Advances in Therapeutics and Diagnostics
Botulinum Neurotoxin Type A
Mauricio R. Delgado, MD, FRCPC
Botulinum neurotoxins (BoNTs) are the most potent acute
lethal toxins known but, paradoxically, are used to man-
age a variety of disorders. BoNTs produce muscle weak-
ness by temporarily inhibiting acetylcholine release in the
neuromuscular junction. Clinical experimentation with
botulinum toxin type A (BoNT/A) began in 1970 as an
alternative to surgical treatment of strabismus.1 In 1989,
the US Food and Drug Administration (FDA) approved
the use of BoNT/A for the management of strabismus,
blepharospasm, and hemifacial spasm. In the last decade,
the therapeutic use of these toxins has grown exponen-
tially to include ophthalmologic, neurologic, gastrointes-
tinal, urologic, dermatologic, and orthopaedic disorders.
In 2000, the FDA approved BoNT/A (Botox; Allergan,
Irvine, CA) and botulinum toxin type B (BoNT/B)
(Myobloc; Elan Pharmaceuticals, San Diego, CA) for the
management of cervical dystonia; in 2002, BoNT/A
(Botox) was approved for the management of frown lines
at the glabella.
Structure and Mechanism of Action
Seven distinct serotypes of BoNTs (ie, A, B, C, D, E, F,
G), produced by various strains of the anaerobic Clostrid-
ium botulinum, have been identified. At the molecular lev-
el, these BoNTs differ mostly in degree of activation, mech-
anism of action, and protein complex size. All serotypes
are initially synthesized as inactive 150-kDa single-chain
proteins, which must be cleaved or nicked by proteases
to become neuroactive.2 The active toxin is a dichain mol-
ecule in which an ~50-kDa light chain and an ~100-kDa
heavy chain are linked by a disulfide bond. The dichain
molecule is composed of three ~50-kDa functional do-
mains—binding, translocation, and catalytic—which are
associated with the different steps for toxicity (Fig. 1). The
final effect of the toxin is a reduction in the release of ace-
tylcholine at the nerve terminal. This chemical denerva-
tion is effective in both striated muscle and eccrine glands.
The extent of denervation is determined largely by dose
and volume of injection. However, there is not always
an apparent direct relationship between the amount of
weakness and the amount of benefit. It is possible to block
a spasm without affecting voluntary movements. After
injection of BoNT/A, recovery of nerve activity occurs
first in newly formed sprouts, then reverts to the parent
Vol 11, No 5, September/October 2003
terminal. In a mouse in vivo study,3 sprouts began to ap-
pear 4 days after injection. By 3 months postinjection, the
morphology and uptake behavior of the parent terminal
“were indistinguishable from those visualized before poi-
soning.”3
Pharmacodynamics
Typically, there is a 24- to 72-hour delay between the ad-
ministration of BoNT/A and onset of clinical effect, al-
though some patients (eg, dystonic) may experience ben-
efits immediately. The delay is attributable to the process
of binding, internalization, and changes at the neuromus-
cular junction. Diffusion of BoNT/A depends on the dose
and volume injected. BoNT/A can diffuse up to 4.5 cm
from the site of a 10-U injection (diluted 100 U/mL).4 This
diffusion is important in determining therapeutic response.
Indications for Use
The uses of BoNT/A relevant to orthopaedic surgeons
and others who manage musculoskeletal disorders are
not FDA-approved. For the last decade, BoNT/A has been
used to treat patients with hypertonia (spasticity or dys-
tonia), which is usually combined with other neurologic
deficits, such as weakness, poor motor control, abnormal
reflexes, or sensory proprioceptive deficits. In addition
to dynamic joint deformities affecting posture and mo-
tion, hypertonia may result in secondary problems (eg,
pain, interference with hygiene, poor self-image).
Dr. Delgado is Professor, Department of Neurology, University of Texas
Southwestern Medical School, and Director, Department of Pediatric Neu-
rology, Texas Scottish Rite Hospital for Children, Dallas, TX.
Dr. Delgado or the departments with which he is affiliated has received re-
search or institutional support from Allergan.
Reprint requests: Dr. Delgado, 2222 Welborn Street, Dallas, TX 75219.
Copyright 2003 by the American Academy of Orthopaedic Surgeons.
J Am Acad Orthop Surg 2003;11:291-294
291
Botulinum Neurotoxin Type A
Figure 1 The four-step neurotoxic process of botulinum neurotox-
ins. A, The binding domain mediates interaction between the toxin
and the presynaptic nerve terminal membrane. B, The toxin is inter-
nalized by receptor-mediated endocytosis. C, The translocation do-
main forms a pore in the endosomal membrane. This pore provides
passage for the catalytic domain into the cytosol of the presynaptic
neuron. D, The final step involves proteolysis of a soluble N-ethyl-
maleimide–sensitive factor attachment receptor protein, a crucial com-
ponent in synaptic vesicle membrane fusion.
BoNT/A has most commonly been used in combination
with physical therapy, casting, splints, and orthoses. For
patients with both focal and regional hypertonia, the most
common treatment goals include prevention and/or re-
duction of dynamic muscle contractures to improve tol-
erance to orthoses and facilitate function. BoNT/A also
has been used to ease care and assist in presurgical eval-
uation. Many series have reported subjective and objec-
tive improvements.
Several published randomized prospective trials have
assessed the effect of BoNT/A on upper extremity hy-
pertonia of adult patients, most of whom had spasticity
secondary to stroke. All found a marked reduction in tone
and passive joint range of motion but little or no improve-
ment in hand function.5,6 Similar results have been re-
ported in children with cerebral palsy (CP) who received
BoNT/A for upper extremity hypertonia.7,8 The most no-
table subjective change was improvement in cosmesis.7
Response to BoNT/A for spastic lower limb problems
in adults with multiple sclerosis, stroke, and traumatic
spinal cord and brain injury has been reported. Range
of motion and hygiene scores improved when BoNT/A
292 Journal of the American Academy of Orthopaedic Surgeons
was used to relieve spasticity in hip adductor muscles
to ease nursing care in patients with multiple sclerosis.
Median time for retreatment was approximately 4
months.9 BoNT/A also has been used to manage spastic
equinovarus secondary to stroke.10 A clear subjective im-
provement as well as a significant reduction in spastic-
ity by Ashworth scores (P < 0.001) at 1 month have been
found in stroke patients. Improvement in gait velocity
and step length also have been demonstrated.11 BoNT/A
is equally effective compared with phenol when used to
manage equinovarus secondary to stroke.12
All published randomized controlled trials of BoNT/A
used to manage spasticity in the lower extremity in chil-
dren have involved patients with CP and have evaluated
the effect of BoNT/A only on dynamic equinus. The larg-
est series13 demonstrated a significant (P < 0.05) improve-
ment in passive ankle dorsiflexion and gait pattern as-
sessed by a physician rating scale; this finding also has
been verified using gait analysis.14 Results were compa-
rable when plaster cast ankle stretching was compared
with injection of BoNT/A into the gastrocnemius mus-
cle for ankle equinus. However, the beneficial effect was
longer in the BoNT/A group than in the cast group.15
Spastic equinus in children with CP is the most com-
mon indication for use of BoNT/A, and its use has de-
creased and/or delayed the need for Achilles tendon
lengthening. Children younger than 6 years without a sig-
nificant fixed contracture (ie, ankle-foot dorsiflexion great-
er than −10°) respond best. Use of an ankle-foot orthosis
for at least 6 h/day seems to prolong the interval between
injections.
Analgesic effects were first noted when patients with
cervical dystonia who were treated with BoNT/A report-
ed a decrease in pain. Other types of muscle hypertonia
also may be associated with spasms and pain. In a cross-
over randomized study, BoNT/A effectively reduced
muscle pain caused by spasticity and rigidity of multi-
ple etiologies.16 In a study of 26 patients with chronic neck
pain from whiplash associated disorder, a total dose of
100 U of BoNT/A (Botox) was injected into five trigger
points. Significant (P < 0.01) reduction from baseline was
achieved at 4 weeks in the 14 patients who received
BoNT/A.17 In another study of 31 patients with chronic
low back pain, a total dose of 200 U of BoNT/A (Botox)
applied in five lumbar paravertebral levels on the side
of maximum discomfort resulted in significant pain re-
duction compared with placebo at 3 (P = 0.012) and 8 (P
= 0.009) weeks postinjection.18 Postoperative pain in pa-
tients with CP who undergo tendon release has been at-
tributed to muscle spasms and has usually been man-
aged with benzodiazepines and opiates. Barwood et al19
reported a significant mean reduction of postoperative
pain (P < 0.003), a reduction in mean analgesic require-
ment (P < 0.005), and a reduction in mean length of hos-

pital admission (P < 0.003) with a dose of 4 U/kg of
BoNT/A (Botox) per side.
In a randomized study of 40 patients with tennis el-
bow that compared the effect of BoNT/A injected into
the extensor carpi radialis brevis muscle with surgery (ie,
Hohmann procedure), 65% of patients treated with
BoNT/A (13/20) and 75% of patients treated surgically
(15/20) had good to excellent results at 1 year.20 No sig-
nificant differences were found between the two groups.
The authors postulated that BoNT/A paresis of the ex-
tensor carpi radialis brevis muscle leads to rest of this
muscle for 2 to 4 months, giving the tendon fibers near
the lateral epicondyle time to heal. Delgado et al21 pub-
lished a preliminary report of the efficacy of BoNT/A in
four infants with severe congenital clubfoot who failed
to respond to stretching, taping, and bracing.
Drug Interactions and Adverse Events
Coadministration of BoNT/A and aminoglycosides or
other agents that interfere with neuromuscular junction
function (eg, curare-like drugs) may result in potentia-
tion of the weakening effects of the toxin. In general,
BoNT/A is safe and well tolerated. Adverse events usu-
ally occur within the first week after injection and, al-
though generally transient, may last several weeks or
months. Localized pain, tenderness, or bruising may be
associated with the injection. Local weakness of the in-
jected muscle or muscles represents the expected phar-
macologic action of BoNT/A. However, weakness of ad-
jacent muscles (eg, ptosis after facial injection) also may
occur because of spread of toxin. Subclinical effects, such
as increased jitter on single-fiber electromyographic ex-
amination, have been reported in muscles distant from
the injection site.22 Uncommonly, clinically notable remote
effects have been reported, such as generalized weak-
ness.23 Side effects in children are also relatively uncom-
mon (5% to 12%), mild, and transient, including exces-
sive facial weakness, local pain, and low-grade fever.
Urinary incontinence and constipation have been report-
ed in isolated cases.24,25 Although rare, there have been
reports of spontaneous death after treatment with BoNT,
sometimes associated with dysphagia, pneumonia, or oth-
er significant debility.
Dosage and Cost
The two commercially available BoNT/A preparations
are Botox and Dysport (Ipsen, Slough, England) (Table
1). Botox is commercially available worldwide, but Dys-
Vol 11, No 5, September/October 2003
Mauricio R. Delgado, MD, FRCPC
Table 1
Cost Comparison of BoNT/A
Substance Dosage Cost*
Botox 100 U £129 ($206 US)†
Botox 100 U $466.97‡
Dysport 500 U £165 ($264 US)†
* Cost estimates as of September 2003.
† Prices are those in England. Prices vary by country.
‡ Price on www.drugstore.com, September 10, 2003.
port is available only in Europe and Asia. A unit is the
amount of BoNT/A that is lethal to 50% (LD50) of female
Swiss-Webster mice (18 to 20 g) injected intraperitoneal-
ly. Although Botox and Dysport use the same biologic
unit, their effect on patients is not equivalent. The United
Kingdom Botulinum Toxin and Cerebral Palsy Working
Party reported a Botox-to-Dysport ratio of 1:2.5 U to 1:5
U.26 Therefore, it is important to identify the product when
interpreting reports on BoNT/A. A dilution of 100 U in
1 to 2 mL of normal saline is typically used for Botox,
but actual dosage varies by muscle size, patient weight,
and desired effect.
Summary
BoNT/A is a powerful and relatively safe pharmacolog-
ic agent that produces focal, temporary, reversible, and
gradual muscular denervation. It is an important tool that
should be used as part of an integral approach for the
management of muscle hyperactivity. BoNT/A is an ef-
fective treatment for children with CP who have spastic
equinus because it delays and may prevent surgery. In
adults with stroke, head and spinal injury, and multiple
sclerosis, it facilitates the acute rehabilitation process and
prevents or manages contractures. BoNT/A also is effec-
tive in patients with chronic musculoskeletal pain syn-
dromes and in those with postoperative pain associated
with muscle spasm. Its full therapeutic effects are not com-
pletely understood, and its potential applications are still
being discovered.
Acknowledgment
The author wishes to recognize Mario Romero, PhD, for
his assistance in the development of Figure 1.
293
Botulinum Neurotoxin Type A
References
1. Scott AB, Rosenbaum A, Collins CC: Pharmacologic weakening
of extraocular muscles. Invest Ophthalmol 1973;12:924-927.
2. Aoki KR: Preclinical update on BOTOX® (botulinum toxin type
A)-purified neurotoxin complex relative to other botulinum
neurotoxin preparations. Eur J Neurol 1999;6(suppl 4):S3-S10.
3. De Paiva A, Meunier FA, Molgó J, Aoki KR, Dolly JO: Functional
repair of motor endplates after botulinum neurotoxin type A
poisoning: Biphasic switch of synaptic activity between nerve
sprouts and their parent terminals. Proc Natl Acad Sci U S A
1999;96:3200-3205.
4. Borodic GE, Ferrante R, Pearce LB, Smith K: Histologic assess-
ment of dose-related diffusion and muscle fiber response after
therapeutic botulinum A toxin injections. Mov Disord 1994;9:31-
39.
5. Smith SJ, Ellis E, White S, Moore AP: A double-blind placebo-
controlled study of botulinum toxin in upper limb spasticity af-
ter stroke or head injury. Clin Rehabil 2000;14:5-13.
6. Bakheit AMO, Pittock S, Moore AP, et al: A randomized,
double-blind, placebo-controlled study of the efficacy and safe-
ty of botulinum toxin type A in upper limb spasticity in patients
with stroke. Eur J Neurol 2001;8:559-565.
7. Corry IS, Cosgrove AP, Walsh EG, McClean D, Graham HK:
Botulinum toxin A in the hemiplegic upper limb: A double-
blind trial. Dev Med Child Neurol 1997;39:185-193.
8. Fehlings D, Rang M, Glazier J, Steele C: An evaluation of
botulinum-A toxin injections to improve upper extremity func-
tion in children with hemiplegic cerebral palsy. J Pediatr 2000;
137:331-337.
9. Snow BJ, Tsui JK, Bhatt MH, Varelas M, Hashimoto SA, Calne
DB: Treatment of spasticity with botulinum toxin: A double-
blind study. Ann Neurol 1990;28:512-515.
10. Burbaud P, Wiart L, Dubos JL, et al: A randomised, double
blind, placebo controlled trial of botulinum toxin in the treat-
ment of spastic foot in hemiparetic patients. J Neurol Neurosurg
Psychiatry 1996;61:265-269.
11. Reiter F, Danni M, Lagalla G, Ceravolo G, Provinciali L: Low-
dose botulinum toxin with ankle taping for the treatment of
spastic equinovarus foot after stroke. Arch Phys Med Rehabil
1998;79:532-535.
12. Kirazli Y, On AY, Kismali B, Aksit R: Comparison of phenol
block and botulinus toxin type A in the treatment of spastic foot
after stroke: A randomized, double-blind trial. Am J Phys Med
Rehabil 1998;77:510-515.
13. Koman LA, Mooney JF III, Smith BP, Walker F, Leon JM: Bot-
ulinum toxin type A neuromuscular blockade in the treatment
294 Journal of the American Academy of Orthopaedic Surgeons
of lower extremity spasticity in cerebral palsy: a randomized,
double-blind, placebo-controlled trial. BOTOX Study Group.
J Pediatr Orthop 2000;20:108-115.
14. Sutherland DH, Kaufman KR, Wyatt MP, Chambers HG,
Mubarak SJ: Double-blind study of botulinum A toxin injections
into the gastrocnemius muscle in patients with cerebral palsy.
Gait Posture 1999;10:1-9.
15. Corry IS, Cosgrove AP, Duffy CM, McNeill S, Taylor TC, Gra-
ham HK: Botulinum toxin A compared with stretching casts in
the treatment of spastic equinus: A randomised prospective tri-
al. J Pediatr Orthop 1998;18:304-311.
16. Grazko MA, Polo KB, Jabbari B: Botulinum toxin A for spastic-
ity, muscle spasms, and rigidity. Neurology 1995;45:712-717.
17. Freund BJ, Schwartz M: Treatment of whiplash associated neck
pain [corrected] with botulinum toxin-A: A pilot study. J Rheu-
matol 2000;27:481-484.
18. Foster L, Clapp L, Erickson M, Jabbari B: Botulinum toxin A and
chronic low back pain: A randomized, double-blind study. Neu-
rology 2001;56:1290-1293.
19. Barwood S, Baillieu C, Boyd R, et al: Analgesic effects of bot-
ulinum toxin A: A randomized, placebo-controlled clinical trial.
Dev Med Child Neurol 2000;42:116-121.
20. Keizer SB, Rutten HP, Pilot P, Morré HH, v Os JJ, Verburg AD:
Botulinum toxin injection versus surgical treatment for tennis
elbow: A randomized pilot study. Clin Orthop 2002;401:125-131.
21. Delgado MR, Wilson H, Johnston C, Richards S, Karol L: A pre-
liminary report of the use of botulinum toxin type A in infants
with clubfoot: Four case studies. J Pediatr Orthop 2000;20:533-
538.
22. Olney RK, Aminoff MJ, Gelb DJ, Lowenstein DH: Neuromus-
cular effects distant from the site of botulinum neurotoxin in-
jection. Neurology 1988;38:1780-1783.
23. Bhatia KP, Münchau A, Thompson PD, et al: Generalised mus-
cular weakness after botulinum toxin injections for dystonia: A
report of three cases. J Neurol Neurosurg Psychiatry 1999;67:90-93.
24. Bakheit AMO, Severa S, Cosgrove A, et al: Safety profile and
efficacy of botulinum toxin A (Dysport) in children with muscle
spasticity. Dev Med Child Neurol 2001;43:234-238.
25. Delgado MR: The use of botulinum toxin type A in children
with cerebral palsy: A retrospective study. Eur J Neurol 1999;
6(suppl 4):S11-S18.
26. Carr LJ, Cosgrove AP, Gringras P, Neville BGR: Position paper
on the use of botulinum toxin in cerebral palsy. UK Botulinum
Toxin and Cerebral Palsy Working Party. Arch Dis Child 1998;
79:271-273.
 Perspectives on Modern Orthopaedics
New-Generation Implant
Arthroplasties of the Finger Joints
Peter M. Murray, MD

Abstract
Prosthetic replacement in the hand must address such unique challenges as pres- In 1979, Linscheid and Dobyns9 de-
ervation of the collateral ligaments, tendon balancing, and stability. Some recently veloped a prototype of a PIP joint pros-
developed implant arthroplasties of the metacarpophalangeal and proximal inter- thesis, which they called surface re-
phalangeal joints have anatomically designed articular components; others have non- placement arthroplasty, that was
cemented, press-fit, carefully contoured intramedullary stems. The rationale behind intended to preserve the collateral lig-
developing the unlinked or semiconstrained prosthesis with anatomic geometry is aments and thus unload the compo-
that it would create balanced forces across the joint. Low-profile, anatomically de- nent stems. Other MCP and PIP joint
signed implants limit the amount of bone removed and preserve the integrity of the designs were subsequently developed,
collateral ligaments. A metacarpophalangeal joint implant with an elliptical meta- including the Keesler, the Hagert, and
carpal head and a nonfixed center of rotation can enhance stability in flexion through the Sibly-Unsworth.5,6 Recent design
greater articular contact. A proximal interphalangeal joint implant that preserves modifications and longer follow-up
the collateral ligaments also can achieve improved stability. Component loosening of these early prototypes has gener-
is not an early complication with these recent designs, and arc of motion is satisfactory. ated continued interest in anatomic,
J Am Acad Orthop Surg 2003;11:295-301 minimally constrained PIP and MCP
joint designs. Other new European
designs, such as the Saffar (Dimso
SA, Mernande, France), the Digitale
The primary goals of finger joint ar- hinged prosthesis initially indicated (Procerati, Paris, France), the WEKO
throplasty are to alleviate pain, re- for the severely traumatized PIP joint.1 Fingergrundgelenk (Implant-Service,
store stability, and preserve or en- Two years later, Flatt4 reported on the Hamburg, Germany), and the DJOA3
hance motion. Early digital implants, use of a more rotationally stable mod- (Landos, Malvern, PA), were devel-
such as the Vitallium cap for arthro- ification of the Brannon prosthesis for oped to improve intramedullary fix-
plasty of the metacarpophalangeal the rheumatoid MCP joint.5 These first- ation rather than anatomic configu-
(MCP) and proximal interphalangeal generation hinged designs failed be- ration of the articular surfaces.7,10,11
(PIP) joints,1 were developed with cause of a nonanatomic center of ro-
concepts similar to those used in suc- tation, a high coefficient of friction at
cessful implant arthroplasty of the the hinge mechanism, metallic implant
lower extremity. However, finger to- debris, and, ultimately, breakage.6,7 The Dr. Murray is Associate Professor, Department
tal joint arthroplasty has been slow second generation of hinged prosthe- of Orthopedic Surgery, Division of Hand and Mi-
to develop, primarily because of ear- ses had a ball-and-socket design, with crosurgery, The Mayo Clinic, Jacksonville, FL.
ly design failures. The Swanson the intent of allowing adduction and
Neither Dr. Murray nor the department with
hinged Silastic spacer is the most abduction in addition to flexion and which he is affiliated has received anything of val-
commonly used implant for PIP and extension.6 These metal-on-plastic ue from or owns stock in a commercial company
MCP joint reconstruction, particular- MCP joint designs included the or institution related directly or indirectly to the
ly for patients with rheumatoid ar- Griffiths-Nicolle, the Schetrumpf, the subject of this article.
thritis, in whom 90% 10-year survi- Steffee, the Walker, and the Schultz.
Reprint requests: Dr. Murray, 4500 San Pablo
vorship has been reported.2,3 These implants were fraught with Road, Jacksonville, FL 32224.
In 1959, Brannon and Klein1 pub- complications, including proximal
lished the results of the first series of phalangeal component failure, hyper- Copyright 2003 by the American Academy of
Orthopaedic Surgeons.
a digital total joint replacement. They trophic bone formation, poor motion,
reported encouraging results with a and instability.7,8

Vol 11, No 5, September/October 2003 295

New-Generation Implant Arthroplasties of the Finger Joints

PIP Joint Implant
Arthroplasty
The principal shortcoming of previ-
ous metallic, metalloplastic, and single-
component polymeric plastic-hinged
designs was the amount of bone re-
section required for implantation. The
extent of resection frequently violated
the origin and insertion of the collat-
eral ligaments. The two primary sta-
bilizing factors of the PIP joint are the
bicondylar geometry of the articula-
tion and the collateral ligaments.12,13
The extensor mechanism also may be
considered a stabilizer.12,13 In the ab-
sence of the two primary stabilizers,
the stems of the monoaxial-hinged de-
sign of the first-generation PIP joint
arthroplasty bore high loads, which
frequently resulted in loosening, cor-
tical penetration, and subsidence.1,4-6,12,14
Subsequent hinged or fully constrained
linked designs were unable to ame-
liorate these shortcomings.
The natural flexibility of the Swan-
son Silastic spacer offers greater lon-
gevity compared with previous
metallic-hinged designs. The hinge
resists prolonged cyclic loading but
is prone to fracture at the stem-hinge
junction. However, these implants
continue to function after breakage in
rheumatoid patients. The Swanson
Finger Joint Implant (Wright Medical
Technology, Arlington, TN) is the
most commonly used PIP joint arthro-
plasty device, but it is generally not
recommended for the index or long
fingers of active individuals.9,15 The
generous resection of the proximal
phalangeal head required by the
Swanson Silastic spacer sacrifices the
radial and ulnar collateral ligaments
of the PIP joint. Resection of the col-
lateral ligaments leaves the Silastic
implants of the index and long dig-
its vulnerable to pinch stresses. Ex-
ternal pinch forces of 70 N are con-
sidered normal, with resultant forces
on the PIP joint postulated to be as
high as six times the externally ap-
plied force.6 A successful arthroplas-
ty must be able to sustain these trans-
mitted forces.
The rationale behind new-generation
arthroplasty of the PIP joint is that a
minimally constrained, unlinked pros-
thesis with an anatomic center of ro-
tation would balance forces acting across
the joint. In theory, preservation of bone
stock and collateral ligaments lends
enhanced stability to the arthroplasty
beyond that which can be accomplished
with a Silastic spacer alone.Also, greater
durability can be expected compared
with earlier hinged designs. The an-
atomic configuration, in combination
with retention of the collateral liga-
ments and PIP joint capsule, should
reduce axial torque from the bone-
prosthesis interface.12 Ash and
Unsworth16 demonstrated that an an-
atomically designed PIP joint surface
replacement arthroplasty could with-
stand pinch force >65 N. They also
showed that an ultra-high–molecular-
weight (UHMW) polyethylene mate-
rial for both weight-bearing surfaces
could produce wear rates similar to
those of metal-on-polymer.16
The SR PIP Finger Prosthesis
(Avanta, San Diego, CA) has a
stemmed, bicondylar proximal pha-
langeal component milled from cobalt-
chromium (CoCr). The middle pha-
langeal component of this PIP joint
implant is machined from UHMW
polyethylene, which is supported by
a thin titanium backing and stem. The
articular surfaces of the components
are congruent. Both components have
stems designed to fit the internal con-
tours of the medullary canal. The low-
profile design of the PIP joint surface
replacement arthroplasty reduces the
amount of bone removed and preserves
the integrity of the lateral collateral
ligaments (Fig. 1). Four different sizes
have been made of each component.
The PIP joint surface replacement im-
plant is approved for revision arthro-

Figure 1 A, Titanium-backed UHMW polyethylene middle phalangeal (left) and bicondylar CoCr proximal phalangeal (right) components
of the SR PIP Finger Prosthesis. (Reproduced with permission from Avanta, San Diego, CA.) Anteroposterior (B) and lateral (C) postop-
erative radiographs of PIP joint surface replacement arthroplasty for posttraumatic degenerative arthritis of the PIP joint. Notice the titanium-
backed, second-generation middle phalangeal component.

296 Journal of the American Academy of Orthopaedic Surgeons


plasty of the PIP joint, for arthroplasty
in the painful osteoarthritic PIP joint,
and for the posttraumatic arthritic PIP
joint. This prosthesis seems less de-
sirable in settings of pronounced bone
loss or when the collateral ligaments
are missing or incompetent.
Other recent PIP joint arthroplas-
ty designs include the Saffar, the
Digitos (Osteo AG, Selzach, Swit-
zerland), the DJOA3, and the WEKO
Fingergrundgelenk prostheses. Al-
though labeled semiconstrained by
their manufacturers, the DJOA3 and
Saffar prostheses have a prominent
stabilizing midline crest between the
proximal and distal components. No-
tably, the DJOA3 (Fig. 2) does not re-
quire preservation of the collateral lig-
aments and is composed of a stainless
steel proximal component and a
polyethylene distal component. The
Saffar is a similarly designed, nonce-
mented semiconstrained titanium-
polyethylene prosthesis.7 The Digitos
prosthesis (Fig. 3) is a modular, fully
constrained second-generation PIP
joint prosthesis specifically designed
for unstable joints without collateral
Figure 2 The DJOA3 PIP (top) and MCP
(bottom) joint prostheses. (Reproduced with
permission from Linscheid RL: Implant ar-
throplasty of the hand: Retrospective and pro-
spective considerations. J Hand Surg [Am]
2000;25:796-816.)
Vol 11, No 5, September/October 2003
Figure 3 The Digitos PIP joint prosthesis.
(Reprinted with permission from Linscheid
RL: Implant arthroplasty of the hand: Retro-
spective and prospective considerations.
J Hand Surg [Am] 2000;25:796-816.)
ligaments. Similarly, the WEKO Fin-
gergrundgelenk prosthesis is a con-
strained design that fits into in-
tramedullary bone sleeves (Fig. 4).
Technique
Several surgical approaches, in-
cluding the dorsal, lateral, and pal-
mar, have been used during the evo-
lution of PIP joint arthroplasty.12
Unique difficulties can occur with
each approach because important
structures must be sacrificed or in-
cised during the exposure. The cen-
tral slip is vulnerable with the dor-
sal approach. The collateral ligaments
are at risk with the traditional lateral
approach. The volar plate and the
flexor tendon sheath are at risk with
the palmar approach. Linscheid et
al12 reported an increased incidence
of late swan-neck deformities in pa-
tients undergoing PIP joint surface re-
placement arthroplasty when the pal-
mar approach was used. In contrast,
Lin et al17 reported no instances of
swan-neck deformity or flexor tendon
bowstring in 69 silicone arthroplas-
ties using the palmar approach.17 The
approach preferred by Linscheid et
al12 for the PIP joint surface replace-
ment is the modified dorsal approach
described by Chamay,18 which offers
a generous exposure of the PIP joint
through a distally based triangular
flap of the extensor mechanism (Fig.
5). Before entering the joint, thin rem-
nants of the dorsal PIP joint capsule
Peter M. Murray, MD
are incised. The radial and ulnar col-
lateral ligaments are protected using
small Homan retractors. Judicious
placement of these retractors brings
the base of the middle phalanx into
full view.
For any type of PIP joint arthro-
plasty performed through a dorsal ap-
proach, an osteotomy of the base of
the middle phalanx is done through
the subchondral bone, perpendicular
to the long axis of the phalanx. The
collateral ligament insertion should
be protected during the osteotomy, al-
though a small portion of the inser-
tion may need to be undermined.19
Minamikawa et al13 have shown in a
cadaveric model that the PIP joint re-
mains stable even after half of the col-
lateral ligament substance is removed.
After preparation of the middle pha-
lanx base, an osteotomy of the prox-
imal phalangeal head is done using
a microsagittal saw. Asmall bur is used
to shape the resected proximal pha-
langeal head to accept the desired
prosthetic device. The proximal and
middle phalanges are appropriately
broached, and trial components are
inserted. The permanent components
are implanted once sizing for best fit
is completed. Polymethylmethacrylate
in a semifluid state is used for the
Avanta SR PIP Finger Prosthesis, but
many of the other new-generation de-
Figure 4 The WEKO Fingergrundgelenk
prosthesis. (Reprinted with permission from
Linscheid RL: Implant arthroplasty of the
hand: Retrospective and prospective consid-
erations. J Hand Surg [Am] 2000;25:796-816.)
297
New-Generation Implant Arthroplasties of the Finger Joints
Figure 5 Chamay approach to the PIP joint, with distally based flap of extensor mechanism
raised to expose the joint. (Adapted with permission from Avanta, San Diego, CA.)
signs are press-fit. Rehabilitation is ini-
tiated by postoperative day 5 in most
cases. A dynamic extension splint is
applied for 4 weeks, permitting ac-
tive flexion and dynamic extension.
Results
The Swanson silicone implant is
the most studied prosthesis for recon-
struction of the rheumatoid PIP joint.
Ashworth et al2 reported on PIP joint
silicone implants at an average
follow-up of 5.8 years. Pain was not
present in 67% of joints, and prosthe-
sis survivorship was 81% at 9 years.
The mean postoperative arc of motion
was 29°, compared with a preoper-
ative mean of 38°. Complications in
this series were negligible. Lin et al17
reported on 69 silicone PIP joint spac-
ers (48 with primary or posttraumat-
ic osteoarthritis) at a mean follow-up
of 3.4 years. Mean postoperative
range of motion was 46° compared
with 44° preoperatively. There were
12 joints with complications.
In 1997, Linscheid et al12 published
initial results for the SR PIP Finger
Prosthesis. Sixty-six joint surface re-
placement arthroplasties were insert-
ed, mostly in patients with osteoar-
298
thritis. There were 32 good results, 19
fair, and 15 poor at a mean follow-up
of 4.5 years. This series combined re-
sults from several generations of the
evolving surface replacement design.
Arthroplasties performed through a
dorsal approach yielded better results
than those done through a lateral or
palmar approach. Complications, in-
cluding instability, ulnar deviation,
swan-neck deformity, flexion contrac-
ture, tenodesis, and joint subluxation,
occurred in 19 of the 66 arthroplas-
ties. No components showed evi-
dence of loosening. Range of motion
at follow-up averaged from −14° ex-
tension to 61° flexion. The postoper-
ative arc of motion was 41°, an im-
provement of 12° over preoperative
motion.
To date, published results are not
available for the Saffar and Digitos
prosthetic devices. Condamine et
al10 reported the results of the DJOA3
implant (Fig. 2), which they consider
a third-generation PIP joint prosthet-
ic device. These results suggest sat-
isfactory function in 110 implanted
prostheses with only 3% loosening.
However, 80% of the patients in this
series had been followed for <1 year.
Journal of the American Academy of Orthopaedic Surgeons
MCP Joint Implant
Arthroplasty
Stability, recurring deformity, loosen-
ing, and tendon balancing are the pri-
mary challenges facing the design of
a replacement for the MCP joint.5,20
A common problem in MCP total
joint designs has been the appropri-
ate location of the center of rotation
for the metacarpal head compo-
nent.5 Incorrect placement of the cen-
ter of rotation hinders joint flexion
and extension. If the center of rota-
tion of an MCP joint prosthesis is
placed too dorsal, digital extension
becomes difficult but flexion is en-
hanced. Placement of the center of ro-
tation in a palmar direction may lim-
it digital flexion but may enhance
digital extension.5 In the native joint,
the center of rotation of the MCP joint
in relation to the metacarpal head is
not fixed because the sagittal contour
of the head is elliptical. The move-
ments of the normal MCP joint pro-
duce both abduction and adduction,
along with some rotation.21 Finally,
three-dimensional models of the
hand have shown that internally
transmitted compression joint forces
can range to as high as six times the
externally applied pinch force.21 The-
oretically, the design of a prosthetic
joint would be superior if the design
closely approached the normal ana-
tomic configuration. Such a design
would allow the sliding and rotation-
al movements typically observed.
However, shortcomings of an ana-
tomically configured design are the
potential for instability or sublux-
ation, particularly when ligamentous
incompetence is present.
The MCP PyroCarbon Total Joint
Prosthesis (Ascension Orthopedics,
Austin, TX) is an unlinked MCP joint
implant. The pyrolytic carbon coat-
ing is applied to a high-strength
graphic substrate to create an implant
that is highly compatible with living
tissue.22 The components have offset
intramedullary stems, which support
hemispheric articulating surfaces

(Fig. 6). The offset intramedullary
stems presumably help neutralize ul-
narly directed forces. These articulat-
ing surfaces resemble, but do not an-
atomically replicate, the metacarpal
head and the articular base of the
proximal phalanx. The implant is
very effective in implant-bone load
transfer because of an elastic modu-
lus similar to that of cortical bone.22
The pyrolytic carbon material has
been shown to be very stable in a pri-
mate model, producing no wear, wear
debris, or inflammatory reaction. The
low profile of the MCP PyroCarbon
Total Joint Prosthesis is designed to
preserve the collateral ligaments.
Based on the same design concepts
used for the development of the SR
PIP Finger Prosthesis, the SR MCP
Finger Prosthesis (Avanta) is a min-
imally constrained, unlinked design
that attempts to reestablish the ana-
tomic geometry of the metacarpal
head. The metacarpal component is
made of CoCr; the proximal phalanx
component is manufactured of
UHMW polyethylene (Fig. 7). The
metacarpal head component is ellip-
tical in an attempt to approximate the
changing center of rotation in the nat-
ural MCP joint. Furthermore, the
metacarpal head prosthesis has vo-
lar flanges, thereby enhancing surface
contact in flexion. This enhanced con-
Figure 6 The MCP PyroCarbon Total Joint
Prosthesis. (Reproduced with permission
from Ascension Orthopedics, Austin, TX.)
Vol 11, No 5, September/October 2003
Figure 7 UHMW polyethylene proximal
phalangeal (left) and CoCr metacarpal (right)
components of the SR MCP Finger Prosthe-
sis. (Reproduced with permission from Avan-
ta, San Diego, CA.)
tact in flexion increases radioulnar
stability.19 This prosthesis has been
designed to help compensate for the
soft-tissue imbalance often encoun-
tered at the MCP joint in the rheuma-
toid patient. The dorsal lip of the
proximal phalangeal component has
been extended to prevent palmar sub-
luxation of the joint. Additionally, the
metacarpal component has a central
raised portion designed to inhibit ul-
nar drift. The metacarpal head also
is offset radially on its stem to help
decrease ulnarly directed moments.7
Perhaps more important than any
other stabilizing design feature, the
low-profile nature of the prosthesis
retains the origin and insertion of the
collateral ligaments. Therefore, the
MCP joint surface replacement ar-
throplasty ultimately may be appro-
priate for both osteoarthritis and
rheumatoid arthritis. However, cer-
tain conditions encountered in pa-
tients with rheumatoid arthritis, such
as severe bone erosion and collateral
ligament incompetence, may create
limitations for the use of this device.
Several other MCP joint prosthe-
ses recently have been developed. The
Saffar implant is a noncemented, semi-
constrained titanium-polyethylene
MCP joint prosthesis with a central
articulating crest for stability. The Digi-
tale MCP prosthesis has titanium-
coated, anatomically shaped, stainless
steel press-fit stems designed to stim-
ulate bony ingrowth. The Mathys
MCP RM Finger System (Mathys, Bett-
lach, Switzerland) uses a polyacetal-
Peter M. Murray, MD
resin proximal component and a poly-
ester distal component. This prosthesis
has the unique feature of a screw-
expanded intramedullary fixation for
enhanced intramedullary fit21 (Fig. 8).
The DJOA3 MCP joint implant (Fig.
2) studied by Condamine et al10 has
a spherical stainless steel head and a
cylindrical polyethylene proximal pha-
langeal component.
Technique
For a single-digit arthroplasty, the
extensor mechanism of the MCP joint
is exposed under tourniquet control
through a longitudinal incision. If mul-
tiple joints are to be replaced, a trans-
verse incision is preferable. The ex-
tensor mechanism is dissected in such
a way that relocation can be accom-
plished at the time of wound closure.
In most situations, it is possible to pre-
serve and imbricate the sagittal bands
separately from the dorsal MCP joint
capsule. In patients with rheumatoid
arthritis, it is necessary to do this to
correct digital ulnar drift. Some sur-
geons prefer to incise the extensor
mechanism along its radial border to
imbricate the extensor tendon on the
radial sagittal band. This can be com-
bined with an incision along the ul-
nar border of the extensor tendon to
facilitate radial mobilization of the ex-
tensor tendon, especially in the con-
tracted state. Alternatively, the exten-
sor mechanism can be incised along
its ulnar border, and the extensor ten-
don can be centralized by creating a
Figure 8 The MCP RM Finger System. (Re-
printed with permission from Linscheid RL:
Implant arthroplasty of the hand: Retrospec-
tive and prospective considerations. J Hand
Surg [Am] 2000;25:796-816.)
299
New-Generation Implant Arthroplasties of the Finger Joints

sling made either of the radial sag-

ittal band or from the extensor ten-
don itself.
The capsule is then longitudinally
incised to fully expose the MCP joint.
In most designs, a metacarpal sizing
template is used to determine the
amount of bone to be resected so that
the collateral ligaments are spared.
Next, the base of the proximal pha-
lanx is prepared by a thin osteotomy
perpendicular to the longitudinal axis
of the phalanx. With this proximal
phalanx osteotomy, only the articu-
lar surface and subchondral bone are
removed (Fig. 9). Awls are used to en-
ter the intramedullary canals of the
metacarpal and the proximal pha-
lanx; the respective intramedullary
canals are sequentially broached un-
til the appropriate fit is obtained. Tri-
al components are inserted and re-
duced, and the joint is tested for
stability and range of motion. De- Figure 9 Thin, transverse subchondral osteotomy of the proximal phalanx in preparation
pending on the prosthesis chosen, the for MCP joint arthroplasty. (Adapted with permission from Avanta, San Diego, CA.)
metacarpal and phalangeal compo-
nents are inserted using polymethyl-
methacrylate or are press-fit. For pa- at 10 years was 90%. Blair et al23 re- series of 151 MCP PyroCarbon Total
tients with ulnar drift, the extensor ported the results of 115 Swanson Si- Joint Prostheses (Ascension Orthope-
mechanism is then centralized using lastic implants at a mean follow-up dics) implanted over an 8-year peri-
an imbrication technique. Postop- of 54 months. Mean MCP joint mo- od was followed up at a mean of 11.7
erative rehabilitation involves a dy- tion was 43° (13° extension to 56° flex- years.22 Most patients had rheumatoid
namic extension outrigger splint per- ion), and ulnar drift recurred in 43% arthritis. The arc of MCP joint motion
mitting active flexion and passive of fingers (49/115). Furthermore, arc improved a mean of 13°. The 10-year
extension for approximately 4 weeks. of motion is known to be in a more survivorship was 81.4%. At long-term
This is often followed by a nighttime extended position after Silastic MCP follow-up, those joints with ulnar drift
resting hand splint for an additional joint spacer placement.23,24 had developed recurrent ulnar drift
6 weeks. The MCP joint surface replacement to the degree identified preoperatively.
arthroplasty has been available in Eu- Complications led to 18 implant re-
Results rope for 8 years and is currently un- visions (12%).22
Clinical experience with the Swan- der clinical trial in the United States.
son Silastic MCP joint spacer is greater No series has been published report-
than with any new-generation MCP ing results. Although theoretically Summary
joint arthroplasty device. The results there are advantages to the use of the
of using a new MCP joint prosthesis MCP joint surface replacement ar- The primary challenges to anatomi-
thus must be compared with the gold throplasty, currently it cannot be con- cally shaped arthroplasties in the fin-
standard, the Silastic MCP joint spacer. sidered a replacement for the Swan- gers are joint stability, rebalancing of
Hansraj et al3 reported the results of son Silastic MCP joint spacer. tendons, and prevention of prosthet-
170 Swanson Silastic MCP joint spac- Primate studies have shown no ev- ic loosening. Surface replacement de-
ers at a mean follow-up of 5.2 years. idence of debris or inflammatory re- signs limit bone resection and preserve
No pain was reported in 54% of these action after implantation of the pyro- the integrity of collateral ligaments.
joints. Mean postoperative arc of mo- lytic carbon MCP joint arthroplasty.25 Preservation of bone stock and col-
tion was 27°, compared with 38° pre- Good bone incorporation of the pros- lateral ligaments maintains stability
operatively. Prosthesis survivorship thesis also was observed. Asubsequent while reducing axial torque at the

