EBSJ Vol1 Nr1 May10

Evidence-Based Spine-Care Journal
volume 1/issue 1 — May 2010
ISSN 1662-3282
Message from the Editor “The journey of a thousand miles…” 5–8
Science in spine Conducting a winning literature search 9–14
Original research Heterotopic ossification in cervical disc arthroplasty: Is it 15–20

clinically relevant?
Treating thoracic-disc herniations: Do we always have to go 21–28

anteriorly?
Fueling the debate: Are outcomes better after posterior lumbar 29–34
interbody fusion (PLIF) or after posterolateral fusion (PLF) in
adult patients with low-grade adult isthmic spondylolisthesis?
The effect of body mass index on lumbar lordosis on the Mizuho 35–40
OSI Jackson spinal table
Dynamic anterior cervical plating for multi-level spondylosis: 41–46

Does it help?
Kyphoplasty: Traditional imaging compared with computer- 47–50

guided intervention—time to rethink technique?
ProDisc-C versus fusion with Cervios chronOS prosthesis 51–56

in cervical degenerative disc disease: Is there a difference
at 12 months?
Systematic reviews Unilateral facet dislocations: 57–66

Is surgery really the preferred option?
Addressing the challenges and controversies of managing os 67–74

odontoideum: results of a systematic review
Case report Metachronous presentation of metastasis from renal cell 75–82

carcinoma: evaluation and management of spinal metastasis
2
Imprint
EBSJ JOURNAL

Editorial
Board
Senior Editor-in-Chief: Co-Managing Editor-in-Chief:
Luiz Roberto Vialle, MD Jeffrey Wang, MD
Grupo de Coluna / Spine Unit, UCLA Spine Center, UCLA School of Medicine,
Universidade Católica do Paraná Department of Orthopaedic Surgery and Neuosurgery
Brigadeiro Franco, 979, 80.430-210 CURITIBA, Brasil 1250 16th Street, 7th floor, Tower #715,
Santa Monica, CA 90404, USA
Editor-in-Chief and Scientific Editor-in-Chief:
Jens Chapman, MD Co-Managing Editor-in-Chief:
Chief of Spine Service, Harborview Medical Center Karsten Wiechert, MD
325 Ninth Avenue, Box 359798, Seattle, Department of Spinal Surgery, Hessingpark-Clinic
Washington 98104-2499, USA Hessingstraße 17, 86199 Augsburg, Germany
Deputy Regional Editors
Deputy Editor Scientific Methodology: AOSpine Africa: Robert Dunn

Andrea C Skelly, PhD AOSpine China: Kenneth Cheung
AOSpine India: KV Menon
Deputy Editor Clinical Sciences:
AOSpine Europe: Marinus DeKleuver
Michael Fehlings, MD, PhD
AOSpine LatinAmerica: Marcello Gruenberg
Deputy Editor Basic Sciences: AOSpine Middle East: Zayed Al-Zayed
Keito Ito, PhD AOSpine North America: Darrel Brodke
AOSpine Pacifica: Bryan Ashman
Deputy Editor Publishing:
Kathrin Lüssi
Disclaimer
Great care has been taken to maintain the accuracy of the infor- Some of the products, names, instruments, treatments, logos, desi-
mation contained in the publication. However, neither AOSpine gns, etc. referred to in this book are also protected by patents and trade
International, nor the editors and/or the authors can be held marks or by other intellectual property protection laws (eg, “AO”, “ASIF”,
responsible for errors or any consequences arising from the use “AO/ASIF”, “TRIANGLE/GLOBE Logo”, “AOSpine” are registered trade-
of the information contained in this publication. marks) even though specific reference to this fact is not always made in
The statements or opinions contained in editorials and articles the text. Therefore, the appearance of a name, instrument, etc. without
in this journal are solely those of the authors there of and not of designation as proprietary is not to be construed as a representation by
AOSpine International. The products, procedures, and therapies the publisher that it is in the public domain.
described are only to be applied by certified and trained medical This publication, including all parts thereof, is legally protected by
professionals in environments specially designed for such proce- copyright. Any use, exploitation or commercialization outside the nar-
dures. No suggested test or procedure should be carried out unless, row limits of copyrights legislation, without the publisher’s consent, is
in the reader’s professional judgment, its risk is justified. Because illegal and liable to prosecution. This applies in particular to photostat
of rapid advances in the medical sciences, we recommend that reproduction, copying, scanning or duplication of any kind, translating,
independent verification of diagnosis, drugs dosages, and operating preparation of micro-films and electronic data processing and storage.
methods should be made before any action is taken. Institutions’ subscriptions allow to reproduce tables of content or pre-
Although all advertising material is expected to conform to pare lists of articles including abstracts for internal circulation within
ethical (medical) standards, inclusion in this publication does not the institutions concerned. Permission of the publisher is required for
constitute a guarantee or endorsement of the quality or value of resale or distribution outside the institutions. Permission of the publisher
such product or of the claims made of it by its manufacturer. is required for all other derivative works, including compilations and
translations.
Permission of the publisher is required to store or use electronically
any material contained in this journal, including any article or part of
an article. For inquiries contact the publisher at the address indicated.
Volume 1/Issue 1 — 2010
3
Table of Contents
EBSJ JOURNAL

May 2010
Volume 1, Issue 1
Message From the editor
5–8 “The journey of a thousand miles…”

Jens Chapman
science in spine—Fundamentals
9–14 Conducting a winning literature search

Erika D Ecker et al
Original research
15–20 Heterotopic ossification in cervical disc arthroplasty: Is it clinically relevant?

Giuseppe M Barbagallo et al
21–28 Treating thoracic-disc herniations: Do we always have to go anteriorly?

Richard J Bransford et al
29–34 ueling the debate: Are outcomes better after posterior lumbar interbody fusion
F
(PLIF) or after posterolateral fusion (PLF) in adult patients with low-grade adult
isthmic spondylolisthesis?
Giovanni Barbanti Bròdano et al
35–40 The effect of body mass index on lumbar lordosis on the Mizuho OSI Jackson spinal
table
Justin Bundy et al
41–46 Dynamic anterior cervical plating for multi-level spondylosis: Does it help?
Ashraf A Ragab et al
47–50 Kyphoplasty: Traditional imaging compared with computer-guided intervention—

time to rethink technique?
Michael P Silverstein et al
51–56 P roDisc-C versus fusion with Cervios chronOS prosthesis in cervical degenerative
disc disease: Is there a difference at 12 months?
Matjaz Vorsic et al
Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

4 Table of Contents—Evidence-Based Spine-Care Journal—May 2010
Systematic reviews
57–66 Unilateral facet dislocations: Is surgery really the preferred option?

Marcel Dvorak et al
67–74 Addressing the challenges and controversies of managing os odontoideum: results of a

systematic review
Jefferson R Wilson et al
Case report
75–82 Metachronous presentation of metastasis from renal cell carcinoma: evaluation

and management of spinal metastasis
Joshua C Patt et al
83 Definition of the different classes of evidence (CoE)
Message from the Editor—“The journey of a thousand miles….” 5
Message from the Editor
“The journey of a thousand miles….”
One common translation of the remainder of this famous ancient quote by Chinese
philosopher Lao-Tzu, is that the journey “starts with a single step.”
It is with this insight that I would like to introduce you to the Evidence-Based Spine-
Care Journal (EBSJ), a new scientific journal organized through AOSpine Interna-
tional, the largest multispecialty, truly global spine society. Of course, a reasonable
first question to ask is “Why another spine journal?” A brief review of where we came
from and where we are intending to head may provide insight as well as reveal its
purpose.
AO (Arbeitsgemeinschaft für Osteosynthesefragen)—the venerable and unique surgi-

cal organization—recently celebrated its 50th year. The simple yet profound concept of
this organization was based on four principles, which later became known as the AO’s
“four pillars of cooperation”:
• Documentation of all patients
• Development of instrumentation and implants (which has been cutting edge!)
• Research
• Teaching (which is still unsurpassed)
Since its inception, the AO has espoused many of the primary principles of evidence-
based medicine (EBM) before it was formally described. AO values and activities have
encouraged evaluation of results, critical appraisal to provide context for drawing in-
ferences and using these as a basis for directing future basic and clinical research. This
integrated and ever-evolving process sets the stage for the AO community to be an ac-
tive player in EBM.

6 Message from the Editor—“The journey of a thousand miles….”
Within the spine-care community we have remained faithful to the core values, yet
have found it frequently difficult to offer highest quality source materials for our hun-
dreds of annual spine courses around the world. This has made it challenging to teach
using an evidence-based approach versus relying mainly on opinion-based resources.
With this inaugural issue, EBSJ is the first step toward assisting the global community
of spine-care professionals with finding, describing and developing the highest quality
evidence in its very own, unique format. In doing so, EBSJ expands on the first step
that was EBSS (Evidence-Based Spine Surgery).
Behind the substitution of one consonant within the title change from EBSS to the
new EBSJ are a number of rather substantial changes:
• Its community-centric approach: EBSJ’s format and publication model will make it
much easier for members of the international community to contribute. With this
approach, EBSJ will be a better vehicle for uniting AOSpine surgeons and other
providers from around the world.
• A more efficient production model: Combined with eventual MEDLINE® indexing,
this model will allow for broader dissemination and stronger impact across
disciplines.
• A directive to enhance the quality of evidence in spine care: With our evidence-based fo-
cus, presentation and distribution, we can expand upon the overall quality of evi-
dence in the field.
This inaugural issue of EBSJ includes research submissions from the first Global Spine
Conference held last June in San Francisco. This is an excellent start to establishing
EBSJ’s commitment to the international community, while featuring some interesting
topics and novel approaches to treatment of spine-related problems.
These lofty goals set forth a literal “journey of (more than) a thousand miles,” with
meetings and gatherings around the world from a group of dedicated individuals who
have committed themselves to the concept of a truly novel and hopefully very worth-
while spine journal that surpasses the scope of existing journals.
Another possible translation of the Chinese proverb is that the journey “…begins be-
neath one’s feet.” This is perhaps a reference to beginning where one is and knowing
where one wants to go from there. Applied to our current situation, we would like to
explore in unprecedented depth the current status of the evidence on any given spine
topic, thus allowing us to consider how to improve its quality. We may not like the
quality of the current evidence, but we can significantly enhance it by setting on a
clear path of many steps and many feet, geared towards providing meaningful and
clinically relevant evidence for spine care.
To set us in motion for this journey, regular features of EBSJ will include:
• S ystematic reviews on timely and sometimes controversial topics. This allows us to see
“where we are at” with regard to the evidence on those topics. We hope this stim-
ulates discussion and encourages researchers to enhance the quality of evidence.
It hopefully will also stimulate new interest in exploring certain topics under a
new light.
Message from the Editor—“The journey of a thousand miles….” 7
• Original research reports. As a reporting requirement for EBSJ, studies must follow
specific guidelines accounting for patients and follow-up based on reporting stan-
dards from CONSORT and others, detailing patient characteristics and reporting of
methods based on the PICO or PPO concept. No other journal makes these specifi-
cations. To require these is a step toward enhanced credibility.
• Appraisal
of original studies. We feature methodological reviews and Class of Evi-
dence (CoE) appraisals of original studies by PhD-level researchers with expertise
in clinical research and epidemiology. This provides authors and the spine com-
munity with insight into how to enhance the quality of future studies. It enhances
understanding of how to perform and use research from an evidence-based per-
spective by including critique of important methodological components that can
bias results. Future researchers can gain significant insight by considering these
areas and planning studies which seek to decrease bias.
• Methodological
assistance. We will offer an opportunity for assistance to authors with
topics of importance to the global community who otherwise would not have ac-
cess to methodological expertise or statistical analysis. This will broaden the
chance for research-minded spine surgeons around the world to contribute in a
meaningful fashion to the knowledge base of spine care.
• Brief
educational pieces on how to effectively use and conduct research. These will facili-
tate an understanding of how to find, understand, create and report the highest
quality research.
• An evidence-based case discussion. This will address the recurring issue of outliers in
actual daily care in the face of ever-changing guidelines and evidence-based prac-
tice recommendations.
You will note that each of the studies in this inaugural issue has limitations; in fact, all
studies have limitations. It is a well-honed academic practice to find fault and limita-
tions in any research, especially clinical research. This applies even to recent multi-
million dollar studies conceived by some of the brightest minds in medicine, public
health, and epidemiology. Only by taking a step back and making constructive sugges-
tions that improve the quality of studies and how they are reported, will we be able to
enhance the credibility and quality of the evidence in spine care overall. Hence I am
proud of the authors who were willing to stick their necks out and contribute to a
brand new spine journal. By sharing their research in EBSJ, they are starting the pro-
cess of systematically and profoundly changing the way we deliver spine care, by in-
corporating evidence into our practices in a clinically meaningful fashion. I very much
respect and appreciate the authors who submitted to our first issue for taking the first
actual steps in our journey.
As spine care professionals, our first priority is to provide our patients with the highest
quality of care possible; this includes consideration of the very basic evidence that what
we do is beneficial and acknowledging what the limitations may be. As teachers—
whether through AOSpine courses, in our hospitals, or at our medical schools—we
should increasingly rely on scientifically founded, evidence-based materials to replace
the more subjective impression-based teaching contents of the past. Our literature
must provide a higher quality of evidence so that we can practice and teach from an
evidence-based perspective, including understanding where the evidence falls short.
Whether we like it or not, as healthcare systems around the world increasingly feel
economic constraints, mounting pressure will be placed on our spine community to

8 Message from the Editor—“The journey of a thousand miles….”
provide high quality evidence of efficacy, effectiveness, safety, and cost-effectiveness

to support funding and reimbursement decisions. This process can be a most painful
one, especially if changes in customary health care delivery models are brought forth.
Unfortunately, there are plenty of examples where the good name of “evidence” is em-
ployed for political decisions rather than using scientific publications as a starting point
for discussion. The discussion should include objective review of the studies, always
with an eye towards leaving room for human exceptions and with an understanding
of the limitations of research to provide conclusive evidence for all situations. Our own
literature must provide a consistently higher quality of evidence that effectively docu-
ments the efficacy, effectiveness and safety of what we do so that policy and decision
makers have a clear picture of the strengths and limitations of what we do.
With EBSJ, we have an opportunity to move forward on the journey, a grand march,
toward higher quality research and evidence in our field. I invite you as a reader, au-
thor or reviewer to join in our mission of setting the stage for evidence-based practice
and influencing the future of spine surgery. Simply go to the EBSJ tab on the AOSpine
website to learn more about this exciting resource. We hope you will become engaged
in and join our “journey of a thousand miles.”
Welcome to the inaugural issue of EBSJ!
Science in spine—Fundamentals—Conducting a winning literature search 9
9—14
Conducting a winning literature search

Authors Erika D Ecker, Andrea C Skelly
InstitutionSpectrum Research, Inc., Tacoma, Washington, USA
So what is a “winning literature search”? Simply put, it is one that provides you with the infor-
mation you need to find the types of articles that will help you with clinical practice or research.
Literature searching is a combination of an art and a science. Understanding the basic anatomy
and physiology of searching can get you started on finding the information you need.
i. ANAtOMy OF A LitERAtURE SEARCh—thE SkELEtON
A. Constructing an appropriate question

• Asking the right question is the primary key to creating a winning search. Your
questions must be answerable. If your question is too broad, your search will yield
more information than you can possibly look through.
• Suppose you are interested in determining the incidence of commonly reported
complications following artificial disc replacement (ADR) as compared with fusion.
too broad More answerable Better yet

What is the incidence of What is the incidence of What is the incidence of complications following
complications following ADR complications following ADR versus ADR versus fusion for degenerative disc disease of the
versus fusion? fusion of the lumbar spine? lumbar spine?
Or even…
What is the incidence of complications following

single-level ADR versus single-level fusion for
degenerative disc disease of the lumbar spine?

10 Science in spine—Fundamentals—Conducting a winning literature search
Using treatment studies as an PICO (Patients, Intervention, Comparator, Outcomes) table for designing
example, the PICO concept can really
help you create an answerable question your question
since, as you will see below, it will help Included Excluded
you create a search strategy.
Patients Degenerative disc disease at one level of the lumbar spine All other diagnoses (ie, tumor, trauma);
2- or 3-level disc disease; cervical disease
Intervention Artificial disc replacement
Comparator Fusion
Outcome Death, infection, subsidence and migration, loss of disc height, All other complications
heterotopic ossification and spontaneous fusion, reoperation
Similarly, if your study was prognostic rather than therapeutic, a PPO table would be
used instead of the PICO table in order to help formulate your question. Thus, the
categories would change to Patients, Prognostic factors, and Outcome.
B. Using the appropriate database(s)

Now that you have honed your question, it is time to focus on efficient article retrieval.
What type of information is needed and what type of articles do you need? What will
you do with the information you gather? This will determine the type of information
you need and types of articles that may provide that information. It also may influence
the type of database you search for that information.
• B ibliographic databases contain references to published literature, such as journals

and newspaper articles, conference proceedings and papers, reports, government
and legal publications, patents, and books.
• T wo types of literature that are often the focus of systematic searches are indexed,
peer-reviewed literature, and gray literature.
–– Peer-reviewed literature is scholarly work that generally represents the latest
original research in a field. These articles undergo expert screening before pub-
lication to ensure meaningfulness within the context of other research in the
discipline and, at least in theory, sound methodology.
–– “Gray” literature refers to material that is not formally published by commer-
cial publishers or peer-reviewed journals, including reports, fact sheets, white
papers, conference proceedings, and other documents from various organiza-
tions and government agencies.
The table below provides brief descriptions of common databases and sources to search
both peer-reviewed and gray literature.
Database & website Description What is included

Indexed, peer-reviewed literature
MEDLINE – National Library of Medicine’s (NLM) premier – Academic journals covering the
http://www.ncbi.nlm.nih.gov/pubmed/ bibliographic database fields of medicine, nursing,
– PubMed is a free search engine and is dentistry, veterinary medicine,
maintained by the National Center for the health care system, preclinical
Biotechnology Information (NCBI) at the sciences
NLM – Much of the literature in biology,
– Contains over 19 million citations biochemistry, molecular evolution
– International in scope
EMBASE (Excerpta Medica Database) – Comprehensive biomedical and – Active, peer-reviewed journals
http://www.embase.com/ pharmacological database – Broad international scope
– Maintained by Elsevier and can be accessed
by subscribed users only
– Contains over 20 million citations
Database & website Description What is included

Cochrane Reviews – Database comprised of formal, extensive – Topics including medications,
http://www.cochrane.org/reviews/ systematic reviews that often contain meta- surgery, technology, education
analysis – A rticles are also indexed in
– Maintained by The Cochrane Collaboration, an PubMed
international nonprofit organization, and
published and hosted by Wiley InterScience
– O ffers free access to abstracts and some full
length articles; however, most full text
reviews require a subscription or
pay-per-view access
– Designed to facilitate clinical decision-making
in healthcare by exploring the evidence for and
against the effectiveness and appropriateness
of treatments in specific circumstances
Cochrane CENTRAL (The Cochrane – Collection of databases in medicine and other – Focuses on randomized or
Central Register of Controlled Trials) healthcare specialties controlled research articles
http://uscc.cochrane.org/en/newPage3.html – Uses a search interface called OVID
AOSpine—EBSS.live – Comprehensive database designed to – Summaries of recently published
http://www.aospine.org/ streamline the search process by providing research articles on a variety of
evidence on treatment of spine problems that topics, including spine therapies,
is organized effectively and graded according prognoses, and diagnoses
to evidence class
– Maintained by AOSpine International and
requires a paid membership for access
Gray literature
AHRQ (Agency for Healthcare – Lead federal agency charged with improving – Focuses on synthesizing the
Research and Quality) the quality, safety, efficiency, and effectiveness evidence and facilitating the
http://www.ahrq.gov/ of health care for the United States translation of evidence-based
– E stablished twelve evidence-based practice research.
centers (EPCs) that develop evidence reports
and HTAs on topics relevant to healthcare
organization and delivery issues, specifically
those common to and/or significant for the
Medicare and Medicaid populations
– Maintained by AHRQ staff and provides free,
online access
NCG – Free, public resource for evidence-based – E vidence-base for guidelines is
(National Clearinghouse Guidelines) clinical practice guidelines described
http://www.guideline.gov/ – Created and maintained by the AHRQ – Syntheses of selected guidelines
that cover similar topic areas
– E xpert commentary on issues of
interest and importance to the
clinical guideline community
INAHTA – Database containing information on HTAs – Research including systematic
(International Network of Agencies from various countries around the world reviews, ongoing and completed
for Health Technology Assessment) – Managed by INAHTA Secretariat in trials, questionnaires, and
http://www.inahta.org/HTA/Database/ collaboration with their United Kingdom economic evaluations
member Centre for Reviews and
Dissemination (CRD)
– Free access
For example, continuing with our question regarding complications after ADR versus
fusion, which type of database listed in the table makes the most sense to search?
I ndexed peer-reviewed articles will give us the best available and most current data
and MEDLINE, which includes millions of citations for biomedical articles and can be
accessed using PubMed for free, seems like a great starting place. Generally speaking,
PubMed will be the best place to begin your search and there are various ways, as you
will see below, to refine and limit your search in order to find exactly what you need.

II. The physiology of literature searching—how it works
Now that you have an answerable question and an idea of what type of database you
need to search (at least to start), let’s talk about the nuts and bolts of searching. For the
purposes of this paper, we will use PubMed as the search engine.
A. Getting specific—the basics of “how to”
1. Quickstart:
• T ype a word or phrase into the query box, including subject, author, and/or journal
• Click on the search button or press the “enter” key
• Results will be displayed in summary format:
Anything which appears Anterior fixation of odontoid fractures in an elderly population.
in blue and is underlined is a link that
reveals more information. Clicking on Dailey AT, Hart D, Finn MA, Schmidt MH, Apfelbaum RI.
the title would bring up the abstract J Neurosurg Spine. 2010 Jan; 12(1):1-8.
(Abstract format). Clicking on “Related PMID: 20043755 [PubMed - indexed for MEDLINE]
articles” would provide a link to Related articles
other similar articles that might be
of interest.
• T o retrieve more information about the search results, use the display settings
menu (upper left corner) to view the abstract or MEDLINE formats, change the
number of items that appear per page, and sort by recently added, publication
date, first author, last author, journal, or title.
• P ubMed also contains links to full-text articles (appears in upper right corner of
page) at participation publishers’ web sites as well as links to other third party
sites such as libraries and sequencing centers.
2. Advanced searching in PubMed—MeSH terms and the MeSH database:

a. Medical subject headings (MeSH)
It is important to understand that PubMed uses a controlled vocabulary to index
journal articles called MeSH and uses “automatic term mapping” to find MeSH
terms when you search. MeSH terms are organized in a hierarchy called a tree, with
more specific (narrower) terms arranged beneath broader terms. By default, PubMed
includes in the search all narrower terms; this is called “exploding” the MeSH term.
Inclusion of MeSH terms enhances and optimizes the search strategy. For example,
if you looked up the term “Spine” in the MeSH database you would see:
All MeSH categories
Anatomy category
Musculoskeletal system
Skeleton
Bone and bones
Spine
Cervical vertebrae
Axis +
cervical atlas
Coccyx
Intervertebral disk
Lumbar vertebrae
Sacrum
Spinal canal
Epidural space
Thoracic vertebrae
Therefore, PubMed would retrieve every article containing any of the terms located
under Spine in the hierarchy.
b. MeSH database features

MeSH vocabulary contains over 25,000 descriptors and is updated weekly and
reviewed annually. You can only search citations that have been indexed for
M EDLINE (92% of the PubMed database) using MeSH terms. Features include:
• A llows you to identify and select appropriate MeSH terms for a search and to see
their definitions
• Builds a PubMed search strategy
• Displays MeSH terms in the hierarchy (MeSH tree) allowing you to broaden/
narrow a search
• L imits MeSH terms to a major concept/topic heading for a search
• A llows you to broaden your search by choosing not to explode a term
• Attaches subheadings for a search creating complex search strategies
–– The list of subheadings includes terms paired at least once with a given heading
in MEDLINE.
• Focuses searches using other types of MeSH terms including publication types [pt],
substance names [nm] or registry numbers [rn], and pharmaceutical actions [pa]
• MeSH Brower for access to annotations: http://www.nlm.nih.gov/mesh/
To access MeSH from PubMed, click on MeSH Database on the PubMed homepage
or click MeSH under “more resources” in “advanced search.”
Once in the MeSH database, if you entered cancer into the search bar and clicked Go
(or hit Enter) you would see:
€ 1: Neoplasms Links
New abnormal growth of tissue….
Clicking on “neoplasms” will bring up the page where you have the option of select-
ing any of the features listed above to help you refine your search.
Also, clicking “links” adjacent to the MeSH term desired, will give you a drop-down
menu which offers several options:
• PubMed: search PubMed with the term
• P ubMed—Major topic: search PubMed with the MeSH term, retrieving only
c itations where the term is a major focus
• Clinical queries: put the MeSH term into the Clinical Queries box where the
search may be further refined
The Mesh database homepage
• N LM MeSH browser: show the MeSH browser descriptor data for this term
includes three brief tutorials on how to i ncluding scope note, allowable qualifiers, and the MeSH tree
search with the MeSH database, combine
MeSH terms, and apply subheadings and
other features of the MeSH database.
B. Too much information! Refining your search
• R eplace general search terms with more specific terms (the MeSH database would
be a great resource for this)
• Add terms or combine search terms with connector words: AND, OR, or NOT using
upper case letters (called Boolean logic)
–– AND between terms returns only records that contain all of the search terms
–– OR between terms returns all records that contain any of the search terms
–– NOT between search terms returns only records that contain the first term and
not the second
• T runcate terms. Place an asterisk (*) at the end of a string of characters to search
for all terms that being with that string. PubMed searches the first 600 variations
of a truncated term.
–– Example: mimic* will find all terms that begin with the letters m-i-m-i-c-; eg,
mimic, mimics, mimicking

