Isolated and Combined Posterior Cruciate

Ligament Injuries
Daniel M. Veltri, MD, and Russell F. Warren, MD

Abstract

Posterior cruciate ligament (PCL) injuries represent 3% to 20% of all knee liga- ally involve multiple ligaments.
mentous injuries, but the diagnosis often is missed at initial evaluation. Diagnos- Forced hyperextension can injure
tic acumen is increased by knowledge of knee biomechanics and selective the PCL, but this usually results in
ligament-cutting studies. The examiner must differentiate the isolated PCL injury combined ligamentous injury
from combined ligamentous injury to determine appropriate treatment. Isolated involving the anterior cruciate liga-
acute PCL tears with less than 10 mm of posterior laxity at 90 degrees of flexion ment (ACL).1,6 Posteriorly directed
should be treated with an aggressive rehabilitative program. This amount of laxity force to the anteromedial tibia with
is found in the majority of isolated acute PCL tears. Isolated acute PCL tears with the knee in hyperextension may also
more than 10 to 15 mm of posterior laxity and PCL tears with combined ligamen- cause a posterolateral corner injury,1
tous injuries should be reconstructed. Large PCL bony avulsions should be fixed which results in varus and external-
internally. Small PCL bony avulsions with more than 10 mm of posterior laxity rotation knee instability. Significant
should be reconstructed. Chronic PCL injuries initially should be treated with an varus or valgus stress will injure the
aggressive rehabilitation program. If such a program is not successful in a patient PCL only after rupture of the appro-
with more than 10 to 15 mm of posterior laxity and no significant radiographic evi- priate collateral ligament.
dence of degenerative changes, the PCL should be reconstructed.
J Am Acad Orthop Surg 1993;1:67-75 Biomechanics

Posterior cruciate ligament injuries

Injury to the posterior cruciate liga-
ment (PCL) is thought to account for
3% to 20% of all knee ligament
injuries.1,2 The true incidence of PCL
injuries remains unknown because
many isolated PCL injuries may be
undetected. Parolie and Bergfeld3
noted a 2% PCL injury rate among
asymptomatic college football play-
ers invited to the National Football
League predraft examination.
Accurate diagnosis of the PCL
injury is the first step in determining
appropriate management. The abil-
ity to differentiate an isolated from a
combined ligamentous injury is
aided by a knowledge of knee bio-
mechanics obtained with the use of
selective ligament-cutting tech-
niques.4 It is also important to under-
stand the natural history of the
PCL-injured knee, the results of non-
operative treatment with aggressive
rehabilitation, and the results of sur-
gical treatment to determine appro-
priate management.5-8 In this article
we will present the current approach
to the diagnosis and management of
isolated and combined PCL injuries.
Mechanism of Injury
Most PCL injuries occur as a result of
athletic, motor vehicle, or industrial
accidents. The mechanism of most
athletic PCL injuries is a fall on the
flexed knee with the foot in plantar
flexion.3,7 This imparts the force to the
tibial tubercle, which drives the tibia
posteriorly and ruptures the ligament,
usually resulting in an isolated PCL
injury. Similarly, in motor vehicle acci-
dents, the knee is flexed, and the tibia
is forced posteriorly on impact with
the dashboard.6 Hyperflexion of the
knee without a direct blow to the tibia
can also cause isolated PCL injury.
The PCL can be involved in other
mechanisms of injury, but these usu-
are commonly overlooked during
the initial evaluation of the acutely
injured knee. The physical examina-
tion findings in isolated PCL injury
are subtle. Knowledge of the biome-
chanics obtained from selective liga-
ment-cutting experiments allows
correlation of a simulated physical
examination with known ligament
injury. Such selective cutting studies
measure the change in knee motion
after transection of a specific liga-
ment. The experimentally produced
change in laxity over a range of knee-
Dr. Veltri is Chief, Department of Orthopaedic
Surgery, Luke Air Force Base, Litchfield Park,
Ariz. Dr. Warren is Professor of Orthopaedic
Surgery, Cornell Medical College, New York
City; and Chief, Sports Medicine and Shoulder
Service, The Hospital for Special Surgery.
Reprint requests: Dr. Warren, Sports Medicine
and Shoulder Service, The Hospital for Special
Surgery, 535 E. 70th St, New York, NY 10021.

Vol 1, No 2, Nov/Dec 1993 67

Posterior Cruciate Ligament Injuries
flexion angles provides an important varus moment at any angle of
basis for clinical knee testing. flexion. In contrast, sectioning of the
Gollehon et al4 used selective liga- LCL and the deep ligament complex
ment-cutting techniques to evaluate resulted in increased varus angula-
the role of the PCL and the postero- tion at all angles of knee flexion and
lateral corner in stability of the knee. was maximal at 30 degrees. Addi-
They found that isolated sectioning tional sectioning of the PCL further
of the PCL increased posterior trans- increased varus angulation at all
lation with posteriorly directed force angles of knee flexion.
at all angles of flexion, but the maxi- Isolated sectioning of the PCL did
mal excursion occurred at 90 degrees not increase external rotation with
of flexion. With an intact PCL, sec- an external rotation moment at any
tioning of the lateral collateral liga- angle of knee flexion. With an intact
ment (LCL) and the deep ligament PCL, sectioning of the LCL and the
complex (arcuate ligament, popli- deep ligament complex increased
teus tendon, fabellofibular ligament, external rotation at all angles of
and posterolateral capsule) pro- flexion and was maximal at 30
duced small but significant increases degrees. Additional sectioning of the
in posterior translation at all angles PCL markedly increased external
of flexion and was maximal at 30 rotation at 60 and 90 degrees of
degrees. The amount of posterior flexion.
translation produced by combined
sectioning of the LCL and the deep
ligament complex with an intact PCL Clinical Examination
was similar to that produced by iso-
lated sectioning of the PCL at 0 and Biomechanical data can be applied
30 degrees of knee flexion. to clinical examination of the knee
Isolated sectioning of the PCL did (Table 1). Changes in posterior trans-
not increase varus angulation with lation, external rotation, and varus
Table 1
Usefulness of Clinical Tests in Detection of Knee Injury
Type of Injury*
PCL and
Posterolateral Posterolateral
Clinical Test PCL Corner Corner
Posterior drawer, 30 degrees + + ++
Posterior drawer, 90 degrees ++++ – +++
Posterior sag, 90 degrees +++ – +++
Quadriceps active ++++ – +++
Prone external rotation,
30 degrees – ++++ +++
Prone external rotation,
90 degrees ++ + +++
Varus stress, 30 degrees – +++ +++
Varus stress, 90 degrees ++ + +++
Reverse pivot shift –/+ ++ ++
* Symbols represent grading scale for usefulness in detecting type of injury,
ranging from – (not useful) to ++++ (most useful).
68 Journal of the American Academy of Orthopaedic Surgeons
angulation are the most useful
findings for detecting injury to the
PCL and the posterolateral corner.1
Isolated PCL injury will allow maxi-
mum posterior translation with pos-
teriorly directed force at 70 to 90
degrees of flexion. Since posterior
translation is greatest at 90 degrees
of flexion, the posterior drawer test
should be performed in this posi-
tion. Achieving 90 degrees of knee
flexion in an acute injury may be
difficult, however. Increased poste-
rior translation, external rotation,
and varus angulation at 30 degrees
of knee flexion that decreases at 90
degrees indicates isolated injury to
the posterolateral corner. Thus, com-
paring posterior translation, exter-
nal rotation, and varus angulation at
30 and 90 degrees can help differen-
tiate PCL injury from posterolateral
corner injury.4 Increased posterior
translation, varus angulation, and
external rotation at 90 degrees of
flexion indicate combined injury to
both the PCL and the posterolateral
corner.4
The posterior drawer test at 90
degrees of flexion is most useful for
documenting PCL insufficiency.
This test is performed with the
patient supine, with both feet on the
table and the knee flexed to 90
degrees. At this angle of flexion, the
anterior tibial condyles should be
well anterior to the corresponding
femoral condyles (approximately 10
mm). The injured knee is compared
with the normal knee. If the tibia can
be moved posteriorly 0 to 5 mm on
the injured side, this is considered a
grade I posterior drawer sign. This
usually corresponds to posterior
displacement of the tibial condyles
to a position that is still anterior to
the femoral condyles. If the tibia can
be displaced 5 to 10 mm posteriorly,
this is a grade II posterior drawer
sign. This corresponds to posterior
displacement of the tibial condyles
until they are flush with the femoral
condyles. If the tibia can be dis-

placed more than 10 mm posteri-
orly, this represents a grade III pos-
terior drawer sign. This corresponds
to displacement of the tibial
condyles posterior to the femoral
condyles.
In addition to posterior displace-
ment, the examiner should usually
assess an endpoint when performing
a posterior drawer test. Most acutely
PCL-deficient knees have an altered
endpoint with a posterior drawer
test. However, the posterior end-
point may return to normal with
time in the chronically PCL-deficient
knee. In this situation we find the
posterior drawer test endpoint less
sensitive than the endpoint in a Lach-
man test done for an ACL injury.
Examination of the injured knee
should always include a Lachman
test at 30 degrees of flexion. In the
PCL-deficient knee, the tibia is sub-
luxated posteriorly, and the Lach-
man test may demonstrate increased
anteroposterior (AP) translation
with a firm anterior endpoint. The
increased AP translation is due to the
posterior subluxation from the PCL
injury and should not be confused
with the findings in an ACL-deficient
knee, which has a soft endpoint.
Vol 1, No 2, Nov/Dec 1993
Daniel M. Veltri, MD, and Russell F. Warren, MD
The posterior drawer test should
also be performed with the foot in
internal and external rotation. Many
patients with a positive posterior
drawer sign in neutral rotation have
decreased excursion when the
drawer test is performed in internal
rotation. 3,7 This finding has been
attributed to PCL injury with an
intact Humphry’s, or Wrisberg’s,
ligament.7 Such a finding also may
indicate maintenance of the integrity
of the posterolateral corner, which
provides the secondary restraint to
posterior displacement.1
The posterior drawer test per-
formed with the foot in external
rotation (the posterolateral drawer
test) has been used to assess postero-
lateral corner injury. The findings
with this maneuver must be com-
pared with those in the intact unin-
jured knee. A positive finding can
indicate injury to the PCL or pos-
terolateral corner but is not specific.1
The quadriceps active test is also
useful in the diagnosis of PCL
injury.9 This test involves placing the
patient supine and flexing the knee
90 degrees with the foot resting on
the table (Fig. 1). In the intact knee, a
quadriceps contraction results in
posterior translation of the tibia rela-
tive to the femur. In the PCL-
deficient knee, the tibia rests in a
posteriorly subluxated position, and
a quadriceps contraction produces
anterior translation of the tibia rela-
tive to the femur. Thus, anterior
translation with quadriceps contrac-
tion with the knee at 90 degrees of
flexion indicates PCL injury. We
consider the quadriceps active test
and the posterior drawer test to be
the most useful tests for diagnosing
PCL injury.
The posterior sag test is similar to
the posterior drawer test.6 The test is
performed at 90 degrees of hip and
knee flexion and uses gravity to
apply a posteriorly directed force to
the tibia. The posterior sag of the
tibia on the injured side is compared
with that on the noninjured side.
Posterior displacement of the tibia
indicates PCL injury.
Passive external rotation of the
tibia relative to the femur with the
knee at 30 and 90 degrees of flexion
should also be examined.1 This is
best evaluated with the patient in the
prone position, but the supine posi-
tion can also be used. The examina-
tion is done by comparing the axis of
Fig. 1 The quadriceps active
test is performed with the
affected hip and knee at 90
degrees of flexion and the
foot resting on the table. One
of the examiner’s hands
restrains the foot of the
affected leg while the patient
attempts to slide the foot
down the table with a quadri-
ceps contraction. In the PCL-
deficient knee, the tibia is
posteriorly subluxated (left).
A quadriceps contraction
causes anterior tibial sublux-
ation, which is visible when
the examiner is observing
the tibial movement from
the affected side (right).
69
Posterior Cruciate Ligament Injuries
the medial border of the foot relative
to the femur. 1 With the patient
placed prone, the foot is forcefully
externally rotated, and the degree of
external rotation of the foot is com-
pared with that on the noninjured
side (Fig. 2). External rotation of the
injured knee 10 degrees or more than
can be achieved in the noninjured
knee is considered significant. In
addition, the tibial condyles are pal-
pated to determine their position rel-
ative to the femur. This component
of the examination ensures that the
increased external rotation is from
posterolateral, not anteromedial,
instability. Increased external rota-
tion at 30 degrees that decreases
at 90 degrees indicates isolated
injury to the posterolateral corner.4
Increased external rotation at both
30 and 90 degrees indicates injury to
both the PCL and the posterolateral
corner.
Varus and valgus stress tests are
performed at full extension and 30
degrees of flexion. Increased varus
opening at 30 degrees of flexion indi-
cates LCL and possibly posterolat-
eral corner injury. Slightly increased
varus opening at full extension is
consistent with combined injury to
the LCL and posterolateral corner.
Fig. 2 The prone external
rotation test with the
patient’s knees flexed 30
degrees. The feet are exter-
nally rotated by the exam-
iner. External rotation of the
affected foot relative to the
thigh is compared with that
on the normal side. The
test result is considered
significant if external rota-
tion on the affected side is 10
degrees or more greater than
that achieved on the normal
side. This test is also per-
formed with the patient’s
knees flexed 90 degrees.
70
Significant varus opening at full
extension indicates additional injury
to the PCL and possibly the ACL.1
Significant valgus opening at 30
degrees of flexion indicates medial
collateral ligament (MCL) injury,
which is commonly seen with PCL
injury.
The external rotation recurvatum
test and the reversed pivot shift test
are also used to identify PCL and
associated injuries.9,10 The external
rotation recurvatum test involves
grasping the great toe with the knee
in extension while the patient is
supine.1 A positive sign occurs when
the knee falls in varus, hyperexten-
sion, and external rotation. This test
was originally thought to indicate
isolated posterolateral injury. How-
ever, when excessive varus and
hyperextension are present, injury to
the ACL and possibly the PCL is also
present. The reverse pivot shift test
has been used to diagnose postero-
lateral instability. 10 This test is
significant only if a positive result is
found to a greater degree in the
injured knee than in the noninjured
knee.1 Normal intact knees may have
a positive reverse pivot shift; this
correlates directly with generalized
ligament laxity.
Journal of the American Academy of Orthopaedic Surgeons
Diagnostic Studies
Instrumented knee testing and mag-
netic resonance (MR) imaging can be
used to confirm the diagnosis of PCL
injury. The most useful application
of instrumented knee testing is the
quadriceps active test performed
with a knee-ligament arthrometer as
described by Daniel et al.9 Magnetic
resonance imaging has proved to be
sensitive and specific in the diagno-
sis of acute PCL injury11 and can be
used to identify meniscal and chon-
dral pathologic changes. Magnetic
resonance imaging can also be used
to detect acute partial PCL tears,
which generally present as painful
knees without significant posterior
instability on physical examination.
Radiographs are useful in docu-
menting PCL avulsion fractures and
degenerative changes associated
with PCL injury. We routinely
obtain standing AP radiographs in
full extension and posteroanterior
(PA) radiographs in 45 degrees of
flexion to assess the presence of com-
partment wear. Merchant views are
used to evaluate the patellofemoral
compartment. Standing weight-
bearing radiographs in full exten-
sion from the hip to the ankle are
obtained in cases of combined PCL
and posterolateral instability to rule
out varus alignment that would
require proximal tibial valgus
osteotomy prior to consideration of
ligament reconstruction.1
Natural History and
Clinical Results
Knowledge of the natural history
and the results of nonoperative and
surgical treatment is important when
deciding on proper treatment of the
PCL-injured knee. Parolie and
Bergfeld3 reported long-term results
of nonoperative treatment of isolated
PCL injuries. At an average follow-
 Daniel M. Veltri, MD, and Russell F. Warren, MD

up of 6.2 years, 80% of the patients
were satisfied with their results, and
84% had returned to their previous
sport. Rehabilitation of the quadri-
ceps on the injured side to 100% of
the strength on the noninjured side
correlated with a successful result of
the rehabilitative treatment. Fowler
and Messieh5 reviewed the results of
treatment of seven complete isolated
PCL tears and five partial tears. All
patients returned to their previous
activity and experienced no limita-
tions in their injured knee. Torg et al8
reviewed the data on 14 patients
with straight posterior instability
and 29 with combined multidirec-
tional instability. The patients with
straight posterior instability had bet-
ter functional results than the
patients with multidirectional insta-
bility. Patients with better functional
results were more likely to have
greater quadriceps strength in the
affected extremity.
Whether the PCL-deficient knee is
at risk for the development of degen-
erative changes is not clear at this
time because there are no pertinent
prospective studies. In such a study,
all patients would be followed up to
determine whether chronic articular
injury occurs subsequent to or inde-
pendent of acute chondral injury.
Despite the lack of prospective stud-
ies, it appears that progressive
degenerative changes may occur in
some PCL-deficient knees.7,8
In theory, compartment degener-
ation could result from acute chon-
dral injury associated with PCL
injury or from increased joint-con-
tact forces created by the absence of
the PCL. Skyhar et al 12 used a
cadaver model to show that isolated
sectioning of the PCL leads to
increased medial and patellofemoral
compartment pressures. Torg et al8
reported that degenerative changes
noted on radiographs were more
common in patients with combined
instability patterns than in those
with isolated PCL injuries. The
degenerative changes in these
patients involved both the medial
and the lateral compartments.
Clancy et al7 noted no articular dam-
age in 15 acute PCL injuries,
although they reported medial com-
partment changes in chronically
PCL-deficient knees. In their series,
nine of ten patients who underwent
PCL reconstruction more than 4
years after their original injury had
moderate to severe articular injury
to the medial compartment.
The long-term results of surgical
reconstructions for PCL instability
also remain unclear.13 Open reduc-
tion and internal fixation of bony
avulsions and reconstruction with
the central third of the patellar ten-
don have provided good objective
and functional results.7,13 Primary
repair of interstitial tears and PCL
reconstructions with the semitendi-
nosus and gracilis, the iliotibial
band, and the medial gastrocnemius
inconsistently produce good func-
tional results and often fail to pro-
vide objective stability.13
Acute PCL Instability
Nonoperative Treatment
Routine reconstruction is usually
not required for the treatment of iso-
Acute PCL avulsions
Large fragment
Posterior tibial
Open reduction translation
and internal fixation <10 mm
Quadriceps
rehabilitation
Fig. 3 Treatment algorithm for PCL avulsion fractures.
lated acute PCL injuries. The degree
of posterior translation is important
in assessing an isolated PCL injury. If
it is less than 10 mm, as in the major-
ity of isolated injuries, a nonopera-
tive aggressive rehabilitative
program should be utilized. If the
posterior translation is greater than
10 to 15 mm, reconstruction is
advised, since it is likely that addi-
tional secondary restraints have been
compromised, although this may not
be apparent on physical examina-
tion. Associated ligament injuries
identified by physical examination
or at surgery should be repaired or
reconstructed. Greater laxity in the
acutely PCL-deficient knee may
increase the risk of development of
degenerative joint disease.
Radiographs are used to docu-
ment the presence of PCL avulsion
fractures and osteochondral injury
(Fig. 3). The PCL is not recon-
structed when small tibial PCL avul-
sion fractures are present and
posterior translation of the tibia at
90 degrees of flexion is less than 10
mm. If the avulsed fragment is small
and posterior translation at 90
degrees of flexion is greater than 10
to 15 mm, the PCL should be recon-
structed. If the avulsed fragment is
large (i.e., can be internally fixed
with a 4.0-mm cancellous screw),
Small fragment
Posterior tibial
translation
>10–15 mm
PCL reconstruction

Vol 1, No 2, Nov/Dec 1993 71

Posterior Cruciate Ligament Injuries

fixation is warranted. For large tibial return to athletic activity. In the Operative Treatment
avulsions this is performed by a authors’ experience, athletes with iso- If an acute PCL injury is present
posterior approach as described by lated acute PCL injuries without and the posterior displacement is
Burks and Schaffer.14 associated chondral or meniscal greater than 10 to 15 mm at 90
Magnetic resonance imaging is injuries can return to their sport in 3 degrees of flexion, reconstruction or
used to document the location of the to 4 weeks, but that return must be augmentation of the PCL should be
PCL tear and the presence of associ- based on the individual patient’s performed (Fig. 4). If a grade III
ated meniscal or chondral injury in progress; on occasion, return to sport MCL, ACL, or posterolateral injury
acute tears that are amenable to non- can take significantly longer. is present in association with a PCL
operative treatment. The finding of
increased signal intensity on the T2
images suggests osseous and possi-
bly chondral injury. If significant Acute PCL tear
chondral injury is suspected, one
should perform arthroscopy to eval-
uate the status of the articular carti- Clinical and arthrometric examination
lage. Meniscal injury is relatively
infrequent in acute isolated PCL rup-
tures. If a vertical longitudinal tear in Isolated PCL tear
Isolated PCL tear
the vascularized portion of the with >10 –15 mm
with <10 mm
medial meniscus is present, we rec- of posterior displacement
of posterior
ommend repair, since isolated sec- displacement
tioning of the PCL has been shown to
increase medial compartment pres- Rehabilitation to
sures in a cadaver model.12 Once the MR imaging regain knee motion
osteochondral and meniscal injuries
have been treated, we proceed with a
rehabilitation program that empha- PCL reconstruction
Isolated
sizes quadriceps strengthening.
PCL tear
Rehabilitation follows the princi-
ples of open- and closed-kinetic- PCL and chondral
chain exercises.15 Open-kinetic-chain or meniscal injury
exercises are performed with the foot
PCL and grade III MCL,
free; knee motion is independent of ACL, or posterolateral
hip and ankle motion. In closed- Examination under injury
kinetic-chain exercises, the foot is anesthesia and
fixed so that knee motion occurs in arthroscopy
concert with hip and ankle motion.
Open-kinetic-chain extension exer-
cises (i.e., seated knee extensions with
No associated
weights) are avoided in PCL rehabil- ligament injury
itation, since they can stress the Acute reconstruction/
patellofemoral joint. The quadriceps repair of all ligament
muscles are rehabilitated with func- injuries
Treat meniscal and
tional closed-chain exercises, such as chondral pathology
squats and leg presses. This nonop-
erative rehabilitative treatment
requires constant maintenance of Rehabilitate quadriceps
quadriceps strength to achieve func- (if adequate strength,
tional success. When the patient’s return to sports)
injured knee has regained 90% of the
quadriceps and hamstring strength Fig. 4 Treatment algorithm for acute PCL injuries other than avulsion fractures.
on the normal side, the patient can

72 Journal of the American Academy of Orthopaedic Surgeons


injury, reconstruction of all ligamen-
tous injuries should be undertaken.
If the knee is grossly unstable, plac-
ing the neurovascular structures at
risk, early reconstruction with a
patellar tendon autograft is per-
formed. In such a case, one must be
concerned that a knee dislocation
might have occurred and sponta-
neously reduced. Prior to surgery,
an angiogram or MR study with vas-
cular imaging capability should be
performed to rule out associated
arterial injury.
With associated posterolateral,
ACL, or grade III MCL injury, it
appears best to operate early (within
1 week) to maximize healing poten-
tial, since late surgery for posterolat-
eral injury has relatively poor
results. Delaying ACL reconstruc-
tion after acute ACL injury to regain
full knee motion and to allow for
capsular healing has been found to
be of benefit in decreasing the inci-
dence of postoperative arthrofibro-
sis. It may be prudent for operative
candidates with acute isolated PCL
tears to undergo a rehabilitative
course to regain knee motion prior to
surgery.
Acute surgical treatment of com-
plete PCL tears can include primary
repair, augmentation, or reconstruc-
tion, depending on the location of
the injury. If the tear is on the bone-
ligament interface, we use the prin-
ciples noted above. Primary repair
of intrasubstance PCL tears should
not be done without augmentation
of the repaired PCL with a semi-
tendinosus and/or gracilis auto-
graft. Alternatively, the defect can be
reconstructed with a patellar tendon
autograft, a semitendinosus or gra-
cilis autograft, or a patellar or
Achilles allograft. The optimal
method for PCL reconstruction is
not clear at this time, but the use of
patellar tendon autografts appears
to result in a higher rate of objective
success. 7,13 Reconstruction with a
patellar tendon autograft is our pre-
Vol 1, No 2, Nov/Dec 1993
Daniel M. Veltri, MD, and Russell F. Warren, MD
ferred method, provided there is
sufficient length of the patellar ten-
don (40 mm or more).
Reconstructions of the PCL can be
performed with open or arthroscop-
ically assisted techniques. If the
arthroscopically assisted technique
is chosen, we recommend fluoro-
scopic control and a posteromedial
portal to assist in tibial tunnel prepa-
ration. 16 This procedure is techni-
cally demanding, particularly
because the patellar tendon graft is
passed at a sharp angle from the
tibia to the femur. This may create
fraying of the patellar tendon graft
and subsequent laxity. If the tibia is
of poor bone quality, the patellar
tendon graft may erode through the
proximal tibia, creating graft laxity.
Most important, the arthroscopically
assisted technique requires a patel-
lar tendon length of 40 mm or more
to maintain the bone blocks within
their tunnels.
Although this procedure can be
done in most cases, in some patients
the autograft patellar tendon will be
too short to allow the bone blocks to
remain in their tunnels, and ade-
quate graft fixation will not be
achieved. A posterior approach can
be used to ensure adequate tendon
length and to avoid an acute angle
for graft passage.14 The femoral PCL
tunnel is prepared with arthro-
scopic assistance. 16 A posterior
arthrotomy is then used to prepare
the proximal tibia for graft place-
ment.14 The tibial bone block is fixed
to the posterior aspect of the tibia
using standard 4.0-mm cancellous
screws. This allows greater length
for passage of the femoral bone
block into its tunnel and a straighter
graft orientation.
In addition to patellar tendon and
semitendinosus or gracilis auto-
grafts, allografts can be used for PCL
reconstruction. Patellar or Achilles
tendon allografts should be longer
than 40 mm to ensure adequate
length for fixation.
If posterolateral or MCL recon-
struction is performed with PCL
reconstruction, additional incisions
are used. The posterolateral corner
can be reconstructed with a biceps
tenodesis or patellar tendon allograft.
The MCL is repaired primarily. If an
ACL reconstruction is needed, this
can also be performed arthroscopi-
cally. The ACL and PCL femoral and
tibial tunnels are prepared first. The
PCL graft is inserted next, followed
by ACL graft insertion. The PCL graft
is fixed with interference screws
while the tibia is centered on the
femur in full extension. The ACL is
then fixed with interference screws
with the knee in 20 degrees of flexion.
If multiple ligament reconstructions
are required, patellar tendon and
semitendinosus/gracilis autografts
can be used. Finally, multiple allo-
grafts can be used to avoid the exten-
sive dissection necessary for multiple
graft harvest.
Postoperative Rehabilitation
Postoperative rehabilitation fol-
lowing PCL reconstruction is
designed to restore range of motion
without stressing the graft. Exercises
that produce posterior tibial transla-
tion are avoided. Limited weight
bearing using crutches is allowed
with a knee brace locked in full exten-
sion to stabilize the joint. Quadriceps
exercises are started on the first post-
operative day with active knee exten-
sion (without weights) from 90 to 0
degrees and straight leg raises. Pas-
sive knee-flexion exercises are used
to gain knee flexion slowly over 6
weeks. Open-kinetic-chain ham-
string exercises (seated leg curls) are
not used, since posterior tibial trans-
lation occurs with open-chain knee
flexion exercises.15 Running begins at
5 months and sport-specific agility
drills at 6 to 7 months following
surgery. Full return to sports is
allowed when adequate quadriceps
and hamstring strength is demon-
strated (90% of that on the noninjured
73
Posterior Cruciate Ligament Injuries
side) and sport-specific agility and
proprioreceptive skills have been
mastered.
Chronic PCL Instability
Treatment of chronic PCL instability
is based on the degree of instability,
the radiographic evidence of degen-
erative changes, and the presence of
symptoms that have not responded
to rehabilitative treatment (Fig. 5).
The surgeon must evaluate the
results of previous surgical or con-
servative treatment. It is important to
note the mechanical alignment, the
patellofemoral function, and the sta-
tus of the medial and lateral com-
partments. Standing AP radiographs
in full extension and 45-degree-
Chronic PCL tear/avulsion
Chronic posterolateral
instability
Standing AP hip-to-ankle
radiograph in extension
Varus Normal alignment
Consider valgus
Standing full-extension AP and
tibial osteotomy
45-degree-flexion PA views
Still symptomatic
posterior instability
No or mild degenerative
changes on radiographs
Consider PCL reconstruction
Fig. 5 Treatment algorithm for chronic PCL injuries.
74
flexion PA radiographs are useful for
documenting early degenerative
knee changes. If the patient’s main
complaint is pain and the symptoms
suggest patellofemoral or medial
compartment disease, a bone scan is
performed. Increased bone-scan
activity may represent the sequelae
of an acute chondral injury or altered
weight-bearing forces due to the
absence of the PCL, or it may be
unrelated to the chronic PCL injury.
Whether the chronically PCL-
deficient knee is at risk for progres-
sive degenerative changes is not
known. However, isolated section-
ing of the PCL has been shown to
increase medial and patellofemoral
compartment pressures in a cadaver
model.12 We consider progressively
Chronic pain and/or instability
with >10 –15 mm of
posterior displacement
Rehabilitate quadriceps
Still symptomatic Improvement
Continue rehabilitation
Severe degenerative
changes
Quadriceps Progressively increased
rehabilitation activity on biennial
or osteotomy bone scans
Journal of the American Academy of Orthopaedic Surgeons
increased activity on serial bone
scans to be secondary to altered knee
biomechanics from the absence of
the PCL.
We recommend nonoperative
treatment with quadriceps rehabili-
tation for the majority of patients
with chronic PCL instability. In
these cases, the degree of posterior
laxity alone is not a criterion for
reconstruction; one must also con-
sider the presence of symptoms, the
results of diagnostic studies, and
the results of nonoperative rehabil-
itation. If posterior displacement is
greater than 10 to 15 mm and non-
operative treatment with aggres-
sive rehabilitation has failed, we
consider reconstruction. Recon-
struction is not performed if there is
radiographic evidence of marked
degenerative changes. If associated
posterolateral instability is present,
a standing AP radiograph from the
hip to the ankle is used to assess
mechanical knee alignment. In
knees with posterolateral instability
and varus knee deformity, a valgus
tibial osteotomy is recommended. If
the patient remains symptomatic
following osteotomy, PCL recon-
struction is considered. Patients
selected for a nonoperative aggres-
sive rehabilitative program are fol-
lowed up closely. In the absence of
radiographic evidence of progres-
sive degenerative changes, bone
scans are performed every 2 years
to see whether bone-scan activity is
increasing.
Although there are no prospective
studies that document that PCL
reconstruction can prevent the devel-
opment of degenerative knee
changes or return bone-scan activity
to normal, we recommend PCL
reconstruction if early radiographic
evidence of mild degenerative
change or progressively increased
bone-scan activity is noted. We have
found that reconstruction can
improve stability and decrease pain
in such cases. The technique for