300 Journal of the American Academy of Orthopaedic Surgeons


bone-cement interface. This is in con-
trast with earlier implants, which were
highly constrained, did not offer suf-
ficient degrees of freedom, and failed
to duplicate the normal center of mo-
tion. When marked bone loss is
present or collateral ligaments have
been rendered incompetent, more con-
strained designs may be more appro-
priate. The best results with the long-
est follow-up of any hand total joint
arthroplasty have been reported with
References
1. Brannon EW, Klein G: Experiences
with a finger-joint prosthesis. J Bone
Joint Surg Am 1959;41:87-102.
2. Ashworth CR, Hansraj KK, Todd AO, et
al: Swanson proximal interphalangeal
joint arthroplasty in patients with rheu-
matoid arthritis. Clin Orthop 1997;342:
34-37.
3. Hansraj KK, Ashworth CR, Ebramza-
deh E, et al: Swanson metacarpopha-
langeal joint arthroplasty in patients
with rheumatoid arthritis. Clin Orthop
1997;342:11-15.
4. Flatt AE: Restoration of rheumatoid
finger-joint function: Interim report on
trial of prosthetic replacement. J Bone
Joint Surg Am 1961;43:753-774.
5. Beevers DJ, Seedhom BB: Metacar-
pophalangeal joint prostheses: A re-
view of the clinical results of past and
current designs. J Hand Surg [Br] 1995;
20:125-136.
6. Beevers DJ, Seedhom BB: Metacar-
pophalangeal joint prostheses: A re-
view of past and current designs. Proc
Inst Mech Eng [H] 1993;207:195-206.
7. Linscheid RL: Implant arthroplasty of
the hand: Retrospective and prospec-
tive considerations. J Hand Surg [Am]
2000;25:796-816.
8. Adams BD, Blair WF, Shurr DG:
Schultz metacarpophalangeal arthro-
plasty: A long-term follow-up study.
J Hand Surg [Am] 1990;15:641-645.
9. Linscheid RL, Dobyns JH: Total joint ar-
throplasty: The hand. Mayo Clin Proc
1979;54:516-526.
10. Condamine J, Marcucci L, Bisson P, Leb-
Vol 11, No 5, September/October 2003
use of the pyrolytic carbon MCP im-
plant, which has successfully complet-
ed formal FDA review and has been
released for general use.
Initial reports of the PIP and MCP
joint surface replacement implants
are encouraging, particularly because
the component loosening typical of
earlier designs has not been a prob-
lem to date. However, recurrent joint
deformity and limited motion remain
challenges for the surface replace-
rieton L: DJOA arthroplasty: Ten years
of experience, in Schuind F, Cooney WP,
An K-N, Garcia-Elias M (eds): Advances
in Biomechanics of the Hand and Wrist. New
York, NY: Plenum Press, 1996, pp 76-83.
11. Saffar P: La fixation prothetique: Gener-
alities, in Table Ronde sur les prosthe-
sies interphalangiennes proximales: Con-
grès de la Société Française de Chirurgie
de la Main. La Main 1997;2:107-109.
12. Linscheid RL, Murray PM, Vidal MA,
Beckenbaugh RD: Development of a
surface replacement arthroplasty for
proximal interphalangeal joints. J Hand
Surg [Am] 1997;22:286-298.
13. MinamikawaY,HoriiE,AmadioPC,Cooney
WP, Linscheid RL, An KN: Stability and
constraint of the proximal interphalangeal
joint. J Hand Surg [Am] 1993;18:198-204.
14. Flatt AE, Ellison MR: Restoration of rheu-
matoid finger joint function: III. A
follow-up note after fourteen years of ex-
perience with a metallic-hinge prosthe-
sis. J Bone Joint Surg Am 1972;54:1317-1322.
15. Amadio PC: Arthroplasty of the prox-
imal interphalangeal joint, in Morrey
BF (ed): Joint Replacement Arthroplasty.
New York, NY: Churchill-Livingstone,
1991, pp 147-157.
16. Ash HE, Unsworth A: Design of a sur-
face replacement prosthesis for the
proximal interphalangeal joint. Proc
Inst Mech Eng [H] 2000;214:151-163.
17. Lin HH, Wyrick JD, Stern PJ: Proximal
interphalangeal joint silicone replace-
ment arthroplasty: Clinical results us-
ing an anterior approach. J Hand Surg
[Am] 1995;20:123-132.
Peter M. Murray, MD
ment prostheses as well as for other
new-generation digital joint implants.
The Swanson Silastic spacer has been
a viable alternative for the patient
with rheumatoid arthritis and has
achieved consistent patient satis-
faction. Nevertheless, the concept of
surface replacement arthroplasty for
finger joints may provide the oppor-
tunity both to extend indications and
to provide more durable functional
results.
18. Chamay A: A distally based dorsal and
triangular tendinous flap for direct ac-
cess to the proximal interphalangeal
joint. Ann Chir Main 1988;7:179-183.
19. Berger RA, Beckenbaugh RD, Lin-
scheid RL: Arthroplasty in the hand
and wrist, in Green DP, Hotchkiss RN,
Pederson WC (eds): Green’s Operative
Hand Surgery, ed 4. New York, NY:
Churchill Livingstone, 1999, vol 1, pp
147-191.
20. Tamai K, Ryu J, An KN, Linscheid
RL, Cooney WP, Chao EY: Three-
dimensional geometric analysis of the
metacarpophalangeal joint. J Hand Surg
[Am] 1988;13:521-529.
21. Beevers DJ, Seedhom BB: Design of a non-
constrained, non-cemented, modular,
metacarpophalangeal prosthesis. Proc Inst
Mech Eng [H] 1995;209:185-195.
22. Cook SD, Beckenbaugh RD, Redondo J,
Popich LS, Klawitter JJ, Linscheid RL:
Long-term follow-up of pyrolytic car-
bon metacarpophalangeal implants.
J Bone Joint Surg Am 1999;81:635-648.
23. Blair WF, Shurr DG, Buckwalter JA:
Metacarpophalangeal joint implant ar-
throplasty with a Silastic spacer. J Bone
Joint Surg Am 1984;66:365-370.
24. Madden JW, De Vore G, Arem AJ: A ra-
tional postoperative management pro-
gram for metacarpophalangeal joint
implant arthroplasty. J Hand Surg [Am]
1977;2:358-366.
25. Cook S, Beckenbaugh R, Weinstein A,
Klawitter J: Pyrolite carbon implants in
the metacarpophalangeal joint of ba-
boons. Orthopedics 1983;6:952-961.
301
 Lower Extremity Angular Malunion:
Evaluation and Surgical Correction
Robert A. Probe, MD

Abstract
The lower extremity has a mechanical axis with joint orientation that allows joint ty in radiographic evaluation, this is
longevity and efficiency in bipedal gait. When normal alignment is lost because of most commonly approximated by us-
trauma or other conditions, deviations from this anatomic norm may be deleterious ing the center of the femoral head and
to long-term joint function. In fractures that have healed with angular malunion, center of the ankle as the outermost
all facets of the deformity must be carefully considered, including alteration in length, points of a line defining the mechan-
rotation, alignment, and translation. Once all elements are fully defined, the effects ical axis. Using these two points as
of the malunion on mechanical axis and joint orientation can be understood. Tech- references, the average mechanical
niques for surgical correction include wedge, dome, and oblique osteotomies and dis- axis crosses the knee 10 mm medial
traction osteogenesis. Each method possesses characteristics appropriate for certain to its frontal plane center (Fig. 1, A).
clinical situations. Judicious patient selection and thoughtful preoperative planning Radiographically, this is approximat-
may allow restoration of normal mechanics. ed by the position of the medial tib-
J Am Acad Orthop Surg 2003;11:302-311 ial spine. In the sagittal plane, the me-
chanical axis from the center of the
femoral head to the center of the an-
kle lies just anterior to the center of
After diaphyseal fracture healing, the zontal joint line orientation. It is crit- rotation of the knee joint (Fig. 1, B).
morphology of an involved bone is ical for the treating physician to fully Functionally, this anterior position of
rarely left unaffected, and some alter- assess and characterize these angu- the mechanical axis is desirable be-
ation in length, rotation, angulation, lar malunions. The physician also cause it allows for passive locking of
and translation is expected. With mod- should understand the implications the knee in full extension.
ern fracture care, such deviations from for the joint, the indications for an os- The mechanical axis of the tibia di-
the original shape are generally small teotomy, and preoperative osteotomy rectly coincides with the anatomic
in magnitude and well tolerated by planning. axis; however, because of the medial
patients. However, on rare occasions,
the change in bone morphology is suf-
ficient to cause concern. Functional im- Normal Biomechanics of Dr. Probe is Chairman, Department of Ortho-
pairment, cosmetic deformity, and the the Lower Extremity paedics, Scott & White Memorial Hospital, Scott,
long-term effect of malalignment on Sherwood and Brindley Foundation, and Associ-
joint integrity and stability are the most To understand the pathomechanics of ate Professor, The Texas A&M University Sys-
important problems. These concerns malunion, the surgeon first must have tem Health Science Center, College of Medicine,
are particularly germane in the lower a thorough knowledge of normal Temple, TX.
extremity, in which the altered distri- lower extremity mechanics. Although
Neither Dr. Probe nor the department with which
bution of weight-bearing stresses leads differing methodologies of measure- he is affiliated has received anything of value from
to abnormal force concentrations across ment and ethnic variation cause some or owns stock in a commercial company or insti-
joints.1 Furthermore, tilting of the knee discrepancies in reported values, a tution related directly or indirectly to the subject
and ankle joint surfaces can lead to few generalizations are appropri- of this article.
detrimental shear stress within artic- ate.4-7
Reprint requests: Dr. Probe, 2401 S 31st Street,
ular cartilage, as well as to changes The mechanical axis of the lower Temple, TX 76508.
in joint contact area.2,3 When these con- extremity passes from the femoral
ditions require management, an os- head through the calcaneal tuberos- Copyright 2003 by the American Academy of
teotomy must be designed to restore ity. Because of the variable position- Orthopaedic Surgeons.
normal alignment, length, and hori- ing of the tuberosity and the difficul-

302 Journal of the American Academy of Orthopaedic Surgeons


Figure 1 A, The frontal plane mechanical
axis (dashed line) of the lower extremity ex-
tends from the center of the femoral head,
across the medial tibial spine, to the center
of the ankle. B, The sagittal plane mechani-
cal axis (dashed line) extends from the cen-
ter of the femoral head, anterior to the cen-
ter of rotation of the knee, to the center of the
ankle.
position of the femoral head relative
to the shaft, there is a difference be-
tween the mechanical and anatomic
axes of the femur. The femoral me-
chanical axis is the line from the cen-
ter of the femoral head to the center
of the knee; the anatomic axis is the
line from the piriformis fossa to the
center of the knee. In an individual
of average size, the anatomic axis is
in 6° of valgus compared with the me-
chanical axis. This angle may be in-
creased in shorter femurs and de-
creased in longer femurs, making
comparison with the contralateral
side beneficial.
Vol 11, No 5, September/October 2003
Robert A. Probe, MD
Joint Orientation
The frontal plane orientation of the
joints also should be defined. The
neck-shaft angle commonly has
been used for the hip and proximal
femur; however, this value is depen-
dent on landmarks that change ra-
diographically with different de-
grees of hip rotation. An alternative
method, described by Chao et al,4 is
the proximal femoral orientation an-
gle. This angle is formed by the di-
vergence of a line from the tip of the
greater trochanter to the center of
the femoral head and the femoral
mechanical axis. Although individ-
ual variation may be present, 90°
(parallel to the floor) is a reasonable
estimate in the absence of compara-
tive contralateral films (Fig. 2). In
stance, the knee joint line is oriented
in 3° of valgus relative to the me-
chanical axis. As a result, the distal
femoral articular surface is in slight
valgus relative to the femoral me-
chanical axis, and the proximal tibial
articular surface is in slight varus
relative to its mechanical axis. This
knee valgus is valuable because, Figure 2 Frontal plane orientation of the
lower extremity joints. The center of the fem-
during gait, the limb assumes a 3° oral head to the tip of the trochanter line
varus position as the foot is planted should be parallel to the floor. The knee is in
3° of valgus, and the ankle typically is ori-
beneath the body’s center of gravity.
ented parallel to the floor.
The medial inclination of the limb
makes the knee axis parallel to the
floor during weight bearing. The three-dimensionally. Leg-length dis-
orientation of the ankle joint is usu- crepancy can be estimated with cal-
ally perpendicular to the mechanical ibrated blocks leveling the anterior
axis.4 Individual patients may dem- superior iliac spines to palpation. The
onstrate deviation from these popu- relative contribution of the tibia to the
lation averages and, when available, length discrepancy can be estimated
the joint inclination of a normal op- with the patient prone and the knees
posite side provides valuable com- flexed 90°. In this position, the dis-
parative information. crepancy in sole height usually can
be attributed to the tibia, with the re-
mainder of the discrepancy account-
Patient Evaluation ed for by the femur. Increased preci-
sion can be obtained with leg-length
Deformity Assessment radiographs on a ruler or with com-
Accurate malunion surgery begins puted tomography.8 Overall limb ro-
with precise definition of the defor- tational differences can be estimated
mity. Implications of a corrective os- by comparing maximal internal and
teotomy cannot be known unless the external rotation of the lower limbs.
malunion is precisely characterized The tibial component can be ascer-
303
Lower Extremity Angular Malunion: Evaluation and Surgical Correction
tained by having the patient sit with
the knees flexed 90° and the ankles
at neutral. In this position, the differ-
ence in the projection of the foot will
describe the rotational difference
within the tibias. Obtaining a comput-
ed tomography scan through the fem-
oral neck, supracondylar femur, and
distal tibia and comparing the rota-
tional position of the two extremities
gives a very accurate measurement of
rotational deformity.9
Angular deformity is usually es-
timated by superimposing intersect-
ing lines centered in the medullary ca-
nal of the proximal and distal
fragments on both anteroposterior
(AP) and lateral radiographs. How-
ever, this technique is subject to er-
ror if a short metaphyseal segment is
present because it can be difficult to
accurately determine the axis. Phys-
iologic bowing of the femur or the tib-
ia also can make drawing accurate
lines along the medullary canal dif-
ficult. Milner10 reviewed a series of
malunited tibial fractures and found
an error range of 11.7° (from −6.2° to
5.5°) in the coronal plane when rely-
ing on the medullary canal as an axis
reference. Increased accuracy may be
achieved by using a reversed radio-
graph of the contralateral side as a
template. The mechanical axis may be
drawn on the uninjured side, fol-
lowed by superimposition of the side
with the deformity on which the
proximal and distal mechanical axes
have been drawn. Divergence of the
proximal and distal mechanical axes
indicates true angular deformity.
The point of intersection of the
proximal and distal axes has been
called the center of rotation of angu-
lation.11 In cases of pure angulation,
this intersection occurs at the apex of
the deformity. In cases of angulation
with translation, the center of rotation
of angulation is moved away from the
site of maximal angular deformity, at
a distance proportional to the amount
of translation (Fig. 3).
One advantage of this method of
biplanar radiographic assessment is
304
Figure 3 In cases of combined translation
and angulation, the center of rotation of
angulation (dark dot) may be defined such
that rotation centered on this point will
correct both deformities. This point is defined
by the intersection of the mechanical axes of
the proximal and distal segments (dashed
lines).
its simplicity: sagittal and frontal
plane deformity can be estimated
with any set of standard radiographs.
A second advantage is that the effects
of the malunion on both the frontal
and sagittal plane mechanical axes
can be estimated. The principal dis-
advantage is that this method rarely
defines the true magnitude and ori-
entation of the angular deformity.12
If angulation is seen on both AP and
lateral radiographs, the true magni-
tude of angulation will be greater
than that seen on either view, and the
plane of deformity usually is some-
where in between. With the deformi-
ty measured on AP and lateral radio-
graphs, trigonometric calculations
can help define the true plane and po-
sition of the deformity, according to
the following formula:12
Journal of the American Academy of Orthopaedic Surgeons
Orientation angle from frontal
plane = arc tan
tan (lateral)
tan (anteroposterior)
True magnitude = arc tan
√tan2(lateral) + tan2(anteroposterior)
Alternatively, the plane and mag-
nitude of the deformity may be de-
fined on fluoroscopic examination.
All extremities with angular deformi-
ty may be rotated into a plane in
which no angular deformity is seen
on fluoroscopy. Orthogonal to this
plane is the plane of maximal angu-
lar deformity. Quantitative assess-
ment of the deformity can be obtained
from the radiographs in this projec-
tion.
Once the magnitude of the defor-
mity is defined, its effects on joint me-
chanics must be assessed. This is a
function of the magnitude and direc-
tion of the angular deformity as well
as its location within the leg. For ex-
ample, a 20° valgus deformity in the
subtrochanteric region of the femur
will result mainly in leg lengthening,
but a 20° valgus deformity of the su-
pracondylar region will result in
shortening and substantial lateral
translation of the mechanical axis.
Computer-modeled effects on
length, mechanical axis, and joint ori-
entation of various 20° malunions are
listed in Table 1. The variability in me-
chanical consequences of these
malunions underscores the necessity
of considering both the magnitude
and location of the malunion. In gen-
eral, as the deformity approaches the
knee, the mechanical axis is translat-
ed medially for a varus deformity and
laterally for a valgus deformity. An-
gular malunions around the knee also
have the greatest effect on leg length.
As the deformity approaches the hip
or ankle, effects on the mechanical
axis are diminished; however, the ad-
jacent joint becomes increasingly
malaligned.

Table 1
Mechanical Implications of Various 20° Frontal Plane Malunions*
Length
Malunion Change
20° subtrochanteric +11 mm
valgus
20° subtrochanteric −18 mm
varus
20° supracondylar −11 mm
varus
20° proximal tibia −9 mm
varus
20° distal tibia −4 mm
varus
* Malunion consequences derived with computer modeling software (LightwaveMod-
eler; Newtek, San Antonio, TX). Leg length changes reflect the absolute distance from
the top of the femoral head to the ankle and not actual bone segment lengthening.
Mathematical formulas and nomo-
grams have been developed to assist
in the understanding of the mechan-
ical effects of particular malunions.13,14
Puno et al13 described trigonometric
methods of calculating these effects;
however, because many physicians are
not familiar with this methodology,
it is more common for the mechan-
ical axis and joint orientation to be
measured radiographically. These
measurements should be taken on a
full-length radiograph from a 51-in
cassette, covering the hip to the an-
kle. The beam should be centered on
the knee, with the patella pointing di-
rectly forward and the x-ray tube 10
feet away. The line from the femoral
head to the center of the ankle defines
the mechanical axis. A perpendicular
line from the mechanical axis to the
medial tibial spine defines the moment
arm of axis deviation. Malalignment
of the knee may be estimated from this
long leg radiograph; however, dedi-
cated AP views of the hip and ankle
are preferred for measurement of their
respective orientation because of the
parallax error on long leg radiographs.
Limb axis translation also has an
effect on mechanical axis shift as well
as joint orientation. Angulation and
translation have been shown to be in-
dependent of each other, with the di-
Vol 11, No 5, September/October 2003
Mechanical Knee Ankle
Axis Change Orientation Orientation
14 mm 2° valgus 2° valgus
lateral
11 mm 2° varus 2° varus
medial
64 mm 10° varus 10° varus
medial
57 mm 8° valgus 12° varus
medial
12 mm 1° valgus 19° varus
medial
rection and magnitude of the two de-
formities unrelated.12 The composite
effects of these two variables are best
discerned by review of long leg,
weight-bearing radiographs.
Symptom Assessment
Other important considerations in
patient evaluation include the status
of local muscle strength, ligamentous
stability, cartilage integrity, and range
of motion of the joints of the affected
extremity. Mild medial mechanical
axis displacement (varus) may be
poorly tolerated with incompetent
lateral ligaments of the knee.15 Sim-
ilarly, if the medial chondral surface
of the knee has been damaged, me-
dial axis deviation is likely to exac-
erbate arthritic symptoms. Ankle
malalignment may be well tolerated
in the setting of a supple subtalar
joint, but a foot with a stiff subtalar
joint will be intolerant to minor de-
grees of ankle malalignment.3 Final-
ly, a well-developed quadriceps mus-
cle may compensate for posterior
displacement of the sagittal plane me-
chanical axis; however, a patient with
a weak muscle likely would experi-
ence symptomatic buckling of the
knee.
Symptoms attributable to mal-
union may be apparent immediately
Robert A. Probe, MD
after fracture healing or may mani-
fest over a longer period of time. Short-
term consequences are infrequent and
usually arise when a malunion is se-
vere enough to exceed the compen-
satory limits of adjacent joints. For ex-
ample, a procurvatum deformity of
the distal femur, which places the sag-
ittal plane mechanical axis posterior
to the knee and thus prevents lock-
ing of the knee during the stance phase
of gait, could be expected to be symp-
tomatic immediately. Likewise, a
patient with a varus distal tibial
malunion that exceeds the compen-
satory valgus effect that can be ob-
tained through the subtalar joint
would experience disruption of gait
mechanics and noticeable symptoms.
Chronic Effects of
Malunions
Delayed-onset symptoms caused by
altered mechanical forces on the joints
are more common than immediate
symptoms. Although a direct causal
relationship between joint deteriora-
tion and altered mechanical loads re-
sulting from malunion has not been
established, an increasing number of
animal, cadaveric, and clinical stud-
ies support this hypothesis. In a rab-
bit model in which 30° angular
malunions were created in the prox-
imal tibia, Wu et al16 observed histo-
logic changes in both cartilage and
bone on the overloaded condyle over
a 34-week period. The cartilage dem-
onstrated irregularity and loss of the
superficial horizontal layer, as well as
clefts extending into the transition
zone. The subchondral bone showed
increased thickness and decreased
porosity. The location of chondral
changes showed direct correlation
with changes in the subchondral
plate, suggesting that increased sub-
chondral plate stiffness may play a
causative role in the overlying carti-
lage changes.
Simulated malunions of varying
degrees and directions in human ca-
305
Lower Extremity Angular Malunion: Evaluation and Surgical Correction
davers have been used to demon-
strate changes in contact pressures
within the joint.1-3 McKellop et al1
used pressure-sensitive film to dem-
onstrate doubling of contact pressure
across the knee with simulated 20°
malunions in both varus and valgus
directions. Tarr et al2 demonstrated
that simulated malunions in the dis-
tal third of the tibia could alter the
shape and diminish the size of tibio-
talar contact. In subsequent experi-
mental work, Ting et al3 demonstrat-
ed that simulated subtalar stiffness
potentiated these mechanical alter-
ations. The conclusions of these stud-
ies are that simulated malunions ap-
pear to alter contact pressure within
adjacent joints and that these chang-
es are maximized as the deformity is
placed closer to the joint.
The finding that may be extrapo-
lated from these animal and cadav-
eric data is that the presence of an al-
tered mechanical environment within
the joint places the joint at increased
risk for degenerative arthropathy.
Puno et al17 reviewed 27 patients with
28 tibial shaft fractures a mean of 8.2
years after injury. They measured ar-
ticular malalignment rather than just
the degree of malunion. This distinc-
tion is important because it takes into
consideration both the magnitude and
location of the malalignment. Regres-
sion analysis showed that the great-
er the ankle malalignment, the poor-
er the ankle function scores. The knee
did not show similar results; howev-
er, mean malalignment in the knee was
only 1.3° compared with 6.6° in the
ankle. Kyro et al18 compared the func-
tion of 17 patients with tibial malunion
to that of 47 patients without malunion
and found significantly (P < 0.05) more
subjective complaints and function-
al limitations in patients with angu-
lation >5°. In another series of 14 pa-
tients with malaligned tibial or femoral
fractures followed up at a mean of 31.7
years, there was progressive knee de-
formity thought to be directly relat-
ed to a combination of the calculated
increased angular force on the tibial
306
plateau and time from original inju-
ry.19 This seems to document the det-
rimental effects of altered mechanical
axis on long-term joint function. van
der Schoot et al20 reviewed 88 patients
at a mean follow-up of 15 years after
tibial fracture. They found a statisti-
cally significant (P < 0.001) relation-
ship between tibial malalignment and
degenerative changes in the knee and
ankle.
Not all of the literature supports
this thesis of long-term detrimental
effect. Merchant and Dietz21 reviewed
37 patients after tibial shaft fracture
at a mean follow-up of 29 years. In
patients with >5° varus, radiograph-
ic arthrosis was noted in the ankles;
however, there was no correlation with
the degree of malunion and knee or
ankle function scores. Milner et al22
reviewed 164 patients at a minimum
of 30 years after treatment of tibial
shaft fracture. They found increased
subtalar stiffness associated with an-
gular malunion but no statistically sig-
nificant association of malunion with
ankle or knee arthritis. This study is
valuable because it confirms that mi-
nor degrees of malunion are tolerat-
ed at the knee and ankle; however, ex-
panding these conclusions to more
severe deformities is not warranted
because only 4% of patients had healed
with coronal plane malunion ≥10°. Be-
cause of these conflicting reports, as-
cribing long-term functional deficit to
malunion remains controversial, es-
pecially because the effect of the orig-
inal extremity trauma on the joint sur-
face cannot be accurately determined.23
Surgical Indications
Common indications for malunion
correction include ligamentous insta-
bility on the convex side of the de-
formity,15 leg-length discrepancy >2
cm, inability to place the foot in a plan-
tigrade position, and unicondylar ar-
thritis of the knee. However, the symp-
toms caused by these mechanical
alterations frequently can be improved
Journal of the American Academy of Orthopaedic Surgeons
nonsurgically. Load-transferring brac-
es, shoe orthoses, shoe lifts, and an-
algesics all may have potential ben-
efit and should be tried before surgical
intervention. If these prove to be in-
effective in relieving symptoms, os-
teotomy may be considered.
The origin of pain in patients with
angular malunion may be multifac-
torial and is often unclear. Potential
etiologies include overloaded liga-
mentous structures, local muscle and
tendon irritation, and tensile strain of
bone. Theoretically, all of these sourc-
es of pain could be improved by cor-
rective osteotomy.
Uncertainty about the long-term
outcome of malunion often has made
decision-making problematic, espe-
cially for patients with asymptomat-
ic angular malunion. The mainstay of
treatment in this group is patient ed-
ucation about future risk of degener-
ative arthritis. At minimum, patients
should be made aware that osteoto-
my is a treatment option should
symptoms develop. There are no de-
finitive criteria to determine wheth-
er osteotomy is indicated; however,
in active individuals, commonly used
guidelines are varus malalignment of
the knee or ankle >10°, valgus mala-
lignment of the knee or ankle >15°,
or a 20-mm medial shift in the me-
chanical axis.
A patient considering osteotomy
for cosmetic reasons must have a clear
understanding of the risk and mag-
nitude of the contemplated procedure
and have realistic expectations re-
garding outcome. Joint fibrosis, mus-
cle weakness, and articular changes
all may contribute to posttraumatic
limb dysfunction and are not gener-
ally improved with malunion correc-
tion.
Surgical Planning
Correction of angular deformity re-
quires decisions to be made about the
location and type of osteotomy and
the method of osteotomy stabiliza-
 Robert A. Probe, MD

tion. Selection of an osteotomy site is the direction, location, and magnitude cutaneously or with an intramedul-
a balance between the geometrically of the deformity. Despite the removal lary saw. The amount of lengthening
ideal position and biologic factors. of bone, a net increase in leg length is equal to half the height of the dis-
Ideally, the angular correction should often results from the correction. No- tracted triangular base. The amount
be centered coincident with the cen- mograms have been developed to as- of linear bone lengthening will be ad-
ter of rotation of angulation, although sist in the estimation of overall leg ditive to any length derived from
other considerations, such as the length change from corrective oste- limb straightening. The primary dis-
quality of the soft-tissue envelope, the otomy.14 Some qualitative conclusions advantages of this technique include
healing potential of the osteotomized can be drawn from these nomograms: the potential for introduction of un-
bone, and the ability to provide rigid (1) correction of varus deformity al- wanted length and creation of a tri-
internal fixation, may justify move- ways results in leg lengthening, (2) the angular bone defect. In adults, trian-
ment of the osteotomy site from the amount of lengthening from varus cor- gular bone defect often must be filled
center of rotation of angulation. How- rection is greatest adjacent to the hip with graft, which incurs the morbid-
ever, if the osteotomy site is moved and diminishes as the osteotomy ap- ity associated with graft harvest and
from the center of rotation of angu- proaches the ankle, (3) correction of a risk of osteotomy nonunion.
lation, the intercalary segment re- valgus deformity in the proximal third Neutral wedge osteotomy (Fig. 4)
mains angulated, which will create a of the femur leads to limb shorten- combines closing and opening wedge
secondary translation deformity. Of- ing, and (4) length gains from correc- osteotomies. A closing wedge osteot-
ten these secondary deformities are tion of distal valgus deformities are omy is done on the convex side of the
clinically insignificant; however, their greatest at the knee and diminish to- deformity, with the apex of the resect-
presence should be anticipated and ward the ankle. ed triangle in the middle of the os-
their consequences considered. If the The advantages of opening wedge teotomy site. Opposing the surfaces
preoperative plan suggests notable osteotomy are regained length and of the closing wedge creates an open-
deformity, accommodating transla- the ability to do the osteotomy per- ing wedge on the contralateral side.
tion may be planned for the distal seg-
ment.
Each type of osteotomy—closing
wedge, opening wedge, neutral wedge,
dome, and oblique osteotomy and dis-
traction osteogenesis—has inherent
characteristics that should be consid-
ered in surgical decision-making. Clos-
ing wedge osteotomy, in which the cen-
ter of rotation is on the concave side
of the deformity, is the most common
method of angular correction. The ad-
vantages of this technique include the
ability to apply it directly at the cen-
ter of rotation of angulation, the re-
sultant contact of viable bone, and the
precision the osteotomy affords. There
are several drawbacks to the closing
wedge osteotomy: extensive surgical
exposure is required; ligaments and
tendons that cross the osteotomy are
functionally lengthened; and the bone
segment is shortened with the removal
of the triangular wedge of bone. The
length lost in the osteotomized bone
segment is equal to half the height of
the triangle’s base. More complex are
the changes in leg length that result Figure 4 Neutral wedge osteotomy combines the features of closing (A) and opening (B)
from correction of the angular defor- wedge osteotomies. The resected wedge may be used on the contralateral side as an osteo-
genic graft.
mity. These changes are dependent on

Vol 11, No 5, September/October 2003 307

Lower Extremity Angular Malunion: Evaluation and Surgical Correction
If the osteotomy apex is moved slight-
ly to the concave side of the middle
of the deformity, the resected trian-
gle of bone may be used as graft for
the resultant opening wedge defect.
The rationale for movement of the
apex in a concave direction is to ac-
commodate for the bone lost in the
resected triangle from the passage of
a saw blade. In this combined osteot-
omy, the point of rotation is the apex
of the closing wedge osteotomy, and
bone segment length should remain
unchanged.
Dome osteotomy uses a bony cut
followed by correctional rotation
across this arced surface (Fig. 5). The
arc of the osteotomy can be consid-
ered to be a portion of a circle, with
the center of the circle defining the
point of rotation of the osteotomy. The
orientation of the concavity of the
dome is critical because its direction
defines the point of rotation. Ideally,
Figure 5 A and B, Dome osteotomies are
created as an arc with the point of rotation
(dot) centered at the center of rotation of an-
gulation. Angular correction is correlated to
translation. The larger the radius, the more
translation produced by a given angular cor-
rection.
308
this point of rotation will coincide
with the center of rotation of angu-
lation so that limb translation is not
introduced. Although the bone will
be restored to normal alignment over-
all, residual angular deformity with-
in the segment of bone between the
point of rotation and the osteotomy
will remain. This residual malaligned
segment creates translation at the os-
teotomy site. Creating a dome with
as short a radius as practicable will
minimize the translational effects of
this segment. Dome osteotomy is tra-
ditionally done in the metaphyseal
portion of long bones. This allows the
osteotomy to be made through can-
cellous bone, with its inherent supe-
rior healing capabilities. The dome os-
teotomy is advantageous because
adjustments can be made in angular
correction, no bone resection is re-
quired, and the contacting metaphy-
seal bone usually heals rapidly. The
primary disadvantages of the dome
osteotomy are that it is generally re-
stricted to metaphyseal sections of
bone, angular correction is tied to
translation across the osteotomy site,
and there is no capacity to correct for
rotation.
Combined deformities of angula-
tion and rotation may be managed by
creating an osteotomy oblique to the
long axis of the bone. If a tibia is split
along the coronal plane, rotation of
the anterior and posterior halves will
result in only an angular change of
these two parts. If a horizontal osteot-
omy is made in the middiaphysis,
rotation of proximal and distal seg-
ments results in only rotational
change. Between these two extremes
are osteotomies that, when rotated,
result in both rotational and angular
correction. This principle may be used
to create the single-cut osteotomy
for correction of angulation and ro-
tation.24,25 The orientation of this os-
teotomy has been defined by math-
ematical formulas;24,25 the most
straightforward determination of ori-
entation is defined by the following
formula:24
Journal of the American Academy of Orthopaedic Surgeons
Angle from long axis in no-angu-
lation view = arc tan
axial rotation
angular deformity
Intraoperatively, the no-angulation
view of the bone is found with fluo-
roscopy. The cut is created with the
width of the blade turned parallel to
this plane, deviating away from the
long axis of the bone by the defined
angle (Fig. 6). In malunions in which
the angular deformity predominates,
this cut becomes steep and difficult
to execute. Advantages of this osteot-
omy include a large surface area for
healing and the ability to perform in-
terfragmentary fixation and add
length by sliding along the osteoto-
my surface after rotation.
Angular correction also can be
achieved with distraction osteogen-
esis,26 using hinges incorporated into
an external fixation frame. The hing-
es are placed on the convex side of
the deformity with the axis serving
as the point of rotational correction,
thus creating a trapezoidal opening
wedge, which has the potential of
adding length. An additional benefit
is that, after angular correction, the
surgeon has the ability to resolve any
residual length discrepancy with fur-
ther distraction. Placement of the
hinges proximal or distal to the apex
of angulation may additionally allow
for simultaneous translational correc-
tion (Fig. 7). Recently developed soft-
ware advancements have expanded
the capabilities of distraction osteo-
genesis by allowing for the simulta-
neous correction of length, angula-
tion, translation, and rotation with a
single frame.27 Despite these frame
advances, distraction osteogenesis
continues to require prolonged peri-
ods of external fixation with its atten-
dant complications, including pin-
tract infections, joint contracture, and
delayed regenerate formation. Be-
cause of these potential difficulties,
this technique is generally reserved
for patients with complex multipla-

Figure 6 A, Simultaneous correction of both angulation and translation is possible with
single-cut osteotomies. B, The saw blade is oriented parallel to the plane of maximum an-
gular deformity at an angle of inclination (short dashed line) derived from a mathematical
formula. Long dashed line = long axis of the bone.
nar deformity, length discrepancy, or
compromised soft tissue.
The final component of surgical
planning is to determine the method
of osteotomy stabilization. Cast im-
mobilization has the potential advan-
tage of allowing postoperative adjust-
ments, although the less rigid control
of osteotomy alignment is a draw-
back. External fixation also allows
postoperative adjustment and pro-
vides increased rigidity relative to
casting. However, pin-tract complica-
tions and the prolonged period of use
necessary for diaphyseal osteotomy
detract from its utility. Locked in-
tramedullary devices can be used to
stabilize an osteotomy; if the medul-
lary canal remains, insertion of a rod
may help restore alignment.28 The
disadvantage is that a second surgi-
cal site is required for insertion. Plate
Vol 11, No 5, September/October 2003
fixation has been used most common-
ly in osteotomy stabilization. Because
open exposure is generally required
for osteotomy, plate application may
be done through that incision. The
plate also may serve as an adjunctive
tool for realignment. If placed on the
convex side, the plate may be fixed
to one end of the bone and a compres-
sion device attached to pull the bone
back into alignment.29 Another ad-
vantage of plate fixation over in-
tramedullary rods is the ability to sta-
bilize short metaphyseal segments.
Results
Because of the relative infrequency of
corrective osteotomy for malunion,
there is a dearth of clinical series dem-
onstrating patient outcomes. Graehl
Robert A. Probe, MD
et al30 reported on supramalleolar
dome and wedge osteotomies done
in eight patients with tibial malunion.
Although complications were fre-
quent, the seven patients who main-
tained correction reported symptom-
atic improvement. Sanders et al31
performed oblique single-cut osteot-
omies in 15 patients with tibial defor-
mity. Mean deformity was 14° in the
coronal plane and 13° in the sagittal
plane, with mean shortening of 2.2
cm. In the 12 patients with adequate
follow-up, osteotomy healed in 10 at
a mean of 4.5 months. All 10 were able
to return to preinjury employment
and were pleased with the surgical re-
sult. Average postoperative deformi-
ty correction was to within 1° in the
coronal plane and 2° in the sagittal
plane, with an average lengthening
of 1.3 cm. Two failures were noted:
one wound dehiscence and one frac-
tured plate. Based on their experienc-
es, the authors recommended this
technique for correction of angular
tibial deformity. They were somewhat
disappointed with their results in re-
gaining limb length and recommend-
ed that distraction techniques be con-
sidered if length discrepancy exceeds
2.5 cm. No mention was made of pre-
operative rotational deformity or the
rotational changes necessarily in-
curred with this technique.32
Sangeorzan et al33 reported on four
patients with combined rotational
and angular deformity of the tibia.
With mathematical planning, all four
patients obtained correction of angu-
lation and rotation with a single-cut
osteotomy. One patient had postop-
erative infection, which resolved with
débridement and delayed primary
closure. In their series of 23 pediat-
ric and adult patients treated for
lower extremity deformity, Tetsworth
and Paley23 demonstrated the effec-
tiveness of Ilizarov methodology in
correcting deformity. The average
mechanical axis deviation was re-
duced from 48 to 8.6 mm; the obliq-
uity of the knee joint improved from
16° to 3°. However, complications
309
Lower Extremity Angular Malunion: Evaluation and Surgical Correction
Figure 7 A and B, The Ilizarov method may be useful in malunions with combined an-
gular and length deformity. In such cases, a hinge position that allows for trapezoidal dis-
traction (shaded areas) is determined.
were frequent, with universal occur-
rence of pin-tract infection and a 36%
incidence of major complications. Av-
References
1. McKellop HA, Sigholm G, Redfern FC,
Doyle B, Sarmiento A, Luck JV Sr: The
effect of simulated fracture-angulations
of the tibia on cartilage pressures in the
knee joint. J Bone Joint Surg Am 1991;73:
1382-1391.
2. Tarr RR, Resnick CT, Wagner KS,
Sarmiento A: Changes in tibiotalar joint
contact areas following experimentally
induced tibial angular deformities. Clin
Orthop 1985;199:72-80.
3. Ting AJ, Tarr RR, Sarmiento A, Wagner
K, Resnick C: The role of subtalar mo-
tion and ankle contact pressure changes
from angular deformities of the tibia.
Foot Ankle 1987;7:290-299.
4. Chao EY, Neluheni EV, Hsu RW, Paley
D: Biomechanics of malalignment.
Orthop Clin North Am 1994;25:379-386.
5. Moreland JR, Bassett LW, Hanker GJ:
Radiographic analysis of the axial
alignment of the lower extremity. J Bone
Joint Surg Am 1987;69:745-749.
6. Paley D, Herzenberg JE, Tetsworth K,
McKie J, Bhave A: Deformity planning
310
erage frame time was 158 days, and
marked improvement occurred in the
latter half of the study. This suggests
for frontal and sagittal plane corrective
osteotomies. Orthop Clin North Am
1994;25:425-465.
7. Tang WM, Zhu YH, Chiu KY: Axial
alignment of the lower extremity in
Chinese adults. J Bone Joint Surg Am
2000;82:1603-1608.
8. Carey RP, de Campo JF, Menelaus MB:
Measurement of leg length by comput-
erised tomographic scanography: Brief
report. J Bone Joint Surg Br 1987;69:
846-847.
9. Wissing H, Buddenbrock B: Determin-
ing rotational errors of the femur by ax-
ial computerized tomography in com-
parison with clinical and conventional
radiologic determination [German].
Unfallchirurgie 1993;19:145-157.
10. Milner SA: A more accurate method of
measurement of angulation after frac-
tures of the tibia. J Bone Joint Surg Br
1997;79:972-974.
11. Paley D, Tetsworth K: Mechanical axis
deviation of the lower limbs: Preoper-
ative planning of uniapical angular de-
Journal of the American Academy of Orthopaedic Surgeons
that, as with many of these complex
techniques, experience contributes to
successful results.
Summary
Normal lower extremity alignment is
ideal for a mechanically efficient bi-
pedal gait. When components of the
lower extremity are traumatically al-
tered, function may be affected. A
thorough understanding of the signif-
icance of the altered mechanics and
a preoperative plan that allows com-
plete rectification of this multifacet-
ed, complex problem is necessary be-
fore attempting surgical management
of lower extremity malunion. The
complexity of osteotomy and distrac-
tion osteogenesis, and the compro-
mise in local tissues imparted by pre-
vious trauma, make these procedures
susceptible to complication. It is im-
perative for the surgeon and patient
to consider fully the risks and ben-
efits during the decision-making pro-
cess.
formities of the tibia or femur. Clin
Orthop 1992;280:48-64.
12. Green SA, Gibbs P: The relationship of
angulation to translation in fracture de-
formities. J Bone Joint Surg Am 1994;76:
390-397.
13. Puno RM, Vaughan JJ, von Fraunhofer
JA, Stetten ML, Johnson JR: A method
of determining the angular malalign-
ments of the knee and ankle joints re-
sulting from a tibial malunion. Clin
Orthop 1987;223:213-219.
14. Wade RH, New AM, Tselentakis G,
Kuiper JH, Roberts A, Richardson JB:
Malunion in the lower limb: A nomo-
gram to predict the effects of osteotomy.
J Bone Joint Surg Br 1999;81:312-316.
15. Badhe NP, Forster IW: High tibial os-
teotomy in knee instability: The rationale
of treatment and early results. Knee Surg
Sports Traumatol Arthrosc 2002;10:38-43.
16. Wu DD, Burr DB, Boyd RD, Radin EL:
Bone and cartilage changes following
experimental varus or valgus tibial an-
gulation. J Orthop Res 1990;8:572-585.