• U se a wildcard. Use a “?” to replace a letter or denote an extra letter where spelling
or word variation is possible.
–– Example: behavio?r will find behaviour or behavior
• Use the “limit” option in PubMed to limit citations by age group, language, publi-
cation type, date, human studies, etc.
• Use the “advanced search” option to look up a term as it is indexed in PubMed
• Use the MeSH database features
Example:
Let’s say we are interested in what the best surgical treatment is for osteoporotic spine frac-
tures. Using some of the tips above, the chart below shows how a typical search might go:
Search term(s) # citations

“Spinal fracture” [MeSH] OR vertebral compression fracture 16023
“Spinal Fracture” [MeSH] OR vertebral compression fracture AND “osteoporosis” [MeSH] 3718
“Spinal Fracture” [MeSH] OR vertebral compression fracture AND “osteoporosis” [MeSH] AND “surg*” 911
“Spinal Fracture” [MeSH] OR vertebral compression fracture AND “ osteoporosis” [MeSH] AND “Surg*” 54
Limits: only items with abstracts, humans, clinical trial, English, publication date from 1990–2010
By combining terms (using Boolean logic), truncating a term, and using the limits
optionwe were able to narrow our search down from 16,023 articles to a more man-
ageable and relevant 54 articles. The “details” tab in the PubMed search window shows
the complete search expression (ie, query translation) employed by PubMed, similar to
what is represented in the table above.
Summary checksheet
Use PICO (or PPO) to come up with an answerable question
Decide what type of literature you will need to search in order to adequately answer your question
Decide which database is most practical to use to start your search
Pick out key terms from your question to enter into the database’s search box
Refine your search as necessary by combining terms and/or using limiting options that the database provides
One of the best resources that PubMed provides for users new to the database is the
online tutorials. They are brief but informative and because they are interactive you are
guided step-by-step through each process. Perhaps consulting the online tutorials and
the fact sheets on PubMed would be a next step for you. Give it a try! In fact, here is the
link to the PubMed Tutorial homepage created by the National Library of Medicine:
http://www.nlm.nih.gov/bsd/disted/pubmedtutorial/. Also, check-out the PubMed
help page which contains a plethora of information regarding all aspects of PubMed:
http://www.ncbi.nlm.nih.gov/bookshelf/br.fcgi?book=helppubmed&part=pubmedhelp
III. Closing thoughts
Remember, literature searching is a combination of an art and a science. It requires

practice, intuition, and some trial and error. While there is a basic structure, a set of
guidelines and many tools for assisting one with basic searches, there are a variety of
nuances and advanced techniques that may be required for more specialized searches.
For systematic reviews as an example, extensive searches are required and may take
numerous hours, involving many databases (including those for gray literature), and a
combination of advanced search strategies in order to be methodologically sound. Use
of personnel with specialized expertise in conducting such searches may provide the
best results and be the most resource effective.
Original research—Heterotopic ossification in cervical disc arthroplasty: Is it clinically relevant? 15
15—20
Heterotopic ossification in cervical disc

arthroplasty: Is it clinically relevant?
Giuseppe M Barbagallo, Leonardo A Corbino, Giuseppe Olindo, Vincenzo Albanese
Authors
Institution
Department of Neurosurgery, Azienda Ospedaliero-Universitaria “Policlinico-Vittorio Emanuele”,
Catania, Italy
Abstract
Methods evaluation and class
of evidence (CoE)
Study design: Retrospective cohort study.
Methodological principle:
Study design: Objective: To analyze the presence and clinical relevance of heterotopic os-
Prospective cohort sification (HO) at 3 years mean follow-up.
Retrospective cohort •
Case-control Methods: Thirty patients suffering from cervical radiculopathy and/or my-
Case series elopathy treated with anterior disc replacement (ADR) were studied.
Methods HO was classified using the McAfee grading system. Range of motion
Patients at similar point in course of • was measured from flexion and extension x-rays. Short-form 36 and
treatment neck disability index (NDI) assessed functional outcome.
Follow-up ≥ 85%
Similarity of treatment protocols for • Results: Forty-five prostheses were implanted in 30 patients with cervical
patient groups radiculopathy and / or myelopathy, mean age 40.9 years. Nineteen pa-
Patients followed for long enough for • tients received 1 level and 11 patients received multilevel disc replace-
outcomes to occur ment. The incidence rate of HO was 42.2% (19 levels). Segmental range
Control for extraneous risk factors* of motion was ≥ 3° in 93.8% of patients with HO. There was no signifi-
Evidence class: III cant difference in functional scores between those who did and those
who did not develop HO. Males tended to develop HO more frequently
* Authors must provide a description of robust than females, though this was not statistically significant. The indica-
baseline characteristics, and control for those tion for surgery (soft disc hernia or spondylosis) was not associated with
that are potential prognostic factors.
the formation of HO.
The definiton of the different classes

Conclusions: Functional improvement is maintained despite the presence
of evidence is available on page 83.
of HO following cervical disc arthroplasty. Indications for arthroplasty
should not be halted by the risk of HO.
See web appendix at www.aospine.org/ebsj

No financial support has been or will be received for this study.
Notation of device status. Prodisc-C and Prestige LP cervical artificial discs: FDA approved

16 Original research—Heterotopic ossification in cervical disc arthroplasty: Is it clinically relevant?
Study RATIONALE and context METHODS

Maintainance of motion following arthroplasty can be
hindered by the development of heterotopic ossification Study design: Retrospective cohort study.
(HO), whose causes remain uncertain [1, 2, 3–5].
Whether the formation of HO following cervical disc ar- Inclusion criteria: Patients suffering from radiculopa-
throplasty is of clinical or functional importance is also thy, myelopathy or myeloradiculopathy secondary to
unknown. The aims of this study are to investigate the soft disc hernia and/or mild degree spondylosis oper-
rate of HO at a longer follow-up (mean 3 years), to deter- ated between October 2004 and December 2006.
mine if the presence of HO negatively affects functional
outcome and to identify factors that may precipitate the Exclusion criteria: Patients with prior surgery at the
formation of HO following cervical disc arthroplasty. instrumented level were excluded.
Patient population and interventions (Fig 1)

Clinical questions • Thirty-nine patients suffering from cervical ra-
1. What is the incidence of HO following cervical diculopathy and/or myelopathy treated with an-
arthroplasty? terior disc replacement (ADR) were selected. Nine
2. Does the development of HO negatively influence the patients were excluded from this analysis because
clinical outcome in patients receiving arthroplasty? of insufficient outcome data at different time in-
3. Are there factors that influence the formation of HO tervals (n = 7) or were not available for follow-up
following cervical arthroplasty? (n = 2).
Outcomes and analysis

• Evidence of HO using the McAfee grading system
[6] (Table 1).
• P rostheses range of motion (ROM). ROM was cal-
culated by using two lines parallel to the Prodisc-
C keels, or to the endplates of the Prestige LP, and
measuring the intervening angle in flexion and
extension on lateral radiographs. Lordosis was as-
signed a negative value and kyphosis a positive
one; the difference between the two values gave
the final ROM. ROM was measured three times
per level and a mean value was obtained.
• C T scan on patients with grade two or higher HO
to assess localization of HO.
• Short-form 36 (SF-36) and neck disability index
(NDI). A comparison of scores was made between
those who developed HO and those who did not
using a two-sided Student t-test.
• A ll images were independently reviewed by both
a radiologist and a neurosurgeon not directly in-
volved in the surgical procedures.
Table 1 McAfee grading of heterotopic

Results
ossification (HO)[6]
• Forty-five disc prostheses (seven Prestige-LP and 38
Grade 0 Absence of HO
Prodisc-C) were implanted in 30 patients (18 males)
ranging in age from 28–63 years (mean age 40.9). Grade 1 Presence of HO in front of vertebral body but not in the
anatomic disc space
The underlying disease included radiculopathy
(n = 13), myelopathy (n = 4) or myeloradiculopathy Grade 2 Presence of HO in the disc space, possibly affecting the
prosthesis’s function
(n = 13), secondary to soft-disc hernia (n = 17), spon-
dylosis (n = 8) or a combination of disc hernia and Grade 3 Bridging HO with prosthesis’s motion still preserved
spondylosis (n = 5). Nineteen patients received one- Grade 4 Complete fusion of the segment with absence of motion in
level disc replacement, seven patients received two- flexion/extension
level and four patients received three level disc re-
placement. The following distribution of prostheses/
level was recorded: C3–4:5; C4–5:4; C5–6:23; C6–
7:13 (Table 2). Table 2 Patient characteristics
• The incidence rate of HO using the number of pros-
Characteristic N = 30
theses as the denominator was 42.2% (19 levels in 16
patients, Fig 6). The risk of a patient developing HO Age, years, mean (range) 40.9 (28–63)
after receiving one or more prostheses was 53.3% Male, n (%) 18 (60)
(16/30). Soft disc hernia, n (%) 17 (57)
• Ten patients showed progressive ossification over Spondylosis, n (%) 8 (27)
time: in six patients de novo HO was demonstrated at
Soft disc hernia & spondylosis, n (%) 5 (17)
23–46 months (mean 33.3) postoperatively (Figs 2, 3).
Radiculopathy, n (%) 13 (43)
• A mong the 16 patients with HO, 15 (93.8%) demon-
strated a range of motion ≥ 3° (Fig 4). Myelopathy, n (%) 4 (13)
• Comparison of pre- and postoperative SF-36 and NDI Myeloradiculopathy, n (%) 13 (43)
scores revealed functional improvement from base- No. of levels treated 45
line to follow-up at 3, 6, 12, 18, 24, and 36 months. 1 level 19 (patients)
There was no significant difference in functional 2 levels 7 (patients)
3 levels 4 (patients)
scores between those who did and those who did not
develop HO, P > .05 (Fig 7). Levels implanted
C3–4 5
• R isks and relative risks by factor related to patient
C4–5 4
characteristics associated with HO formation were C5–6 23
analyzed: males tended to develop HO more than fe- C6–7 13
males (twice as likely), though this did not reach sta-
tistical significance (Table 3).
• No patients required revision surgery for adjacent
segment disease or persisting/recurrent symptoms.
Table 3 The risk (%) and unadjusted relative risk (RR)
of HO by patient characteristics
n/N (%) RR 95% CI P-value

Fig 1 Patient population and intervention
Sex
Total patients receiving intervention
Female 4/12 (33.3) 1.0
during time period
(n = 39) Male 12/18 (66.7) 2.0 0.8, 4.7 .13
Excluded (n = 9) Indication
Reasons:
• insufficients data (n = 7) Soft disc hernia 8/17 (47.1) 1.0
Available for • not available for follow-up (n = 2) Spondylosis 4/8 (50.0) 1.1 0.4, 2.5 1.0
analysis
(n = 30) Soft disc hernia & spondylosis 3/5 (60.0) 1.3 0.5, 3.1 1.0
Neurological involvement
Myelopathy 2/4 (50.0) 1.0
Patients with HO Patients without HO Radiculopathy 7/13 (53.8) 1.1 0.4, 3.2 1.0
(n = 16) (n = 14) Myeloradiculopathy 7/13 (53.8) 1.1 0.4, 3.2 1.0

Discussion • I ndications for arthroplasty are not affected by the

• Our rate of grade two and three HO (42.2%) is con- risk of HO.
sistent with that previously reported [7]. It is higher • P reventative measures such as prophylactic non-
than in other series [8, 9] but lower than Sola’s et al steroidal medications, local application of bone wax
[10]. or intentional early range of motion measures were
• HO is prevalent on the anterolateral surface of verte- not routinely utilized in our study. While these are
bral bodies (Fig 5). interesting factors, actual prospective studies will be
• HO does not influence the clinical and functional needed to compare their effect. Our experiences may
outcome. serve as a valuable baseline comparison.
• Strengths: This study reports 3 years mean follow-up
data, to date the longest available. For all included
patients complete follow-up data are available. Fur-
thermore, we have shown that a delayed progression,
or de novo onset, of HO is possible.
• L imitations: This study was not conducted prospec-
Fig 3 Double-level arthroplasty with Prodisc-C at
tively and is based on a relatively small albeit con-
C5–6 and C6–7. Note the progression of HO at the
secutive sample size. Lack of potential prognostic
C4–5 disc space at 1 month (left), 12 (middle) and 48
factors available for analysis and a 77% follow-up
(right) months, respectively, postoperatively (arrow).
rate are further limitations.
• A longer follow-up is necessary to assess whether ar-
throplasty can reach the goal of maintaining motion
and if development of high-grade HO at the target
level infers an increased risk of adjacent segment
degeneration.
CONCLUSIONS
• Clinical and functional improvement following cer-
vical disc arthroplasty is maintained despite the
1 month 12 months 48 months
presence of HO.
• The indication for surgery (soft-disc herniation or
“hard-disc” spondylosis) is not associated with the
Fig 4 Flexion and extension x-rays (same patient as in
formation of HO.
Fig 3) showing mobile disc prostheses at 48-months
follow-up despite the presence of heterotopic
Fig 2 Double-level arthroplasty with Prodisc-C at ossification.
C5–6 and C6–7. Absence of HO at 1 month after
surgery (left), presence of grade two HO 1 year
postoperatively (middle) and of grade three HO
(right) 3 years postoperatively, respectively.
1 month 1 year 3 years

HO 2 HO 3
Fig 5a–b Same case as in Fig 4. Axial CT scan showing Fig 7 Differences of pre- and postoperative functional
the presence of laterally sited HO (arrows) at C4–5 outcome in the two groups
level (a) and at C6–7 level (b)
PCS HO Patients
PCS Non HO Patients
120
100
80
60
a
40
20
0
Pre 3 6 12 18 24 36
SF-36 Physical component mean score (PCS)
MCS HO Patients
b MCS Non HO Patients
120
100
80
60
40
Fig 6 Number of ossified levels and stratification for 20
McAfee ossification grades
(HO: heterotopic ossification) 0
Pre 3 6 12 18 24 36
10
10 SF-36 Mental component mean score (MCS)
8
8
n. levels
NDI HO Patients
6
NDI Non HO Patients
4 120
2 100
1
0
HO 2 HO 3 HO 4 80
60
40
20
0
Pre 3 6 12 18 24 36
Neck disability index (NDI) mean scores

REFERENCES throplasty is appreciated. As the authors note, the causes of

HO in disc arthroplasty remain unclear and are probably
1. Bartels RHMA, Donk R (2005) Fusion around cervical multifactorial.
disc prosthesis: case report. Neurosurgery; 57:194.
2. Heller JG, Tortolani PJ, Park AE, et al (2003) Computed HO frequency: The authors report a HO rate of 42% fol-
Tomography (CT) scan assessment of paravertebral lowing artificial disc replacement in 32 patients, a higher
bone after total cervical disc replacement: prevalence, overall rate than is reported in most other studies. HO rates
temporal relationships, and the effects of NSAIDs. in the May 2008 EBSS Special Edition on cervical arthro-
Proc Cervical Spine Res Soc. Read at the Annual Meeting plasty were summarized as follows: There were no HO cases
of the Cervical Spine Research Society, European reported in two RCTs, with follow-up of 24 months in one
Section, Jun 19-20; Barcelona, Spain study and other being a preliminary report with most pa-
3. Parkinson J, Sekhon LHS (2005) Cervical arthroplasty tients having only 12-months follow-up. HO was, however,
complicated by delayed spontaneous fusion. J Neuro- common in two case series with one reporting an overall 18%
surg Spine; 2:377–380. rate or 7% cases when restricted to grades III or IV. The other
4. Pickett GE, Sekhon LHS, Sears WR, et al (2006) Compli- series of patients with 1, 2, or 3 level disease reported that
cations with cervical arthroplasty. J Neurosurg Spine; 8% of segments had grade I, 39% of segments had grade II,
4:98–105. 10% had HO leading to restricted movement and 9% experi-
5. Pimenta L. McAfee PC, Cappuccino A (2004) Clinical enced spontaneous fusion.
experience with the new artificial cervical PCM (Cer-
vitech) disc. Spine J; 4(suppl 6):315–321. There are several factors which may partially explain dis-
6. McAfee PC, Cunningham BW, Devine J, et al (2003) crepancies in HO rates across studies, aside from differences
Classification of heterotopic ossification (HO) in arti- in length of follow-up. It is possible that increased attention
ficial disk replacement. J Spinal Disord Tech; 16:384– to radiographic detail may go hand in hand with increased
389. reporting. Grading the severity of HO is not an exact science.
7. Mehren C, Suchomel P, Grochulla F, et al (2006) Hetero- There is some subjectivity which may translate into difficulty
topic ossification in total cervical artificial disc re- in distinguishing between adjacent grades (eg, between
placement. Spine; 31:2802–2806. grades 2 and 3) and overlap in classification. The grading of
8. Heidecke V, Burkert W, Brucke M, et al (2008) Interver- HO has so far not been subjected to inter- and intra-observer
tebral disc replacement for cervical degenerative dis- reliability evaluation and is not known to correlate with any
ease—clinical results and functional out come at two health-related quality of life (HrQoL) outcomes measures.
years in patients implanted with the Bryan cervical
disc prosthesis. Acta Neurochir;150:453–459.
9. Leung C, Casey ATH, Goffin J, et al (2005) Clinical sig- Study specifications and protocols: Use of structured
nificance of heterotopic ossification in cervical disc measures and protocols in a prospective study, which are con-
replacement: a prospective multi center clinical trial. sistently applied to all patients, is important to decrease study
Neurosurgery; 57:759–763. bias. While the authors suggest that structured protocols were
10. Sola S, Hebecker R, Knoop M, et al (2005) Bryan cervi- used, no details were provided. Reporting detail about proto-
cal disc prosthesis—three years follow-up. Eur Spine cols assists in determining the extent to which various factors
J; 14 (suppl 1):38. may or may not have influenced the results. For studies of
HO, important protocol details should include whether or not
bone waxing was done at the osteotomy site as well as specif-
ics for radiographic measurements and use of antiinflamma-
tory medications. Description of post-operative rehabilitation
is also important. It is unclear whether early initiation of
Editorial staff perspectives range of motion creates an inflammatory response which
This is a CoE III prognostic study. contributes to HO formation or if it facilitates functional
outcome.
Disc replacement offers a new interventional option and
long-term data on the potential complications to artificial Strengths: The use of validated outcomes measures (NDI
disc replacement compared with fusion are very much need- and SF-36) is commendable as is the authors’ acknowledg-
ed. This prospective study’s contribution to the literature re- ment of small sample size and a relatively low follow-up rate
garding heterotopic ossification (HO) following cervical ar- (77%) as study limitations.
Original research—Treating thoracic disc herniations: Do we always have to go anteriorly? 21
21—28
Treating thoracic-disc herniations:

Do we always have to go anteriorly?
Richard J Bransford1, Fangyi Zhang 2 , Carlo Bellabarba1, Michael J Lee3
Authors
Institutions 1
Harborview Medical Center, Department of Orthopedics and Sports Medicine, Seattle, WA , USA
2
Harborview Medical Center, Department of Neursosurgery, Seattle, WA, USA
3
University of Washington, Department of Orthopedics and Sports Medicine, Seattle, WA, USA
Abstract
of evidence (CoE)
Study design: Objective: To determine if there is a difference in outcome and complica-
Randomized contolled trial tions in surgically managed patients with thoracic-disc herniations
Cohort study • (TDH) undergoing a modified transfacet pedicle-sparing decompression
Case control and fusion (posteriorly) compared to those undergoing anterior transt-
Case series horacic discectomies (anteriorly).
Statement of concealed allocation*
Intent to treat * Methods: Thirty-five consecutive operatively managed TDH underwent
Independent or blind assessment • operative management between March 2003 and November 2009. Out-
Complete follow-up of ≥ 85% comes and complications were reviewed from patient records and x-rays
Adequate sample size assessing differences between those treated posteriorly and those treat-
Controlling for possible confounding ed anteriorly.
Evidence class: III
Results: Twenty-four patients underwent posterior management for 35
* Applies to randomized controlled trials only.
TDH and ten patients underwent anterior management for twelve TDH.
Mean age was 50 years in both groups. Body mass index (BMI) averaged
28.8 in the anterior group and 32.0 in the posterior group. Follow-up
averaged 38 weeks with four patients lost to follow-up (all posterior).
Major complications secondary to surgery occurred in three patients
(30%) in the anterior group (pulmonary embolus, pneumonia, and
wrong level surgery) and in seven patients (35%) in the posterior group
(seroma, misplaced instrumentation requiring revision, recurrence
requiring an additional operation, and four infections). No neurological
complications occurred and all patients noted improvement from base-
line. Average length of stay was 7.3 days in the anterior group and 4.2
days in the posterior group (P < .003). Final pain as assessed by visual
analog scale (VAS) improved from 6.7 to 4.3 in the anterior group and
6.9 to 2.3 in the posterior group (P = .05).
Conclusions: Complication rates are similar between groups and are ap-
proach related. Posteriorly managed patients had greater improvement
in pain and shorter length of stay.
No financial or any other support was received for this work. University of Washington IRB Approval # 37607

22 Original research—Treating thoracic disc herniations: Do we always have to go anteriorly?
StUdy RAtiONALE MEthOdS
Patients with symptomatic thoracic-disc herniations Study design: Retrospective cohort study
(TDH) not amenable to conservative measures have clas-
sically been treated with a thoracotomy and anterior dis- Inclusion criteria: All patients with TDH treated with
cectomy. A modified transfacet pedicle-sparing decom- either a modified transfacet pedicle-sparing decom-
pression and fusion has recently been proposed as an pression and fusion or an anterior thoracotomy
alternative option in the management of TDH [1]. A between March 2003 and November 2009.
comparison between anteriorly based and posteriorly
based approaches has not previously been done. Exclusion criteria: Patients with TDH who were treat-
ed operatively with other techniques such as
laminectomy or complete costotransversectomy
with corpectomy during this collection period
OBJECtivE were not included in this study.
The objective of this study is to determine whether there Patient population

is a difference in outcomes and complications in patients • Thirty-four consecutively managed patients with
treated with a posterior transfacet decompression and TDH met the criteria. Patients with radicular symp-
fusion compared to those treated with an anterior thora- toms and/or pure axial back pain had a minimum
cotomy and discectomy for symptomatic TDH. of 6 months of conservative therapy prior to surgery.
Twenty-four were treated posteriorly and ten were
treated anteriorly (Fig 1).
• Treatment technique was based on surgeon
preference and was not influenced by patient
demographics or herniation location or type. Of
eight fellowship trained spine surgeons, four used
an anterior approach and four used a posterior
approach (Fig 1).
Interventions
• The anterior technique consisted of a lateral trans-
thoracic approach through the chest in the lateral
position with the assistance of a thoracic access
surgeon in all cases but one. Eight of the ten patients
also underwent fusions; two had discectomies with-
out fusion.
• The posterior technique consisted of a modified
transfacet pedicle-sparing decompression and fu-
sion in the prone position as previously described
in detail by Bransford [1] in 24 patients. With this
technique, there is no retraction of the neural ele-
ments and no sacrifice of the nerve roots and the
pedicles are spared. All patients are instrumented
with posterior pedicle screws and an interbody
T-PLIF (Synthes, Paoli, PA) allograft placed into the
disc space.