chronic reconstruction is the same as
that outlined for arthroscopically
assisted acute reconstruction. 16 If
patellofemoral degenerative changes
are present, one can use a contralat-
eral patellar tendon autograft, a
semitendinosus or gracilis autograft,
or a patellar or Achilles tendon allo-
graft for reconstruction to avoid any
effect of graft harvest on the
patellofemoral joint. Rehabilitation is
similar to that after acute reconstruc-
tion.
Summary
Although PCL tears are estimated to
account for 3% to 20% of all knee lig-
ament injuries, these injuries are
commonly missed at initial evalua-
tion. 1,2 The natural history of the
References
1. Cooper DE, Warren RF, Warner JJP: The
posterior cruciate ligament and postero-
lateral structures of the knee: Anatomy,
function, and patterns of injury. Instr
Course Lect 1991;40:249-270.
2. Clendenin MB, DeLee JC, Heckman JD:
Interstitial tears of the posterior cruciate
ligament of the knee. Orthopedics
1980;3:764-772.
3. Parolie JM, Bergfeld JA: Long-term
results of nonoperative treatment of iso-
lated posterior cruciate ligament
injuries in the athlete. Am J Sports Med
1986;14:35-38.
4. Gollehon DL,Torzilli PA, Warren RF:
The role of the posterolateral and cruci-
ate ligaments in the stability of the
human knee: A biomechanical study. J
Bone Joint Surg Am 1987;69:233-242.
5. Fowler PJ, Messieh SS: Isolated poste-
rior cruciate ligament injuries in ath-
letes. Am J Sports Med 1987;15:553-557.
6. Kannus P, Bergfeld J, Jarvinen M, et al:
Injuries to the posterior cruciate ligament
of the knee. Sports Med 1991;12:110-131.
Vol 1, No 2, Nov/Dec 1993
Daniel M. Veltri, MD, and Russell F. Warren, MD
PCL-injured knee and the results of
nonoperative and surgical treatment
provide some guidelines for man-
agement of these injuries.5-8,16 In acute
isolated PCL tears with less than 10
mm of posterior laxity at 90 degrees
of flexion, current knowledge sug-
gests nonoperative treatment that
stresses aggressive quadriceps reha-
bilitation. In acute PCL tears with
more than 10 to 15 mm of posterior
laxity at 90 degrees of flexion or
combined ligamentous injury, the
PCL should be reconstructed with a
patellar tendon autograft, a semi-
tendinosus or gracilis autograft, or,
in selected cases, a patellar or
Achilles tendon allograft. We recom-
mend a patellar tendon autograft for
the majority of PCL reconstructions.
In combined-ligament injuries, all
ligamentous injuries should be
7. Clancy WG Jr, Shelbourne KD, Zoellner
GB, et al: Treatment of knee joint instabil-
ity secondary to rupture of the posterior
cruciate ligament: Report of a new proce-
dure. J Bone Joint Surg Am 1983;65:
310-322.
8. Torg JS, Barton TM, Pavlov H, et al: Nat-
ural history of the posterior cruciate lig-
ament-deficient knee. Clin Orthop
1989;246:208-216.
9. Daniel DM, Stone ML, Barnett P, et al:
Use of the quadriceps active test to diag-
nose posterior cruciate-ligament disrup-
tion and measure posterior laxity of the
knee. J Bone Joint Surg Am 1988;70:
386-391.
10. Jakob RP, Hassler H, Staeubli HU:
Observations on rotatory instability of
the lateral compartment of the knee:
Experimental studies on the functional
anatomy and the pathomechanism of
the true and the reversed pivot shift
sign. Acta Orthop Scand Suppl 1981;
52:1-32.
11. Grover JS, Bassett LW, Gross ML, et al:
reconstructed. Small acute PCL
avulsion fractures with more than 10
mm of posterior laxity are treated
with PCL reconstruction. All large
PCL avulsion fractures are treated
with internal fixation. All chronic
PCL injuries are initially treated
with a nonoperative aggressive
rehabilitation program. Reconstruc-
tion should be performed in chronic
PCL injuries when laxity is more
than 10 to 15 mm at 90 degrees of
knee flexion, minimal radiographic
degenerative changes are present,
and a nonoperative aggressive reha-
bilitation program has failed. Proper
diagnosis, the knowledge of the nat-
ural history, and the results of surgi-
cal and nonoperative treatment
provide the rationale for current
management of the PCL-injured
knee.
Posterior cruciate ligament: MR imag-
ing. Radiology 1990;174:527-530.
12. Skyhar MJ, Warren RF, Ortiz GJ, et al:
The effects of sectioning of the posterior
cruciate ligament and the posterolateral
complex on the articular contact pres-
sures within the knee. J Bone Joint Surg
Am 1993;75:694-699.
13. Veltri DM, Warren RF, Silver G: Com-
plications in posterior cruciate ligament
surgery. Operative Techniques Sports Med
1993;1:154-158.
14. Burks RT, Schaffer JJ: A simplified
approach to the tibial attachment of the
posterior cruciate ligament. Clin Orthop
1990;254:216-219.
15. Lutz GE, Palmitier RA, An KN, et al:
Comparison of tibiofemoral joint forces
during open-kinetic-chain and closed-
kinetic-chain exercises. J Bone Joint Surg
Am 1993;75:732-739.
16. Warren RF, Veltri DM: Arthroscopically
assisted posterior cruciate ligament
reconstruction. Operative Techniques
Sports Med 1993;1:136-142.
75
JAAOS Home Page

Table of Contents

Search

Help
Metastatic Tumors of the Spine: Diagnosis and Treatment
Kevin D. Harrington, MD

Abstract

Metastatic disease of the spine occurs in as many as 70% of patients with dissem- vertebral destruction; and (5) devel-
inated cancer and may result in vertebral collapse, spinal instability, and progres- opment of spinal instability from
sive neurologic compromise. Today, magnetic resonance imaging is the most such a fracture, particularly when
effective means of differentiating benign from malignant causation of vertebral col- associated with lytic destructive
lapse, based on the imaging patterns and extent of marrow ablation. The more changes in the posterior elements.
rapid the onset of the neurologic deficit, the worse the prognosis for recovery, no Spinal cord and/or nerve-root
matter what treatment is instituted. The majority of vertebral lesions requiring compression occurs in approximately
decompression and stabilization emanate from the vertebral body and are best man- 5% of patients with widespread can-
aged by anterior decompression and stabilization alone. With posterior element cer. The most common cause of this
destruction, spinal subluxation through the involved segment, or involvement of compression is the extrusion of tumor
the lumbar spine, a combination of both anterior and posterior stabilization is tissue and detritus of bone or disk into
required. The author’s preference is to perform anterior vertebral replacement with the spinal canal following the partial
methylmethacrylate incorporating a Knodt distraction rod. This construct affords collapse of a vertebral body that has
instantaneous stability that is not adversely affected by postoperative irradiation. been infiltrated and weakened by a
Many devices can provide adequate posterior stabilization, but the author prefers metastatic deposit.
to use Luque rods with sublaminar wire fixation. In a series of 77 patients with
major neurologic compromise treated with this technique, 62% showed improve-
ment by at least two Frankel grades, compared with fewer than 5% who improved Radiographic Findings
after laminectomy decompression with or without irradiation. Nineteen of the 77
patients remained alive more than 4 years postoperatively. Plain radiographs of a symptomatic
J Am Acad Orthop Surg 1993;1:76-86 patient typically will demonstrate
either an anterior compression
deformity with secondary kyphosis
The spine is the most common site Approximately 70% of patients (Fig. 1) or a more uniform vertebral
for skeletal metastases, irrespective who die of cancer have evidence of collapse usually associated with
of the primary tumor involved. The vertebral metastases apparent on posterior column destruction and
vertebral body typically is affected careful postmortem examination. focal spinal instability (Fig. 2). Of
first because of its rich blood supply Three fourths of these lesions origi- course, either of these bony deformi-
and sinusoidal vascular distribution. nate from carcinoma of the breast, ties can also result from osteopenic
However, the initial radiographic prostate, kidney, or lung or from changes unrelated to malignancy,
finding often is destruction of a less myeloma or lymphoma. However, due to a variety of causes. Primary
well vascularized pedicle. This para- vertebral metastases often are vertebral neoplasms or indolent ver-
dox is explainable by the fact that asymptomatic and may be discov- tebral osteomyelitis also may
between 30% and 50% of a vertebral ered only on routine bone scans. progress to cause vertebral collapse
body must be destroyed before any When symptoms do develop, they and a lesion difficult to differentiate
changes can be recognized radio- are a consequence of one or more of
graphically, unless there is a blastic the following: (1) an enlarging mass
or sclerotic reaction. In contrast, within the vertebral body, which
minimal lysis of pedicular bone can may break through the cortex and Dr. Harrington is Clinical Associate Professor,
Department of Orthopaedic Surgery, University
be appreciated because the cortex of invade paravertebral soft tissues; (2)
of California, San Francisco.
the pedicle tends to be involved early compression or invasion of adjacent
and because the pedicle can be seen nerve roots; (3) compression of the Reprint requests: Dr. Harrington, 3838 Califor-
well in cross section on conventional spinal cord; (4) development of a nia Street, Suite 516, San Francisco, CA 94118.
anteroposterior radiographs. pathologic fracture secondary to

76 Journal of the American Academy of Orthopaedic Surgeons


Fig. 1 Radiograph of a 66-year-old woman
with known breast cancer and scintigraphi-
cally demonstrable metastases to T-11 and
T-12. Although the wedge compression
fractures demonstrated presumably are sec-
ondary to metastases, their appearance on
plain radiography is indistinguishable from
that of benign pathologic fractures sec-
ondary to osteoporosis.
from metastatic disease. Even
patients with known metastatic dis-
ease of the spine may develop col-
lapse or instability at other spinal
levels due to nonmalignant causes.
All of these processes initially
present as back pain of sudden or
insidious onset, with or without
neurologic compromise. A history
of progressive quadriparesis or even
of specific radiculopathy is of mini-
mal benefit in helping to differenti-
ate among the various potential
causes of spinal deformity. The oft-
quoted maxim that sudden fracture
myelopathy invariably is the result
of acute trauma has been repeatedly
proved invalid, just as the concept
that acute trauma never results in
gradual or progressive neurologic
compromise has been proved
wrong.
Vol 1, No 2, Nov/Dec 1993
Other Diagnostic Studies
The availability today of a variety of
imaging modalities has enhanced our
ability to differentiate between
benign and malignant spinal defor-
mity on the basis of distribution of
abnormalities in the spine as well as
specific patterns of focal bony
destruction. Technetium-99m scintig-
raphy often will demonstrate multi-
ple sites of radioisotope uptake in
other vertebrae, long bones, ribs, or
the skull typical of generalized skele-
tal metastases, even when a patient’s
symptoms and plain radiographs
suggest isolated involvement of a sin-
gle spinal level (Fig. 3).
The most helpful and sensitive
study, however, has been magnetic
resonance (MR) imaging, because
this technique most effectively
delineates the extent and pattern of
Fig. 2 Spontaneous fracture of L-1 from
known metastatic breast cancer. Osteolysis
of all three columns of the spine resulted in
symmetrical vertebral collapse and focal
instability.
Kevin D. Harrington, MD
marrow involvement within an
affected vertebra. Characteristically,
the malignant pathologic fracture
occurs because virtually the entire
vertebral body has been infiltrated
by tumor. The tumor spreads ini-
tially through the hematopoietic tis-
sue and only later progressively
destroys bone. In contrast, benign
compression fractures occur because
the bone substance itself has been
lost or weakened, with hematopoi-
etic tissue remaining relatively
intact. In both instances, the disk
remains unaffected, thus helping to
differentiate either lesion from
osteomyelitis (Fig. 4).
An MR image of a benign com-
pression fracture typically reveals
preservation of the normal marrow
signal, although there may be dis-
placement of the marrow along vec-
tors created by the compression
deformity. This phenomenon is par-
ticularly apparent in the T1-weighted
image, where the combination of the
hematopoietic tissue, edema, and
bleeding increases the focal water
signal and the consequent intensity of
that signal (Fig. 5)
An acute benign compression
fracture of the superior endplate
typically causes temporary linear
striation of the marrow distribution
in the rest of the vertebra, particu-
larly on T1 imaging. This finding
usually occurs in a uniform pattern
and is reversible as fracture healing
occurs. 1 The T2-weighted image
shows bone-marrow signal intensity
in the fractured bone similar to that
in the rest of the vertebral body.
In contrast, the MR imaging of a
compression fracture secondary to
metastatic malignancy reveals total
or subtotal replacement of the nor-
mal bone by tumor. This is reflected
by a decreased-signal-intensity
(darker) image on T1-weighted
images (Fig. 6) and increased inten-
sity on T2 images. There may be
incomplete replacement of marrow,
but its pattern will be irregular,
77
Metastatic Tumors of the Spine
Fig. 3 Anterior whole-body radionuclide
image of a patient with prostatic carcinoma
reveals multiple foci of increased tracer
deposition in the shoulders, ribs, lumbar
spine, pelvis, and proximal femora.
reflecting focal destruction rather
than uniform compression of
hematopoietic tissue and fat.
Although MR imaging has a high
level of sensitivity, its specificity
may become blurred when an acute
benign fracture is associated with
marked edema and bleeding into
the marrow space. The T1 signal
may mimic the typical tumor pat-
tern (Fig. 7). Bulging of the partially
collapsed vertebral body and dif-
fuse marrow signal changes extend-
ing into the pedicles may be
78
strongly suggestive of tumor
infiltration. In these instances, or in
any situation in which an occult
symptomatic vertebral metastasis is
suspected, early biopsy of the lesion
is warranted.
Computed tomography (CT)-
directed needle biopsy is accurate
and safe and has virtually replaced
open or percutaneous trocar biopsy
in most centers. In the event of an
equivocal or nondiagnostic speci-
men, the CT-directed biopsy should
be repeated at different areas of the
affected vertebra before resorting to
open biopsy techniques.
Clinical Course
Once the presence of spinal metas-
tases has been established, treatment
options can be considered. As
already noted, it is common for ver-
tebral metastases to be asymp-
Fig. 4 Sagittal MR image of the lumbar
spine of a 66-year-old man receiving
chemotherapy for metastatic prostatic carci-
noma. Spontaneous hematogenous
osteomyelitis developed at L4-5.
Journal of the American Academy of Orthopaedic Surgeons
Fig. 5 Sagittal T1-weighted MR image
shows two benign compression fractures
with incomplete bone marrow replacement
and peripheral low-signal-intensity band
(arrows).
tomatic and to be diagnosed only
with the use of routine bone scintig-
raphy. Such a finding may prompt
the oncologist to alter the patient’s
chemotherapy or hormonal manipu-
lation, but no specific additional
measures are indicated. If spinal
pain develops, it is essential to clar-
ify whether it is attributable to
tumor destruction or to local phe-
nomena such as osteoporosis or
arthritis, particularly because corti-
costeroids or chemotherapy given as
part of systemic cancer treatment
may result in marked osteopenia
(Fig. 1). Insufficiency fractures of the
spine due to local irradiation may
appear years after treatment has
been completed. Debilitated cancer
patients who are receiving che-
motherapy typically become chroni-
cally pancytopenic and are at
increased risk for hematogenous
osteomyelitis involving the spine
(Fig. 4).
When spinal metastases truly are
the source of pain, that pain is usu-
ally of gradual onset, is relentlessly

Fig. 6 Sagittal T1-weighted MR image of
the cervical spine of a 69-year-old woman
with widely metastatic breast carcinoma.
Multiple foci of abnormal replacement of
the marrow signal are particularly apparent
in the C-1, C-2, C-4, and C-8 vertebral bod-
ies.
progressive over weeks or months,
is worse at night, and is unassociated
with significant elevations of white
blood cell count or sedimentation
rate. This type of pain has been
attributed to stretching of the perios-
teum by direct pressure of the
expanding tumor or to microfrac-
tures occurring sequentially within
weakened bone. Another potential
source of pain is from compression
of the ventral aspect of the dura,
which is richly innervated with noci-
ceptor fibers. Such pain can occur
before there is evidence of neuro-
logic involvement. Pain can also
result from invasion of paraverte-
bral structures, sometimes produc-
ing neurologic symptoms from
involvement of the lumbosacral
plexus.
Not infrequently, the patient will
localize the pain at a level below the
actual metastatic lesion. This may
lead the unsuspecting physician to
attribute initial symptoms to arthri-
tis or disk disease and to continue
Vol 1, No 2, Nov/Dec 1993
conservative and ineffective treat-
ment in the face of progressive neu-
rologic compromise. The presence
of radicular pain may help to locate
the level of vertebral involvement.
Fig. 7 Images of a 72-year-old woman with
sudden onset of severe thoracolumbar pain
without trauma. Top, Sagittal T1-weighted
image shows marked homogeneously
decreased signal intensity with posterior
bulging of the vertebral cortex into the
canal. Bottom, Axial T1-weighted image
shows that abnormal signal changes extend
into both pedicles. Both T2-weighted
images were interpreted as suggestive of
tumor infiltration of the vertebral body, but
biopsy revealed only osteoporosis.
Kevin D. Harrington, MD
Approximately 50% of patients with
thoracic cord impingement com-
plain of radicular pain before they
develop symptoms of cord involve-
ment. Such pain often is described
as “girdle pain,” particularly with
lesions at T-9 or below, and may not
be recognized as reflective of inter-
costal root irritation.2
With more central neural involve-
ment, motor deficits usually precede
sensory changes because of the typi-
cally anterior location of cord com-
pression. Loss of sphincter control is
thought to be a late phenomenon,
and usually occurs only in patients
with profound cord involvement.
However, cauda equina involve-
ment can occur acutely or subtly in
patients with involvement of the
conus medullaris. Sphincter func-
tion should be carefully and sequen-
tially evaluated. The sensory level
often is not a reliable indicator of the
level of cord compression, com-
monly being recorded several seg-
ments below the site of fracture or
tumor extrusion into the spinal
canal.
The rapidity of onset of muscle
weakness has considerable bearing
on the prognosis. Constans et al3
reported that 166 of 600 patients
(28%) had an acute onset with a
delay of less than 48 hours between
the manifestation of initial symp-
toms and the appearance of maximal
neurologic compromise. These
patients had the worst prognosis for
recovery, no matter what treatment
was rendered. Patients with a slower
evolution of neurologic compro-
mise, indicating in most instances a
slower growth rate of the metastasis
and a sparing of the anterior spinal
artery, had a decidedly better prog-
nosis. Tarlov and Herz 4 demon-
strated experimentally that even
major neurologic compromise
caused by gradual cord compression
was reversible for a longer period
than was compromise due to an
acute cord lesion. Conversely, a
79
Metastatic Tumors of the Spine
sudden onset of paralysis is almost
invariably associated with a poor
prognosis, probably primarily
attributable to vascular compromise.
Nonoperative Treatment
The philosophy of treatment for ver-
tebral metastases has changed con-
siderably in the past two decades.
With improvement in chemotherapy
and hormonal manipulation, many
patients with bony metastases now
survive for long periods without
premorbid involvement of vital
organs. Consequently, progressive
vertebral metastases are often
apparent in patients with a pro-
longed life expectancy, and the
prospect of ultimate spinal instabil-
ity and neurologic compromise
becomes of increasing concern.
Most patients with spinal metas-
tases do not develop progressive
spinal instability or neurologic
involvement and can be treated suc-
cessfully with systemic chemother-
apy, local irradiation, or temporary
bracing. Primary tumor types vary
in radiosensitivity after metastasis
(Table 1). Even those who sustain a
pathologic compression fracture of
one or more vertebral bodies often
can be treated effectively with tem-
porary bed rest and soft bracing, as
is done for pathologic compression
fractures due to osteoporosis. In my
experience, approximately 80% of
patients with spinal metastases can
be treated effectively with one of
these nonoperative modalities.2,5
When metastases are causing
minimal bone destruction and pain
appears to be the result of periosteal
expansion or reaction within the
bone to tumor, radiation therapy
alone often is the ideal means of
achieving relief. If the tumor
extends into the epidural space,
causing early neurologic compro-
mise, radiation therapy usually
leads to recovery unless the cord or
80
Table 1
Radiosensitivity of Common
Metastases
High sensitivity
Myeloma
Lymphoma
Moderate sensitivity
Colon
Breast
Prostate gland
Lung
Squamous cell
Low sensitivity
Renal
Thyroid
Melanoma
Metastatic sarcoma
nerve roots are compressed by frag-
ments of bone or disk detritus. Radi-
ation therapy also should be the
primary treatment modality in
patients with an anticipated survival
of 4 months or less or with vertebral-
body lesions affecting multiple lev-
els of the spine.
The threshold for radiation com-
plications, including myelopathy,
radiation osteitis, interference with
wound healing, and interference
with graft incorporation consistently
appears to be between 3,000 and
3,500 cGy. Because the control of
local tumor recurrence in the spine
does not seem to improve with doses
in excess of 3,000 cGy, it is generally
recommended that local irradiation
be limited to this dose level. In any
case, adjunctive irradiation should
be postponed for a minimum of 3 to
4 weeks after any operative interven-
tion to limit interference with wound
healing and graft incorporation.
Operative Management
The principal indications for opera-
tive intervention are progressive neu-
rologic compromise and intractable
mechanical spine pain unresponsive
Journal of the American Academy of Orthopaedic Surgeons
or unlikely to be responsive to irradi-
ation or bracing. Decompression is
particularly indicated when cord or
root compression is due to retropulsed
bone or disk fragments or when spinal
instability or malalignment causes
neural compromise. Other specific
indications include radioinsensitive
tumors, recurrence of cord compres-
sion following adequate local irradia-
tion, and presumed metastases when
the primary tumor is occult.
Two decades ago, “operative
intervention” usually meant lam-
inectomy decompression. The
results of this procedure for the
management of advanced spinal
metastases were dismal. The major-
ity of patients with neurologic com-
promise did not improve. Instead,
progressive spinal deformity and
instability frequently developed as a
result of, rather than in spite of, the
decompression. In a large retro-
spective series, Gilbert et al6 demon-
strated that radiation therapy alone
was as effective as decompressive
laminectomy (with or without radi-
ation) in the treatment of epidural
cord compression. After either
treatment, fewer than 50% of
patients regained the ability to walk.
It was only after the evolution of
anterior spinal decompression and
stabilization techniques that the clin-
ical results showed dramatic
improvement.2,7 In the vast majority
of patients, tumor originates from
the vertebral body or soft tissue ante-
rior to the spinal cord and cannot be
decompressed adequately from a
posterior laminectomy approach.
When the entire vertebral body (both
anterior and middle columns)
becomes weakened by tumor lysis,
the vertebral body begins to collapse,
and the bending moment of the spine
shifts posteriorly. As this worsens,
the compression load on the remain-
ing vertebral body increases geomet-
rically, leading to a progressive
kyphotic deformity and ultimately to
extrusion of tumor tissue, disk, and

bony detritus posteriorly into the
spinal canal (Fig. 8).
Ordinarily the posterior elements
(posterior column) are minimally
involved, and posterior tensile sta-
bility remains intact. In such a situ-
ation, overall spinal stability can be
restored entirely through an anterior
approach. However, if tumor
destruction of the posterior elements
(particularly the pedicles) is
advanced, the greatly increased ten-
sile loads posteriorly cannot be
resisted. Typically, a forward-shear-
ing deformity will develop (Fig. 2),
further compromising the spinal
canal and necessitating both anterior
and posterior decompression and
stabilization.
If the previously mentioned indi-
cations for operative intervention
are present, the surgeon must con-
sider separately the issues of decom-
pression and stabilization. For any
given patient with spinal cord or
cauda equina compromise, decom-
pression should be recommended as
soon as a clear-cut motor deficit is
Fig. 8 Replacement of the vertebral body
by tumor results in collapse of the body,
increasing kyphosis, and extrusion of tumor
and bone fragments into the epidural space.
Vol 1, No 2, Nov/Dec 1993
Kevin D. Harrington, MD
apparent, but only if that deficit cor-
relates with a demonstrable focus of
spinal canal intrusion by tumor or
bony debris. In my experience, nei-
ther systemic corticosteroids nor
emergency local irradiation is
beneficial in such circumstances.
The rare syndrome of progressive
sensory loss in the absence of motor
deficit may respond to local irradia-
tion, particularly if a peridural tumor
mass is apparent without major
spinal instability or bony debris
within the canal. However, the sur-
geon must be aware of the fact that
numbness and paresthesias, particu-
larly if peripheral, more often are
attributable to the neurotoxic effect
of certain chemotherapeutic agents.
One must also be wary of attribut-
Fig. 9 Unusual “napkin-ring” constriction
ing progressive motor compromise of the cord caused by a metastatic tumor
to irradiation-induced transverse within the spinal canal growing around the
myelitis unless a gadolinium- dura and compressing the cord circumfer-
entially. In such cases both anterior and pos-
enhanced MR imaging study clearly terior decompression and stabilization are
demonstrates changes consistent usually necessary.
with that diagnosis. In my experi-
ence, it is far more likely for progres-
sive motor deficits to be caused by
gradual spinal instability or local functionally intact, at least in the cer-
tumor recurrence than by the late vical and thoracic spine, restoration
effects of irradiation. Patients with of stability can be achieved by ante-
intractable pain secondary to spinal rior vertebral reconstruction alone.
instability who do not have neuro- If all three columns are severely
logic compromise do not require weakened, combined anterior and
emergency operative intervention. posterior stabilization is essential.
Such patients may enjoy sufficient The only exception to this general
relief from external bracing, render- rule pertains to the lumbar spine.
ing spinal stabilization unnecessary. Because of its lordotic curvature and
If elective surgery is required, the extent of weight-bearing torque
chemotherapy must be discontinued and lateral bending forces to which
early enough to allow correction of it is subjected, I believe that both
anemia and recovery of white blood anterior and posterior stabilization
cell and platelet counts. are necessary in all instances in
Spinal canal compromise from which spinal decompression is
posterior extrusion of the vertebral required (Fig. 10).
body can be decompressed only The surgeon should strive to
from an anterior approach. Com- achieve instantaneous and rigid
bined anterior and posterior cord intraoperative stability and should
compression (so-called napkin-ring not depend on gradual incorpora-
compression) usually must be tion of bone grafts to restore late
relieved by both anterior and poste- local rigidity. There is abundant evi-
rior approaches (Fig. 9). If the poste- dence that, with rare exceptions,
rior column structures remain bone grafts will not be incorporated
81
Metastatic Tumors of the Spine