17. Puno RM, Vaughan JJ, Stetten ML,
Johnson JR: Long-term effects of tibial
angular malunion on the knee and an-
kle joints. J Orthop Trauma 1991;5:247-254.
18. Kyro A, Tunturi T, Soukka A: Conser-
vative treatment of tibial fractures: Re-
sults in a series of 163 patients. Ann Chir
Gynaecol 1991;80:294-300.
19. Kettelkamp DB, Hillberry BM, Murrish
DE, Heck DA: Degenerative arthritis of
the knee secondary to fracture malunion.
Clin Orthop 1988;234:159-169.
20. van der Schoot DK, Den Outer AJ,
Bode PJ, Obermann WR, van Vugt AB:
Degenerative changes at the knee and
ankle related to malunion of tibial
fractures: 15-year follow-up of 88 pa-
tients. J Bone Joint Surg Br 1996;78:
722-725.
21. Merchant TC, Dietz FR: Long-term
follow-up after fractures of the tibial
and fibular shafts. J Bone Joint Surg Am
1989;71:599-606.
22. Milner SA, Davis TR, Muir KR, Green-
Vol 11, No 5, September/October 2003
wood DC, Doherty M: Long-term out-
come after tibial shaft fracture: Is
malunion important? J Bone Joint Surg
Am 2002;84:971-980.
23. Tetsworth K, Paley D: Malalignment
and degenerative arthropathy. Orthop
Clin North Am 1994;25:367-377.
24. Rab GT: Oblique tibial osteotomy for
Blount’s disease (tibia vara). J Pediatr
Orthop 1988;8:715-720.
25. Sangeorzan BP, Judd RP, Sangeorzan
BJ: Mathematical analysis of single-cut
osteotomy for complex long bone de-
formity. J Biomech 1989;22:1271-1278.
26. Tetsworth KD, Paley D: Accuracy of
correction of complex lower-extremity
deformities by the Ilizarov method.
Clin Orthop 1994;301:102-110.
27. Rozbruch SR, Helfet DL, Blyakher A:
Distraction of hypertrophic nonunion
of tibia with deformity using Ilizarov/
Taylor Spatial Frame: Report of two
cases. Arch Orthop Trauma Surg 2002;
122:295-298.
Robert A. Probe, MD
28. Behrens FF, Sabharwal S: Deformity
correction and reconstructive proce-
dures using percutaneous techniques.
Clin Orthop 2000;375:133-139.
29. Helfet DL, Jupiter JB, Gasser S: Indirect
reduction and tension-band plating of
tibial non-union with deformity. J Bone
Joint Surg Am 1992;74:1286-1297.
30. Graehl PM, Hersh MR, Heckman JD:
Supramalleolar osteotomy for the treat-
ment of symptomatic tibial malunion.
J Orthop Trauma 1987;1:281-292.
31. Sanders R, Anglen JO, Mark JB: Ob-
lique osteotomy for the correction of
tibial malunion. J Bone Joint Surg Am
1995;77:240-246.
32. Marsh JL: Letter: Oblique osteotomy
for correction of tibial malunion. J Bone
Joint Surg Am 1996;78:151-152.
33. Sangeorzan BJ, Sangeorzan BP, Hansen
ST Jr, Judd RP: Mathematically directed
single-cut osteotomy for correction of
tibial malunion. J Orthop Trauma 1989;3:
267-275.
311
 Common Rotational Variations in Children
Todd L. Lincoln, MD, and Patrick W. Suen, MD

Abstract
Most rotational variations in young children, such as in-toeing, out-toeing, and tor- gests a possible pathologic develop-
ticollis, are benign and resolve spontaneously. Understanding the normal variations mental or neurologic disorder. The
in otherwise healthy children is vital to identifying true structural abnormalities physician should also determine
that require intervention. A deliberate assessment of the rotational profile is nec- whether the rotational problem has
essary when evaluating children who in-toe or out-toe. In-toeing is usually attrib- caused a functional impairment such
utable to metatarsus adductus in the infant, internal tibial torsion in the toddler, as tripping, pain, or shoe wear diffi-
and femoral anteversion in children younger than 10 years. Out-toeing patterns culties. Perceived gait disturbances
largely result from external rotation hip contracture, external tibial torsion, and ex- must be interpreted in the context of
ternal femoral torsion. Although congenital muscular torticollis is the most com- the normal immature pattern of walk-
mon explanation for the atypical head posture in children, more serious disorders, ing that is characteristic of young chil-
including osseous malformations, inflammation, and neurogenic disorders, should dren. Relevant birth history should be
be excluded. noted, including gestational age,
J Am Acad Orthop Surg 2003;11:312-320 length of labor, complications, Apgar
scores, birth weight, and number of
days in the hospital. These details
may heighten the suspicion for pos-
Benign rotational variations such as internal tibial torsion axis, an exter- sibility of cerebral palsy. The family
in-toeing, out-toeing, and torticollis nal contracture at the hip, or flexible history should include a careful as-
are seen in many healthy children. Al- foot deformities. The external hip sessment of rotational disorders in
though the physical appearance of contracture initially masks the high other family members and the pres-
these conditions may initially be degree of femoral anteversion also ence of hereditary disorders (eg, vi-
alarming, spontaneous resolution oc- characteristic of normal infants at tamin D–resistant rickets, muco-
curs in most cases. A thorough under- birth. Postnatally, the lower extrem- polysaccharidoses, achondroplasia,
standing of the normal rotational ities continue to rotate externally un- epiphyseal or metaphyseal dysplasia)
variations that may occur in children til adult values are reached (between that may affect the rotational profile.
younger than 10 years is essential to ages 8 and 10 years). During this pe- Evaluation of postural conditions
properly reassure and educate fam- riod of rapid growth, the tibia typi- requires both a static and a dynamic
ilies, as well as to identify more se- cally externally rotates 15° while fem- physical examination. The static ex-
rious underlying structural problems oral anteversion decreases an average
that might exist. of 25°.1 Normal rotational profiles in
childhood therefore are variable and Dr. Lincoln is Assistant Professor, Department
age-dependent. of Orthopaedic Surgery, Stanford University
Medical Center, Lucile Salter Packard Children’s
In-toeing and Out-toeing Hospital, Palo Alto, CA. Dr. Suen is in private
Evaluation practice at Kaiser Permanente, Oakland, CA.
Natural History Normal variability in young chil-
Limb buds appear in the fifth week dren must be differentiated from None of the following authors or the departments
in utero. The great toes develop in a more serious structural problems. The with which they are affiliated has received anything
preaxial position, rotating medially in clinical history should delineate the of value from or owns stock in a commercial com-
pany or institution related directly or indirectly
the seventh week to bring the hallux onset and duration of a structural
to the subject of this article: Dr. Lincoln and Dr.
to midline. Subsequent intrauterine problem and any evidence of progres- Suen.
molding causes external rotation at sion. Whereas the typical natural his-
the hip, internal rotation of the tibia, tory of benign rotational conditions Copyright 2003 by the American Academy of
and variable positioning of the foot. would suggest improvement over Orthopaedic Surgeons.
Thus, many infants are born with an time, a progressive deformity sug-

312 Journal of the American Academy of Orthopaedic Surgeons


amination should begin with an eval-
uation of the overall appearance of
the patient before focusing on the
lower extremities. Short stature or
disproportionate body-to-limb ratio
may suggest skeletal dysplasia. The
child’s rotational profile, as described
by Staheli,2 should be recorded. The
five components in this profile are in-
ternal and external hip rotation,
thigh-foot axis, transmalleolar axis,
heel-bisector angle, and foot progres-
sion angle during gait.
Hip rotation is most easily mea-
sured with the patient in the prone
position. A parent can hold a fearful
or uncooperative younger child face
to face to soothe her or him during
the examination. Infants have an av-
erage of 40° of internal rotation
(range, 10° to 60°) and 70° of exter-
nal rotation (range, 45° to 90°). By age
10 years, internal hip rotation aver-
ages 50° (range, 25° to 65°) and ex-
ternal rotation, 45° (range, 25° to
65°).2 Internal rotation measuring 70°,
80°, or 90° is evidence respectively of
a mild, moderate, or severe increase
in femoral torsion.1 Increased femo-
ral torsion may be evident during
gait, with medially facing patellar
alignment.
The thigh-foot axis also is best ex-
amined with the child in the prone
position and the knee flexed 90° (Fig.
1). This angle consists of the rotation
of the tibia and hindfoot in relation
to the longitudinal axis of the thigh
and indicates the amount of tibial tor-
sion present. In infants, the thigh-foot
angle averages 5° internal (range,
−30° to +20°). Excessive internal tib-
ial torsion spontaneously resolves by
age 3 or 4 years in most children. By
age 8 years, the thigh-foot axis aver-
ages 10° external (range, −5° to +30°)
and usually changes very little after
that.2
Measurement of the transmalle-
olar axis also aids in determining the
amount of tibial torsion. This axis is
the angle formed at the intersection
of an imaginary line from the lateral
to the medial malleolus, and a second
Vol 11, No 5, September/October 2003
Todd L. Lincoln, MD, and Patrick W. Suen, MD
Figure 1 The thigh-foot axis is best evalu-
ated with the child in a prone position. The
angle subtended by the longitudinal axis of
the thigh and the foot defines the degree of
internal or external tibial torsion present.
line from the lateral to the medial
femoral condyles. At gestational age
5 months, the fetus has approximate-
ly 20° of internal tibial torsion. The
tibia then rotates externally, and most
newborns have an average of 4° of in-
ternal tibial torsion. As a child grows,
the tibia continues to rotate external-
ly. Adults have an average of 23° of
external tibial torsion (range, 0° to
40°).1
The foot should be examined for
additional causes of apparent in-
toeing or out-toeing. The heel-bisector
line, the line drawn through the mid-
line axis of the hindfoot and the fore-
foot, is helpful in evaluating forefoot
adduction and abduction.3 In a neu-
tral foot, the heel-bisector line passes
through the second web space.
Assessment of the foot progression
angle during gait is the fifth and fi-
nal component of a child’s rotational
profile. The foot progression angle is
the angle of the foot relative to an
imaginary straight line in the patient’s
path. Patients who in-toe are assigned
a negative angular value; patients
who out-toe are given a positive val-
ue. This value represents the sum to-
tal effect of the child’s structural align-
ment (ie, femoral torsion, tibial
torsion, foot contour) as well as any
dynamic torsion forces resulting from
muscle forces. Some pathologic con-
ditions will have characteristic gait
patterns. For example, a patient with
mild cerebral palsy may demonstrate
mild equinus and in-toeing, whereas
in-toeing with a Trendelenburg gait
suggests hip dysplasia.
Children with rotational profiles
two standard deviations outside the
mean for their age are considered ab-
normal.1 In such children, further di-
agnostic studies (eg, plain radio-
graphs) should be considered,
depending on the specific abnormal-
ity. For example, foot radiographs
may help diagnose skewfoot in a
child with severe in-toeing. Others
have suggested extremity radio-
graphs for children presenting with
short stature (<25th percentile), a
worrisome hip examination, marked
limb asymmetry, or pain.1 To rule out
hip dysplasia, some advocate a pel-
vic radiograph for any patient pre-
senting with a gait abnormality that
is not easily explained by the rotation-
al profile, asymmetric hip motion, or
hip pain.2
In-toeing usually is caused by be-
nign conditions such as metatarsus
adductus, excessive internal tibial tor-
sion, and excessive femoral torsion.
Less frequently, patients have patho-
logic conditions such as clubfoot,
skewfoot, hip disorders, and neuro-
muscular diseases. Metatarsus ad-
ductus, with or without internal tib-
ial torsion, is the most common cause
of in-toeing from birth to 1 year. In
toddlers, internal tibial torsion caus-
es most in-toeing. After age 3 years,
in-toeing usually is caused by in-
creased femoral anteversion. More se-
vere in-toeing suggests a combination
of deformities, such as internal tibial
torsion and excessive femoral ante-
version.4
Out-toeing typically is caused by
external rotation contracture of the
313
Common Rotational Variations in Children

hip, external tibial torsion, or exter-
nal femoral torsion. External rotation
contracture of the hip capsule is a
common finding during infancy,
whereas external tibial or femoral tor-
sion is more commonly seen in older
children and adolescents who out-
toe.2 Severe pes planovalgus also has
been associated with out-toeing.
More serious conditions, such as a
slipped capital femoral epiphysis, hip
dysplasia, or coxa vara, are less com-
mon but should be considered.
Active treatment of childhood ro-
tational disorders is unnecessary in
most cases. Prudent care consists of
reassurance and education about the
natural history of the condition. Brac-
ing and shoe modifications are unnec-
essary and should be actively dis-
couraged for these normal children.
Many published studies have shown
that such interventions have no de-
monstrable effect on the natural his-
tory or on spontaneous resolution.5
One study even indicated an associ-
ation of brace use for benign torsion-
al variations during childhood with
lower self-esteem scores during
adulthood.6
positioning causes the deformity. This
theory is supported by the high rate
of spontaneous resolution of metatar-
sus adductus as well as its associa-
tion with twin pregnancies.9 Sleeping
position also may contribute to the
development of metatarsus adductus.
Many babies sleep in a prone posi-
tion with the hip and knees flexed and
the feet adducted. Other authors have
proposed anatomic differences as the
primary cause. Surgical findings have
indicated that a muscle imbalance
from a tight anterior tibial tendon or
an anomalous insertion of this tendon
could cause metatarsus adductus.
However, others were unable to re-
produce metatarsus adductus in still-
born fetuses by using traction on the
anterior tibial tendon.10 Furthermore,
in patients with cerebral palsy, a spas-
tic anterior tibial tendon leads to
hindfoot varus. Such findings chal-
lenge the muscle imbalance concept.
Another theory is that the medial cu-
neiform is abnormally shaped in pa-
tients with metatarsus adductus.
Morcuende and Ponseti11 found a
trapezoid-shaped medial cuneiform
with a broadened and medially tilt-
ed articular surface at the metatarsal-
medial cuneiform articulation in fe-
tuses with metatarsus adductus.
Metatarsus adductus usually is
seen in the first year of life and oc-
curs more frequently on the left side.
Presenting complaints include cosme-
sis, an in-toeing gait, or excessive shoe
wear. On physical examination, the
foot appears C-shaped, with a con-
cave medial border and a convex lat-
eral border (Fig. 2). Pressure sites dur-
ing shoe wear may include the medial
border of the first metatarsopha-
langeal joint or a prominent lateral
border at the base of the fifth meta-
tarsal. Hyperactivity of the abductor
hallucis muscle also may contribute
an additional dynamic component to
this foot position, particularly in chil-
dren younger than 18 months. The
hindfoot will be neutral or in valgus,
but never in varus. Range of motion
of the ankle and subtalar joint will be
normal.
Metatarsus adductus has been
classified by Smith et al3 as mild,
moderate, or severe, depending on
the heel-bisector angle. Greene12 also
developed a classification scheme

Other Postural Conditions

Metatarsus Adductus
Metatarsus adductus consists of
medial deviation of the forefoot on
the hindfoot with a neutral or slight-
ly valgus heel (Fig. 2). This condition,
described by Henke in 1863, is the
most common pediatric foot problem
referred to orthopaedic surgeons. It
occurs in 1:5,000 live births and in 1:20
siblings of patients with metatarsus
adductus. The rate of metatarsus ad-
ductus is higher in males, twin births,
and preterm babies.7 Earlier studies
suggested a relationship between
metatarsus adductus and hip dyspla-
sia, but recent studies indicate no
such correlation.8
Although the exact cause of meta-
tarsus adductus is unknown, numer-
Figure 2 Typical clinical appearance of a child with metatarsus adductus.
ous theories exist. One is that in utero

314 Journal of the American Academy of Orthopaedic Surgeons


based on the heel-bisector angle and
the visual appearance of the lateral
border of the foot. However, because
flexibility appears to correlate more
closely with treatment and progno-
sis, classification systems based on
flexibility of the deformity may be
preferable.13 A later classification sys-
tem described by Bleck14 designated
a flexible forefoot as one that could
be abducted beyond the midline heel-
bisector angle, a partially flexible fore-
foot as one that could be abducted to
midline, and a rigid forefoot as one
that could not be abducted to mid-
line. The classification system of
Crawford and Gabriel15 also is based
on flexibility of the forefoot.
Routine imaging studies are not
necessary in infants with metatarsus
adductus but may be indicated in
children older than 4 or 5 years with
unresolved deformity and pain. The
usefulness of radiographs before age
4 years is limited by the lack of suf-
ficient ossification in the bones of the
foot. In older children, forefoot ad-
duction, excessive medial deviation
at the tarsal-metatarsal joint, and a
neutral or valgus heel will be evident
on a standing radiograph. Although
classification systems of metatarsus
adductus based on radiographic cri-
teria exist, they have poor intraob-
server and interobserver agreement
and no prognostic significance.16
Most cases of flexible metatarsus
adductus resolve spontaneously and
do not require use of splinting, brac-
es, or special shoes. Rushforth17 did
a prospective study of 83 children
with 130 cases of flexible metatarsus
adductus. At follow-up with no treat-
ment (mean, 7 years), 58% had no re-
sidual deformity, 28% had mild de-
formity, 10% had moderate deformity,
and 4% had severe adductus.17 Pon-
seti and Becker18 studied 335 children
with flexible metatarsus adductus
who received no treatment. All pa-
tients improved in 3 to 4 years. In a
series of 21 patients (31 feet) with
partly flexible or inflexible metatar-
sus adductus treated with serial cast-
Vol 11, No 5, September/October 2003
Todd L. Lincoln, MD, and Patrick W. Suen, MD
ing, 20 patients (95% [29 feet]) had
painless normal feet as adults; 1 pa-
tient (5% [2 feet]) had residual adduc-
tus and pain only after strenuous ac-
tivity.13 Most evidence indicates that
flexible metatarsus adductus com-
monly resolves without treatment
and that even when it does not, it
rarely leads to pain in adulthood.
Patients with rigid metatarsus ad-
ductus deformities should undergo
early casting. Although some authors
claim that below-knee casting is less
effective than long leg casting, no data
support this claim.12 In a study of 37
feet with inflexible moderate metatar-
sus adductus and 48 feet with severe
metatarsus adductus, Katz et al19 dem-
onstrated that below-knee casting can
improve metatarsus adductus defor-
mities. Correction of the foot defor-
mity was achieved by 6 to 8 weeks in
all cases. At 2- to 6-year follow-up,
moderate deformity had recurred in
six feet with initial severe inflexible
deformity; one additional patient had
developed a severe deformity.
Uncommonly, resistant cases of in-
flexible metatarsus adductus may re-
quire surgery because of painful shoe
wear. Surgical options include release
of the abductor hallucis tendon, me-
dial midfoot capsulotomy, tarsometa-
tarsal joint capsulotomy and release
of the intermetatarsal ligaments, or
osteotomy at the metatarsal bases and
cuneiforms. Lengthening of the ab-
ductor hallucis with medial capsulot-
omy of the naviculocuneiform and
cuneiform first metatarsal joints is
technically simple and was shown to
be effective in a recent series of 29 feet
in 18 children.20 Capsulotomy of the
tarsometatarsal joints and release of
intermetatarsal ligaments (the
Heyman-Herndon procedure) has a
41% failure rate and complications
such as skin slough, osteonecrosis of
the cuneiforms, dorsal prominence of
the first metatarsal-cuneiform joint,
and early degenerative arthritis.21 Os-
teotomy at the metatarsal bases is as-
sociated with shortening of the first
metatarsal in 5% to 30% of patients.22
In contrast, an opening wedge osteot-
omy of the medial cuneiform, com-
bined with a closing wedge osteoto-
my of the cuboid or osteotomies at the
base of the second through fourth
metatarsals, has been shown to be
safe and effective.23 Thus, this appears
to be the most effective surgical op-
tion in patients older than 3 years
with persistent rigid metatarsus ad-
ductus deformities.
Metatarsus Primus Varus
Metatarsus primus varus is an iso-
lated adducted first metatarsal. In
contrast with simple metatarsus ad-
ductus, in metatarsus primus varus
the lateral border of the foot has a nor-
mal alignment, and there is often a
deepened vertical skin crease on the
medial border of the foot at the tar-
sometatarsal joint. In general, meta-
tarsus primus varus is a more rigid
deformity than simple metatarsus ad-
ductus, and early casting is recom-
mended. Persistent deformity in
childhood is associated with progres-
sive hallux valgus. Opening medial
cuneiform osteotomy has been de-
scribed for selective use in children
with a severe deformity.22
Dynamic Hallucis Abductus
Dynamic hallux abductus, other-
wise known as the wandering or at-
avistic toe, also can cause in-toeing.
The great toe deviates medially dur-
ing ambulation while the remainder
of the forefoot remains straight. Dy-
namic hallucis abductus usually pre-
sents after a child begins walking and
is thought to be caused by an imbal-
ance of the great toe abductor and ad-
ductor muscles. Dynamic hallux ab-
ductus usually resolves with age and
subsequent fine motor coordination
development.
Skewfoot
Skewfoot, also called congenital
metatarsus varus or serpentine meta-
tarsus adductus, is characterized by
adducted metatarsals combined with
a valgus deformity of the heel and
315
Common Rotational Variations in Children
plantarflexion of the talus (Fig. 3). Lit-
tle is known of the pathogenesis of
this disorder. Improper casting of
metatarsus adductus or clubfoot de-
formities may result in a skewfoot be-
cause of failure to support the hind-
foot while abducting the forefoot in
the cast. However, most cases are
thought to be idiopathic.24
The amount of hindfoot valgus
necessary to classify a foot as a true
skewfoot rather than as the more
common metatarsus adductus is not
strictly defined. As a result, limited
epidemiologic information about this
deformity is available. Determining
hindfoot valgus in infants is difficult
because of their small size; common-
ly, skewfoot is not diagnosed until lat-
er in childhood. Pain or callus forma-
tion under the head of the talus and
the base of the fifth metatarsal may
be reported, and uneven shoe wear
may develop. Standing radiographs
confirm the presence of an adducted
forefoot and a valgus hindfoot.
The natural history of this defor-
mity is unclear. Although some feet
undergo spontaneous correction, oth-
ers clearly continue to have pain, cal-
losities, and problems with shoe wear.
Surgery is indicated for children with
Figure 3 A, Clinical appearance of a skewfoot. B, Anteroposterior radiograph of a skew-
foot showing hindfoot valgus, talar plantarflexion, midfoot abduction, and forefoot adduc-
tion.
316
a persistently symptomatic foot de-
formity. Mosca25 reported successful
outcomes in 9 of 10 children treated
after age 6 years with an opening
wedge osteotomy on the calcaneus
and the medial cuneiform.
Positional Calcaneovalgus
Positional calcaneovalgus is a flex-
ible foot deformity characterized by
dorsiflexion at the ankle and mild sub-
talar joint eversion. It may be the most
common pediatric foot deformity, with
an estimated incidence ranging from
0.1% to 50% in some series.8 Position-
al calcaneovalgus is most common in
girls, first-born children, and children
of young mothers. Intrauterine mal-
positioning is thought to cause this
deformity. Imaging studies are not nec-
essary for diagnosis but may help rule
out the presence of a more serious un-
derlying disorder, such as congenital
vertical talus or posteromedial bow-
ing of the tibia. Treatment of position-
al calcaneovalgus does not alter the
natural history of this deformity.26 All
cases appear to resolve spontaneous-
ly, with or without manipulation and
bandaging. Therefore, no treatment is
recommended for positional calca-
neovalgus.
Journal of the American Academy of Orthopaedic Surgeons
Rotational Deformities of
the Lower Extremity
Tibial Torsion
Internal tibial torsion is the most
common cause of in-toeing from ages
1 to 3 years. In two thirds of affected
children, the increased torsion is bi-
lateral. When unilateral, internal tib-
ial torsion usually affects the left side.
Most cases are thought to be caused
by intrauterine positioning. Accurate
clinical recognition relies on measure-
ment of the thigh-foot and transmal-
leolar axes. Although most children
with increased tibial torsion are nor-
mal, excessive internal tibial torsion
is also associated with tibia vara,
while increased external tibial torsion
is often associated with neuromuscu-
lar conditions such as myelodyspla-
sia and polio.
Parents of children with increased
internal tibial torsion often report that
the child is clumsy and trips frequent-
ly. Treatment with splinting, shoe
modifications, exercises, and braces
has proved to be ineffective.5 Because
the natural history of internal tibial
torsion strongly favors spontaneous
resolution by age 4 years, expectant
observation is recommended instead.
Disability from persistent residual in-
ternal tibial torsion is rare, and it is
not a risk factor for degenerative joint
disease. Some have even suggested
that in-toeing improves sprinting
ability.27
In contrast to internal tibial torsion,
excessive external tibial torsion tends
to increase with age. It is usually dis-
covered in late childhood or adoles-
cence, tends to be unilateral, and
more often affects the right side.2 Dis-
ability from external tibial torsion is
more common and includes patel-
lofemoral pain and patellofemoral in-
stability.2,28 Some have found an as-
sociation between external tibial
torsion and degenerative joint disease
in the knee, but most believe it is not
a risk factor.28
Surgical treatment of tibial torsion
is rarely indicated and should be re-
 Todd L. Lincoln, MD, and Patrick W. Suen, MD

served for children older than 8 years tervention may be indicated in a child disorders. In a series of 288 children
with marked functional or cosmetic older than 8 years with a marked cos- with torticollis, congenital muscular
deformity and a thigh-foot angle great- metic or functional deformity, ante- torticollis was the cause in 82% of cas-
er than three standard deviations be- version >50°, and internal hip rota- es.33 Of the remaining 18%, most had
yond the mean (eg, thigh-foot angle tion >80°. Surgeries to correct femoral Klippel-Feil syndrome or a neurologic
>15°).2 Both proximal and supramal- torsion include proximal and distal disorder. Klippel-Feil syndrome is
leolar tibial derotational osteotomies femoral osteotomies. A proximal fem- characterized by congenitally fused
have been used to manage tibial tor- oral osteotomy may be considered if cervical vertebrae and a short neck.
sion. However, most surgeons prefer the patient has a concomitant varus Osseous malformations that cause
the supramalleolar osteotomy because or valgus deformity. Otherwise, a dis- torticollis include basilar impression;
of its lower complication rate.29 In the tal femoral osteotomy through a lat- atlanto-occipital anomalies; and a uni-
skeletally mature adolescent, derota- eral approach is the preferred treat- lateral absence of C1, familial cervi-
tional osteotomy with intramedullary ment. A small compression plate may cal dysplasia, and atlantoaxial rota-
fixation is also an option. be used to treat skeletally immature tory displacement. Any of a variety
patients and a blade plate for skele- of neurologic disorders may be the eti-
Femoral Torsion tally mature patients.32 ologic agent, including posterior fossa
Femoral torsion is the angular dif- tumors (Fig. 5) and cervical tumors,
ference between the femoral neck axis syringomyelia, Arnold-Chiari mal-
and the transcondylar axis of the Torticollis formations, ocular dysfunction, and
knee. At birth, neonates have an av- paroxysmal torticollis of infancy. A
erage of 40° of femoral anteversion. Torticollis is any deformity in which formal ophthalmologic examination
By age 8 years, average anteversion the head is tilted and abnormally ro- frequently is indicated when the ster-
decreases to the typical adult value tated. The differential diagnosis of tor- nocleidomastoid muscle is not clearly
of 15°. Most cases of femoral torsion ticollis includes typical congenital tight on examination. Acute-onset tor-
are idiopathic, although a familial as- muscular torticollis as well as torti- ticollis in the setting of a pharyngitis
sociation is identified in some pa- collis secondary to osseous malforma- or recent adenoidectomy may indi-
tients. tions, inflammation, and neurogenic cate Grisel’s syndrome. Ballock and
Increased femoral anteversion is
the most common cause of in-toeing
in early childhood, tends to occur in
females, and is symmetric. Children
with excessive femoral anteversion
characteristically sit with their legs in
the W position (Fig. 4) and run with
an eggbeater-type motion (because of
internal rotation of the thighs during
swing phase). In-toeing from exces-
sive femoral anteversion usually in-
creases until age 5 years and then
resolves by age 8. On physical ex-
amination, internal hip rotation is in-
creased and external hip rotation de-
creased. No association between
increased femoral anteversion and
degenerative joint disease has been
proved; however, some association
with knee pain has been suggested.30
Knee pain may be particularly prev-
alent in children with concomitantly
increased femoral anteversion and ex-
ternal tibial torsion (so-called miser-
able malalignment syndrome).31
No treatment is necessary for most Figure 4 Characteristic ability of a 6-year-old child with increased femoral anteversion to
sit in the W position. The child’s patellas are outlined by the dotted circles.
cases of femoral torsion. Surgical in-

Vol 11, No 5, September/October 2003 317

Common Rotational Variations in Children
Figure 5 Axial computed tomography im-
age of a large posterior fossa tumor (astrocy-
toma) (arrow) in a 4-year-old child who pre-
sented for evaluation of torticollis and recent
change in gait.
Song33 outlined a useful diagnostic al-
gorithm in 1996 based on their retro-
spective review of children with non-
muscular causes of torticollis.
Congenital muscular torticollis, a
painless deformity associated with
contracture of the sternocleidomastoid
muscle, is the most common cause of
torticollis and typically is identified
in the first 2 months of life. This con-
tracture of the sternocleidomastoid
muscle leads to a head tilt toward the
involved side and head rotation to-
ward the opposite side. It is associ-
ated with breech and difficult deliv-
eries as well as other musculoskeletal
disorders, such as metatarsus adduc-
tus, hip dysplasia (Fig. 6), or talipes
equinovarus. The authors of one clin-
ical study reported a 7% to 20% in-
cidence of developmental dysplasia
of the hip in patients with congenital
muscular torticollis.34 Multiple theo-
ries regarding the etiology of congen-
ital muscular torticollis have been pro-
posed, including fibrosis of the
sternocleidomastoid muscle after a
peripartum intramuscular bleed, fi-
brosis caused by a compartment syn-
drome of the sternocleidomastoid
318
muscular compartment,35 intrauterine
crowding, and primary myopathy of
the sternocleidomastoid muscle.36
Congenital muscular torticollis is
more commonly seen on the right
side. A painless mass may be palpa-
ble in the sternocleidomastoid region
in the first 2 weeks of life, reaching
maximum size in 4 weeks, then re-
gressing. By age 4 to 6 months, tor-
ticollis and contracture of the ster-
nocleidomastoid are the only clinical
findings. Persistent torticollis may
lead to skull and facial deformities (ie,
plagiocephaly). A child who sleeps
prone usually lies with the affected
side down, resulting in flattening of
the face on that side. If the child sleeps
supine, flattening of the contralater- Figure 6 Torticollis in an infant with devel-
al skull occurs. This plagiocephaly opmental dysplasia of the hip. (Courtesy
will become permanent if the torticol- Texas Scottish Rite Hospital for Children,
Dallas, TX).
lis persists and is left untreated.33
Treatment usually is nonsurgical.
For infants younger than 1 year, a pro- one series, 11 of 12 patients had a sat-
gram of sternocleidomastoid muscle isfactory result with a bipolar proce-
stretching is recommended. The par- dure combined with Z-plasty of the
ents should be taught to stretch the sternal attachment.38 In another series
child’s contralateral ear to the shoul- of 55 patients, >50% had satisfactory
der and gently push the chin to touch improvement of their plagiocephaly
the shoulder on the same side as the and a 2% recurrence rate.40 Potential
contracted sternocleidomastoid. Nine- surgical complications include inju-
ty percent of cases resolve with such ry to the spinal accessory nerve, jug-
treatment.37 After age 2 years, nonsur- ular veins, carotid vessels, and the fa-
gical treatment is unlikely to be effec- cial nerve. In the postoperative
tive. It is preferable to surgically treat period, patients may do some simple
children with persistent torticollis and stretching exercises, but they often re-
an unacceptable amount of facial quire bracing to maintain corrected
asymmetry before age 3 years.38 How- alignment.
ever, some improvement in facial
asymmetry has been shown even in
children surgically treated as late as Summary
8 years.39
Current surgical options are uni- Understanding the spectrum of pos-
polar or bipolar release. Middle third tural variations that can occur in chil-
transection and complete resection dren younger than 10 years is requi-
are no longer recommended because site to avoid the needless treatment
of risk to the spinal accessory nerve. of benign conditions as well as to dis-
Unipolar release consists of division tinguish true pathologic structural ab-
of the distal portion of the sterno- normalities. Referral of a child to an
cleidomastoid muscle and typically orthopaedic surgeon for in-toeing or
is done for a mild deformity. Bipolar out-toeing is commonplace; for most
release entails division of both the of these children, the etiology of the
sternocleidomastoid origin and inser- complaint can be quickly diagnosed
tion for more notable involvement. In by a systematic assessment of the
Journal of the American Academy of Orthopaedic Surgeons

child’s rotational profile. Knowledge
of the natural history of metatarsus
adductus, tibial rotation, and femo-
ral anteversion is the basis for the ap-
propriate education and reassurance
of families and primary care provid-
ers who are unnecessarily worried
about children with such physiolog-
ic conditions. A need for diagnostic
imaging or active intervention is rel-
References
1. Staheli LT, Corbett M, Wyss C, King H:
Lower-extremity rotational problems in
children: Normal values to guide manage-
ment. J Bone Joint Surg Am 1985;67:39-47.
2. Staheli LT: Rotational problems in chil-
dren. J Bone Joint Surg Am 1993;75:939-949.
3. Smith JT, Bleck EE, Gamble JG, Rinsky
LA, Pena T: Simple method of docu-
menting metatarsus adductus. J Pediatr
Orthop 1991;11:679-680.
4. Fabry G, Cheng LX, Molenaers G: Nor-
mal and abnormal torsional development
in children. Clin Orthop 1994;302:22-26.
5. Wenger DR, Mauldin D, Speck G, Mor-
gan D, Lieber RL: Corrective shoes and
inserts as treatment for flexible flatfoot
in infants and children. J Bone Joint Surg
Am 1989;71:800-810.
6. Driano AN, Staheli L, Staheli LT: Psy-
chosocial development and corrective
shoewear use in childhood. J Pediatr
Orthop 1998;18:346-349.
7. Hunziker UA, Largo RH, Duc G: Neo-
natal metatarsus adductus, joint mobil-
ity, axis and rotation of the lower extrem-
ity in preterm and term children 0-5 years
of age. Eur J Pediatr 1988;148:19-23.
8. Mosca VS: The foot, in Morrissy RT, Wein-
stein SL (eds): Lovell and Winter’s Pedi-
atric Orthopaedics, ed 5. Philadelphia, PA:
Lippincott Williams & Wilkins, 2001, vol
2, pp 1151-1215.
9. Wynne-Davies R, Littlejohn A, Gorm-
ley J: Aetiology and interrelationship of
some common skeletal deformities. (Tal-
ipes equinovarus and calcaneovalgus,
metatarsus varus, congenital dislocation
of the hip, and infantile idiopathic scolio-
sis.) J Med Genet 1982;19:321-328.
10. Reimann I, Werner HH: The pathology
of congenital metatarsus varus: A post-
mortem study of a newborn infant. Acta
Orthop Scand 1983;54:847-849.
11. Morcuende JA, Ponseti IV: Congenital
metatarsus adductus in early human fe-
tal development: A histologic study.
Clin Orthop 1996;333:261-266.
Vol 11, No 5, September/October 2003
Todd L. Lincoln, MD, and Patrick W. Suen, MD
atively uncommon and should be re-
served for children who fall two stan-
dard deviations outside the mean
rotational profile for their age. A sec-
ond common source of orthopaedic
referral consists of a wide variety of
postural pediatric foot abnormalities.
Familiarity with these conditions,
ranging from the routine infant with
a calcaneovalgus foot posture to the
12. GreeneWB:Metatarsusadductusandskew-
foot. Instr Course Lect 1994;43:161-177.
13. Farsetti P, Weinstein SL, Ponseti IV: The
long-term functional and radiographic
outcomes of untreated and non-
operatively treated metatarsus adductus.
J Bone Joint Surg Am 1994;76:257-265.
14. Bleck EE: Metatarsus adductus: Classi-
fication and relationship to outcomes of
treatment. J Pediatr Orthop 1983;3:2-9.
15. Crawford AH, Gabriel KR: Foot and an-
kle problems. Orthop Clin North Am
1987;18:649-666.
16. Cook D, Breed A, Cook T, DeSmet AD,
Muehle CM: Observer variability in ra-
diographic measurement classification
of metatarsus adductus. J Pediatr Orthop
1992;12:86-89.
17. Rushforth GF: The natural history of
hooked forefoot. J Bone Joint Surg Br
1978;60:530-532.
18. Ponseti IV, Becker JR: Congenital meta-
tarsus adductus: The results of treatment.
J Bone Joint Surg Am 1966;48:702-711.
19. Katz K, David R, Soudry M: Below-
knee plaster cast for the treatment of
metatarsus adductus. J Pediatr Orthop
1999;19:49-50.
20. Asirvatham R, Stevens PM: Idiopathic
forefoot-adduction deformity: Medial
capsulotomy and abductor hallucis
lengthening for resistant and severe de-
formities. J Pediatr Orthop 1997;17:496-500.
21. Stark JG, Johanson JE, Winter RB: The
Heyman-Herndon tarsometatarsal cap-
sulotomy for metatarsus adductus: Re-
sults in 48 feet. J Pediatr Orthop 1987;7:
305-310.
22. Lynch FR: Applications of the opening
wedge cuneiform osteotomy in the sur-
gical repair of juvenile hallux abducto
valgus. J Foot Ankle Surg 1995;34:103-123.
23. McHale KA, Lenhart MK: Treatment of
residual clubfoot deformity—the “bean-
shaped” foot—by opening wedge me-
dial cuneiform osteotomy and closing
wedge cuboid osteotomy: Clinical review
rare child presenting with a skewfoot
deformity, is needed to properly se-
lect those children who require treat-
ment. Similarly, recognition of the
high prevalence and clinicial findings
of congenital muscular torticollis,
along with awareness of other, less
common etiologies of torticollis in
children, assists the proper selection
of diagnostic studies and treatment.
and cadaver correlations. J Pediatr Orthop
1991;11:374-381.
24. Mosca VS: Flexible flatfoot and skew-
foot, in Drennan JC (ed): The Child’s Foot
and Ankle. New York, NY: Raven Press,
1992, pp 355-376.
25. Mosca VS: Skewfoot deformity in chil-
dren: Correction by calcaneal neck
lengthening and medial cuneiform
opening wedge osteotomies. J Pediatr
Orthop 1993;13:807-812.
26. Larsen B, Reimann I, Becker-Andersen
H: Congenital calcaneovalgus: With
special reference to treatment and its re-
lation to other congenital foot deformi-
ties. Acta Orthop Scand 1974;45:145-151.
27. Fuchs R, Staheli LT: Sprinting and in-
toeing. J Pediatr Orthop 1996;16:489-491.
28. Turner MS: The association between
tibial torsion and knee joint pathology.
Clin Orthop 1994;302:47-51.
29. Krengel WF III, Staheli LT: Tibial rota-
tional osteotomy for idiopathic torsion:
A comparison of the proximal and dis-
tal osteotomy levels. Clin Orthop 1992;
283:285-289.
30. Reikerås O: Patellofemoral characteris-
tics in patients with increased femoral an-
teversion. Skeletal Radiol 1992;21:311-313.
31. Delgado ED, Schoenecker PL, Rich
MM, Capelli AM: Treatment of severe
torsional malalignment syndrome.
J Pediatr Orthop 1996;16:484-488.
32. Schoenecker PL, Rich MM: The lower
extremity, in Morrissy RT, Weinstein SL
(eds): Lovell and Winter’s Pediatric Ortho-
paedics, ed 5. Philadelphia, PA: Lippin-
cott Williams & Wilkins, 2001, vol 2, pp
1059-1104.
33. Ballock RT, Song KM: The prevalence of
nonmuscular causes of torticollis in chil-
dren. J Pediatr Orthop 1996;16:500-504.
34. Hummer CD, MacEwen GD: The coex-
istence of torticollis and congenital dys-
plasia of the hip. J Bone Joint Surg Am
1972;54:1255-1256.
35. Davids JR, Wenger DR, Mubarak SJ:
319
Common Rotational Variations in Children
Congenital muscular torticollis: Se-
quela of intrauterine or perinatal com-
partment syndrome. J Pediatr Orthop
1993;13:141-147.
36. Tang S, Liu Z, Quan X, Qin J, Zhang D:
Sternocleidomastoid pseudotumor of
infants and congenital muscular torti-
collis: Fine-structure research. J Pediatr
Orthop 1998;18:214-218.
320
37. Binder H, Eng GD, Gaiser JF, Koch B:
Congenital muscular torticollis: Results
of conservative management with
long-term follow-up in 85 cases. Arch
Phys Med Rehabil 1987;68:222-225.
38. Ferkel RD, Westin GW, Dawson EG,
Oppenheim WL: Muscular torticollis: A
modified surgical approach. J Bone Joint
Surg Am 1983;65:894-900.
Journal of the American Academy of Orthopaedic Surgeons
39. Coventry H: Congenital muscular tor-
ticollis in infancy: Some observations
regarding treatment. J Bone Joint Surg
Am 1959;41:815-822.
40. Wirth CJ, Hagena FW, Wuelker N, Sie-
bert WE: Biterminal tenotomy for the
treatment of congenital muscular torti-
collis: Long-term results. J Bone Joint
Surg Am 1992;74:427-434.
 Chronic Massive Rotator Cuff Tears:
Evaluation and Management
Andrew Green, MD

Abstract
Most studies of rotator cuff repairs report high success rates. However, the majority Classification
of these studies combine the results of surgical management of rotator cuff tears of
various sizes; few published reports specifically evaluate the management of chronic Rotator cuff tears usually are classi-
massive tears. Chronic massive rotator cuff tears may be acute traumatic, chronic fied according to the chronicity and
atraumatic, or acute-on-chronic. A detailed history and thorough physical exami- size of the tear. The chronicity of a ro-
nation often are sufficient to establish the diagnosis. Radiographic evaluation can tator cuff tear can refer to the dura-
reveal osseous changes suggestive of pathology. Magnetic resonance imaging can tion of either symptoms or patholo-
determine the size of rotator cuff tears and status of the muscles but generally is not gy, although the former is easier to
necessary for patients who are not candidates for surgery. Chronic massive rotator quantify. The duration of time from
cuff tears without glenohumeral arthritis can be managed nonsurgically or with sub- injury to surgery is important; mobi-
acromial débridement, rotator cuff repair, or rotator cuff reconstruction. However, lization and repair of the retracted
treatment of these patients is challenging, and results are comparatively inferior to tendon can be difficult even as early
those of treating patients with smaller rotator cuff tears. as 6 weeks after traumatic massive ro-
J Am Acad Orthop Surg 2003;11:321-331 tator cuff tear.5 Patients who present
without a history of injury usually
have chronic intermittent pain and
dysfunction. In such patients, the du-
Codman1 accurately described rota- more likely to have persistent defects ration of pathology is more difficult
tor cuff pathology in 1934; subse- and weakness and that the size of the to determine because the onset usu-
quently, others reported on the tech- persistent defect correlated with the ally predates that of symptoms.
nical aspects of rotator cuff repair result. Only a few investigators have There are several approaches for
and methods of addressing acromial specifically studied the management classifying the size of rotator cuff
impingement. Recognizing that in- of massive rotator cuff tears. Some tears. The most common is based on
dividuals with rotator cuff tears have advocated either arthroscopic or the largest dimension of the tear:
could be asymptomatic, McLaugh- open débridement of massive small tears measure <1 cm; medium
lin2 stated that surgery for rotator tears.9,10 Others report satisfactory tears, 1 to <3 cm; large tears, 3 to <5
cuff disorders was indicated not by long-term results in as many as 90%
the diagnosis but by pain and dis- of patients who have undergone re-
ability. Early reports3,4 of surgical pair of a massive rotator cuff
Dr. Green is Associate Professor, Department of
management showed only a 50% to tear.6,7,11 Other authors have consid- Orthopaedic Surgery, Brown University School
60% success rate, whereas more re- ered alternative surgical options, such of Medicine, Providence, RI.
cent studies have shown successful as tendon and muscle mobilization
outcomes (ie, good pain relief and and transfers, synthetic and other tis- Reprint requests: Dr. Green, Suite 200, 2 Dud-
functional restoration) in 80% to sue interposition, arthrodesis, and ar- ley, Providence, RI 02905.
90% of cases.5-7 throplasty, for the management of Neither Dr. Green nor the department with which
Harryman et al8 used ultrasound massive rotator cuff tears considered he is affiliated has received anything of value from
to evaluate the effect of rotator cuff to be irreparable. Determining the or owns stock in a commercial company or insti-
tution related directly or indirectly to the subject
tear size and repair integrity on the best treatment can be difficult. No of this article.
outcome of surgical management. Al- randomized prospective studies com-
though tear size and repair integrity paring nonsurgical and surgical treat- Copyright 2003 by the American Academy of
did not correlate with pain relief, the ment exist, nor are there studies com- Orthopaedic Surgeons.
authors found that larger tears were paring the various surgical options.

