM, FIPPa,b,*, Umeshraya T. Pai, MD, FIPPc
a

Pain Management Services, Tennessee Valley Healthcare Systems, Department of Veterans Aairs,
1310 24th Ave South, Nashville, TN 37212, USA
b
Department of Anesthesiology, Vanderbilt University Medical School, Nashville, TN 37212, USA
c
Departments of Anesthesiology and Anatomy, University of Cincinnati Medical Center,
PO Box 670531, Cincinnati, OH 45267-0531, USA

Pain is present in up to 80% of patients with

cancer of the head and neck. Several factors make
pain management dicult in patients with head
and neck cancer, including the erosive nature of
the neoplasms that invade the region; the rich
innervation of the head and neck; the dynamic
pain evoked by unavoidable movements, such as
swallowing, talking, or chewing; and the neuropathic pain induced by chemotherapy or radiation.
Often, a multidisciplinary approach is indicated
initially, which includes use of medication trials
(systemic analgesic therapy with nonsteroidal
anti-inammatory drugs, anticonvulsants, antidepressants, phenothiazines, oral and parenteral
opioids, and other adjuvants); surgery; radiation
therapy or chemotherapy directed at the neoplasm; or use of transcutaneous nerve stimulation,
acupuncture, or therapy with hormones or steroids. Another important approach is the use of
interventional pain medicine therapies. Interventional approaches to this complex pain problem
include the use of peripheral nerve blocks, blockade of relevant ganglia, and use of central
neuraxial techniques. The goals of this review
are to review the anatomy of the head and neck
region and then to detail relevant techniques for
management of cancer pain of the head and neck.

This work was supported in part by the

Department of Veterans Affairs, Tennessee Valley
Healthcare System.
* Corresponding author.
E-mail address: sdattamd@gmail.com (S. Datta).

Mechanisms of cancer pain of the head and neck

Head and neck tumors produce pain through
several mechanisms, including stimulation of
nerve endings in mucous membranes, compression or invasion of sensory nerves, ulceration and
infection, bony invasion and destruction, and as
a complication of oncologic therapy [1]. Stimulation of free nerve endings in the mucosa and submucosa is often one of the initial signs of tumor
growth. Stimulation of these nociceptors generally
produces supercial pain, often causing a local
constant burning sensation. Malignant tumors
that arise from the mucous membranes are primarily squamous cell carcinomas. These tumors
tend to ulcerate, although the ulceration, per se,
does not necessarily produce the pain. Additionally, ulcerated tissue often becomes secondarily infected, leading to inammation and edema that
further potentiate the pain. Invasion or compression of the trigeminal nerve is another cause of
pain. The pain is generally felt in the area innervated by the particular portion of the nerve that
is involved. Bone invasion by tumor does not initially produce pain. As the tumor grows, however,
it causes the periosteum to expand, which stimulates the nociceptive aerents that innervate this
structure. The pain is localized to the site of involvement but can be referred to a more distal
area. Another way that bony involvement causes
pain is secondary bacterial invasion with the development of osteomyelitis. Finally, pain may occur as a complication or sequelae of treatment of
the underlying tumor. Radical surgical resection

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.014

oralmaxsurgery.theclinics.com

628

DATTA & PAI

of head and neck cancer often requires that nerves

be transected to achieve a total resection. This can
lead to the formation of painful neuromas. Mucositis is a common complication of chemotherapy
administered alone or in combination with radiation therapy. Some patients also develop an outbreak of herpes zoster in the area involved by tumor
and go on to develop postherpetic neuralgia.

Anatomy of head and neck

For the purpose of this review, relevant
anatomic considerations include the peripheral
nervous system, the autonomic ganglia (sphenopalatine and stellate), and the central neuraxis.
They are detailed in sequence, and relevant
interventional pain medicine techniques are described. The dermatomes are well dened, with
good options for approaching and blocking most
peripheral nerves in this region of the body
(Fig. 1).

trigeminal nerve and the autonomic nervous

system, including the ciliary, sphenopalatine,
otic, and submaxillary ganglia (Figs. 2 and 3).
Proximally, the sensory component of the trigeminal nerve is connected to the ventral aspect of the
pons. Distally, the sensory component leaves from
the medial concave border of the trigeminal (semilunar) ganglion. The trigeminal ganglion is located at the apex of the petrous temporal bone
in the posterior medial part of the middle cranial
fossa. The ganglion is also related to the inferior
lateral aspect of the cavernous sinus. The minor
motor component of the trigeminal nerve is located at the medial side of the nerve at the attachment to the pons and runs inferior to the ganglion
to exit through the foramen ovale, with the mandibular division as the motor branch.
Trigeminal (gasserian) ganglion

The trigeminal nerve is the fth cranial nerve

and has three divisions: ophthalmic, maxillary,
and mandibular. It is predominantly sensory, with
a small motor component in the mandibular
nerve. Communication exists between the

The gasserian ganglion is formed from two

roots that exit the ventral surface of the brain
stem at the midpontine level. These roots pass in
a forward and lateral direction in the posterior
cranial fossa across the border of the petrous part
of the temporal bone. They then enter Meckels
cave, which is formed by invagination of the
surrounding dura mater into the middle cranial
fossa. The dural pouch that lies just behind the
ganglion, called the trigeminal cistern, contains

Fig. 1. Dermatomes indicating innervation of the neck.

(From Rosenberg M, Phero JC. Regional anesthesia and
invasive techniques to manage head and neck pain. Otolaryngol Clin North Am 2003;36:1202.)

Fig. 2. Trigeminal nerve anatomy. (From Rosenberg M,

Phero JC. Regional anesthesia and invasive techniques
to manage head and neck pain. Otolaryngol Clin North
Am 2003;36:1208.)

Trigeminal system

INTERVENTIONAL APPROACHES TO ORAL CANCER PAIN MANAGEMENT

Fig. 3. Trigeminal nerve sensory distribution. (From

Rosenberg M, Phero JC. Regional anesthesia and invasive techniques to manage head and neck pain. Otolaryngol Clin North Am 2003;36:1209.)

cerebrospinal uid (CSF). The three sensory divisions of the gasserian ganglion are ophthalmic
(V1), maxillary (V2), and mandibular (V3) divisions. A small motor root joins the mandibular
division as it exits the cranial cavity via the
foramen ovale.
Ophthalmic nerve
The ophthalmic nerve is purely sensory in
function. It enters the orbit via the superior
orbital ssure. Its branches are the frontal,
nasociliary, and lacrimal nerves. The terminal
branches of the frontal nerve are the supraorbital
and supratrochlear nerves. These terminal
branches exit the orbital cavity anteriorly and
provide innervation to the upper eyelid, forehead,
and scalp. The terminal branches of the nasociliary nerve consist of the infratrochlear and external nasal nerves, which provide cutaneous and
mucosal innervation to the apex and ala of the
nose and anterior nasal cavity. The lacrimal nerve
continues to innervate the lacrimal gland and
outer canthus of the eye.
Maxillary nerve
The maxillary nerve (V2) is the second division
of the trigeminal nerve. It is purely sensory. The
maxillary nerve exits from the cranial cavity
through the foramen rotundum. From this point,
the nerve traverses the superior part of the
pterygopalatine fossa and swings laterally to
traverse the inferior orbital ssure toward the
maxillary sinus. As the nerve runs along the roof

629

of the maxillary sinus, it supplies the maxillary

sinus itself and the anterior teeth of the upper jaw
via the anterior and middle superior alveolar
nerves. The nerve then exits through the infraorbital foramen to innervate the skin of the face and
the underlying mucosa extending from the lower
eyelid to the upper lip. Although the nerve is
located at the pterygopalatine fossa, it is connected
to the pterygopalatine ganglion, through which it
gives o branches to the nasal cavity, pharynx, and
palate. In addition, the nerve gives o the zygomatic nerve and the posterior superior alveolar
nerve. The zygomatic nerve supplies the lateral
portion of the face, and the posterior superior
alveolar nerve supplies the upper molar region.
To summarize, the branches of the maxillary
nerve may be subdivided into four regional
groups:
1. Intracranial group: including the middle
meningeal nerve, which supplies the dura mater of the middle cranial fossa
2. Pterygopalatine group: the zygomatic nerve
and sphenopalatine, greater and lesser palatine and posterior superior alveolar branches
3. Infraorbital canal group: the anterosuperior
alveolar branch, which innervates the incisors and canines, anterior wall of the maxillary antrum, and oor of the nasal cavity,
and the middle superior branch, which supplies the premolars
4. Infraorbital facial group: the inferior palpebral, external nasal, and superior labial
branches

Mandibular nerve
The mandibular nerve is the third division of
the trigeminal nerve (V3). It arises from the lower
part of the distal convexity of the trigeminal
ganglion and then joins the motor component.
From this point, the nerve exits the foramen ovale
to enter the infratemporal fossa. It then divides
into a smaller anterior division and a larger
posterior division. The anterior division is predominantly motor, except for the buccal branch,
which provides sensation to the cheek. The motor
branches innervate the muscles of mastication.
The posterior division is predominantly sensory,
except for the mylohyoid branch, which provides
motor innervation to the mylohyoid muscle and
the anterior belly of the digastric muscle. The
sensory portion of the mandibular nerve innervates the meninges (via the recurrent meningeal