A B C

Fig. 10 Radiographs of a 65-year-old woman with multiple myeloma, progressive tumor infiltration, and collapse of the L-3 vertebral body.
A, The patient presented with a rapidly progressive cauda equina syndrome (Frankel grade C) despite 4,500 cGy of local irradiation. After
anterior L-3 vertebrectomy and replacement by methylmethacrylate incorporating a Knodt rod, a posterior four-level stabilization was
accomplished with Luque rods and sublaminar wire fixation. The patient enjoyed a complete neurologic recovery. B, Six years later, a new
compression fracture appeared at L-1, again associated with a progressive cauda equina syndrome. C, The L-1 vertebral body was replaced
using methylmethacrylate incorporating a Rezinian vertebral distractor. The original Luque rods were replaced with longer rods and sub-
laminar wiring spanning seven levels. Pathologic examination of the resected L-1 vertebral body revealed that it had collapsed because of
radiation osteitis, not myeloma.

in the face of postoperative irradia-
tion of the affected area. For these
reasons, I advocate the technique of
replacing the resected vertebral
body with methylmethacrylate,
polymerizing in situ, and incorpo-
rating a distraction-fixation device
that secures the cement mass into the
adjacent normal vertebral endplates.
In my hands, the most effective
device is the Knodt distraction rod
with hooks (Zimmer), which jacks
open the collapsed vertebral space to
its appropriate height and can be
buried entirely within the long axis of
the spine. This fixation construct
does not protrude beyond the verte-
bral bodies, thus protecting adjacent
soft tissues from injury (Fig. 11). The
combination of the methylmethacry-
late and the Knodt rod very effec-
tively resists compression and torque
loads in the cervical and thoracic
spine but requires adjunctive poste-
rior stabilization devices in the lum-
bar spine.
The Rezinian distraction device
functions in a similar manner and
also does not extend beyond the
confines of the vertebral bodies.
However, in my experience, it offers
no advantages over the Knodt rod
and is many times more expensive.
The distraction hook-rod system is
similar in concept to the Knodt rod
but is much bulkier and extends into
the perivertebral soft tissues, caus-
ing a risk of soft-tissue erosion.
Alternative anterior-fixation
devices that depend on screw
fixation across the vertebral bodies
are more complicated to insert, pro-
trude well outside the vertebral col-
umn, and are subject to a higher
incidence of failure because their
means of screw fixation to the verte-
bral bodies is at right angles to the
axial compression load on the spine.
If posterior fixation is necessary, a
variety of devices are available. Their
selection should be based on the
severity of posterior bony destruction
demonstrable in any given patient.
Most commonly, patients with a
metastatic malignant neoplasm
extensive enough to require posterior
stabilization have advanced lysis of
one or more pedicles (in addition to
the vertebral body), which precludes
secure fixation by pedicle screw-and-
rod systems. Distraction or compres-
sion rods with hooks may be used but
have the disadvantage of focusing
the fixation stress at only a few levels
where progressive tumor lysis may
cause late instability. For this reason,
I have usually chosen to use Luque
rods with sublaminar (not spinous
process) wire fixation three levels
above and three below the span of
laminectomy decompression. On
occasion, when the strength of lami-
nar bone at any level is suspect, com-
bining the sublaminar wires with
methylmethacrylate may help to
reduce the tendency of an individual
wire to cut through soft bone at that
level (Fig. 12).

82 Journal of the American Academy of Orthopaedic Surgeons

 Kevin D. Harrington, MD

lapse of the ipsilateral lung for

improved exposure. A chest tube is
required postoperatively for a period
of 48 to 72 hours for pleural drainage
and lung reexpansion. Occasionally,
overnight intubation will be expedi-
ent, particularly for the patient who
is moderately debilitated, has chest
wall or pleural metastases that inter-
fere with ideal ventilation, or shows
evidence of pleural metastases.
The thoracotomy incision is made
one level higher than the highest
affected vertebra, and the rib at that
level is removed. The vertebral bod-
ies are easily visualized through the
thin overlying parietal pleura. By
transecting but not removing one or
two additional ribs below the inci-
sion, it is possible to expose multiple
vertebrae above or below the tumor
focus. By incising the posterolateral
crura of the diaphragm and then
approaching the lumbar spine
retroperitoneally, we have been able
A B to expose from T-8 to L-4 through
Fig. 11 Images of a patient with metastatic breast carcinoma 51⁄2 years after a midthoracic the same thoracotomy incision with
vertebrectomy and anterior stabilization with a Knodt rod and methylmethacrylate. A, Lat- a single rib resected.
eral radiograph demonstrates that the height of the vertebral space has been reconstituted
fully and remains so without evidence of displacement of the construct despite the absence The parietal pleura is incised, ele-
of posterior stabilization. B, CT scans. Top, Section through the vertebral body just above the vated, and reflected to expose the
cement construct. Note that the tip of the Knodt rod hook protrudes slightly in front of the segmental vessels (Fig. 13, A).
anterior longitudinal ligament. Bottom, Section through the methylmethacrylate recon-
struction. Despite the diffraction artifact from the metal rod (arrow), the normal dimensions These are ligated and transected as
of the spinal canal can be appreciated. close to the aorta as possible, thus
minimizing disturbance of the par-
avertebral anastomoses. In more
than 60 such approaches, I have
Operative Technique embolized preoperatively. Olerud seen no evidence clinically of cord
et al8 have described the indications vascular compromise after division
The technique of anterior decom- and technique for this procedure in of up to nine vessels on one side;
pression and stabilization of the detail. In essence, using standard some surgeons, however, feel that
thoracic spine is illustrated in Fig- arteriographic techniques, the spinal evoked potential monitoring
ure 13. Before undertaking the pro- major feeder vessels supplying the is essential as the vessels are sequen-
cedure, the surgeon should attempt tumor focus are catheterized, and a tially ligated. After division of these
to anticipate how aggressive the thickened paste made of moistened vessels, the aorta can be retracted
tumor appears radiographically and morcellized absorbable gelatin carefully, facilitating exposure of
and how vascular the lesion is sponge (Gelfoam) is injected, which the entire anterior aspect of the ver-
likely to be. Large osteolytic effectively obstructs blood flow. tebral bodies involved (Fig. 13, B).
lesions with minimal host bony Anterior stabilization of the tho- Careful blunt dissection is contin-
response are likely to be extremely racic spine requires a thoracotomy, ued subperiosteally to expose the
vascular, particularly if the pri- with exposure of the pericardium, lateral aspect of the affected verte-
mary malignant neoplasm is one lung, and the great vessels. A bra on the opposite side.
myeloma or metastatic hyper- double-lumen endotracheal tube All remnants of the affected verte-
nephroma. Such lesions should be may be employed, permitting col- bra should be resected, together with

Vol 1, No 2, Nov/Dec 1993 83

Metastatic Tumors of the Spine
all tumor tissue. Only by performing
a complete vertebrectomy can the
surgeon be sure of removing every
bit of debris forced into the spinal
canal by the posterior vector of the
kyphotic deformity. The anterior two
thirds of the vertebra can be removed
rapidly with a gouge and rongeur
(Fig. 13, C). When only a thin shell of
bone and tumor tissue remains in
front of the spinal canal, an angled
curet is used to avoid inadvertent
penetration of the dura or damage to
the cord and nerve roots (Fig. 13, D).
Great care is taken to decompress the
canal completely, using the angled
curet to undercut the posterior cor-
ners of the intact vertebrae above and
below the level of resection.
After complete decompression, a
high-speed bur is used to cut a well
into the intact vertebral endplates of
sufficient depth and width to seat
84
Fig. 12 For posterior stabi-
lization, the Luque rods are
cut to appropriate lengths,
interdigitated along the lam-
inar sulcus, and secured by
doubled 16-gauge wires
at e a c h l e v e l ( l e f t ) .
Stability above and below
the laminectomy can be
enhanced by packing meth-
ylmethacrylate into the
areas of wire-rod fixation
(right). This forms a rigid
construct that allows sub-
laminar wire fixation at any
single level to reinforce
every other level.
the Knodt rod and hooks (Fig. 13, E).
As the rod is twisted, the hooks will
become seated firmly into the verte-
brae, and the kyphotic angulation
will be corrected (Fig. 13, F).
A malleable retractor is placed
across the back of the defect to pro-
tect the dura from the heat of poly-
merization and, more important,
from compression by the expanding
cement mass. Methylmethacrylate
then is packed about the rod and
hooks and into the defects in the ver-
tebral endplates (Fig. 13, G). Before
polymerization is complete, all
excess cement is removed from out-
side the confines of the vertebral
bodies. A CT scan of the vertebral
construct should show that the
cross-sectional diameter of the
acrylic-metal construct is nearly
identical to that of the normal verte-
bra, with no encroachment of
Journal of the American Academy of Orthopaedic Surgeons
cement into the spinal canal (Fig. 11,
B). In patients who have a good
prognosis for prolonged survival
and who will not require further
irradiation, cancellous autogenous
bone or allograft may be packed
around the vertebral construct to
enhance the likelihood of bony
arthrodesis.
The decompression-stabilization
procedure in the cervical spine is
much simpler than that in the tho-
racic spine, because an essentially
avascular interval is used for the
approach between the sternomas-
toid and carotid sheath laterally and
the strap muscles, trachea, and
esophagus medially. Ordinarily,
the only vascular structure requir-
ing ligation and transection is the
middle thyroid vein. The technique
for vertebrectomy and distraction-
stabilization is similar to that
described for the thoracic spine and
has been discussed extensively else-
where.2,9
In my experience, the lumbar
spine is the least common location
for metastatic lesions requiring
anterior decompression. This is
fortunate, since it is also the area
where anterior exposure is most
difficult, at least for the L-4, L-5,
and S-1 vertebral bodies. Anterior
stabilization is also most problem-
atic for these lower lumbar levels.
Exposure is best accomplished
through a flank incision, parallel-
ing the inferior costal margin. Dis-
section is retroperitoneal, with the
transversalis fascia and abdominal
contents being displaced medially
until the ureter, vena cava, aorta,
and iliac vessels are encountered.
In patients who have previously
undergone local irradiation, it may
be very difficult to mobilize the
great vessels overlying the L-4 and
L-5 vertebral bodies, and great care
must be taken to avoid tearing the
vena cava. This approach has also
been described extensively else-
where.2 As already noted, because
 Kevin D. Harrington, MD

A B C D

E F G H

Fig. 13 Technique for anterior decompression and stabilization of the thoracic spine. A, Decompression is accomplished by means of a tho-
racotomy with the patient in the lateral decubitus position. B, The aorta is retracted gently, the segmental vessels are ligated and transected,
and the affected vertebral body is easily approached. The presence of a prominent paravertebral extrapleural tumor mass will often assist in
locating the focus of destruction. C, Most of the tumor and bone-disk debris can be removed with a small periosteal elevator. D, As the level
of the posterior cortical margin is approached, further decompression is achieved with an angled gouge. All material adherent to the adja-
cent vertebral body is removed. E, The vertebral space is recreated with a lamina spreader. A small angled curet is used to complete decom-
pression of the spinal canal and to round off the edges of the posterior cortices of adjacent vertebrae. F, The endplates of the adjacent vertebrae
are undercut with a high-speed bur to allow the ends of the Knodt rod and the bodies of its hooks to be buried within the vertebral bone. G,
The Knodt rod has been positioned within the resected space. Twisting distracts its hooks, and their bodies become firmly impacted within
the adjacent vertebral bone. Only the tips of the hooks extend anterior to the vertebral cortex. H, The defect is filled with methylmethacry-
late that polymerizes in situ, incorporating the rod and hooks. To avoid compression of the cord, a malleable retractor is placed between the
expanding mass and the spinal canal.

of the lordotic configuration of the
lumbar spine and because of the
torque and lateral bending
moments encountered there, I
advocate a combination of anterior
decompression-stabilization and
posterior stabilization for all lum-
bar spinal metastases requiring
surgical treatment.
Results
It is essential to discuss, at least
briefly, the overall results for the treat-
ment of patients with spinal instabil-
ity and neurologic compromise from
metastatic malignancy. Only by such
an assessment can the reader deter-
mine for himself or herself whether
the aggressive techniques described
here for selected instances of cord and
root decompression and for spinal
stabilization seem justified.
Frankel et al 10 established a
classification system for quanti-
tating neurologic compromise
(Table 2). With the use of this sys-
tem the extent of sensory and
motor dysfunction can be conve-
niently discussed and the results
of various treatment regimens can
be compared. Although the Frankel
classification relates primarily to
acute traumatic, rather than gradu-
ally progressive, spinal cord com-
promise, it is nevertheless useful as
a means of comparing the efficacy of
different techniques for treating
metastatic spine disease.
Using this system, Nather and
Bose11 reported that fewer than 5%
of patients with Frankel grade A, B,
or C lesions recovered normal
(grade E) or near-normal (grade D)
function after laminectomy decom-
pression. By comparison, in my
series of 77 patients treated by the
techniques of anterior decompres-
sion described herein, 62%
improved to the level of either
grade D or grade E.5 Of 14 patients
with complete paraplegia or quad-
riplegia (grade A), eight improved
at least two grades, and six regained
the ability to walk and have normal
bowel and bladder function.2 The
mean postoperative survival period
for patients with breast metastases,
myeloma, and lymphoma was

Vol 1, No 2, Nov/Dec 1993 85

Metastatic Tumors of the Spine

approximately 28 months. At the

other extreme, patients with lung
cancer metastases had a mean post-
operative survival period of only 8
months. Nineteen patients sur-
vived for more than 4 years postop-
eratively. Twelve had had major
neurologic compromise preopera-
tively, and all 12 had improved by
at least two grades postoperatively.
As expected, the long-term sur-
vivors had primary malignant con-
ditions with good prognoses for
survival, including breast carci-
noma in ten patients and multiple
myeloma in six.
Ten of the 19 survivors required
additional operations for the se-
quelae of other bony metastases,
including four with distant spinal
metastases and two with late local
recurrence. Two patients suffered
posterior wound sloughs through
previously irradiated tissues. There
were no wound-healing problems
with anterior spine approaches. My
Table 2
Frankel Classification System for Neurologic Compromise
Grade A Complete motor and sensory loss
Grade B Complete motor loss; incomplete sensory loss
Grade C Some motor function below the level of involvement; incomplete
sensory loss
Grade D Useful motor function below the level of involvement; incomplete
sensory loss
Grade E Normal motor and sensory function
experience seems comparable with cacy for surgical management of all
that of other clinical investigators spinal metastases. Most patients do
who used similar decompression not continue to suffer severe pain
and stabilization techniques.12-17 after vertebral collapse once they
Based on these results, I believe have completed an initial period of
that patients with major neurologic rest and a course of local irradiation.
compromise or intractable mechani- Most do not experience significant
cal spine pain from vertebral collapse neurologic compromise, and many
or instability should be considered with spinal involvement, even when
for decompression and stabilization. associated with severe local pain or
The majority can be treated with the neurologic compromise, do not
anterior approach alone. However, enjoy a sufficiently long life
my enthusiasm for this procedure expectancy to warrant operative
must not be construed as an advo- intervention of this magnitude.

References
1. Yuh WTC, Zachar CK, Barloon TJ, et al:
Vertebral compression fractures: Dis-
tinction between benign and malignant
causes with MR imaging. Radiology
1989;172:215-218.
2. Harrington KD: Orthopaedic Manage-
ment of Metastatic Bone Disease. St Louis:
CV Mosby, 1988.
3. Constans JP, de Divitiis E, Donzelli R, et
al: Spinal metastases with neurological
manifestations: Review of 600 cases. J
Neurosurg 1983;59:111-118.
4. Tarlov IM, Herz E: Spinal cord com-
pression studies: IV. Outlook with com-
plete paralysis in man. AMA Arch
Neurol Psychiatry 1954;72:43-59.
5. Harrington KD: Anterior decompression
and stabilization of the spine as a treat-
ment for vertebral collapse and spinal
cord compression from metastatic malig-
nancy. Clin Orthop 1988;233:177-197.
6. Gilbert RW, Kim JH, Posner JB:
Epidural spinal cord compression from
metastatic tumor: Diagnosis and treat-
ment. Ann Neurol 1978;3:40-51.
7. Cooper PR, Errico TJ, Martin R, et al: A
systematic approach to spinal recon-
struction after anterior decompression
for neoplastic disease of the thoracic and
lumbar spine. Neurosurgery 1993;32:1-8.
8. Olerud C, Jonsson H Jr, Lofberg AM,
et al: Embolization of spinal metas-
tases reduces peroperative blood loss:
21 patients operated on for renal cell
carcinoma. Acta Orthop Scand 1993;
64:9-12.
9. Harrington KD: Anterior cord decom-
pression and spinal stabilization for
patients with metastatic lesions of the
spine. J Neurosurg 1984;61:107-117.
10. Frankel HL, Hancock DO, Hyslop G, et
al: The value of postural reduction in the
initial management of closed injuries of
the spine with paraplegia and tetraple-
gia: Part I. Paraplegia 1969;7:179-192.
11. Nather A, Bose K: The results of decom-
pression of cord or cauda equina com-
pression from metastatic extradural
tumors. Clin Orthop 1982;169:103-108.
12. Bohlman HH, Sachs BL, Carter JR, et al:
Primary neoplasms of the cervical spine:
Diagnosis and treatment of twenty-
three patients. J Bone Joint Surg Am
1986;68:483-494.
13. Fidler MW: Anterior decompression and
stabilisation of metastatic spinal frac-
tures. J Bone Joint Surg Br 1986;68:83-90.
14. McAfee PC, Bohlman HH, Ducker T, et
al: Failure of stabilization of the spine
with methylmethacrylate: A retrospec-
tive analysis of twenty-four cases. J Bone
Joint Surg Am 1986;68:1145-1157.
15. Siegal T, Tiqva P, Siegal T: Vertebral
body resection for epidural compres-
sion by malignant tumors: Results of
forty-seven consecutive operative
procedures. J Bone Joint Surg Am
1985;67:375-382.
16. Sundaresan N, Scher H, DiGiacinto GV,
et al: Surgical treatment of spinal cord
compression in kidney cancer. J Clin
Oncol 1986;4:1851-1856.
17. Weinstein JN, Kostuik JP: Differential
diagnosis and surgical treatment of
metastatic spine tumors, in Frymoyer
JW (ed): The Adult Spine: Principles and
Practice. New York: Raven Press, 1991,
vol 1, pp 861-888.

86 Journal of the American Academy of Orthopaedic Surgeons

JAAOS Home Page

Table of Contents

Search

Help
 Tendon Disorders of the Foot and Ankle
Donald C. Jones, MD

Abstract

Attritional and traumatic injuries to the tendons around the foot and ankle are not Flexor Hallucis Longus
uncommon. Treatment of overuse-type injuries (tendinitis) remains straightfor- Tendon
ward. However, surgical treatment of peroneal subluxation, Achilles tendon rup-
tures, and posterior tibial tendon insufficiency remains somewhat controversial. The flexor hallucis longus tendon is
Generally speaking, soft-tissue reconstruction of the superior peroneal retinaculum most frequently affected in athletes
is superior to bony procedures for peroneal dislocation. Open repair of a torn Achilles and other individuals who are
tendon is more predictable than closed treatment. Good clinical judgment is needed involved in repetitive push-off
in determining the best treatment for posterior tibial tendon problems. The painful maneuvers (e.g., ballet dancers, in
os peroneum syndrome is a newly described spectrum of posttraumatic conditions whom the sur les pointes position is fre-
that may be the cause of lateral foot pain, which is frequently difficult to identify. quently implicated). These activities
J Am Acad Orthop Surg 1993;1:87-94
transmit tremendous forces across the
tendon and its sheath, resulting in irri-
Every tendon around the foot and Localized swelling, tenderness, tation and tenosynovitis.
ankle can cause symptoms as the and crepitus over this tendon indi- The flexor hallucis longus tendon
result of overuse or injury. Most fre- cate a diagnosis of anterior tibial originates from the lower part of the
quently affected are the anterior tib- tenosynovitis. Treatment includes posterior surface of the fibula lateral
ial, flexor hallucis longus, Achilles, ice, rest, and, on rare occasions, to the medial crest, with a portion of
peroneal, and posterior tibial ten- immobilization. Spontaneous rup- the origin arising from the covering
dons. In this article I will review the tures are rare and usually painless. fascia and the adjacent fascial sep-
relevant anatomy, discuss the clini- An incomplete rupture with mini- tum that it shares with other mus-
cal syndromes involving these ten- mal dorsiflexion weakness does not cles. At the ankle, the tendon lies in
dons, and outline the appropriate require repair. A complete rupture the most posterior lateral compart-
treatment. may be overlooked or confused with ment of the flexor retinaculum,
a foot drop from a lumbosacral whence it travels distally to insert on
radiculopathy or peroneal palsy. the distal phalanx of the great toe. As
Anterior Tibial Tendon Such ruptures usually occur in the the tendon courses behind the
sixth and seventh decades. In cases medial malleolus, it passes through
The anterior tibial muscle originates of complete rupture with foot drop, a fibro-osseous tunnel located on the
from the proximal two thirds of the end-to-end surgical repair should posterior aspect of the talus, bor-
tibia, the lateral tibial condyle, and be performed, although elderly dered anteriorly by the body of the
the interosseous membrane, with patients may choose to use a talus, medially by the medial tuber-
insertion onto the navicular, the first dorsiflexion-arrestive brace. Early cle of the talus, laterally by the lat-
metatarsal base, and the medial diagnosis and treatment are impor- eral tubercle of the talus, and
cuneiform. This musculotendinous tant. The proximal end of the tendon posteriorly by the flexor retinacu-
unit supplies 80% of the dorsiflexion retracts to the superior retinaculum
power of the ankle. The straight but can be reapproximated if rup- Dr. Jones is Clinical Senior Instructor, Oregon
course of the tendon under the supe- ture is diagnosed early.1 Health Sciences University, Portland; and
rior extensor retinaculum results in In cases diagnosed late, half of the Orthopedic Consultant, Athletic Department,
University of Oregon, Eugene.
minimal mechanical demands; thickened proximal end of the ten-
therefore, an overuse syndrome is don may be used as a free graft and Reprint requests: Dr. Jones, Orthopedic and
less common than would be routed under the cruciate ligament, Fracture Clinic of Eugene, 1200 Hilyard Street,
expected in a muscle of this size and or a free extensor tendon graft may Suite 600, Eugene, OR 97401.
power. be utilized.2