Outcomes Analysis
• Major complications were defined as those requiring • Categorical baseline variables and complication
unanticipated additional surgery, infection, readmis- rates were compared using a Chi-square test.
sion, or life-threatening complications. • Changes from preoperative to postoperative ASIA
• P rimary outcomes included a change in neurologi- motor scores and VAS pain scores were compared
cal status as graded by the American Spinal Injury within and between treatment groups using a two
Association (ASIA) spinal cord injury grade and tailed t-test.
motor score and change in pain. • Other comparisons including length of hospital stay,
• Pain was graded using a visual analog scale (VAS) intensive care admission (ICU), and estimated blood
as part of the patient intake forms and was recorded loss (EBL) were analyzed using a two tailed t-test.
as a numerical number from 0–10. VAS was defined • We defined statistical significance as P < .05. Statisti-
as general body pain as opposed to specifying for cal analysis was performed using SAS 9.2 software
back pain, chest pain, or radicular pain. (SAS Inc., Cary, NC).
Fig 1 Patient sampling and selection flow chart
Group A Lost to follow-up (n = 4)

(Posterior) Reasons: Analyzed (n = 20)
(n = 24) 3 due to lengthy distance Excluded from
with follow-up with local analysis (n = 0)
spine surgeon, 1 with
serious mental health issues
requiring institutionalization
Assessed Enrollment Group or

for eligibility (n = 34) treatment Allocation Follow-up
assignment Analysis
(n = 39)
(based on individual
surgeons preference)
Excluded (n = 5)
Group B Lost to follow-up
Not meeting inclusion Analyzed (n = 10)
(Anterior) (n = 0)
criteria (n = 5) Excluded from
(n = 10)
analysis (n = 0)
Treated with a different
surgical technique
(ie, laminectomy or
costotransversectomy)

Table 1 Demographic characteristics comparing

Results
treatment groups at study entry
• M ean age (50 years), sex (70% male anterior / 58% Posterior (N = 24) Anterior (N = 10)
Mean (range) or n (%) Mean (range) or n (%) P-value
male posterior), BMI (28.8 anterior / 32 posterior),
and comorbidities were not statistically different be- Age (years) 50.4 (18–71) 49.9 (35–57) .92
tween groups (Table 1). Male 14 (58) 7 (70) .70
• The overall follow-up rate was 88% (30/34) with 83% BMI 32 (24–47) 28.8 (18–42) .25
follow-up in the posteriorly treated group versus
100% in the anteriorly treated group. The mean follow- Levels
up was 41 weeks (6–168) in the posteriorly treated T1–2 1 0 .51
group and 34 weeks (6–112) in the anteriorly treated T2–3 1 0 .51
group.
T3–4 0 0 1.0
• There was not a significant difference in EBL between
the two groups. Average length of stay was 7.3 ± 3.2 days T4–5 1 0 .51
with 1 ICU day in the anterior group and 4.2 ± 2.0 days T5–6 1 0 .51
(excluding two with unusual circumstances) with 0 ICU T6–7 2 3 .06
days in the posterior group (P < .003) (Table 2). T7–8 7 3 .67
• No patient had a worsening neurological exam post- T8–9 6 1 .32
operatively and most with a motor score less than
T9–10 3 3 .11
100 improved by 3.2–3.4 points (Table 3).
• VAS improved from a mean of 6.7 ± 1.4 preoperative- T10–11 3 1 .84
ly to 4.3 ± 2.5 at last clinic visit in the anterior group T11–12 10 1 .41
and 6.9 ± 3.2 preoperatively to 2.3 ± 2.0 at last clinic
Comorbidities
visit in the posterior group (P = .05 for change from
Morbid obesity 5 (21) 2 (20) 0.96
baseline to final follow-up between treatment
groups) (Table 3). Diabetes 3 (13) 2 (20) 0.62
• Major complications in those with follow-up occurred COPD* 2 (8) 2 (20) 0.56
in three (30%) of anteriorly treated patients and seven Mean follow-up 41 (6–168) 34 (6–112) 0.65
(35%) of posteriorly treated patients and appeared to (weeks)
be related mainly to approach. The types of complica- * Chronic obstructive pulmonary disease.
tions are outlined in Table 4.
Table 2 Immediate postoperative measures

comparing treatment groups
Posterior (N = 24) Anterior (N = 10) P-value
Measure Mean ± SD Mean ± SD
EBL Total (cc) 740 ± 812 691 ± 371 .86
EBL per level (cc) 493 ± 524 633 ± 346 .4
Length of stay 4.2 ± 2.0 * 7.3 ± 3.2 < .003
* Refer to online appendix for explanation of two patients excluded in

LOS from posterior group.
Table 3 Comparison of neurological and pain Fig 2 Preoperative CT myelogram in 46-year-old

outcomes comparing treatment groups woman with large calcified T8–9 thoracic disc
Neurological herniation and myelopathy who underwent posterior
pain improvement decompresion. Fig 2a sagittal cut and Fig 2b axial cut.
(vAS score) (ASiA score)
Posterior Anterior Posterior Anterior
N = 20 N = 10 N = 7* N = 6*
Baseline (points) 6.9 (± 3.2) 6.7 (± 1.4) 87.4 (± 9.9) 92.1 (± 3.9)
Follow-up (points) 2.3 (± 2.0) 4.3 (± 2.5) 90.6 (± 11.3) 95.5 (± 5.6)
Change (points) 4.6 2.4 3.2 3.4
Within group < .0001 .01 .67 .37
P-value†
Between group .05 .98
P-value‡
* Neurology was compared in patients with a motor score less than 100
(N = 13). The remaining patients had motor scores of 100
preoperatively and at final follow-up. a b
† P-value associated with change from baseline to 12 months within
each treatment group.
‡ P-value comparing baseline to 12 month changes between posterior
and anterior approaches. Fig 3a Intraoperative fluoroscopy images showing
endplate shaver used to prepare the disc space for the
graft and Fig 3b intraoperative lateral showing
placement of pedicle screws with T-PLIF allograft in
Table 4 Comparison of complication rates between disc space for patient presented in Fig 2.
treatment groups
posterior Anterior
N = 20 N = 10
n (%) n (%) P-value*
Number of patients 7 (35) 3 (30) .96
with complications
Infection† 5 (25) 0 .08
Pneumonia‡ 0 1 (10) .15
Wrong level surgery 0 1 (10) .15
Recurrence 1 (5) 0 .47
a b
Pulmonary embolism 0 1 (10) .15
Implant complication 1 (5) 0 .47
* Chi-square test. Fig 4 Postoperative CT scan demonstrating placement

† One patient with an infection developed osteomyelitis leading to a of instrumentation and placement of graft with
fracture which required a revision fusion.
‡ Pneumonia with 1.5 L effusion.
excision of calcified thoracic disc for patient presented
in Fig 2 and 3 (Fig 4a sagittal cut and Fig 4b axial cut).
a b

Fig 5 Preoperative MRI of 56-year-old woman with

diSCUSSiON
T9–10 thoracic disc herniation and myelopathy who
underwent an anterior decompression.
• Thoracic disc herniations are rare in comparison with
Fig 5a sagittal cut and Fig 5b axial cut
their cervical or lumbar counterparts and are thought
to comprise 0.1–4% of all disc herniations [2,3].
• Anterior transthoracic decompressions of thoracic
disc herniations are considered the gold standard
[4–7], but this has not been compared with a posterior
transfacet posterior decompression and fusion with
respect to outcomes and complications. There are a
limited number of retrospective case series discussing
operative management. [1, 8–18] (table 5).
• Strengths: This is the fi rst study comparing transtho-
racic anterior discectomies to posterior transfacet
pedicle sparing discectomies in the management of
thoracic disc herniations.
a b • Limitations: This is a retrospective study with a rela-
tively small sample of patients. There is the possibility
that the small sample size may have limited the power
Fig 6 Postoperative AP (Fig 6a) and lateral (Fig 6b) x-rays
to make meaningful comparisons, particularly of
demonstrating placement of anterior graft and
major complications. There is also the possibility that
instrumentation in patient with images presented in
the follow-up of 41 weeks in the posterior group versus
Fig 5.
34 weeks in the anterior group may have biased the
outcomes of VAS improvement and motor score
improvement. Another limitation is the 17% loss to
follow-up in the posterior group compared to 0% in
the anterior group; this unequal balance in loss to
follow-up may influence the outcomes if those lost
to follow-up were more likely to have improved or
have had poorer outcomes.
• Since individual procedures were based on surgeon
preference, there is the possibility of bias. However,
each of the eight surgeons chose only one of the
a b techniques which was their standard of care for
management of all thoracic disc herniations. Base-
line differences such as BMI, level of herniation,
Fig 7 Visualization of transfacetal thoracic discec- type of herniation, and comorbidities were unlikely
tomies can be enhanced by use of an arthroscope. to have confounded the interpretation of the out-
This patient received a thoracic discectomy at T7–8. come comparisons, though a stratified analysis or
Completeness of decompression was verified with a multiple regression was not possible to control for
70° arthroscope under dry technique. these factors due to the small sample size.
• Both techniques allowed for adequate decompression
and equal improvement neurologically. There was a
statistically longer length of stay in the anterior
group compared to the posterior group. There was
greater improvement in pain as measured by VAS in
the posterior group compared to the anterior group.
• Each technique appears more susceptible to compli-
cations related to the approach.

Table 5 Table of published retrospective case series describing operative management of thoracic disc herniations
Posterior Year N* Approach Complications n (%) Neuro deterioration N (%) Class of evidence
Maiman [8] 1984 23 Lateral extracavitary 0 (0%) 0 (0%) IV
Simpson [9] 1993 21 Costotransversectomy 0 (0%) 0 (0%) IV
Le Roux [10] 1993 20 Transpedicular 1 (5%) 0 (0%) IV
Levi [11] 1999 35 Transpedicular 2 (5.7%) 1 (2.9%) IV
Bilsky [12] 2000 20 Transpedicular 3 (15%) 0 (0%) IV
Bransford [1] 2010 18 Transfacet 6 (33%) 1 (5.5%) IV
Anterior Year N* Approach Complications n (%) Neuro deterioration N (%) Class of evidence
Otani [13] 1988 23 Transthoracic 0 (0%) 0 (0%) IV
Bohlman [14] 1988 19 Transthoracic 2 (11%) 2 (11%) IV
Fujimara [15] 1997 33 Transthoracic 2 (6%) 0 (0%) IV
Regan [16] 1998 29 Video assisted 4 (13.8%) 0 (0%) IV
Ayhan [17] 2010 27 Transthoracic 6 (21.4%) 2 (7.4%) IV
Combined series Year N* Approach Complications n (%) Neuro deterioration N (%) Class of evidence
Stillerman [18] 1998 71 82 disc herniations 12 (14.6%) 1 (1.4%) III
49 Transthoracic

23 Transfacet
8 Lateral extracavitary
2 Transpedicular
* Articles had to have a minimum of 15 patients in order to be included.
Summary and conclusion REFERENCES
• B oth treatments appear to improve pain and neuro- 1. Bransford R, Zhang F, Konodi M, et al (2010) Experi-
logical status. ence with treatment of thoracic disc herniation using
• There was a significantly shorter length of stay and a a modified transfacet pedicle-sparing decompression
substantial improvement in pain with the posterior and segmental fusion. Journal of Neurosurgery: Spine;
approach over the anterior approach. 12: 221–231.
• Complication rates are similar between techniques 2. Arce CA, Dohrmann GJ (1985) Herniated thoracic disks.
and are largely approach related. Infections appear Neurol Clin; 3:383–392.
to be more frequently associated with a posterior 3. Ridenour TR, Haddad SF, Hitchon PW, et al (1993) Her-
approach. niated thoracic disks: treatment and outcome. J Spinal
• Either technique is effective in decompressing the Disord; 6:218–224.
neural elements. Prospective comparative studies 4. Fessler RG, Sturgill M (1998) Review: complications of
with larger samples which are designed to limit con- surgery for thoracic disc disease. Surg Neurol; 49(6):
founding and bias are needed to further determine 609–618.
the superiority of one technique over the other. 5. McCormick WE, Will SF, Benzel EC (2000) Surgery for
thoracic disc disease. Complication avoidance: over-
view and management. Neurosurg Focus; 9(4): e13.
6. Mulier S, Debois V (1998) Thoracic disc herniations:
transthoracic, lateral, or posterolateral approach? A
review. Surg Neurol; 49(6):599–606; discussion 606–608.

7. Vollmer DG, Simmons NE (2000) Transthoracic ap- Editorial staff perspectives

proaches to thoracic disc herniations. Neurosurg Focus; This is a CoE III treatment study.
9(4): e8.
8. Maiman DJ, Larson SJ, Luck E, et al (1984) Lateral Bransford et al compare approaches to treating thoracic disc
extrac avitary approach to the spine for throacic disc herniation, which is rare relative to cervical and lumbar her-
herniation: report of 23 Cases. Neurosurgery; 14(2): niations. This study provides an example of the challenges
178–182. related to studying rare conditions, the primary and most ob-
9. Simpson JM, Silveri CP, Simeone FA, et al (1993) Thoracic vious being small numbers of patients available for study
disc herniation. Re-evaluation of the posterior approach and the long period of time required to accumulate enough
using a modified costotransversectomy. Spine; 18(13): cases to study. These factors make prospective studies (ran-
1872–1877. domized controlled trials or traditional prospective cohort
10. Le Roux PD, Haglund MM, Harris AB (1993) Thoracic studies) more difficult to design and implement, and use of
disc disease: Experience with the transpedicular retrospective cohort studies or other study designs more ap-
approach in twenty consecutive patients. Neurosur- pealing and feasible.
gery; 33(1):58–66.
11. Levi N, Gjerris F, Dons K (1999) Thoracic disc hernia- Timing: It may take a long time for a single surgical center to
tion. Unilateral transpedicular approach in 35 con- accrue a sufficient number of cases; however, changes in tech-
secutive patients. J Neurosug Sci; 43(1):37–42. nology, treatment options and perspectives continue over that
12. Bilsky MH (2000) Transpedicular approach for tho- time period and may be rapid. Endoscopic transthoracic dis-
racic disc herniations. Neurosurg Focus; 9(4):e3. cectomy provides an example: After initial enthusiasm for it
13. Otani K, Yoshida M, Fujii E, et al (1988) Thoracic disc as a less invasive, more benign alternative to open transtho-
herniation. surgical treatment in 23 patients. Spine; racic discectomy, lack of comparative studies or evidence of its
13(11):1262–1267. effectiveness and safety in real life applications resulted in a
14. Bohlman HH, Zdeblick TA (1988) Anterior excision of rapid decline in its popularity. Thus, it may have been of
herniated thoracic discs. JBJS (Am); 70(7): 1038–1047. questionable value to have this as an intervention arm in a
15. Fujimura Y, Nakamura M, Matsumoto M (1997) Anterior lengthy or time-sensitive study.
decompression and fusion via the extrapleural ap-
proach for thoracic disc herniation causing myelopathy. Study design: With rare conditions such as TDH, or other
Keio J Med; 46(4):173–176. instances where randomized controlled trials are not feasible
16. Regan JJ, Ben-Yishay A, Mack MJ (1998) Video-assisted or ethical, outcomes from methodologically rigorous nonran-
thoracoscopic excision of herniated thoracic disc: domized cohort studies may provide the best approximation
description of technique and preliminary experience of what might be observed in RCTs. Such studies can take
in the first 29 cases. J Spinal Disord; 11(3): 183–191. surgeon training and preferences and patient preferences
17. Ayhan S, Nelson C, Gok B, et al (2010) Transthoracic into account. These studies must, however, have carefully de-
surgical treatment for centrally located thoracic disc fined exclusion criteria, and document and consider prognos-
herniations presenting with myelopathy: a 5-year in- tic factors. In addition, the patient and surgeon preferences
stitutional experience. J Spinal Disord Tech (e-pub ahead must be well understood [King]. Regardless of the study de-
of print). sign chosen, attention must be given to methods of reducing
18. Stillerman CB, Chen TC, Couldwell WT, et al (1995) Ex- bias and accounting for potentially confounding factors.
perience in the surgical management of 82 symptom-
atic herniated thoracic discs and review of the litera- While single studies like the one by Bransford will not change
ture. J Neurosurg; 83(6):971–976. the current standard of care (anterior transthoracic surgery)
for symptomatic TDH causing cord compression, it provides a
foundation for further study. A multi-center, collaborative
study may provide an opportunity to aggregate a larger num-
ber of patients to further evaluate treatment options for TDH.
Reference
King M, Nazareth I, Lampe F, et al (2005) Impact of par-
ticipant and physician intervention preferences on
randomized trials: a systematic review. Jama; 293:
1089–1099.
Original research—Fueling the debate: Are outcomes better after posterior lumbar interbody fusion (PLIF) or (…) 29
29—34
Fueling the debate: Are outcomes better after

posterior lumbar interbody fusion (PLIF) or
after posterolateral fusion (PLF) in adult
patients with low-grade adult isthmic
spondylolisthesis?
Authors G Barbanti Bròdano1, F Lolli 2, K Martikos2, A Gasbarrini1, S Bandiera1, T Greggi 2, P Parisini†2, S Boriani1
Institution 1 Istituti Ortopedici Rizzoli, Oncologic and Degenerative Spine Department, Bologna, Italy
2
Istituti Ortopedici Rizzoli, Deformity Spine Department, Bologna, Italy
Abstract
Clinical question: Do more adult patients affected by low grade isthmic

spondylolisthesis have significant clinical and radiological improve-
ment following posterior lumbar interbody fusion (PLIF) than those
who receive posterolateral fusion (PLF)?
Methods: One hundred and fourteen patients affected by adult low grade
isthmic spondylolisthesis, treated with posterior lumbar interbody fu-
sion or posterolateral fusion, were reviewed. Clinical outcome was as-
sessed by means of the questionnaires ODI, RMDQ and VAS. Radio-
graphic evaluation included CT, MRI, and x-rays. The results were
analyzed using the Student t-test.
Results: The two groups were similar with respect to demographic and sur-
gical characteristics. At an average follow-up of 62.1 months, 71 pa-
tients were completely reviewed. Mean ODI, RMDQ and VAS scores
didn’t show statistically significant differences. Fusion rate was similar
between the two groups (97% in PLIF group, 95% in PLF group). Major
complications occurred in 5 of 71 patients reviewed (7%): one in the
PLIF group (3.6%), four in the PLF group (9.3%). Pseudarthrosis oc-
curred in one case in the PLIF group (3,6%) and in two cases in PLF
group (4.6%).
Conclusions: In our series, there does not appear to be a clear advantage of

posterior lumbar interbody fusion (PLIF) over posterolateral fusion
(PLF) in terms of clinical and radiological outcome for treatment of
adult low grade isthmic spondylolisthesis.
This research has received no financial support.

30 Original research—Fueling the debate: Are outcomes better after posterior lumbar interbody fusion (PLIF) or (…)
STUDY RATIONALE AND CONTEXT METHODS
The choice of correct surgical treatment of adult low- Study design: Retrospective cohort study.
grade isthmic spondylolisthesis remains a topic of de-
bate. Many studies in the literature analyze clinical and Inclusion criteria: All adult patients who had under-
radiological outcome of different fusion techniques by gone posterior lumbar interbody fusion (PLIF) (Figs
various approaches, including posterolateral fusion 2, 3) or posterolateral fusion (PLF) (Fig 4) for low
(PLF) and lumbar interbody fusion, but considerable grade isthmic spondylolisthesis (Meyerding grade
controversies regarding what is the “gold standard” ap- 1 or 2) between February 2003 and April 2005, and
proach still exist [1–16]. who had a minimum of 4 years of follow-up.
Exclusion criteria: Previous spine surgery, age less

than 40 years, etiology other than isthmic, high-
CLINICAL QUESTION grade spondylolisthesis, concomitant conditions
which could compromise outcomes.
Do more adult patients affected by low grade isthmic Patient population and interventions compared (Fig 1):
spondylolisthesis have significant clinical and radiologi- • One-hundred-and-fourteen consecutive patients
cal improvement following posterior lumbar interbody met the inclusion criteria, and were divided into
fusion (PLIF) than those who receive posterolateral fu- two groups, according to the surgical treatment
sion (PLF)? they received: PLIF group (posterior lumbar inter-
body fusion) and PLF group (posterolateral fusion)
(Table 1). Patients were evaluated preoperatively,
postoperatively and at final follow-up.
• At the time of surgery all patients complained of
low back and leg pain.
• Posterior pedicle screw instrumentation alone was
used as support to fusion in the PLF group. Carbon
fiber, titanium and peek cages were added in the
PLIF group. A laminectomy was performed in all
cases. All patients received allograft bone and au-
tograft bone obtained from decompression.
Outcome and analysis:

• Demographic, preoperative, perioperative and
postoperative data were collected.
• Clinical outcome was assessed by means of the Os-
westry disability index (ODI), Roland Morris Dis-
ability Questionnaire (RMDQ) and visual ana-
logue scale (VAS), for back and leg pain
respectively, filled in by patients preoperatively
and at last follow-up.
• R adiographic evaluation included preoperative CT
(performed to assess the isthmic nature of the le-
sion) and MRI of the lumbar spine, as well as
standing plain and functional films with flexion
and extension views before and after surgery and
during the follow-up, when requested. Fusion was
defined as radiographic evidence of bone bridging,
the absence of lucency around the implant, and
no motion during functional films.
• Overall complications were noted. Major compli-
cations were those that needed revision surgery or
resulted in permanent neurological deficit.
• T
he results were analyzed using the Student t-test.
Results are expressed as the mean (range), with a
P-value of < 0.5 considered as being statistically
significant.
Fig 1 Patient sampling and selection
Group PLF Lost to follow-up Analyzed

(n = 67) (n = 24): (n = 43)
Received allocated Reasons:
intervention (n = 67) Failed to show (n = 24)
Did not receive allocated
intervention (n = 0)

for eligibility (n = 114) treatment Allocation Follow-up Analysis
(n = 205) assignment
Excluded (n = 91)
Reasons: Group PLIF Lost to follow-up Analyzed

previous spine surgery (n 13) (n = 47) (n = 19): (n = 28)
age < years (n = 35) Received allocated Reasons:
etiology other than isthmic (n = 27) intervention (n = 47) Death (n = 1)
high grade spondylolisthesis (n = 12) Did not receive allocated Failed to show (n = 18)
severe osteoporosis (n = 4) intervention (n = 0)
Table 1 Characteristics of intervention groups

RESULTS
PLIF group PLF group
• T he two groups were similar with respect to demo- All enrolled Patients at All enrolled Patients at
graphic and surgical characteristics (Table 1). N = 47 follow-up N = 67 follow-up
N = 28 N = 43
• At an average follow-up of 62.1 months (range 51–
78), 71 patients (62.3%), 28 (59.6%) of the PLIF Age, years (mean 54.8 ±8.6 55.1 ±9.2 51.6 ±8.6 49.3 ±7.4
±SD)
group and 43 (64.2%) of the PLF group, were com-
pletely reviewed. n (%) n (%) n (%) n (%)
• Clinical outcome. Both techniques ensured improve- Female gender 25 (53.2) 14 (50.0) 39 (58.2) 24 (55.8)
ment of clinical outcome, without statistically signif- Spondylolisthesis 21 (44.7) 12 (42.9) 28 (41.8) 16 (37.2)
icant differences between the two groups (P > .05). grade I
Unsatisfactory clinical results were achieved in four Spondylolisthesis 26 (55.3) 16 (57.1) 39 (58.2) 28 (65.1)
patients (14.3%) in the PLIF group and in eight pa- grade II
tients in the PLF group (18.6%) (Table 2). L3–4 3 (6.4) 2 (7.1) 0 (0) 0 (0)
• R adiologic outcome. The x-rays performed at final spondylolisthesis
follow-up showed a fusion rate of 97% in the PLIF L4–5 16 (34.0) 14 (50.0) 25 (37.3) 14 (32.6)
group, 95% in the PLF group, without statistically spondylolisthesis
significant differences (P > .05). L5–S1 28 (59.6) 12 (42.9) 42 (62.7) 29 (67.4)
spondylolisthesis
1 level fusion 30 (63.8) 19 (67.9) 50 (74.6) 32 (74.4)
2 level fusion 13 (27.7) 7 (25.0) 16 (23.9) 11 (25.6)
3 or more level fusion 4 (8.5) 2 (7.1) 1 (1.5) 0 (0)

Table 3 Major complications requiring revision

• C
omplications. Complications requiring revision sur-
surgery
gery occurred in 5 of 71 patients reviewed (7%), one
in the PLIF group (3.6%) and four in the PLF group PLIF group PLF group
(n = 28) (n = 43)
(9.3%). Pseudarthrosis occurred in one case in the n (%) n (%)
PLIF Group, in two cases in the PLF group (Table 3).
Major complications, n (%)
More information on complications is available in the Revision surgery 1 (3.6)* 4 (9.3)†

web appendix at www.aospine.org/ebsj. Pseudarthrosis 1 (3.6) 2 (4.6)
* Revision due to pseudarthrosis

† Two revisions due to pseudarthrosis
Table 2 Clinical outcome
PLIF group PLF group

Between
Percent Within group Percent Within group group
Baseline Follow-up change P-value* Baseline Follow-up change P-value* P-value†
ODI (mean % ±SD) 53.2 ±18.8 25.6 ±18.1 57.7 ±24.4 < .05 52.1 + 19.0 24.5 + 18.3 59.2 + 24.9 < .05 > .05
RMDQ 13.9 ±6.1 7.2 ±6.2 57.3 ±26.5 < .05 13.4 ±6.2 6.9 ±6.3 58.3 ±27.4 < .05 > .05
VAS “leg score” 7.4 ±1.4 4.1 ±2.8 49.3 ±30.0 < .05 7.6 ±1.4 3.5 ±3.2 58.5 ±33.6 < .05 > .05
VAS “back score” 7.7 ±1.3 3.0 ±2.0 62.8 ±21.8 < .05 7.8 ±1.3 3.8 ±2.7 55.3 ±29.1 < .05 > .05
Persistent low-back 100% 14.3% n.a. n.a. 100% 18.6% n.a. n.a. n.a.
pain (%)
Persistent sciatica 100% 3.6% n.a. n.a. 100% 11.6% n.a. n.a. n.a.
* P-value associated with change from baseline to follow-up in each treatment group
† P-value comparing change in baseline to follow-up between PLIF and PLF groups
Discussion
• I n our series, there does not appear to be a clear ad-

vantage of posterior lumbar interbody fusion over
posterolateral fusion in terms of clinical and radio-
logical outcome.
• A higher incidence of complications requiring surgi-
cal revision (9.3% versus 3.6%) was found in the PLF
group. Pseudarthrosis occurred in one case in the
PLIF group (3.6%) and in two cases in the PLF group
(4.6%).
• Despite nerve root manipulation required to insert
the cages into the intervertebral space, in our series
we found only one case of sciatica at last follow-up in
the PLIF group.
• L imitations. The present series should be interpreted
in the context of its limitations, including the retro-
spective nature of the review, the fact that patients
were not randomized between posterior lumbar in-
terbody fusion and posterolateral fusion, the low fol-
low-up rate and the small sample size.
Fig 2 Posterior lumbar interbody fusion for low grade

CONCLUSIONS
isthmic spondylolisthesis, 3 and 15 months after
surgery. Note the segmental sagittal alignement
• I n case of adult low grade isthmic spondylolisthesis,
(kyphosis) that could compromise long term clinical
posterior lumbar interbody fusion doesn’t seem to
and radiographic outcome (risk of negative effect on
provide advantages in terms of mechanical stability
adjacent disc).
and fusion rate (pseudarthrosis incidence: 3.6%
verses 4.6%).
• I n our series, both treatments ensured good clinical
results, without statistically significant differences
between the two techniques.
References
1. Möller H, Hedlund R. (2000) Surgery versus conserva-

tive management in adult isthmic spondylolisthesis - a
prospective randomized study: part 1. Spine;
1;25(13):1711–1715.
2. Kaneda K, Satoh S, Nohara Y, et al. (1985) Distraction
Fig 3 CT scan 15 months after surgery rod instrumentation with posterolateral fusion in isth-
mic spondylolisthesis. 53 cases followed for 18–89
months. Spine; 10(4):383–389.
3. Hanley EN, Jr, Levy JA. (1989) Surgical treatment of isth-
mic lumbosacral spondylolisthesis. Analysis of variables
influencing results. Spine; 14(1):48–50.
4. Boos N, Marchesi D, Heitz R, et al. (1992) [Surgical treat-
ment of spondylolisthesis with mild displacement by
pedicular fixation and posterolateral fusion in adults].
Rev Chir Orthop Reparatrice Appar Mot; 78(4):228–235. Re-
view. French.
5. Thalgott JS, Sasso RC, Cotler HB, et al. (1997) Adult
spondylolisthesis treated with posterolateral lumbar fu-
Fig 4 Posterolateral fusion for low grade isthmic sion and pedicular instrumentation with AO DC plates. J
spondylolisthesis. Note the postoperative disc height Spinal Disord.; 10(3):204–208.
gain that could compromise fusion. 6. Nooraie H, Ensafdaran A, Arasteh MM (1999) Surgical
management of low-grade lytic spondylolisthesis with
C-D instrumentation in adult patients. Arch Orthop Trau-
ma Surg; 119(5–6):337–339.
7. McGuire RA, Amundson GM (1993) The use of primary
internal fixation in spondylolisthesis. Spine;
15;18(12):1662–1672.
8. Thomsen K, Christensen FB, Eiskjaer SP, et al (1997)
1997 Volvo Award winner in clinical studies. The effect
of pedicle screw instrumentation on functional outcome
and fusion rates in posterolateral lumbar spinal fusion: a
prospective, randomized clinical study.
Spine; 15;22(24):2813–2822.
9. France JC, Yaszemski MJ, Lauerman WC, et al (1999) A
randomized prospective study of posterolateral lumbar
fusion. Outcomes with and without pedicle screw in-
strumentation. Spine; 15;24(6):553–560.