630

DATTA & PAI

branch), the temporomandibular joint, the ear,

and the outer surface of the tympanic membrane;
the anterior two thirds of the tongue and adjoining oor of the mouth; and the mandible with its
associated teeth. The sensory portion terminates
as the mental nerve, which supplies the chin and
mucous membrane of the lower lip.
Peripheral nerve blocks in the management of oral
cancer pain
Maxillary nerve block
History
There are several dierent approaches to
blocking the maxillary nerve. An oral approach
is commonly practiced by dentists. Levy and
Baudoni [2] advocated the approach commonly
used by pain physicians (ie, the lateral approach).
Maxillary nerve blockade is indicated in the case
of an oral cavity with involvement in the distribution of the maxillary nerve [3,4]. An indwelling
catheter has been described.
Landmarks
1. Midpoint of the zygomatic arch of the temporal bone
2. Condyle of the mandibular head
3. Coronoid process of the mandible; the level
of the coronoid notch is at the external auditory meatus.
4. Mandibular notch between the condyle and
the coronoid process; this is most easily
done by having the patient open and close
the mouth.
Commonly used approaches in performing
a maxillary nerve block are described (Fig. 4).
Lateral approach
The patient is positioned supine with the head
turned away from the side of the intended block.
The side of the face is prepared and draped in
a sterile fashion. The landmarks are palpated, and
the midpoint of the zygomatic arch is marked.
The skin is inltrated with local anesthetic in the
area of the mandibular notch between the condyle
and the coronoid process of the mandible below
the midpoint of the zygomatic arch. The level of
the coronoid notch is at the level of the external
auditory meatus.
After inltration, a 22-gauge, 3-inch needle is
introduced perpendicular to the midpoint of the
zygomatic arch and walked onto the lateral
pterygoid plate. A depth marker can be placed

on the needle to the anticipated depth (approximately 0.51 cm from the initial depth to the
lateral pterygoid plate). The needle is then withdrawn and redirected and advanced anteriorly
and superiorly at an angle of approximately 45
toward the root of the nose (see Fig. 4). A nerve
stimulator may be helpful if available, because
paresthesias may or may not be elicited. After aspiration to rule out intravascular placement of the
needle, 2 to 3 mL of local anesthetic is deposited
for desired eect. To minimize a CSF injection,
the needle should not be advanced further than
1.5 cm past the lateral pterygoid plate. In a human
and osteologic study, Singh and colleagues [5]
found that the needle should not be advanced by
more than approximately 0.25 cm beyond the distance to the pterygoid plate while performing
a maxillary nerve block by the lateral extraoral
approach.
Anterolateral approach
The patient is positioned and prepared in the
manner described previously. The angle between
the inferior border of the zygomatic bone and the
coronoid process of the mandible is located and
marked. After a skin wheal is raised at this angle,
a 3-inch, 22-gauge needle is directed medially,
superiorly, and posteriorly to lie along the posterior surface of the maxilla and is further advanced
approximately 4 to 5 cm depending on the extent
of the subcutaneous tissue. Once the needle tip
walks o the maxilla, a paresthesia may be elicited
when the needle tip reaches the pterygopalatine
fossa. As before, a nerve stimulator may be
helpful to verify the position of the needle.
Suprazygomatic approach
A suprazygomatic approach has been advocated by Fujii [6] and popularized by Okuda and
coworkers [7]. The patient is placed supine, and
a wheal of local anesthetic is placed in the skin
overlying the angle formed by the superior edge
of the zygomatic arch and its anterior portion.
The 23-gauge, 5-cm needle is inserted perpendicularly above the weal and advanced to reach the
greater wing of the sphenoid. The needle is then
reinserted in a slightly anterior direction to advance approximately 0.5 cm deeper than the depth
of initial contact. Paresthesia may or may not be
elicited.
The suprazygomatic approach may have
advantages in preventing hemorrhagic complications. This approach limits penetration of the
maxillary artery, because the maxillary artery in

INTERVENTIONAL APPROACHES TO ORAL CANCER PAIN MANAGEMENT

631

Fig. 4. Maxillary and mandibular nerve block. (Upper panel) Coronoid notch is located at the midpoint of the zygoma.
A nger is placed at this point, and the patient is asked to open the mouth. The condyle of the mouth should be palpable
immediately, deep to the ngertip as the mouth opens. The ngertip should then sink into the coronoid notch as the
mouth is closed. (Lower panel) Maxillary and mandibular nerves are approached by way of the coronoid notch below
the midpoint of the zygoma. (1) Needle passes through the infratemporal fossa to reach the lateral pterygoid plate. The
initial direction of the needle should be medial and slightly anterior. (2) Needle is then walked anteriorly until it passes
into the pterygomaxillary (pterygopalatine) fossa, where the maxillary nerve is blocked. (3) Needle is then walked from
position (1) posteriorly until it passes just posterior to the lateral pterygoid plate to block the mandibular nerve as it
emerges from the foramen ovale. The needle point is kept at the same depth as the lateral pterygoid plate to prevent
accidental introduction of the needle into the posterior pharynx. (Adapted from Murphy TM. Somatic blockade of
head and neck. In: Cousins MJ, Bridenbaugh PO. Neural blockade in clinical anesthesia and management of pain.
3rd edition. Philadelphia: Lippincott Raven; 1998. p. 500; with permission.)

the pterygopalatine fossa lies ventral to the

maxillary nerve and its branch and there is
minimal distance between the point of entry of
the needle and the pterygopalatine fossa.

Intraoral approaches
The various intraoral approaches to the maxillary nerve include the anterior and middle
superior maxillary anesthesia (AMSA) technique

632

DATTA & PAI

[8], palatal approach-anterior superior alveolar

(P-ASA) technique [9], and pterygopalatine canal
technique [10].
Complications
1. If the needle is placed too deep and anterior,
direct injection into the optic nerve is possible, resulting in permanent blindness with
the use of neurolytic agents. Transient diplopia has been reported [11].
2. The exceedingly vascular nature of the compartment in which the maxillary nerve lies
makes intravascular injection quite possible,
and meticulous aspiration tests are essential.
3. Because the maxillary nerve injection site is
quite vascular, hematoma formation is common. A suprazygomatic approach may decrease the incidence of hematoma formation.
4. Aspiration of air usually indicates that the
needle has been placed too far posteriorly
and the pharynx has been entered. If this
happens, the needle direction needs to be
changed.
Mandibular Nerve Block
Landmarks
Landmarks are similar to those described for
the maxillary nerve block.

Lateral extraoral technique

The patient is positioned and prepared in
a manner similar to that described for the
maxillary nerve block (Fig. 5). The landmarks
are palpated, and a point below the midpoint of
the zygomatic arch in the mandibular notch is
marked. After inltration of the overlying skin
with local anesthetic, a 22-gauge, 3-inch needle is
inserted perpendicular to the skin and advanced
posteromedially to a depth of approximately 4
to 5 cm; at this point, paresthesias in the distribution of the nerve may be elicited. Local anesthesia
is deposited after negative aspiration to achieve
the desired eect. Conversely, if bone (lateral pterygoid plate) is contacted, the needle is withdrawn
and redirected more posteriorly in an attempt to
elicit paresthesias (see Fig. 4). Singh and colleagues [12] postulated that there is no osteologic
basis to advance the needle beyond the distance to
the lateral pterygoid plate to reach the mandibular
nerve. Because the needle may contact the nerve at
variable points rather than the shortest distance to
the nerve, however, the needle may be advanced
by 0.07 cm on the right side and by 0.11 cm on
the left side in patients.
Intraoral techniques for blockade of mandibular
nerve
Successful anesthesia of the mandibular teeth
and soft tissues is more dicult to achieve than

Fig. 5. Lateral extraoral approach for maxillary and mandibular nerve block. Please note that a suprazygomatic approach is also possible. (Adapted from Pai UT, Nayak R. Maxillary, mandibular, and glossopharyngeal nerve blocks.
In: Benzon HT, Raja SN, Borsook D, editors. Essentials of pain medicine and regional anesthesia. New York: Churchhill Livingston; 1999. p. 10; with permission.)

INTERVENTIONAL APPROACHES TO ORAL CANCER PAIN MANAGEMENT

anesthesia of maxillary structures. Some factors

involved are the greater anatomic variation in the
mandible and the depth of soft tissue penetration
required. The Gow-Gates technique requires a single intraoral puncture site, and the anesthetic
solution is deposited in a single target area
without changing the needle position. The GowGates technique is indeed a true mandibular nerve
block because it provides sensory anesthesia to
virtually the entire distribution of the mandibular
nerve, including the inferior alveolar, lingual,
mylohyoid, mental, incisive, auriculotemporal,
and buccal nerves.
Traditional intraoral technique
The target needle area is the imaginary line
drawn through the coronoid notch parallel to the
mandibular occlusal plane and vertically drawn
two thirds to three fourths of the distance between
the coronoid notch and the posterior part of the
ramus. The solution is deposited slightly superior
to the lingula, which houses the orice for the
neurovascular bundle entering the mandible.
Gow-Gates intraoral technique
The tip of the needle is placed just below the
mesiolunal cusp of the maxillary second molar
and is moved to a point just distal to the molar.
The puncture point is located just medial to the
deep tendon of the temporalis muscle and as close
to cusps of the maxillary second molar as the
syringe allows. Externally, the border of the
tragus and the corner of the mouth provide a guide
for the line of injection.
Akinosi technique
Akinosi [13] developed a procedure for application in situations in which the patient is
unable to open the mouth, thereby precluding
the performance of the conventional mandibular
block. In the Akinosi technique, the mouth of
the patient remains partially closed, without
establishing occlusal contact, thus allowing relaxation of cheek muscles and facilitating cheek
separation for improved visualization of the
zone. The needle is inserted between the ascending ramus of the mandible at its medial surface
and the maxillary tuberosity at the mucogingival
junction adjacent to the maxillary second or
third molar if present. The patient is required
to remain seated and semireclined, with support
of the head, neck, and shoulders. The index
nger or thumb is used to separate the cheek
laterally to aord improved visualization of the