Vol 11, No 5, September/October 2003 321

Chronic Massive Rotator Cuff Tears
cm; and massive tears, ≥5 cm. Rota-
tor cuff tears also are classified ac-
cording to the number of tendons in-
volved. Additionally, the extent of
tendon retraction and tissue quality
are important factors that are not gen-
erally accounted for by the various
classification systems. The tear size
expressed in area is probably the most
useful classification method.
Presentation
The incidence of rotator cuff tearing
increases with age. This has been con-
firmed by cadaveric studies as well
as by magnetic resonance imaging
(MRI) and ultrasonographic studies
of patients with asymptomatic shoul-
ders.12,13 Sher et al12 reported a 28%
incidence of full-thickness rotator cuff
tears in asymptomatic individuals
older than 60 years.
Massive rotator cuff tears present
in three clinical settings: acute trau-
matic, acute-on-chronic, and chronic
atraumatic. Although massive tearing
occurs in a broad age range, patients
with acute traumatic massive tears tend
to be younger than those with chronic
atraumatic massive tears. Patients with
acute traumatic massive tears do not
have the typical signs of chronicity,
such as substantial supraspinatus or
infraspinatus muscle atrophy.
Patients with acute-on-chronic
tears present with two different clin-
ical scenarios. The first are those with
preexisting chronic symptomatic ro-
tator cuff tearing who sustain an in-
jury that causes an acute extension of
the tear. The second are those with a
preexisting asymptomatic massive
rotator cuff tear and an acute injury
that results in the onset of shoulder
pain. After an acute traumatic injury,
patients often are unable to actively
elevate the injured arm. Some pa-
tients present with extensive anteri-
or arm ecchymosis. Patients who
deny preexisting symptoms but have
the typical clinical features of chron-
ic rotator cuff tearing, such as spina-
322
tus muscle atrophy, external rotation
weakness, and plain radiographic
findings, are likely to have had pre-
existing rotator cuff tearing. Each of
these scenarios is unique and requires
a different management approach.
Many patients with chronic mas-
sive rotator cuff tears do not have a
history of a significant shoulder in-
jury but instead report an insidious
onset of symptoms. They compose
the group with chronic atraumatic
massive rotator cuff tear and are of-
ten older, less active, and present with
complaints of shoulder pain. Because
of the chronicity of rotator cuff tear-
ing, these patients usually have in-
fraspinatus muscle atrophy and may
have a chronic biceps tendon rupture.
Despite severe rotator cuff deficien-
cy, some of these patients have good
elevation strength because of com-
pensatory deltoid muscle strength.
Patients with profound deltoid weak-
ness may be unable to actively elevate
the arm. MRI scans can be helpful be-
cause they can demonstrate fatty re-
placement of the rotator cuff muscles
and confirm chronic pathology.14
Some patients with chronic mas-
sive rotator cuff tears have a history
of a traumatic injury that caused the
tear. These individuals tend to be
younger and complain of pain and
weakness. They usually report initial
inability to raise the arm, with grad-
ual recovery of the capacity to elevate
the arm overhead.
Physical Examination
A detailed physical examination is es-
sential to accurately determine the
status of the rotator cuff. The size of
the tear usually can be determined by
physical examination alone. Because
cervical spondylosis and radiculop-
athy can cause shoulder pain that
mimics the pain of rotator cuff pathol-
ogy, the evaluation should include a
careful examination of the cervical
spine and a thorough neurologic as-
sessment. Shoulder girdle weakness
Journal of the American Academy of Orthopaedic Surgeons
and atrophy can be the result of cer-
vical stenosis, brachial plexus disor-
ders (eg, Parsonage-Turner syndrome,
brachial neuritis, tumor), or supra-
scapular neuropathy.
Visual inspection of the patient is
very important. Both shoulders, in-
cluding the scapulas, must be exposed.
Several findings are consistent with
chronic massive rotator cuff tearing,
including anterior-superior sublux-
ation and prominence of the humeral
head, infraspinatus muscle atrophy,
and rupture of the proximal tendon
of the long head of the biceps. Su-
praspinatus atrophy is more difficult
to detect beneath the trapezius mus-
cle. Deltoid detachment is rare but
usually is visible as a defect at the an-
terior aspect of the origin of the mid-
dle deltoid. Spontaneous deltoid rup-
ture is a rare but disabling sequela of
chronic massive rotator cuff tearing,
probably resulting from erosion of the
deltoid origin by the humeral head
as it articulates with the acromion. If
not identified and repaired early, func-
tional outcome typically is dismal.15
Amassive rotator cuff tear occasion-
ally can be detected as a palpable de-
fect in the supraspinatus tendon at the
anterolateral aspect of the shoulder.1
Swelling caused by subdeltoid syn-
ovial fluid may be present. Similarly,
when the deltoid is disrupted, there
is a sulcus and palpable defect at the
edge of the acromion.
Loss of passive shoulder motion is
uncommon in the presence of a large
or massive rotator cuff tear. It is im-
portant to recognize shoulder stiff-
ness because strength is more diffi-
cult to evaluate in its presence.
Glenohumeral stiffness can be caused
by adhesive capsulitis, capsular con-
tracture, and glenohumeral arthritis.
Although idiopathic adhesive capsu-
litis is not usually associated with
massive rotator cuff tearing, stiffness
can occur with massive chronic tears
as a result of injury or failure to move
the shoulder.
External rotation weakness is char-
acteristic of massive rotator cuff tear-

ing. Elevation weakness is a less con-
sistent finding. Some patients have
sufficient deltoid strength to mask the
absence of supraspinatus strength.
The Jobe empty can test, which as-
sesses strength with the shoulder el-
evated approximately 90° and inter-
nally rotated with the thumb pointing
downward, usually will cause pain
and elicit weakness. Pain also can be
the cause of inability to elevate the
arm. A subacromial injection of 10 mL
of 1% lidocaine can eliminate the pain
and allow a better assessment of ro-
tator cuff strength.
In contrast to the findings associ-
ated with chronic massive rotator cuff
tearing, patients with acute traumat-
ic massive rotator cuff tears may have
more obvious signs of acute injury (eg,
swelling, ecchymosis). Patients with
acute traumatic tears usually have
good infraspinatus muscle bulk.
Rotator cuff strength can be as-
sessed with manual muscle testing
(Fig. 1, A and B). External rotation
weakness and external rotation lag
are signs of massive rotator cuff tear-
ing involving the infraspinatus ten-
don. An external rotation lag sign is
elicited by passively positioning the
arm in maximal external rotation.
When there is marked weakness, the
patient is unable to hold the arm in
this position and the hand falls to-
ward the abdomen. The hornblower’s
sign, an inability to externally rotate
Figure 1 A, Patient with a massive rotator cuff tear whose right shoulder has an external rotation lag sign. The arm can be passively ex-
ternally rotated. B, The patient cannot actively externally rotate the arm. C, Positive liftoff maneuver in a patient with a subscapularis tendon
tear of the right shoulder. (Panel C courtesy of Tom R. Norris, MD.)
Vol 11, No 5, September/October 2003
the elevated arm, also demonstrates
severe infraspinatus weakness.
Associated subscapularis muscle
tearing is relatively uncommon but
can be easily identified if one is aware
of the problem. Patients with sub-
scapularis disruption have internal
rotation weakness, excessive passive
external rotation, and a positive lift-
off test (Fig. 1, C) or belly-press test.16
The liftoff test is difficult to do when
pain or limited shoulder motion pre-
vents positioning of the arm and hand
behind the back. Subscapularis tears
may be overlooked by the inexperi-
enced examiner.
The combination of a detailed his-
tory and thorough physical examina-
tion often provides sufficient informa-
tion to establish a diagnosis of massive
rotator cuff tear. Imaging studies pro-
vide information to confirm the diag-
nosis and assist in treatment selection.
Radiographic Evaluation
Plain Radiographs
A complete evaluation of the
shoulder includes a series of five plain
radiographic views: anteroposterior,
anteroposterior in both internal and
external rotation, axillary lateral, and
outlet. Although plain radiographs
do not visualize soft tissues, they
show osseous changes that suggest
the presence of rotator cuff patholo-
Andrew Green, MD
gy and are particularly helpful in as-
sessing patients with chronic massive
rotator cuff tears. Elevation of the hu-
meral head relative to the glenoid and
narrowing of the acromiohumeral
space are findings consistent with
long-standing rotator cuff pathology
(Fig. 2, A). It has been suggested that
an acromiohumeral space <7 mm is
consistent with a rotator cuff tear, and
that a space <5 mm indicates a mas-
sive tear.
The true anteroposterior and ax-
illary lateral radiographs can demon-
strate even very mild glenohumeral
arthritis. The axillary lateral view also
demonstrates the relative anteropos-
terior position of the humeral head.
In advanced cases, the humeral head
can be seen articulating with the cora-
coid. The outlet view is used to dem-
onstrate the acromial morphology.
Several studies have reported that
acromial spurring is associated with
rotator cuff tearing. The anteroposte-
rior view in internal and external ro-
tation assists evaluation of the acro-
mioclavicular joint.
The typical radiographic findings
of rotator cuff tear arthropathy in-
clude loss of joint space, elevation of
the humeral head, erosion and round-
ing off of the greater tuberosity, and
articulation of the humeral head with
the acromion (Fig. 2, B). If a complete
series of radiographs is not obtained,
the articular involvement of the hu-
323
Chronic Massive Rotator Cuff Tears
Figure 2 A, Neutral anteroposterior radiograph of a shoulder with a chronic massive ro-
tator cuff tear. Although there is reduction of the acromiohumeral space and the humeral
head is elevated relative to the glenoid, there is no glenohumeral arthritis. B, Neutral an-
teroposterior radiograph of a shoulder with rotator cuff tear arthropathy. There is no acro-
miohumeral space, there are degenerative changes of the glenohumeral joint, and the greater
tuberosity is rounded off.
meral head can be obscured, the di-
agnosis missed, and inappropriate
treatment (eg, acromioplasty, rotator
cuff repair) instituted.
Arthrography
In the past, arthrography was fre-
quently cited as the best diagnostic
tool for evaluation of a rotator cuff
tear, but it has several limitations and
disadvantages. It is invasive and does
not provide as much additional an-
atomic information as an MRI. Imag-
ing modalities such as ultrasonogra-
phy or MRI provide more detailed
information about the size of the tear
and the condition of the rotator cuff
muscles, and they have replaced ar-
thrography as the imaging tests of
choice.17,18
Ultrasonography
Acceptance of ultrasonographic
imaging of the rotator cuff has been
variable. Centers with extensive ex-
perience report high rates of sensitiv-
ity and specificity,18 although these
findings have not been confirmed by
others. Some of the variability is re-
lated to inexperience and some to dif-
324
fering diagnostic criteria. More recent
technologic advances are encourag-
ing a reevaluation of the value of ul-
trasonography as a tool for imaging
the rotator cuff.
Magnetic Resonance Imaging
MRI is highly accurate and dem-
onstrates detailed anatomic informa-
tion of the rotator cuff.17 Unlike arthrog-
raphy, MRI can be used to determine
the size of rotator cuff tears and the
status of the rotator cuff muscles. The
latter information can help establish
the chronicity of the tearing and aid
in treatment decision making.
Coronal oblique MRI scans are used
to evaluate the supraspinatus tendon
and muscle. The extent of retraction
and the size and quality of the su-
praspinatus muscle can be determined
(Fig. 3, A). Fatty replacement of the
supraspinatus muscle in the su-
praspinatus fossa indicates chronic pa-
thology.14 The size of the supraspina-
tus tear in the anterior-posterior
direction can be assessed by noting
the tear on sequential images. Sagit-
tal oblique images demonstrate the
anterior-posterior extent of supraspina-
Journal of the American Academy of Orthopaedic Surgeons
tus tearing as well as the quality of
all of the rotator cuff muscles (Fig. 3,
B). Axial images can demonstrate the
condition of the biceps tendon as well
as that of the subscapularis and in-
fraspinatus tendons and muscles. Mag-
netic resonance arthrography is not
typically indicated for the evaluation
of the rotator cuff, although it can be
helpful in imaging the rotator cuff af-
ter surgery, when it is difficult to dif-
ferentiate scar tissue from tendon.
Although many studies have con-
firmed the accuracy of MRI for assess-
ing the rotator cuff, specific indications
have rarely been addressed.17 Over-
use of MRI is a pervasive problem,
and careful clinical evaluation can help
define the appropriate application of
this excellent imaging technique. Pa-
tients with an insidious onset of shoul-
der pain and dysfunction do not re-
quire advanced imaging until after
appropriate nonsurgical treatment has
failed. Patients for whom surgery is
not a consideration do not need MRI
scans unless there is concern about an-
other pathologic entity, such as an in-
fection or neoplasm.
Management
Successful outcome depends on se-
lecting the best treatment option for
the specific patient. Chronic massive
rotator cuff tears present in a variety
of patients and can be treated with
different approaches, including non-
surgical methods, débridement, re-
pair, and reconstruction. High success
rates have been reported with each
of these approaches.6-11,19-27 In some
cases, more than one of the options
may be appropriate. Thus, careful
consideration of the many related fac-
tors is important.
Nonsurgical Management
Many patients with chronic mas-
sive rotator cuff tears can be treated
successfully without surgery. Activ-
ity modification, oral nonsteroidal
anti-inflammatory medications, and

Figure 3 MRI scans of the left shoulder of a 64-year-old man with a chronic rotator cuff tear. A, T2-weighted coronal oblique image demon-
strates the supraspinatus tendon tear (arrows), atrophy of the muscle, and fatty replacement in the supraspinatus fossa. B, T1-weighted sag-
ittal oblique image demonstrates fatty replacement in the supraspinatus fossa (asterisk) and lipoatrophy of the infraspinatus muscle (arrow).
subacromial corticosteroid injections
can help manage symptoms. In their
study of nonsurgical treatment of pa-
tients with full-thickness rotator cuff
tears, Bokor et al19 noted improve-
ment in 50% to 85% of patients. Du-
ration of symptoms seemed to corre-
late with the long-term success of
nonsurgical management because pa-
tients with symptoms for longer than
6 months had poorer outcomes.19
Although some studies have
shown a negative correlation between
number of preoperative corticoste-
roid injections and results of rotator
cuff repair,28 others have not.6 Patients
with a longer history of shoulder pain
are more likely to have had multiple
injections. Subacromial injections can
be very helpful in the initial phases
of physical therapy and rehabilitation
because with less pain, patients are
better able to comply with a physi-
cal therapy regimen. For most pa-
tients, the total number of injections
should be limited to three because, in
general, repeated corticosteroid injec-
Vol 11, No 5, September/October 2003
tions have a detrimental effect on the
rotator cuff and articular surfaces.
Physical therapy is used to restore
shoulder motion and strengthen the
intact portions of the rotator cuff,
periscapular muscles, and deltoid. Mo-
tion can be improved with passive
stretching exercises. Strengthening of
the internal and external rotators is
best achieved with resistance exercises
done with the arms below chest level.
Deltoid strengthening should be ini-
tiated in the supine position with the
effects of gravity minimized, progress-
ing to a seated or standing position.
In addition, scapular muscle strength-
ening can enhance the function of a
weak rotator cuff. Strengthening should
be progressed gradually and within
the patient’s comfort level.
Surgical Management
Surgical management of chronic
massive rotator cuff tears encompass-
es a spectrum of complexity, from
minimally invasive arthroscopic ap-
proaches to major reconstructive sur-
Andrew Green, MD
gery. In some instances, subacromial
decompression and rotator cuff dé-
bridement alone may be adequate to
manage massive irreparable rotator
cuff tears;9,10,20 satisfactory results
have been reported in patients under-
going this procedure primarily and
in those with persistent cuff defects
after the repair.8,11
Débridement and Acromioplasty:
Open and Arthroscopic
Débridement and acromioplasty
are best suited for lower-demand in-
dividuals and those unwilling to co-
operate with the prolonged and rig-
orous rehabilitation required after
rotator cuff repair and reconstruction.
Because the results of débridement of
full-thickness rotator cuff tears may
deteriorate with time, active individ-
uals who do not improve with non-
surgical treatment are probably bet-
ter served by attempted rotator cuff
repair.29 The ideal candidate for dé-
bridement is a relatively inactive in-
dividual with shoulder pain who has
325
Chronic Massive Rotator Cuff Tears
good elevation strength, can actively
elevate the arm overhead, and can ex-
ternally rotate the arm with gravity
eliminated. Exhibiting these abilities
suggests that there are good shoulder
kinematics and balanced internal and
external rotators.20
The arthroscopic approach for sub-
acromial débridement is easier and has
a more rapid rehabilitation than the
open approach because the deltoid or-
igin is preserved.9 With a limited ac-
romioplasty, undersurface spurring
and rough excrescences are removed
and the greater tuberosity is smoothed.
Thus, the coracoacromial arch is main-
tained by avoiding excessive acromio-
plasty and by preserving the cor-
acoacromial ligament, which helps
prevent loss of the restraint to supe-
rior humeral head subluxation. Loss
of the coracoacromial arch with ac-
companying anterior-superior dislo-
cation or subluxation of the humeral
head is an unsolved problem.
The open approach is preferred when
the preoperative evaluation suggests
that the rotator cuff tear is partially
reparable. Alternatively, the rotator cuff
tear can be assessed with arthroscopy,
and a partial repair can be done by
arthroscopy, a mini-open technique,
or a traditional open technique.
Biceps tenotomy or tenodesis also
has been recommended as an adjunct
to arthroscopic débridement of chron-
ic massive rotator cuff tears. If there
is subluxation, dislocation, or partial
tearing of the tendon of the long head
of the biceps, tenotomy or tenodesis
may enhance the ability to alleviate
shoulder pain. Arthroscopic tenoto-
my is a minimally invasive procedure
that does not require the postopera-
tive immobilization or protection re-
quired after tenodesis, making it pref-
erable for the older patient.30
Rotator Cuff Repair
Early procedures for rotator cuff re-
pair encompassed a variety of tech-
niques, including partial and total
acromionectomy, to treat acromial
impingement and supraspinatus
326
syndrome. Satisfactory results of ro-
tator cuff repair in these early reports
occurred in only 50% to 60% of pa-
tients.3,4 Neer’s31 anterior acromio-
plasty became the preferred technique
for treating impingement syndrome
and for decompressing the subacro-
mial space in conjunction with rota-
tor cuff repair. Neer and Marberry32
also highlighted the problems with
acromionectomy, especially those re-
lated to deltoid detachment. More re-
cent studies of rotator cuff repair have
reported good function and pain re-
lief in 80% to 90% of cases.5-7
The goal of rotator cuff surgery is
to repair the rotator cuff tendons to
the proximal humerus and to decom-
press the subacromial space without
disrupting the coracoacromial arch.
Several techniques have been used to
mobilize the rotator cuff tendons and
facilitate repair. The ability to repair
the tear is primarily related to the size
and chronicity of the tear. Although
the preoperative evaluation provides
considerable information to help de-
termine the reparability of a chronic
massive rotator cuff tear, no tear can
be considered irreparable until a re-
pair is attempted, even though at-
tempted repair is not indicated in all
cases. Tendon loss, retraction, and scar-
ring are commonly encountered and
must be addressed to achieve a sat-
isfactory repair. In the past, these tears
were best managed with open rather
than arthroscopic approaches. More
recently, arthroscopy has been used
to repair larger chronic rotator cuff tears.
For an open repair, a skin incision
is made over the top of the lateral third
of the acromion along Langer’s lines.
The incision can be extended anteri-
orly to permit a deltopectoral approach
to repair a subscapularis tear if need-
ed. The skin and subcutaneous tissue
are elevated as full-thickness flaps to
expose the acromion and the origins
of the anterior and middle heads of
the deltoid muscle. The anterior del-
toid is elevated off the anterior acro-
mion and the deltoid fibers are split
longitudinally in a lateral direction just
Journal of the American Academy of Orthopaedic Surgeons
posterior to the deltoid raphe for ap-
proximately 3 to 4 cm. By doing this,
the coracoacromial ligament is released
from the acromion but not resected.
This approach allows ample access to
the subacromial and subdeltoid spac-
es as well as to the supraspinatus, in-
fraspinatus, and teres minor tendons.
Although some authors recommend
distal clavicle resection to improve ex-
posure, this should be done only when
there is clinically relevant acromioclav-
icular arthritis.
Subacromial and subdeltoid adhe-
sions are released, and excess bursal
tissue can be excised to allow visu-
alization of the rotator cuff tear. Ad-
hesions under the anterior deltoid are
often more tenacious and require for-
mal dissection to clearly visualize the
rotator cuff interval and subscapularis
tendon.
Anterior acromioplasty can be ac-
complished with a variety of tech-
niques using chisels, burs, and rasps.
The goal is to flatten the undersurface
of the anterior acromion, decompress
the subacromial space, and create a
smooth acromial surface. Spurring on
the undersurface of the distal clavi-
cle also is removed. Colman et al33
demonstrated that removal of 5.4 mm
of the undersurface of the anterior
acromion reduces the contact pressure
of the acromion on the supraspina-
tus tendon. Care should be taken to
avoid excessive resection, especially
shortening of the acromion, to prevent
anterior-superior instability. The sub-
scapularis tendon then is inspected for
tearing. Tears of the upper third of the
subscapularis often can be repaired
through this exposure. More extensive
tears may require a separate deltopec-
toral approach.
The supraspinatus and infraspina-
tus tendon edges are identified and
traction sutures are placed. In contrast
to cases of acute massive tears that
usually can be easily mobilized to the
greater tuberosity, additional steps are
required to mobilize chronic massive
tears. Adhesions superficial to the su-
praspinatus and infraspinatus mus-
 Andrew Green, MD

cles are bluntly released. Release of

the rotator cuff interval and the cor-
acohumeral ligament at the base of
the coracoid helps to mobilize the su-
praspinatus tendon. Capsular releas-
es superiorly and posteriorly also can
improve mobility (Fig. 4). Occasion-
ally, the interval between the su-
praspinatus and infraspinatus ten-
dons, the so-called posterior interval,
is released to allow differential mo-
bilization of the supraspinatus and in-
fraspinatus tendons (Fig. 5). The lim-
it of mobilization of the supraspinatus
and infraspinatus muscles is deter-
mined by the extent to which the su-
prascapular nerve can be mobilized.
Warner et al34 found that the standard
anterosuperior approach allowed only
1 cm of lateral advancement of the ten-
dons, whereas the Debeyre technique, Figure 4 Glenohumeral capsular releases done to mobilize the rotator cuff in a right shoul-
der. A, The dotted line indicates the site of capsular release peripheral to the glenoid labrum.
in which the supraspinatus muscle is B, A blunt elevator is used to separate the capsule and rotator cuff tendons from the labrum.
mobilized from the scapula, permit- (Adapted with permission from Warner JJP, Gerber C: Massive tears of the posterior-superior
ted up to 3 cm of lateral advancement. rotator cuff, in Warner JJP, Iannotti JP, Gerber C [eds]: Complex and Revision Problems in Shoul-
der Surgery. Philadelphia, PA: Lippincott-Raven, 1997, pp 177-202.)
Once the rotator cuff is fully mo-
bilized, the insertion site on the prox-
imal humerus is prepared by decor- periosteum,middledeltoid,andacro- and reattached to the acromion as
ticating the bone just lateral to the mioclavicular joint capsule. A secure part of the deltoid repair.
articular surface. Tendon reattach- repair of the deltoid is critical to avoid The rotator cuff should be sutured
ment can be accomplished by pass- postoperative deltoid avulsion. The with the arm at the side because a re-
ing no. 2 braided nonabsorbable su- coracoacromial ligament is preserved pair done with the shoulder abducted
tures through transosseous tunnels or
by using suture anchors. Gerber et
al35 found that the modified Mason-
Allen suture technique has the best
holding strength in tendon tissue of
the many suture constructs they me-
chanically tested. More recent stud-
ies demonstrate that some suture an-
chors have better strength in bone
than do transosseous sutures. Aug-
mentation of the bone of the greater
tuberosity with small plates or
washer-type devices is advocated to
improve the strength of the suture-
bone interface.11
After the rotator cuff repair is com-
plete, the anterior deltoid origin is re-
attached to the anterior acromion Figure 5 Superior view of a left shoulder showing anterior and posterior rotator cuff in-
with nonabsorbable sutures passed terval releases to mobilize a large tear of supraspinatus and infraspinatus tendons. A, Dot-
ted lines indicate the releases. B, The supraspinatus tendon is mobilized (black arrow) sep-
through drill holes in the acromion. arately from the infraspinatus. C, The tendon is sutured into place to complete the repair.
This repair is reinforced with addi- (Adapted with permission from Codd T, Flatow E: Anterior acromioplasty, tendon mobili-
tional nonabsorbable sutures that fix zation, and direct repair of massive rotator cuff tears, in Burkhead WZ [ed]: Rotator Cuff Dis-
orders. Baltimore, MD: Williams & Wilkins, 1996, p 330.)
the anterior deltoid to the acromial

Vol 11, No 5, September/October 2003 327

Chronic Massive Rotator Cuff Tears
will have excessive tension and be likely
to fail. If such positioning is not pos-
sible, a partial repair is done or ro-
tator cuff reconstruction is considered.21
Burkhart et al21 advocated partial re-
pair that reestablishes the cable con-
struct of the rotator cuff. In part, this
is achieved by repairing the cuff de-
fect side-to-side. Often the infraspina-
tus muscle can be mobilized laterally
and superiorly to be repaired to the
greater tuberosity.Although some have
advocated reattachment of the rota-
tor cuff more medially into the artic-
ular surface, this is not widely accepted.
Postoperative recovery and reha-
bilitation is lengthy. The repair is pro-
tected with an arm sling or abduction
immobilizer for 6 to 8 weeks. Abduc-
tion positioning is used to relieve ten-
sion on repairs done with the arm at
the side. Postoperative abduction po-
sitioning should not be used to allow
repair of an irreparable tear. Passive
stretching exercises to regain shoul-
der motion are begun the day after
surgery. After repair of chronic mas-
sive tears, passive internal rotation and
horizontal adduction are avoided for
the first 6 weeks to protect the in-
fraspinatus repair. Light active use and
active-assisted range-of-motion move-
ment are initiated after 6 weeks. For-
mal strengthening is delayed until 12
weeks after surgery. Overall recovery
can take more than 12 months. Over-
ly aggressive early rehabilitation has
been implicated as a cause of failure.
No one has specifically studied the
relation between length of time from
occurrence of massive rotator cuff tears
to repair and surgical outcome. How-
ever, Bassett and Cofield5 found bet-
ter results when tears were repaired
within 6 weeks of injury. In the case
of acute massive rotator cuff tears, ear-
ly repair is technically easier and prob-
ably more likely to restore shoulder
strength. Consequently, most authors
recommend early repair for these tears
in active, healthy individuals.
Few researchers have specifically
analyzed the results of repair of mas-
sive rotator cuff tears. Most series com-
328
bine the results of a variety of tear siz- tator cuff tear, that more than 1 year
es; in general, outcomes for repair of was required for the restoration of
larger tears are inferior to those for strength, and that the final strength
smaller ones.8 Harryman et al8 and Ger- was less than that in the contralateral
ber et al11 reported that repair integ- shoulder.
rity, not original tear size, best corre-
lated with the functional outcome of Rotator Cuff Reconstruction
rotator cuff repair. In their study of Approaches used to reconstruct the
repair of chronic massive rotator cuff muscles and tendons damaged by ir-
tears, Bigliani et al6 reported 85% good reparable massive rotator cuff tears in-
and excellent long-term results. clude transfers of the rotator cuff ten-
Björkenheim et al36 reported that the dons, other muscle and tendon
results of repair of large and massive transfers, and tissue and synthetic ma-
rotator cuff tears were markedly in- terial substitution.22-27,38-41 Subscapu-
ferior to the results of repair of smaller laris tendon transfer can be used to
tears. Most recently, Jost et al37 stud- achieve complete rotator cuff repair
ied the clinical outcome of patients with when repair of the supraspinatus and
rerupture after rotator cuff repair and infraspinatus muscles leaves a resid-
found that the outcome correlated sig- ual superior defect. The upper third
nificantly with the size of the remain- of the subscapularis tendon is sepa-
ing tear (P < 0.005) and the extent of rated from the anterior capsule and
fatty degeneration (P < 0.05) of the in- then transferred superiorly22 (Fig. 6).
fraspinatus and subscapularis mus- Most series do not report using sub-
cles. The size of the tear at follow-up scapularis tendon transfer because such
was related to the size of the original transfer risks internal rotation weak-
tear; larger persistent tears were as- ness or internal rotation contracture.
sociated with larger initial tears. How- Debeyre et al23 first described lat-
ever, Rokito et al7 reported that all of eral advancement of the supraspina-
their patients were satisfied after re- tus muscle and found that repair of
pair of a chronic large or massive ro- large rotator cuff tears was improved
Figure 6 Subscapularis tendon transfer for a supraspinatus tear that cannot be repaired pri-
marily. A and B, The upper portion is elevated off the anterior capsule and transferred su-
periorly. The inferior muscular insertion of the subscapularis is left intact. (Adapted with
permission from Warner JJP, Gerber C: Massive tears of the posterior-superior rotator cuff,
in Warner JJP, Iannotti JP, Gerber C [eds]: Complex and Revision Problems in Shoulder Surgery.
Philadelphia, PA: Lippincott-Raven, 1997, pp 177-202.)
Journal of the American Academy of Orthopaedic Surgeons
 Andrew Green, MD

with this procedure. Ha’eri and Wiley38
also reported good results with lat-
eral advancement of the supraspina-
tus muscle and, importantly, noted that
the muscle was not denervated by the
procedure. Although Warner et al34
demonstrated that the supraspinatus
could be safely mobilized up to 3 cm
laterally, formal lateral advancement
of the supraspinatus muscle usually
is not done.
Latissimus dorsi muscle transfer is
used to substitute for loss of the in-
fraspinatus and supraspinatus ten-
dons (Fig. 7). Latissimus dorsi trans-
fer to the rotator cuff has been used
to treat individuals with obstetric bra-
chial plexus palsy and residual shoul-
der weakness. A healthy, strong latis-
simus dorsi muscle is used to restore
external rotation and head depression
forces that are lost with chronic mas-
sive rotator cuff tears. Gerber26 report-
ed that the results of latissimus dorsi
transfer for massive rotator cuff tear
were better with an intact subscapu-
laris tendon. Miniaci and MacLeod27
reported 82% satisfactory results (14/
17 patients) after latissimus dorsi
transfer in patients with previously
failed surgical treatment of massive
rotator cuff repair. They noted that
primary latissimus dorsi transfer for
an irreparable massive rotator cuff
tear is rarely indicated, and they still
recommended subacromial decom-
pression and repair with local tissue
as the initial surgical procedure.
Teres minor transfer, deltoid mus-
cular flap transfer, and trapezius
transfer have been described but are
infrequently used.23-25,41 Although all
of these procedures are used to cre-
ate a substitute for the absence of su-
praspinatus function, they do not ad-
dress or restore the balance between
the anterior and posterior force cou-
ples of the rotator cuff.
Tissue substitution with synthetic
materials, as well as autologous and
autogenous tissue implants, has been
attempted,39,40 but there are a lim-
ited number of published results.
Neviaser et al39 reported 88% good
and excellent results (14/16 patients)
with freeze-dried cadaveric rotator
cuff tissue used to repair chronic mas-

Figure 7 Latissimus dorsi muscle transfer for massive irreparable rotator cuff tear. The patient is in the lateral decubitus position. A, The
latissimus dorsi muscle and tendon are dissected (dotted line). B, The latissimus dorsi tendon is released from the humerus. C, Sutures are
placed into the tendon. D, The tendon is passed inferior to the posterior deltoid and the acromion to the superior wound. E and F, Superior
view. The transferred tendon is sutured to the edge of the mobilized but deficient cuff edge and the greater tuberosity. (Adapted with per-
mission from Warner JJP, Gerber C: Massive tears of the posterior-superior rotator cuff, in Warner JJP, Iannotti JP, Gerber C [eds]: Complex
and Revision Problems in Shoulder Surgery. Philadelphia, PA: Lippincott-Raven, 1997, pp 177-202.)

Vol 11, No 5, September/October 2003 329

Chronic Massive Rotator Cuff Tears
sive rotator cuff tears. The disadvan-
tages of the material are the poten-
tial for foreign body reaction to
synthetics and tissue rejection. Also,
such materials do not replace the atro-
phic and weakened rotator cuff mus-
cles that typically are present with
chronic massive rotator cuff tears.
Glenohumeral Arthrodesis
Glenohumeral arthrodesis, origi-
nally indicated for glenohumeral ar-
thritis, polio, and brachial plexus
palsy,42 is used infrequently. Shoul-
der arthrodesis is indicated with
loss of deltoid and rotator cuff func-
tion. Arthrodesis can be done for
painful chronic massive irreparable
rotator cuff tears if the goal is a
strong, stable shoulder girdle. It
may be appropriate for the patient
with a painful shoulder with anteri-
or-superior dislocation because of
loss of the coracoacromial arch. How-
ever, rotator cuff reconstruction is
preferred if the articular surfaces are
intact. Otherwise, shoulder arthrod-
esis may result in undesirable loss of
upper extremity function below the
chest level. Arthrodesis also is diffi-
cult to achieve in the osteopenic eld-
erly patient with a chronic massive
rotator cuff tear. Arntz et al43 re-
ported that the results of humeral
head replacement for rotator cuff
References
1. Codman EA: The Shoulder: Rupture of the
Supraspinatus Tendon and Other Lesions
in or About the Subacromial Bursa. Bos-
ton, MA: Thomas Todd, 1934.
2. McLaughlin HL: Lesions of the muscu-
lotendinous cuff of the shoulder: I. The
exposure and treatment of tears with re-
traction. J Bone Joint Surg 1944;26:31-51.
3. Bosworth DM: An analysis of 28 consec-
utive cases of incapacitating shoulder le-
sions radically explored and repaired.
J Bone Joint Surg 1940;22:369-392.
4. Hammond G: Complete acromionecto-
my in the treatment of chronic tendini-
tis of the shoulder. J Bone Joint Surg Am
1962;44:494-504.
5. Bassett RW, Cofield RH: Acute tears of
330
tear arthropathy were better than
those with arthrodesis.
Shoulder Arthroplasty
Although rotator cuff tears are not
commonly associated with severe gle-
nohumeral arthritis, Neer et al44 coined
the term rotator cuff tear arthropathy
to distinguish glenohumeral arthritis
in conjunction with chronic massive
rotator cuff tears from more typical
types of glenohumeral arthritis. Hu-
meral head replacement is indicated
when nonsurgical modalities fail to
provide pain relief. Currently, uncon-
strained humeral head replacement
is the preferred treatment. Total shoul-
der arthroplasty in these patients is
associated with early glenoid loosen-
ing. Although the reported function-
al results of humeral head replacement
for rotator cuff tear arthropathy can
be limited, pain relief is excellent.45
Reverse shoulder arthroplasty designs
recently have been introduced to treat
patients with irreparable massive ro-
tator cuff tears. Currently, there are in-
sufficient clinical outcome data avail-
able to assess this procedure.
Summary
Chronic massive rotator cuff tears
can cause substantial shoulder pain
the rotator cuff: The timing of surgical
repair. Clin Orthop 1983;175:18-24.
6. Bigliani LU, Cordasco FA, McIlveen SJ,
Musso ES: Operative repair of massive
rotator cuff tears: Long-term results.
J Shoulder Elbow Surg 1992;1:120-130.
7. Rokito AS, Cuomo F, Gallagher MA,
Zuckerman JD: Long-term functional
outcome of repair of large and massive
chronic tears of the rotator cuff. J Bone
Joint Surg Am 1999;81:991-997.
8. Harryman DT II, Mack LA, Wang KY,
Jackins SE, Richardson ML, Matsen FA
III: Repairs of the rotator cuff: Correla-
tion of functional results with integrity
of the cuff. J Bone Joint Surg Am 1991;73:
982-989.
Journal of the American Academy of Orthopaedic Surgeons
and dysfunction. Because several
treatment options are appropriate
for different clinical scenarios, care-
ful patient evaluation and treatment
selection are critically important.
Many patients with chronic massive
rotator cuff tears can be treated non-
surgically. The goals of surgical
treatment must be considered in the
context of the individual patient and
the complexity of the procedure it-
self, as well as postoperative recov-
ery and rehabilitation. Complete
primary repair is possible in some
patients, while in others only a par-
tial repair can be achieved. Recon-
struction of the rotator cuff is most
appropriate in young, active pa-
tients for whom functional restora-
tion is important. Latissimus dorsi
muscle transfer is the preferred re-
constructive option for active indi-
viduals disabled by shoulder pain
who have weakness of elevation and
external rotation and have good del-
toid muscle strength. Most other re-
constructive options have more lim-
ited indications and have not been
shown conclusively to be superior to
débridement. Older, inactive pa-
tients, especially those with signifi-
cant medical comorbidities, are bet-
ter served by less complicated
treatment approaches that provide
reasonably predictable pain relief.
9. Gartsman GM: Massive, irreparable
tears of the rotator cuff: Results of op-
erative débridement and subacromial
decompression. J Bone Joint Surg Am
1997;79:715-721.
10. Rockwood CA Jr, Williams GR Jr, Burk-
head WZ Jr: Débridement of degener-
ative, irreparable lesions of the rotator
cuff. J Bone Joint Surg Am 1995;77:857-866.
11. Gerber C, Fuchs B, Hodler J: The results
of repair of massive tears of the rotator
cuff. J Bone Joint Surg Am 2000;82:505-515.
12. Sher JS, Uribe JW, Posada A, Murphy
BJ, Zlatkin MB: Abnormal findings on
magnetic resonance images of asymp-
tomatic shoulders. J Bone Joint Surg Am
1995;77:10-15.

13. Tempelhof S, Rupp S, Seil R: Age-
related prevalence of rotator cuff tears
in asymptomatic shoulders. J Shoulder
Elbow Surg 1999;8:296-299.
14. Nakagaki K, Ozaki J, Tomita Y, Tamai
S: Alterations in the supraspinatus mus-
cle belly with rotator cuff tearing: Eval-
uation with magnetic resonance imag-
ing. J Shoulder Elbow Surg 1994;3:88-93.
15. Blazar PE, Williams GR, Iannotti JP:
Spontaneous detachment of the deltoid
muscle origin. J Shoulder Elbow Surg
1998;7:389-392.
16. Gerber C, Krushell RJ: Isolated rupture
of the tendon of the subscapularis mus-
cle: Clinical features in 16 cases. J Bone
Joint Surg Br 1991;73:389-394.
17. Iannotti JP, Zlatkin MB, Esterhai JL,
Kressel HY, Dalinka MK, Spindler KP:
Magnetic resonance imaging of the
shoulder: Sensitivity, specificity, and
predictive value. J Bone Joint Surg Am
1991;73:17-29.
18. Teefey SA, Hasan SA, Middleton WD,
Patel M, Wright RW, Yamaguchi K: Ul-
trasonography of the rotator cuff: A
comparison of ultrasonographic and
arthroscopic findings in one hundred
consecutive cases. J Bone Joint Surg Am
2000;82:498-504.
19. Bokor DJ, Hawkins RJ, Huckell GH,
Angelo RL, Schickendantz MS: Results
of nonoperative management of full-
thickness tears of the rotator cuff. Clin
Orthop 1993;294:103-110.
20. Burkhart SS: Arthroscopic treatment of
massive rotator cuff tears: Clinical re-
sults and biomechanical rationale. Clin
Orthop 1991;267:45-56.
21. Burkhart SS, Nottage WM, Ogilvie-
Harris DJ, Kohn HS, Pachelli A: Partial
repair of irreparable rotator cuff tears.
Arthroscopy 1994;10:363-370.
22. Cofield RH: Subscapular muscle trans-
position for repair of chronic rotator
cuff tears. Surg Gynecol Obstet 1982;154:
667-672.
23. Debeyre J, Patte D, Elmelik E: Repair of
ruptures of the rotator cuff of the shoul-
Vol 11, No 5, September/October 2003
der: With a note on advancement of the
supraspinatus muscle. J Bone Joint Surg
Br 1965;47:36-42.
24. Dierickx C, Vanhoof H: Massive rotator
cuff tears treated by a deltoid muscular
inlay flap. Acta Orthop Belg 1994;60:
94-100.
25. Gazielly DF: Deltoid muscular flap
transfer for massive defects of the rota-
tor cuff, in Burkhead WZ Jr (ed): Rota-
tor Cuff Disorders. Baltimore, MD: Wil-
liams & Wilkins, 1996, pp 356-367.
26. Gerber C: Latissimus dorsi transfer for
the treatment of irreparable tears of the
rotator cuff. Clin Orthop 1992;275:152-160.
27. Miniaci A, MacLeod M: Transfer of the
latissimus dorsi muscle after failed re-
pair of a massive tear of the rotator cuff:
A two to five-year review. J Bone Joint
Surg Am 1999;81:1120-1127.
28. Watson M: Major ruptures of the rota-
tor cuff: The results of surgical repair in
89 patients. J Bone Joint Surg Br 1985;67:
618-624.
29. Zvijac JE, Levy HJ, Lemak LJ: Arthro-
scopic subacromial decompression in
the treatment of full thickness rotator
cuff tears: A 3- to 6-year follow-up.
Arthroscopy 1994;10:518-523.
30. Kempf JF, Gleyze P, Bonnomet F, et al:
A multicenter study of 210 rotator cuff
tears treated by arthroscopic acromio-
plasty. Arthroscopy 1999;15:56-66.
31. Neer CS II: Anterior acromioplasty for
the chronic impingement syndrome in
the shoulder: A preliminary report.
J Bone Joint Surg Am 1972;54:41-50.
32. Neer CS II, Marberry TA: On the disad-
vantages of radical acromionectomy.
J Bone Joint Surg Am 1981;63:416-419.
33. Colman WW, Kelkar R, Flatow EL, et al:
Abstract: The effect of anterior acromio-
plasty on rotator cuff contact: An exper-
imental and computer simulation. J Shoul-
der Elbow Surg 1996;5(suppl):S8-S9.
34. Warner JJP, Krushell RJ, Masquelet A,
Gerber C: Anatomy and relationships of
the suprascapular nerve: Anatomical con-
straints to mobilization of the supraspina-
Andrew Green, MD
tus and infraspinatus muscles in the man-
agement of massive rotator-cuff tears.
J Bone Joint Surg Am 1992;74:36-45.
35. Gerber C, Schneeberger AG, Beck M,
Schlegel U: Mechanical strength of re-
pairs of the rotator cuff. J Bone Joint Surg
Br 1994;76:371-380.
36. Björkenheim JM, Paavolainen P, Aho-
vuo J, Slätis P: Surgical repair of the ro-
tator cuff and surrounding tissues: Fac-
tors influencing the results. Clin Orthop
1988;236:148-153.
37. Jost B, Pfirrmann CWA, Gerber C: Clin-
ical outcome after structural failure of
rotator cuff repairs. J Bone Joint Surg Am
2000;82:304-314.
38. Ha’eri GB, Wiley AM: Advancement of
the supraspinatus muscle in the repair
of ruptures of the rotator cuff. J Bone
Joint Surg Am 1981;63:232-238.
39. Neviaser JS, Neviaser RJ, Neviaser TJ:
The repair of chronic massive ruptures
of the rotator cuff of the shoulder by use
of a freeze-dried rotator cuff. J Bone Joint
Surg Am 1978;60:681-684.
40. Ozaki J, Fujimoto S, Masuhara K, Tamai
S, Yoshimoto S: Reconstruction of
chronic massive rotator cuff tears with
synthetic materials. Clin Orthop 1986;
202:173-183.
41. Paavolainen P: Teres minor transfer, in
Burkhead WZ Jr (ed): Rotator Cuff Dis-
orders. Baltimore, MD: Williams &
Wilkins, 1996, pp 342-348.
42. Cofield RH, Briggs BT: Glenohumeral
arthrodesis: Operative and long-term
functional results. J Bone Joint Surg Am
1979;61:668-677.
43. Arntz CT, Matsen FA III, Jackins S: Sur-
gical management of complex irrepara-
ble rotator cuff deficiency. J Arthroplasty
1991;6:363-370.
44. Neer CS II, Craig EV, Fukuda H: Cuff-
tear arthropathy. J Bone Joint Surg Am
1983;65:1232-1244.
45. Williams GR Jr, Rockwood CA Jr:
Hemiarthroplasty in rotator cuff-
deficient shoulders. J Shoulder Elbow
Surg 1996;5:362-367.
331
 Pediatric Soft-Tissue Tumors
Kamran Aflatoon, DO, Albert J. Aboulafia, MD, FACS, Edward F. McCarthy, Jr, MD,
Frank J. Frassica, MD, and Alan M. Levine, MD

Abstract
Soft-tissue tumors in children (<18 years) are a heterogeneous group of lesions. Mass- clude the nature and location of the
es may be asymptomatic or associated with pain or discomfort. Although most le- pain, duration of symptoms, and any
sions are benign, developing an appropriate differential diagnosis requires knowl- aggravating or alleviating factors, al-
edge of the clinical and radiographic characteristics of tumors and tumorlike conditions though this information may be dif-
in children. A thorough history and physical examination, followed by appropriate ficult to elicit from a very young child.
imaging studies, when indicated, can establish a correct diagnosis and help deter- The parents may be able to provide
mine appropriate treatment recommendations. additional information, but the onset
J Am Acad Orthop Surg 2003;11:332-343 of crying or wanting to be carried
may be the only symptom.
The most common soft-tissue tu-
mors likely to present with pain in-
Soft-tissue tumors in children repre- most types of tumor other than fibro-
sent a heterogeneous group of lesions. histiocytic and lipocytic are more
The mass may be congenital, traumat- common in children than in adults7
ic, inflammatory, benign neoplastic, (Table 1). Compared with those of Dr. Aflatoon is Instructor, Department of Ortho-
or malignant neoplastic. Benign le- adults, pediatric sarcomas usually paedic Surgery, Johns Hopkins University School
sions may have limited growth, such have a higher histologic grade and of Medicine, Baltimore, MD. Dr. Aboulafia is
Assistant Professor, Department of Orthopaedic
as a lipoblastoma, or they may be ag- greater metastatic potential and are
Surgery, University of Maryland, Baltimore.
gressive, such as a desmoid tumor or more responsive to chemotherapy.9 Dr. McCarthy is Professor, Department of Pathol-
extra-abdominal fibromatosis. Ap- The clinical history is critical to ogy, and Professor, Department of Orthopaedic
proximately 6,000 new cases of soft- establish a preliminary diagnosis. Surgery, Johns Hopkins University School of Med-
tissue sarcoma are diagnosed each After a thorough history and phys- icine. Dr. Frassica is Chairman and Professor,
Department of Orthopaedic Surgery, Johns Hop-
year in adults, compared with 600 in ical examination, radiographic eval-
kins University School of Medicine. Dr. Levine
children (<18 years),1,2 or five to nine uation is used to narrow the differ- is Director, Alvin and Lois Lapidus Cancer
cases per million children annually.3 ential diagnosis. Any lesion that cannot Institute, Department of Orthopaedic Oncology,
Soft-tissue sarcomas are the fifth most be differentiated by physical exami- Sinai Hospital, Baltimore.
common pediatric malignancy (4%) nation and radiographic findings is
None of the following authors or the departments
after leukemia (31%), central nervous considered indeterminate and requires
with which they are affiliated has received anything
system tumors (18%), lymphoma a biopsy10 (Fig. 1). of value from or owns stock in a commercial com-
(12%), and neuroblastoma (8%).4 In pany or institution related directly or indirectly
a review of more than 900 benign and to the subject of this article: Dr. Aflatoon, Dr.
malignant soft-tissue tumors in pa- Aboulafia, Dr. McCarthy, and Dr. Frassica. Dr.
Clinical Presentation Levine, or the department with which he is affil-
tients in the first two decades of life,
iated, serves as a consultant to or is an employee
Kransdorf5,6 found that approximate- Typically, the child presents after dis- of Wright Medical.
ly 30% were vascular, 15% were neu- covery of a palpable mass by the par-
rogenic, and 14% were fibroblastic ents or pediatrician. A recent trauma Reprint requests: Dr. Aflatoon, c/o Elaine P.
and myofibroblastic or were myogen- may have brought the mass to the Henze, Medical Editor, Department of Ortho-
paedic Surgery, Johns Hopkins Bayview Medical
ic. Certain soft-tissue tumors occur parents’ attention. There may be as-
Center, Room A672, 4940 Eastern Avenue,
more commonly in specific age sociated symptoms, such as pain, Baltimore, MD 21224-2780.
groups and some in certain locations. joint contracture, or impaired func-
For instance, malignant fibrous his- tion. When a soft-tissue mass is symp- Copyright 2003 by the American Academy of
tiocytoma is the most frequent soft- tomatic, pain is the most common Orthopaedic Surgeons.
tissue sarcoma in adults, whereas complaint. The history should in-

332 Journal of the American Academy of Orthopaedic Surgeons

 Kamran Aflatoon, DO, et al

Table 1
Relative Incidence of the Most Common Neoplastic Soft-Tissue Tumors4,7,8

Occurrence (%)

Tumor Type Examples Children Adults

Vascular Hemangioma, arteriovenous malformation, lymphangioma, 29 9

glomus tumor, angiomatosis, hemangiopericytoma
Neurogenic Neurofibroma, malignant peripheral nerve sheath, schwannoma, 15 9
primitive neuroectodermal, clear cell sarcoma, neurothekeoma
Myogenic Rhabdomyosarcoma, leiomyosarcoma, leiomyoma, angiomyoma, 14 5
epithelioid leiomyoma
Fibroblastic/ Superficial fibromatosis, deep fibromatosis, fibrosarcoma, fibrous 12 7
myofibroblastic hamartoma
Fibrohistiocytic Fibrous histiocytoma, reticulohistiocytoma, xanthoma, malignant 12 17
fibrous histiocytoma, dermatofibrosarcoma protuberans,
angiomatoid fibrous histiocytoma
Lipocytic Lipoma, lipoblastoma, angiolipoma, spindle-cell lipoma, 6 16
myolipoma, myelolipoma, atypical lipoma, liposarcoma
Other — 12 37

clude hemangiomas and glomus tu- when the extremity is in a dependent volve the gastrocnemius muscle, the
mors. Patients with hemangiomas position. Dilatation of venous cavities child can present with toe walking.
typically complain of pain and dis- can cause muscular cramping and Symptoms may be relieved after the
comfort after a period of standing or discomfort. When hemangiomas in- extremity is elevated or put at rest.
These indications can be especially
helpful in identifying hemangiomas
Child with a soft-tissue mass
involving the plantar aspect of the
foot (Fig. 2).
History and physical examination
The character of the pain also may
be helpful. Paroxysms of pain radi-
Radiography/MRI
ating from a lesion on exposure to
cold temperatures suggest a glomus
tumor.11 Unlike glomus tumors in
adults, those in children may be mul-
Determinate lesion Indeterminate lesion tifocal. Pediatric glomus tumors com-
(eg, hemangioma,
lipoma, ganglion cyst) monly are in the superficial soft tis-
Biopsy

Treatment based
on symptoms

Malignant lesion Benign lesion

Superficial lesion Deep lesion, lesion close to neu- Treatment based

or size ≤5 cm rovascular bundles, or size >5 cm on symptoms

Neoadjuvant therapy (preoperative

Wide surgical chemotherapy if tumor is sensitive [eg, Wide surgical
reconstruction rhabdomyosarcoma] and/or radiation therapy resection Figure 2 An 18-year-old man who had mul-
[eg, tumor is close to neurovascular bundles]) tiple operations on the medial and plantar as-
pects of the foot for hemangioma presented
with recurrence of the hemangioma and
Figure 1 Presentation, diagnosis, and management of soft-tissue tumors. pain.