Vol 1, No 2, Nov/Dec 1993 87

Tendon Disorders of the Foot and Ankle
lum. The associated tendon sheath
courses behind the medial malleolus
and forms a separate compartment.
Because the tendon runs through
the confined space in the fibro-
osseous tunnel, it is particularly pre-
disposed to mechanical irritation
and inflammation. Prolonged in-
flammation or stenosis of the fibrous
tunnel may cause the tendon to
develop a partial rupture, some-
times accompanied by snapping or
triggering. Patients complain of
pain, tenderness, and a snapping
sensation posteromedially. Occa-
sionally, inability to flex the great toe
is reported as well. A useful clinical
test is to compare the amount of pas-
sive extension of the great toe
metatarsal joint with the foot in the
neutral and plantar-flexed positions.
Contracture or triggering is sug-
gested when the patient is unable to
extend the metatarsophalangeal
joint beyond neutral with the foot
and ankle in the neutral position, but
passive extension is possible with
the ankle plantar-flexed.
Conservative treatment consists
of ice, nonsteroidal anti-inflamma-
tory agents, strapping of the foot,
and longitudinal arch supports
placed in firm-sole shoes. These
measures will frequently alleviate
pain and over time facilitate restora-
tion of function. Surgery is required
when severe stenosis of the fibro-
osseous tunnel is present, accompa-
nied by pain, triggering, and tendon
contracture. The procedure consists
of release of the constrictive flexor
retinaculum and resultant decom-
pression of the flexor hallucis ten-
don. Even after successful surgical
release, the patient should be
informed that there is a possibility of
future complete or partial rupture of
the central fibers of the tendon.
Peroneal Tendons
The peroneal tendons (i.e., those of
the peroneus brevis and longus
88
muscles) pass posterior to the fibula
and are restrained by the superior
peroneal retinaculum (SPR). Most
fibulae have a definite sulcus, but
significant anatomic variations do
exist.3 The sulcus width ranges from
5 to 10 mm; 7% of these grooves are
convex, 11% are flat, and 82% are
concave. Regardless of the sulcus,
peroneal tendon stabilization pri-
marily depends on the SPR. Most
anatomy texts illustrate the SPR as a
single band originating from the
posterior ridge of the fibula and
inserting onto the lateral wall of the
calcaneus. However, insertions onto
the Achilles tendon and the lateral
calcaneus sometimes are present.
Most of the attention concerning
peroneal pathology has been
directed toward subluxation and
dislocation of the peroneal tendons.
Recently, however, incomplete tears
of the peroneus brevis and the
painful os peroneum syndrome 4
have been recognized as important
clinical entities.
Subluxation or Dislocation
Subluxation or dislocation of the
peroneal tendons can be an occult
event and is probably often over-
looked. Either condition can be
difficult to diagnose but should
always be considered a possibility
following any injury that results in
sudden and forceful contraction of
the peroneal muscles in association
with rapid plantar flexion and inver-
sion of the foot and ankle.
Although acute subluxation of
the peroneal tendons is frequently
confused with ankle sprain, the
examiner can differentiate them by
the location of tenderness. While
ankle sprains cause tenderness over
the anterior talofibular ligament,
subluxation of the peroneal tendons
results in tenderness over the retro-
malleolar area. Because of the early
traumatic swelling, palpating the
tendons during dislocation is quite
difficult. Radiographs are seldom
Journal of the American Academy of Orthopaedic Surgeons
helpful. However, if an avulsion
fracture of the lateral ridge of the dis-
tal fibula is present on an x-ray film,
the diagnosis of subluxation or dis-
location of the peroneal tendons can
be made with certainty. Ankle
arthrograms are seldom beneficial,
and radiographic views that show a
shallow peroneal groove are seldom
of diagnostic significance.
Treatment of acute dislocation of
the peroneal tendons remains con-
troversial. Some orthopaedists
believe that conservative treatment is
of benefit, while others recommend
early surgical intervention, particu-
larly for active, competitive athletes.
Conservative treatment consists
of a compression dressing fabricated
from a felt pad cut in the shape of a
keyhole and strapped over the lat-
eral malleolus. Gentle pressure is
placed on the peroneal tendons and
the SPR. This is reinforced with a
plaster splint. Once the acute symp-
toms have resolved, a well-molded
cast is applied for a total of 6 weeks.
After cast removal, an aggressive
ankle rehabilitation program
emphasizes both strengthening and
proprioception education.
If surgery is selected, numerous
procedures have been described for
stabilizing the peroneal tendons.5 The
procedures fall into two general cate-
gories, bony procedures and soft-tis-
sue procedures. Bony procedures are
of historical interest only and include
a variety of methods to increase the
depth of the peroneal groove.6-9 Soft-
tissue procedures include that of
Bonnin,10 which plicates the attenu-
ated retinaculum; that of Eckert and
Davis,11 who reattach the retinaculum
to the malleolar ridge; and that of
Jones,12,13 wherein a sling is fabricated
from a small strip of the adjacent
Achilles tendon. Sarmiento and
Wolf 14 describe rerouting the per-
oneal tendons beneath the calcaneal
fibular ligament.
Two types of pathologic lesions
are most frequently found when

surgery is being performed. The first
lesion is simple attenuation of the
SPR. If this is the only abnormality,
surgical treatment consists of
advancing the SPR to the posterior
edge of the fibula, where it is sutured
through drill holes. Anatomic recon-
struction of the attenuated SPR is the
desired goal. Failure to advance the
stretched-out SPR adequately allows
continued partial subluxation of the
peroneus brevis tendon, while over-
tightening the SPR can lead to
painful stenosis.
On other occasions, a “Bankart-
type” lesion15 is found (Fig. 1). The
SPR is lifted from its posterolateral
fibular attachment, creating a pouch
that allows anterior subluxation of
the peroneal tendons. The repair
consists of reattaching the SPR to the
posterolateral aspect of the fibula,
thereby obliterating the offending
pouch.
Rarely, an anomalous peroneus
brevis muscle will cause subluxa-
tion. The peroneus muscle belly may
extend into the fibular groove, caus-
ing encroachment, or there may be a
bifid tendon.16,17
Fig. 1 Repair of a Bankart-type lesion. Left, Elevation of the SPR from the posterior fibula
creates a pouch. Peroneal tendons subluxate or dislocate into this pouch. Right, Reattach-
ment of the SPR to the fibula obliterates the pouch, stabilizing the peroneal tendons.
Vol 1, No 2, Nov/Dec 1993
Longitudinal Tears of the
Peroneus Brevis
Historically, little mention has
been made about tears of the per-
oneus brevis tendon. However, this
condition has now been recognized
as a distinct clinical entity. Tears of
the peroneus brevis occur when the
anterior portion of the tendon slips
forward, out of the groove, and over
the sharp posterior ridge of the
fibula. A complete or incomplete
longitudinal tear can measure from
2 to 5 cm (Fig. 2). In all cases, the cen-
tral portion of the longitudinal split
is centered over the distal tip of the
fibula. The tear usually involves the
middle or anterior portion of the
tendon.
Patients generally present with
retromalleolar pain and tenderness.
They may have a history of multiple
ankle sprains or chronic ankle insta-
bility.
In patients with documented
tears of the peroneus brevis, con-
servative treatment is generally
unsuccessful. Surgical repair is
accomplished through a curved 7-
cm incision along the posterior third
Donald C. Jones, MD
of the fibula. The competence of the
SPR is assessed. The SPR is then
opened in such a way that it can be
tightened if attenuated.
If the split is through the anterior
third of the tendon and the smaller
portion of the tear is frayed in any
way, I excise the anterior third. If the
tear is in the middle third and both
fragments are without degenerative
change, I repair the tendon with
buried nonabsorbable suture. If, how-
ever, the entire width of the peroneus
brevis tendon is involved and there is
significant fraying, the degenerated
segment of the tendon is excised in
toto, and tenodesis of the proximal
and distal stumps to the peroneus
longus tendon is performed (Fig. 3)
Following surgery on the per-
oneus brevis tendon, the SPR is
advanced and imbricated onto a
fresh bony bed. If there is associated
ankle instability, this should be
repaired as well.
Painful Os Peroneum Syndrome
The painful os peroneum syn-
drome is a spectrum of posttrau-
matic conditions, including one or
more of the following4: (1) an acute
os peroneum fracture or diastasis of
a multipartite os peroneum; (2) a
chronic os peroneum fracture or
diastasis of a multipartite os per-
oneum fracture associated with
stenosing peroneus longus tenosyn-
ovitis; (3) attrition or partial rupture
of the peroneus longus tendon prox-
imal or distal to the os peroneum; (4)
frank rupture of the peroneus
longus tendon; or (5) the presence of
a gigantic peroneal tubercle on the
lateral wall of the calcaneus that
traps the peroneus longus tendon
and its os peroneus during peroneus
longus tendon excursion.
Patients with the painful os per-
oneum syndrome have a history of
either direct trauma to the lateral
side of the foot or a supination-
inversion ankle injury. Symptoms
usually include tenderness along the
89
Tendon Disorders of the Foot and Ankle
peroneus longus tendon distal to the
fibula. Pain is usually exacerbated
by resisted plantar flexion of the first
ray and the heel-rise phase of gait.
Fig. 3 Irreparable tear of the peroneus brevis necessitates excision of the tear and tenode-
sis to the adjacent peroneus longus.
90
Fig. 2 Complete longitudi-
nal tear of the peroneus bre-
vis. The peroneus longus
remains reduced, while the
torn peroneus brevis dislo-
cates over the tip of the
fibula.
Patients also have weakness or pain
with forced foot eversion.
Diagnosis may be based on ra-
diographic or magnetic resonance
Journal of the American Academy of Orthopaedic Surgeons
(MR) imaging data or the findings
on exploration motivated by a high
degree of suspicion.
Conservative treatment consists
of cast immobilization, with or with-
out corticosteroid injections. I
always cast the extremity at least
once for 4 to 6 weeks. Corticosteroid
administration is optional; if palpa-
ble synovitis is present, however, I
routinely perform an injection. Sur-
gical treatment consists of (1) exci-
sion of the os peroneum and the
giant peroneal tubercle with pri-
mary repair of the peroneus longus
tendon; (2) excision of the os per-
oneum and degenerated peroneus
longus tendon with tenodesis of the
peroneus brevis to the peroneus
longus tendon; or (3) excision of the
os peroneum with primary repair of
the peroneus longus tendon.
Achilles Tendon
The gastrocnemius originates from
the lateral and medial femoral
condyles, while the soleus origi-
nates from the posterior surface of
the tibia and the fibula. The soleus
and gastrocnemius contribute sepa-
rately to the formation of the
Achilles tendon, with the gastrocne-
mius segment measuring 11 to 26
cm and the soleus portion measur-
ing 3 to 11 cm. The blood supply to
the Achilles tendon comes from
both proximal and distal sources
(Fig. 4). The least vascular area is 2
to 6 cm above the tendon insertion
into the calcaneus, which is the
usual location of chronic inflamma-
tion and rupture.
The primary etiologic factor
resulting in damage to the Achilles
tendon is training errors, such as a
sudden increase in training mileage,
a single severe competitive session
(a 10-km race or a marathon), a sud-
den increase in training intensity,
repetitive heel running, recom-
mencement of training after an
extended period of inactivity, and

running on uneven or slippery ter-
rain. Hindfoot and leg malalign-
ments may also contribute.
Peritendinitis
Peritendinitis is inflammation
within the peritenon without associ-
ated Achilles tendinosis (Fig. 5). The
symptoms consist primarily of pain,
which is aggravated by activity and
relieved by rest. Tenderness is pre-
sent several centimeters proximal to
the insertion of the Achilles tendon
into the calcaneus.
Fig. 4 Blood supply to the Achilles tendon.
Note that longitudinal vessels supply the
tendon proximally and distally, while trans-
verse vessels vascularize the middle por-
tion.
Vol 1, No 2, Nov/Dec 1993
Initially, treatment includes ice,
massage, contrast baths, and non-
steroidal anti-inflammatory medi-
cations. If hindfoot alignment
problems are present, an orthosis is
prescribed. In more advanced or
chronic cases, the peritenon of the
Achilles tendon becomes fibrotic
and stenosed. While the use of
steroidal injections for tendon
injuries is generally considered quite
hazardous, we have found that sub-
peritenon infiltration of lidocaine is
effective in relieving symptoms.
Mechanical lysis of adhesions can be
achieved by rapid injection of 15 ml
of local anesthetic into the sub-
peritenon space.
If conservative means fail, open
lysis of adhesions is performed
through a medial incision exposing
the involved area of tendon. One
should be very careful to protect the
anterior fatty tissue, as this is a source
of the blood supply of the tendon.
However, the medial, lateral, and
posterior peritenon can be excised.
Tendinosis
The pathology of Achilles tendi-
nosis is interstitial microscopic fail-
ure or obvious central tissue necrosis
with subsequent mucoid degenera-
tion. This is usually the result of
accumulated repetitive microtrauma
Fig. 5 Thickened inflamed
Achilles peritenon.
Donald C. Jones, MD
and evolves pathologically in a rela-
tively predictable manner. Initially,
the peritenon sheath becomes
inflamed. If the overuse continues,
the tendon itself may become
inflamed or hypovascular secondary
to restriction of blood flow through
the scarred peritenon. Degenerative
changes in the tendon then follow.
With tendinosis, the tendon has a
noninflammatory histologic appear-
ance with collagen fiber disorienta-
tion, hypocellularity, scattered
vascular ingrowth, and occasional
areas of necrosis or calcification.
Despite these changes, the condition
can be asymptomatic. Patients will
frequently note a palpable but pain-
less mass in the Achilles tendon
approximately 4 to 6 cm proximal to
the insertion of the tendon. Those
patients who become symptomatic
usually have peritenous inflamma-
tion along with the intratendinous
mucoid degeneration.
Nonsurgical treatment includes
(1) a 1- to 2-week period in a non-
weight-bearing cast if the symptoms
are severe, (2) anti-inflammatory
agents and ice, (3) heel-cord stretch-
ing within limits of comfort, and (4)
careful assessment of the foot and
leg alignment, with orthotic correc-
tion if necessary. Conservative treat-
ment is provided for 6 months. If
91
Tendon Disorders of the Foot and Ankle
symptoms persist, surgery is recom-
mended.
Surgery consists of first debriding
the overlying inflamed peritenon. The
tendon is then palpated in the area of
fusiform thickening or nodular
enlargement. If there are only minor
changes on palpation but significant
preoperative symptoms, several lon-
gitudinal incisions are made into the
tendon. The purposes of these longi-
tudinal incisions are to visualize areas
of central tendon necrosis, which
should be excised, and to stimulate a
healing reaction. If a significant area
of degeneration, characterized by a
glossy homogeneous appearance
(Fig. 6), is found, it is excised. The area
of elliptical excision is then closed. If
the defect is large and the excision
extensive, the Achilles tendon is rein-
forced using the plantaris tendon, the
flexor digitorum communis, or a turn-
down flap.
The period of postoperative
immobilization depends on the size
of the defect. If a small defect is
excised, the patient is immobilized
for 2 weeks. If a larger defect is
excised, 4 to 6 weeks of immobiliza-
tion may be necessary.
Partial Rupture
Partial ruptures of the Achilles
tendon were thought to be rare until
Ljungqvist described 24 cases in
1968.18 Unlike total ruptures, which
tend to occur in middle-aged decon-
ditioned persons, partial ruptures
occur in well-trained athletes. Partial
tears usually involve the lateral
aspect of the Achilles and may be
longitudinal, transverse, or both.
Diagnosis of partial tears, although
frequently difficult, has been
enhanced considerably through the
use of MR imaging (Fig. 7).
If a large partial tear is identified,
immediate repair should be under-
taken. However, if a small defect is
present, conservative treatment con-
sisting of heel lifts, ice, and rest is
92
ing the calf does not cause passive
ankle plantar flexion), and the
patient is usually unable to perform
a single heel rise. However, the
patient is frequently able to plantar-
flex the foot when it is not bearing
weight because of the plantar-
flexion action of the posterior tibial,
toe flexor, and peroneal tendons.
The main objective of treatment is
to provide the patient with a tendon
as close to normal in length and
strength as possible. Nonoperative
measures can achieve this objective
provided the length of treatment is
sufficient to allow the tendon to
reestablish adequate intrinsic
Fig. 6 Nodular Achilles mucoid degenera-
tion. Note the smooth, glossy appearance of strength and to avoid elongation
the involved tendon. with future activities.
The controversy of closed versus
open treatment of Achilles tendon
usually adequate. If a small defect ruptures has been ongoing for
fails to respond to conservative years. When deciding between
treatment, surgical excision of the operative and nonoperative treat-
involved area or repair of the tear is ment, the physician and the patient
undertaken. should weigh carefully the risks
Complete Rupture
Complete rupture of the Achilles
tendon occurs most frequently in the
middle-aged, competitive male
involved in intermittent athletic
activities. There may be a history of
prerupture intermittent heel pain
suggestive of long-standing mild
chronic Achilles tendinosis. More
frequently, however, rupture occurs
without preexisting complaints.
Two theories are suggested to
explain the cause of acute Achilles
tendon rupture: (1) chronic tendon
degeneration and (2) acute mechani-
cal overload.19 In fact, both of these
factors are usually involved.
The symptoms of rupture are fairly
classic. The middle-aged athlete will
often hear or feel a pop while experi-
encing minimal discomfort. Immedi-
ate weakness in push-off is noted,
followed by pain and swelling. Fig. 7 Magnetic resonance image of a par-
On physical examination, the tial longitudinal tear of the Achilles tendon.
Thompson test is positive (squeez-
Journal of the American Academy of Orthopaedic Surgeons

and benefits. The risks of closed
treatment utilizing prolonged non-
weight-bearing cast immobilization
are decreased strength, rerupture,
and stiffness.20 The strength of the
tendon is approximately 30% of
normal in the conservatively
treated patient group, compared
with near normal in the surgical
group.21 The average rerupture rate
in the conservatively treated patient
is 18%, while the patient who
undergoes open surgical treatment
has a 2% rerupture rate.22,23 The risks
of surgery include pulmonary
embolism, sural nerve injury,
suture granulomas, skin problems,
stiffness, and infection.
A compromise between open and
closed treatment is the technique of
Ma and Griffith.24 They plantar-flex
the ankle, bring the Achilles tendon
ends together, and percutaneously
repair the tendon. They have reported
their results as excellent. The advan-
tage of open repair is that it provides
direct visualization of the disrupted
tendon ends and allows restoration of
the tendon to its normal length.
Postoperatively, two methods of
treatment are available. One is the
standard rigid immobilization. This
is accomplished by using a short-
leg cast for 6 to 8 weeks, followed by
wearing a shoe with an elevated
heel for 1 month. The second
method is functional postoperative
treatment, which is now gaining
popularity. Mahan and Carter 21
have described the use of a postop-
erative functional orthosis rather
than cast immobilization. Saltzman
and Thermann22 have described the
use of a modified boxer’s boot,
which provides protection as well
as functional treatment.
References
1. Stuart MJ: Traumatic disruption of the
anterior tibial tendon while cross-coun-
try skiing: A case report. Clin Orthop
1992;281:193-194.
Vol 1, No 2, Nov/Dec 1993
Posterior Tibial Tendon
Inflammation of the posterior tibial
tendon is more common than anterior
tibial tenosynovitis. As the tendon
curves behind the medial malleolus, it
functions much like a rope being
pulled through a pulley. Attrition
occurs at the bone-tendon points of
contact. It should also be noted that
during the pronation phase of run-
ning gait, the mechanical demands
placed on this structure are quite
high. As a result, microtrauma occurs,
and the tendon may become inflamed
and undergo degeneration and rup-
ture. Posterior tibial tenosynovitis is
also commonly seen in systemic
inflammatory diseases such as
rheumatoid arthritis.
The primary complaints of patients
with posterior tibial tenosynovitis,
partial rupture, or complete rupture
are pain, weakness, and eventual
deformity. The pain is aggravated by
activity and is partially relieved by
rest and anti-inflammatory agents.
When the tendon ruptures com-
pletely, the pain may be referred from
its usual medial location to the lateral
aspect of the ankle in the sinus tarsi
region. This type of pain is associated
with the development of a valgus
deformity. The anterior process of the
talar articular surface of the posterior
facet impinges on the superior aspect
of the calcaneus. The eventual severe
clinical deformity secondary to poste-
rior tibial tendon rupture is a combi-
nation of hindfoot valgus with
forefoot abduction and pronation.
Treatment is determined by the
degree of involvement of the poste-
rior tibial tendon. If the patient has
swelling and inflammation of only a
few weeks’ duration, an aggressive
2. Lapidus PW: Indirect subcutaneous
rupture of the anterior tibial tendon:
Report of two cases. Bull Hosp Jt Dis
1941;2:119-127.
Donald C. Jones, MD
conservative treatment program
should be instituted. This program
consists of 4 to 6 weeks of cast immo-
bilization holding the foot slightly
inverted and plantar-flexed. If immo-
bilization is unsuccessful, one can
consider a corticosteroid injection into
the tendon sheath without injecting
steroid into the tendon itself.
If the patient is unresponsive to
conservative treatment after a few
months, surgical treatment is indi-
cated to prevent further damage to
the tendon. At the time of surgery a
thorough tenosynovectomy is under-
taken. The tendon is also thoroughly
inspected. If the tendon is intact but
minimal longitudinal rents in the ten-
don are found, the rents are either
sutured or debrided.
If the tendon is detached from the
navicular, severely attenuated, or
ruptured, the treatment plan should
be based on the degree of deformity.
If the patient has minimal or no defor-
mity, attempts should be made either
to reattach the tendon to the navicular
through a bony tunnel or to augment
the ruptured tendon. If augmentation
is chosen, the tendon of choice is the
flexor digitorum communis. If the
deformity is severe and well estab-
lished, the patient usually has had a
moderate degree of pain for a period
of years. A subtalar arthrodesis is
used in this setting. An isolated talona-
vicular arthrodesis or a talonavicular
arthrodesis in combination with a cal-
caneal cuboid arthrodesis may also be
performed to stabilize the hindfoot.
An unusual problem is recurrent
subluxation. The diagnosis is based
on the patient’s symptoms, physical
examination findings, and MR imag-
ing evaluation. Surgical repair is gen-
erally indicated.25
3. Edwards ME: The relations of the per-
oneal tendons to fibula, calcaneus,
and cuboideum. Am J Anat 1928;42:
213-253.
93
Tendon Disorders of the Foot and Ankle
4. Sobel M, Mizel MS: Peroneal tendon
injury, in Pfeffer GB, Frey CC (eds): Cur-
rent Practice in Foot and Ankle Surgery.
New York: McGraw-Hill, 1993, vol 1, pp
30-56.
5. Brage ME, Hansen ST Jr: Traumatic sub-
luxation/dislocation of the peroneal
tendons. Foot Ankle 1992;13:423-431.
6. DuVries HL (ed): Surgery of the Foot, ed
2. St Louis: CV Mosby, 1965.
7. Kelly RE: An operation for the chronic
dislocation of the peroneal tendons. Br J
Surg 1920;7:502-504.
8. Watson-Jones R: Fractures and Joint
Injuries, ed 5. Baltimore: Williams &
Wilkins, 1976, pp 951-961, 987-993.
9. Zoellner G, Clancy W Jr: Recurrent dis-
location of the peroneal tendon. J Bone
Joint Surg Am 1979;61:292-294.
10. Bo nnin JG: In j ur i e s t o t h e Ank l e .
Darien, Conn: Hafner Publishing,
1970, p 302.
11. Eckert WR, Davis EA Jr: Acute rupture
of the peroneal retinaculum. J Bone Joint
Surg Am 1976;58:670-672.
12. Jones E: Operative treatment of chronic
dislocation of the peroneal tendons. J
Bone Joint Surg 1932;14:574-576.
94
13. Thomas JL, Sheridan L, Graviet S: A
modification of the Ellis Jones proce-
dure for chronic peroneal subluxation. J
Foot Surg 1992;31:454-458.
14. Sarmiento A, Wolf M: Subluxation of
peroneal tendons: Case treated by
rerouting tendons under calcaneofibular
ligament. J Bone Joint Surg Am 1975;57:
115-116.
15. Arrowsmith SR, Fleming LL, Allman
FR: Traumatic dislocations of the per-
oneal tendons. Am J Sports Med
1983;11:142-146.
16. Sobel M, Bohne WH, O’Brien SJ: Per-
oneal tendon subluxation in a case of
anomalous peroneus brevis muscle.
Acta Orthop Scand 1992;63:682-684.
17. Sobel M, Warren RF, Brourman S: Lat-
eral ankle instability associated with
dislocation of the peroneal tendons
treated by the Chrisman-Snook proce-
dure: A case report and literature
review. Am J Sports Med 1990;18:539-543.
18. Ljungqvist R, Eriksson E: Partial tears
of the patellar tendon and the Achilles
tendon, in Mack RP (ed): American
Academy of Orthopaedic Surgeons Sympo-
sium on the Foot and Leg in Running
Journal of the American Academy of Orthopaedic Surgeons
Sports. St Louis: CV Mosby, 1982, pp
92-98.
19. Di Stefano VJ: Pathogenesis and diagno-
sis of the ruptured Achilles tendon.
Orthop Rev 1975;4:17-18.
20. Lea RB, Smith L: Rupture of the achilles
tendon: Nonsurgical treatment. Clin
Orthop 1968;60:115-118.
21. Mahan KT, Carter SR: Multiple ruptures
of the tendo Achillis. J Foot Surg
1992;31:548-559.
22. Saltzman CL, Thermann H: Achilles
tendon problems, in Pfeffer GB, Frey CC
(eds): Current Practice in Foot and Ankle
Surgery. New York: McGraw-Hill, 1993,
vol 1, pp 194-218.
23. Wills CA, Washburn S, Caiozzo V, et al:
Achilles tendon rupture: A review of the
literature comparing surgical versus
nonsurgical treament. Clin Orthop
1986;207:156-163.
24. Ma G, Griffith T: Percutaneous repair of
acute closed ruptured achilles tendon: A
new technique. Clin Orthop 1977;128:
247-255.
25. Ouzounian TJ, Myerson MS: Disloca-
tion of the posterior tibial tendon. Foot
Ankle 1992;13:215-219.
 Locked Femoral Nailing
Robert A. Winquist, MD

Abstract

Locked intramedullary nailing has become the standard of care for most femoral frac- anticipated preoperatively, and that
tures. Originally designed to prevent rotation and shortening in comminuted frac- either missed fractures or comminu-
tures of the midshaft, its application has been extended proximally and distally to tion caused by surgery led to short-
nearly all femoral fractures from the lesser trochanter to the supracondylar area. ening and rotation in an additional
Achieving a closed reduction and selecting the proper starting point in the piriformis 10% of patients treated with
region are crucial to a successful result. Following the proper surgical technique for unlocked femoral nailing. To pre-
the specific nail used is more important than nail material or design. Large-diameter vent these complications, their rec-
reamed nails provide greater strength than unreamed nails. Static locking has been ommendation, with which I concur,
shown to yield nearly the same high union rates as dynamic locking and is now the was that static locking (locking at
accepted standard. Distal targeting of the interlocking screw remains the most both ends of the nail) be used in all
difficult aspect of the surgical technique; most surgeons prefer freehand targeting femoral shaft fractures.
with a sharp trocar. Second-generation (reconstruction) nails, with screws directed The patient’s age is important in
toward the femoral head, has extended the indications for locked nailing proximally determining the appropriateness of
to subtrochanteric fractures and combined femoral neck-shaft fractures. locked nailing. My preference is to
J Am Acad Orthop Surg 1993;1:95-105 use locked intramedullary nails in
most female patients aged 12 years
and older and in most male patients
Traditional treatment of femoral shaft femoral shaft fractures but demands aged 13 years and older. In patients
fractures has been traction or cast experience on the part of the surgical below these ages, treatment is indi-
bracing. Unfortunately, the use of team. vidualized, with greater use of inter-
these techniques typically led to a nal fixation in younger patients with
high rate of malunion and knee stiff- multiple trauma and additional ipsi-
ness.1,2 The advent of plate fixation Indications lateral injuries. One should consider
improved both alignment and knee flexible intramedullary nails, such as
motion but resulted in a higher rate of Interlocking nails were initially indi- Ender nails or Rush rods, in younger
infection, nonunion, and implant fail- cated for femoral fractures with insta- patients. In the growing child, the nail
ure. Closed Küntscher nailing 3 bility of length, rotation, and must stop short of the distal femoral
allowed both excellent function and angulation. Originally, ideal indica- epiphysis. Apophyseal arrest of the
an extremely low nonunion and infec- tions were femoral shaft fractures with trochanter has not been a problem in
tion rate. Winquist type III comminution this population, but avascular necro-
Only two problems remained: (greater than 50 percent of the cortex sis of the femoral head has been noted
shortening and rotation.4 The solution comminuted) and Winquist type IV in teenagers. Therefore, in younger
to these problems appeared to be the fractures (segmental comminution).7 patients a starting point for nail inser-
development of an intramedullary As experience was gained with these tion a little farther anterior and lateral
nail with holes for screw fixation. locked nails, indications were than the standard piriformis fossa
Modny, Halloran, and Huckstep all extended to segmental fractures, spi- starting point should be considered.
developed this concept,5 but the first ral fractures, fractures below the lesser
published report detailing the use of trochanter, and infraisthmal fractures, Dr. Winquist is Clinical Professor, Department
an interlocking (locked) femoral nail including some minimally displaced of Orthopaedics, University of Washington,
Seattle.
came from Gerhard Küntscher.6 Use fractures extending into the knee.8-11
of the locked femoral nail inserted In a large series, Brumback et al12 Reprint requests: Dr. Winquist, 1229 Madison
with a closed technique has become clearly demonstrated that the degree Street, Suite 1600, Seattle, WA 98104.
the standard of care for treatment of of comminution could not always be