633

zone, positioning the needle parallel to the

occlusal plane of the upper jaw. In this position,
the needle is advanced into the soft tissue to
a depth of 25 mm from the maxillary tuberosity,
where the anesthetic solution is instilled in the
mesial zone of the pterygomandibular space
near the main branches of the inferior alveolar
nerve. Negative blood aspiration is conrmed,
and the anesthetic solution is injected at a rate
of 1.8 mL in approximately 60 seconds.
Comparison of the three techniques
Chances of intravascular injection are the
lowest with the Akinosi technique, because only
a few blood vessels are located in the injection
zone [14]. The Gow-Gates technique was the most
eective in inducing anesthesia, whereas the
Akinosi technique seemed to be most acceptable
among patients [15].
Complications
1. Intravascular injection is a potential complication because of the proximity of the pterygoid plexus of veins, the maxillary artery,
and its branch, the middle meningeal artery.
2. If the needle is inserted too deep, the superior
constrictor muscle can be pierced, resulting in
entry into the pharynx.
3. Temporary complete paralysis of cranial
nerves III, IV, and VI has been reported after
a Gow-Gates injection [16]. To avoid such
complication, it is important to take the following precautions when using the GowGates technique: because of the proximity
of the internal maxillary artery (accessory
and middle meningeal) and the pterygoid
plexus of veins, and because the anesthetic
is injected quickly, it is paramount that careful aspiration be performed before administration of the local anesthetic; injections
should be on or within 1 mm to 2 mm of
the condylar neck.
Glossopharyngeal nerve block
Anatomy
The glossopharyngeal nerve is the ninth cranial
nerve. It contains motor and sensory bers. It
arises from the cranial part of the medulla and
exits from the cranial cavity through the intermediate compartment of the jugular foramen. The
nerve then runs between the internal carotid
artery and internal jugular vein; after that point,

634

DATTA & PAI

it swings around the stylopharyngeus muscle toward the pharynx and the tongue. During its
course, it supplies sensory bers to the middle ear,
posterior third of the tongue, and the pharynx. It
also innervates the carotid sinus and the carotid
body. The nerve is in close proximity to the vagus
nerve, accessory nerve, and sympathetic trunk.
Parasympathetic bers pass via the glossopharyngeal nerve to the otic ganglion. Postganglionic
bers from the ganglion carry secretory bers to
the parotid gland.
Technique
Extraoral approach. Appropriate monitoring and
intravenous access are required before proceeding
with the block. The patient is positioned supine
with the head turned to the side opposite the
block. The lateral face and portion of the neck
below the ear are prepared in sterile manner. The
angles of the mandible and the mastoid process
are marked. The point midway between these two
landmarks inferior to the ear corresponds to the
position of the styloid process of the temporal
bone. The skin overlying the styloid process is
inltrated with local anesthetic. A 22-gauge, 3inch needle is then inserted perpendicular to the
skin and advanced. The styloid process should be
encountered within 3 cm. After contact is made,
the needle is withdrawn and walked o the styloid
process posteriorly. This corresponds to the location of the glossopharyngeal nerve as it curves
around the stylopharyngeus muscle (Fig. 6). Local
anesthetic (23 mL) is injected after negative
aspiration.
Intraoral approach. The patient opens the mouth
widely, and the tongue is retracted downward
with a tongue depressor. A 22-gauge, 3-cm spinal
needle with its tip bent is inserted through the
mucosa at the lower lateral portion of the
posterior tonsillar pillar. The needle is advanced
approximately 0.5 cm. After careful aspiration,
local anesthetic is injected.
Complications
1. Intravascular injection into the internal carotid artery or internal jugular vein is a potential risk. Injection into the carotid artery can
result in seizures and possible cardiovascular
collapse.
2. Hematoma from trauma to these vessels can
occur.
3. Proximity to the vagus and accessory nerves
can result in blockade of these nerves as

Fig. 6. Glossopharyngeal nerve block. (Upper panel)

Needle is inserted at a point midway between the mastoid process and the angle of the mandible. (Lower
panel) Needle is inserted at a right angle to the skin.
At a depth of 2 to 3 cm, the styloid process should be
contacted (if present). The needle is then walked posteriorly o the styloid process. Local anesthetic deposited at
this point should block glossopharyngeal, accessory, and
vagus nerves. Note the proximity of the internal carotid
artery and the internal jugular vein. (Adapted from Murphy TM. Somatic blockade of head and neck. In:
Cousins MJ, Bridenbaugh PO, editors. Neural blockade
in clinical anesthesia and management of pain. 3rd
edition. Philadelphia: Lippincott Raven; 1998. p. 503;
with permission.)

well. Inadvertent blockade of the hypoglossal

and spinal accessory nerves during a glossopharyngeal nerve block results in weakness
of tongue and trapezius muscle.
4. Bilateral blockade of the glossopharyngeal
nerves can result in total pharyngeal paralysis
with an associated risk of aspiration.

INTERVENTIONAL APPROACHES TO ORAL CANCER PAIN MANAGEMENT

Techniques of blockade of relevant ganglia

in cancer pain of head and neck
Sphenopalatine ganglion block and neurolysis
The sphenopalatine ganglion (SPG) block has
been used sporadically in management of pain of
oral cancer. Prasanna and Murthy [17] reported
immediate short-term pain relief with intranasal
SPG blockade in 10 patients with intractable
pain from cancer of the tongue and the oor of
the mouth.
Anatomy
The SPG is the largest group of neurons
outside the cranial cavity (Fig. 7). It lies in the
pterygopalatine fossa and resembles a vase on
lateral uoroscopic view. The pterygopalatine
fossa is bordered anteriorly by the posterior wall
of the maxillary sinus, posteriorly by the medial
plate of the pterygoid process, medially by the
perpendicular plate of the palatine bone, superiorly by the sphenoid sinus, and laterally by communication with the infratemporal fossa [18].
The foramen rotundum, through which the maxillary branch of the trigeminal nerve passes, is located on the superolateral aspect of the
pterygopalatine fossa. The opening to the pterygoid canal, which houses the vidian nerve, is located in the inferomedial portion of the fossa.

Fig. 7. Anatomy of the SPG and its immediate connections. (From Raj PP, Lou L, Erdine S, et al. Radiographic imaging for regional anesthesia and pain
management. New York: Churchill Livingston; 2003.
p. 66; with permission.)

635

The ganglion within the fossa is located posterior

to the middle turbinate of the nose and lies a few
millimeters deep to the lateral nasal mucosa. The
maxillary artery is also located in the pterygopalatine fossa. The SPG is suspended from the maxillary branch of the trigeminal nerve at the
pterygopalatine fossa via the pterygopalatine
nerves and lies medial to the maxillary branch
when viewed in the sagittal plane. Posteriorly, it
is connected to the vidian nerve, which is formed
by the greater petrosal and deep petrosal nerves.
Caudally, the ganglion is in direct connection
with the greater and lesser palatine nerves. The
SPG has autonomic, sensory, and motor components. The sensory bers arise from the maxillary
nerve; pass through the SPG; and are distributed
to the nasal membranes, soft palate, and parts
of the pharynx. The autonomic innervation
includes sympathetic and parasympathetic components. The sympathetic component begins
with preganglionic sympathetic bers originating
in the upper thoracic spinal cord, forming the
white rami communicantes, and coursing through
the sympathetic ganglion, where the preganglionic
bers synapse with the postganglionic ones. The
postganglionic bers then join the carotid nerves
before branching o and traveling through the
deep petrosal and vidian nerves. The postganglionic sympathetic nerves continue their path
through the SPG on their way to the lacrimal
gland and the nasal and palatine mucosa. The
parasympathetic component has its preganglionic
origin in the superior salivatory nucleus and then
travels through a portion of the facial nerve before
forming the greater petrosal nerve. The greater petrosal nerve, in turn, joins the deep petrosal nerve
to form the vidian nerve, which ends in the SPG.
Technique
The SPG is situated in the pterygopalatine
fossa, which can be reached by inserting a needle
through the mandibular notch and pterygomaxillary ssure. The pterygomaxillary ssure, which
forms the gateway to the pterygopalatine fossa,
can be visualized on a lateral radiograph of the
skull as an elongated wedge, which lies just below
the anterior part of the sphenoid sinus (Fig. 8).
The patient should lie supine. A lateral uoroscopic image is obtained so as to have total superimposition of both rami of the mandible. The
pterygomaxillary ssure is then identied just below the anterior part of the sphenoid sinus and
just posterior to the maxillary sinus (see Fig. 8).
The index nger is used to palpate the lower

636

DATTA & PAI

Fig. 8. (A) Lateral-oblique uoroscopic view of skull to show the pterygomaxillary ssure. (B) Relations of the pterygomaxillary ssure and the pterygopalatine fossa. (Adapted from Gauci CA. Sphenopalatine ganglion RF/PRF. In:
Gauci CA. Manual of RF techniques. Amsterdam, The Netherlands: FlivoPress, 2004. p. 75; with permission.)

margin of the zygomatic arch to identify the mandibular notch (Fig. 9). This is the entry point. A
3.5-inch, 22-gauge spinal needle is introduced
and directed slightly cranially and ventrally and
is then advanced cranially and ventrally through
the pterygomaxillary ssure and into the pterygopalatine fossa. A give is usually felt at approximately 4 cm. An anteroposterior view of the skull
is then obtained, and the needle is advanced until
the tip comes to lie adjacent to the lateral wall of
the nose (Fig. 10). Care is taken to avoid advancing the needle through the lateral nasal wall. If

Fig. 9. Identication of the mandibular notch. (Adapted

from Gauci CA. Sphenopalatine ganglion RF/PRF. In:
Gauci CA. Manual of RF techniques. Amsterdam, The
Netherlands: FlivoPress, 2004. p. 76; with permission.)

paresthesias are elicited in the hard palate (stimulation of greater and lesser palatine nerves), the
needle is anterior and lateral and should be redirected in a more posterior and medial direction. If
paresthesias are elicited in the teeth, the maxillary
branch of the trigeminal nerve is being stimulated
and the needle must be directed more caudal.
Trigeminal ganglion block
The patient is placed supine on the table with
the head in an extended position. The direction of

Fig. 10. Needle inserted through the pterygopalatine

fossa. (Adapted from Gauci CA. Sphenopalatine ganglion
RF/PRF. In: Gauci CA. Manual of RF techniques.
Amsterdam, The Netherlands: FlivoPress, 2004. p. 77;
with permission.)