Vol 11, No 5, September/October 2003 333

Pediatric Soft-Tissue Tumors
sues of the arms and legs rather than
the subungual region, as in adults.11
A history of trauma may be con-
fusing rather than defining. When
trauma brings a previously unrecog-
nized mass to attention, diagnosis
may be delayed because the mass is
thought to be posttraumatic in origin
(eg, hematoma, myositis ossificans,
retained foreign body) (Figs. 3 and 4).
The opposite also may be true: a post-
traumatic lesion may simulate a ma-
lignant process and lead to unneces-
sary anxiety and diagnostic studies
if a patient fails to recall the causative
traumatic episode.
Physical Examination
The physical examination should in-
clude an accurate assessment of the
soft-tissue mass, including size, char-
acter, depth, and mobility. Lesions
that are firm, deep (ie, below the fas-
cia), and >5 cm in diameter are sus-
picious for malignancy. The shape
Figure 3 Lateral radiograph showing a firm
mass in the right upper extremity of a 10-year-
old boy. The parents could not recall any spe-
cific trauma. The mass was diagnosed as myo-
sitis ossificans circumscripta after computed
tomography revealed a zonal calcification in
the periphery and no calcifications in the cen-
ter of the lesion.
334
and girth of the extremity should be
measured, recorded, and compared
with those of the opposite side. Le-
sions such as lymphangioma or hem-
angioma may present with asymmet-
ric enlargement of the involved
extremity.12,13 The regional and ma-
jor lymph nodes should be palpated
because rhabdomyosarcoma, syn-
ovial sarcoma, and epithelial sarco-
ma potentially can metastasize
through the lymphatic system. Range
of motion of the adjacent joints should
be recorded because flexion contrac-
ture may be secondary to lesions such
as desmoid tumor or intramuscular
hemangioma.
Some common lesions in pediat-
ric patients have characteristic phys-
ical findings. These lesions often do
not require additional imaging for di-
agnosis and may be treated symp-
tomatically. Superficial hemangio-
mas frequently present with blue skin Figure 4 Two-centimeter mass on the right
lower extremity of a 5-year-old boy with no
discoloration and subcutaneous history of trauma. Fat-saturated T2-weighted
swelling. The swollen area increases spin echo coronal MRI scan demonstrated a
in size when the extremity is in a de- high signal intensity in the mass. Excisional
biopsy revealed a retained thorn.
pendent position because it fills with
blood. The patient may complain of
pain with ambulation secondary to Neurofibromas in individuals with
engorgement of the hemangioma. von Recklinghausen’s neurofibroma-
Subcutaneous lipomas account tosis are multiple, and growth of the
for 15% of adipose tumors in the lesions may be accelerated during pu-
first two decades of life.14 Typically, berty or pregnancy16 (Figs. 5 and 6).
the mass is doughy and nonmobile The café-au-lait spots in prepubertal
and does not cause discomfort. The patients are smaller (approximately
most common site of involvement is 5 mm) than those in adults (approx-
the trunk. Subcutaneous lipomas imately 15 mm).17 MRI of neurofibro-
usually do not grow, but when they mas is not always required unless the
do, magnetic resonance imaging patient complains of pain or the mass
(MRI) can help determine whether a is enlarging.
lesion is a lipoma. Deep lesions other than myositis
Ganglia are not uncommon in pe- ossificans require additional MRI
diatric patients; approximately 10% scans. In pediatric patients, rhab-
to 15% of all ganglia occur in the first domyosarcoma (which accounts for
two decades of life.15 The mass is a 45% to 50% of all malignant soft-tissue
well-defined, firm nodule composed sarcomas in children18), synovial sar-
of mucoid cyst surrounded by a fi- coma, fibrous tumors, neurogenic tu-
brous capsule. Ganglia usually are lo- mors, fibrohistiocytic tumors, and des-
cated adjacent to or are attached to a moid tumors (or extra-abdominal
joint capsule or tendon sheath. The fibromatosis) are deep to fascia and
dorsomedial aspect of the hand is the usually have a firm consistency.
most common site. Diagnosis often Malignant lesions are categorized
can be confirmed by aspiration. into three grades (Table 2). Some soft-
Journal of the American Academy of Orthopaedic Surgeons
 Kamran Aflatoon, DO, et al

Table 2
Grading for Pediatric Nonrhabdomyosarcoma Soft-Tissue Sarcoma19

Grade Characteristic

1 Myxoid and well-differentiated liposarcoma

Dermatofibrosarcoma protuberans
Well-differentiated and infantile fibrosarcoma
Well-differentiated and infantile hemangiopericytoma
Extraskeletal myxoid chondrosarcoma
Angiomatoid (malignant) fibrous histiocytoma
Well-differentiated malignant peripheral nerve sheath tumor
2 Sarcomas (neither grade 1 nor 3)
<15% area of necrosis
<5 mitoses per high-power field
No substantial nuclear atypia
Not markedly cellular
3 Pleomorphic or round cell liposarcoma
Mesenchymal chondrosarcoma
Extraskeletal osteosarcoma
Alveolar soft part sarcoma
Malignant Triton tumor
Figure 5 Café-au-lait spots in a 16-year-old
boy with neurofibromatosis.
Marked atypia and cellularity

tissue malignant tumors may be su-
perficial, readily palpable, and small,
but they should not be ignored. The
most common are dermatofibrosar-
coma protuberans (Fig. 7), epithelioid
sarcoma, plexiform fibrohistiocytic tu-
mor, and angiomatoid fibrous histio-
cytoma.
Because of their activity level and
tendency to play outdoors, children
are prone to penetration injuries from
sharp objects such as glass, wood, or
thorns. Incomplete removal of a for-
eign body may leave material behind
to induce future granulomatous re-
action or abscess formation.20 Foreign-
body granuloma may lack the com-
mon findings of an infectious process
and clinically and radiographically
may mimic a soft-tissue sarcoma.20,21
Radiographic Examination
Plain radiographs and MRI are the
most commonly used modalities for
the evaluation of soft-tissue tumors.
Other potentially useful imaging
methods include computed tomogra-
phy (CT) and ultrasound.
teroposterior and lateral). To accen-
tuate soft-tissue details, the technician
should obtain radiographs using a
low-kilovolts peak (<50 kVp) tech-
nique (Fig. 8).
The radiodensity of the lesion
(ranging from radiolucency to radio-
pacity) provides important clues. Deep
lipoblastoma or other fatty tumors
may present as a radiolucent mass.
Air is rarely seen in soft-tissue tumors,
except in cases of abscess. Radiodense
foreign bodies within the soft tissues

Figure 6 Coronal T2-weighted MRI scan of
the knee demonstrating neurofibroma along
the course of the peroneal nerve (arrow) in a
14-year-old boy.
Plain Radiographs
Plain radiographs are critical in im-
aging any soft-tissue mass and are the
first step in formulating the differen-
tial diagnosis. In some cases, radio-
graphs provide more useful informa-
tion than do advanced imaging
techniques. They may occasionally re-
veal an underlying osseous process Figure 7 Axial T1-weighted MRI scan of a
(eg, osteochondroma, osteomyelitis). subcutaneous, nonpalpable mass in the left
Radiographs should be obtained in leg of an 18-year-old man. Biopsy revealed a
dermatofibrosarcoma protuberans.
two orthogonal planes (usually an-

Vol 11, No 5, September/October 2003 335

Pediatric Soft-Tissue Tumors
Figure 8 Anteroposterior low-kVp radio-
graph showing a soft-tissue mass (arrow)
along the ulnar aspect of the right hand of a
10-year-old girl.
can be easily detected on plain radio-
graphs. Other findings include ma-
trix mineralization, periosteal reaction,
or cortical erosions. The calcification
or mineralization pattern should be
carefully studied. Chondroid matrix
on plain radiographs appears as rings
and arcs, whereas osteoid matrix pre-
sents as a cloudlike density (eg, ex-
traskeletal osteosarcoma) or mature
trabecular bone (eg, myositis ossifi-
cans), based on rate of growth. Radio-
graphically, myositis ossificans pre-
sents with a well-developed zonal
calcification pattern. In contrast with
extraskeletal osteosarcoma, the calci-
fication is more mature at the periph-
ery than at the center of the lesion.
Multiple small spotty calcifications are
seen in 15% to 20% of synovial sar-
comas.22 The presence of phleboliths
on plain radiographs is characteris-
tic of vascular lesions, such as heman-
gioma13 (Fig. 9).
Ossification in the subcutaneous
fat, muscles, tendons, and ligaments,
forming a network of interconnecting
bridges between adjacent bones and
joints, is diagnostic of fibrodysplasia
ossificans progressiva.23 This hered-
itary disease presents with a slowly
progressive calcification of fibroblas-
tic tissue that principally affects chil-
336
dren younger than 10 years. The os-
sification leads to eventual ankylosis
of the involved joints.
Occasionally, a soft-tissue tumor
may erode into the bone and be con-
fused for a primary bone tumor. How-
ever, a primary soft-tissue tumor pre-
sents with the epicenter outside the
bone, whereas a primary bone tumor
has an epicenter inside the medullary
cavity. Periosteal new bone formation Figure 9 Lateral radiograph showing mul-
tiple phleboliths in the upper extremity of a
may or may not be present. Promi- 14-year-old boy with hemangiomas.
nent periosteal reaction is common-
ly encountered in hypervascular soft-
tissue neoplasms.24 Solid periosteal weighted fat-suppression sequences
reaction is more characteristic of a be- can be highly specific. For example,
nign lesion, whereas interrupted peri- using fat-suppression sequences, be-
osteal reaction is associated with ma- nign lipoblastoma, the most common
lignant soft-tissue tumors invading the fatty tumor in children, can be iden-
bone. Radiographic evidence of ini- tified with a high degree of confi-
tial erosion of the outer cortex of the dence because of signal attenuation:
bone, followed by beveling of the cor- fat within the tumor has the same
tex toward the medullary cavity, in- characteristics as subcutaneous fat.
dicates a soft-tissue lesion. Malignant lesions (eg, myxoid) and
infection tend to have low signal in-
Magnetic Resonance Imaging tensity on T1-weighted images and
MRI is the most useful modality high signal intensity on T2-weighted
for imaging soft-tissue tumors because images. One distinctive feature of in-
it can visualize the mass in multiple fection is the involvement of multi-
orthogonal views. It also provides a ple fascial planes. Occasionally, ma-
detailed anatomic picture of the mass, lignant tumors contain areas of
including location and the relation- hemorrhage, which make the lesion
ship to anatomic structures such as appear heterogeneous.
nerves, arteries, fascia, and periosteum. Tumors with low signal intensity
Soft-tissue masses vary in size and on T1- and T2-weighted images, such
may arise in the skin, subcutis, mus- as a desmoid tumor, are predominant-
cle, or other deep soft tissues. In con-
trast with malignant bone tumors,
most soft-tissue tumors, whether be-
nign or malignant, tend to appear
round or oval on MRI (Fig. 10); ex-
ceptions include extra-abdominal des-
moid tumors and dermatofibrosar-
coma protuberans. Both benign and
malignant soft-tissue tumors are en-
capsulated, grow by centrifugal ex-
tension, and tend to respect anatom-
ic planes as they grow. This organized
pattern may appear misleadingly be-
nign to those unfamiliar with soft-
tissue tumors. Figure 10 Axial T2-weighted MRI scan dem-
Although MRI does not provide onstrating a spherical mass (fibrosarcoma [ar-
histologic diagnosis, the characteris- row]) in the medial aspect of the thigh of a
16-year-old girl.
tic features of a lesion on T1- and T2-
Journal of the American Academy of Orthopaedic Surgeons

ly fibrous (Fig. 11). A desmoid tumor
is a benign, nonencapsulated tumor
that arises from the musculoapo-
neurotic structures and tends to ex-
tend along fascial planes. It has a lo-
cal infiltrative growth pattern without
potential for distant metastasis. The
appearance of desmoid tumors on
MRI is that of somewhat ill-defined
intramuscular soft-tissue masses. Bony
erosion and/or neurovascular encase-
ment are common features. Desmoid
tumors usually affect the muscles of
the shoulder and hip. Such masses are
isointense and do not demonstrate
marked heterogeneity on T1-weighted
images. However, on T2-weighted se-
quences, the tumors appear hyperin-
tense with variable degrees of hetero-
geneity and fibrosis.25 T2-weighted
images are superior to T1-weighted
images for demonstrating the ill-
defined borders and infiltrative nature
of desmoid tumors.
The use of contrast medium such
as gadolinium is helpful for differen-
tiating cystic lesions from lesions that
appear to be bright on T2-weighted
images but have solid components
(ie, tumor) within larger cystic areas.
Malignant lesions are vascular and
tend to enhance with the use of in-
travenous contrast. A benign tumor,
such as a hemangioma, cyst, or an ab-
Figure 11 A, A 16-year-old boy with a desmoid tumor involving the proximal tibia. B, On the T1-weighted axial MRI scan, the desmoid
tumor appears dark. C, On the T2-weighted, fat-suppressed axial image, the tumor is mixed in signal intensity, with areas of dark and bright
signal.
Vol 11, No 5, September/October 2003
scess, characteristically shows con-
trast enhancement of the peripheral
rim but no enhancement of the cav-
ity because of its internal dysvascu-
larity.
Occasionally, the location of the
mass, which may be better defined by
MRI than by physical examination,
may provide diagnostic clues. For in-
stance, nodular masses along the ten-
don sheath with low signal intensity
on both T1- and T2-weighted images
suggest tenosynovial pigmented vil-
lonodular synovitis of the tendon
sheath, whereas an intra-articular pe-
dunculated mass with the same sig-
nal characteristics suggests intra-
articular pigmented villonodular
synovitis. Masses with fusiform ends
that appear to be in continuity with
a nerve are either schwannomas or
neurofibromas. Table 3 lists selected
soft-tissue masses, their presenting
complaints, and radiographic find-
ings.
Biopsy
The most important factor determin-
ing the need for biopsy is certainty
of the diagnosis after completion of
the physical examination and imag-
ing studies. Lesions such as lipoma,
Kamran Aflatoon, DO, et al
hemangioma, arteriovenous malfor-
mations, ganglion, myositis ossificans,
plantar fibromatosis, and fibromato-
sis colli are considered determinate
lesions and do not require biopsy. All
other lesions that cannot be accurate-
ly categorized and diagnosed are
deemed indeterminate and require a
biopsy.
Preoperative imaging related to
the tumor should be completed be-
fore biopsy because postoperative
changes impair subsequent MRI eval-
uation of the mass. In addition, pre-
operative MRI helps the surgeon
evaluate the extent of the tumor, lo-
cate the optimal site for biopsy by
avoiding areas that appear necrotic,
and plan future surgery.
The surgeon should be familiar
with the indications for, as well as the
advantages and disadvantages of, the
various biopsy techniques. These
techniques include needle (fine and
core), excisional, and incisional. Inci-
sional biopsy once was the preferred
method.26-28 However, compared
with the incisional technique, needle
biopsy has several advantages. It can
be done in an outpatient setting with
local anesthetic, is associated with a
reduced incidence of infection and
wound complications, affords the op-
portunity to sample various portions
337
Pediatric Soft-Tissue Tumors

of the tumor, and may have less po-

Table 3
tential for local contamination by tu-
Selected Common Tumors
mor cells.
Excisional biopsy is rarely indicat-
Tumor Type Presentation
ed for indeterminate lesions but may
be considered for small (<1 to 2 cm) Benign neoplastic
superficial masses located in anatom- Superficial Superficial blue mass, occasional pain
ic areas amenable to adequate exci- hemangioma
sion with tumor-free margins. The Deep hemangioma Muscular pain, enlarging mass
characteristics of the lesion and the
extent of infiltration should be eval- Digital fibromatosis Nonpainful mass, usually dorsal or on the sides
of the long, ring, small digits; first year of life
uated based on MRI scans before ex-
cisional biopsy. Fibrous hamartoma Nonpainful, small, rapidly growing mass in
of infancy dermis and subcutis
An incisional biopsy may be done
Schwannoma Slowly growing mass, no pain except with
to obtain an adequate sample of the
larger ones
tumor, but several principles should
Glomus Paroxysm of pain radiating from a lesion,
be followed. A tourniquet can help cold sensitive
decrease blood loss during biopsy
Neurofibroma Slowly growing nonpainful mass
and improve visualization of the sur-
gical field. However, before inflation Infantile fibromatosis Patients usually age <8 years; nonpainful mass,
of the tourniquet, the extremity (desmoid) rapid growth
should be elevated above heart level
to diminish its blood supply. Use of Fibrous histiocytoma Solitary slowly growing nodule
an Esmarch tourniquet, which would
compress the tumor, is discouraged. Myofibroma (MF) and MF: single mass in the dermis or subcutaneous,
A transverse incision for biopsy al- myofibromatosis <2 cm in diameter, nonpainful; MFT: multiple
ways should be avoided because it in- (MFT) lesions
creases the extent of tissue loss at the Fibromatosis colli Nodule in the sternocleidomastoid (SCM) muscle
presenting as torticollis; rapid growth in the
time of definitive resection and may
first few months, then stationary
lead to contamination of unaffected
Lymphangioma Extremity heaviness, asymmetric enlargement
muscle compartments. Creation of of one extremity
flaps at the time of biopsy should be Nonmalignant
avoided because more extensive dis- Foreign body Gradually growing mass in upper or lower
section allows microscopic tumor granuloma extremity
cells to spread to previously uncon- Heterotopic Growing mass in an active child, may provide
taminated areas. Necrotic areas ossification history of trauma
should not be entered during biopsy Malignant
because results would be of no diag- Dermatofibrosarcoma Slow and persistent growth over a long period
nostic value. Damage to the local neu- protuberans of time, most commonly in the trunk and
rovascular structures may be prevent- proximal extremities
ed by careful evaluation of the MRI Malignant peripheral Patients with neurofibromatosis 1 at increased
scan. To prevent hematoma formation nerve sheath tumor risk, present with a painful neurofibroma
and tumor spread, meticulous hemo- Synovial sarcoma Palpable mass in extremity (generally
asymptomatic)
stasis should be obtained at the con-
clusion of the biopsy before closure. Infantile fibrosarcoma Rapidly growing mass in the first year of life,
mostly in extremities
Frequently, a drain placed in line with
Rhabdomyosarcoma Upper or lower extremity, rapidly growing non-
the incision is necessary.
painful mass, most common in lower extremity
For all incisional biopsies, intraop-
erative frozen section should be done
to ensure adequate quality of the tis-
sue sample, which is crucial in cases same procedure based on the result ing and genetic analysis, may require
with substantial areas of necrosis. of frozen section alone. Complex cas- a larger than usual amount of spec-
However, the definitive surgical pro- es requiring more extensive studies, imen and occasionally necessitate an
cedure usually is not done during the such as immunohistochemical stain- open biopsy.

338 Journal of the American Academy of Orthopaedic Surgeons


Physical Findings
Benign neoplastic
Compressible mass with blue skin discoloration
None
Rarely >2 cm, hemispheric or dome-shaped, with shiny
skin surface
Freely movable, rarely fixed to the fascia or muscle
Freely movable except in the long axis of the nerve
Lesion found in the subcutis of the extremity
Usually multiple, may present with superficial
and deep lesions
Deeply seated, poorly circumscribed solitary mass in
muscle or fascia; may infiltrate muscles and nerves as
it grows and cause pain and functional disturbance
Elevated or pedunculated lesions <2 cm; overlying
skin is red to red-brown secondary to hemosiderin
MFT has multiple nodules and involvement of muscle,
bone, and internal organs (lung, heart, gastrointestinal
system)
Firm mass in the SCM close to the sternal or
clavicular portion; should not be confused with
acquired (traumatic) torticollis
Asymmetric enlargement of the extremity
Nonmalignant
Occasionally painful, may present with erythema
Hard mass most commonly seen in the anterior thigh
Malignant
Firm plaquelike lesion with red to blue discoloration
of the surrounding skin
Large, deeply situated mass commonly associated
with major nerve trunk
Deep mass, close proximity to major joints
Size from a few to >20 cm
Solid deep mass; lymph nodes should be inspected
for evidence of metastasis
Minimally invasive techniques in- ple aspirations through the tumor us-
clude fine-needle aspiration biopsy ing a 23- to 27-gauge needle. It should
and core needle biopsy. Fine-needle be done only in a center that has pa-
aspiration biopsy is done by multi- thologists with extensive experience
Vol 11, No 5, September/October 2003
Kamran Aflatoon, DO, et al
Radiographic Findings
Bright on T1- and T2-weighted MRI; may have
phlebolith on radiograph
Bright on T1- and T2-weighted MRI; may have
phlebolith on radiograph
Low signal on T1- and T2-weighted MRI
Low signal on T1- and T2-weighted MRI
Low signal on T1-weighted MRI, high signal
on T2-weighted MRI
High signal on T1- and T2-weighted MRI
Low signal on T1-weighted MRI, high signal on
T2-weighted MRI
Inhomogeneous signal intensity on MRI; poor margin-
ation; involvement of neurovascular bundles
Low signal on T1- and T2-weighted MRI
Soft-tissue swelling; multiple circumscribed lytic bony
lesions, some marginal sclerosis; no cortical violation
Low signal on T1- and T2-weighted MRI
Multiloculated osteolytic lesions in bone; lymphangio-
gram localizes blind pouch
Some materials are radiodense and easily seen on radio-
graphs; ultrasound may be helpful in some cases
Zonal pattern of calcification with more peripheral
mature bone compared with center of mass
When subcutaneous, MRI shows more tissue involvement
than clinical picture
Low signal on T1-weighted MRI, high signal on
T2-weighted MRI; MRI can identify nerve trunk
Low signal intensity on T1-weighted MRI, high signal
on T2-weighted MRI
Inhomogeneous low signal on T1-weighted MRI,
high signal on T2-weighted MRI
Inhomogeneous low signal on T1-weighted MRI,
high signal on T2-weighted MRI
with the technique.29,30 Under even
optimal circumstances, fine-needle
aspiration is less effective than core
biopsy, especially for fibrous lesions
339
Pediatric Soft-Tissue Tumors
and those that require accurate grad-
ing to determine the best treatment.
The larger core needle (1 mm in di-
ameter) provides a sliver of tissue for
microscopic analysis. In an outpatient
setting with the patient under local
anesthesia, the needle is passed per-
cutaneously several times through the
tumor. To improve the accuracy of the
sample, necrotic areas should be
avoided. Such areas can be identified
before biopsy by careful evaluation
of the MRI scan. CT or ultrasound guid-
ance can be used for deep lesions. Usu-
ally there is sufficient tissue for spe-
cific histochemical staining and grading.
Diagnostic accuracy approaches 90%
with core needle biopsy.31
Staging
A patient with a soft-tissue sarcoma
should be staged before initial treat-
ment. The presence or absence of
skeletal or visceral metastases should
be assessed during this evaluation
phase. Because the lungs are the most
common site for metastasis, CT scans
of the chest are essential. Tumors such
as rhabdomyosarcoma, epithelial sar-
coma, or synovial sarcoma may me-
tastasize to the regional lymph nodes.
Suspicious enlargement of the local
and regional nodes should be eval-
uated clinically and with MRI or CT
scans. However, even when the clin-
ical examination and radiographic
findings are negative for lymphatic
metastasis, biopsy may reveal the
presence of micrometastasis to the re-
gional lymph nodes.32,33
Jager et al34 found a 7% incidence
of skeletal metastasis in patients with
soft-tissue sarcoma (8/109 patients),
and Kim et al35 reported a 4% inci-
dence (1/23 patients). The presence
of skeletal metastasis should be ini-
tially investigated with bone scan
and plain radiographs and, when
indicated, with MRI.36 Bone marrow
aspiration is indicated as part of the
staging for patients with rhabdomyo-
sarcoma.31
340
Management
Management of a soft-tissue tumor is
based on its natural history, tenden-
cy for local recurrence, and potential
for or presence of distant metastasis.
The choice of nonsurgical or surgical
approach is dictated by tumor type.
The appearance of lipoma on MRI
(the same signal characteristics as
subcutaneous fat on all sequences) is
so characteristic that, in most cases,
biopsy is not needed to establish the
diagnosis with reasonable certain-
ty.35,37 Although the characteristic ap-
pearance of lipomas is well described,
MRI also is useful in identifying less
common lipomalike variants that are
unique to children younger than 5
years and that are predominantly
seen as fat in MRI but exhibit vari-
able amounts of inhomogeneity. Wide
local excision is used to treat lipoblas-
toma and lipoblastomatosis.
Asymptomatic hemangiomas can
simplybeobserved.Thosethataresymp-
tomatic (pain or increasing deformity)
can be treated by surgical excision if
function will not be severely impaired
or by embolization and/or direct in-
jection in areas in which surgical ex-
cision will have notable morbidity.
Some tumors, such as hemangio-
mas or arteriovenous malformations,
can be managed with minimally in-
vasive treatment. Injection of ethanol
into the hemangioma cavity may oblit-
erate the tumor mass, although fail-
ure (ie, local recurrence) rates may be
more than 50%.38 An arteriovenous mal-
formation may be managed by plac-
ing coils in the major feeding vessels.
Marginal resection may be used to
treat most benign tumors, such as a
benign neurofibroma or lipoblas-
toma. In this type of resection, the
tumor is removed but the inflamma-
tory zone around the tumor, with po-
tential microscopic tumor, is left. For
malignant tumors (eg, rhabdomyo-
sarcoma, synovial sarcoma), the sur-
gical approach is to achieve wide re-
section, which involves removal of the
tumor, its reactive zone, and a rim of
Journal of the American Academy of Orthopaedic Surgeons
healthy tissue around it. However, some
malignant soft-tissue tumors in chil-
dren require multimodality therapy.
Desmoid Tumor
Desmoid tumors are not truly ma-
lignant because they do not exhibit
distant metastases; however, patients
frequently have local recurrences that
may be difficult to manage. There is
no consensus regarding preferred
treatment. Several protocols have been
recommended, including radiation,
surgery, chemotherapy, benign neglect,
and combinations thereof.39-42 The
most commonly used therapies are
surgical resection and radiation. Nei-
ther Merchant et al43 nor Spiegel et
al44 recommended radiation therapy
alone to treat desmoid tumors in pe-
diatric patients. Spiegel et al44 showed
a better result with wide surgical re-
section and adjuvant chemotherapy
than with radiation therapy alone.
Similarly, Merchant et al43 showed 77%
failure with radiation therapy alone
(10/13 patients). Surgical resection is
done with the intent of achieving a
wide or, in some cases, a radical mar-
gin. With negative margins, local con-
trol is improved with adjuvant radi-
ation. However, high rates of local
recurrence, even in the setting of neg-
ative surgical margins and the use of
adjuvant radiation, have created in-
terest in developing nonsurgical and
non–radiation-based treatments. This
is especially important for skeletally
immature patients because radiation
may be associated with risks of growth
disturbance, contractures, and second-
ary malignancies.44
Alternative therapies include ob-
servation, hormones, hormone an-
tagonists, cyclooxygenase inhibitors,
interferon alpha, and cytotoxic che-
motherapy. The use of vinblastine
and methotrexate has been reported
to control disease in most children
with desmoid tumors who were not
amenable to surgical or radiation
treatment.45 Future treatment may fo-
cus on the molecular basis of fibro-
blast mitogenesis.46

Table 4
Statistically Significant Predictors of Failure-Free Survival32
Tumor Characteristic Measurement Statistical Difference
Number of lymph nodes 0 versus 1 or more P < 0.001
affected
Size <5 cm versus ≥5 cm P = 0.04
Invasiveness Superficial or single ver- P < 0.001
sus deep or multiple
Synovial Sarcoma sarcoma has been determined to be
Synovial sarcoma is the second a chromosomal translocation t(2;13)
most common malignant soft-tissue (q35;q14) involving the PAX3 gene on
tumor in pediatric patients.46,47 Wide band 2q35 and the FKHR gene on
surgical resection with adjuvant ra- band 13q14.54,55
diation has a 5-year failure-free sur- The treatment protocol for a pa-
vival rate of approximately 75%.47-49 tient with rhabdomyosarcoma is
The usefulness of chemotherapy in based on several factors, the most im-
conjunction with surgery and radia- portant of which is presence or ab-
tion therapy for long-term survival sence of distant metastasis. Current
remains controversial.47,49,50 treatment of children with rhabdo-
myosarcoma that has metastasized
Rhabdomyosarcoma beyond the regional lymph nodes in-
Rhabdomyosarcoma is more com- cludes multiagent chemotherapy and
mon in the lower extremity (approx- radiation therapy.32,56-59 Such patients
imately 19% of all cases) than upper have a 5-year survival rate, ranging
extremity32,51 and has multiple histo- from 20% to 30%,32,56-59 based on the
logic subtypes with distinct prognoses. histologic subtype of the tumor.
The occurrence of the two subtypes The Intergroup Rhabdomyosar-
most commonly seen in the extrem- coma Study IV has shown that the fol-
ities, embryonal and alveolar, is age lowing factors are statistically signif-
dependent. Embryonal rhabdomyo- icant predictors of failure-free survival
sarcoma occurs more commonly in in patients without distant metasta-
children, whereas alveolar rhab- sis: presence of a tumor in local lymph
domyosarcoma is more commonly nodes, tumor size, and invasiveness
found in adults.52 Another subtype, of the tumor (ie, involvement of the
pleomorphic rhabdomyosarcoma, is neurovascular bundles, multiple com-
more common in adults aged 21 to partments)32 (Table 4). Patients with
81 years (mean age, 51 years).53 nonmetastatic rhabdomyosarcoma
Gene alteration for most embryo- have a mean 3-year overall failure-free
nal rhabdomyosarcoma has been survival rate of 71% when treated
determined to be a deletion of chro- with multimodality therapy, includ-
mosome band 11p15.5. The gene al- ing wide surgical resection, intensive
teration in most alveolar rhabdomyo- multiagent chemotherapy, and radi-
Vol 11, No 5, September/October 2003
Kamran Aflatoon, DO, et al
ation therapy.56,60 Radiation therapy
is recommended for patients with re-
sidual tumor at the site of resection,
involvement of local lymphatics, and
alveolar or undifferentiated rhab-
domyosarcoma.61,62
Summary
Evaluation of the pediatric patient
with a soft-tissue mass can be chal-
lenging. To formulate a differential di-
agnosis and, ultimately, diagnose a
presenting lesion, the clinician should
have an organized and systematic ap-
proach to the evaluation. There are
five basic categories of soft-tissue
masses: congenital, traumatic, inflam-
matory, benign neoplastic, and malig-
nant neoplastic. Some lesions may re-
semble each other, and the diagnosis
may be indeterminate without doing
a biopsy. Before biopsy, the clinician
should obtain all relevant clinical and
radiographic diagnostic information.
To optimize patient outcome, biopsy
should follow all the principles of on-
cologic surgery.
Most traumatic and benign soft-
tissue lesions can be managed by ei-
ther observation or marginal excision.
Survival rates for patients with ma-
lignant sarcomas continue to improve
because of advances in chemothera-
py and use of multimodality therapy
(eg, radiation therapy, surgical resec-
tion). To optimize their management
and survival, patients with malignant
sarcomas should be treated at centers
with special interest in these condi-
tions as well as with the resources to
render multimodality therapy and
provide appropriate follow-up for de-
tecting recurrence or metastasis after
resection.
341
Pediatric Soft-Tissue Tumors
References
1. Gurney JG, Severson RK, Davis S,
Robison LL: Incidence of cancer in chil-
dren in the United States: Sex-, race-,
and 1-year age-specific rates by histo-
logic type. Cancer 1995;75:2186-2195.
2. Miller RW, Young JL Jr, Novakovic B:
Childhood cancer. Cancer 1995;75
(1 suppl):395-405.
3. Stiller CA, Parkin DM: International
variations in the incidence of childhood
soft-tissue sarcomas. Paediatr Perinat
Epidemiol 1994;8:107-119.
4. Hayani A, Mahoney DH Jr, Hawkins
HK, Steuber CP, Hurwitz R, Fernbach
DJ: Soft-tissue sarcomas other than
rhabdomyosarcoma in children. Med
Pediatr Oncol 1992;20:114-118.
5. Kransdorf MJ: Malignant soft-tissue tu-
mors in a large referral population: Dis-
tribution of diagnoses by age, sex, and
location. AJR Am J Roentgenol 1995;164:
129-134.
6. Kransdorf MJ: Benign soft-tissue tu-
mors in a large referral population: Dis-
tribution of specific diagnoses by age,
sex, and location. AJR Am J Roentgenol
1995;164:395-402.
7. Harms D: Soft tissue sarcomas in the
Kiel Pediatric Tumor Registry. Curr Top
Pathol 1995;89:31-45.
8. Dillon P, Maurer H, Jenkins J, et al: A
prospective study of nonrhabdomyo-
sarcoma soft tissue sarcomas in the pe-
diatric age group. J Pediatr Surg 1992;27:
241-245.
9. Felgenhauer J, Hawkins D, Pendergrass
T, Lindsley K, Conrad EU III, Miser JS:
Very intensive, short-term chemothera-
py for children and adolescents with met-
astatic sarcomas. Med Pediatr Oncol 2000;
34:29-38.
10. Frassica FJ, Thompson RC Jr: Evalua-
tion, diagnosis, and classification of be-
nign soft-tissue tumors. Instr Course
Lect 1996;45:447-460.
11. Tsuneyoshi M, Enjoji M: Glomus tumor:
A clinicopathologic and electron micro-
scopic study. Cancer 1982;50:1601-1607.
12. Levine C: Primary disorders of the lym-
phatic vessels: A unified concept. J Pe-
diatr Surg 1989;24:233-240.
13. Weber TR, Connors RH, Tracy TF Jr,
Bailey PV: Complex hemangiomas of
infants and children: Individualized
management in 22 cases. Arch Surg
1990;125:1017-1021.
14. Coffin CM, Williams RA: Congenital lip-
oblastoma of the hand. Pediatr Pathol
1992;12:857-864.
15. Nelson CL, Sawmiller S, Phalen GS:
Ganglions of the wrist and hand. J Bone
342
Joint Surg Am 1972;54:1459-1464.
16. Coffin CM, Dehner LP: Peripheral neu-
rogenic tumors of the soft tissues in
children and adolescents: A clinico-
pathologic study of 139 cases. Pediatr
Pathol 1989;9:387-407.
17. Lowman RM, Livolsi VA: Pigmented
(melanotic) schwannomas of the spinal
canal. Cancer 1980;46:391-397.
18. Hays DM: Rhabdomyosarcoma. Clin
Orthop 1993;289:36-49.
19. Costa J, Wesley RA, Glatstein E, Rosen-
berg SA: The grading of soft tissue sar-
comas: Results of a clinicohistopatho-
logic correlation in a series of 163 cases.
Cancer 1984;53:530-541.
20. Weston WJ: Thorn and twig-induced
pseudotumours of bone and soft tis-
sues. Br J Radiol 1963;36:323-326.
21. Aflatoon K, Aboulafia AJ, McCarthy EF
Jr, Frassica FJ: Soft tissue mass in a
5-year-old boy. Clin Orthop 2003;411:
348-354.
22. Sanchez Reyes JM, Alcaraz Mexia M,
Quinones Tapia D, Aramburu JA: Ex-
tensively calcified synovial sarcoma.
Skeletal Radiol 1997;26:671-673.
23. Bridges AJ, Hsu KC, Singh A, Churchill
R, Miles J: Fibrodysplasia (myositis) os-
sificans progressiva. Semin Arthritis
Rheum 1994;24:155-164.
24. Moser RP Jr, Parrish WM: Radiologic
evaluation of soft tissue tumors, in
Enzinger FM, Weiss SW (eds): Enzinger
and Weiss’s Soft Tissue Tumors, ed 4. St.
Louis, MO: Mosby-Year Book Inc, 2001,
pp 45-102.
25. Hartman TE, Berquist TH, Fetsch JF:
MR imaging of extraabdominal des-
moids: Differentiation from other neo-
plasms. AJR Am J Roentgenol 1992;158:
581-585.
26. Enzinger FM, Weiss SW: Benign fibrous
tissue tumors, in Enzinger FM, Weiss
SW (eds): Soft Tissue Tumors, ed 3. St.
Louis, MO: Mosby-Year Book Inc, 1995,
pp 165-199.
27. Shives TC: Biopsy of soft-tissue tumors.
Clin Orthop 1993;289:32-35.
28. Simon MA, Biermann JS: Biopsy of
bone and soft-tissue lesions. J Bone Joint
Surg Am 1993;75:616-621.
29. Liu K, Layfield LJ, Coogan AC, Ballo MS,
Bentz JS, Dodge RK: Diagnostic accura-
cy in fine-needle aspiration of soft tis-
sue and bone lesions: Influence of clin-
ical history and experience. Am J Clin
Pathol 1999;111:632-640.
30. Kilpatrick SE, Ward WG, Cappellari JO,
Bos GD: Fine-needle aspiration biopsy
of soft tissue sarcomas: A cytomorpho-
Journal of the American Academy of Orthopaedic Surgeons
logic analysis with emphasis on histo-
logic subtyping, grading, and thera-
peutic significance. Am J Clin Pathol
1999;112:179-188.
31. Heslin MJ, Lewis JJ, Woodruff JM, Bren-
nan MF: Core needle biopsy for diagno-
sis of extremity soft tissue sarcoma.
Ann Surg Oncol 1997;4:425-431.
32. Neville HL, Andrassy RJ, Lobe TE, et al:
Preoperative staging, prognostic fac-
tors, and outcome for extremity rhab-
domyosarcoma: A preliminary report
from the Intergroup Rhabdomyosarco-
ma Study IV (1991-1997). J Pediatr Surg
2000;35:317-321.
33. Khirwadkar N, Dey P, Das A, Gupta SK:
Fine-needle aspiration biopsy of meta-
static soft-tissue sarcomas to lymph
nodes. Diagn Cytopathol 2001;24:229-232.
34. Jager PL, Hoekstra HJ, Leeuw J, van
Der Graaf WT, de Vries EG, Piers D:
Routine bone scintigraphy in primary
staging of soft tissue sarcoma: Is it
worthwhile? Cancer 2000;89:1726-1731.
35. Kim EE, Valenzuela RF, Kumar AJ,
Raney RB, Eftekari F: Imaging and clin-
ical spectrum of rhabdomyosarcoma in
children. Clin Imaging 2000;24:257-262.
36. Yoshikawa H, Myoui A, Ochi T, et al:
Bone metastases from soft tissue sarco-
mas. Semin Musculoskelet Radiol 1999;3:
183-190.
37. Reiseter T, Nordshus T, Borthne A,
Roald B, Naess P, Schistad O: Lipoblas-
toma: MRI appearances of a rare paedi-
atric soft tissue tumour. Pediatr Radiol
1999;29:542-545.
38. Beham A, Fletcher CD: Intramuscular an-
gioma: A clinicopathological analysis of
74 cases. Histopathology 1991;18:53-59.
39. Jelinek JA, Stelzer KJ, Conrad E, et al:
The efficacy of radiotherapy as postop-
erative treatment for desmoid tumors.
Int J Radiat Oncol Biol Phys 2001;50:
121-125.
40. Samuels BL: Management of recurrent
desmoid tumor after surgery and radi-
ation: Role of cytotoxic and non-
cytotoxic therapies. Surg Oncol 1999;8:
191-196.
41. Azzarelli A, Gronchi A, Bertulli R, et al:
Low-dose chemotherapy with metho-
trexate and vinblastine for patients
with advanced aggressive fibromatosis.
Cancer 2001;92:1259-1264.
42. Ballo MT, Zagars GK, Pollack A, Pisters
PW, Pollack RA: Desmoid tumor: Prog-
nostic factors and outcome after sur-
gery, radiation therapy, or combined
surgery and radiation therapy. J Clin
Oncol 1999;17:158-167.