Vol 1, No 2, Nov/Dec 1993 95

Locked Femoral Nailing
Timing of Surgery
The timing of surgery is an impor-
tant consideration. Closed reduc-
tion and intramedullary nailing
with a locked nail is a personnel-
and equipment-dependent opera-
tion. For a successful outcome, it is
mandatory that skilled, experienced
personnel be available to perform
the operation and that the proper
equipment be on hand. Therefore,
timing may be dictated by the avail-
ability of staff and implants.
The ideal timing for intra-
medullary nailing is immediately
after patient resuscitation. Immedi-
ate nailing appears to be even more
important in the patient with multi-
ple injuries. Bone et al13 have clearly
demonstrated a decreased incidence
of adult respiratory distress syn-
drome with primary fixation of
femoral shaft fractures compared
with delayed fixation.
Preoperative Planning
Operating room planning must take
place long before the first case of
locked intramedullary nailing is
undertaken. The surgeon must main-
tain up-to-date knowledge of the best
available image intensifiers and must
participate in the selection of this
expensive device. The proper frac-
ture table is also crucial. The best frac-
ture table has a radiolucent perineal
post, allows adequate visualization
of the fracture with the patient in both
the lateral and the supine position,
and is small and easy for the operat-
ing staff to manage. The table should
also be chosen for its usefulness for all
intramedullary nailing techniques.
Interlocking nails and screws in a
range of appropriate sizes must be
available.
In addition to the operating sur-
geon, another surgeon should be
available to reduce the fracture.
Closed reduction of the fracture is
96
the most important and difficult part
of the procedure and requires the
most experience. A technician
trained in the use of the C-arm image
intensifier is the other critical mem-
ber of the surgical team.
Traction
When nailing is immediate, a trac-
tion pin is unnecessary, since the
foot can be placed in temporary
traction and the femur can be
nailed. In patients in whom there is
concern about applying excessive
traction, a femoral pin can be
inserted for use during the surgical
procedure. The knee is flexed to
protect the sciatic nerve. In teaching
institutions with changing and
inexperienced staff, it may be safer
in most cases to use the femoral pin
with the knee flexed to avoid sciatic
and peroneal injuries. If surgery is
delayed, a tibial traction pin is
placed, and heavy traction will be
necessary to maintain the femur at
length, which can be monitored on
the lateral radiograph. The use of
preoperative traction makes the
surgical procedure much easier.
To prevent nerve palsy, it is
Fig. 1 Lateral positioning for intramedullary nailing.
Journal of the American Academy of Orthopaedic Surgeons
extremely important that traction be
used only during those portions of
the case when it is necessary. Traction
is used initially during closed reduc-
tion while the unscrubbed surgeon is
determining whether the reduction
can be achieved. It is released before
the incision is made and is reapplied
when the bulb-tipped guide has been
passed. It is then relaxed and applied
a final time during driving of the nail.
Many surgeons apply traction and
maintain it during the entire proce-
dure. Such prolonged traction is not
necessary and can be associated with
an increased risk of sciatic and
pudendal nerve palsies.
Patient Positioning
Lateral Positioning
Placing the patient in the lateral
position on the fracture table allows
much easier access to the greater
trochanter than use of the supine
position does and facilitates
intramedullary nailing (Fig. 1). The
fracture table should be equipped
with a radiolucent perineal post to
allow visualization of the femoral
neck and shaft. Also, there must be

adequate room for the image
intensifier to be maneuvered proxi-
mally without bumping the upright
stand supporting the table. A
padded support on the anterior
portion of the post is needed to
cushion the iliac crest and prevent
pressure on the anterolateral
femoral cutaneous nerve.
The patella should be internally
rotated 20 to 30 degrees toward the
floor to prevent an external rotation
deformity at the fracture site. Rota-
tion is best checked by rotating the
leg gently and observing the skin
lines in the supracondylar region.
Evaluating the fracture on the image
intensifier is a poor method of judg-
ing rotation of the fracture. The
potential exists for valgus sag at the
fracture site, particularly in infraisth-
mal fractures. To prevent a valgus
reduction, the unscrubbed surgeon
must support the fracture both dur-
ing insertion of the bulb-tipped guide
and during insertion of the
intramedullary nail.
Supine Positioning
Another popular method is
supine positioning of the patient
Fig. 2 Supine positioning for intramedullary nailing. Note adduction of left (operative) leg.
Vol 1, No 2, Nov/Dec 1993
(Fig. 2). Surgeons and other operat-
ing room staff are generally more
familiar with this technique than
with lateral positioning because it is
commonly used for fixation of
intertrochanteric and femoral neck
fractures. Unfortunately, access to
the trochanter is much more difficult.
It requires adduction of the leg,
which creates a varus deformity in
high subtrochanteric fractures. This
adduction also places increased pres-
sure on the pudendal nerve, leading
to an incidence of temporary puden-
dal nerve palsy that can rise to as
high as 10%.14,15 A common error with
supine positioning is rotation of the
knee too far inward, creating internal
rotation deformities. I recommend
that the surgical team select a frac-
ture table and C-arm image
intensifier that are appropriate for
lateral positioning, and that once
they have gained sufficient experi-
ence with this positioning, they use it
for most patients undergoing locked
femoral nailing. An exception is the
patient with multiple injuries, partic-
ularly those involving the contralat-
eral lung, for whom the supine
position is more appropriate.
Robert A. Winquist, MD
Use of a Distractor
Another method of reduction is
with a distractor instead of a fracture
table.16 It is difficult to place the prox-
imal distraction pin anterior to the
medullary canal. Once the device
has been placed, the fracture can be
distracted. The distractor may be
beneficial in patients with multiple
injuries, but the proponents of
locked femoral nailing prefer use of
the fracture table.
Determining Length
Regardless of the patient position-
ing used, judging the adequate
length of the comminuted femur is
extremely difficult.17 Errors can be
made that either leave the femur too
short or overlengthen it by applying
too much traction. In comminuted
fractures it is best to try to select a
fragment that locks into place prox-
imally and distally for use in judg-
ing adequate length. Measuring the
opposite femur to obtain a compar-
ative length is possible, but at best
this method is accurate only to
within 1 cm.
97
Locked Femoral Nailing
Closed Reduction
Closed reduction should be per-
formed as soon as the patient is posi-
tioned on the fracture table, before
preparation and draping. The
unscrubbed surgeon, who should be
familiar with the maneuvers neces-
sary to reduce the fracture, may gain
insight into the vectors needed for
reduction by studying anteroposte-
rior and lateral radiographs. Use of
leaded gloves as well as a crutch may
be helpful. Fracture tables with built-
in clamps for reduction are available,
but unfortunately these bulky
clamps impede movement of the
image intensifier and create prob-
lems during distal targeting of the
interlocking screw. Once surgery has
begun, a reduction rod may be
placed in the proximal femur to
allow manipulation of the proximal
fragment. Some surgeons drape the
entire thigh into the sterile field,
allowing reduction of the fracture by
a member of the scrubbed team. This
can be facilitated by use of a sterile
“reduction wrench” (Fig. 3).
Incision
The incision should start at least 2 cm
proximal to the greater trochanter and
should be about 3 cm long. In obese
patients it must extend even farther
proximally. The dissection is carried
down through the fasciae, and the
trochanter is palpated. Visualization
of the trochanter is not necessary; the
image intensifier is used to locate the
starting point for nail insertion.
Starting Point for Nail
Insertion
Accomplishing the closed reduction
and locating the entry portal in the
femur for nail insertion are the two
most important steps in the surgical
98
Fig. 3 Use of a sterile
“reduction wrench”
(inset) assembled from
the bars used for over-
head traction.
procedure. A piriformis starting
point appears to be the best, as the
piriformis fossa tends to align with
the longitudinal axis of the medullary
canal.18 Küntscher originally advised
against this medial starting point
because of the risk of avascular necro-
sis, intracapsular infection, and stress
fracture of the femoral neck, but these
complications have all been rare. The
use of the piriformis starting point
becomes even more important with
nails that are more rigid than the slot-
ted interlocking nails, because their
greater rigidity increases the risk of
comminution during nail insertion.
Nails with an increased curvature
require a slightly more posterior
starting point. For second-generation
(reconstruction) interlocking nails,
which have screws that extend prox-
imally into the femoral head, a start-
ing point 5 mm anterior to the
piriformis fossa allows easier place-
ment of the screws into the femoral
neck and head.
An awl is placed on the proposed
starting point, and its placement is
checked on both anteroposterior and
lateral views with the image
intensifier. Before the cortex is pene-
Journal of the American Academy of Orthopaedic Surgeons
trated, the awl must be well visual-
ized in both views and, most impor-
tant, must be seen to be aligned with
the medullary canal. An alternative
method is to place a Steinmann pin in
the appropriate starting position and
to check the two planes with the
image intensifier (Fig. 4) The pin is
then drilled into the proximal femur,
and a reamer is used over the pin to
enlarge the starting point.
Reaming
Reaming of the medullary canal pro-
vides uniformity in the canal diame-
ter and allows insertion of a
larger-diameter intramedullary nail.
Increasing the nail diameter dramati-
cally augments nail strength and also
permits the use of interlocking screws
with a larger core diameter, which
further increases strength. The use of
a larger-diameter intramedullary nail
also enhances alignment in midshaft
fractures with minimal comminution,
but is not as effective in the large canal
of infraisthmal and subtrochanteric
fractures.
Although reaming damages the
endosteal blood supply, its restora-

tion within 6 to 8 weeks has been
well documented. Clinically, ream-
ing of the femur has not been found
to cause a higher infection rate or a
lower union rate. Fat embolism may
result from reaming, but the risk of
this sequela is partially dependent
on reamer design and the degree of
reaming. Clinically, the risk of fat
embolism is slight except in the mul-
tiply injured patient with a chest
injury. In patients with such injuries,
the use of an unreamed nail may be
indicated, but unreamed nails have
smaller diameters and unfortunately
carry a higher risk of later fatigue
failure than do reamed nails.
A bulb-tipped guide should
always be used when reaming to
allow extraction of broken reamers.
The reaming should progress in 1-
mm increments until cortical contact
is made, after which reaming in 0.5-
mm increments is advisable. Inter-
locking nails are stiffer than flexible
Küntscher nails and frequently
require overreaming in the range of 1
to 2 mm.19 It is vitally important that
the surgeon study the specific tech-
nique advocated by the manufacturer
for each nail with regard to over-
reaming.
Jamming of Reamers
and Nails
Jammed reamers can usually be freed
from the femur by applying power
Vol 1, No 2, Nov/Dec 1993
Fig. 4 The piriformis entry site
should align with the medullary
canal.
and then quickly twisting the wrist to
free the reamer. It may be necessary
to use a vise grip to back the reamer
out and free it from the femur. Flexi-
ble reamers should never be run in
reverse, as the spiral windings can
uncoil to become hopelessly tangled
within the medullary canal. Inability
to extract the reamer generally indi-
cates that an infraisthmal fracture has
caused a piece of bone to obstruct the
intramedullary canal and block the
exit of the reamer. A guide rod must
then be moved down the canal to
push the fragment out of the canal
through the fracture site before the
reamer can be removed.
If the nail fits too tightly during
insertion, further reaming or a reduc-
tion in nail size is necessary. The nail
should advance with each blow of the
mallet; if it does not do so, it should
be immediately removed before it
becomes incarcerated. A large mallet
is very helpful in removing incarcer-
ated intramedullary nails. If this is
not successful, it may be necessary to
saw a slot into the lateral cortex of the
femur, over the portion of the isth-
mus where the nail tip is incarcer-
ated, to allow bone expansion.
Nail Selection
Nail Design
With the growth in popularity of
interlocking nails, the number of
Robert A. Winquist, MD
available designs has burgeoned. In
the face of union rates of 98% to
100% and infection rates of 1% with
the use of these nails, it has been
difficult to substantiate the clinical
advantage of one design over the
other.20 Stainless steel and titanium
nails appear to give equal results.
Nails with a closed section (circular
nails) and those with an open section
(slotted nails) also provide similar
results. Closed-section nails offer
increased torsional rigidity,21 but this
property has no clinical significance
and may lead to increased com-
minution at the fracture site.22 Wall
thickness has been studied in detail,
and attempts have been made to
increase the strength and augment
the fatigue resistance of the nail.
However, there is little evidence that
these differences translate into a
higher clinical success rate.
The only important factor related
to nail design is that more rigid nails
require further overreaming and
perfectly placed trochanteric start-
ing points to prevent comminution.
The radius of curvature of the
femoral nail varies among manufac-
turers. This difference is of no
significance except that nails with an
increased curvature require a
trochanteric entry point that is a lit-
tle farther posterior than the stan-
dard piriformis starting point in
order to avoid shaft comminution.
There are subtle differences
among nails in the proximal and dis-
tal placement of holes within the
nail. A more proximal placement of
the interlocking screw holes allows
expansion of the indications for nail-
ing to higher fractures, but it also
causes the screw to be placed in the
femoral neck, with some risk of
femoral neck fatigue. A quite distal
placement allows expansion of the
indications to more distal fractures,
but placing the screws through the
wide metaphysis to reach the hole in
the nail creates targeting difficulties.
Unreamed femoral nails have
99
Locked Femoral Nailing
relatively few indications. The
increased strength and fatigue
resistance of the larger-diameter
reamed femoral nails have played
an extremely important role in the
attainment of high union rates in
nailed fractures. In the femur, pre-
sent indications for the smaller-
diameter unreamed nails, with their
increased failure rate, are confined
to fractures in multiply injured
patients with severe chest injuries
and Gustilo grade IIIB and IIIC
open fractures.23 In these two set-
tings, the risks of fat embolism and
damage to the blood supply out-
weigh the risk of nail failure.
An important aspect of nail
design involves the area in which the
screw holes penetrate the nail. Nail
failure usually occurs through the
screw holes,24 yet all bending tests
comparing various products are
conducted on the midshafts of the
devices. Increased wall thickness of
the nail in the vicinity of the hole
provides increased strength.25 Cold
working of the interlocking holes
has also helped increase strength
and is especially important in nails
with small diameters.
In summary, there is little evi-
dence that either material or design
makes a significant difference in the
performance of interlocking nails.
More important than either of these
features is the need for the surgeon
to study the technique outlined by
the manufacturer for each nail and to
carry it out carefully. With few
exceptions, the use of reamed nails is
still the standard.
Interlocking Screw Design
The design of interlocking screws
is somewhat more important than
nail design. Confining the threads to
the distal tip of the screw has been
thought to provide additional
strength to the screw. Unfortunately,
the weakness of the interlocking
screw is at the shaft-thread junction,
and thus little advantage is gained
100
from a partially threaded screw.
Also, this type of screw is less easily
inserted than the fully threaded
screw and is difficult to extract. Fur-
thermore, the partially threaded
screw gains purchase on only one
cortex, comes loose more often, and
backs out more frequently; thus, its
use necessitates the placement of two
screws distally. The fully threaded
screw appears to have the more logi-
cal design and is easier to use.
A more important feature than the
threads is the core diameter of the
screw. Screw failure is a common
complication of locked nails, and a
larger core diameter reduces this
risk. Materials such as titanium and
22-13-5 stainless steel also improve
screw strength.
Static Versus Dynamic
Locking
Early in the development of static
locking (locking the nail at each
end), there were concerns that this
technique would hinder impaction
and lead to an increased nonunion
rate, but many clinical investiga-
tors have since demonstrated that
this is not the case.20,26 Conversely,
dynamic locking (locking the nail
at only one end) has been found to
result in an increased rate of short-
ening and rotation and a higher
complication rate. Dynamization
(removal of the interlocking screws
at one end of the nail during the
healing process) was also popular
early in the use of interlocking
nails, but it also led to shortening
and rotation at the fracture site and
did not increase the union rate.27 In
light of adequate evidence of the
benefits of static locking,26 I recom-
mend static locking of all femoral
fractures from below the lesser
trochanter to the supracondylar
area, with dynamization reserved
for those fractures that have failed
to show healing at 4 to 6 months.
Journal of the American Academy of Orthopaedic Surgeons
Distal Targeting
Accurate targeting of the distal inter-
locking screws in their passage into
the screw holes has been the most
difficult operative feature of inter-
locking nailing. Many attempts have
been made to create proximal jigs to
aid in distal targeting, but these
devices have had limited value. Mag-
netic and light sources have also
proved to be of little use. Goulet et al28
have described the attachment of a
laser beam to a C-arm image
intensifier; although the device
appears attractive, it has not gained
widespread clinical use. 28 C-arm-
mounted targeting devices have also
been of limited benefit.29 Offset-power
equipment with radiolucent drill
chucks has provided a slight benefit.
Freehand targeting is still the most
popular method employed by sur-
geons experienced in this field.30 The
image intensifier is tilted and rotated
until the hole appears completely
round, indicating coaxial alignment.
The placement of the skin incision is
then determined fluoroscopically,
and the fascia is split beneath it. The
point of a sharp, elongated trocar
with a radiolucent handle is then
fluoroscopically placed at the point
on the lateral cortex that coaxially
aligns with the middle of the screw
hole (Fig. 5). Once this point is
located, the trocar or pin is driven
into the lateral cortex and is then
replaced with a drill bit. The drill bit
can be gently tapped through the nail
to the medial cortex before drilling to
prevent nicking the nail with the bit
and weakening it. This freehand tech-
nique has proved to be very success-
ful and requires only slight surgical
experience. It is currently the recom-
mended method for distal placement
of interlocking screws.
Number of Distal Screws
In most femoral shaft fractures,
placement of a single distal screw

Fig. 5 The sharp trocar is brought in
obliquely and aligned coaxially with the
screw hole.
provides adequate fixation and
decreases time spent in targeting. It
appears to be unimportant whether
this screw is placed in the proximal
or the distal screw hole. A fully
threaded screw is preferred, as a
screw with distal threads tends to
back out and necessitates the use of
two screws.
The use of two screws is generally
indicated in infraisthmal fractures to
prevent rotation around the nail and
flexion/extension about a single
screw. Two screws are also indicated
in severely comminuted femoral
fractures, as well as in unreliable
patients who refuse to limit weight
bearing and in head-injury patients.
Postoperative Management
In patients with unstable fractures,
protected weight bearing is neces-
sary until callus formation is evi-
dent. Patients with stable fractures
are allowed early weight bearing
with crutch support. Each patient’s
weight-bearing status is progressed
according to healing noted on fol-
low-up films and clinical progress.
Quadriceps rehabilitation is gen-
erally started 1 day postoperatively.
Chondromalacia is a common
sequela of these injuries, and the
early institution of vigorous physical
Vol 1, No 2, Nov/Dec 1993
therapy appears to exacerbate this
condition. Therefore, the patient
should begin with gentle quadriceps
muscle sets, straight leg lifts, and ter-
minal knee extensions. Progressive
quadriceps muscle work should be
added only as the patient improves.
There is no evidence that a continu-
ous-passive-motion machine is nec-
essary to obtain good results.
Nail Removal
The indications for nail removal are
unclear.31,32 There are no long-term
studies suggesting that removal of
the nail or interlocking screws is nec-
essary. At present, the indications
for removal are symptoms of hip
pain and pain over the screw heads.
Screws with greater head heights
tend to produce more symptoms,12,20
as do screws in subcutaneous areas.
Except in cases of delayed union and
nonunion, early or late dynamiza-
tion no longer appears necessary.
Open Fractures
In the treatment of open femoral frac-
tures with interlocking nails, two
important questions remain. The
first is whether the nailing should be
performed primarily or secondar-
ily.33 Little difference in the infection
rate has been found between frac-
tures nailed primarily and those
nailed in a delayed manner.34 The
second question is whether the
medullary canal should be reamed
or left unreamed. Many reports now
suggest that in open fractures caused
by low-velocity gunshot wounds35
and in Gustilo grade I, II, and IIIA
open femoral fractures, 23 reamed
locked intramedullary nailing is the
treatment of choice. Controversy
persists, however, about the treat-
ment of Gustilo grade IIIB and IIIC
open femoral fractures.23 These frac-
tures may be an indication for the use
of unreamed interlocking nails to
Robert A. Winquist, MD
avoid further damage to the blood
supply.
Second-Generation
Interlocking Nails
Second-generation interlocking
nails are used for fractures of the
proximal femur and combined
femoral neck-shaft fractures. These
nails are available with screws of
various sizes and with differing
angles of placement. Use of the
larger screws is unnecessary and
leads to an increased rate of nail fail-
ure because these screws require
larger screw holes. Screws may be
placed at a 135-, a 130-, or a 125-
degree angle to the femoral shaft.
The normal femoral neck-shaft
angle is 125 to 130 degrees, and
placement of the screws at the 135-
degree angle increases the difficulty
of screw insertion but facilitates
sliding.
Proximal targeting is much more
difficult with reconstruction nails
than with standard interlocking
nails, and the use of a radiolucent
plastic guide is helpful. The most
important technique is the place-
ment of a percutaneous Steinmann
pin along the anterior surface of the
femoral neck to define femoral
anteversion. As the nail is driven
into the bone, it must be rotated
properly so that the proximal jig is
parallel to the anterior pin. Correct
placement of the proximal screw in
the anteroposterior and lateral
planes is necessary. Because the
femoral neck and head project from
the anterior two thirds of the femoral
shaft, the starting point for nail
insertion in the proximal femur is 5
mm anterior to the usual piriformis
fossa starting point. This starting
point places the screws in better
alignment with the femoral neck and
greatly facilitates proximal target-
ing. However, a starting point
placed too far anteriorly leads to
101
Locked Femoral Nailing

fracture of the femoral shaft and fur- carried out with a reconstruction nail, nail insertion, a third screw, which is
ther comminution.18 and the two interlocking screws are cannulated, is added over the anterior
placed into the femoral head. After stabilizing pin.
Femoral Neck-Shaft
Fractures

Femoral neck fractures are found in

combination with approximately 1%
of all femoral shaft fractures. As a
precaution, preoperative radio-
graphs of the hip should be taken in
all patients with a femoral shaft frac-
ture. If the proximal fragment is
rotated, a femoral neck fracture may
be difficult to detect on film; thus, it
is helpful to examine the femoral
neck under fluoroscopy during nail
insertion. The majority of these
femoral neck fractures are high-
angle Pauwels type III fractures sus-
tained at the time of injury, not
during intramedullary nailing. It is
very important to recognize the
anterior location of the femoral neck Group 1 Group 2 Group 3
Neck nondisplaced Neck missed Neck displaced
relative to the femoral shaft, which
makes it possible to place femoral
neck pins and screws anteriorly but
not posteriorly.
Femoral neck-shaft fractures can
be divided into three clinical patterns:
group 1, nondisplaced femoral neck
fractures; group 2, missed femoral
neck fractures; and group 3, dis-
placed femoral neck fractures (Fig. 6).

Group 1: Nondisplaced Femoral

Neck Fracture
This fracture combination includes
a femoral shaft fracture with a nondis-
placed femoral neck fracture and pro-
vides an ideal indication for
second-generation locked nailing.
The surgical technique involves ini-
tially placing a temporary Steinmann Group 1 Group 2 Group 3
pin in the anterior portion of the Second-generation nail Add screws in neck Screws in neck, plate on shaft
femoral neck so that it will not
Fig. 6 Femoral neck-shaft fractures. Top, Classification. Top left, Group 1: Nondisplaced
obstruct the medullary canal during femoral neck fracture. Top center, Group 2: Missed femoral neck fracture. Top right, Group
nail placement. The medullary canal 3: Displaced femoral neck fracture. Bottom, Treatment. Bottom left, Group 1: Locked nail-
must be reamed to a diameter 1.5 to 2 ing is carried out with a reconstruction nail, and the two interlocking screws are placed into
the femoral head. Bottom center, Group 2: Placement of two additional screws in the femoral
mm larger than the reconstruction neck anterior to the intramedullary nail. Bottom right, Group 3: Open anatomic reduction of
nail to prevent displacement of the the femoral neck and multiple-screw fixation. The femoral shaft is then managed with a plate
femoral neck fracture during inser- or, in the case of a diaphyseal fracture, with a retrograde intramedullary nail.
tion of the nail. Locked nailing is then

102 Journal of the American Academy of Orthopaedic Surgeons

 Robert A. Winquist, MD

Group 2: Missed Femoral Neck can be managed with a standard trochanteric fractures for whom sec-
Fracture (first-generation) interlocking nail.36,37 ond-generation nails are indicated
In this group of fractures, the The only patients with true sub- are those with severe osteoporosis or
femoral neck fracture has been
missed initially and is discovered
intraoperatively or postoperatively,
after the femoral shaft fracture has
been nailed. The best form of treat-
ment involves returning the patient to
the operating room and placing two
additional screws in the femoral neck
anterior to the intramedullary nail.

Group 3: Displaced Femoral

Neck Fracture
This group of fractures includes a
femoral shaft fracture and a dis-
placed neck fracture that is identified
initially. The complications of
nonunion and avascular necrosis
that arise in femoral neck fractures
are extremely difficult to manage,
whereas the typical complications of Type 1 Type 3
Type 2
femoral shaft fractures are of a lower Lesser trochanter Lesser trochanter Greater trochanter
magnitude and easier to manage. intact fractured fractured
The recommended treatment for this
fracture combination is an anterior
capsular decompression with an
open anatomic reduction of the
femoral neck and multiple-screw
fixation. The femoral shaft is then
managed either with a plate or, in the
case of a diaphyseal fracture, with a
retrograde intramedullary nail.