10. Möller H, Hedlund R (2000) Instrumented and nonin- follow-up may have different clinical or other characteristics
strumented posterolateral fusion in adult spondylolis- (and outcomes!) that could influence the evaluation of the
thesis—a prospective randomized study: part 2. Spine; study outcome and thus bias results. For example, if those
1;25(13):1716–1721. who are lost to follow-up are more likely to have a good out-
11. Ekman P, Möller H, Tullberg T, et al (2007) Posterior come for a one of the treatments, the analysis would not po-
lumbar interbody fusion versus posterolateral fusion in tentially include as many patients with a good outcome for
adult isthmic spondylolisthesis. Spine; 15;32(20):2178– that treatment and the results may be biased to show that it is
2183. less effective than its comparator.
12. Suk SI, Lee CK, Kim WJ, et al (1997) Adding posterior
lumbar interbody fusion to pedicle screw fixation and Treatment allocation: How treatment was allocated was
posterolateral fusion after decompression in spondylolyt- not well described in this paper, ie, what factors determined
ic spondylolisthesis. Spine; 15;22(2):210–219; discussion whether a patient received PLIF versus PLF aside from what
219–220. appears to be institutional preference (see web appendix).
13. Kim NH, Lee JW (1999) Anterior interbody fusion ver- Ideally, patients would be randomized to treatment groups
sus posterolateral fusion with transpedicular fixation for using an appropriate method of concealed allocation. It is
isthmic spondylolisthesis in adults. A comparison of common for studies to describe treatment allocation based on
clinical results. Spine; 15;24(8):812–816; discussion 817. surgeon preference or patient presentation. This has the po-
14. Swan J, Hurwitz E, Malek F, et al (2006) Surgical treat- tential to bias study results. For example if patients with more
ment for unstable low-grade isthmic spondylolisthesis in severe disease are more likely to receive one treatment over
adults: a prospective controlled study of posterior instru- the other and also have the potential for worse outcomes, the
mented fusion compared with combined anterior-poste- results may not be an accurate reflection of either treatment
rior fusion. Spine J; 6(6):606–614. Epub 2006 Oct 2. in patients with the same disease severity. Allocation based
15. Videbaek TS, Christensen FB, Soegaard R, et al (2006) on the institution’s preference may also bias results as other
Circumferential fusion improves outcome in comparison factors may also differ across institutions. Factors such BMI
with instrumented posterolateral fusion: long-term re- and previous surgery may influence choice of procedure and
sults of a randomized clinical trial. Spine; 1;31(25): 2875– therefore outcomes and need to be described.
2880.
16. Carragee EJ (1997) Single-level posterolateral arthro- Retrospective versus prospective approaches: In this
desis, with or without posterior decompression, for the study (and most retrospective studies), it isn’t clear that a con-
treatment of isthmic spondylolisthesis in adults. A pro- sistent perioperative protocol (for clinical care or outcomes
spective, randomized study. J Bone Joint Surg Am; measurement) was used in both study groups. With prospec-
79(8):1175–1180. tive study design, there is the potential to decrease study bias
compared with retrospective designs. Protocols for patient se-
lection and treatment allocation, perioperative care, collec-
tion of data and follow-up that are specified prospectively
help assure less biased allocation of patients to treatment and
similarity of care and measurement for both groups.
Editorial staff perspectives
This is a CoE III treatment study. Outcomes: Definition and evaluation of fusion status is
long-held area of controversy. In this study, it is unclear how
Comparing outcomes from patients treated with PLIF with fusion was determined and if its assessment was indepen-
those treated with PLF is a commendable goal and important dent. Factors such as use of BMP or grafts which may influ-
in the debate about the best treatment options for low-grade ence fusion and functional outcomes need to be detailed and
adult spondylolisthesis. In order to improve the quality of evi- evaluated for their potential to influence the outcomes. In ad-
dence available to settle the debate, future studies need to addition factors such as reduction of deformity, disc height and
dress a number of key factors. restoration of lordosis should be evaluated.
Patient selection: Methodologically, selecting patients based Final comments: This study’s use of validated outcomes
on the completeness of follow-up at a specific time or studies measures and length of follow-up are two primary strengths.
where > 85% are lost to follow-up creates the possibility of The authors’ acknowledgement of the significant loss to fol-
selection bias. By selecting patients with a minimum of lowup and limitations imposed by retrospective, nonrandom-
4-years follow-up, it is possible that those with less complete ized studies is commendable.
Original research—The effect of body mass index on lumbar lordosis on the Mizuho OSI Jackson spinal table 35
35—40
The effect of body mass index on lumbar

lordosis on the Mizuho OSI Jackson spinal table
Authors Justin Bundy, Tommy Hernandez, Haitao Zhou, Norman Chutkan
Institution Orthopaedic Department, Medical College of Georgia, Augusta, Georgia, USA
Abstract
of evidence (CoE)
Study design: Prospective cohort study.
Study design: Clinical question: Does the patients’ body mass index (BMI) influence the
Prospective cohort • degree of intraoperative lumbar lordosis in patients undergoing opera-
Retrospective cohort tive treatment on the Mizuho Orthopedic Systems Incorporated (OSI)
Case control Jackson spinal table?
Case series
Methods Methods: Twenty-four consecutive patients undergoing posterior spinal
Patients at similar point in course of • instrumentation and fusion on the Jackson table, excluding those with
treatment sagittal malalignment, underwent standing preoperative and prone in-
Follow-up ≥ 85% • traoperative lateral x-rays. Intervertebral body angle measurements
Similarity of treatment protocols for • were obtained from L1–S1 using the modified method of Cobb. Changes
patient groups in angle measurements were compared to BMI using linear regression
Patients followed for long enough for • and ANOVA.
outcomes to occur
Control for extraneous risk factors Results: We found a mean lordosis of 52.6° in standing preoperative x-rays
Evidence class: II compared to a prone position mean lordosis of 61.5° on the Jackson ta-
ble. The mean change was 8.88° with a range of 0°–18°. A linear associ-
* Authors must provide a description of robust ation between lordosis and BMI was demonstrated (P < .0022). As BMI
baseline characteristics, and control for those increased, so did lordosis (correlation coefficient, 0.59).
that are potential prognostic factors.
Conclusions: The current study is the first in which a correlation of patient

body mass and use of the Jackson table has been evaluated. These data
suggest that BMI influences lumbar lordosis on the Jackson table and
that care must be used when dealing with a population with large BMI
on the Jackson table.
There is no conflict of interests related to this article.

No financial support was received for this research.
This is study is approved by IRB.

36 Original research—The effect of body mass index on lumbar lordosis on the Mizuho OSI Jackson spinal table
STUDY RATIONALE AND CONTEXT METHODS
Prone patient position can influence lumbar spine surgi- Study design: Prospective cohort study.
cal techniques and reconstruction results due to changes
in spinal alignment. Studies have investigated patient Inclusion criteria (Fig 1): From July 2005 to December
positioning, especially in regard to various operative 2005, all patients who underwent instrumented
frames and their effect on sagittal alignment. In general, posterior lumbosacral fusion by the corresponding
mechanical decompression of the abdomen is desirable author were included in the study.
for spinal procedures carried out in a prone position in
order to decompress the epigastric plexus and hopefully Exclusion criteria: Patients with preoperatively pres-
thus diminish epidural bleeding. Prolonged prone posi- ent sagittal or coronal plane deformities greater
tion without external mechanical decompression of the than 10° in either coronal or sagittal direction diag-
lower torso could also lead to damage to internal organs. nosed in the preoperative workup were excluded.
These concerns are amplified in an overweight patient Patients with previous lumbar fusion or spon-
population. Unfortunately, there is an absence of infor- dylolisthesis were excluded.
mation regarding patient body habitus as it relates to in-
traoperative alignment of spinal surgery done in a prone Outcomes and prognostic (risk) factors to be evaluated:
position. • During preoperative evaluation, all patients were
weighed on a single digital scale and measured us-
ing a wall tape by the same clinical nurse. BMI was
determined by taking weight over the square of
CLINICAL QUESTION height (kg/m2) [1].
• On the preoperative standing lateral and intraop-
Does a patient’s body mass index (BMI) affect lumbar erative prone lateral x-rays, lumbar lordosis mea-
lordosis of patients undergoing lumbosacral posterior fu- surements were performed from L1–S1 using the
sion surgery in a prone position on a commonly used modified method of Cobb with images centered on
spinal table (Mizuho Orthopedic Systems Incorporated the vertebral body of L3 (Fig 2a–b).
(OSI) Jackson table)? • T wo examiners performed measurements. All
measurements were completed in each patient
by one examiner to maintain consistency.
Analysis:
• I ntraobserver measurement error was evaluated
and found to be less than 3°.
• Statistical analysis was performed using the paired
t-test, ANOVA and linear regression.
Additional information is available in the web appen-

dix at www.aospine.org/ebsj.
RESULTS
• T
he mean BMI in this population was 32.5 (± 4.4) and
88% of patients were considered overweight or obese
(Table 1).
• The mean lumbar lordosis angle from L1 to the sa-
crum with subjects in a standing position was 52.6°
(35°–75°) The mean lumbar lordosis on the Jackson
table was 61.5° (38°–80°) (Table 2).
Fig 1 Patient sampling and selection Table 1 Patient BMI distribution in this population
Adult BMI Category [1] n (%)

Total patients receiving posterior Underweight (< 18.5) 0
lumbosacral fusion (n = 32)
Normal (18.5—24.9) 2 (8.3)
Sagittal or coronal plane deformity
Overweight (25.0–24.9) 3 (12.5)
greater than 10° (n = 8)
Obese (≥ 30) 19 (79.2)
Patients available
for analysis (n =24)
Fig 2a–b Table 2 Standing and prone lumbar lordosis, change

a Standing preoperative lateral lumbar spine film of lordosis and BMI of each patients
b Intraoperative fluoroscopic film in the prone Standing Prone Change in
position showing an increase in lumbar lordosis of 5° Patient lordosis lordosis lordosis BMI
1 58 62 4 28
2 60 68 8 34
3 75 80 5 32
4 50 60 10 32
5 60 76 16 36
6 41 41 0 30
7 51 53 2 32
8 50 58 8 22
9 45 58 13 46
10 47 50 3 30
11 44 56 12 31
12 50 52 2 30
a b 13 35 38 3 30
14 47 65 18 34
15 60 75 15 37
16 58 74 16 34
17 60 77 17 49
18 67 69 2 25
19 54 68 14 42
20 54 68 14 29
21 56 60 4 31
22 42 52 10 29
23 50 62 12 34
24 50 55 5 24
Mean (± sd) 52.6 (± 6.9) 61.5 (± 8.8) 8.9 (± 5.0) 32.5 (± 4.4)
sd = standard deviation

• T
he increase in lordosis was statistically significant. Discussion
Measurements of total lordosis preoperatively and
postoperatively by the same observer were very re- • I n lumbar spine posterior fusion surgery it is desirable
producible and not significantly different. to maintain or achieve physiologic lordosis. This ef-
• A linear association between increasing BMI and in- fort can be helped by thoughtful intraoperative posi-
creasing lordosis was seen (P = .00215). An R-squared tioning and patient selection. Inadequate restoration
value of 0.345 suggests that 35% of the change in an- of sagittal balance has been implicated as a factor in
gle noted may be due to BMI (Fig 3). post fusion surgery low back pain [2, 3].
• Long-term studies have shown that excessive kypho-
sis of the lumbar spine causes a flat back syndrome
Fig 3 Scatter plot of BMI measurements and change in and compensatory hyperlordosis below the levels may
lordosis from standing to prone position, derived from predispose patients to accelerated degenerative
linear regression, showing strong correlation between changes [4–6].
change and increasing BMI • Conversely, fusion in lumbar hyperlordois can lead to
increased compensatory malalignment of other
20 levels.
18 • A statistically significant increase in lumbar lordosis
16 in patients with increasing BMI was seen. It appears
that this is caused by the combination of increased
Change in degrees
14
patient trunk weight and the table configuration,
12 which leaves the prone patient’s lower torso half sus-
10 pended. To our knowledge this finding has not been
8 reported before.
• Care must be taken when selecting overweight or
6
obese patients for positioning on a Jackson spinal ta-
4 ble due to its effects on increasing lumbar lordosis. Ef-
2 forts to restore physiologic sagittal balance of the spi-
0 nal column can include preoperative repositioning of
0 5 10 15 20 25 30 35 40 45 50 55 patients and intraoperative corrective measures.
BMI
• Clinical correlations of iatrogenically induced hyper-
lordosis and its longterm sequela in overweight pa-
tients using this positioning technique are not yet
fully understood.
• A s result of our study we have heightened our aware-
ness of lumbar alignment changes in overweight pa-
tients receiving spinal fusion surgery in a prone posi-
tion. We have expanded our study focus to include a
variety of spinal table alternatives and patient weight
categories (Figs 4, 5).
SUMMARY AND CONCLUSIONS: KEY POINTS
• O
verweight or obese patients have a significant radio-
graphic and possibly clinical lordotic change of align-
ment when placed prone on a Jackson table.
• The increase in lumbar lordosis on the Jackson table
with increasing BMI noted in our study should be
considered by an operating surgeon during posterior
lumbar fusion surgery as it might affect postoperative
lumbar alignment and clinical outcome.
References 4. Davies AG, McMaster MJ (1992) The effect of Luque

rod instrumentation on the sagittal contour of the
1. Centers for Disease Control and Prevention: Health lumbosacral spine in adolescent idiopathic scoliosis
weight. It’s not a diet, it’s a lifestyle. http://www.cdc. and the preservation of a physiologic lumbar lordosis.
gov/healthyweight/assessing/bmi/adult_bmi/; last accessed Spine; 17: 112–115.
March 11, 2010. 5. LaGrone MO (1988) Loss of lumbar lordosis: Compli-
2. Cochran T, Irstam L, Nachesmson A (1993) Long term cations of spinal fusion for scoliosis. Orthop Clin North
anatomic and functional changes in patients with ad- Am; 19: 383–393.
olescent idiopathic scoliosis treated by Harrington 6. Phillips WA, DeWald RL (1985) A comparison of Luque
rod fusion. Spine; 8: 577–584. segmental instrumentation with Harrington rod in-
3. Hayes M, Tomkins S, Herndon W, et al (1988) Clinical strumentation in the management of idiopathic sco-
and radiographic evaluation of lumbosacral motion liosis. Orthop Trans; 9:437–438.
below fusion levels in idiopathic scoliosis. Spine;
13:1161–1167.
Fig 4 Positioning overweight patients for prone spine Fig 5 An alternative to a Jackson table is the Mizuho
surgery is a well known and, as of yet, incompletely OSI Wilson frame which bends the trunk forward to
resolved challenge. While the primary concerns often ease surgical exposure. Depicted here is a Wilson
center around cardiopulmonary function for patients frame which affords free suspension of the abdominal
with large body mass, decompression of the epigastric structures, however is limited by the width of the
plexus by avoiding any direct pressure of the abdomi- aperture, which may be too narrow, as well as height.
nal contents is a major concern as well. The Mizuho Difficulties may also arise while positioning large
OSI Jackson spinal table offers excellent abdominal patients with short necks due to inability to achieve a
organ decompression as depicted, but may accentuate neutral neck position. Ultimately the choice of spinal
lumbar lordosis by anterior pull on the lumbar spine table in large patients is a compromise between
created by a large panniculus. spinal alignment needs and patient safety.

Editorial staff perspectives

This is a CoE II prognostic study.
This is a novel study which provides a valuable perspective

on the need to consider patient body habitus and its potential
impact on maintaining appropriate lordosis. Certainly the
finding that lumbar lordosis disproportionally increases in
patients with higher BMI’s when positioned prone on a Jack-
son spinal table, which leaves the abdomen freely suspended,
is noteworthy for intraoperative consideration. To further
evaluate this phenomenon and provide context for these find-
ings, a few methodological points deserve consideration:
What is a “strong” correlation? A correlation of 0.59 may not

be considered a “strong” correlation. The sample size is small
and addition of a correlation line to the scatter plot would
confirm that there is a lot of variation around it. Particularly
in a cohort where potentially confounding factors (eg, age,
sex) were not formally evaluated, the estimate of correlation
(and R-squared value reported) should be interpreted
cautiously.
Statistical methodology: While authors report an R-

squared based on linear regression, no information on the
regression model is provided. Details of the model used to gen-
erate the R² and P-value should be described. Is this based on
regression model that only has BMI in the model? If there
were other variables in the model, it should be stated what
was included (additional variables also influence R²). R² is
probably not a clinically meaningful number. It tells you that
for the particular model, a percent of the change is explained
by the combination of factors in the model...and the rest is not
explained by the model. R² is model dependent and there are
number of other aspects of the model (and fit) that need to be
considered.
Lack of comparators: Although the mean BMI of the pa-

tients is high (33), the evaluation was not exclusively done in
obese or overweight patients (there are some patients with
BMI < 30) and there isn’t a comparison of change in lordosis
among obese with non-obese patients, including those con-
sidered normal with respect to BMI. While BMI is a com-
monly used indicator of obesity, it measures total body mass
and doesn’t take into account lean muscle mass, which varies
for men and women and with age. Factors other than obesity
may affect lordosis, such as trunk length and ligamentous
laxity (eg, Marfan’s syndrome and Ehlers-Danlos patients).
These potential confounding factors should be considered in
further studies. To the extent that clinical factors may influ-
ence both BMI and change in lordosis, these factors should be
measured and evaluated.
Original research—Dynamic anterior cervical plating for multi-level spondylosis: Does it help? 41
41—46
Dynamic anterior cervical plating for

multi-level spondylosis: Does it help?
Authors Ashraf A Ragab, F Spain Hodges, Clint P Hill, Robert A McGuire, Michelle Tucci
Institution University of Mississippi Medical Center, Jackson, MS, USA
Abstract
of evidence (CoE)
Study design: Randomized controlled trial.
Study design: Objective: To compare fusion rates, time to fusion, complication rates and
RCT • subsidence between 1) a static, 2) a dynamic angulation, and 3) a dy-
Cohort namic translation plate in anterior cervical discectomy and fusion for
Case control symptomatic degenerative cervical disease.
Case series
Concealed allocation (RCT) Methods: Thirty-six patients with two level, symptomatic cervical degen-
Intent to treat (RCT) • erative changes requiring surgery were randomized in a blinded fash-
Blinded / independent evaluation of • ion to receive a statically locked plate, Cervical Spine Locking Plate
primary outcome (CSLP) (Synthes, Paoli, PN, USA), an Atlantis Vision® Anterior Cervical
Complete follow-up of ≥ 85%* • Plate System (Medtronic, Memphis, TN, USA) which allows angular
Adequate sample size dynamization, or a Premier® Anterior Cervical Plate System (Medtron-
Control for confounding† • ic) which allows translational dynamization. Structured data collection
Evidence class: II and measurement protocols were used. Intervertebral composite al-
lograft cages were used in all groups. Identical external immobilization
* Reliable data are data such as mortality or and antiinflammatory medication protocols were followed. X-rays were
reoperation. obtained at preset time points postoperatively. Assessment of the pri-
† Authors must provide a description of robust
mary outcomes was blinded. Rate of and time to fusion, graft/instru-
baseline characteristics, and control for those
that are unequally distributed between mentation complications, subsidence, and reoperation for adjacent level
treatment groups. disease were measured. Paired t-test and three-way Analysis of Vari-
ance test (ANOVA) were used to assess statistical differences between
groups.
of evidence is available on page 83. Results: The three groups were similar demographically. Fusion rates in
the CSLP, Atlantis and Premier plate groups were 100%, 91%, and 92%
respectively. Mean time to fusion was 6.1, 8.3 and 6.3 months respec-
tively but differences were not statistically significant. Mean subsidence
in the groups was 1.9, 1.6, and 2.6 mm respectively. Subsidence was
found even for the static (CSLP) plate, but no statistically significant dif-
ferences were found.
Conclusions: We found no clinical advantage of dynamic plates over static

plates with regards to fusion rates, time to fusion, subsidence, complica-
tions, or adjacent-level surgery. Static plating allows for subsidence at
similar levels to dynamic plating.
There is no financial disclosure for this study.

University of Mississippi IRB approval 2004.