43. Merchant TE, Nguyen D, Walter AW,
Pappo AS, Kun LE, Rao BN: Long-term
results with radiation therapy for pedi-
atric desmoid tumors. Int J Radiat Oncol
Biol Phys 2000;47:1267-1271.
44. Spiegel DA, Dormans JP, Meyer JS, et
al: Aggressive fibromatosis from infan-
cy to adolescence. J Pediatr Orthop 1999;
19:776-784.
45. Skapek SX, Hawk BJ, Hoffer FA, et al:
Combination chemotherapy using vin-
blastine and methotrexate for the treat-
ment of progressive desmoid tumor in
children. J Clin Oncol 1998;16:3021-3027.
46. Shields CJ, Winter DC, Kirwan WO,
Redmond HP: Desmoid tumours. Eur J
Surg Oncol 2001;27:701-706.
47. Andrassy RJ, Okcu MF, Despa S, Raney
RB: Synovial sarcoma in children: Sur-
gical lessons from a single institution
and review of the literature. J Am Coll
Surg 2001;192:305-313.
48. Ferrari A, Casanova M, Massimino M,
et al: Synovial sarcoma: Report of a se-
ries of 25 consecutive children from a
single institution. Med Pediatr Oncol
1999;32:32-37.
49. Okcu MF, Despa S, Choroszy M, et al:
Synovial sarcoma in children and ado-
lescents: Thirty three years of experi-
ence with multimodal therapy. Med Pe-
diatr Oncol 2001;37:90-96.
50. Womer RB, Pressey JG: Rhabdomyosa-
rcoma and soft tissue sarcoma in child-
hood. Curr Opin Oncol 2000;12:337-344.
51. Pappo AS, Shapiro DN, Crist WM,
Vol 11, No 5, September/October 2003
Maurer HM: Biology and therapy of pe-
diatric rhabdomyosarcoma. J Clin Oncol
1995;13:2123-2139.
52. Weiss SW, Goldblum JR: Rhabdomyo-
sarcoma, in Weiss SW, Goldblum JR
(eds): Enzinger and Weiss’s Soft Tissue
Tumors, ed 4. St. Louis, MO: Mosby,
2001, pp 785-836.
53. Furlong MA, Mentzel T, Fanburg-
Smith JC: Pleomorphic rhabdomyosar-
coma in adults: A clinicopathologic
study of 38 cases with emphasis on
morphologic variants and recent skele-
tal muscle-specific markers. Mod Pathol
2001;14:595-603.
54. Tobar A, Avigad S, Zoldan M, Mor C,
Goshen Y, Zaizov R: Clinical relevance
of molecular diagnosis in childhood
rhabdomyosarcoma. Diagn Mol Pathol
2000;9:9-13.
55. Anderson J, Gordon T, McManus A, et
al: Detection of the PAX3-FKHR fusion
gene in paediatric rhabdomyosarcoma:
A reproducible predictor of outcome?
Br J Cancer 2001;85:831-835.
56. Baker KS, Anderson JR, Link MP, et al:
Benefit of intensified therapy for pa-
tients with local or regional embryonal
rhabdomyosarcoma: Results from the
Intergroup Rhabdomyosarcoma Study
IV. J Clin Oncol 2000;18:2427-2434.
57. Sandler E, Lyden E, Ruymann F, et al:
Efficacy of ifosfamide and doxorubicin
given as a phase II “window” in children
with newly diagnosed metastatic rhab-
domyosarcoma: A report from the Inter-
Kamran Aflatoon, DO, et al
group Rhabdomyosarcoma Study Group.
Med Pediatr Oncol 2001;37:442-448.
58. Breitfeld PP, Lyden E, Raney RB, et al:
Ifosfamide and etoposide are superior
to vincristine and melphalan for pedi-
atric metastatic rhabdomyosarcoma
when administered with irradiation
and combination chemotherapy: A re-
port from the Intergroup Rhabdomyo-
sarcoma Study Group. J Pediatr Hematol
Oncol 2001;23:225-233.
59. Donaldson SS, Meza J, Breneman JC, et
al: Results from the IRS-IV randomized
trial of hyperfractionated radiotherapy
in children with rhabdomyosarcoma: A
report from the IRSG. Int J Radiat Oncol
Biol Phys 2001;51:718-728.
60. Crist WM, Anderson JR, Meza JL, et al:
Intergroup rhabdomyosarcoma study-
IV: Results for patients with nonmeta-
static disease. J Clin Oncol 2001;19:3091-
3102.
61. Wolden SL, Anderson JR, Crist WM, et
al: Indications for radiotherapy and
chemotherapy after complete resection
in rhabdomyosarcoma: A report from
the Intergroup Rhabdomyosarcoma
Studies I to III. J Clin Oncol 1999;17:
3468-3475.
62. Smith LM, Anderson JR, Qualman SJ, et
al: Which patients with microscopic
disease and rhabdomyosarcoma expe-
rience relapse after therapy? A report
from the soft tissue sarcoma committee
of the children’s oncology group. J Clin
Oncol 2001;19:4058-4064.
343
 Use of Physical Forces in Bone Healing
Fred R. T. Nelson, MD, Carl T. Brighton, MD, PhD, James Ryaby, PhD, Bruce J. Simon, PhD,
Jason H. Nielson, MD, Dean G. Lorich, MD, Mark Bolander, MD, PhD, and John Seelig, MD

Abstract
During the past two decades, a number of physical modalities have been approved erate physical forces to influence
for the management of nonunions and delayed unions. Implantable direct current fracture healing is necessary for
stimulation is effective in managing established nonunions of the extremities and their optimal clinical application
as an adjuvant in achieving spinal fusion. Pulsed electromagnetic fields and capac- (Table 2).
itive coupling induce fields through the soft tissue, resulting in low-magnitude volt-
age and currents at the fracture site. Pulsed electromagnetic fields may be as ef-
fective as surgery in managing extremity nonunions. Capacitive coupling appears Dr. Nelson is Director of Resident Education, Hen-
ry Ford Hospital, Detroit, MI. Dr. Brighton is Paul
to be effective both in extremity nonunions and lumbar fusions. Low-intensity ul-
B. Magnuson Professor Emeritus of Bone and Joint
trasound has been used to speed normal fracture healing and manage delayed unions. Surgery, Department of Orthopaedic Surgery,
It has recently been approved for the management of nonunions. Despite the dif- University of Pennsylvania, Philadelphia, PA. Dr.
ferent mechanisms for stimulating bone healing, all signals result in increased in- Ryaby is Senior Vice President, OrthoLogic,
tracellular calcium, thereby leading to bone formation. Tempe, AZ. Dr. Simon is Director of Research,
EBI, Parsippany, NJ. Dr. Nielson is Chief Resi-
J Am Acad Orthop Surg 2003;11:344-354
dent, Department of Orthopaedic Surgery, Jaco-
by Medical Center, Bronx, NY. Dr. Lorich is As-
sociate Director, Orthopaedic Trauma Surgery,
Hospitals for Special Surgery, New York, NY. Dr.
Nonunion has been defined as no These options include direct current Bolander is Professor of Surgery, Mayo Clinic,
Rochester, MN. Dr. Seelig is Doctor of Neurosur-
demonstrated change in healing on (DC), pulsed electromagnetic fields
gery, San Diego, CA.
serial radiographs over a 3-month pe- (PEMFs), capacitive couplings, and
riod.1 Delayed union is defined as a ultrasound. None of the following authors or the departments
speed of fracture healing that is slow- Over the past two decades, an es- with which they are affiliated has received any-
er than anticipated, with no implied timated 400,000 fracture nonunions, thing of value from or owns stock in a commer-
cial company or institution related directly or in-
expectancy of either eventual healing delayed unions, and fusions have
directly to the subject of this article: Dr. Nelson,
or eventual nonunion. Of approxi- been managed by physical fields. In Dr. Nielson, Dr. Lorich, and Dr. Seelig. Dr.
mately 6 million extremity fractures January 2000, the Society for Physi- Brighton or the department with which he is af-
that occur annually in the United cal Regulation in Biology and Med- filiated has received research or institutional sup-
States,2,3 between 5% and 10% result icine sponsored a symposium to re- port from Biolectron. Dr. Brighton or the depart-
ment with which he is affiliated has received
in either nonunion or delayed view the clinical applications and
royalties from Biolectron. Dr. Brighton or the de-
union.3 Assuming an average cost in mechanisms of action for these var- partment with which he is affiliated serves as a
lost wages and additional medical ious modalities. The core material consultant to or is an employee of Biolectron. Dr.
treatment for each of these cases of from that symposium has been orga- Ryaby or the department with which he is affil-
$10,000, the annual economic loss is nized into a format to help clinicians iated serves as a consultant to or is an employee
of OrthoLogic. Dr. Simon or the department with
$3 to $6 billion. In an attempt to min- become more effective in and
which he is affiliated has stock or stock options
imize problems with fracture healing, knowledgeable about application of held in Biomet. Dr. Bolander or the department
improved methods of internal and ex- these physical signals. Physicians with which he is affiliated has received research or
ternal fracture immobilization have should be familiar with commonly institutional support from Simth & Nephew and
been combined with appropriately used terms and their definitions (Ta- Exogen.
timed early transmission of physio- ble 1) and appreciate the history of
Reprint requests: Dr. Nelson, K-12, 2799 W.
logic forces across the fracture sites.4 the clinical use of these physical Grand Boulevard, Detroit, MI 48202.
Additionally, a number of adjunctive forces. A thorough understanding of
treatment options to stimulate normal the mechanisms of action, indica- Copyright 2003 by the American Academy of
fracture healing, delayed unions, and tions for use, and clinical outcomes Orthopaedic Surgeons.
nonunions have been developed.5 of commonly used devices that gen-

344 Journal of the American Academy of Orthopaedic Surgeons

 Fred R. T. Nelson, MD, et al

Table 1
Terms and Definitions

Term
Physical forces
Direct electrical stimulation
Pulsed electromagnetic
fields (PEMFs)
Capacitive coupling
Definition
Include any mechanical, electrical, or sonic force applied to an area of bone fracture heal-
ing. This is in distinction from biochemical osteoinductive therapies.
Involves an implanted cathode placed in the area of expected bone stimulation and a
battery-based anode placed subcutaneously. A constant 20-µA direct current is deliv-
ered.
Use magnetic coils that receive a specific pulsed electrical current that results in a mag-
netic flux density ≈0.1 to 18 G in the form of a pulse train with a 15-Hz or sinusoidal
76-Hz frequency. A pulse train is a rapid sequence, typically of twenty 220-µsec repeat-
ing spikes. A gauss (G) is a unit of electromagnetic flux. (The earth’s geomagnetic field
is approximately 0.6 G.)
Requires two surface electrodes placed on the skin across a fracture site. A 60-kHz si-
nusoidal wave signal is generated by a 9-V battery; this results in an internal field of 0.1
to 20 mV/cm and a current density of 300 µA/cm2 that is not felt by the patient.

History of Development of
Physical Fields
In 1841, Hartshorne6 described a case
of fracture nonunion that was treat-
ed with “shocks of electric fluid
passed daily through the space be-
tween the ends of the bone.” Lente7
in 1850 described three cases of de-
layed unions or nonunions treated
with galvanic current. More than 100
years later, electrical stimulation of
bone regained clinical scientific prom-
inence when Fukada and Yasuda8 de-
scribed “piezoelectric potentials”
generated by mechanical stress on the
crystalline structure of bone. At the
same time, ultrasound began to show
promise as a method of stimulating
fracture healing. In 1953, Corradi
used continuous wave ultrasound to
stimulate fracture healing, producing
an increase in periosteal callus.9
A central hypothesis in the appli-
cation of physical forces is that strain-
generated electrical potentials may be
a regulatory signal for cellular pro-
cesses of bone formation. The idea
that electrical fields might be impor-
tant in the repair process was de-
scribed in the early 1960s by Bassett
and Becker.10 AFourier transform was
used to break down the electromag-
netic signal into its major and minor
components to predict the biological-
ly important rate of generation of
electric potentials in bone by mechan-
ical stress.10 This was used as the ba-
sis for selecting one of the currently
used PEMFs. Additional knowledge
of the nature of endogenous electric
fields in biology led to the develop-
ment of the direct electric fields now
in use.11 Subsequently there was fur-
ther development of PEMFs as well

Table 2
Devices That Generate Physical Forces

Device Wave Form Tissue Electrical Field

Direct current
Pulsed electromagnetic
field
Capacitive coupling
Pulsed electromagnetic
field, modified
Combined magnetic field
Ultrasound
20 µA As delivered
4.5-msec–long bursts of twenty 220-µsec 1.5 mV/cm; 10 µA/cm2
18-G pulses repeated at 15 Hz
60 kHz, 10 µA (rms), 6 V peak to peak 0.1 to 20 mV/cm and
delivered by 9-V battery 300 µA/cm2 at 60 kHz
790-mG field of a burst of twenty-one 4 mV/cm peak to peak
260-µsec pulses with repetition rate of 15 Hz
76.6-Hz sinusoidal 40-µT (400 mG) peak-to- Magnetic field effect, not
peak AC magnetic field superimposed on induced field
20-µT DC magnetic field
Sinusoidal N/A

rms = root-mean-square

Vol 11, No 5, September/October 2003 345

Use of Physical Forces in Bone Healing
as combined (DC and AC) magnetic
fields.
Arabbit fibular fracture model was
used to define the dose-response
curve for capacitive coupling in frac-
ture healing. An internal field of 220
mV with a current density of 250 µA
was the most effective for induction
of healing.12 The effects of ultrasound
on fracture callus stimulation were
studied by numerous investigators
using a variety of animal models.9
Pilla et al13 found that brief periods
(20 min/day) of pulsed ultrasound (a
200-µsec burst of 1.5-MHz sinusoidal
waves repeated at 1 kHz) at a low in-
tensity (30 mW/cm2) accelerated the
recovery of torsional strength and
stiffness in a midshaft fibular osteot-
omy of the rabbit.
Although most human clinical
studies conducted during the devel-
opment of these devices were retro-
spective, prospective controlled stud-
ies now exist. However, most of these
record only the presence or absence
of healing as an end point. Outcomes
such as return to work or specific ac-
tivities have not been reported but are
important for assessing the role of
these devices compared with alterna-
tive techniques to stimulate fracture
repair. Revascularization, as in core
decompression for osteonecrosis of
the femoral head, and stimulation of
articular cartilage repair in osteoar-
thritis are potential new applications
for these methodologies that are cur-
rently under investigation.
Direct Current
Basic Science
In 1981, Brighton et al11 showed
that with direct electrical stimulation,
the pO2 is lowered and pH raised in
the vicinity of the cathode. A low
pO2 is favorable to bone formation;
Brighton et al11 found lower pO2 at
the bone-cartilage junction of the
growth plate and in newly formed
bone and cartilage in fracture callus.
Among the cellular mechanisms of
346
electric current–induced osteogenesis established nonunion is one that
are increased proteoglycan and col- shows no visible progressive signs of
lagen synthesis14 (Table 3). healing. The FDA’s original definition
stipulated no visible signs of healing
Clinical Data for at least 3 months after at least 9
After the initial clinical demonstra- months since injury.) Originally the
tion of fracture healing in 1971 by anode was placed on the skin, with
Friedenberg et al,57 Brighton et al22 in a battery pack worn at the waist. Im-
1977 reported the use of DC by per- plantable batteries acting as anodes
cutaneous wire placement for tibial were later developed to deliver a con-
nonunions that had been present for sistent 20-µA current. The cathode
an average of 3.3 years. Treated with now can be wrapped in a spiral and
a field of 10 to 20 µA over 12 weeks, shaped to match the area of interest.
39 of 57 nonunions healed. Based on In contrast with surface induction,
this study and animal models, 20 µA implanted DC stimulation eliminates
was determined to be the preferred the problem of patient compliance
current. In 1981, Brighton et al11 re- when used in conjunction with a sur-
ported on 178 nonunions managed gical procedure for internal fixation
with 4 percutaneously inserted cath- or bone grafting. Direct electrical
odes, each delivering a 20-µA DC, re- stimulation also has been approved
sulting in 149 successful unions. Suc- by the FDA for use in spinal fusion.
cess rates were 83.3% for tibial An open exposure is required, and the
nonunions, 66.7% for clavicular battery/anode is removed 6 months
nonunions, and 61.5% for humeral after implantation.
nonunions. The presence of a syno-
vial-lined pseudarthrosis prevented
healing. Pulsed Electromagnetic
Fields
Current Indications
In 1979, the Food and Drug Ad- Basic Science
ministration (FDA) approved the use The PEMF signal was developed
of DC in established nonunions. (An to induce electrical fields in bone sim-
Table 3
Physical Forces in Bone Healing: Mechanisms of Action
Device (Clinical Studies) Mechanism*
Direct current11-14 O2↓,11,14 synthesis collagen and proteoglycan11
Pulsed electromagnetic Cytokines12,13,21-25
field (PEMF)15-20
Capacitive Bone cell proliferation,29 activate voltage-gated
coupling26-28 calcium channels, PGE2, cytosolic calcium,
activated calmodulin,26 mRNA TGF-β30
Modified PEMF31,32 Vascular ingrowth, osteoblast migration, matrix
calcification33
Combined magnetic Ion transport across cell membranes and ion
field15,34 dependent cell signaling in tissues,35-37
growth cytokines38-47
Ultrasound9,48-51 Influx and efflux of K+,52 cartilage bone Ca++,53
adenylate cyclase activity,54 TGF-β,54 PGE2,55
aggrecan and vascularity,9 PDGF-AB56
* Increases, stimulates, or activates
Journal of the American Academy of Orthopaedic Surgeons

ilar in magnitude and time course to
the endogenous electrical fields pro-
duced in response to strain. These
fields are thought to underlie the abil-
ity of bone to respond to a changing
mechanical environment, as de-
scribed by Wolff’s law. The signal con-
sists of 4.5-msec–long bursts of twen-
ty 220-µsec 18-G pulses repeated at
15 Hz. This results in a time-varying
extracellular and intracellular electri-
cal field.
Research with PEMFs has focused
on regulation of messenger RNA
(mRNA) and protein synthesis of the
transforming growth factor-beta
(TGF-β)/bone morphogenetic protein
(BMP) gene family because these cy-
tokines have been shown to modu-
late cellular activity of osteochondral
progenitor cells, chondrocytes, and
osteoblasts. In many animal studies
and recently in human clinical trials,
TGF-β, BMP-2, and BMP-7 have been
shown to enhance fracture repair. In
an endochondral ossification model
using demineralized bone matrix–in-
duced osteogenesis, PEMF treatment
caused an increase in chondrogene-
sis concomitant with an up-regulation
of TGF-β.23 A several-fold increase in
BMP-2 and BMP-4 mRNA occurred
in chick calvarial osteoblasts in vivo
after 15 days of stimulation with this
same signal.58 In a rat calvarial osteo-
blast culture, 1 hour of stimulation re-
sulted in a threefold increase in
BMP-2 mRNA and a sixfold increase
in BMP-4 mRNA.59
Two recent studies describe the ef-
fects of PEMF on TGF produc-
tion.25,60 In one,25 confluent cultures
of MG63 human osteoblast-like cells
were stimulated for 8 hours a day for
4 days and showed a significant (P <
0.05) increase in TGF levels in stim-
ulated versus control cells after 1 and
2 days of stimulation. PEMF enhanc-
es differentiation of MG63 cells, as ev-
idenced by decreased proliferation
and increased alkaline phosphatase
activity and osteocalcin and collagen
production. These results support
earlier observations in the endochon-
Vol 11, No 5, September/October 2003
dral bone model that PEMF stimula-
tion increases chondrogenesis by
enhancing differentiation of osteo-
chondral precursor cells into a chon-
drogenic lineage without affecting
proliferation.24 In a second study,60
nonunion cells derived from patients
undergoing surgery were successful-
ly cultured, and PEMF stimulation of
these cells resulted in significant (P
< 0.05) increases in TGF-β production
compared with nonstimulated con-
trol cells.16 Cells derived from hyper-
trophic nonunion tissue were more
responsive than cells derived from
atrophic tissue, a result that supports
the clinical observation that patients
with hypertrophic nonunions re-
spond more favorably to electromag-
netic stimulation than do patients
with atrophic nonunions.
Clinical Data
More than 250 published basic re-
search and clinical investigations
have evaluated the efficacy of PEMF
stimulation.21 In 1990, Sharrard re-
ported a double-blind trial of delayed
unions in 45 tibial shaft fractures
managed by plaster cast, with active
PEMF units (n = 20) or identical dum-
my control units (n = 25) for a period
of 12 weeks.19 Nine of 20 fractures
(45%) in the active group healed,
compared with 3 of 25 fractures (12%)
in the control group (P < 0.01).19 Bas-
sett et al61 reported on a series of 127
diaphyseal tibia nonunions treated
with PEMFs that yielded an overall
success rate of 87%. A year later, Bas-
sett et al62 reported the results of
PEMF treatment with surgery and
bone grafting in 83 nonunions with
wide fracture gaps, synovial pseudar-
throsis, and malalignment. These pa-
tients achieved an 87% success rate.
In a broad literature review compar-
ing PEMF treatment of nonunions
with surgical therapy, Gossling et
al16 noted that 81% of reported cases
healed with PEMF versus 82% with
surgery. Also, the success of surgical
treatment for infected nonunions was
69%, whereas 81% of the PEMF-
Fred R. T. Nelson, MD, et al
treated group healed.16 In open frac-
tures, surgical healing exceeded
PEMF (89% and 78%, respectively),
but in closed injuries, PEMF-
managed fractures healed more fre-
quently than did surgically treated
fractures (85% and 79%, respective-
ly). This study indicates the efficacy
of PEMF treatment to be comparable
to that of surgical intervention for
fracture nonunion.
PEMF treatment has applications
in the upper extremity, as well. Fryk-
man et al17 reported that 35 of 44
scaphoid nonunions (80%) were man-
aged successfully by PEMFs with cast
immobilization. However, in a con-
tinuation of that study published 6
years later, the overall success rate
had decreased to 69% because of
breakdown of some of the fractures
originally reported as unions. Prox-
imal pole fractures healed in 50%.63
The daily dosage of PEMF treat-
ment is important in the healing pro-
cess. A dose-response study demon-
strated that an increase in daily
treatment time correlates with a re-
duction in the time to healing of non-
union fractures.18 Patients treated for
10 hours per day healed an average
of 76 days earlier than did those treat-
ed fewer than 3 hours per day.
Current Indications
PEMF treatment is recommended
as an adjunct to standard fracture
management. Indications for use in-
clude nonunions, failed fusions, and
congenital pseudarthrosis. Recently,
the definition of a nonunion has been
modified to failure to exhibit visibly
progressive signs of healing.64 This
definition thus permits all forms of
electrical stimulation intervention to
take place earlier in the treatment
than previously and removes contro-
versy regarding when a delayed
union may be considered a nonunion.
Generally, a fracture gap >5 mm,
suspected or documented synovial
pseudarthrosis, and severe devascu-
larization are contraindications for
the use of PEMFs. Patients typically
347
Use of Physical Forces in Bone Healing
are treated for 3 to 9 months depend-
ing on fracture location, severity, and
time from injury. Some difficult frac-
tures may require management for
longer periods. The fracture should
progress to healing within 3 to 6
months. If surgery is needed, some
patients choose to continue use of the
stimulator to enhance healing after
surgery.
Capacitive Coupling
Basic Science
Use of capacitive coupling for frac-
ture healing stimulation involves the
application of two surface electrodes
placed on the skin with the fracture
between the electrodes. The induced
field is driven by an oscillating elec-
tric current, as opposed to the elec-
tromagnetic field induction of PEMF.
In an in vitro rat calvarial bone cell
model, Brighton et al29 found that
field strength was the dominant fac-
tor affecting bone cell proliferative re-
sponse to a capacitive coupled field.
Field strengths calculated at 0.1 to 20
mV/cm (60 kHz and 300 µA/cm2),
with various pulse configurations as
well as continuous signals, are effec-
tive in stimulating bone cell prolifer-
ation.29 The clinical effect of electri-
cally induced osteogenesis is easily
recognized. However, the basic phys-
iology of how electrical signals stim-
ulate bone is more difficult to dem-
onstrate in the laboratory. Using
various metabolic inhibitors, Lorich
et al26 showed that signal transduc-
tion in capacitive coupling stimula-
tion activated voltage-gated calcium
channels, leading to increases in pros-
taglandin E2 (PGE2), cytosolic calci-
um, and activated calmodulin. This
is in contrast to signal transduction
of indirect coupling and combined
magnetic fields (CMFs), in which the
cytolsolic calcium is secondary to re-
lease of calcium from intracellular
stores. This leads to an increase in ac-
tivated calmodulin. Although the ini-
tial signal transduction of capacitive
348
coupling is different from inductive
coupling of a combined DC and
pulsed electromagnetic field, there
appears to be a common pathway.65
In addition, Zhuang et al30 demon-
strated that an appropriate capaci-
tively coupled electrical field in-
creased levels of mRNA for TGF-β1
in osteoblastic cells by a mechanism
involving the calcium/calmodulin
pathway.
Clinical Data
In a prospective, nonrandomized
multicenter study comparing patients
with 17 recalcitrant nonunions (who
had undergone prior surgery or elec-
trical stimulation) with 5 who had
routine nonunions (no previous treat-
ment), Brighton and Pollack1 report-
ed a mean healing rate of 77.3% with
capacitive coupling after a mean of
22.5 weeks. Brighton et al66 used lo-
gistic regression analysis in a retro-
spective study of the healing rate of
271 tibial nonunions treated by DC,
capacitive coupling, or bone graft.
The authors identified seven risk fac-
tors that adversely affected the heal-
ing rate of nonunions managed with
capacitive coupling: duration of non-
union, prior bone graft surgery, pri-
or electrical stimulation, open frac-
ture, osteomyelitis, comminuted or
oblique fracture, and atrophic non-
union. With no or one risk factor
present, there were no significant dif-
ferences among the three treatment
methods (96% to 99%). With the pres-
ence of two to five risk factors, capac-
itive coupling yielded poorer results
in managing atrophic nonunion; oth-
erwise, results were similar regard-
less of treatment modality. With six
or seven risk factors, all three forms
of treatment provided poor results.
Unfortunately, this study did not
evaluate smoking as a possible risk
factor.
Scott and King27 reported the re-
sults of a small, prospective double-
blind study using capacitive coupling
in the management of established
nonunions. They found a statistical-
Journal of the American Academy of Orthopaedic Surgeons
ly significant association between the
use of capacitive coupling and even-
tual union. Six of the 10 nonunions
in the actively managed group
healed, compared with none of the 11
in the placebo group (P = 0.004). There
also have been two double-blind pro-
spective lumbar fusion studies using
capacitive coupling. Goodwin et al28
studied 179 patients randomized into
groups assigned active or nonactive
coils after lumbar fusion. The authors
reported a statistically significant (P
= 0.0043) increased rate of fusion in
the active group (84.7%) compared
with the placebo group (64.9%). Pos-
terolateral bone graft combined with
concurrent instrumentation of the af-
fected levels had a higher rate of fu-
sion than did graft without instru-
mentation. Within the instrumented
group, stimulated patients showed
higher fusion rates than did the pla-
cebo control subjects.
Current Indications
Capacitive coupling is indicated
for nonunions of long bones and the
scaphoid and as an adjunct treatment
in spinal fusions. In applying capac-
itive coupling, cast immobilization
typically is used. Two small windows
are cut out for the application of the
electrodes, which are positioned
across the approximate site of the
fracture and moistened before appli-
cation. When the pads dry, the mon-
itor detects the loss of contact and sets
off an alarm, indicating that the pads
need to be remoistened. Currently
available electrodes last up to 1 week
without requiring reapplication of
gel. The pads are worn 24 hours a day
and are changed weekly, or more of-
ten as required for hygiene. The de-
vice uses a 9-V battery that should be
replaced daily. Skin reaction is usu-
ally mild. If necessary, electrodes can
be moved to a new skin site. Treat-
ment is discontinued if there is severe
skin reaction. Serial anteroposterior,
lateral, and oblique radiographs are
used to monitor progression of heal-
ing, as in normal fracture manage-

ment. Device usage is typically 25
weeks and is discontinued when the
fracture heals or after 3 months of no
progression in healing.
Pulsed Electromagnetic
Field, Modified
Basic Science
A modified PEMF was developed
to reduce energy requirements. It de-
livers an average 790-mG field of a
burst of twenty-one 260-µsec pulses
repeated at 15 Hz. The devices are
horseshoe-shaped, flattened sole-
noids; some use a saddle-shaped coil.
There are several suggested mecha-
nisms of action. Using the original
PEMF signal (also with a repetition
rate of 15 Hz), Yen-Patton et al33
showed that this modified PEMF in-
creased the number of vessels, or
“sprouting,” in endothelial tissue by
a factor of 10 to 15. The neovascular-
ization occurs in vitro after 5 to 8
hours of stimulation. The authors also
noted increased migration of osteo-
blasts and an enhanced mineraliza-
tion of new fibrocartilage.33 A differ-
ent field was developed for the spine,
delivered by dual coils that encom-
pass the entire lumbar area. This is a
160-mG field of ninety-nine 260-µsec
pulses.
Clinical Data
Amulticenter open trial of the mod-
ified PEMF device was conducted
with 139 patients who had one or more
fractures that had not healed for at
least 9 months (some >5 years).31 The
lengthy time of nonunion served as
the baseline because spontaneous frac-
ture healing was unlikely to occur. The
only intervention applied was the ad-
dition of PEMF therapy prescribed for
8 hours a day for at least 90 days. Frac-
ture healing was judged by four cri-
teria: cortical bone bridging and ab-
sence of motion on stress radiographs,
no or minimal pain, no or minimal
edema, and no need for casting. On
completion of the course of treatment,
Vol 11, No 5, September/October 2003
patients who wore the device for at
least 3 hours a day for a minimum of
90 days had a significantly (P < 0.05)
better healing rate than did patients
who complied to a lesser degree with
the treatment regimen (80% versus
19.2%). There was no significant dif-
ference in fracture healing rate for the
average wear times of 3 to 6 hours, 6
to 9 hours, and >9 hours. Healing oc-
curred in the presence of fracture gaps
≥6 mm whether the patient was a
smoker or had comminution, an open
fracture, prior infection, or multiple
surgical procedures. Long-term follow-
up 4 years later revealed essentially
the same healing rate with no long-
term adverse effects.
Mooney32 reported the results of
a prospective, multicenter, random-
ized, placebo-controlled clinical trial
of PEMF stimulation for lumbar spine
fusion. One hundred ninety-five pa-
tients underwent interbody fusion (an-
terior and posterior approaches). (In-
terbody fusions are easier to evaluate
than posterolateral fusions.) Spine fix-
ation was by hook and rod, predat-
ing the use of pedicle screws. Patients
were prescribed the device for a to-
tal of 8 hours a day for a minimum
of 90 days or until healed. An anal-
ysis of usage versus fusion success
demonstrated that a dosage of only
4 hours a day for 90 days was enough
to significantly (P = 0.005) increase fu-
sion rates. Consistent use at this lev-
el resulted in an overall fusion rate
of 92% in the PEMF group compared
with 64.9% in the placebo group. In
a second phase of this study, 126 pa-
tients with a failed fusion who were
at least 9 months from prior surgery
were given an active device to use for
8 hours a day for at least 90 days. No
additional surgery was done. The
study included both interbody and
posterolateral fusions at one or more
levels. Of patients who wore the de-
vice for at least 2 hours, 67% achieved
solid fusion.32 In a historical cohort
study of 42 patients treated with PEMF
stimulation and 19 nonstimulated pa-
tients, Marks67 found that the rate of
Fred R. T. Nelson, MD, et al
fusion enhancement (97.6% and 52.6%,
respectively) was statistically signif-
icant (P < 0.001).
Current Indications
The use of modified PEMF devic-
es is indicated for fracture nonunions
that demonstrate no radiographic ev-
idence of progression of bony heal-
ing. The recommended dose is 3
hours of daily usage until healing oc-
curs, typically 3 to 6 months. Use of
the Spinal-Stim (Orthofix, McKinney,
TX) is indicated as an adjunct to spi-
nal fusion surgery to increase the
probability of fusion success and as
a nonsurgical treatment to salvage a
failed spinal fusion. The recommend-
ed dose is at least 2 hours a day until
the patient is healed, typically 3 to 9
months.
Combined Magnetic Fields
Basic Science
The scientific basis of CMFs is
predicated on theoretic physics con-
firmed by experimental demonstra-
tions that combinations of dynamic
and static magnetic fields affect ion
transport across cell membranes and
affect ion-dependent cell signaling
in tissues.35-37 Specifically, combined
AC and DC magnetic fields are pre-
dicted to couple to calcium-depen-
dent and magnesium-dependent
cellular signaling processes in tis-
sues.
Cellular studies of CMFs have ad-
dressed effects on both signal trans-
duction pathways and growth factor
production. The resulting working
model from the studies of Fitz-
simmons and colleagues38-40 is the
proposal that short-duration CMF
stimulus of 30 minutes activates se-
cretion of growth factors (eg, insulin-
like growth factor-II [IGF-II]). The
clinical benefit on bone repair is the
result of this up-regulation of growth
factor production, with the short-term
(30-minute) CMF stimulus acting as
a triggering mechanism that couples
349
Use of Physical Forces in Bone Healing
to the normal molecular regulation of
bone repair mediated by growth fac-
tors. The studies underlying this
working model have shown effects of
CMFs on calcium ion transport38 and
cell proliferation.39 In 1995, Fitzsim-
mons and colleagues40,41 reported
IGF-II release and increased IGF-II re-
ceptor expression in osteoblasts. Ef-
fects of CMFs on IGF-I and IGF-II in
rat fracture callus were reported by
Ryaby et al.42 Recent studies have
shown effects of CMFs on experimen-
tal fracture healing43,44 and on os-
teopenic animal models,45,46 possibly
mediated by attenuation of tumor ne-
crosis factor α–dependent signaling
in osteoblasts.47 However provoca-
tive, the role of growth factors in
transduction of CMFs in cells and tis-
sues, and the link to the observed clin-
ical benefit of CMFs, require further
investigation.
Clinical Data
In a prospective, randomized pi-
lot study of patients with acute, phase
1 Charcot neuroarthropathy, 10 con-
trol subjects and 11 patients treated
with CMFs were followed weekly
and treated until the difference in
temperature between the two feet
was less than 2°C, foot volumes were
within 10% of each other, and frac-
ture consolidation had occurred.68
Subsequently, 10 more patients were
added to the CMF-treatment group.
Results showed that the mean time
to consolidation in the control group
was 23.2 ± 7.7 weeks. In contrast,
treatment with the CMF device de-
creased time to consolidation to 11.1
± 3.2 weeks (P < 0.001). There was no
statistically significant difference in
entry criteria between the control and
CMF groups.
The most recent application of
CMFs has been as an adjunctive stim-
ulation device for spinal fusion.69 A
prospective, randomized, double-
blind, placebo-controlled trial was
conducted on primary uninstrument-
ed lumbar spine fusion. Patients had
one- or two-level fusions (between L3
350
and S1) with either autograft alone or
in combination with allograft. The
CMF device configured for spinal fu-
sion has a single posterior coil cen-
tered over the fusion site. Treatment
was applied for 30 minutes a day for
9 months. The primary end point was
assessment of fusion at 9 months,
based on radiographic evaluation by
a blinded panel consisting of the treat-
ing physician, a musculoskeletal ra-
diologist, and a spine surgeon. This
panel evaluation differed from those
of other spinal fusion studies with
noninvasive bone growth stimulators
in that the treating surgeon’s assess-
ment of fusion could be overruled by
the blinded panel. Of the 243 patients
enrolled, 201 were available for eval-
uation. Of the patients with active de-
vices, 64% healed at 9 months; only
43% of placebo-device patients healed
(P = 0.003 by Fisher’s exact test).
Among female patients, 67% of those
with active devices achieved fusion
compared with 35% of those with
placebo devices (P = 0.001 by Fisher’s
exact test). Of the 201 patients,
repeated-measures analyses of fusion
outcomes showed a main effect of
treatment favoring the active treat-
ment (P = 0.030) in a model with only
a main effect. In a model with main
effect and a time-by-treatment inter-
action, the time-by-treatment interac-
tion was significant (P = 0.024), indi-
cating acceleration of healing. The
investigators concluded that the ad-
junctive use of the CMF device for
noninstrumented fusions results in
higher fusion rates and in earlier fu-
sions. This was the first randomized
clinical trial of noninstrumented pri-
mary posterolateral lumbar spine fu-
sion with evaluation by a blinded, un-
biased panel. The fusion rates in this
study were lower than those of other
noninstrumented studies reported in
the literature. The lower success rates
are thought to be because of the high-
risk patient group (average age, 57
years) coupled with the use of non-
instrumented technique with pos-
terolateral fusion only.
Journal of the American Academy of Orthopaedic Surgeons
Current Indications
Application of CMFs for 30 min-
utes a day has been shown to be ef-
fective for management of nonunions
and as adjunctive stimulation for pri-
mary spinal fusion. Future indications
for CMFs may include osteoarthritis
and neuroarthropathy, but adoption
of additional applications will require
increased knowledge of the tissue-
level mechanisms combined with well-
designed clinical trials.
Ultrasound
Basic Science
Azuma et al70 confirmed the in-
creased efficiency of the 200-µsec
burst (versus 100-µsec and 400-µsec
bursts) of 1.5-MHz sinusoidal waves
repeated at 1 kHz (versus 2 kHz) at
a low intensity of 30 mW/cm2. Ad-
ditional animal data suggest that the
biology of fracture healing can be ac-
celerated by the use of ultrasound but
that no specific stage of healing is
more sensitive than another.70 There
is a wide range of proposed mecha-
nisms by which low-intensity ultra-
sound stimulates fracture healing.9
Minimal heating effect (well below
1°C) may increase some enzymes,
such as matrix metalloproteinase 1
(interstitial collagenase), which are
exquisitely sensitive to small varia-
tions in temperature.71 Ultrasound
has been shown to change the rate of
influx and efflux of potassium ions,
increase calcium incorporation in
both differentiating cartilage and
bone cell cultures, and increase sec-
ond messenger activity paralleled by
the modulation of adenylate cyclase
activity and TGF-β synthesis in osteo-
blastic cells.52 In primary chondro-
cytes, the application of ultrasound
at 50 mW/cm2 increased release of
cellular calcium.53 Increased PGE2
production via the induction of
cyclooxygenase-2 mRNA occurs in
mouse osteoblasts in a manner sim-
ilar to that which is effected by fluid
shear stress and tensile force stimu-

li.55 Ultrasound has been shown to in-
crease the expression of genes in-
volved in the inflammation and
remodeling stages of fracture repair.
Low-intensity ultrasound stimulates
an up-regulation of aggrecan gene ex-
pression in cultured chondrocytes
and stimulates proteoglycan synthe-
sis in rat chondrocytes by increasing
aggrecan gene expression.72 This
might explain the role of ultrasound
in augmenting endochondral ossifi-
cation and thus increasing the me-
chanical strength and overall repair
of the fractured bone. Given the ef-
fect of low-intensity ultrasound on
hundreds of genes working in a com-
plex biologic system to achieve the
healing response, it would likely be
misleading to overemphasize the im-
pact of a single gene. Low-intensity
ultrasound treatment over a 10-day
period stimulated a greater degree of
vascularity in an osteotomized dog
ulna model of fracture healing.73 It is
generally believed that greater blood
flow serves as a principal factor in the
acceleration of fracture healing. In-
deed, one of the main biologic goals
of the inflammatory response is to re-
establish the blood supply to the in-
jured area.
Clinical Data
The initial clinical trials for ultra-
sound were focused on reduction of
healing time. A randomized, double-
blind, placebo-controlled study of 67
closed or grade 1 open tibial fractures
using ultrasound treatment of 20 min-
utes a day at 30 mW/cm2 led to a sig-
nificant (P < 0.01) 24% reduction in
the time of clinical healing (86 ± 5.8
days in the active-treatment group
compared with 114 ± 10.4 days in the
control group).48 Using both clinical
and radiographic criteria, a 38% de-
crease in the time to overall healing
was apparent. Twelve of 34 placebo-
treated patients (35%) developed de-
layed union, whereas only 2 of 33
ultrasound-treated patients (6%) had
delayed union (P < 0.01). In another
multicenter, prospective, randomized,
Vol 11, No 5, September/October 2003
double-blind, placebo-controlled clin-
ical trial of 61 dorsally angulated frac-
tures of the distal radius, the mean
time to union was significantly (P <
0.0001) reduced by 38% for ultra-
sound-treated patients (61 ± 3 days)
compared with placebo-treated patients
(98 ± 5 days).49 Ultrasound treatment
resulted in a significantly (P < 0.01)
smaller loss of reduction (20 ± 6%) com-
pared with placebo (43 ± 8%).49 Other
successful clinical trials have demon-
strated reduction of healing time with
ultrasound, including leg-lengthening
procedures.9 Ultrasound treatment of
nonunions resulted in an 85% heal-
ing rate in 385 nonunions, with a mean
healing time of 14 months.9
Ultrasound is not effective in all
settings requiring bone healing (ie,
tibial fractures stabilized with in-
tramedullary fixation). Other clinical
studies have demonstrated enhanced
rate of fracture healing in smokers,
patients with diabetes, and patients
with renal insufficiency or who are us-
ing steroids.
Current Indications
In October 1994, low-intensity ul-
trasound was approved for the stim-
ulation of healing of fresh fractures.
In February 2000, approval was ex-
tended to the treatment of established
nonunions. The device requires a dai-
ly 20-minute application of the ultra-
sound head on the skin through a win-
dow in the immobilization device. The
device is not portable; it must be at-
tached to a wall power source while
in use. With the depth of penetration
at 3.5 cm, the device must be close to
the bone to be effective.
Clinical Management
In the management of nonunions
with physical fields, the degree of im-
mobilization required for patient
comfort is usually similar to that for
gradual healing without stimulation.
Nonunions should be adequately sta-
bilized and have good healing poten-
Fred R. T. Nelson, MD, et al
tial (adequate soft-tissue coverage
and evidence of a good blood sup-
ply). The presence of a synovial
pseudarthrosis (articular-like surface)
is a contraindication for all physical
stimulation devices. A fracture with
palpable motion is generally immo-
bilized a joint above and below; how-
ever, some humeral, forearm, and leg
fractures may be more effectively im-
mobilized in a fracture brace. Delayed
unions and nonunions that are mala-
ligned require surgical correction be-
fore healing can occur. Weight bear-
ing is determined by the same criteria
as those used for nonstimulated man-
agement of a slow-healing fracture.
If physical stimulation is to be used
after internal fixation and/or grafting
of a nonunion, postoperative man-
agement is generally the same as for
cases in which no external stimula-
tion is used. The cost effectiveness of
any fracture stimulation device de-
pends on knowing which fractures re-
spond best, the requirements for fix-
ation or grafting, and the patient’s
employment, personal, and social cir-
cumstances.
Summary
Physical stimulation in the form of
electrical fields and ultrasound is
important in orthopaedic applica-
tions, including for nonunions and
spinal fusions. The common effect of
these forces appears to be an in-
crease in intracellular calcium by a
variety of cellular mechanisms. This
results in an increase in osteoblastic
function in cells capable of bone for-
mation. In selected cases, the success
rate approximates that of surgical
procedures. Physical forces also can
be used to enhance open techniques
such as bone grafts for fracture heal-
ing, arthrodeses, and spinal fusions.
Outcomes such as return to specific
activities or work have not yet been
reported. This information will be
important to assess these devices
comparatively with alternative tech-
351
Use of Physical Forces in Bone Healing
niques of stimulating fracture repair.
Future research directions for elec-
trical fields will include fractures at
risk, failed fusion, porous ingrowth,
osteoporosis, and revascularization
References
1. Brighton CT, Pollack SR: Treatment of
recalcitrant non-union with a capaci-
tively coupled electrical field: A prelim-
inary report. J Bone Joint Surg Am 1985;
67:577-585.
2. Praemer A, Furner S, Rice DP: Muscu-
loskeletal Conditions in the United States,
ed 2. Rosemont, IL: American Academy
of Orthopaedic Surgeons, 1999.
3. Praemer A, Furner S, Rice DP (eds):
Musculoskeletal Conditions in the United
States. Park Ridge, IL: American Acad-
emy of Orthopaedic Surgeons, 1992.
4. Buckwalter JA, Grodzinsky AJ: Load-
ing of healing bone, fibrous tissue, and
muscle: Implications for orthopaedic
practice. J Am Acad Orthop Surg 1999;5:
291-299.
5. Einhorn TA: Enhancement of fracture-
healing. J Bone Joint Surg Am 1995;77:
940-956.
6. Hartshorne E: On the causes and treat-
ment of pseudarthrosis and especially
the form of it sometimes called super-
numerary joint. Am J Med 1841;1:
121-156.
7. Lente RW: Cases of un-united fracture
treated by electricity. New York State J
Med 1850;5:317-319.
8. Fukada E, Yasuda I: On the piezoelec-
tric effect of bone. J Phys Soc Japan 1957;
12:1158-1162.
9. Rubin C, Bolander M, Ryaby JP, Had-
jiargyrou M: The use of low-intensity
ultrasound to accelerate the healing of
fractures. J Bone Joint Surg Am 2001;83:
259-270.
10. Bassett CAL, Becker RO: Generation of
electric potentials by bone in response
to mechanical stress. Science 1962;137:
1063-1064.
11. Brighton CT, Black J, Friedenberg ZB,
Esterhai JL, Day LJ, Connolly JF: A mul-
ticenter study of the treatment of non-
union with constant direct current.
J Bone Joint Surg Am 1981;63:2-13.
12. Brighton CT, Hozack WJ, Brager MD, et
al: Fracture healing in the rabbit fibula
when subjected to various capacitively
coupled electrical fields. J Orthop Res
1985;3:331-340.
13. Pilla AA, Mont MA, Nasser PR, et al: Non-
invasive low-intensity pulsed ultrasound
352
after core decompression for os-
teonecrosis of the femoral head.
Stimulation of articular cartilage
synthesis in osteoarthritis currently
is being investigated. Ultrasound is
accelerates bone healing in the rabbit.
J Orthop Trauma 1990;4:246-253.
14. Brighton CT, McCluskey WP: Cellular
response and mechanisms of action of
electrically induced osteogenesis, in
Peck WA (ed): Bone and Mineral
Research/4. Amsterdam, The Nether-
lands: Elsevier Science Publishers BV,
1986, pp 213-254.
15. Bassett CA: Fundamental and practical
aspects of therapeutic uses of pulsed
electromagnetic fields (PEMFs). Crit
Rev Biomed Eng 1989;17:451-529.
16. Gossling HR, Bernstein RA, Abbott J:
Treatment of ununited tibial fractures:
A comparison of surgery and pulsed
electromagnetic fields (PEMF). Orthope-
dics 1992;15:711-719.
17. Frykman GK, Taleisnik J, Peters G, et al:
Treatment of nonunited scaphoid frac-
tures by pulsed electromagnetic field
and cast. J Hand Surg [Am] 1986;11:
344-349.
18. Pethica BA, Brownell J: Abstract: The
dose-response relationship in PEMF
therapy of ununited fractures. Transac-
tions of the Eighth Annual Meeting of the
Bioelectrical Repair and Growth Society.
Washington, DC, 1988, p 28.
19. Sharrard WJ: A double-blind trial of
pulsed electromagnetic fields for de-
layed union of tibial fractures. J Bone
Joint Surg Br 1990;72:347-355.
20. Mammi GI, Rocchi R, Cadossi R, Mas-
sari L, Traina GC: The electrical stimu-
lation of tibial osteotomies: Double-blind
study. Clin Orthop 1993;288:246-253.
21. Ryaby JT: Clinical effects of electromag-
netic and electric fields on fracture
healing. Clin Orthop 1998;355(suppl):
S205-S215.
22. Brighton CT, Friedenberg ZB, Mitchell
EI, Booth RE: Treatment of nonunion
with constant direct current. Clin Or-
thop 1977;124:106-123.
23. Ciombor DM, Aaron RK: EMF stimu-
lates cartilage differentiation in endo-
chondral ossification coincident with
an increase in TGF expression, in Ber-
sani F (ed): Electricity and Magnetism in
Biology and Medicine. New York, NY:
Kluwer Academic/Plenum Press, 1999,
pp 139-144.
Journal of the American Academy of Orthopaedic Surgeons
being evaluated for stimulation of
fresh fractures in patients with co-
morbidities including older patient
age, diabetes, active smoking status,
vascular insufficiency, and obesity.
24. Aaron RK, Ciombor DMcK, Polk C:
Therapeutic potential of electric fields
in skeletal morphogenesis, in Buck-
walter JA, Ehrlich MG, Sandell LJ, Trip-
pel SB (eds): Skeletal Growth and Devel-
opment: Clinical Issues and Basic Science
Advances. Rosemont, IL: American
Academy of Orthopaedic Surgeons,
1998, pp 589-610.
25. Lohmann CH, Schwartz Z, Liu Y, et al:
Pulsed electromagnetic field stimulation
of MG63 osteoblast-like cells affects dif-
ferentiation and local factor production.
J Orthop Res 2000;18:637-646.
26. Lorich DG, Brighton CT, Gupta R, et al:
Biochemical pathway mediating the re-
sponse of bone cells to capacitive cou-
pling. Clin Orthop 1998;350:246-256.
27. Scott G, King JB: A prospective, double-
blind trial of electrical capacitive cou-
pling in the treatment of non-union of
long bones. J Bone Joint Surg Am 1994;
76:820-826.
28. Goodwin CB, Brighton CT, Guyer RD,
Johnson JR, Light KI, Yuan HA: A
double-blind study of capacitively cou-
pled electrical stimulation as an adjunct
to lumbar spinal fusions. Spine 1999;24:
1349-1357.
29. Brighton CT, Okereke E, Pollack SR,
Clark CC: In vitro bone-cell response to
a capacitively coupled electrical field:
The role of field strength, pulse pattern,
and duty cycle. Clin Orthop 1992;285:
255-262.
30. Zhuang H, Wang W, Seldes RM, Taher-
nia AD, Fan H, Brighton CT: Electrical
stimulation induces the level of TGF-
beta 1 mRNA in osteoblastic cells by a
mechanism involving calcium/
calmodulin pathway. Biochem Biophys
Res Commun 1997;237:225-229.
31. Garland DE, Moses B, Salyer W: Long-
term follow-up of fracture nonunions
treated with PEMFs. Contemp Orthop
1991;22:295-302.
32. Mooney V: A randomized double-blind
prospective study of the efficacy of pulsed
electromagnetic fields for interbody lum-
bar fusions. Spine 1990;15:708-712.
33. Yen-Patton GP, Patton WF, Beer DM,
Jacobson BS: Endothelial cell response
to pulsed electromagnetic fields: Stim-