Subtrochanteric Fractures
The availability of second-generation
nails extends the benefits of locked
nailing to fractures of the extreme
proximal regions of the femur. There
are three clinical patterns: type 1, true
subtrochanteric fractures; type 2,
reverse intertrochanteric fractures; Type 1 Type 2 Type 3
and type 3, intertrochanteric-sub- First-generation nail Second-generation nail Hip screw
trochanteric fractures (Fig. 7). Fig. 7 Subtrochanteric fractures. Top, Classification. Top left, Type 1: True subtrochanteric
fracture (lesser trochanter is intact). Top center, Type 2: Reverse intertrochanteric fracture
(lesser trochanter is fractured, but the greater trochanter and piriformis fossa are intact). Top
Type 1: True Subtrochanteric right, Type 3: Reverse intertrochanteric fracture (lesser trochanter is fractured, but the greater
Fractures trochanter and piriformis fossa are intact). Bottom, Treatment. Bottom left, Type 1: Treatment
is with a standard (first-generation) interlocking nail. Bottom center, Type 2: Treatment is with
The lesser trochanter is intact in a second-generation interlocking nail, which is statically locked. Bottom right, Type 3: Stan-
these fractures. True subtrochanteric dard treatment is with a compression hip screw.
fractures below the lesser trochanter

Vol 1, No 2, Nov/Dec 1993 103

Locked Femoral Nailing
with a metastatic lesion that may
extend into the intertrochanteric area.
Type 2: Reverse
Intertrochanteric Fractures
In this pattern the lesser trochanter
is fractured, but the greater trochanter
and piriformis fossa are still intact.
These fractures provide an ideal indi-
cation for a second-generation inter-
locking nail, which is statically locked
if there is any distal comminution.
Type 3: Intertrochanteric-
Subtrochanteric Fractures
In this group the fracture extends
into the greater trochanter and the
piriformis fossa. Standard treatment
is with a compression hip screw.
Only in those cases with minimal
displacement of the trochanteric
fracture and extensive shaft com-
minution should the use of a second-
generation nail be considered.
Routine use of second-generation
References
1. Johnson KD, Johnston DWC, Parker B:
Comminuted femoral-shaft fractures:
Treatment by roller traction, cerclage
wires and an intramedullary nail or an
interlocking intramedullary nail. J Bone
Joint Surg Am 1984;66:1222-1235.
2. Webb LX, Gristina AG, Fowler HL:
Unstable femoral shaft fractures: A com-
parison of interlocking nailing versus
traction and casting methods. J Orthop
Trauma 1988;2:10-12.
3. Küntscher G: Die Marknagelung von
Knochenbrüchen. Arch Klin Chir
1940;200:443-455.
4. Winquist RA, Hansen ST Jr: Commi-
nuted fractures of the femoral shaft
treated by intramedullary nailing.
Orthop Clin North Am 1980;11:633-648.
5. Browner BD, Cole JD: Current status of
locked intramedullary nailing: A
review. J Orthop Trauma 1987;1:183-195.
6. Küntscher G: Practice of Intramedullary
Nailing. Springfield, Ill: Charles C
Thomas Publishers, 1967.
7. Winquist RA, Hansen ST Jr, Clawson
DK: Closed intramedullary nailing of
femoral fractures: A report of five hun-
dred and twenty cases. J Bone Joint Surg
Am 1984;66:529-539.
104
nails in these fractures has led to a
high incidence of varus deformity
and failure. The incidence of varus
deformity is increased by supine
positioning of the patient and
adduction of the hip.
Summary
Closed intramedullary nailing with
reamed, statically locked nails is the
treatment of choice for the large
majority of femoral fractures from
the lesser trochanter to the supra-
condylar area. Closed reduction and
proper location of the piriformis
starting point for nail insertion are
the most important aspects of the
surgical technique. Nail design plays
a much smaller role. Distal targeting
of the interlocking screws continues
to be the most difficult surgical step,
and the freehand technique with a
sharp trocar is commonly used.
Static nailing is appropriate for
8. Butler MS, Brumback RJ, Ellison TS, et
al: Interlocking intramedullary nailing
for ipsilateral fractures of the femoral
shaft and distal part of the femur. J Bone
Joint Surg Am 1991;73:1492-1502.
9. Wiss DA, Fleming CH, Matta JM, et al:
Comminuted and rotationally unstable
fractures of the femur treated with an
interlocking nail. Clin Orthop 1986;212:
35-47.
10. Wiss DA, Brien WW, Stetson WB: Inter-
locked nailing for treatment of segmen-
tal fractures of the femur. J Bone Joint
Surg Am 1990;72:724-728.
11. Wu CC, Shih CH: Interlocking nailing of
distal femoral fractures: 28 patients fol-
lowed for 1-2 years. Acta Orthop Scand
1991;62:342-345.
12. Brumback RJ, Reilly JP, Poka A, et al:
Intramedullary nailing of femoral shaft
fractures: Part I. Decision-making errors
with interlocking fixation. J Bone Joint
Surg Am 1988;70:1441-1452.
13. Bone LB, Johnson KD, Weigelt J, et al:
Early versus delayed stabilization of
femoral fractures: A prospective ran-
domized study. J Bone Joint Surg Am
1989;71:336-340.
14. Brumback RJ, Ellison TS, Molligan H, et
Journal of the American Academy of Orthopaedic Surgeons
nearly all femoral shaft fractures, and
a single distal screw is adequate. The
use of unreamed nails is appropriate
only in Gustilo grade IIIB and IIIC
open femoral fractures and in
femoral fractures in patients with
multiple injuries, particularly those
involving the chest.
Second-generation interlocking
nails provide an ideal treatment for
combined femoral neck-shaft frac-
tures in which the neck is nondis-
placed. These nails are also indicated
for pathologic fractures in the
intertrochanteric and subtrochanteric
regions. In subtrochanteric fractures
they are best used when the lesser
trochanter is fractured but the piri-
formis fossa is intact. A standard
interlocking nail can be used in sub-
trochanteric fractures below the
lesser trochanter. For fractures
extending into the greater trochanter,
the traditional compression hip
screw is still the treatment of choice.
al: Pudendal nerve palsy complicating
intramedullary nailing of the femur. J
Bone Joint Surg Am 1992;74:1450-1455.
15. Lyon T, Koval KJ, Kummer F, et al:
Pudendal nerve palsy induced by frac-
ture table. Orthop Rev 1993;22:521-525.
16. McFerran MA, Johnson KD: Intra-
medullary nailing of acute femoral shaft
fractures without a fracture table: Tech-
nique of using a femoral distractor. J
Orthop Trauma 1992;6:271-278.
17. Wiss DA, Brumback RJ, Kyle RF, et al:
Current concepts in femoral nailing.
Contemp Orthop 1993;26:177-214.
18. Johnson KD, Tencer AF, Sherman MC:
Biomechanical factors affecting fracture
stability and femoral bursting in closed
intramedullary nailing of femoral shaft
fractures, with illustrative case presen-
tations. J Orthop Trauma 1987;1:1-11.
19. Browner BD: Pitfalls, errors, and compli-
cations in the use of locking Küntscher
nails. Clin Orthop 1986;212:192-208.
20. Cameron CD, Meek RN, Blachut PA, et
al: Intramedullary nailing of the femoral
shaft: A prospective, randomized study.
J Orthop Trauma 1992;6:448-451.
21. Russell TA, Taylor JC, LaVelle DG, et al:
Mechanical characterization of femoral

interlocking intramedullary nailing sys-
tems. J Orthop Trauma 1991;5:332-340.
22. Alho A, Moen O, Husby T, et al: Slotted
versus non-slotted locked intramedullary
nailing for femoral shaft fractures. Arch
Orthop Trauma Surg 1992;111:91-95.
23. Brumback RJ, Ellison PS Jr, Poka A, et al:
Intramedullary nailing of open fractures
of the femoral shaft. J Bone Joint Surg Am
1989;71:1324-1331.
24. Bucholz RW, Ross SE, Lawrence KL:
Fatigue fracture of the interlocking nail
in the treatment of fractures of the distal
part of the femoral shaft. J Bone Joint
Surg Am 1987;69:1391-1399.
25. Franklin JL, Winquist RA, Benirschke
SK, et al: Broken intramedullary nails.
J Bone Joint Surg Am 1988;70:1463-
1471.
26. Brumback RJ, Uwagie-Ero S,
Lakatos RP, et al: Intramedullary
nailing of femoral shaft fractures:
Part II. Fracture-healing with static
Vol 1, No 2, Nov/Dec 1993
interlocking fixation. J Bone Joint Surg
Am 1988;70:1453-1462.
27. Thoresen BO, Alho A, Ekeland A, et al:
Interlocking intramedullary nailing in
femoral shaft fractures: A report of
forty-eight cases. J Bone Joint Surg Am
1985;67:1313-1320.
28. Goulet JA, Londy F, Saltzman CL, et al:
Interlocking intramedullary nails: An
improved method of screw placement
combining image intensification and
laser light. Clin Orthop 1992;281:199-
203.
29. Kempf I, Grosse A, Beck G: Closed locked
intramedullary nailing: Its application to
comminuted fractures of the femur. J
Bone Joint Surg Am 1985;67:709-720.
30. Knudsen CJM, Grobler GP, Close REW:
Inserting the distal screws in a locked
femoral nail. J Bone Joint Surg Br
1991;73:660-661.
31. Miller R, Renwick SE, DeCoster TA, et
al: Removal of intramedullary rods after
Robert A. Winquist, MD
femoral shaft fracture. J Orthop Trauma
1992;6:460-463.
32. Brumback RJ, Ellison TS, Poka A, et al:
Intramedullary nailing of femoral shaft
fractures: Part III. Long-term effects of
static interlocking fixation. J Bone Joint
Surg Am 1992;74:106-112.
33. Chapman MW: The role of intramed-
ullary fixation in open fractures. Clin
Orthop 1986;212:26-34.
34. Lhowe DW, Hansen ST: Immediate nail-
ing of open fractures of the femoral shaft.
J Bone Joint Surg Am 1988;70:812-820.
35. Wiss DA, Brien WW, Becker V Jr: Inter-
locking nailing for the treatment of
femoral fractures due to gunshot wounds.
J Bone Joint Surg Am 1991;73:598-606.
36. Wiss DA, Matta JM, Sima W, et al: Sub-
trochanteric fractures of the femur.
Orthopedics 1985;8:797-800.
37. Wu CC, Shih CH, Lee ZL: Sub-
trochanteric fractures treated with inter-
locking nailing. J Trauma 1991;31:326-333.
105
JAAOS Home Page

Table of Contents

Search

Help
 Elbow Arthritis: Treatment Options
Shawn W. O’Driscoll, MD, PhD, FRCS(C)

Abstract

The treatment of elbow arthritis is conceptually similar to that for arthritis of associated with some loss of flexion
other major joints. The treatment of elbow arthritis has been evolving rapidly due as well. There may be crepitus in the
to advances in arthroscopic techniques and surgical treatment for contractures elbow, but the characteristic finding
and improved prosthetic designs. The reliability of total elbow replacement is is pain on forced extension or flexion.
approaching that of total replacement of the knee, hip, and shoulder. There remain On the radiographs there are
a number of controversies and unanswered questions that require further experi- osteophytes on the olecranon and
ence and longer follow-up for resolution. coronoid processes, osteophytes
J Am Acad Orthop Surg 1993;1:106-116 filling in the olecranon and coronoid
fossae, and usually loose bodies
(which may not actually be loose)
Although pain is the most com- soft tissues, causes joint laxity that (Fig. 1). In the advanced stages the
mon complaint, patients with elbow results in mechanical wearing and radioulnar joint and finally the radio-
arthritis may also complain of stiff- further destruction due to malalign- humeral joint may become involved.
ness, weakness, instability, or cos- ment or subluxation. Eventually, the The etiology of this condition is
metic deformity. The combination of elbow can become flail, with exces- still not known. The fact that both
complaints and their relative severity sive motion in the coronal plane. degenerative arthritis and osteochon-
determine the treatment options and dritis dissecans are so prevalent in
the likelihood of patient satisfaction. throwing athletes suggests a link
Osteoarthritis
between the two. Also, many patients
Rheumatoid Arthritis Primary osteoarthritis of the elbow, with osteoarthritis have loose bodies,
only recently recognized and indicating that loose bodies might be
Rheumatoid arthritis affects the described in the English-language causally related to the arthritis.
elbow less frequently than other literature, is characteristic in its clini-
joints, but when it does occur, it cal and radiographic presentations.1 Posttraumatic Arthritis
results in painful impairment of Originally recognized in Japan,
function that for years we have where its treatment was also first Posttraumatic arthritis can occur fol-
tended to overlook or minimize described, osteoarthritis of the elbow lowing various injuries, but is most
because of a general pessimism is most commonly seen in men with common with distal humeral frac-
regarding treatment options and a history of heavy use of the arm, tures that involve intra-articular
results. The severity of the disability weight lifters, and throwing athletes. comminution. Stiffness is common.
is profoundly realized by patients In fact, it is a disorder almost exclu- Nonunions in this region usually
who have had bilateral elbow sive to men. They present in their result in a flail dysfunctional elbow.
involvement for an extended period third to eighth decades with a char- Treatment is dictated by the patho-
of time and then have one elbow acteristic history of mechanical-
replaced. They usually request impingement pain at the extremes of
surgery on the contralateral side motion, classically in extension more
within a few months. so than in flexion. Carrying any- Dr. O’Driscoll is Associate Professor of Orthope-
The pattern of involvement of the thing, such as a briefcase, with the dics, Department of Orthopedics, Mayo Clinic
elbow is similar to that of other joints, elbow extended is painful. Pain in and Mayo Medical School, Rochester, Minn.
with the primary involvement in the the midportion of the arc of motion is Reprint requests: Dr. O’Driscoll, Orthopedic
ulnohumeral articulation. Loss of present only in the late stage. A Research, Mayo Clinic, Medical Science Build-
bone stock, with or without associ- flexion contracture of approximately ing, 3rd Floor, Rochester, MN 55905.
ated destruction of the periarticular 30 degrees is typical and may be

106 Journal of the American Academy of Orthopaedic Surgeons

 Shawn W. O’Driscoll, MD, PhD, FRCS(C)

A B

Fig. 1 Primary degenerative arthritis of the elbow has a classic pattern of radiographic changes, characterized by osteophytes on the coro-
noid and olecranon processes (arrows); coronal osteophytes encroaching on the margins of the coronoid and olecranon fossae, with thick-
ening of the normally thin bone separating these two fossae; and eventually loss of the articular cartilage and involvement of the radioulnar
and radiohumeral joints. Loose bodies (often adherent to the soft tissues) are common, though not seen on these anteroposterior (A) and lat-
eral (B) radiographs.

logic findings, complaints, and age
of the patient.
Nonsurgical Treatment
The nonsurgical management of
elbow arthritis includes the standard
medical treatment and physical ther-
apy for most other joint disorders.
Acetylsalicylic acid and nonsteroidal
anti-inflammatory agents are used
unless precluded by gastrointestinal
side effects. More potent agents,
including antimalarial agents, gold
salts, immunosuppressive drugs,
and corticosteroids, are resorted to
when necessary. Intra-articular injec-
tions of corticosteroids are easily per-
formed and should be considered
before surgery. Radioactive synovec-
tomy, performed by sterile intra-
articular injection of a radioisotope, is
also minimally invasive and should
probably be recommended as a more
conservative treatment option to
young patients with inflammatory
arthritis, those with early inflamma-
tory arthritis, and those who are can-
didates for surgical synovectomy.
Physical therapy includes pain-
control measures, such as avoidance
of activities that place excessive
stresses on the elbow, intermittent
periods of rest, and application of
heat or cold. Splinting is sometimes
useful. Lightweight hinged splints
that permit active range-of-motion
exercises protect the elbow from
varus-valgus stresses and minimize
pain. Resting or night splints also
can be helpful. Gentle exercises
should be performed on a regular
basis to maintain mobility and
strength in the muscles. Occupa-
tional therapy interventions with
aids for activities of daily living are
useful. These would include handle
extensions to cope with elbow-
flexion contractures.
Surgical Treatment
Options
Surgery is indicated following fail-
ure of nonsurgical management.
There are a number of surgical
options, including arthroscopy,
open synovectomy, osteotomy,
resection and interpositional arthro-
plasty, arthrodesis, and total elbow
arthroplasty (TEA). Total elbow
arthroplasty provides the most con-
sistent results. However, the stage of
the disease, the age of the patient,

Vol 1, No 2, Nov/Dec 1993 107

Elbow Arthritis

and the presence of other joint
involvement are important determi-
nants of treatment choice.
Arthroscopy
Arthroscopy is assuming a greater
role in diagnosis and management
of elbow problems, as it is in other
joint disorders. It is useful to per-
form a synovial biopsy. Undiag-
nosed painful snapping of the elbow
can be associated with cartilaginous
loose bodies that do not appear on
radiographs, posttraumatic arthritis,
primary degenerative arthritis,
dense soft-tissue adhesions (e.g., fol-
lowing radial-head excision), and
ulnohumeral rotatory instability.
Patients with spontaneous onset of
contracture are often found to have a
form of inflammatory arthritis.
Patients with localized posttrau-
matic arthritis sometimes benefit
from debridement of the area and
localized synovectomy. A complete
synovectomy is technically possible
for the management of inflamma-
tory or septic arthritis, although
technically highly demanding and
associated with a theoretical risk to
neurovascular structures. One must
be constantly aware of the fact that
the nerves may be within a few mil-
limeters of the operating instru-
ments in the anterior part of the
elbow. Although the safety of this
procedure has not yet been proved,
we believe that the risks are minimal
if certain safety precautions are
observed. The advantages of arthro-
scopic over open synovectomy are
impressive. It is done as an outpa-
tient procedure, causes minimal
morbidity, and permits rapid return
of motion, and a complete synovec-
tomy is technically possible. Treat-
ment of primary degenerative
arthritis is possible in the early
stages by removal of the osteophytes
from the olecranon and coronoid as
well as from the olecranon fossa (Fig.
2).2,3 Removal of osteophytes from
the coronoid fossa is more difficult.
Open Synovectomy
Synovectomy with or without radial-
head excision is a well-recognized
and accepted form of treatment for
rheumatoid arthritis. Satisfactory
pain relief is obtained in about 70% to
90% of patients.4 The good results are
reported to persist. Increased range
of motion is less likely than pain
relief. There is controversy regarding
its success in later stages after joint
destruction has occurred. Also
unclear is the role of radial-head exci-
sion. Progressive articular destruc-
tion following synovectomy and
radial-head excision has been noted
and is thought to be due to increased
ulnohumeral loading. Late valgus
instability has been a problem in the
experience of some surgeons.
In general, surgeons experienced
with both TEA and synovectomy
favor TEA in the later stages
because the patients are so much
more satisfied and the functional
improvement is so much greater.
Osteotomy
Treatment of osteoarthritis consists of
decompressing the impinging areas.
Currently this is being performed
with use of the Outerbridge-Kashi-
wagi (ulnohumeral) arthroplasty,

A B C
Fig. 2 Arthroscopic treatment of osteoarthritis. A, Osteophytes are removed with a small osteotome and graspers. A bur is used to smooth
off the olecranon (B) and to recreate the olecranon fossa, removing any osteophytes and thickened bone (C). (Reproduced with permission
from O’Driscoll SW, Morrey BF: Arthroscopy of the elbow, in Morrey BF (ed): The Elbow and Its Disorders. Philadelphia: WB Saunders, 1993,
p 128.)

108 Journal of the American Academy of Orthopaedic Surgeons

 Shawn W. O’Driscoll, MD, PhD, FRCS(C)

which is really a core osteotomy of
the distal humerus and osteotomies
of the tips of the olecranon and coro-
noid1 (Fig. 3). It is performed through
a triceps-splitting approach using the
Cloward drill to go through the
humerus (Fig. 4).
This procedure is indicated for
primary osteoarthritis in patients
with pain at the extremes of motion,
but not in the midportion of the arc
of motion or at rest. The procedure
characteristically relieves impinge-
ment pain and frequently permits
some improvement in range of
motion, especially when the rehabil-
itation program involves the use of
patient-adjusted static braces post-
operatively. Successful results (pain
and motion improved) have been
reported in 85% of patients.1
Resection and
Interpositional
Arthroplasty
Resection arthroplasty is an option
for salvaging an elbow, particularly
following failed TEA. Its success
(relatively pain-free functional arc of
motion with reasonable stability) is
more likely if the medial and lateral
columns of the distal humerus and
the olecranon and coronoid remain
in place.5 If the elbow becomes flail
or grossly unstable, the limb remains
nonfunctional, and the result is
unsatisfactory.
For younger patients (typically
less than 60 years of age), interposi-
tion arthroplasty is recommended
for posttraumatic arthritis if bone
loss does not preclude it.6,7 The pro-
cedure involves removal and/or
reshaping of the articular surfaces
and resurfacing with an interposi-
tion tissue such as autogenous fascia
lata or dermis. Distraction arthro-
plasty involves the use of a hinged
external fixation device that holds
the elbow joint slightly distracted,
stable, and aligned while permitting
full motion in the first few weeks fol-
lowing interposition arthroplasty
(Fig. 5). The results are satisfactory
in most cases, although the tech-
niques are demanding and require
substantial expertise.
In young patients I have used
periosteum from the proximal tibia
for “biologic resurfacing” because of
its potential to regenerate articular
cartilage (Fig. 6). The indications and
contraindications as well as results to
be expected are not yet fully known;
thus, it remains experimental.
Arthrodesis
Arthrodesis of the elbow is incom-
patible with satisfactory function
due to the fact that range of motion
of the elbow is essential for use of the
hand. There is no single optimal
position. It is indicated when
intractable sepsis is present and
when reconstruction by revision
TEA is no longer possible. It is prob-
ably never indicated as a primary
procedure, although controversy
exists in the case of young male
patients who perform heavy labor.
Fortunately, this situation is rare.
Total Elbow Arthroplasty
The evolution of TEA has had simi-
larities to that of total knee arthro-
plasty. Biomechanically, there are
three types of prosthetic joint

A B C

Fig. 3 Outerbridge-Kashiwagi (ulnohumeral) arthroplasty (same patient as in Fig. 1). A, Procedure involves excision of the osteophyte from
the olecranon (arrows), core osteotomy of the humerus to remove the marginal osteophytes from the olecranon and coronoid fossae, and
excision of the coronoid osteophytes through the hole in the humerus. Loose bodies are removed anteriorly and posteriorly. In the elbow
shown, there are also osteophytes on the capitellum and radial head. B, Fenestration created by the arthroplasty mimics a congenital fenes-
tration seen in some patients (C) and does not significantly weaken the humerus.

Vol 1, No 2, Nov/Dec 1993 109

Elbow Arthritis
A B
Fig. 4 Surgical technique of ulnohumeral arthroplasty. A, Olecranon is exposed through a triceps-splitting approach, and osteophytes are
removed. B, Large trephine (large Cloward drill) is used to fenestrate the distal humerus, angling it proximally to exit at the margin of the
joint. C, Coronoid osteophyte is removed under direct vision through the fenestration.
designs: nonconstrained, semicon-
strained, and constrained.
Over two decades ago, it was
observed that satisfactory pain relief
could be provided to patients with
arthritis by replacing the elbow joint
with a hinged prosthesis. This type
of constrained prosthesis transfers
all of the stresses directly to the pros-
thesis-cement-bone interfaces. It is
110
therefore associated with a very high
failure rate due to mechanical loos-
ening. The same was found to be
true of hinged designs in the knee
and ball-and-socket designs for the
shoulder. A major degree of bone
destruction accompanies such loos-
ening, making salvage difficult.
Although it is rare in medicine to be
able to state categorically that there
Journal of the American Academy of Orthopaedic Surgeons
C
is no indication for a certain proce-
dure, this is true for arthroplasty
with the constrained-hinge type of
elbow prosthesis, which has now
been abandoned. All the theoretical
advantages of a constrained
arthroplasty can be provided by a
semiconstrained design with a per-
manent coupling-bolt type of articu-
lation.
Fig. 5 The hinged elbow
distraction device designed
by Morrey permits stable
alignment of the elbow, vari-
able distraction, and motion
in both flexion-extension
and pronation-supination
arcs. (Reproduced with per-
mission from Morrey BF:
Post-traumatic contracture
of the elbow: Operative
treatment, including distrac-
tion arthroplasty. J Bone Joint
Surg Am 1990;72:601-618.)

Fig. 6 The patient, a 22-year-old woman, had a painful stiff elbow with posttraumatic arthri-
tis secondary to an open fracture-dislocation 4 months earlier. Photographs obtained 3 weeks
after surgery show active motion from 20 to 130 degrees with the hinged elbow distractor in
place. (Reproduced with permission from O’Driscoll SW: Surgery of elbow arthritis, in
McCarty DJ, Koopman WJ [eds]: Arthritis and Allied Conditions, 12th ed. Philadelphia: Lea &
Febiger, 1993, p 957.)
Less-constrained prostheses
should be less prone to mechanical
loosening, because the stresses are
absorbed by the soft tissues rather
than being transferred to the bone-
prosthesis interface. A true noncon-
strained joint replacement provides
little or no inherent stability by virtue
of its shape and articulation, there-
fore relying solely on the periarticular
soft tissues for stability (Fig. 7). The
current surface-replacement prosthe-
ses are not truly nonconstrained and
would be better termed “minimally
constrained,” as there is a degree of
constraint afforded by the articula-
tion itself. Examples include those
designed by Ewald (capitellocondy-
lar) and by Pritchard, the two most
popular in North America, as well as
those by Sorbie, Souter, Lowe, Liver-
pool, London, Wadsworth, and
Kudo. These designs have been in use
since 1972.
There was an initial trend to sim-
ply replace the articular surfaces of
the distal humerus and proximal
ulna, but these components without
intramedullary stems had a ten-
dency to loosen and displace. Kudo
Vol 1, No 2, Nov/Dec 1993
and Iwano 8 reported a 70% inci-
dence of loosening for nonstemmed
humeral components. The majority
of components now available have
intramedullary stems that help to
prevent the rocking or tilting type
of motion that causes loosening.
Loosening is no longer a common
problem with nonconstrained
replacements. Instability (disloca-
tion, subluxation, or maltracking)
has been a problem in 5% to 20% of
nonconstrained TEAs. This is par-
ticularly true when loss of bone or
soft-tissue integrity is significant.
A loose-hinge or sloppy-hinge
semiconstrained prosthesis offers a
compromise between the stability
provided by a hinged prosthesis and
the low incidence of loosening of a
nonconstrained surface replacement.
In most designs the ulnar and
humeral components are linked so
that they do not dislocate, but the link-
age allows for a degree of laxity that
permits the soft tissues to absorb
some of the stresses that would nor-
mally be applied to the prosthesis-
cement-bone interface. Such designs
include the Pritchard-Walker,
Shawn W. O’Driscoll, MD, PhD, FRCS(C)
Pritchard Mark II, Coonrad II, Mor-
rey-Coonrad (Mayo-modified Coon-
rad)(Fig. 8), GSB III, triaxial, and
AHSC (Volz). This is the most com-
monly used class of elbow replace-
ments today.
The indications for use of a semi-
constrained prosthesis include all
cases in which bone-stock or soft-tis-
sue integrity is not adequate for use of
a minimally constrained device.
Although it might be theoretically
more likely to loosen than a minimally
constrained device, this is not turning
out to be so in clinical experience and
reports in the literature.6,8-12 Thus, some
consider a semiconstrained prosthesis
to be indicated in any patient requir-
ing TEA. Others reserve minimally
constrained devices for patients under
the age of 60.
Indications
The general indication for
surgery is the same as that for
replacement of the hip, knee, or
shoulder—improvement in the
quality of life by restoration of pain-
free function (motion, stability, and
strength) in a joint that is causing
functional impairment. This is indi-
cated when such a goal cannot be
met by nonsurgical means or other,
less invasive surgical options.
The most common diagnosis for
which TEA is performed is rheuma-
toid arthritis. The typical patient
undergoing TEA is in American
Rheumatism Association class III or
IV (i.e., capable of performing only
some or none of the usual occupa-
tional or daily activities).13 Other indi-
cations include the treatment of
supracondylar or intercondylar
nonunions of the distal humerus,
severely comminuted acute supra-
condylar or intercondylar fractures of
the distal humerus in elderly patients
with osteoporotic bone that cannot be
reduced and fixed adequately, and
flail elbow caused by posttraumatic
loss of bone or structural integrity.
111
Elbow Arthritis

Fig. 7 Patients with adequate bone stock and soft tissues for stability can be treated with a
nonconstrained arthroplasty such as the capitellocondylar (Ewald) prosthesis. This is the old-
est elbow prosthesis still in use and is reported by the originator to have excellent long-term
results. It does not include a radial head component. Though a radial head might increase
stability, its insertion would require precise alignment and sizing, making the operation
more complicated. (Reproduced with permission from Ewald FC, Simmons ED Jr, Sullivan
JA, et al: Capitellocondylar total elbow replacement in rheumatoid arthritis: Long-term
results. J Bone Joint Surg Am 1993;75:498-507.)