INTERVENTIONAL APPROACHES TO ORAL CANCER PAIN MANAGEMENT

the needle is towards the pupil when one looks

from the front and midpoint of the zygomatic arch
when one looks from the side. A nger may be
placed in the mouth. This helps guide the needle
and prevents penetration of the oral mucosa.
There is a denite risk of meningitis if the needle
enters the oral mucosa. The direction of the needle
should be veried under uoroscopy in submental,
lateral and PA views. In many patients it is possible
to see the foramen ovale. When the foramen ovale
is seen, the needle is directed in a coaxial view
down the foramen ovale. When the needle enters
the foramen ovale, the uoroscope is turned laterally. The lateral view is important to verify the
depth of the needle inside the Meckels cave. For
conrming that the pain generator is the trigeminal
ganglion, after negative aspiration, up to 1 mL of
local anesthetic is injected. The patient should have
pain relief if the pain generator is present. The
physician should monitor that the solution has not
entered the cranial cerebrospinal uid. The brain
stem function should be evaluated by asking for
complaints of bilateral headache or fourth or sixth
nerve palsy or if papillary changes occur [19].
Stellate ganglion block
Anatomy
The cervical sympathetic chain is composed of
the superior, middle, and inferior cervical ganglia.
Often, the inferior cervical ganglion is fused with
the rst thoracic ganglion, forming the stellate
ganglion, which commonly measures 2.5 cm long,
1 cm wide, and 0.5 cm thick. It is usually located
in front of the neck of the rst rib and extends to
the interspace between the seventh cervical vertebra and the rst thoracic ganglion, which rests
over the neck of the rst rib. The stellate ganglion
lies medial to the scalene muscles, lateral to the
longus colli muscle, anterior to the transverse
process and prevertebral fascia, and superior to
the subclavian artery and posterior aspect of the
pleura. The initial portion of the vertebral artery
lies anterior to the stellate ganglion. Preganglionic
sympathetic bers originate from cell bodies in the
anterolateral column of the spinal cord. Nerves
supplying the head and neck arise from the rst
and second thoracic spinal segments. The preganglionic axons leave the T1 and T2 ventral
roots; pass through the white rami communicantes; join the sympathetic chain; and ultimately
synapse at the inferior (stellate), middle, or
superior cervical ganglion. Postganglionic sympathetic bers pass the gray rami and join the

637

cervical or upper cervical plexus. Most of the

sympathetic bers for the head and neck travel
along the common and then internal or external
carotid artery. Some of the bers leave the stellate
ganglion, form the vertebral plexus, and innervate
cranial structures supplied by the vertebral plexus,
however. Thus, blockade of the sympathetic innervation of the head and neck must incorporate
inclusion of the stellate ganglion.
Techniques of stellate ganglion block
Multiple techniques have previously been described [2023]. The anterior paratracheal technique is performed with the patient positioned
supine and the needle inserted vertically at 90 between the cricoid cartilage and the carotid artery
to touch the anterior tubercle on the transverse
process of C6. Careful aspiration before injection
is mandatory, because injection of only a small
volume of local anesthetic into the vertebral artery
can lead to immediate convulsion, blindness, and
loss of consciousness [24]. This procedure is usually performed blindly and requires a high volume
to block the stellate ganglion adequately. This
may lead to blockage of adjacent nerve structures
as well as possible complications. This has led to
the search for a technique to overcome these issues. Currently, a new technique using uoroscopy is becoming popular [25]. This is our
current method of choice. The patient is placed
supine with the neck slightly extended, and the
head is rotated slightly to the opposite side to be
blocked. The uoroscopy beam is directed in an
anteroposterior direction until the C5-C6 disc is
well visualized. This usually requires caudocranial
angulation of the uoroscope. The C-arm is then
rotated obliquely, ipsilateral to the side where
blockade is desired. The rotation must occur to allow adequate visualization of the neural foramina.
A skin weal is raised at the surface point where the
junction of the uncinate process and the vertebral
body is seen on the uoroscope. Under real-time
imaging, a single pass is made with a 25-gauge spinal needle to contact bone at this point. Care
should be exercised to avoid passage of the needle
toward the neural foramina and the thecal sac,
which are located posteriorly; the disc, which is
located cephalad; and the esophagus, which is
located medial to the ultimate target point. In its
nal position, the needle tip comes to rest in the
junction between the uncinate process and the vertebral body. The stylet is removed, the extension
set is attached, and radiopaque contrast (12
mL) is injected to visualize the longus colli muscle.

638

DATTA & PAI

The syringe containing the contrast is exchanged

for the one that contains the local anesthetic.
After ensuring that negative aspiration is performed, a 0.5-mL test dose is injected to rule
out intravascular injection into the vertebral artery. This is followed by slow injection of local
anesthetic at a rate of 3 to 5 mL into the ganglion. Advantages of this technique include eliminating pushing or pulling the vascular system
out of the way; eliminating pressure on the C6
tubercle, which can be painful; minimizing chances of injury to the esophagus or recurrent laryngeal nerve or attributable to intravascular
injection; and reducing the volume of local anesthetics needed to cover the lower cervical
through upper thoracic areas. A representative
block is demonstrated in Fig. 11.

Central neuraxial techniques for management

of oral cancer pain
Intraventricular opiates
The therapeutic basis underlying local administration into the intraventricular CSF is based on
studies that have demonstrated a high concentration of opiate receptors in the walls of the third
ventricle and the periaqueductal gray matter. The
use of intraventricular opiates should be based on
strict criteria. The therapy should be restricted to
patients with terminal head and neck cancer in
whom oral narcotic therapy has failed. In general,
patients with bilateral midline or diuse pain that
is not amenable to percutaneous or destructive
procedures are the best candidates for this therapy. Intraventricular opiates are generally administered as a daily bolus injection through an
implanted ventricular catheter connected to a subcutaneous reservoir. A ventricular catheter is
placed into the lateral ventricle in a standard
fashion and positioned as close as possible to the
foramen of Monro or, if possible, is threaded into
the third ventricle. The catheter is connected to
a subcutaneous access port. The port is accessed
percutaneously with a 25-gauge needle, and several milliliters of CSF are gently aspirated; the
CSF is kept to ush the catheter after instillation of the drug. It is critical that meticulous
sterile technique be used each time the reservoir
is accessed so as to avoid contamination. Preservative-free morphine is injected into the reservoir and gently ushed using the CSF that is
withdrawn. The onset of analgesia occurs between
15 and 30 minutes after administration, with the

Fig. 11. Stellate ganglion block performed for facial

pain using uoroscopy. The needle has been placed on
the body of C7 at the junction of the uncinate and
body of the C7 vertebra, and dye outlines spread above
and below the stellate ganglion. L, left.

peak analgesic eect being reached by 60 minutes.

The mean range of analgesia is approximately 28
hours (range: 1270 hours) [26,27]. Lazorthes and
colleagues [26] reported their experience with intraventricular opioid therapy in 82 patients with
cancer pain. The average follow-up was 9 weeks.
The initial daily dose of morphine was 0.3 mg
(range: 0.12 mg). All patients reported complete
analgesia within 1 hour of injection of drug. The
end point of the study was death attributable to
underlying cancer. Final assessment of the patients in this series over an average of 2.5 months
indicated excellent or good analgesia in 66 patients (98%), moderate pain control in 14 patients
(17%), and failure of treatment in 2 patients. As is
the case with spinal opiates, dose escalation was
common. For patients who survived longer than
60 days, there was a 10-fold increase in the average nal dose of morphine compared with the initial dose. The side eects of therapy were minor
for the most part and included nausea, vomiting,
headache, pruritus, urinary retention, dizziness,
and disorientation. Major side eects, including
respiratory depression, occurred in only 3 (4%)
of the 82 patients and were immediately reversed
by naloxone, with only minimal eects on the induced analgesia.
Intrathecal pump
Applengren and coworkers [28] described a series of patients with malignant or nonmalignant

INTERVENTIONAL APPROACHES TO ORAL CANCER PAIN MANAGEMENT

chronic head and neck pain that was refractory to

conventional analgesic therapies. In these patients, pain control was inadequate despite aggressive treatment, including surgery, radiation
therapy, or chemotherapy directed at neoplasms
and pain relief; systemic analgesic therapy with
nonsteroidal anti-inammatory drugs, anticonvulsants, antidepressants, phenothiazines, benzodiazepines, and oral and parenteral opioids; cranial
nerve neurolysis with such techniques as diathermic coagulation of the gasserian ganglion or
through injection of glycerol; peripheral neurolysis with such techniques as epidural phenol injections; local anesthetic-induced neural blockade
with such techniques as continuous epidural analgesia or stellate ganglion blockade; insertion of
a dorsal column stimulator at a cervical level;
transcutaneous nerve stimulation; acupuncture;
and therapy with hormones or steroids. Patients
had inadequate pain relief with these regimens
or experienced intolerable side eects from the
systemic opioid therapy. Numerous invasive techniques have been used in an attempt to treat similar patients with refractory head and neck pain.
Previous therapies, such as intracisternal [29] or
intraventricular [30] injection of morphine, neurolysis of the gasserian ganglion or other cranial
nerves, rhizotomy, cordotomy, or thalamotomy,
often provided inadequate pain relief or produced
unacceptable side eects, however. For example,
in one study, control of pain was attempted with
injection of phenol into the cisterna magna [31].
Eighteen percent of patients had long-lasting, disabling neurologic decits, and 71% had less severe
complications. More extensive neoplastic disease
causing severe pain has been treated by performing
a posterior craniotomy and sectioning the roots of
cranial nerves VII, IX, and X and the sensory roots
of the upper cervical nerves. Many patients cannot
tolerate a major operation, however, and sectioning of the glossopharyngeal and vagus nerves produces paralysis of the pharyngeal and laryngeal
muscles. In contrast to the poor results with these
therapies, Appelgrens group [28] reported on 13
patients with complex refractory pain who received
continuous intracisternal or high cervical subarachnoid infusions of bupivacaine as a method
to control pain in the head, face, mouth, neck,
and upper extremities. For most patients, infusion
of bupivacaine provided satisfactory pain relief, decreased systemic opioid consumption, and improved nocturnal and overall sleep patterns.
In patients with diuse upper neck and head
pain, the source of pain involves multiple nerves.