ulation of growth rate and angiogene-
sis in vitro. J Cell Physiol 1988;134:37-46.
34. Zoltan JD, Ryaby JT: Exogenous signal
generators: A review of the electrical
stimulation of bone. International Jour-
nal of Orthopaedic Trauma 1992;2:25-30.
35. Smith SD, McLeod BR, Liboff AR,
Cooksey K: Calcium cyclotron reso-
nance and diatom mobility. Bioelectro-
magnetics 1987;8:215-227.
36. McLeod BR, Liboff AR: Cyclotron reso-
nance in cell membranes: The theory of
the mechanism, in Blank M, Findl E
(eds): Mechanistic Approaches to Interac-
tions of Electric and Electromagnetic Fields
with Living Systems. New York, NY: Ple-
num Press, 1987, pp 97-108.
37. McLeod BR, Liboff AR, Smith SD: Elec-
tromagnetic gating in ion channels.
J Theor Biol 1992;158:15-31.
38. Fitzsimmons RJ, Ryaby JT, Magee FP,
Baylink DJ: Combined magnetic fields
increased net calcium flux in bone cells.
Calcif Tissue Int 1994;55:376-380.
39. Fitzsimmons RJ, Baylink DJ, Ryaby JT,
Magee FP: EMF-stimulated bone-cell
proliferation, in Blank M (ed): Electric-
ity and Magnetism in Biology and Medi-
cine. San Francisco, CA: San Francisco
Press, 1993, pp 899-902.
40. Fitzsimmons RJ, Ryaby JT, Mohan S,
Magee FP, Baylink DJ: Combined mag-
netic fields increase insulin-like growth
factor-II in TE-85 human osteosarcoma
bone cell cultures. Endocrinology 1995;
136:3100-3106.
41. Fitzsimmons RJ, Ryaby JT, Magee FP,
Baylink DJ: IGF-II receptor number is
increased in TE-85 osteosarcoma cells
by combined magnetic fields. J Bone
Miner Res 1995;10:812-819.
42. Ryaby JT, Fitzsimmons RJ, Khin NA, et
al: The role of insulin-like growth fac-
tor in magnetic field regulation of bone
formation. Bioelectrochem Bioenerg 1994;
35:87-91.
43. Deibert MC, Mcleod BR, Smith SD, Li-
boff AR: Ion resonance electromagnetic
field stimulation of fracture healing in
rabbits with a fibular ostectomy. J Or-
thop Res 1994;12:878-885.
44. Ryaby JT, Huene D, Magee FP, Nasser
PR: Effects of combined AC/DC mag-
netic fields on healing in a closed fem-
oral fracture model. Trans Orthop Res
Soc 1993;18:575.
45. Ryaby JT, Magee FP, Weinstein A,
Fitzsimmons RJ, Baylink DJ: Prevention
of experimental osteopenia by use of
combined AC/DC magnetic fields, in
Blank M (ed): Electricity and Magnetism
in Biology and Medicine. San Francisco,
CA: San Francisco Press, 1993, pp 807-
810.
Vol 11, No 5, September/October 2003
46. Ryaby JT, Magee FP, Haupt DL, Kinney
JH, Fitzsimmons RJ, Baylink DJ: Com-
bined treatment with magnetic fields
and parathyroid hormone to reverse
osteopenia in ovariectomized rats.
Trans Orthop Res Soc 1996;21:247.
47. Ryaby JT, Cai FF, DiDonato JA: Com-
bined magnetic fields inhibit IL-1∝ and
TNF-∝ dependent NF-kB activation in
osteoblast-like cells. Trans Orthop Res
Soc 1997;22:180.
48. Heckman JD, Ryaby JP, McCabe J, Frey
JJ, Kilcoyne RF: Acceleration of tibial
fracture-healing by non-invasive, low-
intensity pulsed ultrasound. J Bone Joint
Surg Am 1994;76:26-34.
49. Kristiansen TK, Ryaby JP, McCabe J,
Frey JJ, Roe LR: Accelerated healing of
distal radial fractures with the use of
specific, low-intensity ultrasound: A
multicenter, prospective, randomized,
double-blind, placebo-controlled study.
J Bone Joint Surg Am 1997;79:961-973.
50. Mayr E, Rudzki M-M, Rudzki M, Bor-
chardt B, Hausser H, Rüter A: Does low
intensity, pulsed ultrasound speed
healing of scaphoid fractures? [Ger-
man]. Handchir Mikrochir Plast Chir
2000;32:115-122.
51. Mayr E, Frankel V, Rüter A: Ultra-
sound: An alternative healing method
for nonunions? Arch Orthop Trauma
Surg 2000;120:1-8.
52. Chapman IV, MacNally NA, Tucker S:
Ultrasound-induced changes in rates of
influx and efflux of potassium ions in
rat thymocytes in vitro. Ultrasound Med
Biol 1980;6:47-58.
53. Parvizi J, Wu CC, Lewallen DG, Green-
leaf JF, Bolander ME: Low-intensity ul-
trasound stimulates proteoglycan syn-
thesis in rat chondrocytes by increasing
aggrecan gene expression. J Orthop Res
1999;17:488-494.
54. Ryaby JT, Mathew J, Duarte-Alves P:
Low intensity pulsed ultrasound affects
adenylate cyclase and TGF-beta syn-
thesis in osteoblastic cells. Trans Orthop
Res Soc 1992;17:590.
55. Kokubu T, Matsui N, Fujioka H, Tsun-
oda M, Mizuno K: Low intensity pulsed
ultrasound exposure increases prosta-
glandin E2 production via the induction
of cyclooxygenase-2 mRNA in mouse
osteoblasts. Biochem Biophys Res Com-
mun 1999;256:284-287.
56. Ito M, Azuma Y, Ohta T, Komoriya K:
Effects of ultrasound and 1,25-
dihydroxyvitamin D3 on growth factor
secretion in co-cultures of osteoblasts
and endothelial cells. Ultrasound Med
Biol 2000;26:161-166.
57. Friedenberg ZB, Harlow MC, Brighton
CT: Healing of nonunion of the medial
Fred R. T. Nelson, MD, et al
malleolus by means of direct current: A
case report. J Trauma 1971;11:883-885.
58. Nagai M, Ota M: Pulsating electromag-
netic field stimulates mRNA expression
of bone morphogenetic protein-2 and
-4. J Dent Res 1994;73:1601-1605.
59. Bodamyali T, Kanczler JM, Simon B,
Blake DR, Stevens CR: Effect of faradic
products on direct current-stimulated
calvarial organ culture calcium levels.
Biochem Biophys Res Commun 1999;264:
657-661.
60. Guerkov HH, Lohmann CH, Liu Y, et
al: Pulsed electromagnetic fields increase
growth factor release by nonunion cells.
Clin Orthop 2001;384:265-279.
61. Bassett CA, Mitchell SN, Gaston SR: Treat-
ment of ununited tibial diaphyseal frac-
tures with pulsing electromagnetic fields.
J Bone Joint Surg Am 1981;63:511-523.
62. Bassett CA, Mitchell SN, Schink MM:
Treatment of therapeutically resistant
non-unions with bone grafts and puls-
ing electromagnetic fields. J Bone Joint
Surg Am 1982;64:1214-1220.
63. Adams BD, Frykman GK, Taleisnik J:
Treatment of scaphoid nonunion with
casting and pulsed elctromagnetic
fields: A study continuation. J Hand
Surg [Am] 1992;17:910-914.
64. Blauvelt CT, Nelson FRT: A Manual of
Orthopaedic Terminology, ed 6. St. Louis,
MO: Mosby, 1998, p 12.
65. Brighton CT, Wang W, Seldes R, Pol-
lack SR: Signal transduction in electri-
cally stimulated bone cells. J Bone Joint
Surg Am 2001;83:1514-1523.
66. Brighton CT, Shaman P, Heppenstall RB,
Esterhai JL Jr, Pollack SR, Friedenberg
ZB: Tibial nonunion treated with direct
current, capacitive coupling, or bone graft.
Clin Orthop 1995;321:223-234.
67. Marks RA: Spine fusion for discogenic
low back pain: Outcomes in patients
treated with or without pulsed electro-
magnetic field stimulation. Adv Ther
2000;17:57-67.
68. Hanft JR, Goggin JP, Landsman A, Sur-
prenant M: The role of combined mag-
netic field bone growth stimulation as
an adjunct in the treatment of neuro-
arthropathy/Charcot joint: An expand-
ed pilot study. J Foot Ankle Surg 1998;37:
510-515.
69. Ryaby JT, Linovitz RJ, Magee FP, Faden
JS, Ponder R, Muenz LR: Abstract:
Combined magnetic fields accelerate
primary spine fusion: A double-blind,
randomized, placebo controlled study,
in Proceedings of the American Academy of
Orthopaedic Surgeons 67th Annual Meet-
ing, Orlando, FL. Rosemont, IL: Ameri-
can Academy of Orthopaedic Surgeons,
2000, vol 1, p 376.
353
Use of Physical Forces in Bone Healing

70. Azuma Y, Ito M, Harada Y, Takagi H,
Ohta T, Jingushi S: Low-intensity
pulsed ultrasound accelerates rat fem-
oral fracture healing by acting on the
various cellular reactions in the frac-
ture callus. J Bone Miner Res 2001;16:
671-680.
71. Welgus HG, Jeffrey JJ, Eisen AZ, Roswit
WT, Stricklin GP: Human skin fibro-
blast collagenase: Interaction with sub-
strate and inhibitor. Coll Relat Res 1985;
5:167-179.
72. Wu CC, Lewallen DG, Bolander ME,
Bronk J, Kinnick R, Greenleaf JF: Expo-
sure to low intensity ultrasound stimu-
lates aggrecan gene expression by cul-
tured chondrocytes. Trans Orthop Soc
1996;21:622.
73. Rawool D, Goldberg B, Forsberg F,
Winder A, Talish R, Hume E: Power
Doppler assessment of vascular chang-
es during fracture treatment with low-
intensity ultrasound. Trans Radiol Soc
North Am 1998;83:1185.

354 Journal of the American Academy of Orthopaedic Surgeons

 Thoracolumbar Fracture Management:
Anterior Approach
John S. Kirkpatrick, MD

Abstract
The surgeon who treats patients with spine trauma must be able to apply a variety Patient Evaluation
of management techniques to achieve optimal care of the patient. The anterior sur-
gical approach is appropriate for some thoracolumbar burst fractures in patients with Appropriate treatment of patients with
neurologic deficit and without posterior ligamentous injury. Surgery is most often thoracolumbar injuries is guided by
indicated for patients with incomplete deficit, especially those with a large retro- a thorough history and physical ex-
pulsed fragment, marked canal compromise, severe anterior comminution, or ky- amination, with particular attention
phosis >30°. This approach provides excellent visualization of the anterior aspect of paid to the neurologic assessment. The
the dura mater for decompression. Reconstruction of the anterior body defect can be history should include the mechanism
done with autograft, allograft, or a cage. Supplementation of the graft with anterior of injury; the presence of pain, weak-
internal fixation helps prevent kyphosis. Clinical results demonstrate improved neu- ness, and loss of sensation; and a re-
rologic function in most patients as well as low pseudarthrosis rates. In patients view of comorbidities. The physical
with incomplete deficit, improvement in neurologic function usually can be expected examination should include log roll-
with few complications. ing to allow visual inspection and pal-
J Am Acad Orthop Surg 2003;11:355-363 pation of the back and spine. Local
tenderness, swelling, gaps between
spinous processes, gibbous deformi-
ty, and ecchymosis should be noted.
The goals of thoracolumbar fracture port of the anterior and middle col- Neurologic examination of the low-
management are to preserve or re- umns of the spine, the anterior ap- er extremities and perineum is crit-
store the neurologic and biomechan- proach provides excellent exposure for ical and should include evaluation of
ical functions of the spine. Both non- reconstruction with structural grafts sensation, motor function, and reflex-
surgical and surgical management or implants. This allows restoration es. Radiographs, computed tomogra-
have a role. Optimal surgical manage- of height and correction of kyphosis phy (CT), and, on occasion, magnet-
ment of thoracolumbar fractures re- while limiting the number of motion ic resonance imaging help delineate
quires understanding the patient’s segments fused. This is especially use- the nature of the fracture and extent
clinical situation, the fracture classi- ful in patients whose general condi-
fication, and the strengths and weak- tion prevents posterior reduction in
nesses of a variety of approaches and the first 7 to 10 days after injury. The Dr. Kirkpatrick is Associate Professor, Division
stabilization techniques. anterior approach also avoids addi- of Orthopaedic Surgery, University of Alabama
The anterior approach can be used tional injury to the paraspinal mus- at Birmingham, Birmingham, AL.
for both management of the neuro- cles and disruption of their innerva-
Neither Dr. Kirkpatrick nor the department with
logic deficit and restoration of stabil- tion. However, comparative studies
which he is affiliated has received anything of val-
ity to the spine. In most patients, neu- between anterior and posterior ap- ue from or owns stock in a commercial company
rologic deficit is caused by impact proaches are limited. Thus, it is dif- or institution related directly or indirectly to the
and/or compression to the ventral sur- ficult to present objective evidence that subject of this article.
face of the spinal cord. The anterior one approach is better than the oth-
Reprint requests: Dr. Kirkpatrick, 940 Faculty
approach provides optimal direct ex- er, especially in the first 7 to 10 days
Office Tower, 510 20th Street South, Birmingham,
posure for visualization of the ven- after injury. Additionally, nonsurgical AL 35294-3409.
tral aspect of the dura mater during management as well as anterior and
surgical decompression.1 Additional- posterior approaches each have their Copyright 2003 by the American Academy of
ly, for fracture patterns involving unique role in the treatment of patients Orthopaedic Surgeons.
marked comminution with loss of sup- with thoracolumbar spine injury.

Vol 11, No 5, September/October 2003 355

Thoracolumbar Fracture Management: Anterior Approach
of injury.2 Magnetic resonance imag-
ing is indicated in the presence of an
unexplained neurologic deficit, pro-
gressive deterioration, or notable soft-
tissue injury.
Patients with fracture in the pres-
ence of multisystem trauma require
special consideration. Blood loss can
be significant (up to 1,500 mL) with
the anterior approach, which can lead
to coagulopathy and hypoperfusion
in the presence of other injuries. Both
the transdiaphragmatic and the tho-
racic approaches can temporarily im-
pair pulmonary function, leading to
marked pulmonary compromise, es-
pecially in the presence of closed
chest injury. Intra-abdominal injury
can lead to notable peritoneal disten-
sion, making retroperitoneal expo-
sure difficult. Patients with multiple
extremity and/or pelvic fractures
may be better served by reduction
and stabilization of unstable thora-
columbar fractures through a poste-
rior approach, thus allowing for more
efficient care of the other injuries.
Assessment of these factors, in ad-
dition to fracture pattern and neuro-
logic status, are critical to making
the appropriate treatment selection
for the patient with thoracolumbar
trauma.
Classification of Injury
The mechanistic fracture classifica-
tion, as modified by McAfee et al,2 is
the preferred thoracolumbar fracture
classification system because it can
provide important insight into reduc-
tion mechanisms and stabilization
needs as well as guide the surgeon
to the appropriate surgical proce-
dure.3 This classification scheme com-
prises compression, stable burst, and
unstable burst fractures; flexion-
distraction injury; and fracture-dis-
location. McAfee et al2 emphasized
the importance of Denis’ middle col-
umn and classified fractures as fail-
ure of the middle column in compres-
sion, distraction, or translation.
356
Indications Controversy exists about whether
patients with complete neurologic in-
The anterior approach is most com- juries warrant decompression in ad-
monly indicated for an unstable dition to stabilization. Patients with
burst fracture from T10 to L3 (al- complete thoracic level paraplegia
though it can be used up to T5) (above T10) have a poor prognosis for
associated with an incomplete neu- recovery and generally are best treat-
rologic deficit and radiologically ed with posterior stabilization. Pa-
demonstrated neural compres- tients with thoracolumbar junction or
sion.1,4 Recent studies5,6 have refined lumbar injuries who initially appear
the radiologic and clinical features to be neurologically complete but are
of patients with incomplete deficits still in spinal shock can be treated
in whom the anterior approach is in- more aggressively with prompt sur-
dicated. These features may include gery. No data indicate a difference in
a large retropulsed fragment with recovery between anterior or poste-
marked (>67%) canal compromise, rior approach when adequate decom-
anterior comminution and kyphosis pression is obtained. Because of a lack
>30°, and time of more than 4 days of definitive studies demonstrating a
from injury.5 Patients with incom- clear advantage of one approach over
plete fracture reduction after a pos- the other, both are considered for
terior approach may be candidates management in appropriate circum-
for anterior decompression if neuro- stances. A summary of relative indi-
logic recovery is incomplete and re- cations for the anterior approach is
sidual compression persists. listed in Table 1.
Parker et al6 reviewed their insti- The anterior approach is more dif-
tution’s experience with the load- ficult for low lumbar fractures (ie, L4
sharing classification, which is and L5) because of anatomic con-
based on the extent of comminution straints, especially for restoring align-
and displacement. Patients who had ment and attaining satisfactory fixa-
fractures with a high degree of com- tion. However, these injuries rarely
minution, displacement, and kypho- require surgery and generally should
sis had a better result with anterior be managed posteriorly in patients in
stabilization because it provided whom stabilization is required. Be-
greater restoration of anterior col- cause of the large ratio of canal area
umn support than did short-seg- to neural element, low lumbar frac-
ment pedicle screw fixation. Others tures behave differently from upper
have found that longer posterior fu- lumbar fractures and generally do
sion constructs provide adequate well with nonsurgical management,
stability. except in cases of instability. For pa-
Table 1
Indications for Anterior Approach
Incomplete paraplegia with stable or unstable burst fracture
Fractures with poor reduction potential with posterior approach
Large retropulsed fragment with >67% compromise5
Anterior comminution with kyphosis >30°5
>4 days since injury5
Fractures with inadequate canal reduction and incomplete neurologic
recovery after posterior stabilization and suboptimal neural recovery
Reconstruction of anterior column after short-segment posterior stabilization
Traumatic disk herniation with flexion-distraction injury
Journal of the American Academy of Orthopaedic Surgeons

tients with instability, the posterior
approach is preferable.
Anterior decompression is rarely
used alone for injuries other than
burst fractures. Most other fracture
types are either well treated nonsur-
gically (ie, compression fractures) or
managed with a posterior approach
to restore the integrity of the poste-
rior elements and prevent kyphosis
(ie, flexion-distraction injuries). In cer-
tain limited circumstances, these in-
juries may require an adjunctive an-
terior decompression because of
herniated disk, marked comminution
of the middle column, or concern
about additional displacement of
fragments into the canal. In such cas-
es, a combined approach should be
considered. Most fracture-disloca-
tions, because of their extreme insta-
bility, are best managed with a pos-
terior approach. Spinal stability of
fracture-dislocations after anterior de-
compression, even with internal fix-
ation, is insufficient; thus, the anteri-
or approach should be avoided.
Contraindications
Preexisting medical conditions and
concurrent traumatic injuries to the
abdomen and chest may represent
relative contraindications to the an-
terior approach. Patients with severe
pulmonary disease may have limit-
ed reserve for pulmonary function
and may not tolerate thoracic or tho-
racoabdominal approaches. Severe
chest or abdominal injuries also may
limit pulmonary reserve or impair ex-
posure. Marked osteoporosis in ad-
jacent vertebrae may result in impac-
tion of the strut graft and failure of
screw purchase, leading to nonunion
and/or kyphosis. Morbid obesity
may impair exposure and lead to in-
adequate visualization for safe de-
compression. When these conditions
are present, the surgeon must balance
the relative merits of anterior and
posterior approaches for the specific
fracture.
Vol 11, No 5, September/October 2003
Timing
The timing of decompression remains
controversial, with a divergence be-
tween the results of animal studies
and clinical reports. Some basic sci-
ence data indicate that early decom-
pression results in improved neuro-
logic recovery. In surgically created
acute cord compression in a canine
model, neurologic recovery was bet-
ter when release was done within 1
hour rather than after longer time pe-
riods.7 Carlson et al8 studied region-
al blood flow, the interface pressure
between the spinal cord and a com-
pressing piston, and somatosensory
evoked potentials in a canine model
with decompression at 5 minutes and
with no decompression. Regional
blood flow returned to normal with-
in 3 hours after onset of compression.
The viscoelasticity of the cord allowed
the interface pressure to decrease to
<20% of the maximum in the first
hour of compression and was approx-
imately 10% by 3 hours. Despite these
changes, somatosensory evoked po-
tentials did not show improvement,
indicating the multifactorial nature of
spinal cord injury. The authors sug-
gested that sustained displacement
initiated a secondary phase of phys-
iologic events. Acorrelation of evoked
potential recovery with regional
blood flow during compression was
later reported, supporting the concept
of an ischemic mechanism of second-
ary injury.9 These factors led to the im-
pression that there is a limited win-
dow of opportunity for obtaining
optimal neurologic recovery in such
injuries. Unfortunately, this window
appears to be too brief to allow clin-
ical rescue, resuscitation, diagnosis,
and induction of anesthesia for urgent
decompression. Recovery of cauda
equina or root injuries does not ap-
pear to be as time-dependent.
Clinical data about the timing of
cord decompression are limited and
mostly relate to cervical injuries. Ear-
ly closed reduction of cervical sublux-
ation has been shown to be possible
John S. Kirkpatrick, MD
as early as 2 hours after injury, with
no neurologic deterioration directly
attributed to traction and reduc-
tion.10 Comparisons of early and late
decompression are few. Vaccaro et
al11 did a prospective, randomized
study of surgery done early (<72
hours) versus late (>5 days) after spi-
nal cord injury. They found no signif-
icant benefit to neurologic function,
length of stay, or length of rehabili-
tation when the surgery was done
early. Authors of a more recent ret-
rospective study comparing experi-
ence at two different institutions
found that early surgery (within 72
hours of injury) was not associated
with a higher complication rate.12
They also suggested that the early
surgery group may have had im-
proved neurologic recovery in spite
of early closed reduction in both
groups. This difference could be ex-
plained by variations in surgeons,
sites, methods of neurologic evalua-
tion, and preoperative function, but
it warrants further study. Although
these studies involve cervical injuries,
they do provide some insight into the
issues related to the timing of surgery
for cord injuries. To date, there is no
clear difference in outcomes based on
timing of surgery.
The timing and use of the anterior
approach in multiply injured patients
requires coordination between trau-
ma surgeons and spine surgeons.
Acute life-threatening injuries, such
as unstable pelvis fractures, head,
chest, and abdominal injuries, and
limb-threatening injuries, such as
open long bone fractures, should be
handled first, followed by spine care.
It is important to consider early de-
compression of incomplete spinal
cord injuries within 24 to 48 hours as
an emergent or urgent procedure, be-
fore the typical onset of pulmonary
complications resulting from the pa-
tient’s injuries. Early decompression
has been shown to be safe and effec-
tive, with the major difference be-
tween urgent and early treatment
being the extent of blood loss.13
357
Thoracolumbar Fracture Management: Anterior Approach
Management of injuries involving the
cauda equina or root deficits are
planned as soon as the patient’s over-
all condition is satisfactory, usually
within the same time period, but oc-
casionally up to 7 to 10 days in pa-
tients with life-threatening chest and
abdominal injuries.
Patients initially treated nonsurgi-
cally or with a posterior approach
who have persistent cord compres-
sion may be candidates for late de-
compression and may obtain clinical
improvement. The anterior approach
can be used months or years after ini-
tial injury.1 Bohlman et al14 studied
patients treated for late pain and/or
paralysis a mean of 4.5 years after in-
jury (range, 3 months to 21 years).
They noted improvement in pain for
41 of 45 patients (91%) and improve-
ment in neurologic function in 21 of
25 patients (84%).
Reconstruction
Patients with burst fractures have fail-
ure of the anterior column in com-
pression, producing a kyphotic defor-
mity and inability of the spine to resist
axial load. When associated with loss
of the posterior column tension band,
the result is an extremely unstable
spinal injury. Anterior decompression
of the neural elements further desta-
bilizes the spinal column by remov-
ing whatever anterior support re-
mains. Thus, the primary principle
for reconstruction after anterior de-
compression is restoration of the an-
terior column so that it can resist ax-
ial compression. If the posterior
tension band also has failed, poste-
rior stabilization may be required, as
well.
Reconstruction generally involves
two components: immediate stabili-
ty and restoration of normal align-
ment. Immediate stability can be ob-
tained with a variety of devices, such
as cages, rod-and-screw or plate-and-
screw constructs, and external brac-
es. Iliac crest autograft was initially
358
used for the strut graft, but more re-
cently, tibial or humeral allograft has
been used.15 Ventral cages containing
cancellous autograft have been devel-
oped, but there have been few stud-
ies of their efficacy. Long-term stabil-
ity can be achieved with fusion of the
strut graft.
Implant types commonly used an-
teriorly for fracture indications in-
clude both rigid and nonrigid plate-
and-screw and rod constructs. Most
studies report the use of either rigid
screw and rod constructs16,17 or rigid
plate constructs.18 Semirigid or dy-
namized constructs using screws and
rods have been reported with satis-
factory outcomes.19 Implants should
be placed laterally to avoid contact
with the aorta because such contact
has been reported to cause late vas-
cular disruption and death.20 In ad-
dition, to avoid problems with the il-
iac vessels, anterior plate-and-screw
or rod-and-screw constructs should
not be used below L4.
Biomechanics of Anterior
Reconstruction
The biomechanics of anterior recon-
struction have been studied in a va-
riety of models, including animal
(both in vivo and in vitro), cadaveric,
and biomechanical synthetic. Devic-
es used anteriorly can be divided into
two categories: interpositional and
splinting. Interpositional devices,
which substitute for the anterior and
middle columns, are usually biolog-
ic (eg, iliac crest strut). Splinting de-
vices are used to stabilize the con-
struct during biologic incorporation
of the interpositional device. The mul-
tifactorial nature of the biomechan-
ical properties of these devices makes
comparison of different studies dif-
ficult. Construct strength and stiffness
are affected by biologic variability in
patient size and bone density, pur-
chase strength of the anchors (usual-
ly screws), load sharing with the graft,
and the mechanical properties of the
Journal of the American Academy of Orthopaedic Surgeons
implants. Biologic variability between
individual patients must be consid-
ered when planning reconstruction
after corpectomy. Although little has
been published about the purchase
strength of anchors, bicortical pur-
chase has been shown to increase
screw pullout strength over unicor-
tical purchase; however, the effect is
less pronounced when bone mineral
density is low.21 Synthetic models
standardized by the American Soci-
ety for Testing and Materials provide
a method for comparing implant stiff-
ness and fatigue strength without the
confounding variables introduced by
biologic variability, specimen avail-
ability, and anchor purchase fail-
ure.22
The load sharing of the graft may
contribute to stability, depending on
the construct chosen. Lee et al23 re-
ported differences in stability testing
depending on the anterior reconstruc-
tion method used. They compared a
polymethylmethacrylate block, iliac
crest bone graft, two small cages, and
one large cage. The large cage was su-
perior in axial rotation and sagittal
motion, and the two small cages and
iliac crest bone graft were superior in
lateral bending. The PMMA block
was approximately the same as iliac
crest bone graft in all modes tested
except lateral bending, where it was
inferior.
Many surgeons use internal fixa-
tion in addition to anterior column re-
construction. The properties of the in-
terpositional and internal fixation
devices may be combined for im-
proved overall stability. In an older
in vitro study, the Kaneda device
(DePuy Acromed, Raynham, MA,
formerly Acromed) was found to re-
store torsional stiffness as well as a
posterior Cotrel-Dubousset con-
struct24 and better than Harrington
rods or the AO fixateur interne (nei-
ther of which is used anymore).25 Oth-
er authors emphasized the impor-
tance of connecting parallel rods used
in anterior or posterior constructs.26
In a canine study, the fusion rate was
 John S. Kirkpatrick, MD

higher with the Kaneda device and
graft (86%) than with graft alone
(29%) at 24 weeks (P = 0.028).27 Ax-
ial, flexural, and torsional stiffness
were tested; only torsional stiffness
was found to be significantly (P <
0.05) stronger in the instrumented fu-
sions. Other studies have included
comparisons of different devices in
animal or cadaveric in vitro models,
with varying results. The models us-
ing stability testing favor the Kaneda
device over plate-and-screw con-
structs in general, especially in
torsion.28-30 Results of stability testing
using a synthetic model have favored
plate over rod constructs, with the ex-
ception of the Z-plate (ZPLATE-ATL;
Medtronic Sofamor-Danke, Mem-
phis, TN), which tends to be the least
stiff of all constructs tested.31 The use
of synthetic models tends to favor
screw anchors over bolt-type anchors,
which may have had some effect on
the results.
The material used in manufactur-
ing the implant affects its stiffness and
strength. There are few direct com-
parison data. Kotani et al31 compared
Kaneda devices made of titanium and
of stainless steel. These devices are as-
sumed to have similar dimensional
tolerances. The titanium implant was
found to have both greater bending
strength and higher stiffness than the
stainless steel implant.
The importance of subtle differenc-
es in stiffness remains to be seen. Clin-
ical studies of all devices demonstrate
high fusion rates and good perfor-
mance, probably making the subtle
differences in biomechanical proper-
ties unimportant. In addition, the
stiffness needed in each plane appears
to be different. One clinical study with
an implant construct that allowed dy-
namic axial compression showed that
no patients had pseudarthrosis, and
kyphosis worsened by only 4° at long-
term follow-up (mean, 42 months;
range, 24 to 84 months).19 Although
the required degree of stiffness re-
mains unclear, it appears that any re-
construction technique that provides
rotational stiffness equal to or great-
er than that of the intact spine will
provide a stable construct that leads
to fusion.
In young patients with qualitative-
ly good bone quality, a biologic strut
(eg, iliac crest bone graft) is optimal
for anterior column support. Alterna-
tively, an appropriately sized allograft
can be used. When preparing the end
plates, a curette should be used to re-
move the cartilaginous end plate to
bleeding bone. This removes less of
the subchondral bone and thus reduc-
es the chance of graft settling. If the
bone quality is poor, or if the patient
prefers to avoid the discomfort of
graft harvest, a tibial allograft pro-
vides a broad base of contact between
the graft and vertebrae. Either type
of graft can be supplemented with
cancellous bone from the vertebrec-
tomy. A low-profile plate construct to
augment the reconstruction is placed
laterally, away from the aorta, avoid-
ing the potential for late erosion of the
aorta (Fig. 1).
Clinical Results
Because of the highly variable nature
of thoracolumbar burst fractures and
the lack of randomized prospective
studies, clinical results after direct an-
terior decompression are difficult to

Figure 1 A, Lateral radiograph of a 48-year-old man who sustained an L2 burst fracture associated with paraparesis. The patient had grade
4 quadriceps strength and dysesthetic pain in his thighs. B, Axial CT scan demonstrates >67% canal compromise and comminution ante-
riorly. Anteroposterior (C) and lateral (D) radiographs after anterior decompression and application of tibial allograft with lumbar plate.
The patient had resolution of leg pain and normal strength after the decompression and returned to full function and work as a contractor.

Vol 11, No 5, September/October 2003 359

Thoracolumbar Fracture Management: Anterior Approach
compare with those of other tech-
niques. However, McAfee et al1 re-
ported that 37 of 42 patients treated
with anterior decompression at a
mean of 60 days after initial injury
had some degree of neurologic im-
provement. Of the 37 patients, 30 pre-
operatively had motor strength of
grade 3 or less. Fourteen of these 30
patients became community ambula-
tors; 9 others regained function ad-
equate for household ambulation, al-
though some required short leg
braces and/or crutches. Radiograph-
ic results indicated that 12 of the 42
patients developed kyphosis >20°
postoperatively. Kaneda et al16 report-
ed a series of 150 patients with burst
fractures and showed comparable re-
sults in 78 with neurologic deficit who
underwent anterior decompression
supplemented with anterior fixation.
The time from injury to decompres-
sion varied from <48 hours to >1 year.
Seventy-two percent of the neurolog-
ically compromised patients (56/78)
recovered completely. Eighty-six per-
cent of all patients who had been em-
ployed preinjury (112/130) returned
to their jobs.16 No patient in either
study1,16 was made neurologically
worse, and those with the most in-
complete deficits recovered one
Frankel grade or more. In both stud-
ies, CT done after decompression
demonstrated adequate decompres-
sion in all but two cases.
Comparative studies between an-
terior and posterior approaches are
few, use multiple surgical techniques,
and have relatively small numbers of
patients in each group. Esses et al32
compared 18 patients treated with an-
terior decompression, iliac strut, and
Kostuik-Harrington anterior fixation
with 22 patients who had posterior
distraction instrumentation with the
fixateur interne and posterolateral fu-
sion. Postoperative CT demonstrated
better canal decompression with the
anterior approach, but this did not
correlate with neurologic recovery,
which was no different between the
groups. Schnee and Ansell33 com-
360
pared 14 patients treated with ante-
rior decompression, allograft strut,
and plate fixation with 9 patients who
underwent combined anterior de-
compression and posterior fixation
and 2 patients who had transpedic-
ular decompression, fusion, and fix-
ation. Choice of technique apparent-
ly was related to the severity of injury
to the spinal column. The authors
concluded that anterior decompres-
sion is critical to success in manag-
ing fractures with significant verte-
bral destruction. Been and Bouma34
studied 27 patients treated with an-
terior decompression and iliac strut
combined with posterior fixation and
compared them with 19 patients treat-
ed with the AO fixateur interne for
posterior distraction and stabiliza-
tion. There was no statistical differ-
ence between the two groups; neuro-
logic recovery of more than one
Frankel grade occurred in 10 of 10 pa-
tients in the anterior group and 7 of
8 in the posterior group. Bladder re-
covery was obtained in 3 of 7 in the
anterior group and in 1 of 3 in the pos-
terior group. Pain relief occurred in
85% of the anterior group and 79%
of the posterior group. Complications
occurred in 15% of the anterior group
and 26% in the posterior group.
Surgical Technique
The right-sided approach is used for
upper thoracic burst fractures to
avoid the aortic arch, common carot-
id artery, esophagus, and trachea.
Thoracic fractures (T6 through T11)
can be approached from the right, al-
though a left-sided approach is pos-
sible because the aorta is easily mo-
bilized. The left side is preferred for
thoracolumbar fractures requiring a
thoracoabdominal (T12-L1) or retro-
peritoneal (L2-3) approach; this
avoids the approach being obscured
by the liver or having to mobilize the
vena cava. However, a right-sided ap-
proach can be used for thoracolum-
bar fractures, if necessary. Lower lum-
Journal of the American Academy of Orthopaedic Surgeons
bar fractures (L4 and L5) usually are
approached posteriorly, although
rarely they may require reconstruc-
tion of the anterior column through
a retroperitoneal (L4) or transperito-
neal (L5) approach.
The patient should be placed in the
true lateral position to help the sur-
geon maintain orientation of the ver-
tebral body. The region of the fracture
is positioned over the break in the ta-
ble because flexing the table will im-
prove exposure. For a fracture at the
thoracolumbar junction (T12-L2), an
oblique incision is made either along
the 12th rib (T12-L1) or just inferior
to it (L1-L2) extending toward the
umbilicus for a retroperitoneal ap-
proach. For T11 and some T12 frac-
tures, especially when internal fixa-
tion is planned, a 10th or 11th rib
approach is used. Special attention is
required in these exposures to remain
extrapleural, and meticulous repair is
needed when the diaphragm is in-
cised. When such an incision is need-
ed, it is important to leave approxi-
mately 1 to 2 cm of the periphery
attached to the chest wall for later re-
pair. Sutures may be placed to mark
normal anatomic locations around
the periphery to aid anatomic closure
after reconstruction. For a 12th rib ex-
posure, the rib is excised and the ret-
roperitoneum identified where the
transversalis fascia, pleura, and dia-
phragm meet near the tip of the 12th
rib. Incision of the abdominal mus-
cles provides access to the retroperi-
toneum. The peritoneum, retroperi-
toneal fat, and kidneys are reflected
anteriorly using blunt dissection,
thereby exposing the quadratus lum-
borum and psoas major muscles. The
level of the fracture is identified and
the psoas gently elevated from the an-
terior portion of the vertebral body.
The segmental vessels then are iden-
tified between the disk spaces on the
vertebral body and are ligated and di-
vided. Subperiosteal dissection of
these structures allows exposure
nearly to the opposite pedicle. A
Finochietto rib retractor or other self-

retaining retractor is placed between
the 11th rib and the iliac crest.
Decompression is accomplished
with the aid of loupe magnification
and a fiberoptic headlight. Diskecto-
my is done above and below the frac-
tured vertebra, followed by corpec-
tomy for the decompression.1 Bone
removed with rongeurs is saved for
later use to supplement the strut graft
with cancellous bone. Direct visual-
ization of the dura through the pedi-
cle resection and the diskectomy sites
helps avoid premature penetration of
the posterior wall of the vertebra and
injury to the dura. The transverse
width of the vertebral body, as noted
through the diskectomy sites, serves
as a guide for the extent of the decom-
pression. The medial wall of the con-
tralateral pedicle is the landmark for
the adequacy of the decompression.
A common mistake is not to decom-
press across the vertebral body to the
contralateral pedicle. The dura will
resume its normal contour after de-
compression.
Some modifications of this tech-
nique are needed for thoracic frac-
tures. A double-lumen endotracheal
tube often is used because some pa-
tients do not tolerate packing of the
lung for exposure. The rib above the
fracture can be used as the landmark
for the approach. Some surgeons re-
move the rib; others use the costal in-
terspace for the exposure. Removal of
the rib has the disadvantage of in-
creased pain, but it provides some lo-
cal autograft and a much wider ex-
posure. The rib head attachment to
the fractured vertebra is removed to
expose the pedicle and foramen.
If the spine is not already in ad-
equate sagittal alignment after decom-
pression, pressure on the skin poste-
riorly at the level of injury and/or
distraction within the corpectomy de-
fect, using a large laminar spreader
or implant instrumentation, can help
restore normal alignment. The verte-
bral end plates should be prepared by
removing the cartilaginous end plate
with a curette or bur. A compromise
Vol 11, No 5, September/October 2003
must be made between obtaining ad-
equate bleeding bone for vascular sup-
ply to the graft and removing so much
of the subchondral bone that the me-
chanical support for the graft is de-
creased. Some surgeons prefer to make
indentations or seating holes into the
vertebral bodies to accommodate the
ends of the graft, but generally this
should be combined with additional
support (eg, internal fixation) to pre-
vent graft impaction and kyphosis.
Measurements are made for the height
and width of the graft, which then are
applied to shaping the iliac crest au-
tograft, allograft tibia or femur, or cage
device. The graft is impacted into po-
sition with direct visualization of the
dura to avoid impingement. Placing
the tricortical portion of the graft on
the contralateral side can help prevent
settling and coronal deformity when
internal fixation is used. The break in
the operating table is then removed,
thereby eliminating the lateral bend-
ing induced in the spine by position-
ing the patient. This tends to lock the
graft in place and prevents the spine
from being left with a coronal plane
deformity. If implants are being used,
they are applied according to man-
ufacturer recommendations. After in-
strumentation, a check should be done
to make sure the hardware does not
impinge on vascular or visceral struc-
tures.
Anatomic closure of the wound is
done after insertion of an appropri-
ate drain or thoracostomy tube. If the
pleural cavity was involved in the ex-
posure, clinical and radiographic
monitoring for pneumothorax should
be done postoperatively. Thoracosto-
my tubes generally are removed
when there is no pneumothorax on
radiograph and no air leak and when
drainage has subsided. In patients
who develop a postoperative ileus,
diet should be restricted and/or na-
sogastric suction used until the return
of bowel sounds. External support
with a total-contact or thoracolum-
bosacral orthosis often is used, and
ambulation is begun once brace fit-
John S. Kirkpatrick, MD
ting is accomplished (usually 48
hours postoperatively). Progressive
exercises for rehabilitation of lumbar
and abdominal muscles are begun ap-
proximately 3 months postoperative-
ly and, when necessary, followed by
work hardening programs.
Complications
Causes of complications can be
grouped into three general categories:
surgical approach, decompression,
and structural (reconstruction).35
Complications may occur as a result
of using the retroperitoneal or thora-
coabdominal approach to the spine.
Pneumothorax, recognized intraoper-
atively and on postoperative radio-
graphs, is managed with insertion of
a chest tube. Atelectasis, and occa-
sionally postoperative pneumonia,
can occur and appears in the con-
tralateral or dependent lung in 3% to
5% of patients.35 Superficial and deep
wound infections are rare; most re-
spond to antibiotic therapy. Infections
unresponsive to initial antibiotic ther-
apy may require surgical débride-
ment, usually with retention of the re-
construction. Genitofemoral nerve
injury, nerve root (eg, lumbar plexus)
traction injury, and injury to the sym-
pathetic plexus occur in approximate-
ly 2% to 4% of patients.1,16,35 Intraop-
erative laceration of the inferior vena
cava has been reported. Blood loss is
variable, but the need for transfusion
should be anticipated. Ileus is com-
mon with retroperitoneal approach-
es but generally resolves within 24
hours. Reported, but rare, complica-
tions include peritoneal entry, dam-
age to the ureter, interruption of
lymphatic channels with resulting
chylothorax or chylous leak, and
splenic rupture. Late complications
from the retroperitoneal approach
also may include incisional hernia
and permanent abdominal swelling
on the side of the approach.
Complications related to decom-
pression are relatively uncommon.
361
Thoracolumbar Fracture Management: Anterior Approach
Iatrogenic neurologic injury is not re-
ported in major series, likely because
of the safety resulting from direct an-
terior visualization of the thecal sac.
Iatrogenic dural lacerations should be
isolated and closed, if possible. Sub-
arachnoid drainage should be consid-
ered for persistent cerebrospinal flu-
id leak.
Kyphosis and pseudarthrosis are
the main structural complications of
anterior arthrodesis. Without internal
fixation, kyphosis occurs in approx-
imately 25% of patients.4 This has not
been felt to be detrimental to the re-
covery of neurologic function because
the kyphosis results from settling of
the graft without compression of the
neural elements. Instrumentation can
reduce this rate of kyphosis to be-
tween 5% and 10%.29 Rates of pseud-
arthrosis are generally 5% to 10% and
appear to be at the lower end of this
range when fixation is used.16,17,32,35
Pain at the iliac crest donor site is re-
References
1. McAfee PC, Bohlman HH, Yuan HA:
Anterior decompression of traumatic
thoracolumbar fractures with incom-
plete neurological deficit using a retro-
peritoneal approach. J Bone Joint Surg
Am 1985;67:89-104.
2. McAfee PC, Yuan HA, Fredrickson BE,
Lubicky JP: The value of computed to-
mography in thoracolumbar fractures:
An analysis of one hundred consecu-
tive cases and a new classification.
J Bone Joint Surg Am 1983;65:461-473.
3. Mirza SK, Mirza AJ, Chapman JR,
Anderson PA: Classifications of thorac-
ic and lumbar fractures: Rationale and
supporting data. J Am Acad Orthop Surg
2002;10:364-377.
4. Bohlman HH: Treatment of fractures
and dislocations of the thoracic and
lumbar spine. J Bone Joint Surg Am 1985;
67:165-169.
5. McCullen G, Vaccaro AR, Garfin SR:
Thoracic and lumbar trauma: Rationale
for selecting the appropriate fusion
technique. Orthop Clin North Am 1998;
29:813-828.
6. Parker JW, Lane JR, Karaikovic EE, Gaines
RW: Successful short-segment instrumen-
tation and fusion for thoracolumbar spine
362
ported in approximately 5% of cases,
although this may be an artificially
low estimate.15,17,35
Implant complications are device-
or technique-related.18 Device-related
complications include screw or bolt
breakage and rod or interconnection
failure, both of which are associated
with progressive kyphosis. In one se-
ries using the first-generation Kane-
da device in 20 patients, three screw
failures and one pseudarthrosis
occurred.17 Implant technique com-
plications are rarely reported, but
concern exists regarding canal pen-
etration and vascular injury to the
vessels on the contralateral side. In-
adequate exposure of the superior
end vertebra may prevent proper
placement of the screw or bolt flush
against the vertebral body, complicat-
ing proper placement of the plate. Al-
though the prominence of implants
is a potential problem in the thoracic
spine, the anterior position of the aor-
fractures: A consecutive 4½-year series.
Spine 2000;25:1157-1170.
7. Delamarter RB, Sherman J, Carr JB:
Pathophysiology of spinal cord injury:
Recovery after immediate and delayed
decompression. J Bone Joint Surg Am
1995;77:1042-1049.
8. Carlson GD, Warden KE, Barbeau JM,
et al: Viscoelastic relaxation and region-
al blood flow response to spinal cord
compression and decompression. Spine
1997;22:1285-1291.
9. Carlson GD, Gorden CD, Nakazowa S,
Wada E, Warden K, LaManna JC:
Perfusion-limited recovery of evoked
potential function after spinal cord in-
jury. Spine 2000;25:1218-1226.
10. Grant GA, Mirza SK, Chapman JR, et al:
Risk of early closed reduction in cervi-
cal spine subluxation injuries. J Neuro-
surg 1999;90(1 suppl):13-18.
11. Vaccaro AR, Daugherty RJ, Sheehan TP,
et al: Neurologic outcome of early ver-
sus late surgery for cervical spinal cord
injury. Spine 1997;22:2609-2613.
12. Mirza SK, Krengel WF III, Chapman JR,
et al: Early versus delayed surgery for
acute cervical spinal cord injury. Clin
Orthop 1999;359:104-114.
Journal of the American Academy of Orthopaedic Surgeons
ta and the overlying psoas major
muscle eliminates this concern at the
thoracolumbar junction.
Summary
The anterior approach for thora-
columbar fractures may be preferred
in patients with incomplete neurolog-
ic deficit from burst fractures with-
out substantial posterior element in-
jury. Excellent visualization of the
anterior dura mater allows safe de-
compression and leads to some de-
gree of neurologic recovery in most
patients. Reconstruction generally in-
cludes the use of iliac crest strut graft,
cages, or allograft. Supplementation
with internal fixation can improve
biomechanical stability and may lead
to improved fusion rates and reduc-
tion in ultimate kyphosis. Complica-
tions are rare and do not generally af-
fect long-term outcome.
13. McLain RF, Benson DR: Urgent surgi-
cal stabilization of spinal fractures in
polytrauma patients. Spine 1999;24:
1646-1654.
14. Bohlman HH, Kirkpatrick JS, Delamar-
ter RB, Leventhal M: Anterior decom-
pression for late pain and paralysis af-
ter fractures of the thoracolumbar
spine. Clin Orthop 1994;300:24-29.
15. Finkelstein JA, Chapman JR, Mirza S:
Anterior cortical allograft in thora-
columbar fractures. J Spinal Disord 1999;
12:424-429.
16. Kaneda K, Taneichi H, Abumi K, Hash-
imoto T, Satoh S, Fujiya M: Anterior de-
compression and stabilization with the
Kaneda device for thoracolumbar burst
fractures associated with neurological
deficits. J Bone Joint Surg Am 1997;79:
69-83.
17. Kirkpatrick JS, Wilber RG, Likavec M,
Emery SE, Ghanayem A: Anterior sta-
bilization of thoracolumbar burst frac-
tures using the Kaneda device: A prelim-
inary report. Orthopedics 1995;18:673-678.
18. Ghanayem AJ, Zdeblick TA: Anterior
instrumentation in the management of
thoracolumbar burst fractures. Clin Or-
thop 1997;335:89-100.