42 Original research—Dynamic anterior cervical plating for multi-level spondylosis: Does it help?
STUDY RATIONALE METHODS
Anterior cervical decompression and fusion is considered Study design: Randomized controlled trial.
a gold standard for the treatment of symptomatic spon-
dylosis following failure of appropriate nonsurgical care. Inclusion criteria: Patients with symptomatic degen-
Favorable clinical outcomes following anterior cervical erative conditions resulting in radiculopathy or
discectomy and fusion (ACDF) have been attributed to myelopathy who would benefit from a two-level
successful healing/fusion of the interbody graft [1, 2]. anterior cervical discectomy and fusion (ACDF)
Anterior fixation with plate and screw devices have been with anterior plating.
recommended for patients requiring multilevel fusions
and may play a beneficial role in maintaining or restor- Exclusion criteria: Exclusion criteria included trauma
ing physiologic alignment of an operated neck while as- to the cervical spine, corpectomy, revision surgery,
suring a best possible fusion rate, especially when non- or previous posterior cervical surgery.
autologous structural bone-graft sources are used.
Studies have investigated various factors that may en- Patient population and interventions compared (Fig 1):
hance fusion rates, including the use of anterior plating • Of 97 patients who were assessed for eligibility, 61
[1, 2, 3, 4]. Rigid plate and screw implants had been a did not meet inclusion criteria or were excluded as
mainstay of fixation. Questions, however, have been described above. None were lost to follow-up,
raised whether these implant designs may adversely af- however, one patient’s x-rays could not be ade-
fect bone healing due to stress shielding and prevention quately measured due to the patient’s morbid obe-
of settling, as has been shown in biomechanical cadav- sity, leaving 35 patients available for a follow-up
eric studies [5, 6]. These concerns have led to the devel- rate of 97% at 12 months.
opment of implants with the stated goal of enhancing • A ll surgeries were performed using the same tech-
fusion rates by providing improved load sharing through nique and postoperative regimen by the same sur-
the anterior interbody grafts and the anterior spinal col- geon, including use of the same intervertebral al-
umn. A variety of implant designs have been introduced, lograft cage without use of supplemental graft
some with angulating screws, others with intrinsic materials.
mechanisms to allow for compression of structural grafts • Patients were randomly assigned to one of three
across the anterior spinal column. These types of im- plate designs: the cervical spine locking plate
plants, however, are invariably more complex and ex- (CSLP, Synthes) or the Atlantis (Medtronic); or
pensive than more conventional rigid locking plates. the Premier (Medtronic). The CSLP plate is a static
These “dynamic” implants have been called into ques- locked plate that theoretically does not allow set-
tion due to their potential to allow for an unwanted col- tling. The Atlantis and the Premier provided con-
lapse and the potential for implant migration. Can the trolled settling.
theoretical advantages of dynamization of plates be sub-
stantiated in the clinical setting based on radiographic Device description
findings? Do “dynamic loading plates” have improved Biomechanical Device name/
fusion rates and radiographic outcomes compared with Implant type concept manufacturer
conventional rigid locking plates? Static locking plate Rigid fixation CSLP / Synthes,
Paoli, PN, USA
Semi rigid Angular screw Atlantis / Medtronic,
locking plate transformation in plate Memphis, TN, USA
Objective
Translational Controlled plate Premier / Medtronic,
loading plate collapse Memphis, TN, USA
To compare the difference in fusion rates, time to fusion,
subsidence levels, complication rates between 1) a static,
• P
atients were randomized in blinded fashion
2) a dynamic angulation, and 3) a dynamic translation
where the assistant nurse selected a folded paper
plate and between dynamized plates in general and stati-
which contained the name of the plate from an
cally locked plates.
envelope, without the knowledge of the surgeon.
Outcomes:
• At follow-up, patients’ charts were analyzed for
age and gender, smoking status and litigation. If
chart data were insufficient, patients were con-
tacted individually and interviewed via telephone Analysis:

and asked to obtain one last x-ray. • Measurements were made from each lateral radio-
• Only radiographic criteria and measurements graph to assess 1) fusion rates, 2) time to fusion, 3)
were evaluated, as the main purpose of this study subsidence, 4) linear translation, and 5) angular
was the effect of the plate on radiographic out- variation [7] (Fig 2).
come and not clinical outcome. X-rays were evalu- • After comparison between the three groups was
ated by observers blinded to patient history at 0 completed, patients with dynamic plates were
and 6 weeks, 3, 6, and 12 months, and at the time grouped together (Premier and Atlantis), and
of most recent follow-up. X-rays were graded as compared with the patients with the static plate
fused or not fused. Criteria for fusion included the (CSLP patients).
presence of bridging trabeculae across the graft • Statistical analysis was performed using ANOVA
site and the lack of radiolucency between the graft for three-way comparison for the different type of
and the adjacent vertebral body. plates. Unpaired Student t-tests were used when
• Standardized radiographic measurements for set- the two dynamic plates were grouped and com-
tling, plate migration, subsidence and linear trans- pared against the static plate for the variable val-
lation were used (see web appendix for details). ues, while chi-square tests were used for categori-
Three independent observers provided the mea- cal values. Results were considered statistically
surements and graded fusion by assessing incor- significant when P < .05.
poration of the graft as “healed” or “unclear or not
healed”. These observers were also asked to asses Additional detail regarding methods can be found in
implant integrity and hardware loosening. the web appendix at www.aospine.org/ebsj.
Group A (CSLP) Lost to follow-up Analyzed

(n = 12) (n = 0) (n = 12)
Received allocated
Assessed Enrollment Randomized Group B (Atlantis) Lost to follow-up Analyzed

for eligibility (n = 36) treatment (n = 12) (n = 1) (n = 11)
(n = 97) assignment
Received allocated Could not assess x-rays
Did not meet inclusion (n = 37) intervention (n = 12) due to obesity
• 36 had one-level ACDF
• 1 had three-level ACDF
Group C (Premier) Lost to follow-up Analyzed
Based on exclusion criteria (n = 23) (n = 12) (n = 0) (n = 12)
• prior revision, corpectomy or instability
Received allocated
due to trauma, posterior surgery,
Refused to participate (n = 1)
Fig 2 Figure showing points of reference used for measurements.
E C
d
B

RESULTS • L
inear translation or migration differences were
found (Table 2). The only statistical difference was
• N o statistical differences with regard to sex, age, fol- found comparing the Premier plate to the Atlantis
low-up period or smokers were found. (Table 1 and where P < .05. Overlap of the upper end of the plate
web appendix). with the adjacent disc mostly occurred with the Pre-
• I n the CSLP group, all levels fused. One pseudoar- mier plate (Fig 3). To date, none of these patients has
throsis was found in each of the dynamic plates. required a repeat or adjacent level surgery. Therefore,
These differences were not statistically significant whether or not a statistical difference was observed,
(P = .61 ). the differences in time to fusion or subsidence were
• No statistical differences between the plates were not clinically significant and no patients required any
seen with regard to time to fusion (P = .59), or total further treatments.
subsidence (P = .63) (Table 2). • A ngular variation of the upper screws in the Atlantis
• No statistical differences were found when dynamic plate changed from 101.7° to 98.7° (mean degrees for
plates were grouped together and compared with the upper and 2.4° for the middle screws).
static plates (Table 3).
Table 1 Demographic and baseline characteristics of intervention groups

CSLP Atlantis Premier
N = 12 N = 11 N = 12
Age (years) 58 ± 15 51 ± 23 47 ± 14
Female (%) 8 (67%) 9 (82%) 6 (50%)
Smokers 1 (1%) 0 0
Litigation 0 0 0
Follow-up (months) 24 (12–42) 29* (12–49) 22 (12–43)
N = number enrolled in study

* One patient returned for follow-up 5 years after his surgery.
Table 2 Summary of radiographic findings at last follow-up

CSLP Atlantis Premier
N = 12 N = 11 N = 12 P - value
Fusion 12 (100%) 10 (91%) 11 (92%) .61
Time to fusion(months) 6.1± 4.5 8.5 ± 6.5 7.7 ± 4.8 .59
Subsidence (mm) 1.9 ± 0.9 2.3 ± 1.6 2.6 ± 2.4 .63
Linear translation/proximal plate migration (mm) 2.2 ± 1.5 1.5 ± 0.7 3.9 ± 2.4 .018
N reflects the number of patients with data available at last follow-up (after loss to follow-up)
P-values are for statistical tests across all three groups. For categorical measures, chi-square analysis was used, and for continuous measures, three way
ANOVA was used.
Table 3 Comparison of static plate to dynamic plates grouped together

Static plate Dynamic plates
N = 12 (grouped) N = 23 P - value
Fusion 12/12 (100%) 22/24 (95%) .44
Time to fusion (months) 6.1 ± 4.5 8.2 ± 5.8 .32
Subsidence (mm) 1.9 ± 0.9 2.1 ± 1.7 .58
Linear translation 2.2 ± 1.5 2.8 ± 2.1 .47
N reflects the number of patients with data available at last follow-up (after loss to follow-up)
P-values are for statistical tests across all three groups. For categorical measures, chi-square analysis was used, and for continuous measures, unpaired
Student t-tests were used.
Fig 3a Fig 3b
Discussion Summary and Conclusions
• T here does not appear to be a clear clinical advantage • T his is the first randomized study to compare the ra-
of dynamic plates over statically locked plates with diographic outcomes of three different types of plates
regard to fusion rates, subsidence and complication (one static and two dynamic) in patients who would
rates. benefit from two-level ACDF.
• I n one retrospective comparative study by Du Bois et • Use of a dynamic or static plate did not make a differ-
al. [7], the fusion rate was higher with dynamic ence on any of the radiographic outcomes measured.
plates. In contrast, our study showed a similar fusion • Conclusions drawn from biomechanical studies were
rate for all comparison groups. This difference could not substantiated in this small randomized con-
be attributed to the retrospective, non-randomized trolled trial.
study design used by Du Bois, which included one, • Additional randomized studies with larger sample
two and three level surgery, with patients receiving sizes are needed to further evaluate the benefits of
allograft or autograft intervertebral grafts. In our dynamic versus static plating.
study, these variables were not present.
• A nother prospective comparative study [8] did not
correlate the fusion rates with the type of plate used.
The radiographic correlation was beyond the scope of
that study. On the contrary, our study was mainly
focused on the radiographic results.
• Strengths: Ours is a randomized controlled trial
where potential differences between comparison
groups (ie, potential confounding factors) other than
those related to anterior fixation choice were kept to
a minimum.
• L imitations: The number of patients in each group
was small and the study may have been underpow-
ered to detect statistical differences between groups.
• P recise time to fusion would probably require week-
ly x-rays to be assessed, which would be ethically
suspect due to increased radiation exposure and lack
of clinical benefit. However, we chose the time points
where x-rays are usually obtained in clinical set-
tings, for patient convenience and from a practical
stand point. Thus, the exact time results obtained in
this study may be limited. Since the same points in
time were used for all three plates to assess fusion,
our results, however, could be considered a trend or
an extrapolation for time to fusion.

References
1. Bohlman HH, Emery SE, Goodfellow DB, et al (1993)

Robinson anterior cervical discectomy and arthrode-
sis for cervical radiculopathy. J Bone Joint Surg [Am];
75:1298–1307.
2. Wang J, McDonough P, Endow K, et al (2000) Increased
fusion rates with cervical plating for two-level ante-
rior cervical discectomy and fusion. Spine; 25:21–25.
3. Emery SE, Bolesta MJ, Banks MA, et al (1994) Robinson
anterior cervical fusion: Comparison of standard and
modified techniques. Spine; 19: 660–663.
4. Samartzis D, Shen FH, Lyon C, et al (2004) Does rigid
instrumentation increase the fusion rate in one-level
anterior cervical discectomy and fusion? Spine J;
4:636–643.
5. Brodke DS, Gollogly S, Alexander Mohr R, et al (2001)
Dynamic cervical plates: Biomechanical evaluation
of load sharing and stiffness. Spine; 26: 1324–1329.
6. Rapoff AJ, O’Brien TJ, Ghanayem AJ, et al (1999) Ante-
rior cervical graft and plate load sharing. J Spinal Dis-
ord; 12: 45–49.
7. DuBois CM, Bolt PM, Todd AG, et al. (2007) Static ver-
sus dynamic plating for multilevel anterior cervical
discetomy and fusion. Spine J; 7(2):188–193.
8. Nunley PD, Jawahar A, Kerr EJ, et al (2009) Choice of
plate may affect outcomes for single versus multilevel
ACDF: results of a prospective randomized single
blind trial. Spine J; 9(2): 121–127.
.
Original research—Kyphoplasty: Traditional imaging compared with computer-guided intervention—time to (...) 47
47—50
Kyphoplasty: Traditional imaging compared

with computer-guided intervention—time to
rethink technique?
Authors Michael P Silverstein1, Michael Mac Millan 2 , Isador H Lieberman 3
Institutions 1 Florida State University, College of Medicine, Tallahassee, FL, USA
2
Department of Orthopaedic Surgery, University of Florida College of Medicine, Gainesville, FL, USA
3
Medical Interventional & Surgical Spine Center, Department of Orthopaedic Surgery,
Cleveland Clinic Hospital, Weston, FL, USA
Abstract
Study design: Equivalence trial (IRB not required for cadaveric studies).
Objective: To compare computer-guided and fluoroscopic kyphoplasty. Fac-

tors of interest were radiation exposure, position of cannula within
pedicles and procedure time.
Methods: Kyphoplasty was performed on two cadavers. Computer-navigat-

ed, cross-sectional images from a cone-beam CT were used for one and
fluoroscopic imaging for the other. In each, T6–9 and T11–L2 vertebrae
were selected. For both imaging methods, anteroposterior and lateral
x-rays were taken. Radiation exposure for both procedures was mea-
sured by four dosimeters. Procedure time, radiation to surgeon and ca-
daver, and position of cannula placement within pedicles were record-
ed. The surgeon wore one under the lead gown, another on the lead
gown at shoulder level, and a third as a ring on the dominant hand. A
dosimeter was also placed on the cadaver.
Results: The radiation from the cone-beam, computer-guided imaging sys-

tem was 0.0 mrem to the surgeon and 0.52 rads to the cadaver. Using
fluoroscopic imaging, surgeon’s and cadaver’s exposure was 5 mrem
and 0.047 rads, respectively. Procedure times were similar and neither
device resulted in cannula malposition.
Conclusions: Cone-beam CT appears as accurate as the fluoroscopy; radia-

tion exposure to the surgeon is eliminated, and radiation levels to the
patient are acceptable.
No financial support was received for this research.

48 Original research—Kyphoplasty: Traditional imaging compared with computer-guided intervention—time to (...)
Study rationale and content Methods
Kyphoplasty is a surgical technique for height restora- Study design: Equivalence trial (IRB not required for
tion and cement augmentation of vertebral bodies fol- cadaveric studies).
lowing osteoporotic vertebral compression fractures or
osteolytic lesions [1–4]. Most studies reveal a low rate of Inclusion criteria: A cadaveric study to examine the
procedure-related adverse events [5–7]. However, when use of cone-beam CT and fluoroscopic assistance
fluoroscopic imaging is used, radiation exposure of the during a kyphoplasty.
surgeon and patient are of concern. Many of the compli-
cations of kyphoplasty have to do with inaccurate can- Exclusion criteria: Cadavers with spinal malformations
nula placement. Computer guidance has reduced radia- and/or deformities were excluded from this study.
tion to surgeon but equivalent imaging accuracy to
fluoroscopy. Techniques compared (Table 1)
• T wo cadavers were utilized for the study. Cone-
beam CT provided imaging assistance on one of the
cadavers and fluoroscopy for the second cadaver. In
Objective each of the two cadavers, the vertebra T6–9 (tho-
racic) and T11–L2 (thoracolumbar) were selected.
To examine the proposition that computer-guided cone- • Kyphoplasty: The thoracic and the thoracolumbar
beam CT technology could be used to perform balloon vertebral pedicles were accessed by placing a Jam-
kyphoplasty and subsequently reduce patient and physi- shidi needle percutaneously in the pedicle and
cian radiation exposure from that occurring with the then a guide wire. Subsequently, both cone-beam
fluoroscopic technique without prolonging the proce- CT and fluoroscopic images in each cadaver re-
dure or foregoing acceptable positioning of cannulas spectively were taken at each level when the can-
within pedicles. nulas were in the proper position.
• Cone-beam CT: The procedure was guided through
images obtained from the Breakaway® O-arm
(Medtronic Inc., Louisville, CO). Navigation was
computer controlled from the Stealth® Station
(Medtronic Inc., Louisville, CO). A high-speed drill,
guided with Navigation system (Medtronic Inc.,
Louisville, CO) was used to percutaneously bore a
pilot hole down the center of the target pedicle and
then place a guide wire in the pilot hole. The can-
nula was then positioned over the guide wire.
• Fluoroscopy: Fluoroscopy was used during the pro-
cedure to determine location. Biplanar imaging was
used for the fluoroscopic procedure. This included
anteroposterior (AP) and a lateral C-arm position.
When the cannulas were in proper placement in the
pedicle of both the cadavers, AP and lateral x-rays
were taken to assess position of the cannulas.
Outcomes and analysis

• P rimary outcomes: Procedure time, radiation ex-
posure to the surgeon and cadaver, and acceptable
positioning of cannula within pedicles was as-
sessed during the procedure for both cone-beam
CT and fluoroscopic assistance.
• Secondary outcomes: none.
Additional information regarding technical and meth-

odological aspects can be found in the web appendix at
www.aospine.org/ebsj.
Original research—Kyphoplasty: Traditional imaging compared with computer-guided intervention—time to (...) 49
Table 1 Radiation to physician, cadaver, and

RESULtS
procedure time
The time required to perform each procedure with fluo- total

imaging Number Radiation Radiation time of
roscopy and cone-beam CT was not significantly differ- devices of levels to physician to cadaver procedure
ent (table 1).
34 minutes,
• Radiation exposure to cadaver from the cone-beam Cone-beam CT 8 0 mrem 0.52 rads
32 seconds
CT was equivalent to the exposure to a patient un-
32 minutes, 51
dergoing a kyphoplasty would experience during Fluoroscopy 8 5 mrem 0.047 rads
seconds
cannula insertion (table 1).
• Assessment of the cannula position revealed no can-
nula malpositions or procedure-related complica-
tions for either technique. Fig 1 Cone-beam CT image as displayed on the
• A cone-beam CT image profi le includes images from Stealth Navigation workstation screen
a single rotation of the cone beam around the cadav-
er, as well as a digital image that projects instrumen-
tal targeting trajectory (Fig 1).
• The fluoroscopic procedure includes an AP (Fig 2a)
and a lateral image (Fig 2b).
diSCUSSiON
Radiation is a concern to both the surgeon and patient

during a kyphoplasty with minimal available informa-
tion regarding cone-beam CT exposure [8, 9].
• Both fluoroscopy and cone-beam CT assisted in al-
lowing the acceptable placement of the cannula
within pedicles, without a significant difference.
• Acceptable placement was assessed based on a two
Fig 2 Position of cannulas were assessed in
condition assessment. These conditions included
anteroposterior (AP) and lateral views during
whether the cannula remained within the pedicle
fluoroscopic imaging.
or if it violated the cortical boundaries of the
a anteroposterior image
pedicle.
b lateral image
• Strengths: Cone-beam CT resulted in zero radiation
exposure to the surgeon. When helpful, the system
provides imaging assistance with projection of tra-
jectory for cannula placement.
• Limitation: The study may have not included enough
cadavers to detect possible surgical performance
problems. With this possibility, there may have been
differences in procedure time.
• Fluoroscopy and cone-beam CT imaging both pro-
vide necessary imaging assistance to perform a a b
kyphoplasty. Cone-beam CT provides additional Cannula placement through Lateral view of vertebral body.
pedicles into vertebral bodies. The location of the cannulas are
tools which may be beneficial in minimally invasive visualized through the pedicles.
procedures and zero radiation exposure to the
surgeon.
• A multicenter trial utilizing both fluoroscopy and
cone-beam CT during a kyphoplasty may provide
stronger evidence for the accuracy and benefit of the
two imaging options.

50 Original research—Kyphoplasty: Traditional imaging compared with computer-guided intervention—time to (...)
Conclusion References:
Using the cone-beam CT kyphoplasty technique: 1. Kim DH, Vaccaro AR (2006) Osteoporotic compression
• Fluoroscopy and cone-beam CT have similar imag- fractures of the spine; current options and consider-
ing capabilities with equivalent time and accuracy. ations for treatment. Spine J; 6(5): 479–487.
• Cone-beam CT offers options such as targeting tools 2. Berlemann U, Franz T, Orler R, et al (2004) Kyphoplasty
and projection lines. for treatment of osteoporotic vertebral fractures: a
• Fluoroscopy results in additional radiation exposure prospective non-randomized study. Eur Spine J; 13(6):
to the surgeon. 496–501.
• O verall, device choice can be based on the surgeon’s 3. Lieberman IH, Dudeney S, Reinhardt MK, et al (2001)
preference. Initial outcome and efficacy of “kyphoplasty” in the
treatment of painful osteoporotic vertebral compression
fractures. Spine; 26:1631–1638.
4. Melton, LJ 3rd, Kan SH, Frye MA, et al (1989) Epidemiol-
ogy of vertebral fractures in women. Am J Epidemiol;
29(5):1000–1011.
5. Garfin SR, Buckley RA, Ledlie J (2006) Balloon kyphop-
lasty outcomes group: balloon kyphoplasty for symp-
tomatic vertebral body compression fractures results in
rapid, significant, and sustained improvements in back
pain, function, and quality of life for elderly patients.
Spine; 31(19): 2213–2220.
6. McArthur N, Kasperk C, Baier M, et al (2009) 1150
kyphoplasties over 7 years: indications, techniques
and intraoperative complications. Orthopedics; 32:90.
7. Taylor RS, Fritzell P, Taylor RJ (2007) Balloon kyphop-
lasty in the management of vertebral compression frac-
tures: an updated systematic review and meta-analysis.
Eur Spine J; 16:1085–1100.
8. Villavicencio AT, Burneikiene S, Bulsara KR, et al (2005)
Intraoperative three-dimensional fluoroscopy-based
computerized tomography guidance for percutaneous
kyphoplasty. Neurosurg Focus; 18(3):e3.
9. Mroz T, Yamashita T, Davros W, et al (2008) Radiation
exposure to the surgeon and the patient during kypho-
plasty. J Spinal Disord Tech; 21(2):96–100.
Original research—ProDisc-C versus fusion with Cervios chronOS prosthesis in cervical degenerative disc (…) 51
51—56
ProDisc-C versus fusion with Cervios chronOS

prosthesis in cervical degenerative disc
disease: Is there a difference at 12 months?
Authors Matjaz Vorsic, Gorazd Bunc
Institution University Hospital Maribor, Department of Neurosurgery; Maribor, Slovenia
Abstract
of evidence (CoE) Study design: Prospective cohort study.
Objective: The aim of the study was to compare clinical results and to de-
Study Design
termine differences in outcomes between anterior cervical discectomy
RCT
and fusion (ACDF) and disc arthroplasty in patients treated for symp-
Cohort study •
tomatic cervical degenerative disc disease.
Case control
Case series
Methods: Forty patients with cervical degenerative disc disease were treat-
Methods
ed with ProDisc-C disc arthroplasty and 40 patients with fusion using
Concealed allocation (RCT)
an intervetebral spacer with integrated fixation (Cervios chronoOS)
Intent to treat (RCT)
implants without additional anterior fixation. Fifty disc prostheses were
Blinded/independent evaluation of •
placed in the first group and 52 intervertebral spacers were implanted
primary outcome
in the second group. Clinical outcomes were assessed before and 12
Follow-up ≥ 85% •
months following the procedure using the neck disability index (NDI)
Adequate sample size •
and visual analog scale (VAS) for neck and arm pain, with 15% im-
Control for confounding
provement in NDI and 20% in VAS defined as a clinically significant.
Evidence class: III
Results: Eighty patients with cervical degenerative disc disease with a
The definiton of the different classes mean age of 49.7 years were included in the study with a minimum fol-
of evidence is available on page 83. low-up of 12 months. The groups were similar at baseline both clinical-
ly and statistically (P > .05) except for age and VAS for arm pain. Both
groups had a statistically significant improvement in NDI and VAS for
neck and arm pain (P < .05) and the arthroplasty group had a better im-
provement according to NDI (74.3% of patients in the arthroplasty
group achieved ≥ 15% improvement in NDI versus 65.7% of patients in
ACDF group).
Conclusions: Both ProDisc C and Cervios chronoOS prostheses resulted in

significant pain reduction and functional outcome for the patients with
slightly better results in the group treated with disc arthroplasty 12
months after the surgery.
The authors have no financial relationships to disclose.

The ProDisc-C prostheses and Cervios chronoOS implants (Synthes) presented in the study are approved for clinical use.

52 Original research—ProDisc-C versus fusion with Cervios chronOS prosthesis in cervical degenerative disc (…)
STUDY RATIONALE AND CONTEXT Objective
Cervical anterior discectomy and fusion (ACDF) is a To compare clinical results 12 months after surgery us-
standard treatment for symptomatic cervical degenera- ing the neck disability index (NDI) and visual analog
tive disc disease in the patients where conservative treat- scale (VAS) for neck and arm pain between a standard
ment has failed. Since fusion may be associated with (ACDF) group and a disc arthroplasty group in patients
progressive degeneration of adjacent motion segments, presenting with symptomatic cervical degenerative disc
disc arthroplasty, which preserves segmental motion disease. The secondary objective was to determine the
and improves load transfer to the adjacent levels, has complications after each procedure.
been introduced in hopes of achieving improved pain
and function without adjacent segment disease [1, 2].
Group A Lost to follow-up Analyzed

(Arthroplasty group) (n = 1) (n = 39)
(n = 40) Reasons: Excluded from
Received allocated Failed to show (n = 1) analysis (n = 1)
intervention Reason:
Lost to follow-up
as above

for eligibility (n = 80) treatment Allocation Follow-up Analysis
(n =114) assignment (12 months)
Excluded (n = 34)
Reasons:
Previous spine surgery (n = 10) Group B Lost to follow-up Analyzed
Refused to participate (n = 12) (ACDF group) (n = 2) (n = 38)
Cardiovascular contraindication (n = 4) (n = 40) Reason: Excluded from
Coagulopathy (n = 3) Received allocated Failed to show (n = 1) analysis (n = 2)
Inability to understand interview (n = 5) intervention Death (n = 1) Reasons:
Lost to follow-up
as above
METHODS • C
ervical collars were not used postoperatively in
either group. NSAIDs were used during the first 2
Study design: Prospective cohort study. weeks selectively for patients with severe postop-
erative neck pain in both groups.
Inclusion criteria: All symptomatic patients with one
or two level soft disc herniations and/or degenera- Outcomes:
tive changes of the cervical spine, not responding • P rimary outcomes: Clinical outcomes included
to conservative treatment from January 2006 to the neck disability index (NDI) and a 10-point vi-
January 2008. sual analog scale (VAS) for neck and arm pain at
baseline and 12 months after surgery. Measure-
Exclusion criteria (Fig 1): Patients with concomitant ments were conducted by study personal blinded
conditions which could confound outcomes assess- to the surgical intervention according to a struc-
ment (eg, previous cervical spinal surgery), con- tured protocol.
traindications to surgery, refusal to participate or • Secondary outcomes: Incidence of complications
inability to complete interviews were excluded. related to device and surgical approach. Radio-
Additionally, patients with cervical instability, os- graphic outcomes included assessment of hard-
teoporosis, malignant disease, infection, spondylo- ware loosening and displacement, malalignment,
discitis, traumatic spine injury, known allergy to heterotopic ossification and hardware failure.
foreign material, myelopathy, were also excluded • I mprovement was measured by calculating the
from the study. change in NDI and VAS scores from baseline (pre-
operative) to the 12 month follow-up. A clinically
Patient population and interventions compared (Fig 1): significant change was considered a 15% improve-
• One-hundred-and-fourteen patients were ap- ment in NDI score and a 20% improvement in VAS
proached to participate in the study. Twenty two score.
did not meet study criteria and 12 refused to par- • A brief description of the measures and statistical
ticipate leaving a total of 80 patients who complet- methods are provided in the web appendix at
ed the informed consent. www.aospine.org/ebsj.
• T reatment was assigned based on instrumentation
availability or surgeon preference but was not re- Analysis:
lated to factors that may have an influence on the • Changes from preoperative to postoperative in
outcome. NDI and VAS scores were compared within treat-
• Enrollment concluded with 40 patients treated ment groups using a paired t-test and between
with ProDisc-C disc arthroplasty and 40 patients treatment groups using an unpaired t-test. Differ-
underwent fusion using Cervios chronoOS im- ences in baseline continuous variable (eg, age)
plants without additional anterior fixation. were tested using unpaired t-tests. Differences in
• A rthroplasty group: The complete anterior cervi- categorical baseline variables (eg, gender) were
cal discectomy was performed in conventional tested using a chi-square test. PASW Statistics 18
fashion under magnification with neural element software was used to provide all the statistical
decompression using high-speed drills and other data.
neural dissection tools. The proper implant bed
was prepared under the fluoroscopic x-ray control
and an appropriately sized ProDisc-C artificial disc
was then implanted.
• ACDF group: Neural element decompression was
carried out using the same technique described
above. An appropriately sized Cervios chronoOS
implant filled with artificial cancellous bone was
inserted using fluoroscopic guidance. No further
anterior fixation (such as plate and screw fixation)
was used.

