Patients with brachial plexus birth palsy presenting to the Hand and
Upper Extremity Program at our institution have been prospectively evaluated
since 1993. Patients with persistent impairment of shoulder function have been
treated according to an evolving surgical algorithm utilizing the following
factors: muscle strength, soft-tissue contractures, and underlying
glenohumeral deformity (Fig.
1). All patients who were treated with an operation had an
internal rotation contracture and concomitant external rotation weakness about
the shoulder. Patients with rotator cuff muscle weakness and imbalance,
minimal joint contracture, and no glenohumeral deformity were treated with
latissimus dorsi and teres major tendon transfers to the rotator cuff, as
previously
described13,26.
Patients with a glenohumeral deformity or dislocation (defined as no contact
between the humeral head and the glenoid) that was deemed to be reducible were
treated with open or arthroscopic reduction of the glenohumeral joint, tendon
transfers, and musculotendinous lengthenings. Preoperative imaging studies
(magnetic resonance imaging and/or computed tomography), examination under
anesthesia, and arthroscopic or open exposure of the joint were utilized to
determine the reducibility of the joint. Patients with an irreducible
dislocation were treated with rotational humeral
osteotomy7,9,10,16,17.
The patients in this study had mild-to-moderate glenohumeral dysplasia and
underwent tendon transfers and concomitant extra-articular soft-tissue
releases with the goal of improving function and hopefully halting or
decreasing the glenohumeral dysplasia, although the ultimate effect of these
rebalancing procedures on joint development was unknown. The results in this
cohort of patients are the focus of the current investigation, which was
approved by the Committee on Clinical Investigation of our institution's
internal review board.
Between 1993 and 2003, 122 patients with brachial plexus birth palsy
underwent tendon transfers about the shoulder for the treatment of persistent
impairment of shoulder abduction and external rotation at our institution. In
patients with concurrent joint contractures, musculotendinous lengthenings of
the pectoralis major, subscapularis, coracobrachialis, and/or short head of
the biceps were performed as deemed appropriate on the basis of preoperative
and intraoperative physical examination. Concomitant open or arthroscopic
reduction and capsulorrhaphy of the glenohumeral joint were performed in
patients with a dislocation that was deemed reducible, as described above. The
upper extremity was immobilized in a spica cast for three to six weeks
postoperatively, with the upper extremity positioned in abduction and external
rotation and care taken to place the shoulder in a reduced position. After
discontinuation of the cast immobilization, the patients underwent physical or
occupational therapy consisting of range-of-motion and strengthening
exercises. All procedures were performed by the senior author (P.M.W.).
Of these 122 patients, forty-three underwent tendon transfers and
soft-tissue lengthenings without concomitant reduction of the glenohumeral
joint or osteotomy of the humerus or glenoid. The tendons of the latissimus
dorsi and teres major were sewn directly into the periosteum of the greater
tuberosity and the tendinous portion of the rotator cuff insertion, as
previously
described26. These
patients were assessed with the usual protocol of preoperative and
postoperative radiographic evaluation of both glenohumeral joints with either
magnetic resonance imaging (patients who were four years of age or younger) or
computed tomography (patients who were five years of age or older) to assess
glenohumeral joint
development24.
Twenty-five patients who were followed for a minimum of two years
postoperatively were included in the current study; their clinical and
radiographic findings are presented in the Appendix.
There were fourteen male and eleven female patients. The mean age at the
time of the surgery was forty-two months (range, fourteen to 168 months). Only
three patients were operated on before the age of two years, and two patients
were older than six years at the time of the surgery. The right shoulder was
affected in sixteen patients and the left shoulder, in nine. The mean duration
of radiographic follow-up was forty-three months (range, twenty-four to 121
months).
Glenoid retroversion and humeral head subluxation were measured according
to standardized techniques (Fig.
2), and the overall radiographic appearance of the glenohumeral
joints was classified according to our previously published grading system
(see Appendix)24.
In this paper, positive values indicate glenoid retroversion and negative
values, anteversion. Humeral head subluxation represents the percentage of the
humeral head, measured in its greatest diameter, lying anterior to the axis of
the scapular spine. Previously published reports have confirmed the validity
and reliability of our measurement techniques; for example, glenoid version
measurements have demonstrated a standard deviation of 5° to
7°24,27.
In addition to these radiographic evaluations, all patients underwent
extensive preoperative and postoperative clinical assessments. The
upper-extremity functions and specifically the active shoulder movement of all
twenty-five patients were classified with the modified Mallet system (see
Appendix)28,29.
The intraobserver and interobserver reliability of this classification system
has been previously
established30.
Aggregate Mallet scores were calculated by summing the individual elements
(maximum score, 25 points).
Comparisons of means for continuous variables were performed with use of
the paired Student t test, with Microsoft Excel 2000 software (Microsoft,
Redmond, Washington). P values were two-tailed, and p values of <0.05 were
considered to be significant.
While the majority of infants with brachial plexus birth palsy
demonstrate spontaneous recovery and attain nearly normal function, some have
persistent neurological deficits resulting in long-term upper-extremity
dysfunction3,5,6.
In addition to causing difficulties with active upper-extremity motion,
persistent muscle weakness and imbalance were found to affect bone growth and
joint development in previous cross-sectional and longitudinal
studies18-24,31,32.
Glenoid retroversion and hypoplasia, posterior subluxation and flattening of
the humeral head, inferior hooking of the coracoid, and clavicular shortening
have all been observed in patients with chronic brachial plexus birth
palsy18-24,31,32.