The best results are often seen in
patients who preoperatively have
little or no use of the limb; postoper-
atively, they frequently have normal
or near-normal motion, strength,
and stability and no pain. Surpris-
ingly, the rehabilitation is faster in a
patient with a supracondylar
nonunion because the operation can
be done with less soft-tissue dissec-
tion and without detaching the tri-
ceps tendon. As a result, the patient
can use the arm without restrictions
immediately following surgery.
Contraindications
The contraindications are similar
to those for replacement of the other
major joints. The only absolute con-
traindication is active infection of the
joint. A history of postseptic arthritis
or osteomyelitis is a relative con-
traindication. Most would recom-
mend reserving TEA for patients over
the age of 60, although lesser age is
not an absolute contraindication.6 Of
course, it is preferable to first exhaust
all other treatment options, including
distraction interposition arthroplasty.
Loss or destruction of bone or soft
tissue is not a contraindication to TEA,
for these problems can be dealt with
surgically. Custom arthroplasties
have been used for treatment of anky-
losis or supracondylar nonunions.9
With appropriate implant selection,
however, custom components are
rarely required, usually being
reserved for revisions or patients with
juvenile rheumatoid arthritis.6
Consideration of Other Joint
Involvement
Patients with rheumatoid arthritis
requiring TEA may have advanced
involvement of the ipsilateral shoul-
der as well. Although the controversy
over which joint should be replaced
first continues, the joint that is more
disabling should probably be oper-
ated on initially. The results for shoul-
der and elbow replacement are similar
to those seen following replacement of
each as an isolated joint.13
Similarly, the contralateral elbow
may require replacement. Again, the
more disabling joint should be oper-
ated on first. The second operation
can be done as soon as the patient is
able to look after himself or herself
with the limb that has recently under-
gone surgery. The results of bilateral
elbow arthroplasties in patients with
rheumatoid arthritis are as good as
those after single-joint replace-
ments.13 My limited experience with
simultaneous bilateral elbow replace-
ments has been very encouraging.
The elbow becomes a true weight-
bearing joint in many patients with
rheumatoid arthritis (as does the
shoulder) because of arthritis in the
lower extremities. Patients who
undergo TEA generally have had pre-
vious operations.13 The need for sub-
sequent lower-extremity surgery,
resulting in requirement of walking
aids, is not a contraindication for
elbow replacement. In fact, some
patients are able to bear weight
through the upper extremities far bet-
ter after joint replacement of the
elbow or shoulder than before.
Technique
“The front door to the elbow is at
the back.” Although there are many

112 Journal of the American Academy of Orthopaedic Surgeons


Fig. 8 Coonrad II elbow
prosthesis, as modified by
Morrey, has a porous-mate-
rial-coated anterior flange,
under which a bone graft is
placed to enhance fixation
and resist the posterior forces
and torsional moments on
the humeral component.
Incorporation of the bone
graft and cortical remodeling
are expected in 80% of cases
or more. This design has
proved highly versatile and
clinically successful.
surgical approaches to the elbow,
each with its own specific advantages
and disadvantages, the versatility of
the posterior approach makes it supe-
rior. A posteriorly placed (slightly
medial or lateral) skin incision per-
mits posteromedial and posterolat-
eral arthrotomies as well as access to
the ulnar nerve and the anterior
elbow via the deep portion of the
Kocher approach. It is therefore the
most useful approach for the elbow.
The skin incision should not cross the
tip of the olecranon in patients with
olecranon bursitis or rheumatoid
arthritis, in whom the soft tissues
over the olecranon are pathologically
altered and more susceptible to
wound breakdown and infection. It is
analogous to the “universal” straight
anterior approach to the knee.
Access to the elbow joint can be
accomplished by reflecting the tri-
ceps with use of the Bryan-Morrey
approach. Others have suggested
reflecting the triceps with a flake of
bone from the tip of the olecranon,
but my personal experience with
this method has been disappointing
due to a high nonunion rate. Some
Vol 1, No 2, Nov/Dec 1993
still advocate a Kocher approach or a
posterior triceps-splitting or triceps-
tongue approach with careful clo-
sure. Ewald et al11 strongly favor a
modified Kocher approach for the
capitellocondylar prosthesis. The
olecranon is never osteotomized as it
is for internal fixation of distal
humeral fractures.
The fine details of surgical tech-
nique will not be discussed here.
However, there are several impor-
tant considerations. Careful han-
dling of the skin and soft tissues is
important, and the skin incision
must not devascularize a compro-
mised region of skin created by pre-
vious incisions. The ulnar nerve is
explored and retracted gently (usu-
ally transposed anteriorly as part of
the procedure). The triceps mecha-
nism is reflected in one of the ways
mentioned unless there is significant
laxity due to bone loss or soft-tissue
laxity, in which case it can be pre-
served. The origin of one ligament is
released, the joint is subluxated or
dislocated, and the bones are pre-
pared for the appropriate compo-
nents. A synovectomy is performed,
Shawn W. O’Driscoll, MD, PhD, FRCS(C)
along with release of any contrac-
tures. The canal is prepared using
current standard cementing tech-
niques, and cement is injected and
pressurized.
If a nonconstrained prosthesis is
used, alignment of the components
and proper soft-tissue balancing are
critical for stability. This includes the
ulnar part of the lateral collateral lig-
ament, which must be properly
repaired to prevent posterolateral
rotatory subluxation of the ulno-
humeral joint.14 Repair of the triceps is
critical for stability of nonconstrained
devices. Some prefer 2 to 4 weeks of
immobilization postoperatively.
With semiconstrained prostheses,
early motion avoiding resisted
extension is probably safe. In such
situations, I start motion 36 hours
after surgery and limit the patient
only from actively extending the
elbow against resistance for 6 weeks.
Positioning of the center of rota-
tion of the prosthesis in alignment
with that of the elbow is important
for proper balancing of the muscle
moment arms. With nonconstrained
devices, it is also important for sta-
bility.
Results
Pain relief is dramatic and as pre-
dictable as that found after total hip
or knee replacement. 10,13,15 At least
90% of patients are highly satisfied
with pain relief. Functional improve-
ment is predictable following
TEA. 6,10,13,15 In a prospective study,
Morrey et al15 showed that strength
increased 90% in flexion and 60% to
70% in pronation-supination. Exten-
sion strength remained relatively
unchanged, which might be
explained on the basis of surgical
approach (detachment and reattach-
ment of the triceps) and offset of the
axis of rotation of the prosthesis.10,15,16
The percentage of improvement in
strength was greater in patients with
rheumatoid arthritis.
Morrey et al have shown that the
113
Elbow Arthritis
functional arcs of motion of the elbow
(i.e., those required to perform the
activities of daily living) are 30 to 130
degrees of flexion and from 50
degrees of supination to 50 degrees of
pronation. Before surgery, patients
usually have less than these func-
tional arcs, with preoperative ranges
of motion averaging 70 degrees of
flexion-extension and 90 degrees of
pronation-supination.13 These aver-
ages increase postoperatively to 100
degrees of flexion-extension and 130
degrees of pronation-supination. The
“functional arcs of motion” are
achieved by most patients. Excellent
motion, close to the functional range,
is also possible in patients with com-
plete ankylosis of the elbow.6
Gains in motion, especially exten-
sion, are usually greater with semi-
constrained prostheses than with
minimally constrained prostheses.
Use of the former permits complete
release of contracted soft tissues and
immediate unrestricted motion
postoperatively, whereas such soft-
tissue releases and unrestricted
extension predispose to dislocation
of surface-replacement prostheses.
Two problems that thwarted early
progress in TEA were mechanical
loosening of constrained (hinged)
designs and dislocation of noncon-
strained designs. The early hinged
design was a fully constrained pros-
thesis that linked the ulnar and
humeral components directly. This
resulted in transfer of all forces and
moments about the elbow directly to
the prosthesis-cement-bone interface.
The failure rate was unacceptably
high, just as it was with this design
concept in knee replacements. 17
Although the elbow has been com-
monly referred to as a non-weight-
bearing joint, the forces that cross it
can exceed three times body weight.
The principal moments (rotational
forces and torques) about the humeral
component are posterior and rota-
tional. These forces can be considered
in the design of a prosthesis.
114
The problem of instability (recur-
rent dislocation or subluxation) of a
nonconstrained elbow prosthesis
appears to have decreased in more
recent reports, but still is in the range
of 5% to 20%. This problem will
likely diminish as our understand-
ing of the mechanism of elbow insta-
bility improves. Until recently, we
were not aware of the fundamental
posterolateral rotatory instability
pattern by which an elbow sublux-
ates or dislocates.14 The important
ulnar part of the lateral collateral lig-
ament complex is violated during
TEA and must be reconstructed.
Also, the soft-tissue constraints
depend on the integrity of the nor-
mal articular architecture to function
properly. If the design of the ulnar
and humeral prosthetic articular
surfaces is not anatomic, the soft-tis-
sue constraints might not maintain
joint stability.
Despite these problems, the mini-
mally constrained TEA prosthesis,
such as the capitellocondylar device,
has been used with satisfactory long-
term success since 1974, with average
follow-up periods of 6 to 7 years.
Ewald et al11 recently reported the
results with 202 capitellocondylar
prostheses after 2 to 15 years (mean,
6 years). Pain relief and functional
improvement were excellent, with
patients scoring an average of 26 pre-
operatively and 91 postoperatively
on a 100-point rating score. Reopera-
tion was required in only 5% of the
cases for loosening, dislocation, and
infection. It was the authors’ impres-
sion that complications seen in ear-
lier years had diminished. This
report from the originator of the
longest-used total elbow is extremely
impressive and indicates that the
results do not deteriorate much with
time.
Both potential problems, loosen-
ing of the constrained-hinge type of
prosthesis and dislocation of the
nonconstrained type, might be over-
come by use of the semiconstrained
Journal of the American Academy of Orthopaedic Surgeons
design.6 The concept of this design
is that the ulnar and humeral com-
ponents are linked by a “loose
hinge,” so that they cannot dislo-
cate or subluxate; however, the lax-
ity built into the sloppy hinge
permits some of the forces and
moments applied across the elbow
to be absorbed by the soft tissues
around it. The static (ligamentous)
and dynamic (muscle) soft-tissue
constraints thus theoretically take
on the role that they play in a non-
constrained design, decreasing the
likelihood of loosening.
This concept has been in clinical
use for over a decade and has pre-
dominated the field of elbow
replacement surgery in the past
decade. There are a number of semi-
constrained designs, and all appear
to be successful. They have been in
use since 1976, and results after fol-
low-up periods averaging up to 9
years have been reported, with
mechanical (nonseptic) loosening
rates of less than 5%.6,9,10,12
The usefulness of the semicon-
strained concept has been confirmed
in laboratory studies. 16 A Mayo-
modified Coonrad design with a
loose hinge (10 degrees of varus/val-
gus and rotational laxity) and an
anterior flange to resist posterior
forces and rotational moments was
tested in cadaver elbows during sim-
ulated active motion and with maxi-
mum varus and valgus moments.
Loading of the biceps, brachialis, and
triceps muscles permitted reproduc-
tion of a nearly normal kinematic pat-
tern and limited varus or valgus
deflections. Thus, at least for the one
type of semiconstrained prosthesis
tested, the concept is feasible and not
just semantically different from that
of a constrained hinge. These data are
thought to at least partially explain
the low rates of loosening observed
clinically in the past decade.
Morrey and Adams12 reported a
95% Kaplan-Meier estimated survival
at 7 years in 68 patients with rheuma-

toid arthritis treated with a Mayo-
Coonrad prosthesis. There were no
cases of mechanical loosening. Longer
follow-up will determine whether the
low incidence of loosening will paral-
lel that in the hip and knee, as it has
after intermediate follow-up.
Controversies and Future
Challenges
The most rapidly evolving aspects of
elbow surgery relate to the use of
arthroscopy and arthroplasty. The
indications are expanding for both
of these procedures. With medium-
term results (5 to 10 years) that are
similar to those for hip and knee
arthroplasty, TEA can be recom-
mended with confidence to patients
with the appropriate indications
(similar to those for arthroplasties of
the knee, hip, and shoulder).
Controversy still remains regard-
ing the timing of shoulder and elbow
replacement in a patient who requires
both. Generally, the more sympto-
matic joint is replaced first.
The indications for minimally
constrained surface-replacement
arthroplasties versus semicon-
strained ones are not clear. At the
present time, loss of bone or liga-
mentous integrity, ankylosis, and
the necessity of soft-tissue releases
are indications for a semiconstrained
prosthesis. The excellent clinical
results with semiconstrained
designs suggest that loosening
might be no more common than
with nonconstrained ones. The theo-
References
1. Morrey BF: Primary degenerative
arthritis of the elbow: Treatment by
ulnohumeral arthroplasty. J Bone Joint
Surg Br 1992;74:409-413.
2. O’Driscoll SW, Morrey BF: Arthroscopy
of the elbow, in Morrey BF (ed): The
Elbow and Its Disorders, 2nd ed. Philadel-
phia: WB Saunders, 1993, pp 120-130.
Vol 1, No 2, Nov/Dec 1993
retical advantage of better preserva-
tion of bone stock with a resurfacing
design is not necessarily true for
elbows. They require more resection
of bone from the ulna and, in some
designs, from the humerus than do
certain semiconstrained designs.
The role of radial-head replacement
in resurfacing designs has never
been determined. Longer-term fol-
low-up will resolve this matter. The
theoretical advantages of a resurfac-
ing design must be considered in
light of the necessity for anatomic
accuracy during insertion to avoid
unbalanced eccentric forces and
moments that can lead to instability
and/or loosening.
The future of TEA is likely to
include modifications to the current
designs of both nonconstrained and
semiconstrained prostheses. Each
will likely continue to have its indi-
cations, with some overlap.
The role of biologic fixation using
a porous coating, such as hydroxy-
apatite, is uncertain. The elbow does
not have a large surface of struc-
turally strong cancellous bone to fix
to such a device, nor to support it
once it is firmly fixed. Further labo-
ratory and clinical research will be
necessary to determine this.
Synovectomy continues to be
used mainly for early stages of
rheumatoid arthritis. There is contro-
versy regarding its success in the
later stages of arthritis and the indi-
cation for arthroplasty versus syn-
ovectomy. In general, the literature
on synovectomy antedates that on
arthroplasty and is from centers
3. Ward WG, Anderson TE: Elbow
arthroscopy in a mostly athletic popula-
tion. J Hand Surg 1993;18A:220-224.
4. Tulp NJA, Winia WPCA: Synovectomy
of the elbow in rheumatoid arthritis:
Long-term results. J Bone Joint Surg Br
1989;71:664-666.
5. Figgie MP, Inglis AE, Mow CS, et al:
Shawn W. O’Driscoll, MD, PhD, FRCS(C)
where arthroplasties have not been
commonly performed on the elbow.
Those surgeons skilled with both
procedures with whom I have dis-
cussed this tend to regard the results
of arthroplasty to be superior in
advanced arthritis. Whether it
should be done by radioactive iso-
tope injection or by arthroscopic or
open techniques is still debated. It
seems wise to offer a trial of isotope
injection, because of its low morbid-
ity, followed, if necessary, by arthro-
scopic synovectomy by those skilled
with this technique. The advantage
of radial-head excision appears to
reside more in the degree of surgical
exposure than in any intrinsic
beneficial effect.
There is also controversy regard-
ing the indications for resection or
interposition arthroplasty versus
TEA in young patients with rheuma-
toid arthritis. Certainly, the former is
more popular in Europe than in
North America, while the opposite is
true for TEA. It is argued that resec-
tion (preserving the epicondyles and
olecranon) is a more conservative
operation that is readily converted
to TEA. However, TEA provides bet-
ter pain relief and function and can
usually be converted to a functional
resection arthroplasty after failure.
Both sides of this argument are
sound, and there is no clear resolu-
tion. I currently favor reserving
resection as a salvage option.
Finally, the role of arthroscopy in
osteoarthritis of the elbow needs
clarification. This will occur as our
skills and experience grow.
Results of reconstruction for failed total
elbow arthroplasty. Clin Orthop
1990;253:123-132.
6. Morrey BF, Adams RA, Bryan RS: Total
replacement for post-traumatic arthritis
of the elbow. J Bone Joint Surg Br 1991;
73:607-612.
7. Morrey BF: Post-traumatic contracture
115
Elbow Arthritis
of the elbow: Operative treatment,
including distraction arthroplasty. J
Bone Joint Surg Am 1990;72:601-618.
8. Kudo H, Iwano K: Total elbow arthro-
plasty with a non-constrained surface-
replacement prosthesis in patients who
have rheumatoid arthritis: A long-term
follow-up study. J Bone Joint Surg Am
1990;72:355-362.
9. Figgie MP, Inglis AE, Mow CS, et al: Sal-
vage of non-union of supracondylar
fracture of the humerus by total elbow
arthroplasty. J Bone Joint Surg Am
1989;71:1058-1065.
10. Goldberg VM, Figgie HE III, Inglis AE,
116
et al: Current concepts review: Total
elbow arthroplasty. J Bone Joint Surg Am
1988;70:778-783.
11. Ewald FC, Simmons ED Jr, Sullivan JA,
et al: Capitellocondylar total elbow
replacement in rheumatoid arthritis:
Long-term results. J Bone Joint Surg Am
1993;75:498-507.
12. Morrey BF, Adams RA: Semicon-
strained arthroplasty for the treatment
of rheumatoid arthritis of the elbow. J
Bone Joint Surg Am 1992;74:479-490.
13. Friedman RJ, Ewald FC: Arthroplasty
of the ipsilateral shoulder and elbow in
patients who have rheumatoid arth-
Journal of the American Academy of Orthopaedic Surgeons
ritis. J Bone Joint Surg Am 1987;69:
661-666.
14. O’Driscoll SW, Bell DF, Morrey BF: Pos-
terolateral rotatory instability of the elbow.
J Bone Joint Surg Am 1991;73:440-446.
15. Morrey BF, Askew LJ, An KN: Strength
function after elbow arthroplasty. Clin
Orthop 1988;234:43-50.
16. O’Driscoll SW, An KN, Korinek S, et al:
Kinematics of semi-constrained total
elbow arthroplasty. J Bone Joint Surg Br
1992;74:297-299.
17. Morrey BF, Bryan RS: Revision total
elbow arthroplasty. J Bone Joint Surg Am
1987;69:523-532.
JAAOS Home Page

Table of Contents

Search

Help
JAAOS Home Page

Table of Contents

Search

Help
 Outcomes Research in Orthopaedics
Robert B. Keller, MD

Abstract

A new agenda in outcomes research has developed in the past decade. The stimu- underlying the outcomes agenda.
lus has come as the result of rapidly increasing health care costs, marked varia- The percentage of gross domestic
tions in utilization of health care services, and deficiencies in the research product spent on health care in the
literature. Outcomes research includes methods such as analysis of large data- United States has risen from 5.2% in
bases, small-area analysis, structured literature reviews (meta-analysis), prospec- 1960 to 14.4% in 1992—the highest
tive clinical trials, decision analysis, and guideline development. Clinical research percentage among the industrialized
should be prospective and should employ modern statistical and assessment meth- nations. In 1990 (the most recent year
ods. The focus of this research is on patient-oriented outcomes of care rather than for which comparable data are avail-
on assessments of the process of care. To illustrate these applications in able), the United States spent 12.2% of
orthopaedics, lumbar spine fusion with internal fixation for “spinal instability” its gross domestic product on health
is presented as an example. Completed large-database analyses, small-area varia- care, compared with 8.5% for
tion studies, and a meta-analysis indicate the need for clinical studies. An outline Canada, 6.3% for Japan, and 8% for
of the form and content of such a study is presented. Germany.1 Various broad measures
J Am Acad Orthop Surg 1993;1:122-129 of health status, such as life
expectancy and infant mortality,
indicate that our extra expenditures
During the past 5 to 10 years a new small-area analysis, meta-analysis, produce no obvious benefit. That is
term has appeared in the medical and decision analysis, have become not to say that the increased expendi-
vocabulary—“outcomes research.” an important part of outcomes tures in the United States do not pro-
The purpose of this article is to define assessment, but clinical research duce higher quality or more effective
and describe this new concept, par- remains the basis of the concept. care. The problem is that we have no
ticularly as it relates to orthopaedic information to prove the point.
surgery. Additionally, by using a
clinical example, the methods under- Factors in Rethinking Practice-Pattern Variations
lying this concept will be clarified. Clinical Research Methods In 1973 Wennberg and Gittel-
What is outcomes research and sohn2 published their first article on
why do we need to be concerned At the outset, we need to understand the subject of variations in practice
about it? Outcomes research can be the factors that have been the stimu- and utilization patterns in medical
simply defined as refined and lus for this major rethinking of the care, which provided a major stimu-
enhanced clinical research. In this clinical research methods we under- lus to more rigorous evaluation of
research there is an important focus on stand and have relied on for so long. clinical practice.
patient-based outcomes as opposed to Several important factors have
measures of process of care. Patient- developed in the past 15 or so years.
based outcomes are assessments that Singly and together, they make it Dr. Keller is Executive Director, Maine Medical
Assessment Foundation, Augusta, Me; Adjunct
measure the results of care as they are clear that we who practice medicine Professor of Surgery and Community and Fam-
perceived by patients. They include need to rethink our current knowl- ily Medicine, Dartmouth Medical School,
factors like pain, function, satisfaction, edge base and how we develop new Hanover, NH; and Associate Professor of Ortho-
and quality of life. Process measures information. pedic Surgery, University of Massachusetts
include such factors as radiographic Medical School, Worcester.
appearance, range of motion, and lab- The Rising Costs of Health Care Reprint requests: Dr. Keller, Maine Medical
oratory results. It seems clear that the dramatic Assessment Foundation, Box 4682, 18 Spruce
Additional new methodologies, increase in health care costs over the Street, Augusta, ME 04330.
such as large-database analysis, past 30 years is the major factor