RESULTS Discussion
• T he patients in both groups had similar baseline • T here was significant functional, neck and arm pain
characteristics (Table 1) with the exception of age and reduction in both groups 12 months postoperatively
baseline NDI and VAS scores. Patients in the ACDF but these differences were not statistically significant
group were slightly older and had slightly lower base- between the arthroplasty and ACDF groups. This
line scores. The 12-month follow-up rate was 97.5% finding is consistent with other studies [6,7].
and 95% for the arthroplasty group and ACDF • Strengths: this is a prospective cohort study with a
groups, respectively. 96% follow-up rate at 12 months using blinded as-
• There was significant functional, neck and arm pain sessment of validated patient reported outcomes
reduction in both groups from baseline to 12 months measures.
after surgery (P < .001) • L imitations: The study was limited by a short follow-
• The mean percent improvement was greater in the up period. Future studies should follow these pa-
arthroplasty group compared to the fusion group in tients for several more years. Since this was not a
each outcome; however, these differences were not randomized trial, we cannot be certain that both
statistically significant (Table 2). groups were similar with respect to all baseline fac-
• The proportion of patients achieving clinically sig- tors that may introduce confounding of the compari-
nificant improvement with each outcome was higher son. The variables we did collect demonstrated two
in the arthroplasty group compared to the fusion relatively similar groups; however, future studies
group (Table 2). These differences were not statisti- should consider other important baseline factors
cally significant. such as American Spinal Injury Association (ASIA)
• One major complication (posterior epidural abscess) score, smoking status, disability claims, etc.
occurred in the arthroplasty group, requiring re- • Clinically, the cervical disc arthroplasty challenges
moval of the prosthesis; otherwise, there were no de- the surgeon to more precise hardware placement. We
vice related complications (such as loosening, migra- found no correlation of a potentially more complex
tion of the implant, material failure, allergic reaction) procedure due to the occurrence of only one compli-
and no approach related complications (dural tears cation. Our study findings, including occurrence of
or leaks, hematomas, esophageal or tracheal injuries, complications, were remarkably similar to previous
laryngeal nerve dysfunction). publications [2,6,7].
• The concern of fusion (ACDF) resulting in progres-
sive degeneration of adjacent segments while disc ar-
throplasty potentially preserves integrity of adjacent
motion segment could not be answered in our limit-
ed follow-up time [3,4,5].
• Studies with longer follow-up (10 years) are neces-
sary to better evaluate the comparative effectiveness,
safety and long term survival of disc arthroplasty
compared to fusion.
Summary and Conclusions
• B oth treatments relieve patient’s pain and improve

functional outcome.
• Despite no statistically significant between group
differences in the primary outcomes, there were
slightly better results in the disc arthroplasty group
compared to the fusion group 12-months postopera
tively.
Table 1 Demographic and baseline characteristics of treatment groups

Arthroplasty group ACDF group
n = 40 n = 40 P-value*
Age (mean ± SD) 48.1 ± 8.1 51.3 ± 8.1 0.02
Female (n, %) 27 (67.5%) 2 (70%) 0.82
Male (n, %) 13 (32.5%) 12 (30%) 0.82
Signs and symptoms duration (mean weeks ± SD) 45.5 ± 31.7 42.2 ± 38.7 0.68
Baseline NDI (mean ± SD) 67.4 ± 9.6 61.7 ± 15.1 0.10
Baseline VAS neck (mean ± SD) 7.1 ± 1.2 6.7 ± 1.5 0.30
Baseline VAS arm (mean ± SD) 6.9 ± 0.7 6.1 ± 1.4 0.004
Number of implants used (n) 50 52 0.64
* Categorical variables compared using chi-square test and continuous variables with unpaired t-test.
Table 2 Summary of NDI and VAS change scores
NDI VAS neck VAS arm

Arthroplasty ACDF Arthroplasty ACDF Arthroplasty ACDF
(n = 38) (n = 38) (n = 39) (n = 38) (n = 39) (n = 38)
Baseline (mean ± SD) 67.4 ± 9.6 61.7 ± 15.1 7.1 ± 1.2 6.7 ± 1.5 6.9 ± 0.7 6.1 ± 1.4
12 months (mean ± SD) 34.9 ± 18.5 38.7 ± 16.1 4.3 ± 1.8 4.7 ± 1.8 3.3 ± 2.2 3.5 ± 1.8
Change (mean ± SD) 32.1 ± 17 22.9 ± 16.9 2.8 ± 1.6 2 ± 1.8 3.5 ± 2.3 2.6 ± 1.6
Improvement
Mean % 48.2% 37.3% 39.4% 29.8% 52.2% 42.6%
≥ 15% in NDI 74.3% 65.7%
≥ 20% in VAS 65.8% 48.6% 71.2% 56.7%
Within treatment P-value* P < .001 P < .001 P < .001 P < .001 P < .001 P < .001
Between treatment P-value† P = .43 P = .38 P = .85
* P-value associated with change from baseline to 12 months within each group.
† P-value comparing baseline to 12 months changes between arthroplasty and ACDF groups.

References
1. Philips FM, Garfin SR (2005) Cervical Disc Replace-

ment. Spine; 30: S27–33.
2. Nabhan A, Ahlhelm F, Shariat K, et al (2007) The Pro-
disc-C Prothesis. Spine; 32: 1935–1941.
3. DiAngelo D, Foley KT, Morrow BR, et al (2004) In vitro
biomechanics of cervical disc arthroplasty with the
Prodisc-C total disc implant. Neurosurgical Focus; 17
(3): 44–54.
4. Bertagnoli R, Yue JJ, Pfeiffer F, et al (2005) Early results
after Prodisc-C cervical replacement. J Neurosurg:
Spine; 2: 403–410.
5. Chang UK, Kim DH, Lee MC, et al (2007) Range of mo-
tion change after cervical arthroplasty with Prodisc-
C artificial discs compared with anterior cervical dis-
cectomy and fusion. J Neurosurg: Spine; 7: 40–46.
6. Kim SW, Shin JH, Arbatin JJ, et al (2008) Effects of a cer-
vical disc prosthesis on maintaining sagittal align-
ment of the functional spinal unit and overall sagittal
balance of the cervical spine. Eur Spine J; 17: 20–29.
7. Nabhan A, Ahlhelm F, Pitzen T, et al (2007) Disc re-
placement using Prodisc-C versus fusion: a prospec-
tive randomized and controlled radiographic and
clinical study. Eur Spine J; 16: 423–430.
Systematic review—Unilateral facet dislocations: Is surgery really the preferred option? 57
57—66
Unilateral facet dislocations:

Is surgery really the preferred option?
Authors Marcel Dvorak1, Alexander R Vaccaro2 , Jeffrey Hermsmeyer3, Daniel C Norvell3
Institutions 1 University of British Columbia, Blusson Spinal Cord Centre, Vancouver BC, Canada
2
Thomas Jefferson University and the Rothman Institute, Philadelphia PA, USA
3
Spectrum Research Inc., Tacoma, Washington, USA
Abstract
Study design: Systematic review.
Objective: To compare the safety and effectiveness of initial surgery versus nonoperative management
of unilateral facet dislocations with or without fractures.
Summary of background: Unilateral facet injuries represent between 6%–10% of all cervical spine
injuries and yet optimal treatment for these injuries has not been established. The surgeon is faced
with the decision of whether to manage the injury operatively or nonoperatively. Providing evi-
dence to support this decision is necessary and is the rationale behind this article.
Methods: A systematic review of the English language literature was undertaken for articles published
between 1970 and August 2009. Electronic databases and reference lists of key articles were
searched to identify studies evaluating surgery and nonoperative management of unilateral facet
dislocations. Bilateral facet dislocations, isolated facet fractures (without dislocation), and com-
plete spinal cord injuries were excluded. Two independent reviewers assessed the level of evidence
quality using the GRADE criteria and disagreements were resolved by consensus.
Results: We identified six articles meeting our inclusion criteria. Treatment failure, neurological dete-
rioration, and persistent pain occurred more frequently in patients treated nonoperatively versus
patients treated with surgery. Surgical patients experienced infections and surgical related compli-
cations not experience by those managed nonoperatively. Patients treated surgically after failed
nonoperative management also experienced better outcomes than those who continued to be man-
aged nonoperatively.
Conclusion: When faced with a patient requesting treatment recommendations for their acute unilat-
eral facet dislocation, the surgeon can state that treatment failure, persistent pain, and neurologi-
cal deteriorationoccur more frequently with nonoperative treatment based on the available litera-
ture. Ultimately it will be the preference of the patient that will decide between these two treatment
approaches.
This systematic review was funded by AOSpine.

58 Systematic review—Unilateral facet dislocations: Is surgery really the preferred option?
STUDY RATIONALE AND CONTEXT MATERIALS AND METHODS
Unilateral facet injuries represent 6%–10% of all cervi- Study design: Systematic review.
cal spine injuries and yet optimal treatment for these in-
juries is frequently in dispute. When faced with a patient Sampling: Search: PubMed, Cochrane collaboration
in the emergency room, the treating spine surgeon is of- database, and National Guideline Clearinghouse
ten asked to recommend either initial surgery or nonop- databases; bibliographies of key articles.
erative treatment based on his/her experience and un- Dates searched: 1970 to August 2009.
derstanding of the literature. While patient preference is
often a strong deciding factor, it is incumbent upon the Inclusion criteria: (1) Unilateral facet dislocations
surgeon to provide the patient with therapeutic advice with or without fractures, (2) Adults 18 years and
that is most likely to return the patient to their pre-inju- older, (3) Studies including ten or more patients in
ry health status with the lowest risk of complications. either arm
This first decision, whether to operate or not, is thus very
important. Often this decision is made not in the con- Exclusion criteria: (1) Bilateral facet dislocation, (2)
trolled environment of an elective office or clinic but complete spinal cord injury, (3) isolated fracture
more frequently in the emergency department and out- without dislocation, (3) less than ten subjects per
side regular hours and may be influenced by resource treatment, (4) data on unilateral facet dislocations
availability, surgeon training, and local practice patterns. not presented separately from other injuries.
P roviding evidence to support this decision and subse-
quent decisions, should initial nonoperative treatment Outcomes: Failed treatment, neurological deteriora-
fail, is necessary and is the rationale behind this article. tion, persistent pain, wound or surgical site infec-
tion, and complications (health-related or surgery-
specific).
OBJECTIVES Analysis: Descriptive statistics.
To compare the safety and efficacy of initial surgery For more details see web appendix at www.aospine.
versus nonoperative management of unilateral facet org/ebsj.
d islocations with or without fractures and, among pa-
tients who experience failed nonoperative management,
to compare outcomes of those who receive surgery ver-
sus those who do not receive surgery. RESULTS
We identified six articles meeting our inclusion criteria

(Fig 1). Four studies evaluated operative or nonoperative
treatment for unilateral facet dislocation. Two studies
evaluated failed nonoperative treatment (inability to
achieve and maintain reduction, a progression in neuro-
logical symptoms, or the presence of late pain and/or
i nstability) that did or did not lead to future surgical
management.
Operative versus nonoperative treatment of unilateral facet dis-

locations (Table 1 and Fig 2)
• One-hundred-and-seventy-six patients undergoing
operative or nonoperative treatment were identified.
• T reatment failure rates were higher in nonoperative-
ly managed patients (80%) than surgically managed
patients (2.6%) [1–4].
• Neurological deterioration happened infrequently,
but occurred more after nonoperative treatment
(5%) versus operative treatment (0%) [1, 2, 4].
Fig 1 Flow chart showing results of literature search

• P osttreatment persistent pain occurred more fre-
quently in nonoperative treatment (30%) than it did 1. Total citations
in operative treatment (10.3%) [1–4]. (n = 66)
• Outcomes were not reported in the nonoperative 2. Title / abstract
review (n =44)
treatment studies, but for operative treatment were
reported as surgical site or deep wound infection 3. Retrieved for full
(7.8%) and general health or surgery specific compli- (n =22)
cations (13.8%) [1–4]. Excluded at
full-text (n =16)
Failed nonoperative treatment that did or did not lead to future 5. Publications
surgical management (Table 2) (n =6)
• Forty-eight patients that had a failed nonoperative
treatment who continued to be managed nonopera-
tively (n = 28) or who subsequently went on to future Fig 2 Outcomes rates comparing surgical to nonopera-
surgery (n = 20) were identified. tive management of unilateral facet dislocations
• Failed anatomical reduction rates were higher among 100
patients with continued nonoperative management 80 Operative
80
(100%) versus those who underwent surgical man- Nonoperative
Percent
agement (30%) [5, 6]. 60
• Neurological deterioration occurred more frequently 40 30
in continued nonoperative treatment (10.7%) versus 20 10.3
operative treatment (0%) [5, 6]. 2.6
0
5
0
• Posttreatment persistent pain occurred more fre- N = 116 N = 60 N = 74 N = 60 N = 116 N = 60
quently in continued nonoperative treatment (70%)
Treatment failure Neurological deterioration Pain
than it did in operative treatment (5%) [5, 6].
Outcomes
Table 1 Subject characteristics of studies evaluating operative versus nonoperative treatment for
unilateral facet dislocation
N = 176 Operative N = 116 Nonoperative N = 60

Outcomes Studies (n) Patients (n) Results (mean) Results (range) Studies (n) Patients (n) Results (mean) Results (range)
Treatment failure* 4 116 2.6% 0–6% 2 60 80% 77–82%
†
Neurological deterioration 3 74 0% 0% 2 60 5% 0–9%
Wound or surgical infection 4 116 7.8% 0–12% NR NR NR NR
Persistent pain 4 116 10.3% 0–14% 2 60 30% 27–32%
Complications ‡ 4 116 13.8% 0–29% NR NR NR NR
* Defined in operative treatment as future subluxation, nonunion, or reoperation; defined in nonoperative treatment as failed anatomical reduction
which may or may not lead to future surgical management.
† Defined as a negative change in neurological status from pre to postoperative.
‡ Including general health complications such as pneumonia and surgery-specific complications such as nerve palsy, dysphagia, difficulty swallowing,
and wound site drainage.
Table 2 Subject characteristics of failed nonoperative treatment that may or may not lead to
future surgical management
N = 48 Operative N = 20 Nonoperative N = 28

Outcomes Studies (n) Patients (n) Results (mean) Results (range) Studies (n) Patients (n) Results (mean) Results (range)
Failed anatomical reduction 2 20 30% 20–40% 2 28 100% 100%
Neurological deterioration* 2 20 0% 0% 2 28 10.7% 0–17%
Persistent pain 2 20 5% 0–10% 1 10 70% 70%
* Defined as a negative change in neurological status from pre to postoperative.

Evidence Summary
Question 1: Compare the safety and efficacy of initial surgery versus nonoperative management of unilateral
facet dislocations
Outcomes Strength of evidence Conclusions/comments

1. Treatment failure Very low Low Moderate High Treatment failure rates are higher in
nonoperatively managed patients compared
to surgically managed patients.
2. Neurological deterioration Very low Low Moderate High Neurological deterioration happened infrequently,
but occurred more frequently in nonoperative treatment
versus operative treatment.
3. Wound or surgical infection Very low Low Moderate High Rate of infection ranged from 0%–12% in
surgically managed patients.
4. Posttreatment pain Very low Low Moderate High Long term persistent pain occurred more
frequently in nonoperative treatment compared
to operative treatment.
5. Complications Very low Low Moderate High Complication rates occurred at a mean of
13.8% in surgically managed patients.
Details about the determination of strength of evidence can be found in the web appendix at www.aospine.org/ebsj
Question 2: Compare the safety and effectiveness of surgery versus nonoperative management after failed
nonoperative management
Outcomes Strength of evidence Conclusions/Comments

1. Failed anatomical reduction Very low Low Moderate High Failed anatomical reduction rates are higher
in nonoperatively managed patients compared to surgically
managed patients.
2. Neurological deterioration Very low Low Moderate High Neurological deterioration was only reported
in patients who received continued nonoperative
management at mean of 10.7%.
3. Posttreatment pain Very low Low Moderate High Long term persistent pain occurred more
frequently in continued nonoperative treatment compared
to operative treatment.
Discussion • A
lthough the quality of publications is lacking,
there is remarkable consistency in the results across
• I n six case series that evaluated isolated unilateral these studies. When faced with a patient requesting
facet dislocations, treatment failure, neurological treatment recommendations for their acute unilat-
deterioration, and persistent pain occurred more fre- eral facet dislocation, the surgeon can state that
quently in patients treated nonoperatively versus pa- treatment failure, persistent pain, and neurological
tients treated with surgery. Surgical patients exper deterioration consistently occur more frequently
ience infections and surgical related complications with nonoperative treatment based on the available
that those treated nonoperatively do not experience. literature. It must be acknowledged that surgical
Patients treated operatively after failed nonoperative treatment carries with it a complication rate likely
management also experience better outcomes than around 10%t–15%. Similarly, there is consistent
those who continue to be managed (Tables 3–4). support for surgical treatment following failure of
nonoperative care. Ultimately is will be the prefer-
• T he existing literature reporting outcomes on the ence of the patient that will decide between these
treatment of unilateral facet dislocations is limited to two treatment approaches.
case series. No studies were identified that compared
operative to nonoperative management in the same
patient population. Pooled rates of treatment failure
from these case series are remarkably higher in
patients who are treated nonoperatively, but the
potential for selection bias in this comparison is
likely and therefore conclusions must be made with
caution. Comparative studies are necessary to estab-
lish the efficacy of operative versus nonoperative
management of these injuries.
• A lthough it is the facet that ultimately generates

d ifficulty for realignment, the disc at the injured
motion segment may also influence surgeon’s deci-
sion making. Three of six studies reported surgical
management of disc pathology.
–– A discectomy at the injured level was performed
in order to facilitate a fusion procedure regardless
of whether or not disc herniation was document-
ed by MRI [1].
–– Discectomy was performed in five patients that
had disc herniation as detected by MRI [4].
–– The disc was excised if there was disc material
dislodged into spinal canal [2].
However, none of these studies reported outcomes

separately for those who had a discectomy versus
those that did not. One study excluded from patient
population all disc herniations documented by MRI
[3]. In two studies there was no mention of disc dis-
ruption or disc herniation [5, 6].

Fig 7 MRI imaging performed preoperatively

Illustrative case (Figs 3–10)
revealed disc material posterior to the anteriorly
subluxed body of C4.
A 43-year-old man was the unrestrained driver and sole
Fig 8 Anterior discectomy was successful at decom-
occupant in a single vehicle roll-over high speed motor
pression, however was not completely successful at
vehicle accident. The patient was found walking around
reducing the dislocated facet joint.
at the scene of the accident. In the emergency room he
complained of neck pain and facial and scalp abrasions.
Neurological examination revealed right shoulder
numbness, but no other neurological abnormality. He
had no other injuries other than his cervical spine facet
subluxation.
Anterior decompression was indicated to remove the
posteriorly displaced disc fragment and combined ante-
rior and posterior fixation provided reduction of the
d islocated facet and stability. In the scenario of a sublux-
ation, as opposed to dislocation, anterior discectomy, 7 8
fusion, and plating are often effective treatment options.
It is anticipated that his C5 radiculopathy would recover
Figs 9–10 Lateral and AP views of combined anterior
after treatment.
and posterior fixation which was ultimately
necessaryto stabilize this injury.
Fig 3 Lateral cervical spine plain x-ray demonstrates
anterior subluxation of C4 on C5 of approximately
25% of the vertebral body diameter.
Fig 4 Axial CT scan
3 4 9 10
Fig 5 Reformatted image confirms a unilateral right

sided facet dislocation at C4–5.
Fig 6 A second reformatted image confirms a unilat-
eral right sided facet dislocation at C4–5.
5 6
Table 3 Subject demographics of studies evaluating operative versus nonoperative treatment

for unilateral facet dislocation
Author Study
(Year) type Population Dislocation characteristics Treatment Follow-up
N = 46 Unilateral cervical facet dislocation ± Surgery: Mean follow-up:
Age: 30 years (19–52) fractures; single level dislocation Posterior reduction and/or internal fixation – Spinous process wire group:
86% male (n = 46, 100%) 102 months
Case series
Shapiro
(1999)
– F irst 24 patients underwent spinous – C able and lateral mass plate

process wire fixation group: 40 months
– Additional 22 patients underwent – Follow-up rate: 86%
interspinous wiring with braided cable – Follow-up range: 12–120
for lateral mass plating months
N = 17 Distractive flexion injuries Anterior fixation Mean follow-up:
Case series
Henriques
(2004)
Age: 47 years (17–82) Stage 2 (unilateral facet) ± fractures (n = 17, 100%) – 15 months
94% male 2-level dislocation n = 3 – Follow-up rate: NR
N = 11 Single-level Anterior cervical discectomy, distraction – Follow-up 30 days

Harrington
(2007)
series
Age: 42 years (22–65) unilateral facet injuries ± fractures reduction with allograft fusion and anterior – Follow–up rate: 100%
Case
55% male cervical plating
N = 42 Single-level unilateral facet injury ± Anterior fixation: Anterior fixation follow-up rate:
Age: 35 years (17–86) fractures anterior cervical discectomy and fusion 14/20 (70%) at 12 months
74% male (n = 20, 47%)
Posterior fixation follow-up
Case series
Anterior fixation n = 20 Posterior fixation: rate: 19/22 (86%) at 12 months

(2007)
Kwon
Age: 38 years (17–86) lateral mass screw-plate fixation, and/or

70% male oblique wiring (n = 22, 53%)
Posterior fixation n = 22

Age: 33 years (17–69)
77% male
N = 26 Isolated Initial treatment consisted of either Mean follow-up:
Age: 37 years (17–74) unilateral facet dislocation skull-tong or halo traction – 3 years
85% male n = 12 (n = 26, 100%) – F ollow-up rate: 100%
– Six (23%) patients experienced
Fracture of facet or associated body reduction with traction
fracture n = 14 - Two treated nonoperatively with
halo thoracic vest
Case series
Rorabeck
(1987)
- Four underwent one-level posterior

fusion
– Twenty (77%) did not experience
reduction
- Ten (38.5%) left in displaced position,
ambulated with external bracing
- Ten (38.5%) underwent open
reduction and single-level posterior
fusion
N = 36* Unilateral facet dislocations or Initial treatment (N = 34): Mean nonoperative follow-up:
Age: 33 years (15-87) fracture-dislocations with – Closed reduction with halo traction nearly 9 years
78% male n = 2 at multiple levels (n = 28), halo traction alone (n = 1),
*2 patients lost to follow-up immobilization with cervical brace Mean operative follow-up:
(n = 4), and no treatment (n = 1) 6.5 years
Nonoperative:
Case series
n = 24 Operative treatment (n = 10): Follow-up rate for both groups:

(1991)
Beyer
Age: 30 years (16–74) Indications for surgery included: 94%

– Persistence or progression of neural
Operative: deficit (n = 4) 2 patients lost to follow-up
n = 10 – Failed reduction (n = 3)
Age: 33 years (15–71) – Loss of reduction (n = 3)
Operative treatment consisted of open
reduction through posterior approach and
fusion with interspinous wiring

Table 4 Subject outcomes of studies evaluating operative versus nonoperative treatment for
unilateral facet dislocation
Author Treatment Neurological
(Year) failure N (%) deterioration N (%) Infection N (%) Pain N (%) Complications N (%) Significant findings
1 (2%) 0 (0%) 3 (6%) 6 (13%) No other complications – No deaths
Experienced – Persistent neck reported – Spinous process wire group
resubluxation pain in 4 (17%) of had 11 (46%) perfect
and underwent spinous process anatomical alignment
Shapiro
(1999)
reoperation of wire group compared to 14 (64%) cable

anterior cervical – Persistent neck and lateral mass plate group
fusion and pain in 2 (9%) of
plating cable and lateral
mass plate group
1 (5.9%) 0 (0%) 0 (0%) 0 (0%) – Developed unilateral – Two (66%) of the three
Patient with recurrent laryngeal nerve patients with 2-level injuries
2-level injury palsy secondary to developed nonunion at one
reoperated with surgical approach in 1 level
Henriques
anterior fusion (2.8%) – One nonunion patient