639

Cranial nerves and peripheral somatic nerves

often coexist to provide nociceptive aerent information. In addition to severe sharp somatic
pain, burning discomfort may result from nociceptive pathways within the sympathetic chain.
The multiplicity of potential nociceptive pathways
can make individual nerve block therapy relatively
ineective [32]. Similarly, local invasion of tumor
can make peripheral neural blockade impractical.
The intrathecal approach can be performed at
a site relatively distant from the site of pain. The
ability of this technique to inhibit multiple nociceptive pathways is probably responsible for the
improved analgesic ecacy observed in a group of
patients for whom previous techniques had been
relatively ineective. Another major advantage of
this approach is that an inpatient titration trial

Fig. 12. Epidural trial for permanent intrathecal pump

implant. The patient has squamous cell cancer of the
oor of the mouth and has undergone four resections.
He is an example of the typical patient with oral cancer
pain. The patient has failed conventional therapy, including megadoses of narcotics (transdermal fentanyl
patch along with immediate-release morphine). He has
multiple sources of pain, including pain from the oral
cancer (with multiple episodes of breakthrough pain),
pain in the shoulder, and pain in the upper abdomen
and lower back from degenerative disc disease. An epidural trial utilizing a tunneled epidural catheter was tried
for 3 days using a local anesthetic (bupivacaine) and narcotic (morphine). The catheter tip was placed at a high
thoracic level (T2) to help with the oral cancer pain
and the shoulder pain. The patient achieved successful
pain control with the trial. After the trial, he was implanted with an intrathecal pump administering the
same medication into his CSF. L, left.

640

DATTA & PAI

Summary

Fig. 13. Close-up view of a programmable implantable

Medtronic Synchromed II intrathecal pump along with
a silastic intrathecal catheter. The catheter is implanted
into the intrathecal space and connected to the pump
through a subcutaneous tunnel. The intrathecal pump
pocket is usually located on the anterior abdominal wall.
Medications (narcotic alone or in combination with local
anesthetic/clonidine/baclofen) are relled through the
access port located in the center of the pump. (Reprinted
with the permission of Medtronic, Inc., 2006.)

with narcotic alone or narcotic combined with

local anesthetic may be performed via an indwelling catheter placed into the epidural or
intrathecal space. A representative patient is
demonstrated in Fig. 12. The trial is expected to
demonstrate whether this modality is eective
for the patient as well as to determine the necessary dosage for placement of the intrathecal
pump. Our preferred method of placing an intrathecal pump system consists of using uoroscopy
to place the catheter into the intrathecal space.
The catheter is silastic and can be maneuvered
into a position that is judged to cover the major
area of pain. For head and neck pain, the best location of the catheter seems to be at the level of
the C1 or C2 vertebra. Patients with shoulder
and arm pain are better served with a catheter
tip located at the C4 or C5 vertebra. The catheter
is then tunneled and connected to an intrathecal
pump (usually the Medtronic Synchromed II;
Medtronic Corporation, Minneapolis, Minnesota), which is usually placed subcutaneously in
the patients abdomen (Fig. 13). The pump is extremely sophisticated and capable of complex programming to meet complex demands.

The management of pain of oral cancer is often

dicult. Causes of pain are complex. A proper
management protocol includes a multidisciplinary
approach, including surgery, radiation therapy, or
chemotherapy directed at neoplasms and pain
relief; systemic analgesic therapy with nonsteroidal anti-inammatory drugs, anticonvulsants, antidepressants, phenothiazines, benzodiazepines,
and oral and parenteral opioids; cranial nerve
neurolysis with such techniques as diathermic
coagulation of the gasserian ganglion or through
injection of glycerol; peripheral neurolysis
with such techniques as epidural phenol injections; local anesthetic-induced neural blockade
with such techniques continuous epidural analgesia or stellate ganglion blockade; insertion of
a dorsal column stimulator at a cervical level;
transcutaneous nerve stimulation; acupuncture;
and therapy with hormones or steroids. Intraventricular, intracisternal, or intrathecal administration of opioids with or without local anesthetics is
indicated for severe unmanageable pain.
References
[1] Bonica JJ. Pain caused by cancer of the head and
neck and other specic syndromes. In: Bonica JJ, editor. The management of pain. 2nd edition. Philadelphia: Lea & Febiger; 1990. p. 793811.
[2] Levy L, Baudoin A. Les injections profondes dans
led traitment de la neuralgie. Presse Med 1906;13:
108.
[3] Kohase H, Miyamoto T, Umino M. A new method
of continuous maxillary nerve block with an indwelling catheter. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 2002;94:1626.
[4] Iwade M, Fukuuchi A, Kawamata M, et al. [Management of severe pain after extended maxillectomy
in a patient with carcinoma of the maxillary sinus.].
Masui 1996;45:825 [in Japanese].
[5] Singh B, Srivastava SK, Dang R. Anatomic considerations in relation to the maxillary nerve block. Reg
Anesth Pain Med 2001;26:50711.
[6] Fujii A. [New technic for blocking the 2nd division
of the trigeminal nerve. Experimental and clinical
studies on suprazygomatic route.]. Masui 1976;
25:137082 [in Japanese].
[7] Okuda Y, Okuda K, Shinohara M, et al. Use of computed tomography for maxillary nerve block in the
treatment of trigeminal neuralgia. Reg Anesth Pain
Med 2000;25:4179.
[8] Friedman MJ, Hochman MN. The AMSA injection:
a new concept for local anesthesia of maxillary teeth
using a computer-controlled injection system. Quintessence Int 1998;29:297303.

INTERVENTIONAL APPROACHES TO ORAL CANCER PAIN MANAGEMENT

[9] Friedman MJ, Hochman MN. Using AMSA and PASA nerve blocks for esthetic restorative dentistry.
Gen Dent 2001;49:50611.
[10] Hawkins JM, Ibsen D. Maxillary nerve block: the
pterygopalatine canal approach. J Calif Dent Assoc
1998;26:65864.
[11] Goldenberg AS. Transient diplopia as a result of
block injections. Mandibular and posterior superior
alveolar. NY State Dent J 1997;63:2931.
[12] Singh B, Srivastava SK, Dang R, et al. Anatomic
considerations in relation to the mandibular nerve
block. Reg Anesth 1993;18:1813.
[13] Akinosi JO. A new approach to the mandibular
nerve block. Br J Oral Surg 1977;15:837.
[14] Martinez Gonzalez JM, Benito Pena B, Fernandez
Caliz F, et al. A comparative study of direct mandibular nerve block and the Akinosi technique. Med
Oral 2003;8:1439.
[15] Cruz EV, Quengua JB, Gutierrez IL, et al. A comparative study: classical, Akinosi, and Gow-Gates
techniques of mandibular nerve block. J Philipp
Dent Assoc 1994;46:139.
[16] Fish LR, McIntire DN, Johnson L. Temporary paralysis of cranial nerves III, IV, and VI after
a Gow-Gates injection. J Am Dent Assoc 1989;
119:1278, 130 [discussion: 129].
[17] Prasanna A, Murthy S. Sphenopalatine ganglion
block and pain of cancer. J Pain Symptom Manage
1993;8:125.
[18] Salar G, Ori C, Iob I. Percutaneous thermocoagulation for sphenopalatine ganglion neuralgia. Acta
Neurochir (Wien) 1987;84:248.
[19] Trigeminal ganglion block and neurolysis. In:
Raj PP, Lou L, Erdine S, et al. Radiographic imaging for regional anesthesia and pain management.
New York: Churchill Livingston; 2003. p. 3748.
[20] Moore DC, Bridenbaugh LD Jr. The anterior approach to the stellate ganglion. JAMA 1956;160:
15862.

641

[21] Bryce-Smith R. Stellate ganglion block. Anesthesia

1952;7:1546.
[22] Davis RM. Stellate ganglion blockda new approach. Anaesthesia 1952;7:1513.
[23] Carrib G, Litwiller R. Stellate ganglion block.
Anesth Analg 1975;54:56770.
[24] Szeinfeld M, Laurencio M, Pallares VS. Total reversible blindness following stellate ganglion block.
Anesth Analg 1981;60:68990.
[25] Abdi S, Zhou U, Patel N, et al. A new and easy technique to block the stellate ganglion. Pain Physician
2004;7:32731.
[26] Lazorthes Y, Sallerin B, Verdi JC. Intracerebroventricular administration of morphine for control of irreducible cancer pain. In: Gildenburg PL,
Tasker RR, editors. Textbook of stereotactic and
functional neurosurgery. New York: McGraw-Hill;
1998. p. 147782.
[27] Osenbach RK. Nontrigeminal craniofacial pain syndromes. In: Follett KA, editor. Neurosurgical pain
management. Philadelphia: Elsevier; 2004. p. 8498.
[28] Appelgren L, Janson M, Nitescu P, et al. Continuous intracisternal and high cervical intrathecal bupivacaine analgesia in refractory head and neck pain.
Anesthesiology 1996;84:25672.
[29] Schoeer PF, Haberer JP, Monteillard CM, et al.
Morphine injections into the cisterna magna for intractable pain in cancer patients [abstract]. Anesthesiology 1987;67:A246.
[30] Lobato RD, Madrid JL, Fatela LV, et al. Intraventricular morphine for control of pain in terminal
cancer patients. J Neurosurg 1983;59:62733.
[31] Bortiluzzi M, Marini G. Phenol injection into the
cisterna magna for relief of advanced intractable
cancer pain in the faciocephalic area. J Neurosurg
Sci 1986;30:16776.
[32] Carpenter RI, Rauck RL. Refractory head and neck
pain. A dicult problem and a new alternative therapy. Anesthesiology 1996;84:24952.