19. Carl AL, Tranmer BI, Sachs BL: Antero-
lateral dynamized instrumentation and
fusion for unstable thoracolumbar and
lumbar burst fractures. Spine 1997;22:
686-690.
20. Jendrisak MD: Spontaneous abdominal
aortic rupture from erosion by a lumbar
spine fixation device: A case report.
Surgery 1986;99:631-633.
21. Breeze SW, Doherty BJ, Noble PS,
LeBlanc A, Heggeness MH: A biome-
chanical study of anterior thoracolum-
bar screw fixation. Spine 1998;23:1829-1831.
22. Standard test methods for static and fa-
tigue for spinal implant constructs in a
corpectomy model. Annual Book of
ASTM Standards 1996;13.01:1097-1112.
23. Lee SW, Lim TH, You JW, An HS: Bio-
mechanical effect of anterior grafting
devices on the rotational stability of spi-
nal constructs. J Spinal Disord 2000;13:
150-155.
24. Gurr KR, McAfee PC, Shih CM: Biome-
chanical analysis of anterior and poste-
rior instrumentation systems after cor-
pectomy: A calf-spine model. J Bone
Joint Surg Am 1988;70:1182-1191.
Vol 11, No 5, September/October 2003
25. Shono Y, McAfee PC, Cunningham
BW: Experimental study of thora-
columbar burst fractures: A radio-
graphic and biomechanical analysis of
anterior and posterior instrumentation
systems. Spine 1994;19:1711-1722.
26. Gaines RW Jr, Carson WL, Satterlee CC,
Groh GI: Experimental evaluation of seven
different spinal fracture internal fixation
devices using nonfailure stability testing:
The load-sharing and unstable-mechanism
concepts. Spine 1991;16:902-909.
27. Zdeblick TA, Shirado O, McAfee PC,
deGroot H, Warden KE: Anterior spinal
fixation after lumbar corpectomy: A
study in dogs. J Bone Joint Surg Am 1991;
73:527-534.
28. Zdeblick TA, Warden KE, Zou D,
McAfee PC, Abitbol JJ: Anterior spinal
fixators: A biomechanical in vitro study.
Spine 1993;18:513-517.
29. An HS, Lim TH, You JW, Hong JH, Eck
J, McGrady L: Biomechanical evaluation
of anterior thoracolumbar spinal instru-
mentation. Spine 1995;20:1979-1983.
30. Hitchon PW, Goel VK, Rogge T, Gros-
land NM, Torner J: Biomechanical stud-
John S. Kirkpatrick, MD
ies on two anterior thoracolumbar im-
plants in cadaveric spines. Spine 1999;
24:213-218.
31. Kotani Y, Cunningham BW, Parker LM,
Kanayama M, McAfee PC: Static and
fatigue biomechanical properties of an-
terior thoracolumbar instrumentation
systems: A synthetic testing model.
Spine 1999;24:1406-1413.
32. Esses SI, Botsford DJ, Kostuik JP: Eval-
uation of surgical treatment for burst
fractures. Spine 1990;15:667-673.
33. Schnee CL, Ansell LV: Selection criteria
and outcome of operative approaches for
thoracolumbar burst fractures with and
without neurological deficit. J Neurosurg
1997;86:48-55.
34. Been HD, Bouma GJ: Comparison of two
types of surgery for thoraco-lumbar burst
fractures: Combined anterior and pos-
terior stabilisation vs. posterior instru-
mentation only. Acta Neurochir (Wien)
1999;141:349-357.
35. McAfee PC: Complications of anterior
approaches to the thoracolumbar spine:
Emphasis on Kaneda instrumentation.
Clin Orthop 1994;306:110-119.
363
 The Unstable Patella After Total Knee Arthroplasty:
Etiology, Prevention, and Management
Michel Malo, MD, FRCSC, and Kelly G. Vince, MD, FRCSC

Abstract
In total knee arthroplasty, most complications related to the extensor mechanism angle to produce a laterally directed
are caused by patellar maltracking or instability. Patellar maltracking may result muscle vector may cause patellofem-
from component malpositioning and limb malalignment, prosthetic design, improp- oral instability.
er patellar preparation, or soft-tissue imbalance. Patellofemoral instability likely re-
sults most frequently from internal malrotation of the femoral or tibial components. Component Malpositioning and
Although a patellofemoral radiograph may display the lateral subluxation of the pa- Limb Malalignment
tella, only computed tomography can quantify rotational malalignment of the fem- Malpositioning of the femoral,
oral or tibial component. Nonsurgical treatment is generally unsuccessful; major tibial, and/or patellar components is
malposition of components is best managed by implant revision. In the absence of the predominant cause of patellofemo-
component malposition, proximal realignments (lateral patellar retinacular release ral instability in TKA.1,4-10 Internal ro-
with lateral advancement of the vastus medialis obliquus muscle) or tibial tubercle tation (Fig. 1) and medialization (Fig.
transfers have been used. Surgical procedures on the patellar tendon itself may risk 2) of the femoral component shifts the
rupture of the extensor mechanism. trochlear groove medially, placing it
J Am Acad Orthop Surg 2003;11:364-371 more distant from the patella, which
tracks laterally relative to the femur.4-8
This increases tension on the lateral
retinacular tissues and exacerbates any
Patellar maltracking has been hy- and rotational positioning is crucial tendency of the patella to subluxate
pothesized as the cause of most pa- to ensure central patellar tracking and or dislocate laterally. Inadvertent in-
tellar complications, such as cata- a successful knee replacement. As ternal rotation of the femoral compo-
strophic wear, component loosening, conformity has increased in implant nent typically is caused by using in-
patellar fracture, and extensor dislo- design to reduce polyethylene wear,
cation. The stability of the patellofem- accurate component positioning has
oral joint in total knee arthroplasty become even more important. Addi- Dr. Malo is Attending Physician, Department of
(TKA) can be influenced by compo- tionally, the patella itself must be pre- Orthopaedic Surgery, Hôpital du Sacré-Coeur de
nent positioning and limb alignment, pared carefully with an accurate bone Montréal, University of Montreal, Montreal, Que-
bec, Canada. Dr. Vince is Associate Professor, De-
component design, patellar prepara- resection and proper soft-tissue bal-
partment of Orthopaedic Surgery, University of
tion, and soft-tissue balancing. The ancing. Understanding the etiology Southern California, Keck School of Medicine, Los
tibiofemoral articulation, which pro- and management of the unstable pa- Angeles, CA.
vides the track for the patella, is like- tella in TKA enables the orthopaedist
ly the most important factor. Malro- to minimize this complication in the Neither Dr. Malo nor the department with which
he is affiliated has received anything of value from
tation of either the femoral or tibial primary surgery.
or owns stock in a commercial company or insti-
component, with or without axial tution related directly or indirectly to the subject
malalignment, has been recognized as of this article. Dr. Vince or the department with
the most frequent cause of problems Etiology which he is affiliated has received royalties from
in the patellofemoral joint.1-5 Patel- Zimmer and Lippincott Williams & Wilkins.
lofemoral complications represent a Patellar instability is not simply a
Reprint requests: Dr. Vince, 1450 San Pablo Street,
high percentage of all complications complication of patellar resurfacing Los Angeles, CA 90033.
in TKA and are a major cause of re- technique. Any alteration in knee ki-
vision surgery. An accurately per- nematics and joint balance that in- Copyright 2003 by the American Academy of
formed arthroplasty with precise creases the tightness of lateral retinac- Orthopaedic Surgeons.
femoral and tibial component axial ular structures or increases the Q

364 Journal of the American Academy of Orthopaedic Surgeons

 Michel Malo, MD, FRCSC, and Kelly G. Vince, MD, FRCSC

coverage of the tibial plateau by in-

ternally rotating a tibial component
that is larger than ideal. Internal ro-
tational malalignment of the tibial
component causes the tibia to rotate
externally during knee flexion, driv-
ing the tubercle laterally. This increas-
es the Q angle and leads to lateral pa-
tellar tracking and subluxation (Fig.
3). Berger et al5 reported that the com-
bined amount of internal rotation of
femoral and tibial components corre-
lates directly with the severity of pa-
tellofemoral instability.
Furthermore, positioning the pa-
tellar button on the most lateral por-
tion of the patella should be avoided
because this tightens the lateral reti-
naculum and increases the tendency
toward lateral subluxation.4,10 Medi-
al positioning of the patellar compo-
Figure 1 A, Normal alignment of both tibial and femoral components yields normal pa- nent reproduces the asymmetric anat-
tellar tracking. B, Internal rotation of the femoral component (arrow) increases tension on
the lateral retinacular tissues, which causes lateral patellar tilt and exaggerates the tendency omy of the normal human patella.
of the patella to track laterally, and consequently increases lateral contact loading stresses on When the patella is everted, anatom-
the patellar component. Dashed line = transepicondylar axis. (Adapted with permission from ic medial placement of the patellar but-
Spitzer AI, Vince KG: Patellar considerations in total knee replacement, in Scuderi G [ed]:
The Patella. New York, NY: Springer-Verlag, 1995, p 313.) ton appears as a lateral position of the
button in the surgical field.
An overall limb malalignment in
strumentation that references from the Internal rotation and medialization >10° of valgus1 or a femoral compo-
posterior femoral articular surfaces of the tibial component also can cause nent positioned in >7° of valgus11 in-
when there are unrecognized deficien- instability and often result from poor creases the Q angle and thereby the
cies of the posterior lateral femoral exposure of the posterolateral tibial lateral vector of the quadriceps mus-
condyle (eg, valgus knee). A more re- plateau, related to difficulty everting cle. This increases the forces that tend
liable guide for rotational alignment the patella and dislocating the exten- to subluxate the patella laterally. The
in such an instance would be the trans- sor mechanism laterally.1,4,5,9 The sur- situation is most commonly seen in
epicondylar axis. geon also may attempt to optimize the patient with degenerative arthri-
tis and a preoperative valgus defor-
mity of the knee that is accompanied
by loss of bone stock of the distal lat-
eral femoral condyle.

Component Design
Several elements of femoral com-
ponent design influence patellar track-
ing. A trochlea oriented laterally fa-
vors engagement of the patella early
in flexion. A high lateral flange and
deep constrained groove provide cen-
tralizing forces to contain the patel-
la, in contrast to the flat unconstrained
Figure 2 A, Component too lateral. B, Component too medial. Medialization of the fem- trochlear surface in the original
oral component has an effect similar to that of internal rotation on lateral patellar tracking. Freeman-Swanson “roller in a trough”
(Adapted with permission from Krackow KA: The Technique of Total Knee Arthroplasty. St. prosthesis.6,10,12-14 Although this may
Louis, MO: CV Mosby, 1990, p 139.)
reduce patellofemoral instability, the

Vol 11, No 5, September/October 2003 365

The Unstable Patella After Total Knee Arthroplasty
Figure 3 Internal rotation of the tibial com-
ponent forces the tibia into external rotation
during knee flexion, increasing the Q angle
and leading to lateral patellar tracking and
subluxation. (Adapted with permission from
Spitzer AI, Vince KG: Patellar considerations
in total knee replacement, in Scuderi G [ed]:
The Patella. New York, NY: Springer-Verlag,
1995, p 315.)
increased patellofemoral contact forc-
es result in abnormal wear, increased
incidence of patellar fracture, and com-
ponent failure.2 Increased constraint
of the trochlear groove alone will not
centralize the patella and in fact may
tighten the lateral retinaculum. Al-
though any patellar button shape may
dislocate, an anatomic or asymmet-
ric component (ie, nondome shapes),
if inaccurately oriented, will more of-
ten adversely affect tracking.
Improper Patellar Preparation
Asymmetric patellar resection also
contributes to patellofemoral instabil-
ity.4,11 The articular facets of a patella
are normally asymmetric, with the
medial facet thicker than the lateral
facet (Fig. 4). Thus, an asymmetric pa-
tellar cut with a thicker lateral facet
and tighter retinaculum may pull the
patella to the lateral side.
Underresection of the patella over-
stuffs the patellofemoral joint and also
366
tightens the lateral retinaculum. Afem-
oral component of excessive size or
one that has been displaced anterior-
ly has a similar effect, overstuffing the
joint. This increases stress on the lat-
eral retinaculum and amplifies the risk
of patellar maltracking (Fig. 5).
Soft-Tissue Imbalance
The soft-tissue structures also can
lead to lateral patellar maltracking
and imbalance in the extensor mech-
anism, characterized by tightness of
the lateral retinaculum and an asso-
ciated weakness of the vastus medi-
alis muscle.15 This is a frequent prob-
lem in patients with longstanding and
severe valgus deformities. Further-
more, a disruption of the capsular re-
pair can result in patella subluxation
or dislocation.15 The disruption can
be caused by suturing technique fail-
ure, hematoma, overly intensive
physical therapy, or traumatic rup-
ture of the medial retinaculum.
Prevention
With advances in surgical technique,
prosthetic design, and instrumenta-
tion, TKA has become an efficient and
durable surgical procedure, provid-
ed that the initial implantation of all
components is carefully done. Sever-
al basic technical guidelines must be
followed to achieve patellofemoral
stability.
The femoral component should be
slightly externally rotated on the fe-
mur relative to the posterior articu-
lar condyles to establish a rectangu-
lar and balanced flexion gap and
to accommodate central patellar
tracking.4-8 To achieve appropriate
femoral component rotational align-
ment, a number of intraoperative land-
marks have been developed: the an-
teroposterior (AP) axis of the femur,
the transepicondylar axis, and the pos-
terior condylar axis. The AP axis of
the femur is a line through the center
of the trochlear groove and the top
of the intercondylar notch; it is also
Journal of the American Academy of Orthopaedic Surgeons
Figure 4 Resection of more bone from the
medial facet is necessary to obtain a symmet-
ric patellar cut parallel to the anterior surface
because the medial facet is thicker than the
lateral facet in a normal patella. (Adapted
with permission from Spitzer AI, Vince KG:
Patellar considerations in total knee replace-
ment, in Scuderi G [ed]: The Patella. New York,
NY: Springer-Verlag, 1995, p 315.)
known as Whiteside’s line. The trans-
epicondylar axis is a line connecting
the lateral epicondylar prominence
and the medial sulcus on the medial
epicondyle, as clearly defined by
Berger et al.16 The posterior condylar
axis is defined as a line that crosses
both posterior femoral condyles. This
will be, on average, at 3° to 5° of in-
ternal rotation relative to the transepi-
condylar axis.
The anterior and posterior cuts of
the femur should be made by resect-
ing the posterior femoral condyles
parallel to the transepicondylar axis,
perpendicular to the AP axis and
approximately 3° externally rotated
from the posterior condylar axis. Al-
though the posterior condylar surface
of the femur is probably the most rou-
tinely used reference in most systems,
the transepicondylar axis is more ac-
curate. This is especially true in a val-
gus knee, in which the posterior lat-
eral femoral condyle is frequently
hypoplastic. If the posterior condy-
lar axis is used as a reference in the
valgus knee, the femoral component
usually will be internally rotated.
The femoral component should be
lateralized maximally (without over-
hang) to reduce tension in the lateral
retinaculum and to place the trochlea
more laterally.4,6,8
Rotational alignment of the tibial
component is equally important.1,4,5,9
The midportion of the tibial compo-
nent should be rotationally aligned
with the tibial attachment of the ex-
tensor mechanism. Guides for this po-
 Michel Malo, MD, FRCSC, and Kelly G. Vince, MD, FRCSC

face of the patellar bone (Fig. 4), with

equal bone thickness remaining prox-
imally, distally, medially, and lateral-
ly to balance the peripatellar retinacu-
lar tissues. To confirm that the patellar
cut is even, careful palpation of the
residual patella between the thumb
and the index finger can be helpful,
as can measurement of the patella in
all four quadrants with a caliper. Ad-
equate exposure with excision of the
synovium and osteophytes around the
patella is essential.
The overall thickness of the patel-
lofemoral joint should be main-
tained.4,11 This involves appropriate
sizing and positioning of components
and reproducing the native patellar
thickness. In sizing the femoral com-
ponent, the AP dimension is more im-
portant than the mediolateral dimen-
sion. Referencing the mediolateral
dimension alone could result in se-
Figure 5 Overstuffing of the patellofemoral joint tightens the lateral retinaculum and in- lecting a femoral component larger
creases the risk of lateral patellar tracking. A, In the knee with optimal femoral component
size (a), this can be caused by increased thickness of the resurfaced patella (b). B, In the knee than is needed in the AP dimension.
with an optimal patella size (d), this can be caused by the use of an oversized femoral com- This would displace the trochlea and
ponent (c). The asterisk indicates the normal level of the patellar articulation on the femoral the patella anteriorly, tightening the
trochlea. (Adapted with permission from Krackow KA: The Technique of Total Knee Arthro-
plasty. St. Louis, MO: CV Mosby, 1990, p 215.) retinaculum and increasing the ten-

sition are the crest of the tibia, or the
medial third of the tibial tubercle, as
initially suggested by Merkow et al.1
If a symmetric tibial component is
used, a few millimeters of the com-
ponent might overhang posterolater-
ally. A common technical error is to
align the posterior margin of the tibial
component parallel to the posterior
cortex of the tibial plateau because it
is a better geometric fit. Although this
increases coverage of the tibia and per-
mits implantation of a larger tibial
component, it also results in internal
rotation of the tibial component, a def-
inite risk factor for maltracking.
Femoral and tibial component im-
plantation should restore the mechan-
ical axis of the limb to neutral. In most
cases, the femoral component should
not be positioned in >7° of valgus
with respect to the long axis of the fe-
mur, and the combined angle of the
femoral and tibial cuts should not ex-
ceed 10° of valgus.1,11 However, in
some patients with a pronounced
femoral bow or malunion of a shaft
fracture, a greater valgus position of
the femoral component may be re-
quired to help ensure the neutral me-
chanical axis.
The patellar component should be
placed on the medial portion of the
patella, leaving some lateral patellar
bone cut uncovered.4,10,17,18 This re-
produces the normal articular config-
uration of the patella in which the
sagittal central ridge is asymmetrical-
ly displaced medially. The lateral
shear forces on the patella are reduced
and patellar tracking is improved by Figure 6 Medialization of the patellar com-
ponent on the cut surface of the patellar bone
allowing the patellar button to artic- allows the patellar button to be centralized
ulate centrally in the trochlear groove in the trochlear groove and improves patel-
(Fig. 6). There is clinical evidence that lar tracking by decreasing lateral patellar sub-
luxation forces. (Adapted with permission
medialization reduces the incidence from Spitzer AI, Vince KG: Patellar consid-
of lateral retinacular release.17,18 erations in total knee replacement, in Scud-
In a symmetric patellar resection,4,11 eri G [ed]: The Patella. New York, NY: Springer-
Verlag, 1995, p 316.)
the cut is parallel to the anterior sur-

Vol 11, No 5, September/October 2003 367

The Unstable Patella After Total Knee Arthroplasty
dency for lateral subluxation. In gen-
eral, the smallest femoral component
that fits without notching the anteri-
or femoral cortex is a good choice to
maintain stability and motion of the
patellofemoral joint. However, the
surgeon must avoid overresection of
the posterior femoral condyles as a re-
sult of choosing too small a compo-
nent, which could lead to flexion in-
stability by creating an increased
flexion gap. To measure and repro-
duce the patellar thickness, a caliper
should be used before resecting the
patellar facets and the measurement
compared with that of the resurfaced
patella with the patellar button in
place. Sufficient thickness of bone
should be left (ideally, 12 to 15 mm)
to anchor the patellar component and
to resist the force applied to the pa-
tella during knee flexion.
With the trial components in place,
an intraoperative assessment of pa-
tellar tracking should be conducted.
Tourniquet inflation can alter intra-
operative patellar tracking during
TKA because it has a binding effect
on the extensor mechanism. Marson
and Tokish19 reported that unneces-
sary lateral retinacular releases could
on occasion be avoided by deflating
the tourniquet in patients in whom
tracking problems were apparent in-
traoperatively. Thus, the potential con-
sequences of an unnecessary lateral
retinacular release might be avoided
(ie, increased postoperative pain and
prolonged rehabilitation, wound-
healing problems, and morbidity as-
sociated with compromised patella
blood supply). Nevertheless, a later-
al retinacular release, when indicat-
ed and judiciously performed, can im-
prove patellar tracking.
The patella must track centrally
during surgery without lateral sub-
luxation or lateral tilt in full flexion.
Patellar tracking can be assessed us-
ing the no-thumb technique (without
manual support from the surgeon’s
thumb to stabilize the patella in the
trochlear groove).20 There may be a
tendency to overestimate the need for
368
lateral retinacular release because the
conventional parapatellar arthrotomy
remains open on the medial side. The
single-suture technique is a modifi-
cation of the no-thumb test in which
the medial retinaculum at the supe-
rior pole of the patella is reapproxi-
mated using a single suture or a tow-
el clip.17,21 This technique reproduces,
albeit statically, the medial vector of
the extensor mechanism. However,
evaluation of patellar tracking during
surgery is relatively crude and can-
not account for the dynamic forces
that act on the patella.
Management of Patellar
Instability
Clinical Evaluation
A complete history and physical
examination is the initial step in eval-
uating a patient with patellar mal-
tracking after TKA. Patient com-
plaints may or may not include the
subjective sensation of subluxation.
Patellar subluxation may manifest
only as peripatellar pain and limited
flexion but without symptoms of
frank instability. The patella that sub-
luxates may prevent the patient from
achieving full flexion during postop-
erative physical therapy. Patellar mal-
tracking may then present as a stiff,
painful knee. Usually pain has been
present since surgery and is described
by the patient as being different from
the pain before TKA. With medial ret-
inacular disruption, there may have
been new-onset pain related to trau-
ma or an episode of popping.
A physical examination of the low-
er extremity may help identify the
cause of patellar maltracking. Over-
all limb alignment in valgus should
be measured to assess the Q angle. Pa-
tellar tracking should be observed
during both active and passive flex-
ion of the knee; in some instances, this
might help differentiate between dy-
namic and static imbalance. Patellar
mobility and lateral retinaculum
tightness should be assessed.
Journal of the American Academy of Orthopaedic Surgeons
Although clinically difficult to ad-
dress precisely, the rotation of the
components can be estimated. With
the patient in a supine or sitting po-
sition and the hips and knees flexed
at 90°, internal rotation to both hips
is applied and measured. Provided
that both hips are disease free, appar-
ent increased internal rotation of the
hip on the TKA limb compared with
the contralateral side suggests an in-
ternally malrotated femoral compo-
nent. Similarly, with the patient sit-
ting on the edge of the table and the
knees side by side, increased exter-
nal rotation of the foot on the TKA
side may indicate an internally mal-
rotated tibial component (Fig. 7).
Radiologic Evaluation
Standard radiographs, including
AP, lateral, and Merchant views, can
provide key clinical information. The
diagnosis can be confirmed by the lat-
erally subluxated patella identified on
the Merchant view. The Merchant
view also shows patellar tilt and the
angle of patellar resection. The angu-
lar position of the femoral component
can be assessed on the AP view; the
size of the femoral component is ap-
parent on the lateral view. Rotational
alignment of the component can be
evaluated, with some difficulty, on
plain radiographs. Eckhoff et al22 have
proposed a method to measure com-
Figure 7 With the patient sitting on the edge
of the table, knees flexed and feet supported,
the relative external rotation of the foot in-
dicates internal rotation of the tibial compo-
nent of the TKA. The internally rotated tib-
ial component rotates the tubercle externally.
 Michel Malo, MD, FRCSC, and Kelly G. Vince, MD, FRCSC

ponent rotation based on the geomet-

ric analysis of implant pegs on a lat-
eral radiograph of the knee in
extension, but its clinical application
is limited by the necessity for the im-
plant to be pegged and by the actual
lack of clinical correlation.
The computed tomography (CT)
protocol described by Berger et al5 is
now the investigation of choice to
quantify the degree of rotational
malalignment based on accurate and
reproducible validated anatomic
landmarks: the surgical transepi-
condylar axis and the tibial tubercle
(Fig. 8). The accuracy of CT for de-
termining rotational alignment of
both components in TKA has been as-
certained by Jazrawi et al.23

Management
Patellofemoral instability is man-
aged on the basis of its etiology. The
precise cause of the instability must
be identified and corrected rather
than simply be compensated by an-
other maneuver. In this respect, rou-
tine lateral retinacular release or a pa-
tellectomy is misguided. Nonsurgical
measures, such as strengthening of
the vastus medialis obliquus muscle
and bracing, generally are unsuccess-
ful because patellofemoral instabili-
ty typically is secondary to a struc-
tural problem. Revision arthroplasty
is needed in most cases, with the type
of procedure dependent on the nature
of component malposition.
In the absence of component mal-
position, a lateral retinacular release
with or without a vastus medialis ad-
vancement (together constituting the
proximal realignment) has centralized
patellar tracking in cases of symptom-
atic patellar tilt or subluxation relat-
ed to tight lateral retinacular struc- Figure 8 A 77-year-old man presented with a painful TKA of the left side, with patello-
femoral maltracking and rotational malalignment of both components. Preoperative radio-
tures.24 This may have been done at graphic AP (A), lateral (B), and Merchant (C) views. D, Preoperative axial CT scan of the
the time of the original TKA. The sur- distal femur showing the internal rotational malalignment of the femoral component. Line
geon should suspect malposition of a extends from the lateral epicondyle to the anterior ridge of the medial epicondyle where
the superficial MCL attaches. Line b extends from the medial epicondylar sulcus (arrow).
femoral and/or tibial components dur- (Line b is the transepicondylar axis.) Line c is the prosthetic posterior condylar line. E, Pre-
ing surgery if a lateral retinacular re- operative axial CT scan showing the excessive internal rotational malalignment of the tibial
lease is necessary in a patient in whom component, that is, the relationship between the tibial component axis (d)—perpendicular to
the posterior margin of the component—and the orientation of the tibial tubercle (e). (Nor-
the patella tracked centrally before sur- mal value is 18°.) Postoperative radiographic AP (F) and Merchant (G) views.
gery. A lateral retinacular release in

Vol 11, No 5, September/October 2003 369

The Unstable Patella After Total Knee Arthroplasty
that patient usually will not reestab-
lish patellofemoral stability if rotation-
al malalignment of the components
is also present. The lateral release can
be done by either an outside-in or
inside-out incision, extending from the
tibial bone to the fibers of the vastus
lateralis, approximately 2 cm from the
lateral border of the patella. The lat-
ter approach does not create a lateral
skin flap. The lateral superior genicu-
late artery should be preserved be-
cause of the potential adverse conse-
quences of compromised blood supply
to the patella, theoretically predispos-
ing to patellar fracture. Nevertheless,
patellar fracture is probably more
closely related to the abnormal forc-
es created by maltracking than to os-
teonecrosis.
Other methods of correcting patel-
lar instability in TKA have been rec-
ommended. Insall described a tech-
nique of proximal realignment of the
extensor mechanism with advance-
ment of the vastus medialis obliquus
and medial plication that could be
combined with a lateral retinacular re-
lease.1 Good results have been report-
References
1. Merkow RL, Soudry M, Insall JN: Patel-
lar dislocation following total knee re-
placement. J Bone Joint Surg Am 1985;67:
1321-1327.
2. Ranawat CS: The patellofemoral joint
in total condylar knee arthroplasty:
Pros and cons based on five- to ten-year
follow-up observations. Clin Orthop
1986;205:93-99.
3. Figgie HE III, Goldberg VM, Figgie MP,
Inglis AE, Kelly M, Sobel M: The effect
of alignment of the implant on fractures
of the patella after total condylar total
knee arthroplasty. J Bone Joint Surg Am
1989;71:1031-1039.
4. Briard JL, Hungerford DS: Patellofem-
oral instability in total knee arthroplas-
ty. J Arthroplasty 1989;4(suppl):S87-S97.
5. Berger RA, Crossett LS, Jacobs JJ,
Rubash HE: Malrotation causing patel-
lofemoral complications after total
knee arthroplasty. Clin Orthop 1998;356:
144-153.
6. Rhoads DD, Noble PC, Reuben JD, Ma-
370
ed with this technique by Merkow et
al1 and Grace and Rand.25 However,
both of these studies predate the cur-
rent comprehensive understanding of
component malrotation and preference
for correcting rotational malalignment.
Some have advocated distal realign-
ment with osteotomy and medial dis-
placement of the tibial tubercle.4,26 Un-
like proximal realignment, this
procedure changes the Q angle, and
it has even been proposed in cases of
mild malrotation of the compo-
nents.26,27 However, many surgeons
avoid procedures on the patella-
tendon attachment because of the po-
tential for serious complications, such
as patellar tendon rupture and non-
union of the osteotomy. Other tech-
niques for osteotomy that use a long
and large osteotomy to allow secure
fixation, combined with maintaining
the distal bony attachment as a hinge
and preserving the lateral periosteum,
recently have been suggested.28 When
substantial malposition of one or both
components exists, revision of the
components is still the procedure of
choice.4
honey OM, Tullos HS: The effect of fem-
oral component position on patellar
tracking after total knee arthroplasty.
Clin Orthop 1990;260:43-51.
7. Anouchi YS, Whiteside LA, Kaiser AD,
Milliano MT: The effects of axial rota-
tional alignment of the femoral compo-
nent on knee stability and patellar
tracking in total knee arthroplasty dem-
onstrated on autopsy specimens. Clin
Orthop 1993;287:170-177.
8. Rhoads DD, Noble PC, Reuben JD, Tul-
los HS: The effect of femoral component
position on the kinematics of total knee
arthroplasty. Clin Orthop 1993;286:
122-129.
9. Nagamine R, Whiteside LA, White SE,
McCarthy DS: Patellar tracking after to-
tal knee arthroplasty: The effect of tib-
ial tray malrotation and articular sur-
face configuration. Clin Orthop 1994;
304:262-271.
10. Yoshii I, Whiteside LA, Anouchi YS:
The effect of patellar button placement
Journal of the American Academy of Orthopaedic Surgeons
Summary
Patellofemoral instability is a major
complication in TKA and remains a
common reason for revision surgery.
Accurate component implantation is
imperative for a successful outcome
in knee replacement. Factors adverse-
ly affecting patellar tracking should
be well understood and recognized;
they can largely be avoided with good
surgical technique. In patients with
a malfunctioning TKA, rotational
malalignment should be suspected in
the absence of axial malalignment,
loosening, or infection. Assessment of
component rotational alignment by CT
is crucial in planning treatment be-
cause the femoral and/or tibial com-
ponent will need to be revised if mal-
positioned. Nonsurgical management
generally is unsuccessful. In the ab-
sence of component malposition, a lat-
eral retinacular release may restore
central patellar tracking. Tibial tuber-
cle transfers also have been used suc-
cessfully. Major malposition of one or
both components requires implant re-
vision.
and femoral component design on pa-
tellar tracking in total knee arthroplas-
ty. Clin Orthop 1992;275:211-219.
11. Hozack WJ: Patellar complications re-
lated to tracking, in Fu FH, Harner CD,
Vince KG (eds): Knee Surgery. Balti-
more, MD: Williams & Wilkins, 1994,
vol 2, pp 1461-1472.
12. Andriacchi TP, Yoder D, Conley A,
Rosenberg A, Sum J, Galante JO: Patel-
lofemoral design influences function
following total knee arthroplasty. J Ar-
throplasty 1997;12:243-249.
13. Chew JT, Stewart NJ, Hanssen AD, Luo
ZP, Rand JA, An KN: Differences in pa-
tellar tracking and knee kinematics
among three different total knee de-
signs. Clin Orthop 1997;345:87-98.
14. Freeman MA, Samuelson KM, Elias SG,
Mariorenzi LJ, Gokcay EI, Tuke M: The
patellofemoral joint in total knee pros-
theses: Design considerations. J Arthro-
plasty 1989;4(suppl):S69-S74.
15. Barnes CL, Scott RD: Patellofemoral

complications of total knee replace-
ment. Instr Course Lect 1993;42:303-307.
16. Berger RA, Rubash HE, Seel MJ,
Thompson WH, Crossett LS: Determin-
ing the rotational alignment of the fem-
oral component in total knee arthro-
plasty using the epicondylar axis. Clin
Orthop 1993;286:40-47.
17. Lewonowski K, Dorr LD, McPherson
EJ, Huber G, Wan Z: Medialization of
the patella in total knee arthroplasty.
J Arthroplasty 1997;12:161-167.
18. Hofmann AA, Tkach TK, Evanich CJ, Ca-
margo MP, Zhang Y: Patellar component
medialization in total knee arthroplasty.
J Arthroplasty 1997;12:155-160.
19. Marson BM, Tokish JT: The effect of a
tourniquet on intraoperative patellofemo-
ral tracking during total knee arthroplasty.
J Arthroplasty 1999;14:197-199.
20. Scott RD: Prosthetic replacement of the
Vol 11, No 5, September/October 2003
Michel Malo, MD, FRCSC, and Kelly G. Vince, MD, FRCSC
patellofemoral joint. Orthop Clin North
Am 1979;10:129-137.
21. Bindelglass DF, Vince KG: Patellar tilt
and subluxation following subvastus and
parapatellar approach in total knee ar-
throplasty: Implication for surgical tech-
nique. J Arthroplasty 1996;11:507-511.
22. Eckhoff DG, Piatt BE, Gnadinger CA,
Blaschke RC: Assessing rotational
alignment in total knee arthroplasty.
Clin Orthop 1995;318:176-181.
23. Jazrawi LM, Birdzell L, Kummer FJ, Di
Cesare PE: The accuracy of computed
tomography for determining femoral
and tibial total knee arthroplasty com-
ponent rotation. J Arthroplasty 2000;15:
761-766.
24. Brassard MF, Insall JN, Scuderi GR:
Complications of total knee arthroplas-
ty, in Insall JN, Scott WN (eds): Surgery
of the Knee, ed 3. New York, NY:
Churchill Livingstone, 2001, vol 2, pp
1801-1844.
25. Grace JN, Rand JA: Patellar instability
after total knee arthroplasty. Clin Or-
thop 1988;237:184-189.
26. Kirk P, Rorabeck CH, Bourne RB,
Burkart B, Nott L: Management of re-
current dislocation of the patella fol-
lowing total knee arthroplasty. J Arthro-
plasty 1992;7:229-233.
27. Nagamine R, Whiteside LA, Otani T,
White SE, McCarthy DS: Effect of me-
dial displacement of the tibial tubercle
on patellar position after rotational
malposition of the femoral component
in total knee arthroplasty. J Arthroplasty
1996;11:104-110.
28. Whiteside LA: Distal realignment of the
patellar tendon to correct abnormal pa-
tellar tracking. Clin Orthop 1997;344:
284-289.
371
 Web and Wireless Review
A Log of Daily Internet Use
For most orthopaedic surgeons, the Internet has many
applications that can be used on a daily basis. On a typ-
ical morning, after making rounds on five patients and
seeing one new consult, I connect to my office computer
from the computer in the recovery room using Go to My
PC (www.gotomypc.com) and review the day’s sched-
ule of patients. (Details on the security of this connec-
tion can be found on its Website.) Through the Go to My
PC site, I can access ChartLogic (www.chartlogic.com),
my electronic medical record program, and enter my new
hospital encounters for billing. Formerly I used my Pock-
et PC for this, syncing the charges when I returned to
the office by using the charge capture program Quick-
Rounds (www.officemedical.com), but a direct Internet
connection approach may be better and is worth eval-
uating.
Once at my office, I access my e-mail with my laptop,
using Spam Inspector (www.giantcompany.com) to fil-
ter out unwanted e-mails. A message from Dr. Jones ask-
ing me to call him today generates an e-mail to my prac-
tice manager, Yvette, to remind me to phone him later.
Since my staff uses standard e-mail on the Internet (not
an Intranet), I can quickly receive and respond to their
messages.
Midmorning, I have a patient with an MRI scan show-
ing a 3-day-old partial distal biceps tendon rupture. Be-
cause I do not often see this problem, I retrieve the in-
formation I need by using MD Consult (www.mdconsult.com)
to access Campbell’s Operative Orthopaedics and Merck Medi-
cus (www.merckmedicus.com) to access Chapman’s Or-
thopaedic Surgery. Later in the morning, I see a patient with
carpal tunnel syndrome. She receives an injection and a
prescription for a night splint. From the Patient/Public
Information section of the AAOS Website (www.aaos.org),
I print out educational material for her to take home.
During lunch, I reply to new e-mail messages, try to
reach Dr. Jones, then return to MD Consult to conduct a
literature search on long-term follow-up studies for a lec-
ture I will be giving on unicompartmental knees. I copy
the abstracts to a new Microsoft Word file, then convert
this document to Microsoft Reader format using a tool
downloaded from www.microsoft.com/reader/downloads/
rmr.asp. I download the entire collection of abstracts to
my Pocket PC for reading later between my two add-on
cases in the OR.
372 Journal of the American Academy of Orthopaedic Surgeons
In the afternoon, my practice manager and I discuss
office overhead cost issues. From the Internet, she has
downloaded prices for all of our commonly used prod-
ucts from six medical suppliers and has developed a cus-
tomized ordering sheet using a mix of the vendors. This
management technique will help produce cost savings
in office overhead.
When I arrive at the OR, I learn that the first surgery
of the evening is delayed, so I read the abstracts down-
loaded earlier in the day. I am told that the OR has not
received my office admission note for this patient. Using
Go to My PC, I connect to my office computer from the
OR computer and access the patient’s chart in ChartLog-
ic. I see that we did fax the note, but I use the Internet
connection to fax it directly to the OR via our office fax
server. Within minutes, the complete consultation is in
the chart and we can proceed.
I arrive home by 10 pm and log onto my computer to
review my e-mail. Dr. Jones is still in the OR. I check my
office Website and add to it a few videos of a unicom-
partmental knee operation for patient viewing. I also set
up a link to MDhub (www.mdhub.com) so that my pa-
tients can request appointments by e-mail or fax.
I visit Web MD (www.webmd.com), where I am the
joint replacement consultant, and answer three questions
on the message board (www.joints.webmd.com). Before
calling it a day, I sign onto the AAOS site again and visit
the Orthopaedic Campus. I make note of an online CME
course concerning the diagnosis and treatment of rota-
tor cuff disease that I intend to complete over the week-
end. I log onto Medscape (www.medscape.com), view the
responses in a discussion group on hip approaches, check
my e-mail for a final time, and at last power down.
Ira H. Kirschenbaum, MD
This article represents solely the opinions of the author and
not those of the AAOS and JAAOS. Any physician who main-
tains or transmits patient medical information electronically
needs to understand and comply with the Administrative Sim-
plification provisions of HIPAA as they become effective. The
Privacy regulations went into effect on April 14, 2003; the
Transactions and Code Sets (TCS) regulations become effec-
tive on October 16, 2003; and the Security regulations will
formally take effect on April 21, 2005.