(2004)
due to nonunion – Transient dysphagia in 1 reoperated with anterior

(2.8%) fusion while reoperation not
– Injury of the lateral necessary in the other as
cutaneous femoral nerve patient free of symptoms
as a result of bone graft
harvesting in 1 (2.8%)
0 (0%) 0 (0%) 1 (9.1%) 0 (0%) Ventilator-related Translational subluxation:
No pain in pneumonia in 1 (9.1%) Preop 5–1 mm;
Preoperative ASIA any patient patient requiring Postop 0–3.5 mm
E-7 (64%) as evidence tracheotomy
D-4 (36%) of stability Sagittal rotation:
Preop -15 to +16°
Postoperative ASIA Postop -11 to +6°
Harrington
(2007)
E-8 (73%)
D-3 (27%)
One preoperative ASIA

score of D improved to E
postoperatively. All other
preoperative ASIA scores
did not change
postoperatively.
1 (2.4%) Anterior fixation: Anterior fixation: Anterior fixation: Anterior fixation: Median time to achieve
Posterior fixation Neurological component 1 (5%) Score of > 5 out of Severe medical discharge criteria:
patient of NASS cervical spine Infection at bone 10 VAS pain in complications acutely Anterior 2.75 (1–24) days;
developed questionnaire: 85.2 (100 graft site at 3 weeks three (15%) postop in one patient (5%) Posterior 3.5 (1.5–42) days
pseudoarthrosis optimal) patients Swallowing difficulties in 11
and required Posterior fixation: (55%) VAS pain score on
anterior revision Posterior fixation: 4 (18%) Posterior fixation: postoperative day 1:
Neurological component Superficial wound Score of > 5 out of Posterior fixation: Anterior 2.6; Posterior 3.6
of NASS cervical spine infection at 2–3 10 VAS pain in No other complications and on postoperative day 2:
(2007)
questionnaire: 83.9 (100 weeks in three three (13.6%) reported Anterior 2.1; Posterior 3.0
Kwon
optimal) patients (14%) patients No difference in anterior versus

MRSA wound posterior regarding SF-36
Baseline neurological infection in one mental and physical scores and
status not measured so patient (4%) for NASS cervical and
deterioration cannot be neurological scores
calculated.
Fusion rate available for
patients at 12-month
follow-up: Anterior 100%
(18/18); Posterior 89% (17/19)
Author Treatment Neurological

(Year) failure N (%) deterioration N (%) Infection N (%) Pain N (%) Complications N (%) Significant findings
20 (77%) 0 (0%) NR 7 (26.9%) – F ive of seven patients left Patients left in displaced
Did not – No patient (0%) in displaced position and position and allowed to heal in
anatomically that experienced having significant that position develop late pain
reduce after reduction with disabling pain went on to
initial treatment traction had pain surgical anterior or
of skull-tong or at follow-up posterior fusion
halo traction – O f ten patients – Two of ten patients
Rorabeck
(1987)
left in displaced receiving open reduction

position seven and single-level fusion did
(70%) had not have a successful
disabling pain reduction
– Of ten patients
that underwent
open reduction
0% had pain at
follow-up
22 (64.7%) Full population: NR Full population: Halo traction not effective – Cervical translation, at or
patients did not 3 (8.8%) 11 (32.4%) as a means of obtaining adjacent to the injury level,
achieve closed reduction seen more frequently w/
anatomical Nonoperative: Nonoperative: – Ten (36%) achieved nonoperative treatment
reduction: 3 (13%) 10 (42%) anatomical reduction (38%) versus operative
– 15 patients had Showed no change or – Seven (25%) remained (20%)
imperfect possibly deterioration at Operative: dislocated – Solid fusion in ten (100%)
reduction follow-up 1 (10%) – Eleven (39%) showed patients treated operatively
– seven patients some improvement – Spontaneous fusion in 13
(1991)
Beyer
(all in patients (54%) in

Operative:
nonoperative nonoperative group
0 (0%)
treatment – A natomical reduction was
group) were attained more frequently by
left in operative intervention (60%
dislocated versus 25%)
position – Less than anatomical
reduction is a risk factor for
cervical translation,
regardless of treatment
References 4. Shapiro S, Snyder W, Kaufman K, et al (1999) Outcome

of 51 cases of unilateral locked cervical facets: inters-
1. Harrington JF, Jr, Park MC (2007) Single level arthrodesis pinous braided cable for lateral mass plate fusion
as treatment for midcervical fracture subluxation: a compared with interspinous wire and facet wiring
cohort study. J Spinal Disord Tech; 20:42–48. with iliac crest. J Neurosurg; 91:19–24.
2. Henriques T, Olerud C, Bergman A, et al (2004) Distrac- 5. Beyer CA, Cabanela ME, Berquist TH (1991) Unilateral
tive flexion injuries of the subaxial cervical spine facet dislocations and fracture-dislocations of the
treated with anterior plate alone. J Spinal Disord Tech; cervical spine. J Bone Joint Surg Br; 73:977–981.
17:1–7. 6. Rorabeck CH, Rock MG, Hawkins RJ, et al (1987) Uni-
3. Kwon BK, Fisher CG, Boyd MC, et al (2007) A prospec- lateral facet dislocation of the cervical spine. An
tive randomized controlled trial of anterior compared analysis of the results of treatment in 26 patients.
with posterior stabilization for unilateral facet injuries Spine (Phila Pa 1976); 12:23–27.
of the cervical spine. J Neurosurg Spine; 7:1–12.

Global Spine Congress
2011
Barcelona, Spain
Catalonia Palace of Congresses
March 24–26, 2011
The inaugural 2009 Global Spine Congress in San Francisco received tremendous acclaim from the
global spine community. Building on that success, we are pleased to announce that the next Global
Spine Congress, designed to meet the specific needs and interests of spine specialists from around
the world, will take place in Barcelona, Spain.
Chairperson: Jeffrey wang USA

Co-chairpersons: Luiz vialle Brazil, Michael Janssen USA, Bartolomé Marré Chile
Local Host: Salvador Fuster Spain
organized by
www.globalspinecongress.org
Systematic review—Addressing the challenges and controversies of managing os odontoideum: results of a (…) 67
67—74
Addressing the challenges and controversies

of managing os odontoideum: results of a
systematic review
Authors Jefferson R Wilson1, Joseph R Dettori 2 , Ellen M VanAlstyne2 , Michael G Fehlings1
Institutions 1 Division of Neurosurgery, University of Toronto, Toronto, Ontario, Canada
2
Spectrum Research, Tacoma, Washington, USA
Abstract
Introduction: Os odontoideum is a rare condition with a controversial pathogenesis and poorly un-
derstood natural history. As a result, it is difficult for clinicians to predict which patients require
surgical fusion to prevent symptomatic progression and potentially devastating neurologic injury.
Methods: We undertook a systematic review of the literature to evaluate the clinical outcomes in the
treatment of asymptomatic and symptomatic cases of os odontoideum. Pubmed, EMBASE, Co-
chrane, National Guideline Clearinghouse Databases as well as bibliographies of key articles were
searched. Articles were reviewed by two independently working reviewers. Inclusion and exclu-
sion criteria were set and each article was subject to a predefined quality rating scheme.
Results: We identified eleven articles meeting our inclusion criteria. Seven studies included patients
with asymptomatic os odontoideum discovered incidentally from which it was possible to examine
data on 18 individuals. Six studies were identified containing more than 15 patients with symp-
tomatic os odontoideum.
Conclusion: The existing literature regarding both asymptomatic or symptomatic os odontoideum and
its treatment is very limited. No definitive conclusions can be drawn from these studies. A basic
approach to the management of os odontoideum is offered based on the findings of this study.
This systematic review was funded by AOSpine.

68 Systematic review—Addressing the challenges and controversies of managing os odontoideum: results of a (…)
STUDY RATIONALE AND CONTEXT MATERIALS AND METHODS
Os odontoideum is a rare condition defined radiographi- Study design: Systematic review.

cally as an ossicle with smooth circumferential cortical
margins representing the odontoid process that has no Sampling:
osseous continuity with the body of C2. It may be classi- • Search: Pubmed, EMBASE, Cochrane, and Na-
fied as stable or unstable based on the extent of excur- tional Guideline Clearinghouse Databases; bibli-
sion of the atlas from the axis on dynamic imaging [1]. ographies of key articles
The pathogenesis of this lesion remains controversial • Dates searched: 1970 to August 2009.
with arguments for both acquired and congenital causes • I nclusion criteria
postulated in the literature. At present, however, most A symptomatic: case reports, case series of any
authors believe that it results from a previous trauma size, adults and children
leading to a chronic nonunion fracture of the odontoid Symptomatic: case series that included 15 or more
process [2–5]. patients, adults and children
Although os odontoideum has a clear radiographic defi- • Outcomes: fusion (%), relief of symptoms
nition, its clinical manifestations are variable with pa- • A nalysis: descriptive statistics
tients existing on a spectrum of symptom severity from
completely asymptomatic or with neck pain to severe Details about methods can be found in the web
spinal cord injury. Due to the paucity of cases and the appendix at www.aospine.org/ebsj.
poorly understood natural history of this condition it is
difficult for clinicians to predict which patients require
surgical fusion to prevent symptomatic progression and
potentially devastating neurologic injury. RESULTS
In order to review the collective experience and to help
augment our current understanding of this condition we e identified eleven articles meeting our inclusion crite-
W
have undertaken a systematic review of the literature to ria (Fig 1). Seven studies included patients with asymp-
evaluate the clinical outcomes in the treatment of as- tomatic os odontoideum discovered incidentally from
ymptomatic and symptomatic cases of os odontoideum. which it was possible to examine data on 18 individuals
[3, 4, 6–10]. Six studies were identified containing more
than 15 patients with symptomatic os odontoideum [2,
3, 5, 10–12].
Asymptomatic incidental os odontoideum

• Eighteen individual asymptomatic patients were
identified, for which the demographics, treatment,
and outcomes are shown in Table 1 [3, 4, 6–10].
• A mong the eleven patients for which demographic
data were supplied, 64% were pediatric, and 64%
were male. Three had Klippel-Feil syndrome, two had
Down’s syndrome, and one had ectodermal dysplasia.
• Follow-up time was reported in only seven of the pa-
tients and ranged from 1.4–11 years.
• Nine patients were treated with spinal fusion proce-
dures to stabilize C1 and C2, and in these cases, solid
fusion was achieved. These patients continued to be as-
ymptomatic, although one paper pointed out that the
neck range of motion was reduced secondary to fusion.
• Nine patients were treated conservatively, and these
were reported to remain stable with no symptoma-
tolgy attributable to their os odontoideum having de-
veloped over the course of follow-up. One patient died
from pulmonary cancer at 11-years follow-up, but his
spinal condition remained stable until that time.
Symptomatic os odontoideum In the one retrospective cohort study, patients were

• Three-hundred-and-forty-four patients in six studies grouped by cord sign status and treatment mode [3].
with symptomatic os odontoideum are summarized In patients with no cord signs, symptoms resolved in
with respect to demographics, treatment, and out- 50% of those patients who were treated conserva-
comes, Table 2 [2, 3, 5, 10–12]. Ages ranged from tively and only 11% of those treated with surgery. In
3–73 years and 62% were males. Forty-seven pa- patients with cord signs, 25% of patients were symp-
tients had a preexisting diagnosis including Klippel- tom free whether they received conservative or op-
Feil syndrome, Down’s syndrome, dysplasia, occipi- erative treatment (Fig 2).
talization, hypertrophy of the anterior atlantal arch,
basilar impression, spondyloepiphyseal dysplasia,
metatrophic dwarfism, or Morquio’s syndrome. All
but 19 patients underwent cervical fusion. CLINICAL GUIDELINES
• Fusion rate
The rate of fusion reported in four studies was 99% In 2001, the American Association of Neurological Sur-
(214/216) [5, 10–12]. geons and the Congress of Neurological Surgeons
• Symptoms (AANS/CNS) provided evidence-based clinical recom-
Three case series report on the change in symptoms mendations for the treatment of os odontoideum [1].
following spinal fusion. After fusion, symptoms were • No clinical evidence was found supporting the rec-
completely resolved or significantly reduced in all pa- ommendation of operative treatment standards or
tients (n = 39) in one study [10], and completely re- guidelines for os odontoideum.
solved in 88% in another [2]. Klimo et al reported • For asymptomatic patients, clinical and radiographic
that symptoms resolved in 68% of patients with pain surveillance may be the appropriate management.
and 39% with neuropathy (20% with myelopathy • Patients with neurological symptoms/signs and C1–2
and 100% with intermittent neuropathy) [5]. It is un- instability are generally managed with posterior fix-
clear how many different patients this represents of ation and fusion.
the total study population as patients may have more • Other operative measures may be warranted in cases
than one symptom. No case series reported worse of irreducible cervicomedullary compression and/or
symptoms or neurological status following treatment. occipitocervical instability.
• The paucity of high quality studies was noted.
Fig 1 Flow chart showing results of literature search
6. A symptomatic cases (n = 7)
1. Total citations (n = 74) 3. Retrieved for full-text 5. Publications

evaluation (n = 41) included (n = 11)
7. Symptomatic series (n = 6)
2. E xcluded after title 4. E xcluded after
/abstract review (n = 33 ) full-text review (n = 30)
Fig 2 Symptom status of patients without and with cord signs after surgical or nonsurgical treatment of
symptomatic os odontoideum [3]
worse unchanged improved resolved
Surgery
No cord (n = 9)
symptoms
(n = 24) No Surgery
(n = 15)
Surgery
Cord (n = 8)
symptoms
(n = 8) No Surgery
(n = 4)
0 10 20 30 40 50 60 70 80 90 100
Percent

Table 1 Asymptomatic patients with incidental discovery of os odontoideum, their treatment and results
Patient Age Preexisting operative Follow-up
no. Author (years) Sex diagnosis Instability fusion (years) Results
1 Dai* nr nr nr nr no nr – Remained stable
6 Dai* nr nr nr nr yes nr – Solid arthrodesis achieved
7 Dai* nr nr nr nr yes nr – Solid arthrodesis achieved
8 Forlin 8 F Down’s syndrome yes yes 5 – Postoperative pin track infection
resolved after halo removal
– Partial resorption of the fusion mass
treated by regrafting at 3 months
– A symptomatic
9 Forlin 9 M Down’s syndrome yes yes 2 – A symptomatic
10 Hickam 9 F none no yes nr – A symptomatic
– Decrease in rotation of head
11 Juhl† 9 M nr yes yes nr – E xtension 75°, lateral flexion right 0°
and left 30°, rotation right 30° and left 45°
12 Juhl† 44 M nr no no 11† – A symptomatic
13 Morgan 39 M Klippel-Feil no no nr – A symptomatic
14 Morgan 64 F Klippel-Feil no no nr – A symptomatic
15 Sankar 13 M None yes yes 1.4 – Fusion achieved at 1.9 months
16 Sankar 12 M Ectodermal yes yes 2.8 – Fusion achieved at 5.9 months
dysplasia
17 Sankar 2 F Klippel-Feil yes yes 2.2 – Fusion achieved at 12.9 months
18 Spierings 28 M None no no 9 – A symptomatic
nr = not reported; Klippel-Feil type fusion of C2 and C3.

* Dai: mean age 24.6 and age range 7–56 years, and follow-up of 6.5 years (1–16) are reported for the total number of asymptomatic and symptomatic
patients, N = 44.
† Juhl: patient died of pulmonary cancer 11 years after diagnosis of os odontoideum.
Table 2 Symptomatic patients treated either nonoperatively or operatively for os odontoideum and their results
Mean Mean
age follow-up
(range) in % Preexisting Insta- Operative (range)
Author N years male diagnosis Symptoms bility treatment in years Results
Dai 39 25 75%* n = 18 Local symptoms: 95% yes 6.5 Fusion: 100%
(7–56) – K lippel-Feil (n = 4) – c ervical pain and/or stiffness (n = 33) (1–16)
– o ccipitalization (n = 6) – weakness of neck (n = 21) All symptoms
– Hypertrophy anterior – dizziness (n = 7) – resolved: 67%
atlantal arch (n = 5) – torticollis (n = 9) – improved: 33%
– Basilar invagination
(n = 2) Neurological findings (n = 30):
– Basilar impression – transient (n = 7)
(n = 1) – immediate, nontransient (n = 4)
– late-onset progressive (n = 19)
– myelopathy (n = 6)
– radiculopathy(n = 2)
– myelopathy & radiculopathy (n = 22)
– cranial nerve defects (n = 8)
Fielding 35 19 60% – Down’s syndrome (n = 1) – Pain (n = 16) 100% yes 3.4 All symptoms
(3–65) – K lippel-Feil (n = 1) – Neurological (n = 17) (1–10) – resolved: 88%
– D ysplasia (n = 4) – O ther (n = 14) – worse: 0%
Gluf 22 10 66%† ns Not described 100% yes 2.8 Fusion: 100%
(4–16) (0.25–8.9)
Klimo 78 21 62% n = 5 – Pain (n = 50) 77% yes 1.2 Fusion: 100%
(1.5–73) –D own’s syndrome (n = 3) – Myelopathy (n = 18) ant: 70% (0.1–9.6) Pain
– Spondyloepiphyseal – Intermittent neuropathy (n = 15) post: 10% – resolved: 68%
dysplasia (n = 1) A-P: 13% – improved: 20%
– Metatrophic dwarfism – unchanged: 12%
(n = 1) Neuropathy
– resolved: 39%
– improved: 50%
– unchanged: 11%
Menezes 134 4–58 55% –M
orquio’s (n = 2) – Acute neurological deterioration yes yes ns Fusion: 98.5%
–D
own’s syndrome following trauma (n = 63) – 2nd fusion required
(n = 10) ‡ – Symptoms insidious (n = 71) to extend initial
C1-C2 fusion (n = 16)
Spierings 36 38 78% – Down’s syndrome Group A: no cord symptoms (n = 15) nr no 8 All symptoms resolved:
(6–62) (n = 1) (0.5–18) – Gp A: 50%
Group B: no cord symptoms (n = 9) nr yes – Gp B: 11%
Group C: cord symptoms (n = 4) nr no – Gp C: 25%
– Gp D: 25%
Group D: cord symptoms (n = 8) nr yes improved:
– Gp A:14%
– Gp B: 45%
– Gp C: 50%
– Gp D: 37%
unchanged:
– Gp A: 29%
– Gp B: 22%
– Gp C: 0%
– Gp D: 13%
worse:
– Gp A: 7%
– Gp B: 22%
– Gp C: 25%
– Gp D: 25%
nr = not reported.
* Dai: % male includes asymptomatic and symptomatic patients.
† Gluf reports % male for a study population of 45 patients with other diagnoses and 22 os odontoideum patients.
‡ Menezes: possibly other patients have preexisting diagnoses; these 12 patients are specifically listed as among those with reducible lesions but
worsening extension.

EVIDENCE SUMMARY
Question 1: Outcomes of treatment in patients with incidental asymptomatic os odontoideum

1. Symptoms Very low Low Moderate High – Patients continued to be symptom free with
or without fusion procedure.
– T he fusion procedure limits neck range of motion,
but other detrimental effects were not noted.
2. Fusion success Very low Low Moderate High – Fusion was successful in most cases.
Question 2: Outcomes of treatment in patients with symptomatic os odontoideum

1. Symptoms Very low Low Moderate High – Symptoms were relieved in the majority
of cases after fusion.
– Symptoms were also seen to improve in the
majority of nonoperative patients.
2. Fusion success Very low Low Moderate High – Fusion was successful in most cases.
Illustrative case was the treatment of choice resulting in successful

• A 24-year-old woman presents with a history of arthrodesis in 99% of the cases. One retrospective
chronic neck pain, worse in the last 2 months, and a cohort study suggests that the outcome from surgery
cervical spine x-ray demonstrating os-odontoideum may not be better than the outcome of conservative
(Fig 3) with 7mm of excursion of the atlas from the axis care.
between flexion and extension. Physical examination • A lthough the etiology of os odontoideum remains
revealed evidence of myelopathy including brisk deep unclear, there are two plausible theories which may
tendon reflexes, as well as positive Hoffman and Babin- account for two different forms of this lesion: 1) it
ski signs bilaterally. MRI performed shortly after pre- represents a congenital aplastic lesion, 2) it is an ac-
sentation was concerning for increased T2 signal in the quired entity related to a chronic non-union from a
upper cervical spinal cord (Fig 4). The patient underwent previous traumatic event. For the purposes of this
C1/2 posterior screw-rod fixation, with iliac crest bone study it is difficult to distinguish between these two
graft which was fixed in situ with C1–2 cable cerclage subpopulations and hence to draw any conclusions
fixation. At 6-months postoperatively, she had experi- about differences in natural history between them.
enced complete resolution of her neck pain symptoms The existing literature regarding either asymptomat-
and has achieved a solid posterior fusion on x-ray (Fig 5). ic or symptomatic os odontoideum and its treatment
The patient remains well at 2-year follow-up. is very limited with only case reports, case series and
one retrospective cohort study available for analysis.
Discussion No definitive conclusions can be drawn from these
• I n 18 case reports, we found no cases of patients with studies. Further comparative studies with more pa-
incidental asymptomatic os odontoideum who later tients would be necessary to address the question of
developed symptoms whether treated conservatively the benefit of surgery relative to the severity of the
or surgically. Among symptomatic patients, fusion symptoms.
Fig 3 Preoperative lateral C-spine x-ray demonstrat-

• F or asymptomatic patients, posterior C1/2 fixation
ing os odontoideum
and fusion has been shown to be a safe procedure,
however there is no convincing evidence to support
its use over conservative therapy. The pros and cons
of operative versus nonoperative management need
to weighed by the clinician and discussed with the
patient.
• For symptomatic patients (severe neck pain, occipital
neuralgia or neurological dysfunction), surgical
treatment is generally recommended. In these cases,
posterior C1/2 fixation and fusion is a safe and effec-
tive procedure that leads to improvement or stabili-
zation of symptoms in the majority of patients.
Fig 4 Sagittal T2 MRI demonstrating increased signal

in the upper cervical spinal cord at the level of the os
odontoideum.
Fig 5 Lateral C-spine x-ray 2-years postoperatively

posterior C1/2 fusion

References
1. The Section on Disorders of the Spine and Peripheral

Nerves of the American Association of Neurological Sur-
geons and the Congress of Neurological Surgeons (2001)
Os Odontoideum. SpineUniverse Web site. http://www.
spineuniverse.com/pdf/traumaguide/19.pdf. Accessed Au-
gust 15, 2009.
2. Fielding JW, Hensinger RN, Hawkins RJ (1980) Os Od-
ontoideum. J Bone Joint Surg Am; 62:376–383.
3. Spierings EL, Braakman R (1982) The management of
os odontoideum. Analysis of 37 cases. J Bone Joint Surg
Br; 64:422–428.
4. Sankar WN, Wills BP, Dormans JP, et al (2006) Os odon-
toideum revisited: the case for a multifactorial etiolo-
gy. Spine (Phila Pa 1976); 31:979–984.
5. Klimo P, Jr., Kan P, Rao G, et al (2008) Os odontoideum:
presentation, diagnosis, and treatment in a series of
78 patients. J Neurosurg Spine; 9:332–342.
6. Juhl M, Seerup KK (1983) Os odontoideum. A cause of
atlanto-axial instability. Acta Orthop Scand; 54:113–118.
7. Morgan MK, Onofrio BM, Bender CE (1989) Familial os
odontoideum. Case report. J Neurosurg; 70:636–639.
8. Hickam HE, Morrissy RT (1990) Os odontoideum de-
tected on a lateral cephalogram of a 9-year-old orth-
odontic patient. Am J Orthod Dentofacial Orthop; 98:
89–93.
9. Forlin E, Herscovici D, Bowen JR (1992) Understanding
the os odontoideum. Orthop Rev; 21:1441–1447.
10. Dai L, Yuan W, Ni B, et al (2000) Os odontoideum: eti-
ology, diagnosis, and management. Surg Neurol; 53:
106–108.
11. Menezes AH (1999) Pathogenesis, dynamics, and
management of os odontoideum. Neurosurg Focus;
6:e2.
12. Gluf WM, Brockmeyer DL (2005) Atlantoaxial transar-
ticular screw fixation: a review of surgical indica-
tions, fusion rate, complications, and lessons learned
in 67 pediatric patients. J Neurosurg Spine; 2:164–169.
Case report—Metachronous presentation of metastasis from renal cell carcinoma: evaluation and (…) 75
75—82
Metachronous presentation of metastasis from

renal cell carcinoma: evaluation and
management of spinal metastasis
Joshua C Patt, Jeffrey S Kneisl
Authors
Institution
Department of Orthopaedic Surgery and Blumenthal Cancer Center, Carolinas Medical Center,
Charlotte, NC, USA
Introduction
Surgical treatment of metastatic cancer in the appendic- As controversial as surgical indications are in the appen-
ular skeleton is well supported in the literature. Straight- dicular skeleton, they are at times even more conten-
forward indications include pathologic fracture and im- tious in the axial skeleton. Issues such as established
pending pathologic fracture [1, 2]. More controversial neurologic deficit as well as impending neurologic de-
indications for operative treatment exist regarding resec- cline are compounded by uncertain criteria for stability
tion or en-bloc removal for solitary metastases or other and pathologic fracture. The purpose of this current re-
painful metastases [3]. Overall, surgical decision mak- port is to describe the unusual presentation of a symp-
ing must be tempered by the patient’s overall perfor- tomatic spinal metastasis in the setting of systemic dis-
mance status (ECOG, Eastern Cooperative Oncology ease, review the indications and treatment and then
Group), tumor type (with relation to expected survival consider what was done using an “evidence-based medi-
and relative radiosensitivity) and the patient’s perceived cine approach”.
ability to recover from surgery [A, 4].
This case report has received no financial support.

Devices described are FDA approved.