Oral Maxillofacial Surg Clin N Am 18 (2006) 643645

End of Life Issues in Oral Cancer

James Bridges, MDa,b,*, John Mulder, MDa,c
a

Alive Hospice, 1718 Patterson Street, Nashville, TN 37203, USA

b
Vanderbilt University School of Medicine, Nashville, TN, USA
c
Pain and Symptom Management Program, Vanderbilt-Ingram Cancer Center,
Vanderbilt University School of Medicine, Nashville, TN, USA

Despite advances in the prevention of head and

neck cancers, approximately 38,530 new cases of
cancer involving the oral cavity, pharynx, and
larynx are diagnosed each year. Although there
have been many advances in the treatment of head
and neck cancer, including improvements in the
quality of life, an estimated 10,270 people died of
the disease in 2004. Curative and palliative
treatments for head and neck cancers are often
begun simultaneously. It is not an either/or
phenomenon. Prescribing analgesics for pain
while chemotherapy or radiation is simultaneously being performed is an example of such
subsequent measures. With the progression of
disease, the likelihood of a curative response
concomitantly diminishes. At this point in the
illness trajectory, palliative, symptomatic treatments become the primary goal. It is important to
note that many dierent treatment modalities are
available to help relieve uncontrolled symptoms of
head and neck cancer. Eective, signicant relief
of symptoms and suering should not be postponed until the last few days of life. Initiating the
conversation about prognosis, symptom management, and issues relating to the end of life can be
dicult and time consuming. The following list is
shows a framework for making medical decisions
in terminal head and neck cancer patients (Box 1).

* Corresponding author. Alive Hospice,

Patterson Street, Nashville, TN 37203.
E-mail address: jbridges@alivehospice.org
(J. Bridges).

1718

Prognostication: be sure the patient and family

understand the diagnosis and prognosis
Recognizing and providing a correct prognosis
is often one of the most dicult objectives that
faces physicians charged with the palliative care of
a patient. Often we are inadequately trained and
feel uncomfortable sharing devastating news with
patients and their families. We also may want to
protect the patient and family from the dicult
reality of their disease. Patients and families are
appreciative when a physician is honest with them
about the extent of their disease and prognosis,
however. They expect the honesty and candor that
is the foundation of a relationship built on trust.
Knowing the extent of disease and likelihood of
treatment success can help guide patients and
families as they plan for the future. This honesty
also facilitates the often painful, dicult transition from a curative treatment plan to one that is
palliative in nature.
The existence of recurrent disease or development of metastasis in head and neck cancer
generally responds poorly to curative treatment.
People with the same tumor burden and cancer
stage may live for dierent lengths of time with or
without treatment, however. One scale currently
in use to help with prognostication is the palliative
performance scale (Table 1). This scale was developed to describe a patients functional level. It
also helps track a patients decline and has some
prognostic value. For example, a patient with
a palliative performance scale of 50% to 60%
with recurrent cancer or widespread metastasis
is, in many instances, likely nearing the last 6
months of life. Similarly, a value of 30% indicates
that a patient likely has 4 to 6 weeks of life. As

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.coms.2006.06.006

oralmaxsurgery.theclinics.com

644

BRIDGES & MULDER

Set new treatment goals and assess new symptoms

Box 1. Framework for making medical
decisions in terminal cancer patients
Be sure the patient and family
understand the diagnosis and
prognosis
Set new treatment goals
Assess new symptoms
Discuss treatment options for the new
symptom
Make a recommendation
Never say There is nothing to be done
Do not abandon patients

a patient declines further and approaches a palliative performance scale score of 10%, he or she
usually has hours to days to live. A simple question to ask is, Would I be surprised if this patient
was dead in 6 months to a year? When a patient
is left with such a short life expectancy, patients
and families are faced with the dicult task of
shifting the focus of care from the curative realm
to the palliative. New goals of treatment clearly
reect this transition.

When the treatment plan changes from curative to palliative, the primary goal is to honor
a patients wishes. To establish goals for treatment
successfully, physicians must ask patients what is
of primary importance to them. Most patients
express a desire to be kept comfortable, whereas
others may have a specic goal, such as going on
a trip or visiting with a relative. Patients must be
assured and reassured that symptoms always will
be addressed. Patients must be asked about
symptoms at each patient encounter. It is not
uncommon for a patient to feel that he or she is
supposed to experience pain or discomfort in
a state of terminal disease. Such feelings may
prevent a patient from discussing specic diculties and discomforts that he or she may be
experiencing. Common diculties experienced
by patients with terminal head and neck cancer
include airway compromise, somatic and neuropathic pain, wounds and stulas, communication
diculties, dysphagia, excessive oral and airway
secretions, and depression with or without anxiety. Assuring a patient and family that all
problems will be addressed appropriately and
that the patient will not suer often helps to
alleviate the fears of patients and their families.

Table 1
Palliative performance scale
PPS level

Ambulation

100

Full

90

Full

80

Full

70

Reduced

60

Reduced

50

Mainly sit/lie

40

Mainly in bed

30

Bed bound

20

Bed bound

10

Bed bound

Death

Activity and evidence

of disease
Normal activity and work;
no evidence of disease
Normal activity and work;
some evidence of disease
Normal activity with eort;
some evidence of disease
Unable to do normal
job/work; signicant disease
Unable to hobby/housework;
signicant disease
Unable to do any work;
extensive disease
Unable to do most activity;
extensive disease
Unable to do any activity;
extensive disease
Unable to do any activity;
extensive disease
Unable to do any activity;
extensive disease

Self-care

Intake

Conscious level

Full

Normal

Full

Full

Normal

Full

Full

Full

Total care

Normal or
reduced
Normal or
reduced
Normal or
reduced
Normal or
reduced
Normal or
reduced
Normal or
reduced
Minimal to sips

Total care

Mouth care only

Full
Occasional
assistance
Considerable
assistance
Mainly assistance
Total care

Full
Full or
confusion
Full or confusion
Full or drowsy;
/
confusion
Full or drowsy;
/
confusion
Full or drowsy;
/
confusion
Drowsy or coma;
/
confusion

END OF LIFE ISSUES IN ORAL CANCER

645

Discuss treatment options and make a treatment

recommendation

Do not abandon patients if their disease is no

longer curable

Many treatment options are available for

symptoms caused by head and neck cancer,
including invasive and noninvasive procedures.
A patients preference should be honored whenever possible. For example, a patient who experiences increasing dysphagia and progressive weight
loss often poses a unique challenge for clinicians.
Faced with this predicament, a physician frequently places a percutaneous gastrostomy feeding tube to provide adequate nutrition. This
approach may not, however, be consistent with
the wishes of every patient. Providing education
regarding a patients individual disease process
and other options, such as changing the diet,
spoon feeding, and providing good oral care, are
other alternatives. A patient may not wish to have
his or her life prolonged by a feeding tube.
Thoughtful discussions with patients through the
use of deep listening and eective communication
skills are necessary when determining the most
appropriate treatment plan for a patient. If, for
example, a patient wishes to proceed with the
placement of an enteral feeding tube, he or she
should be told that this intervention may need to
be weaned or discontinued as the severity of the
disease worsens or other problems arise. Patients
and families facing terminal head and neck cancer
may be confused by the treatment options.
Surgeons must understand fully the wishes of
their patients to provide the most appropriate
recommendations.

Patients may perceive a message of abandonment if the head and neck surgeon delegates the
care of terminal patients to others. When and if
a patient is being referred to a palliative care team
or hospice, the patient and family must be assured
that the surgeon will be with them through the
end of life. Patients often receive an emotional
benet from seeing their surgeon even after
curative treatments have been abandoned. If the
care of other specialists is needed, it is important
for the primary clinician to emphasize to the
patient and family that he or she still cares about
them and is involving a specialist to continue to
provide the best care available. When a surgeon
remains involved in a patients care, he or she has
the opportunity to serve as a resource to physicians providing palliative measures who may not
share their exact wealth of knowledge.

Never say There is nothing we can do

Patients are often worried about uncontrolled
pain and other symptoms at the end of life. It is
never appropriate to tell patients and families
there is nothing that can be done. Patients must
be assured that although there may be nothing
that can be done to cure their cancer, something
always can be done to help with the symptoms
caused by the cancer. This reassurance alone can
alleviate much anxiety on behalf of patients and
families. Although they may not always share this
concern openly, many patients worry deeply
about the pain and suering they may face at
the end of life.

Discussion
The head and neck surgeons role is important
throughout the entire illness trajectory of a patient
who has cancer. Even if patients with advanced
head and neck cancer are referred to hospice or
palliative care, they still present with major
problems, such as pharyngocutaneous stulas,
dysphagia, and compromised airways that head
and neck surgeons are uniquely qualied to
palliate. Continuing involvement in patient care
is rewarding for patients, families, and their
surgeons. Successfully alleviating a terminal patients suering also can bring an enormous sense
of professional accomplishment. Patients and
families are just as grateful for a well-managed
death as they are for treatments that result in
a cure.

Further readings
Anderson F, Downing GM, Hill J. Palliative Performance Scale (PPS): a new tool. J Palliat Care 1996;
12(1):511.
American Cancer Society. Cancer facts and gures, 2004.
Atlanta (GA): American Cancer Society; 2004.
Thompson RA, Krouse R. Terminal care in head and
neck cancer. J Am Coll Surg 2004;5:83742.

Oral Maxillofacial Surg Clin N Am 18 (2006) 647654

Index
Note: Page numbers of article titles are in boldface type.