It has been hypothesized that, as a result of muscular weakness and
imbalance, eccentric forces are imparted across the developing glenohumeral
joint, leading to abnormal joint formation. These changes are progressive and
may lead to functional
limitations23,24.
Ultimately, the degree of glenohumeral deformity influences the choice and
success of secondary reconstructive procedures, as soft-tissue procedures may
result in a suboptimal outcome when there is underlying osseous deformity.
In the growing child, restoration of more normal joint forces may bring
about joint remodeling. While this principle has been extensively studied and
applied in the treatment of developmental dysplasia of the hip, less is known
about its application in the shoulder and particularly in patients with
brachial plexus birth
palsy33-35.
In a prospective analysis by Hui and Torode, twenty-three patients with
brachial plexus birth palsy and posterior glenohumeral subluxation or
dislocation underwent open reduction of the glenohumeral joint and
musculotendinous lengthenings at a mean age of 2.4
years25. At the
time of follow-up, at a mean of 3.5 years, there was improvement in the mean
glenoid retroversion of the affected shoulder. The side-to-side difference in
glenoid retroversion similarly decreased over time.
To our knowledge, however, there has been no report on the effects of
rebalancing the dynamic muscular forces across dysplastic
glenohumeral joints with extra-articular procedures. In the present
investigation, we attempted to determine whether restoring more normal forces
across the developing shoulder retarded and/or reversed the progression of
glenohumeral deformity in patients with brachial plexus birth palsy.
Our results demonstrate that, in appropriately selected patients,
latissimus dorsi and teres major tendon transfers to the rotator cuff combined
with musculotendinous lengthenings not only improve upper extremity function,
but also halt the progression of glenohumeral dysplasia. In the current
investigation, mean glenoid retroversion improved minimally, from 22° to
16.5°, following muscle rebalancing and softtissue releases. Humeral head
subluxation was also improved minimally. Thus, any improvements in
glenohumeral dysplasia were modest at best.
It should be noted that, although we found a mean improvement of 5.4°
in glenoid version following soft-tissue rebalancing procedures, some of this
improvement may reflect normal changes in glenoid morphology that occur with
growth. Glenoid version varies with age. During normal growth and development,
mean glenoid version progresses from 7° of retroversion at birth to 1°
by ten years of
age36. Thus, it is
difficult to conclude that the improvement in glenoid version in the affected
shoulders in our series was due to the soft-tissue rebalancing.
More detailed examination of our results reveals several trends. First,
progressive glenohumeral joint deformity may be arrested if appropriate
surgical procedures are performed when the patient is relatively young and the
shoulder is in an early stage of development. While only three patients were
operated on before the age of two years, it should be emphasized that the mean
age of the patients was forty-two months. Given that only two of the patients
were older than seventy-two months of age, and neither of them demonstrated
improvement in the radiographic appearance of the glenohumeral joint, the
results of this study may not be applicable to patients of all ages.
Additional analysis and longerterm follow-up is needed to determine the
natural history and results of tendon transfer surgery in older patients with
brachial plexus birth palsy.
Similarly, it should be noted that no patients in this study had a type-5
or more severe glenohumeral deformity. For this reason, no conclusions can be
made regarding the effect of dynamic muscle-rebalancing on glenohumeral
development in the presence of severe osseous deformity. Furthermore, while
glenohumeral joint deformity may not progress following tendon transfers
performed in patients with brachial plexus birth palsy, it is still unclear
whether these procedures reliably reverse pre-existing glenohumeral dysplasia.
Our early results suggest that no profound remodeling can be expected. We
believe that, biomechanically, the glenohumeral joint is not subjected to the
same frequency or magnitude of forces as are the weight-bearing joints of the
lower extremity, and for this reason the glenohumeral joint is not as
predisposed to articular surface remodeling.
On the basis of these observations, we believe that surgical intervention
in the form of tendon transfers with or without soft-tissue releases should be
performed in patients with early radiographic evidence of glenohumeral
dysplasia (mild deformity) to halt progressive deformity and improve shoulder
function. Pediatricians, neurologists, and physical and occupational
therapists must be aware of the risk of shoulder dysplasia and its potential
sequelae. Appropriate radiographic evaluation and, when indicated, surgical
treatment should be performed before glenohumeral dysplasia becomes too
advanced.
Future efforts should be directed toward the evaluation of a larger number
of patients with longer follow-up. As greater numbers of patients are
assessed, analysis may reveal specific risk factors for progressive
glenohumeral joint deformity following tendon transfers. In addition, it is
necessary to analyze patients with mild-to-moderate glenohumeral deformity who
have undergone both tendon transfers and arthroscopic or open joint reduction
to determine the utility of these intra-articular procedures in altering the
natural history of glenohumeral dysplasia. Finally, long-term outcomes must be
assessed to determine if glenohumeral deformity, even when treated early and
arrested, portends a worse prognosis in terms of upper-extremity dysfunction,
joint instability, or arthrosis. A multicenter prospective study is currently
under way to address these important issues.
In summary, our results demonstrate that latissimus dorsi and teres major
tendon transfers to the rotator cuff, combined with appropriate
extra-articular musculotendinous lengthenings, may halt the radiographic
progression of glenohumeral deformity in young patients with brachial plexus
birth palsy, although they do not lead to significant joint remodeling.
Function, however, significantly improves following these procedures.
A schematic representation of the modified Mallet classification, a table
showing the preoperative and postoperative radiographic data on all patients,
and a table presenting the classification of glenohumeral deformity in
brachial plexus palsy are available with the electronic versions of this
article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?