122 Journal of the American Academy of Orthopaedic Surgeons


Epidemiologists have typically
expressed the incidence of disease in
terms of the rate of occurrence of a
condition (the number of episodes
per 100,000 population). Wennberg
and Gittelsohn applied similar meth-
ods to study the utilization or con-
sumption of health care services.
They further refined the method by
developing “small areas,” geo-
graphic regions surrounding hospi-
tals at which the majority of local
residents receive care. It turned out,
contrary to what one might think,
that there are marked differences in
hospital admission and surgical rates
between small areas within states. As
one looks more widely, there are also
significant differences between
states, regions, and nations. It is also
important to note that all health care
systems, regardless of their organiza-
tion or financing design, demon-
strate this kind of variation.
Within orthopaedics, there are
few conditions that do not show
variations. Hip fracture and multi-
ple trauma are examples of low-vari-
ation conditions. Essentially every
other condition or procedure in the
specialty shows striking variations
in hospital and surgical use rates.3
The conclusion reached by those
who have carried out these studies is
that after careful statistical adjust-
ments for factors such as age and
sex, the wide variations that exist are
not appropriate. If the high rate of
utilization represents the “right
rate,” then those below that level are
being underserved. If the low rate is
correct, then those above it are
receiving excessive care. The prob-
lem is that we do not know what the
so-called right rate is, but it does
seem clear that all the rates cannot be
correct. Outcomes research hopes to
answer this conundrum.
Deficiencies in the Clinical
Literature
The major source of information
for clinicians is the published litera-
Vol 1, No 2, Nov/Dec 1993
ture. Directly or indirectly, almost all
knowledge in orthopaedics is based
on information that has appeared in
journals and texts. Researchers and
investigators write them, teachers
teach from them, students read
them, board examinations are based
on them, and those in practice rely
on them in their daily practice of
orthopaedics. Without the core of
information based in the scientific
literature, we would practice folk
medicine.
In recent years that fundamental
basis of knowledge and learning has
come into question. The questions
come from two sources. First,
authors have critically reviewed cer-
tain areas of the clinical literature
regarding its quality and accuracy.
They have found significant prob-
lems. Gartland 4 and Gross 5 have
both analyzed the literature of hip
arthroplasty. Each found significant
flaws in it. Faulty research design,
erroneous statistical analysis, and a
lack of focus on patient-oriented out-
comes of treatment were noted.
The second source comes from a
new technique of scientific literature
review known as meta-analysis. In
this method, data from many articles
are pooled to form a larger mass of
information for statistical analysis.
Ideally, only randomized trials qual-
ify for meta-analysis, but few of
these have occurred in orthopaedics.
With care, one can broaden the crite-
ria to include other reports. Meta-
analyses have been published for
several orthopaedic conditions,
including hip fracture, lateral
epicondylitis, and lumbar spine
fusion.6-9
The consistent finding in these
reports has been the lack of random-
ized trials, inadequate study design,
lack of standardized definitions and
measures, poor descriptions of
patients, inadequate and unclear fol-
low-up, and little or no evaluation of
patent-oriented outcomes of care.
Indeed, some attempts at meta-
Robert B. Keller, MD
analysis have not been possible
because the available literature is so
weak.7,9
If the literature on which we so
heavily depend has such significant
deficiencies, it is perhaps not sur-
prising that practice-pattern varia-
tions exist. There is simply not a firm
knowledge and research base on
which the clinician can rely in clini-
cal decision making.
Outcomes Research
Methodologies
Outcomes research in its broadest
context involves a number of differ-
ent methods—literature review,
large-database analysis, small-area
analysis, prospective clinical trials,
decision analysis, and development
of clinical guidelines. In the large,
federally funded Patient Outcomes
Research Teams studies,10 essentially
all of these methods are utilized.
However, these techniques may be
used independently. For instance,
meta-analysis is one method within
outcomes research, but this kind of
analysis is often undertaken as an
independent effort.
Literature Review
An important step in all research
is the need to review what is known
about a subject up to the current
time. Ideally, one would carry out a
meta-analysis of the literature for
each and every project.11 The object
of meta-analysis is to gather compa-
rable data from a number of differ-
ent sources and combine those data
to create a larger and more statisti-
cally significant pool of information
for analysis. In each analysis, strict
rules for inclusion and exclusion of
data from different sources must be
developed. Reader bias in selection
and interpretation of articles is thus
avoided. Because meta-analysis is
time consuming and expensive
($30,000 to $50,000 per analysis is not
123
Outcomes Research in Orthopaedics
unusual), and the literature may be
so deficient as to defy a high-quality
meta-analysis, this step may not be
necessary or useful. A “structured
literature review” in which one
applies many of the rules of meta-
analysis may suffice. The more typi-
cal “narrative review,” in which an
author picks and chooses which arti-
cles to quote and emphasize, is sub-
ject to significant bias.
Large-Database Analysis
This method utilizes analyses of
large databases, such as the
Medicare files. It should be noted
that these are primarily claims data,
which may be subject to significant
error and may require great skill to
interpret. Other claims databases
and state-level hospital discharge
data abstracts can also be useful.
From these sources one can carry out
epidemiologic studies and limited
outcomes analyses on factors such as
mortality, length of stay, complica-
tions, and reoperations.12
None of these databases is per-
fect, and in carrying out analyses
and drawing conclusions, analysts
must be experienced and must exer-
cise caution. However, there is a
tremendous amount of valuable
information in them.
Small-Area Analysis
This form of analysis is a method-
ologic subset of large-database
analysis, in that one needs to access
a large database to carry it out.
Small-area analysis is of specific
interest because it demonstrates to
physicians (and others) that there
are significant inconsistencies in
their practice patterns. It serves the
important and useful purpose of
engaging practitioners in the process
of analysis, feedback, research, and
change in practice patterns.
Prospective Clinical Trials
Clinical research should be con-
ducted prospectively, ideally through
124
randomized clinical trials. Recogniz-
ing that it is not always possible to
randomize patients for many kinds of
medical and surgical treatments,
there are several other study designs13
that can reasonably effectively con-
trol for various biases. Retrospective
studies should be avoided. It is
extremely difficult to recover valid
and accurate outcomes information
from records that were not set up for
the purpose of a specific study.
Numerous methodologic problems
can occur.
It is most important to carefully
plan the study so that the hypothe-
ses one wishes to test will, in fact,
be tested. This implies that the
investigators will design proper
data-collection instruments, calcu-
late adequate sample size, plan
careful follow-up protocols for all
patients, collect information rele-
vant to patient-oriented outcomes
of care, and conduct proper statisti-
cal analyses. Patient outcomes
assessment includes categories
such as satisfaction, function, pain,
utility, and quality of life. Evalua-
tion instruments are available to
accurately measure many general
health factors. 14 It remains for the
specialties to develop standardized
and valid instruments for the con-
ditions they treat.
Ideally, outcomes studies should
involve alternative forms of treat-
ment (e.g., a comparison of surgery
and medical treatment for a given
condition). Case-series reports (the
most common in the literature) pro-
vide very biased information
because one never knows how
patients might have fared with
another treatment, or perhaps no
treatment at all.
Decision Analysis
This is a relatively new concept
adapted from the business world.
The statistical results of clinical
research can be translated into a
series of probabilities and placed
Journal of the American Academy of Orthopaedic Surgeons
into an algorithm or decision tree,
enabling one to numerically esti-
mate the likelihood of various treat-
ment outcomes based on patients’
health states, complications, and
specific outcomes. Outcomes can be
weighted according to their desir-
ability (e.g., from perfect health to
death). Combining the probabilities
and the values assigned to various
outcomes can help to determine the
optional strategies that are most
likely to maximize good results.
The analysis may also point out
where critical information is missing
(and research is needed) or which
decisions are most critical in
influencing clinical results. Decision
analysis provides a numerical prob-
ability of a given outcome.
Clinical Guidelines
Guidelines are an important
product that can be developed from
outcomes research. One of the major
problems in developing valid and
useful guidelines is the fact that
accurate information and data to
inform the guideline process have
not been available. Thus, the
deficiencies of clinical research also
restrain the development of guide-
lines. As the results of improved
clinical outcomes research become
available, they can be used to
develop high-quality practice guide-
lines.
A Clinical Case Example:
Lumbar Spine Fusion
To demonstrate the components and
methods involved in outcomes
research, it would be helpful to use a
specific clinical example. I have cho-
sen instrumented lumbar spine
fusion because it represents a new
technology, shows wide variation in
utilization, and is a controversial
procedure.
With the development of several
spinal fixation devices in the past

decade, there has been rapid growth
in the rates of lumbar fusion. The
increased utilization of this proce-
dure has outpaced the population
growth or any known risk factors
that might produce increased
patient need or demand for the pro-
cedure.15 It would appear that the
increase in utilization of the proce-
dure has been driven in part by the
availability of a new technology. The
question remains: Has the availabil-
ity of this new technology improved
patient outcomes in a way that can
justify the increase in utilization?
This clinical situation appears
ideally suited for outcomes research.
In fact, although additional clinical
research is required to establish
more precisely the role of this proce-
dure in patient care, several steps in
the research process have already
been undertaken.
Large-Database Analysis
Analyses of spine fusion rates
across large national regional areas
have been performed. As with all
elective surgical procedures, sig-
nificant variations are seen. Rates of
spine fusion across the four major
regions of the nation have been
determined for the years 1988 to
1990 through analyses of the
National Center for Health Statistics
database. 15 They indicate a 56%
greater likelihood of spine fusion for
Midwesterners than for residents of
the Northeast (Taylor V, Deyo R:
written communication, August
1993) (Fig. 1). An analysis of 1990
fusion rates among residents of the
five largest counties in the state of
Washington reveals a variation of
240% (Fig. 2). In another analysis of
the Washington database, Deyo et
al16 determined that the rate of in-
hospital complications for disk exci-
sion procedures was 5.4%, which
increased to 12.1% when fusion was
combined with diskectomy.17
The only way that these variable
rates would be reasonable is if there
Vol 1, No 2, Nov/Dec 1993
25
20
15 14
15
Rate, %
10
5
0
Northeast West
Fig. 1 Average annual rates of perfor-
mance of lumbar fusion per 100,000 adults
(age- and sex-adjusted to the 1990 US popu-
lation) for the four large geographic regions
of the United States in the period 1988 to
1990. Fusion was performed 180% more fre-
quently in the Midwest than in the North-
east.
were major differences in underly-
ing spine pathology, work condi-
tions, or injury rates, but these
differences were not identified. The
more likely explanation relates to
the differing practice styles of
orthopaedists (and, more recently,
neurosurgeons) as they reflect their
beliefs about the efficacy and effec-
tiveness of this procedure. These
data do not indicate that any one of
these regional or state rates is prefer-
20
15
Percentage
11
10
8
7
5
0 high per capita rates of surgery for
Spokane Snohomish Yakima
Fig. 2 Percentage of patients in the five
most highly populated counties in the state
of Washington who underwent lumbar
spine procedures in 1990 who also under-
went fusion (unpublished data provided by
Victoria M. Taylor, MD, Seattle).
Robert B. Keller, MD
able. Indeed, none of them may be
24
25 the so-called right rate. But it would
be difficult to defend all of them as
being appropriate. There is a strong
implication that there may be
overuse of the procedure in the Mid-
west or underutilization in the
Northeast, or perhaps both.
Small-Area Analysis
We have studied the utilization of
lumbar fusion across the 72 hospital-
South Midwest service areas of Maine, New Hamp-
shire, and Vermont. Each of these
areas contains at least one hospital
and has one or more orthopaedic
surgeons or neurosurgeons practic-
ing within it. While lumbar disk
excision and cervical procedures
vary only minimally among the
three states, the rate of lumbar
fusion across the region varies by a
factor of 3.6 (Fig. 3). Two clusters of
service areas in the three states have
significantly higher (P<.01) utiliza-
tion rates than the rest of the region.
A study group of orthopaedic
surgeons from the three states has
evaluated these data and cannot
explain the variations on the basis of
population, injury, disease, or other
demographic factors. The only obvi-
ous variable is the presence of spine-
fellowship-trained surgeons in those
service areas where the rates are
high. Subspecialty orthopaedists are
located only in the areas with the
highest fusion rates, with two excep-
17
tions. There are fellowship-trained
14 surgeons in the two academic med-
ical centers located in Vermont and
New Hampshire. Because of the
wide referral areas of these centers, it
is possible for spine surgeons to be
busy in their subspecialty without
King Pierce
the populations they treat. However,
this factor is often not the case for
community-based surgeons.
Fellowship-trained spine sur-
geons have the greatest expertise in
this procedure, and one would prop-
erly expect them to perform most of
125
Outcomes Research in Orthopaedics

study lumbar fusion should be based

on what the literature can tell us

,,,,,,,,,
1.60 1.56* 1.60
1.50 1.50 about the procedure. A meta-analy-

,
1.40 1.40
sis of this literature has been pub-
1.30 1.30
lished.8 The authors used standard
Observed-Expected Ratio

1.20 1.20
,,,,,
1.14
1.09
,,,,, meta-analytic techniques in their
1.10 1.08* 1.10
1.02
1.00 ,,,,, ,,,,, 1.00 review. Their conclusions and rec-

,
,,,,, ,,,,, ,,,,,
0.94 0.94
0.89
0.90 0.90
,,,,, ,,,,, ,,,,,

,
,
ommendations are similar to those in
0.80
,,,,, ,,,,, ,,,,, 0.80
0.70 ,,,,,
,,,,, ,,,,,
,,,,, ,,,,,
,,,,,
0.70 other published meta-analyses. Their

,
,,,,,
,
,,,
analysis of the available literature
,,,,, ,,,,, ,,,,,
0.60 0.60
, 0.50
,,,,,
,,,,,
,,,,,
,,,,,
,,,,,
,,,,,
0.50 pertaining to lumbar fusion revealed

,
0.43*
,
0.40 0.40
0.30 ,,,,,
,,,,, ,,,,,
,,,,, ,,,,,
,,,,, 0.30 that there were no randomized clini-

,
,,,,, ,,,,, ,,,,,
,
0.20 0.20 cal trials of the procedure. They
,,,,, ,,,,, ,,,,,
,
0.10 0.10 found an average of 68% satisfactory
,,,,, ,,,,, ,,,,,

,
,
0.00 0.00
Lumbar Disk Lumbar Fusion Cervical Disk results (range, 16% to 95%), a
pseudarthrosis rate of 14%, and a
Fig. 3 Ratios of observed to expected rates for three commonly performed spine procedures rate of painful donor graft sites of
for Maine (solid bars), New Hampshire (hatched bars), and Vermont (dotted bars)(* = P<.01).
If the surgical practice patterns in the three states were similar, the ratios would be 1.00, indi- 9%. The study also indicated similar
cating no variation among the states. There are only minor differences in the utilization of clinical success rates for instru-
lumbar diskectomy and cervical disk surgery; however, there is a 360% greater utilization of mented as opposed to noninstru-
fusion procedures in New Hampshire compared with Maine. (Adapted with permission
from Taylor VM, Deyo RA, Cherkin DC, et al: Low back pain hospitalization: Recent U.S. mented fusions.
trends and regional variations. Spine [in press].) The conclusion of this review is
that better research is urgently
needed on both the effectiveness
(does the technology work when
the fusion surgery in their hospitals. ulation being served (as in the aca- broadly applied at the community
If the provision of this service were demic centers). The converse is also level?) and the appropriateness (is
consistent, and patients from areas true. A surgeon doing a small or the technique being utilized for the
outside the practice locations of the moderate number of procedures proper patients?) of lumbar fusion.
spine specialists were referred to might have a high per capita rate As should now be clear, the same
them, one would anticipate that sur- because the population served is questions can be asked about most
gical rates across the region might be small (as in the northern New Eng- orthopaedic procedures.
fairly level. That is because the uti- land service areas noted).
lization of a service is counted back We are left with the same question
to the area of residence of the raised by the large-database analy- Designing a Prospective
patient. What our data demonstrate ses: which of these rates is the right Clinical Study
is that fusion rates vary according to rate? With small-area analyses, how-
where the experts are in practice. ever, the questions of appropriate- To frame a prospective study, one
Patients who reside in service areas ness of treatment are even more must first develop a hypothesis. It
where spine surgeons are in practice compelling. Why do residents of one would be difficult in one study to
have a much greater likelihood of service area have over three times the evaluate all aspects of lumbar fusion.
undergoing a fusion than those who likelihood of undergoing a lumbar However, it is always desirable to
reside in adjacent service areas. fusion as those in a community 20 evaluate alternative methods of
It should be apparent that one can- miles away? Until these analyses are treatment. One might wish to study a
not draw conclusions from evaluat- undertaken and presented to physi- condition for which fusion may be a
ing the volume of surgery performed cians, they have no idea that the vari- treatment option. Fusion or nonsur-
by an individual practitioner. Only ations exist and how their practice gical treatment for “spinal instabil-
when population-based rates are patterns compare with those of their ity” is an example. One could also
determined can the rate of utilization colleagues in the region. evaluate different fixation devices
of the procedure be calculated. A applied to similar cohorts of patients
given surgeon could perform a large Literature Review to learn whether some are preferable
number of operations but provide a As with all investigations, consid- to others. There are numerous
low rate of those services to the pop- eration of outcomes research to hypotheses that can be generated.

126 Journal of the American Academy of Orthopaedic Surgeons


One can make the case that these
kinds of studies should have been
conducted prior to the wide dissem-
ination of spinal instrumentation
technology, but this pattern of broad
dissemination and utilization of new
technologies is very common, and
the questions still need to be
answered.
It is important to emphasize at
this point that outcomes research is a
team effort. One of the reasons for
the deficiency of the current litera-
ture is that many research efforts
have been carried out without the
benefit of a team approach. One will
need the support of a research
methodologist, a biostatistician, and
perhaps a survey methodologist and
an epidemiologist. A recent review
of methodologies and statistical
methods in the spine literature noted
statistical deficiencies in 54% of
studies and questionable conclu-
sions based on misleading sig-
nificance testing in 46%.17 The need
for expert support in these disci-
plines is clear. For more complex
studies, colleagues such as health
economists, sociologists, and others
may be required. Clinicians are criti-
cal to the research, but they cannot
design and carry out these studies
alone.
Assume that we wish to study the
outcomes of spine fusion for spinal
instability, and we wish to compare
patients who undergo fusion for this
diagnosis with a group who are
treated nonoperatively. The first
step is to find out how many patients
are required in each treatment arm.
That will require the assistance of a
research methodologist who can cal-
culate the number of patients
required to measure meaningful dif-
ferences in outcome—an exercise
known as power analysis.
Next, one must decide how to
select the patients for each treatment
group. If physicians and patients
were completely uncertain about
which treatment is better, it would
Vol 1, No 2, Nov/Dec 1993
be possible to randomize patients
into different treatment groups. As
an example, Herkowitz and Kurz18
have carried out a prospective study
of patients with spinal instability
secondary to degenerative spondy-
lolisthesis. Alternating patients were
assigned prospectively to an instru-
mented fusion group or a nonfusion
group. While alternating is not a
pure form of randomization, this
study does demonstrate the impor-
tant principle of prospectively eval-
uating patients undergoing different
treatments. True randomization is
difficult in most clinical situations
because physicians and patients
may have distinct preferences for a
specific treatment and might there-
fore be uncomfortable with random-
ization.
An alternative method would be
to randomize the physician rather
than the patient.13 In that situation
patients would be randomly
assigned to surgeons, who would
apply the treatments they prefer.
Another design is the cohort
study. In this concept, patients and
physicians arrive at treatment deci-
sions in the usual way. At that point,
patients are enrolled in a prospective
protocol. In the case of spinal insta-
bility, patients who elect to undergo
spine fusion are enrolled in the sur-
gical cohort and those being treated
nonoperatively are entered in the
other cohort. Data are collected
prospectively from both groups.
By carefully collecting patient-
specific information in a cohort
study, it may be possible to stratify
reasonable comparison groups to
contrast the outcomes of the differ-
ent procedures. In some situations,
the two groups may be sufficiently
different in their presenting condi-
tions that comparisons become
impossible. If appropriate data are
carefully collected, analysts will be
able to make this important determi-
nation and indicate which set of
analyses is possible.
Robert B. Keller, MD
Of greatest importance is deter-
mining the kinds of information to
collect. Some of this may be obvious,
but much is not. One of the great
deficiencies in current publications is
that the correct information is not
solicited from patients at the time of
the study. Clinicians generally know
what they would like to learn from
patients, but they frequently do not
have the skill to frame questions in
order to get the information they seek.
In addition, clinicians may not know
what is really important to patients
about the results of their care. Survey
methodologists play an important
role in developing and testing patient
questionnaires. They may need to
interview focus groups of patients
who have the condition or who have
undergone spine surgery in order to
learn what their concerns are.
There should be an emphasis on
patient-oriented outcomes of care.
For example, patients are not partic-
ularly interested in whether they
have a solid spine fusion, but they
are interested in factors such as pain,
function, and quality of life. The
degree of satisfaction and quality of
life is more relevant to patients than
is range of motion or radiographic
evidence of fusion. Certainly, there
is ample evidence that good clinical
outcomes can occur despite failed
fusion, and vice versa. Process mea-
sures such as strength and range of
motion may not be related to out-
comes measures. Often, both process
and outcome need to be evaluated.
An additional problem is that
there are few, if any, standardized
definitions and measurements that
all investigators have agreed to use.
Thus, even if an article contains
valid information, it is difficult to
compare with others. For example,
there is no broadly accepted
definition of spinal instability. There
are various radiographic criteria,19-22
which are felt to be of variable valid-
ity. Others advocate intraoperative
measurements 23 or physical mea-
127
Outcomes Research in Orthopaedics
sures. 24 The point is that none of
these measures has been broadly
accepted and validated. If one can-
not define the condition being stud-
ied, the research effort becomes most
difficult to undertake.
One of the urgent needs in out-
comes research is the creation of high-
quality, standardized, broadly
accepted, validated survey instru-
ments. This single step would
improve the quality of all reports and
make possible meaningful compar-
isons of various treatments and con-
ditions. In part, this issue is being
addressed in the field of low back
pain and lumbar spine surgery. The
North American Spine Society has
supported the development of a
patient-oriented outcomes question-
naire. Its broad adoption and use
across many clinical investigations
will provide a common set of out-
comes information. Thus, investiga-
tors will shortly have available at least
some of the instruments they need to
evaluate the outcomes of lumbar
surgery in a consistent manner, but
much work remains to be done.
Even with adoption of standard-
ized measures, additional data will
be required by specific outcomes
projects. Those comparing the out-
comes of fusion and nonoperative
treatment for instability will need to
collect very specific information
(e.g., fusion rates, implant failures,
surgical and medical complications,
reoperations, and drug reactions)
that might not be part of another
study. The important thing is to uti-
lize tested and broadly accepted
instruments whenever they are
available and to obtain expert assis-
tance in designing and implement-
ing new measures when necessary.
Prospective collection of data is
essential. Only in this manner can the
investigator be sure that all essential
information is collected, that patients
are appropriately categorized, and
that data are collected at consistent
time intervals for every patient. It is
128
very important to attempt to follow
up all patients. If a number of
patients are lost to follow-up, it is
very difficult to draw proper conclu-
sions. For instance, if a large number
of patients with excellent results
from spine fusion fail to return for
follow-up, the results will be biased
in favor of those who do poorly.
One of the problems in analyzing
the outcomes of spine surgery is that
long follow-up is necessary. While
information can be reported at vari-
ous intervals, one must attempt to
carry out long-term studies.
Expert assistance is required in
performing data analyses in out-
comes projects.18 Relatively few clin-
icians have the expertise to
independently conduct the various
analyses and statistical significance
testing. Careful statistical analysis is
a critical step. Given modern statisti-
cal techniques, it may be possible to
carry out manipulations such as
multiple regression analysis and
obtain statistically significant but
clinically meaningless information.
Conversely, clinically important dif-
ferences might be overlooked if sta-
tistical significance is lacking;
high-quality statistical analysis
might be more revealing.
Finally, when information is
reported, the research methods,
patient-group selection process, and
analyses utilized in the study must
be clearly stated so that readers can
clearly understand and extract the
material, and perhaps even attempt
replication of the results. Common
definitions and standardized report-
ing methods will permit comparison
of different techniques and method-
ologies and aggregation of data
across reports.
Conclusions
In considering outcomes research as
applied to spinal instability, we have
been able to describe many of the
Journal of the American Academy of Orthopaedic Surgeons
methodologies of this discipline as
they might apply to a specific
orthopaedic condition and surgical
procedure. In formulating a research
approach to this clinical entity, two
aspects have become clear. First, we
can see that outcomes research is not
markedly different from clinical
research as we know it. The differ-
ences relate primarily to improved
research methodologies and a focus
on patient-oriented outcomes of
care. Second, in considering research
on fusion for spinal instability, we
find that there are major hurdles to
overcome before one can even begin
such an effort. At the outset, there is
no agreement on how to define and
measure the condition referred to as
“spinal instability.”
The issues discussed in this article
put policy makers, patients, and
payers in a position to make a pow-
erful argument: “Demonstrate to us
that this highly variable, very expen-
sive, and complicated surgery for
spinal instability is cost effective and
really makes patients better. If you
cannot, we will no longer pay for it.”
At present, we cannot agree on what
spinal instability is, and there are no
accurate data about patient out-
comes. How can we presume to
know who should undergo this pro-
cedure and justify to payers and
patients the significant expendi-
tures, complications, and uncertain
outcomes associated with this kind
of major surgery?
It thus seems imperative to per-
form careful studies and analyses to
determine whether the entity that
appears to demonstrate radio-
graphic or imaging evidence of
instability is in fact correlated with a
measurable clinical presentation of
pain, other symptoms, and disabil-
ity. Having accomplished that task,
one must then proceed to assess
whether lumbar fusion produces a
better outcome for patients than
might result from other treatment
approaches.

Finally, it should be clear that car-
rying out outcomes research is not
an easy task, but it should also be
evident that there are no real alter-
natives to conducting this kind of
investigation. The urgent challenge
References
1. Health Care Resource Book. Washington,
DC: House Committee on Ways and
Means, 1993. US Government Printing
Office, publication WMCP:103-4.
2. Wennberg J, Gittelsohn A: Small area
variations in health care delivery. Sci-
ence 1973;182:1102-1108.
3. Keller R, Soule DN, Wennberg JE, et al:
Dealing with geographic variations in
the use of hospitals: The experience of
the Maine Medical Assessment Founda-
tion Orthopaedic Study Group. J Bone
Joint Surg Am 1990;72:1286-1293.
4. Gartland JJ: Orthopaedic clinical
research: Deficiencies in experimental
design and determinations of out-
come. J Bone Joint Surg Am 1988;70:
1357-1364.
5. Gross M: A critique of the methodolo-
gies used in clinical studies of hip-joint
arthroplasty published in the English-
language orthopaedic literature. J Bone
Joint Surg Am 1988;70:1364-1371.
6. Lu-Yao GL, Keller RB, Littenberg B, et
al: Outcomes after displaced femoral
neck fractures: A meta-analysis of 106
published reports. J Bone Joint Surg Am
(in press).
7. Labelle H, Guibert R, Joncas J, et al: Lack
of scientific evidence for the treatment
of lateral epicondylitis of the elbow: An
attempted meta-analysis. J Bone Joint
Surg Br 1992;74:646-651.
8. Turner JA, Ersek M, Herron L, et al:
Vol 1, No 2, Nov/Dec 1993
is for orthopaedic surgeons to
become involved in these initiatives.
Acknowledgments: The author gratefully
acknowledges the advice and assistance of
Richard A. Deyo, MD, MPH, and Victoria M.
Patient outcomes after lumbar spinal
fusions. JAMA 1992;268:907-911.
9. Turner JA, Ersek M, Herron L, et al:
Surgery for lumbar spinal stenosis:
Attempted meta-analysis of the litera-
ture. Spine 1992;17:1-8.
10. AHCPR Program Note: Medical treat-
ment effectiveness research. Rockville,
Md: Agency for Health Care Policy and
Research, US Dept of Health, Education,
and Welfare, March 1990.
11. L’Abbé KA, Detsky AS, O’Rourke K:
Meta-analysis in clinical research. Ann
Intern Med 1987;107:224-233.
12. Wennberg JE, Roos N, Sola L, et al: Use
of claims data systems to evaluate
health care outcomes: Mortality and
reoperation following prostatectomy.
JAMA 1987;257:933-936.
13. Rudicel S, Esdaile J: The randomized
clinical trial in orthopaedics: Obligation
or option? J Bone Joint Surg Am 1985;
67:1284-1293.
14. Liang MH, Fossel AH, Larson MG:
Comparisons of five health status
instruments for orthopedic evaluation.
Med Care 1990;28:632-642.
15. Taylor VM, Deyo RA, Cherkin DC, et al:
Low back pain hospitalization: Recent
U.S. trends and regional variations.
Spine (in press).
16. Deyo RA, Cherkin DC, Loeser JD, et al:
Morbidity and mortality in association
with operations on the lumbar spine:
Robert B. Keller, MD
Taylor, MD, MPH, in reviewing this manu-
script and in providing spine surgery data.
Supported by grant No. HS 06344 (The Back
Pain Outcome Assessment Team) and grant
No. HS 06813 (Outcomes Dissemination: The
Maine Study Group Model) from the Agency
for Health Care Policy and Research.
The influence of age, diagnosis, and pro-
cedure. J Bone Joint Surg Am 1992;74:
536-543.
17. Vrbos LA, Lorenz MA, Peabody EH, et
al: Clinical methodologies and inci-
dence of appropriate statistical testing
in orthopaedic spine literature: Are sta-
tistics misleading? Spine 1993;18:
1021-1029.
18. Herkowitz HN, Kurz LT: Degenerative
lumbar spondylolisthesis with spinal
stenosis: A prospective study compar-
ing decompression with decompression
and intertransverse process arthrodesis.
J Bone Joint Surg Am 1991;73:802-808.
19. Dupuis PR, Yong-Hing K, Cassidy JD, et
al: Radiologic diagnosis of degenerative
lumbar spinal instability. Spine
1985;10:262-276.
20. Friberg O: Lumbar instability: A
dynamic approach by traction-compres-
sion radiography. Spine 1987;12:119-129.
21. Stokes AF, Frymoyer JW: Segmental
motion and instability. Spine 1987;12:
688-691.
22. Dvor̆ák J, Panjabi MM, Novotny JE, et
al: Clinical validation of functional
flexion-extension roentgenograms of
the lumbar spine. Spine 1991;16:943-950.
23. Ebara S, Harada T, Hosono N, et al:
Intraoperative measurement of lumbar
spinal instability. Spine 1992;17:S44-S50.
24. Paris SV: Physical signs of instability.
Spine 1985;10:277-279.
129