A
Access surgery, for oral cancer, 565571
maxillary approach, 565566
transmandibular approach, 566569
visor ap, 569570
Adjunct, diagnostic, for oral cancer and
precancerous lesions, 468472
brush cytology (biopsy), 468469
cytology, 468
tissue uorescence, 469471
toluidine blue, 471472
AJCC stage groupings, for oral cancers, 438
Anatomy, of head and neck, for interventional
pain management in oral cancer,
628630
Gasserian ganglion, 630631
mandibular nerve, 629630
maxillary nerve, 629
ophthalmic nerve, 629
trigeminal nerve, 628
Antiplatelet therapy, with reconstructive surgery
in oral cancer patients, 583
Antithrombotic therapy, with reconstructive
surgery in oral cancer patients, 583
Aspirin, prevention of thrombosis with
reconstructive surgery in oral cancer patients,
583584

B
Biochemoprevention, of oral cancer,
502503
Biopsy, in diagnosis of oral cancer and
precancerous lesions, 472473
sentinel lymph node, in staging of oral cancer,
547563
Bowman-Birk inhibitor, prevention of oral cancer
with, 503
Brush biopsy. See Brush cytology.

Brush cytology, in diagnosis of oral cancer and

precancerous lesions, 468469

C
Cancer pain. See also Pain management., of head
and neck, mechanisms of, 627628
Carbon dioxide laser, for oral leukoplakia,
426427
Carcinogenesis, of oral cancer, 493494
multifocal process, 493494
multistep process, 493
Central neuraxial techniques, for management of
oral cancer pain, 637638
intrathecal pump, 638639
intraventricular opiates, 637638
Cervical lymph nodes, assessment of, in relation
to oral cancer, 535536
Chemoprevention, as strategy for oral cancer
prevention, 499501
Chemoradiation therapy, for oral cancer,
605614
concomitant chemoradiation, 609610
induction therapy before surgery, 608609
overview of combined modality therapy,
605608
concurrent chemoradiation, 606607
future directions, 607608
induction chemotherapy, 605606
supportive care issues, 610611
Chemotherapy, for oor of mouth and tongue
cancer, 522
Combined modality therapy, for oral cancer,
overview of, 605608
Computed tomography, in initial evaluation for
oral cancer, 440441
Cryosurgery, for oral leukoplakia, 426
Curcumin, prevention of oral cancer with, 504

1042-3699/06/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/S1042-3699(06)00085-9

oralmaxsurgery.theclinics.com

648

INDEX

never say theres nothing we can do, 645

prognostication, 643644
set new treatment goals and assess new
symptoms, 644

Cyclo-oxygenase-2 inhibitors, as target for

prevention of oral cancer, 506
Cytochemical and molecular studies, in diagnosis
of oral cancer and precancerous lesions,
475478
abnormal DNA segregation, 476477
DNA alteration, 477478
nodular organizing regions, 475476
Cytology, in diagnosis of oral cancer and
precancerous lesions, 468
ne-needle aspiration, 468

D
Diagnosis, and detection of oral cancer,
456482
clinical features, 465468
precancerous lesions, 465467
squamous cell carcinoma, 467468
cytochemical and molecular studies,
475478
abnormal DNA segregation, 476477
DNA alteration, 477478
nodular organizing regions, 475476
diagnostic adjuncts, 468472
brush cytology (biopsy), 468469
cytology, 468
tissue uorescence, 469471
toluidine blue, 471472
diagnostic methods, 472475
ne-needle aspiration cytology,
473474
punch biopsy, 472
scalpel biopsy, 472473
sentinel node biopsy and cytology,
474475
future of, 478479
Dietary agents, prevention of oral cancer with,
504

Epidermal growth factor receptor, as target for

prevention of oral cancer, 505
Erythroplakia, oral, 427429
clinical aspects, 427428
denition and classication, 427
incidence and prevalence, 427
management, 428429
rate of malignant transformation, 428
Evaluation, patient, in oral cancer, 438442
history, 439
initial, 438439
metastatic evaluation, 441442
physical examination, 439440
radiographic evaluation, 440441
Extracapsular spread, in oor of mouth and oral
tongue cancer, assessment of risk of recurrence
and metastasis with, 526527

F
Field cancerization, in oral cancer, 486
Fine-needle aspiration cytology, in diagnosis of
oral cancer and precancerous lesions,
473474
Flaps. See Reconstruction techniques.
Floor of mouth cancer, 521531
assessment of risk for recurrence and
metastasis, 525527
treatment modalities, 521522
role of chemotherapy, 522
surgery and radiotherapy, 521522
Fluorescence, tissue, in diagnosis of oral cancer
and precancerous lesions, 469471

DNA alteration, in diagnosis of oral cancer and

precancerous lesions, 477478
DNA aneuploidy. See DNA segregation, abnormal.
DNA segregation, abnormal, in diagnosis of oral
cancer and precancerous lesions, 476477

E
End of life issues, in oral cancer, 643645
discuss pretreatment options and make
treatment recommendation, 645
do not abandon patients if disease is no
longer curable, 645

G
Ganglia, blockade of, in management of oral
cancer pain, 634637
sphenopalatine ganglion block and
neurolysis, 634636
stellate ganglion block, 636637
Garlic, prevention of oral cancer with, 505
Genomics, of oral cancer, 486488
Glossopharyngeal nerve block, in management of
oral cancer pain, 633634

649

INDEX

surgical excision, 426

H
H-ras gene, as target for prevention of oral cancer,
505
Heparin, prevention of thrombosis with
reconstructive surgery in oral cancer patients,
583
Histopathologic features, of oral cancer,
483485
evaluation of margins, 483484
invasive pattern, 485
perineural invasion, 485
tumor grading, 484
tumor size, 484
tumor thickness, 484485
Histopathologic grading, of oral cancers, 438
History, patient, in initial evaluation for oral
cancer, 439

I
Imaging, in initial evaluation for oral cancer,
440441
of oor of mouth and oral tongue cancer,
assessment of risk of recurrence and
metastasis with, 525
of oral cavity squamous cell carcinoma, by
region, 446454
buccal and alveolar ridge, 450452
oor of mouth, 447450
hard palate, 453454
lips, 446
oral tongue, 446447
retromolar region, 452453
choice of modality, 445446
lymphadenopathy, 454456
positron emission tomography, 456460
Induction chemotherapy, for oral cancer,
605606, 608609

K
Ketolorac, prevention of oral cancer with,
503504

Lichen planus, oral, 429430

classication, 429
malignant potential, 430
management, 429430
Lip, cancer of, TNM staging for, 436437
Locoregional recurrence, in oral squamous cell
carcinoma, 615625
considerations for salvage, 620621
distant recurrence, 619620
local recurrence, 616617
locoregional or regional recurrence,
617619
Loss of heterozygosity. See DNA alteration.
Lymph nodes, cervical, assessment of, in relation
to oral cancer, 535536
regional/cervical, TNM staging for oral
cancers with involvement of, 437
sentinel node biopsy and cytology in diagnosis
of oral cancer and precancerous lesions,
474475
Lymphatic mapping, in oor of mouth and oral
tongue cancer, assessment of risk of recurrence
and metastasis with, 525526

M
Magnetic resonance imaging, in initial evaluation
for oral cancer, 440441
Malignant transformation, of oral premalignant
lesions, rate of, 428, 430
Mandibular nerve block, in management of oral
cancer pain, 632633
Margins, surgical, in oral cancer, molecular
analysis of, 485486
Maxillary approach, for access in oral cancer
surgery, 565566
midfacial degloving approach, 566
posterior maxillary approach, 566
Weber-Fergusson maxillectomy incision,
565566

Laser, CO2, for oral leukoplakia, 426427

Maxillary nerve block, in management of oral

cancer pain, 630632

Leukoplakia, oral, 425427

CO2 laser, 426427
cryosurgery, 426
management, 425426
nonsurgical therapies, 427

Metastases, of oral cancer, evaluation for,

441442
TNM staging for, 437438
distant, 438
regional/cervical, 437

650

INDEX

Metastasis, of oor of mouth and oral tongue

cancer, assessment of risk of, 525527

Nodal involvement, regional/cervical, TNM

staging for oral cancers with, 437

Microarrays, in genomics of oral cancer, 486488

Nuclear factor kappa beta, as target for

prevention of oral cancer, 505506

Microvascular surgery, for postablative

reconstruction in oral cancer patients, 584590
end-to-end anastomosis, 586587
end-to-side anastomosis, 587588
ap salvage, 588590
recipient vessel selection, 585
vascular anastomosis, 585586
Molecular biology, and clinical behavior of oral
cancer, 483491
genomics, 486488
applications of microarrays, 486488
molecular alterations in, 485486
analysis of surgical margins, 485486
eld cancerization, 486
proteomics, 488489
traditional histopathologic features,
483485
evaluation of margins, 483484
invasive pattern, 485
perineural invasion, 485
tumor grading, 484
tumor size, 484
tumor thickness, 484485
Molecular studies, in diagnosis of oral cancer and
precancerous lesions, 475478
abnormal DNA segregation, 476477
DNA alteration, 477478
nodular organizing regions, 475476

N
Neck dissection, surgical management in oral
cancer, 533546
anatomy of cervical lymph nodes in relation
to oral cancer, 535536
classication of, 536542
functional, 536540
modied radical, 536
supraomohyoid, 540541
considerations for future management of,
544545
decision making on which to perform,
542544
history of neck dissection, 533535
Nerve blocks, in management of oral cancer pain,
630634
glossopharyngeal nerve block, 633634
mandibular nerve block, 632633
maxillary nerve block, 630632