76 Case report—Metachronous presentation of metastasis from renal cell carcinoma: evaluation and (…)
Case report The femoral lesions were treated with simple intramed-
ullary nailing. The ischial lesion was embolized preop-
A 64-year-old woman was seen in surgical orthopedic eratively and then treated with curettage and local adju-
oncologic consultation for a new metastasis to the right vant (hydrogen peroxide [5] and electrocautery)
distal femur. She was originally diagnosed with renal followed by packing of the cavity with PMMA. Opera-
cell carcinoma 8 years prior, with metastatic disease to tive approach for the ischial lesion was a straight poste-
her mediastinal lymph nodes discovered 5 years later but rior approach as utilized for hamstring avulsion repairs.
had been considered in a stable disease state on systemic The sciatic nerve was identified and manipulation mini-
therapy with a tyrosine kinase inhibitor. In addition to mized. Postoperatively the patient had severe left-leg sci-
her femoral disease she had a several year history of low atic symptoms requiring escalating doses of narcotics
back pain and occasional radiating ipsilateral right lower and gabapentin therapy.
extremity pain. Her femur showed a lytic metastasis of
the distal femoral metastasis. This was treated with cu- After several weeks of minimal improvement, the pa-
rettage, local adjuvant (hydrogen peroxide and electro- tient was given an L4/5 translaminar epidural steroid in-
cautery) and poly-methyl-methacrylate (PMMA) aug- jection. This did provide her with moderate pain relief
mentation. Her back pain was treated nonoperatively at and allowed her to slowly wean her narcotic require-
this time. Approximately 1 year later, the patient report- ment. After a slow return of her left leg pain, a second
ed new pain in the left thigh associated with activity. A injection was given 3 months later with similar but
new technetium bone scan identified two distinct lesions slightly less pain relief. This postoperative course was
in the left femur and an additional lesion in the left ischi- complicated by the patient also receiving radiation ther-
um. Options were discussed with the patient and it was apy to 3000 Gy to her ischium and left femur.
decided to proceed with surgical treatment of the femo-
ral lesion for a symptomatic metastasis and to address Six months following her last surgery the patient re-
the ischial lesion at the same surgical setting. turned for an unscheduled visit with 2 weeks of crescen-
Fig 1 Coronal MRI femur‘07 Fig 2 Right femur postoperative
Fig 3 Sagittal MRI spine ‘07 Fig 4 Left ischial metastasis
do low back pain and radiation down her bilateral poste- verse process and pedicle down to the vertebral body of
rior thighs. Noting the significant change in her pain L5. Bleeding was well controlled due to a thorough pre-
without obvious inciting event, a new CT and MRI were operative embolization and a controlled and methodical
ordered of her lumbar spine. These studies revealed a tumor resection.
lytic lesion in her right L5 pedicle with expansion of the
pedicle and right-sided nerve root impingement. Options After local adjuvant neoplasia treatment with peroxide,
were discussed with the patient including radiation ther- electrocautery and a high speed diamond tip burr, the
apy, embolization, surgical treatment and various com- L4/5 and L5/S1 disks were removed and posterior lumbar
binations of the above. Based on her baseline degenera- interbody fusion was carried out with transforaminal in-
tive lumbar spine problems, acute pain exacerbation, terbody allograft cages, local autograft and cancellous al-
and the relatively poor radiation sensitivity of this tu- lograft bone and posterolateral arthrodesis completed
mor, it was decided to proceed with surgery. with decortication, bonegraft placement and placement
of rods and crosslink. Meticulous wound closure with
Surgical intervention nonresorbable sutures, intended to be left in place for an
Following preoperative embolization within 24 hours of extended time, was carried out. The patient’s postopera-
planned surgery, we performed a wide posterior approach tive course was unremarkable. She did receive postopera-
and placed pedicle screws bilaterally at L4 and S1 and tive radiation therapy beginning at 3 weeks postopera-
unilaterally on the left side of L5. Inferior facetectomy of tively. The patient had an unremarkable postoperative
L4 was performed along with laminectomy of L5 to allow course, was off of narcotic pain medicines by the 2-week
isolation of the L5 posterior elements and pedicle. At this follow-up visit and has had durable pain relief of her
point the tumor was excised in an intralesional manner. baseline back pain and her lower extremity radicular
Gross total excision of the tumor was performed with re- symptoms with no evidence of tumor persistence or re-
section of the entire pars, inferior articular facet, trans- currence at short-term (6-month) follow-up.
Fig 5 Intraoperative left femur and ischium (patient

in prone position) Fig 6 Sagittal MRI spine ‘09
Fig 7 Para-sagittal MRI spine ‘09 Fig 8 CT with L5 metastasis

Learning points cinoma and melanoma are remarkably insensitive for

Patients with metastatic cancer should undergo sched- the treatment of bulky disease [10].
uled surveillance staging studies relevant to their given
disease. Timing intervals and choice of studies can be se- In terms of clinical progression a patient with an estab-
lected based on established guidelines such as the Na- lished cancer diagnosis deserves a low threshold for or-
tional Comprehensive Cancer Network [6]. dering advanced imaging studies with the onset of new or
crescendo pain. This patient had confounding baseline
Surgical treatment of metastatic disease does not reliably low back pain and then experienced nonspecific exacer-
provide a cure for patients in isolation. All patients should bation of radicular pain following surgical intervention
be managed by a multidisciplinary team which can infor her tumor disease in another region. Listening to the
clude surgeons as well as medical oncologists and radia- patient and not relying on a negative result from a bone
tion oncologists. Metastatic disease represents a systemic scan prompted the treating physician to order a new MRI
disease and the only meaningful opportunity for a cure when the character and severity of the pain changed.
will necessitate systemic therapy which can include tra-
ditional cytotoxic agents as well as hormonal therapies or The choice of surgical intervention is frequently chal-
novel targeted chemotherapy agents (for instance ty- lenging in these patients. Attempted en-bloc resection in
rosine kinase inhibitors). the appendicular skeleton has been discussed extensive-
ly and does not convincingly result in a durable cure.
Each type of primary cancer has a unique natural histo- Consideration can be given to this type of resection in
ry. Patients with certain tumors such as lung cancers the setting of solitary metastases [B, 11]. Unfortunately
have a typically short expected life span versus those the likelihood of being able to achieve an en-bloc resec-
whose tumors such as breast or renal cell which can have tion with true negative margin in the spine is low and
a much more indolent but progressive course [7, 8, 9]. the added risk is typically not justified by the expected
Any operation considered should include a thoughtful benefit. Patients also almost uniformly get postoperative
discussion with the patient regarding the expected re- radiation therapy and this can also adversely affect fu-
covery from the planned intervention in the context of sion rate [12]. The length of the reconstruction construct
their expected longevity. For a major intervention such should be considered and a longer construct is typically a
as spine surgery, patient life expectancy less than 3 better option. Despite thoughtful interventions, local re-
months has been considered a contraindication for spine currence is more of the rule than the exception, with
surgery. implications for possible future surgery in case of tumor
recurrence. For example, in the present case the rods
The radiosensitivity of the offending tumor is a very im- were left long caudally intentionally. The side-loading
portant variable. For example, patients with myeloma or system utilized (Synthes USS, Synthes, Paoli, PA) in this
lymphoma rarely require surgical intervention due to case would allow the placement of iliac bolts with the
their response to chemotherapy and radiation therapy. current construct left intact and proximal extension also
Breast cancer is also frequently responsive to traditional possible without extensive local re-exploration in a vas-
external beam radiation therapy whereas renal cell car- cular, postirradiated bed.
Fig 9 Postoperative spine lateral Fig 10 Postoperative spine AP
Conclusion References
The current gold standard regarding the surgical treat-
ment of metastatic spine disease largely relies on a study Key references
by Patchell et al, which recommended direct decompres- A. Tokuhashi Y, Matsuzaki H, Toriyama S, et al (1990) Scor-
sive surgery for non-myeloid spine tumors with epidural ing system for the preoperative evaluation of meta-
spinal cord compression and either neurologic deficit or static spine tumor prognosis. Spine; 15:1110–1113.
impending compromise [C]. Unfortunately, this article B. Sundaresan N, Rothman A, Manhart K, et al (2002) Sur-
didn’t address disease manifestation at the root level. gery for solitary metastases of the spine; Rationale
Our further resources are mainly limited to the natural and results of treatment. Spine; 27:1802–1806.
(treated or untreated) history of cancers from a variety C. Patchell RA, Tibbs PA, Regine WF, et al (2005) Direct
of different origins, anticipated responsiveness of these decompressive surgical treatment of spinal cord com-
tumors to radiation therapy and guidelines regarding the pression caused by metastatic cancer: A randomized
pre-operative evaluation and selection of appropriate trial. Lancet; 366:643–648.
surgical candidates. Unfortunately this is a complex de-
cision making process, without concrete answers in vali- Other cited references:
dated reference materials. This leads to individualized 1. Mirels H (1989) Metastatic disease in long bones. A
treatment approaches using multidisciplinary resources, proposed scoring system for diagnosing impending
including radiation and medical oncologists. pathologic fractures. Clin Orthop Relat Res; 249: 256–
264.
2. Harrington KD (1986) Impending pathologic fractures
from metastatic malignancy: evaluation and man-
agement. InstrCourse Lect; 35:357–381.
3. Bickels J, Kollender Y, Wittig JC, et al (2005) function
after resection of humeral metastases: analysis of 59
consecutive patients. Clin Orthop Relat Res; 437:201–
208.
4. Oken MM, Creech RH, Tormey DC, et al (1982) Toxicity
and response criteria of the Eastern cooperative On-
cology Group. Am J Clin Oncol; 5:649–655.
5. Nicholson NC, Ramp WK, Kneisl JS, et al (1998) Hydro-
gen peroxide inhibits giant cell tumor and osteoblast
metabolism in vitro. Clin Orthop Relat Res; 347:250–
260.
6. www.nccn.org
7. Althausen P, Althausen A, Jennings LC, et al (1997)
Prognostic factors and surgical treatment of osseous
metastases secondary to renal cell carcinoma. Cancer;
80:1103–1109.
8. Koizumi M, Yoshimoto M, Kasumi F, et al (2003) Com-
parison between solitary and multiple skeletal meta-
static lesions of breast cancer patients. Ann Oncol;
14:1234–1240.
9. Sugiura H, Yamada K, Sugiura T, et al (2008) Predictors
of survival in patients with bone metastasis of lung
cancer. Clin Orthop Relat Res; 466:729–736.
10. Frassica DA (2003) General principles of external
beam radiation therapy for skeletal metastases. Clin
Orthop Relat Res; 415 Suppl:S158–164.
11. Tomita K, Kawahara N, Kobayashi T, et al (2001) Surgi-
cal strategy for spinal metastases. Spine; 26:298–306.
12. Bouchard JA, Koka A, Bensusan JS, et al (1994) Effects
of irradiation on posterior spinal fusions: A rabbit
model. Spine; 19:1836–1841.

Evidence-based discussion of case report tient has had a relatively indolent renal cancer as evi-
denced by 10-year survival following diagnosis. Her dis-
Laurence D Rhines
Author ease cadence has accelerated in recent years with
University of Texas, MD Anderson Cancer
Institution progressive bony metastasis, however, her overall dis-
Center, Department of Neurosurgery, Houston, Texas ease burden remains limited. Her visceral disease re-
mains controlled (lymph node disease). Her bony dis-
This is a case of a 64-year-old woman with a long history ease, while progressive, is limited. Finally, she has good
of renal cell carcinoma. This was originally diagnosed performance status. Her predicted survival warrants in-
approximately 9.5 years prior to her spinal surgery and, tervention and said intervention must provide reason-
following her nephrectomy, she remained disease free able local tumor control.
for 5 years. At this time mediastinal lymph node disease
was identified and controlled successfully with tyrosine Treatment options
kinase inhibitors. More recently her disease has become 1. Surgical resection
more active, with progressive osseous metastasis. A right a) Intralesional resection
femoral lesion was treated with curettage 8 years after b) En-bloc resection
diagnosis, two femoral lesions and an ischial lesion were 2. Radiation therapy
treated with nailing and curettage 9 years after diagno- a) Standard external beam radiotherapy
sis, and 6 months later she presents with a symptomatic b) Spinal stereotactic radiotherapy
lesion at L5 causing crescendo back pain radiating to the 3. Chemotherapy
thighs. This was treated with preoperative embolization 4. Combination /sequential /pulsed therapy
followed by intralesional resection and posterior stabili- (surgery, radiation, chemotherapy)
zation. She then received standard external beam radio-
therapy 3-weeks postoperatively. Six months later, her Successful treatment of metastatic spine disease requires
pain is relieved and she has no evidence of tumor an understanding of the nature of the patient’s symp-
recurrence. toms and their etiology. The type of symptoms will often
dictate the method of treatment. Metastatic epidural spi-
Critical review of this case centers around two key nal cord compression is best managed with surgical de-
issues, prognosis and treatment options: compression and stabilization except in the most radio/
chemo-sensitive cases (lymphoma and myeloma) [C, in
Patient prognosis factors key references]. The treatment of root compression is
a. Histology (single greatest predictor of survival) more flexible due to lack of available higher-level evi-
i. Poor survival—lung, colorectal, melanoma dence. Local/biological tumor pain typically responds to
ii. Long survival—breast, prostate steroids and any treatment that shrinks the tumor. Me-
iii. Renal cell carcinoma (intermediate, chanical pain and pain from instability are less likely to
variable natural history) be relieved with radiotherapy and/or chemotherapy and
b. Extent of metastatic disease will typically need surgical intervention ranging from
i. Visceral metastasis cement augmentation to surgical stabilization depending
ii. Bony metastasis on the degree of instability.
c. Performance status of patient
i. Karnovsky performance score In the present case, surgical resection and stabilization
ii. Neurologic status was used to treat this L5 metastasis with excellent results
at 6 months. The rationale for proceeding with surgery
Clearly the goal of treatment for metastatic spine disease was the acuity of the pain, baseline degenerative disease,
is effective palliation of patient’s symptoms reflective of and the radiation resistance of renal cell carcinoma. Ap-
anticipated survival time. In deciding on a plan of treat- propriately, embolization was performed prior to surgery
ment, an appreciation of the patient’s overall prognosis to reduce intraoperative blood loss. It is unclear that
is critical in determining how to proceed [1, 2]. This pa- there was any overt spinal instability requiring surgical
stabilization and EBRT was used postoperatively to im- Discussion References

prove local control. En bloc resection was not utilized
presumably based on the patients overall disease status 1. Tomita K, Kawahar N, Kobayashi T, et al (2001) Surgical
and progressive bony metastasis. strategy for spinal metastases. Spine; 298–306.
2. Tokuhashi Y, Matsuzaki H, Oda H, et al (2005) A revised
Of the available treatment options, chemotherapy is the scoring system for preoperative evaluation of meta-
least likely to provide a timely response to the patients static spine tumor prognosis. Spine; 30(19): 2186–
presenting symptoms. Moreover, having failed tyrosine 2191.
kinase inhibitors, the probability of subsequent systemic 3. Gerszten PC, Burton SA, Ozhasoglu C, et al (2007) Ra-
therapy being efficacious is reduced. Therefore, most diosurgery for spinal metastases: clinical experience
centers would not opt for systemic therapy alone to treat in 500 cases from a single institution. Spine; 32(2):
this patient’s L5 disease. From the surgical standpoint, 193–199.
en-bloc excision has been advocated as a treatment for 4. Yamada Y, Bilsky MH, Lovelock DM, et al (2008) High-
solitary metastasis from renal cell carcinoma [1, 2]. This dose, single-fraction image-guided intensity-modu-
patient’s disease, however, is not isolated to the spine. lated radiotherapy for metastatic spine lesions. Int J
She has nodal disease as well as increasing bony involve- Radiation Oncology Biol Phys 71(2): 484–490.
ment over the last 2 years. Intralesional excision, as per- 5. Chang EL, Shiu AS, Mendel E, et al (2007) Phase I/II
formed in this case, is clearly a reasonable choice and has study of stereotactic body radiotherapy for spinal me-
led to a successful outcome in this patient at 6-month tastasis and its pattern of failure. J Neurosurg Spine; 7:
follow-up. Finally, one must consider radiation options. 151–160.
Renal cell carcinoma is a radio-resistant histology in the
context of standard external beam radiotherapy (typi-
cally administered 30 Gy in 10 fractions). In fact, this
relative radioresistance was cited by the authors as a
reason for not using XRT up front, although it was used
as a postoperative adjuvant. It does not appear that spi-
nal stereotactic radiosurgery was considered, and this
would have been a very reasonable alternative to sur-
gery in this patient. Spinal stereotactic radiosurgery al-
lows for the specific targeting of the radiation to the tu-
mor, while sparing the adjacent sensitive structures such
as the neurological elements, bowel, kidneys, vessels,
etc. This allows the tumor to receive a higher dose per
fraction in fewer fractions (consider 24 Gy in a single
fraction) dramatically increasing the biological effective-
ness of the radiation. This high-dose conformal radiation
has been shown to achieve rapid and significant pain re-
lief (> 85% of select cases), as well as high rates of local
tumor control (85%–90%), with very little morbidity
[3, 4, 5]. Therefore based upon these more recent publi-
cations, spinal stereotactic radiosurgery, where avail-
able, might have been another reasonable option for this
patient. This case report demonstrates the complexities
of trying to apply principles derived from published lit-
erature to an individual. Systematic care delivery is de-
sirable but is put to a test under the myriad of clinical
variations.

Additional EBM notes FROM THE Studies included in EBSS report

EDITORIAL STAFF
Patchell RA, Tibbs PA, Regine WF, et al (2005)
A 2009 EBSS report summarized studies published in Direct decompressive surgical resection in the treatment
the previous 20 years that compared outcomes in pa- of spinal cord compression caused by metastatic cancer:
tients with spinal cord compression secondary to meta- a randomised trial. Lancet; 366(9486): 643–648.
static spine tumor (of various types) who received radia-
tion plus surgery with radiation alone [EBSS Volume 7 Falavigna A, Righesso Neto O, Polesso MA, et al (2007)
Issue 1, 2009]. The summary below speaks to the limita- Metastatic tumor of thoracic and lumbar spine: prospec-
tions of the literature from an “evidence” perspective tive study comparing the surgery and radiotherapy vs
and to some of the challenges related to studying the external immobilization with radiotherapy. Arq Neurop-
question of surgical benefit in metastatic disease. siquiatr; 65(3B): 889–895.
Limited evidence from one moderate quality random- Ghogawala Z, Mansfield FL, Borges LF, et al (2001)
ized controlled trial (CoE II) [Patchell] and three retro- Spinal radiation before surgical decompression adversely
spective cohort studies (CoE III) [Falavigna, Ghogawala, affects outcomes of surgery for symptomatic metastatic
Sorensen] suggests that survival, ability to walk, conti- spinal cord compression. Spine; 26(7): 818–824.
nence, and functional status may be improved and pain
and complications reduced in patients experiencing met- Sorensen S, Borgesen SE, Rohde K, et al (1990)
astatic spinal cord compression by adding surgical inter- Metastatic epidural spinal cord compression. Results of
vention to radiation therapy. A higher percentage of pa- treatment and survival. Cancer; 65(7):1502–1508.
tients receiving combined surgery and radiation versus
radiotherapy alone tended to be able to walk, and to
walk longer, with statistical significance being reached
in two of the three studies that examined this outcome.
These potential benefits should be weighed against the
costs, rigors of recovery from major surgery in patients
whose health is already compromised and life expectan-
cy. In cohort studies, treatment choice based on patient
presentation may bias (confounding by indication) re-
sults comparing treatments. A methodologically rigor-
ous multicenter study may help confirm whether sur-
gery followed by radiotherapy will improve outcomes for
metastatic spinal cord compression patients.
While consideration of evidence is important, as seen in

this case, attention to individual patient presentation and
circumstance must inform the ultimate course of action.
83
Definition of the different classes

of evidence (CoE)
Articles on treatment
Class Study type Criteria
I Good quality RCT • Concealment

• Blind or independent assessment for important outcomes
• Follow-up rate of 85% +
• Adequate sample size
II Moderate or poor quality RCT • Violation of any of the criteria for good quality RCT
Good quality cohort • B lind or independent assessment in a prospective study or use of reliable data* in a retrospective study
• Adequate sample size
• Controlling for possible confounding †
III Moderate or poor quality cohort • Violation of any of the criteria for good quality cohort
Case control • Any case-control design
IV Case series • Any case-series design
* Reliable data are data such as mortality or reoperation.

† Authors must provide a description of robust baseline characteristics, and control for those that are unequally distributed between treatment groups.
Articles on prognosis or risk

Class Study type Criteria
I Good quality RCT • Prospective design
• Patients at similar point in the course of their disease
• Similarity of treatment protocols for patient groups
• Patients followed long enough for outcomes to occur
• Controlling for extraneous prognostic factors*
II Moderate quality RCT • Prospective design, with violation of one of the other criteria for good quality cohort study
• Retrospective design, meeting all the rest of the criteria in class 1
III Poor quality cohort • P rospective design with violation of 2 or more criteria for good quality cohort, or
• R etrospective design with violation of 1 or more criteria for good quality cohort
Case control • Any case-control design
IV Case series • Any case-series design
* Authors must provide a description of robust baseline characteristics, and control for those that are potential prognostic factors.
A unique global community
of spine professionals
AOSpine’s unique structure enables it to connect with surgeons at global, regional, and local
levels. The AOSpine International Board works closely with regional boards and individual country
councils to meet the specific needs of surgeons and other member groups from every corner of
the world, while at the same time upholding AOSpine standards of academic excellence.
The core strength of AOSpine stems from its mem- to peers and colleagues from around the world. Members
bers. Whether surgeon, radiologist, or intern, AOSpine can contribute toward and benefit from numerous academic
embraces diversity of background and culture in an educational and professional events across the relevant
organization committed to education, knowledge creation, therapeutic areas. This way AOSpine is seen as a dynamic
sharing, and growth. AOSpine membership opens the networking platform for all spine professionals at any stage
door to this organization, sparking an instant connection in their career.
join us at www.aospine.org

EBSJ Vol1 Nr1 May10

Caricato da

Informazioni sul documento

Descrizione originale:

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

EBSJ Vol1 Nr1 May10

Caricato da

Copyright:

Formati disponibili

Evidence-Based Spine-Care Journal

volume 1/issue 1 — May 2010

Message from the Editor “The journey of a thousand miles…” 5–8

Science in spine Conducting a winning literature search 9–14

Original research Heterotopic ossification in cervical disc arthroplasty: Is it 15–20

Treating thoracic-disc herniations: Do we always have to go 21–28

Dynamic anterior cervical plating for multi-level spondylosis: 41–46

Kyphoplasty: Traditional imaging compared with computer- 47–50

ProDisc-C versus fusion with Cervios chronOS prosthesis 51–56

Systematic reviews Unilateral facet dislocations: 57–66

Addressing the challenges and controversies of managing os 67–74

Case report Metachronous presentation of metastasis from renal cell 75–82

Evidence-Based Spine-Care Journal

Deputy Regional Editors

Deputy Editor Scientific Methodology: AOSpine Africa: Robert Dunn

Evidence-Based Spine-Care Journal

Message From the editor

5–8 “The journey of a thousand miles…”

9–14 Conducting a winning literature search

15–20 Heterotopic ossification in cervical disc arthroplasty: Is it clinically relevant?

21–28 Treating thoracic-disc herniations: Do we always have to go anteriorly?

47–50 Kyphoplasty: Traditional imaging compared with computer-guided intervention—

Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

57–66 Unilateral facet dislocations: Is surgery really the preferred option?

67–74 Addressing the challenges and controversies of managing os odontoideum: results of a

75–82 Metachronous presentation of metastasis from renal cell carcinoma: evaluation

83 Definition of the different classes of evidence (CoE)

Message from the Editor

“The journey of a thousand miles….”

AO (Arbeitsgemeinschaft für Osteosynthesefragen)—the venerable and unique surgi-

Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

provide high quality evidence of efficacy, effectiveness, safety, and cost-effectiveness

Welcome to the inaugural issue of EBSJ!

Conducting a winning literature search

i. ANAtOMy OF A LitERAtURE SEARCh—thE SkELEtON

A. Constructing an appropriate question

too broad More answerable Better yet

What is the incidence of complications following

Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

B. Using the appropriate database(s)

• B ibliographic databases contain references to published literature, such as journals

Database & website Description What is included

Database & website Description What is included

Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

II. The physiology of literature searching—how it works

A. Getting specific—the basics of “how to”

2. Advanced searching in PubMed—MeSH terms and the MeSH database:

b. MeSH database features

Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

Search term(s) # citations

III. Closing thoughts

Remember, literature searching is a combination of an art and a science. It requires

Heterotopic ossification in cervical disc

The definiton of the different classes

See web appendix at www.aospine.org/ebsj

Evidence-Based Spine-Care Journal Volume 1/Issue 1 — 2010

Study RATIONALE and context METHODS

Patient population and interventions (Fig 1)

Outcomes and analysis

Table 1 McAfee grading of heterotopic

n/N (%) RR 95% CI P-value

15–20 Heterotopic ossification in cervical disc arthroplasty: Is it clinically relevant?

21–28 Treating thoracic-disc herniations: Do we always have to go anteriorly?

47–50 Kyphoplasty: Traditional imaging compared with computer-guided intervention—

67–74 Addressing the challenges and controversies of managing os odontoideum: results of a

75–82 Metachronous presentation of metastasis from renal cell carcinoma: evaluation

• B ibliographic databases contain references to published literature, such as journals

Discussion • I ndications for arthroplasty are not affected by the

7. Vollmer DG, Simmons NE (2000) Transthoracic ap- Editorial staff perspectives

• I n our series, there does not appear to be a clear ad-

References 4. Davies AG, McMaster MJ (1992) The effect of Luque