Nucleolar organizing regions, in diagnosis of oral

cancer and precancerous lesions, 475476

O
Oral cancer, 425644
chemoradiation therapy for, 605614
concomitant chemoradiation, 609610
induction therapy before surgery, 608609
overview of combined modality therapy,
605608
concurrent chemoradiation, 606607
future directions, 607608
induction chemotherapy, 605606
supportive care issues, 610611
detection and diagnosis of, 456482
clinical features, 465468
precancerous lesions, 465467
squamous cell carcinoma, 467468
cytochemical and molecular studies,
475478
abnormal DNA segregation, 476477
DNA alteration, 477478
nodular organizing regions, 475476
diagnostic adjuncts, 468472
brush cytology (biopsy), 468469
cytology, 468
tissue uorescence, 469471
toluidine blue, 471472
diagnostic methods, 472475
ne-needle aspiration cytology, 473474
punch biopsy, 472
scalpel biopsy, 472473
sentinel node biopsy and cytology,
474475
future of, 478479
end of life issues in, 643645
discuss pretreatment options and make
treatment recommendation, 645
do not abandon patients if disease is no
longer curable, 645
never say theres nothing we can do, 645
prognostication, 643644
set new treatment goals and assess new
symptoms, 644
evaluation and staging, 435444
patient evaluation, 438442
history, 439
initial, 438439

INDEX

metastatic evaluation, 441442

physical examination, 439440
radiographic evaluation, 440441
staging, 436438
AJCC stage groupings, 438
distant metastases, 438
histopathologic grading, 438
lip and oral cavity, 437
R classication, 438
regional/cervical lymph node
involvement, 437
TNM system, principles and general
rules, 436437
oor of mouth and tongue cancer, 521531
assessment of risk for recurrence and
metastasis, 525527
treatment modalities, 521522
role of chemotherapy, 522
surgery and radiotherapy, 521522
imaging of, by region, 446454
buccal and alveolar ridge, 450452
oor of mouth, 447450
hard palate, 453454
lips, 446
oral tongue, 446447
retromolar region, 452453
choice of modality, 445446
lymphadenopathy, 454456
positron emission tomography, 456460
locoregional recurrence in squamous cell
carcinoma, 615625
considerations for salvage, 620621
distant recurrence, 619620
local recurrence, 616617
locoregional or regional recurrence,
617619
molecular biology and clinical behavior of,
483491
genomics, 486488
molecular alterations in, 485486
proteomics, 488489
pain management in, 625639
anatomy of head and neck, 628630
blockade of relevant ganglia, 634637
central neuraxial techniques, 637639
mechanisms of cancer pain of head and
neck, 627628
peripheral nerve blocks, 630634
premalignant lesions, 425433
erythroplakia, 427429
leukoplakia, 425427
lichen planus, 429430
prevention of, 493511

651

biochemoprevention, 502503
chemoprevention, 499501
dietary agents, 504505
garlic, 505
oral carcinogenesis, 493494
other agents, 503504
Bowman-Birk inhibitor, 503
curcumin, 504
ketolorac, 503504
sulindac, 503
potential new targets for chemoprevention,
505506
cyclo-oxygenase-2 inhibitors, 506
epidermal growth factor receptor, 505
H-ras gene, 505
nuclear factor beta kappa, 505506
p53 gene, 506
retinoids to prevent second primary tumor,
501502
risk factor reduction, 495496
risk factors for, 495
screening and early detection and
treatment, 497499
strategies for, 495
sentinel lymph node biopsy in staging of,
547563
diagnostic ecacy of, 555558
premise for staging with, 548552
remaining challenges, 558560
techniques, 552555
surgical management, access surgery for,
565571
maxillary approach, 565566
transmandibular approach, 566569
visor ap, 569570
of the neck in, 533546
anatomy of cervical lymph nodes in
relation to oral cancer, 535536
classication of neck dissection, 536542
considerations for future management
of, 544545
decision making on which neck
dissection to perform, 542544
history of neck dissection, 533535
postablative reconstruction techniques for,
573604
advantages and disadvantages of free
tissue transfer, 579580
classication of aps, 573
microvascular techniques, 584590
patient management, 580584
regional aps, 573579
selected aps, 590601

652

Oral (continued )
verrucous carcinoma, 513519
clinical features, 513516
microscopic features, 516517
treatment, 517518
Oral tongue cancer, 521531
assessment of risk for recurrence and
metastasis, 525527
treatment modalities, 521522
role of chemotherapy, 522
surgery and radiotherapy, 521522

P
p53 gene, as target for prevention of oral cancer,
506
Pain management, in oral cancer, interventional
approaches to, 627641
anatomy of head and neck, 628630
blockade of relevant ganglia, 634637
sphenopalatine, 634635
stellate, 636637
central neuraxial techniques, 637639
intrathecal pump, 638639
intraventricular opiates, 637638
mechanisms of cancer pain of head and
neck, 627628
peripheral nerve blocks, 630634
glossopharyngeal, 633634
mandibular, 632633
maxillary, 630632
Palliative care, end of life issues in oral cancer,
643645

INDEX

erythroplakia, 427429, 467

leukoplakia, 425427, 466467
lichen planus, 429430
Prevention, of oral cancer, 493511
biochemoprevention, 502503
chemoprevention, 499501
dietary agents, 504505
garlic, 505
oral carcinogenesis, 493494
other agents, 503504
Bowman-Birk inhibitor, 503
curcumin, 504
ketolorac, 503504
sulindac, 503
potential new targets for chemoprevention,
505506
cyclo-oxygenase-2 inhibitors, 506
epidermal growth factor receptor, 505
H-ras gene, 505
nuclear factor beta kappa, 505506
p53 gene, 506
retinoids to prevent second primary tumor,
501502
risk factor reduction, 495496
risk factors for, 495
screening and early detection and
treatment, 497499
strategies for, 495
Prognosis, in oral cancer, making sure patient and
family understand, 643644
Proteomics, in oral cancer, 488489

Patient evaluation. See Evaluation, patient.

Punch biopsy, in diagnosis of oral cancer and

precancerous lesions, 472

Perineural invasion, in oral cancer, 485

Peripheral nerve blocks, in management of oral

cancer pain, 630634
glossopharyngeal nerve block, 633634
mandibular nerve block, 632633
maxillary nerve block, 630632

R classication, of oral cancers, 438

Physical examination, in initial evaluation for oral

cancer, 439440
Positron emission tomography, in initial
evaluation for oral cancer, 440441
Postablative reconstruction. See Reconstruction
techniques.
Precancerous lesions. See Premalignant lesions.
Premalignant lesions, oral, 425433
clinical features of, 465467

Radiographs, in initial evaluation for oral cancer,

440441
Radiotherapy. See also Chemoradiation., for floor
of mouth and tongue cancer, 521522
Reconstruction techniques, postablative, for oral
cancer, 573604
advantages and disadvantages of free tissue
transfer, 579580
classication of aps, 573
microvascular techniques, 584590
patient management, 580584
antiplatelet and antithrombotic
pharmacotherapy, 583584
postoperative, 581583

653

INDEX

practicality of, 558559

support from long-term follow-up data,
560
validity of, 559560
techniques, 552555
excision of, 554
intraoperative lymphatic scanning,
553554
pathologic examination of, 554555
preoperative lymphoscintigraphy,
552553

preoperative, 580581
regional aps, 573579
deltopectoral ap, 578579
latissimus dorsi myocutaneous ap,
577578
pectoralis major myocutaneous ap,
574576
selected aps, 590601
composite bone-containing aps,
597601
soft tissue aps, 590597
Recurrence, locoregional, in oral squamous cell
carcinoma, 615625
considerations for salvage, 620621
distant recurrence, 619620
local recurrence, 616617
locoregional or regional recurrence,
617619
of oor of mouth and oral tongue cancer,
assessment of risk of, 525527

Sphenopalatine ganglion block, and neurolysis for

oral cancer pain, 634636
Squamous cell carcinoma, oral. See also Oral
cancer.
clinical features of, 467468
imaging of, 445463
locoregional recurrence in, 615625

Staging, of oral cancer, 436438

AJCC stage groupings, 438
distant metastases, 438
histopathologic grading, 438
lip and oral cavity, 437
R classication, 438
regional/cervical lymph node involvement,
437
TNM system, principles and general rules,
436437
with sentinel lymph node biopsy, 547563

Salvage, considerations for, in locoregional

recurrence of oral cancer, 620621

Stellate ganglion block, for oral cancer pain,

634635

Scalpel biopsy, in diagnosis of oral cancer and

precancerous lesions, 472473

Sulindac, prevention of oral cancer with, 503

Retinoids, chemoprevention of oral cancer with,

499501
prevention of second primary tumor with,
501502
Risk factors, for oral cancer, 495
reduction of, as prevention strategy,
495496

Screening, for oral cancer, and early detection and

treatment as prevention strategy, 497499
Sentinel lymph node biopsy, and cytology in
diagnosis of oral cancer and precancerous
lesions, 474475
in staging of oral cancer, 547563
diagnostic ecacy of, 555558
accuracy of, 557558
detecting occult metastasis by, 556557
in identication of sentinel node,
555556
premise for staging with, 548552
concept of, 549
evolution of, 549550
in cancers of head and neck, 550552
rationale for, 547548
remaining challenges, 558560

Supportive care, for oral cancer patients

undergoing chemoradiation therapy, 610611
Surgery, radiotherapy and, for oor of mouth and
tongue cancer, 521522
Surgical management, of oral cancer, access
surgery for, 565571
maxillary approach, 565566
transmandibular approach, 566569
visor ap, 569570

T
Terminal illness. See End of life issues.
Tissue uorescence, in diagnosis of oral cancer
and precancerous lesions, 469471
TNM system, for staging of oral cancers, 436437
principles and general rules, 436

654

INDEX

pull-through technique,
568569

Tolonium chloride. See Toluidine blue.

Toluidine blue, in diagnosis of oral cancer and
precancerous lesions, 471472
Tongue, oral, cancer of, 521531
assessment of risk for recurrence and
metastasis, 525527
treatment modalities, 521522
role of chemotherapy, 522
surgery and radiotherapy, 521522
Transmandibular approach, for access in oral
cancer surgery, 566569
lip-split mandibulotomy approach, 566568
non-lip-splitting mandibulotomy approach,
568

V
Verrucous carcinoma, 513519
clinical features, 513516
microscopic features, 516517
treatment, 517518
Visor ap technique, for access in oral cancer
surgery, 569570

W
Weber-Fergusson maxillectomy incision, for
access in oral cancer surgery, 565566