Patients
Between May 2000 and July 2002, 342 arthroscopic rotator cuff repairs were
performed at our institution. Seventeen patients (4.9%) had an isolated
subscapularis tear and were managed with arthroscopic repair of the tendon by
the senior surgeon (L.L.); these patients constituted the study group. Fifteen
of the seventeen tears in the present study were isolated full-thickness
lesions of the subscapularis. During the period of study, a total of 169
patients were identified with subscapularis tears, including those that were
associated with other torn rotator cuff tendons. Patients with tears involving
any of the other rotator cuff tendons were excluded from the study. The
average interval from the onset of symptoms to the time of surgery was
twenty-four months (range, three to forty-four months). The indication for
surgical repair in this patient population was the identification of a
symptomatic subscapularis tear that presented with pain and weakness after a
traumatic injury or the failure of nonoperative management in patients with
degenerative tendon tears.
The study group included thirteen men and four women who had an average age
of forty-seven years (range, twenty-nine to fifty-nine years) at the time of
surgery. The dominant side was involved in sixteen of the seventeen patients.
Thirteen patients had a traumatic tear that had been caused by combined forced
abduction and external rotation (six patients), direct heavy trauma or a blow
to the shoulder (four), heavy lifting (two), or severe traction on the arm
(one). Six of these thirteen tears were work-related injuries. In the
remaining four patients, symptoms developed progressively over a mean of
fifteen months without a specific history of trauma.
Clinical and Radiographic Evaluation
Each patient in the present series was evaluated by an independent examiner
(Y.R.) preoperatively and postoperatively with use of the Constant and
Murley14 and
University of California at Los Angeles
(UCLA)15 scoring
systems. Patients were also asked to rate their level of satisfaction with the
result as very satisfied, satisfied, or unsatisfied at the time of the most
recent postoperative evaluation. Clinically, the diagnosis of a subscapularis
tendon tear was made with the lift-off and belly-press
tests1. We modified
the belly-press test by pushing against the elbows of the patient
(Fig. 1-A). In order to better
quantify the strength of the subscapularis, we also modified the lift-off test
by asking the patient to push against the hand of the examiner while the arm
of the patient was in maximum internal rotation behind the body
(Fig. 1-B). Pain was graded on
a scale of 0 to 15 according to the scheme from the Constant scoring system,
and strength was assessed on a scale from 0 to 5 according to the
internationally accepted classification of neurological assessment. On the
modified lift-off test, eleven patients had a positive result (meaning that
they were able to perform the maneuver although they had less-than-normal
strength), one patient had a normal result, and five patients were unable to
perform the test because of either pain or stiffness. On the modified
belly-press test, nine of the seventeen patients, including four of the five
patients who were unable to perform the lift-off test, had a positive result
(indicating weakness).
With regard to the biceps tendon, four patients had tenderness over the
bicipital groove or a positive Speed
test16. In
addition, two patients had clinical evidence of biceps rupture with the
so-called Popeye deformity.
Each patient was evaluated with a standard series of plain radiographs
(including an axillary view and true anteroposterior views with the arm in
neutral, internal rotation, and external rotation) before and after surgery.
The presence of glenohumeral osteoarthritis in each patient was classified
according to the criteria of Samilson and Prieto for dislocation
arthropathy17 on
the basis of the anteroposterior radiograph made with the arm in neutral
rotation. None of the patients in the present series had preoperative evidence
of arthritis on plain radiographs. In addition, all seventeen patients had
preoperative and postoperative computed tomographic arthrography studies to
confirm the diagnosis of a subscapularis tear and to evaluate the structural
integrity of the repair after surgery. The computed tomographic arthrography
studies also were used to classify the degree of fatty infiltration according
to the system of Goutallier et
al.18. As was done
in the study by Edwards et
al.2, we simplified
the classification in order to facilitate the analysis of our data by
considering stages 0 and 1 together as minimal fatty infiltration and stage 2
as intermediate fatty infiltration. None of our patients exhibited stage-3 or
4 fatty infiltration. Preoperatively, fifteen of the seventeen patients were
classified as having stage-0 or 1 fatty infiltration and two were classified
as having stage-2 fatty infiltration.
Preoperatively, computed tomographic arthrography was used to further
classify each subscapularis tear according to size into three categories by
dividing the tendon into thirds. Seven patients had a tear involving the
superior third of the tendon, six had a tear involving the superior two-thirds
of the tendon, and four had complete separation of the subscapularis from its
insertion on the lesser tuberosity. None of the patients in the present series
had complete disruption of the subscapularis insertion with an eccentric
humeral head resulting in subcoracoid impingement.
Operative Technique and Intraoperative Evaluation of Subscapularis
Tears
All patients had regional anesthesia with an interscalene block and general
anesthesia. The arthroscopic repair was performed with the patient in the
beach-chair position with the arm in 60° of forward flexion and with the
application of 1.5 to 3 kg of longitudinal traction. We believe that this
position has several advantages over the lateral decubitus position when
performing arthroscopic repair of the subscapularis. First, we agree with the
conclusion of Nove-Josserand et al., who reported that the lateral decubitus
position allows for visualization of only the superior 44% of the
subscapularis tendon rather than the superior two-thirds that can be
visualized with the patient in the beach-chair
position4.
Furthermore, the arthroscopic repair of the subscapularis tendon is often
greatly facilitated by manipulation of the arm into varying degrees of
rotation and flexion (depending on the tear size and configuration), and this
is much easier with the patient in the beach-chair position. Access to the
anterior aspect of the shoulder, an important component of our arthroscopic
subscapularis repair technique, is also greatly facilitated with the patient
in the beach-chair position.
Three to four portals were used, depending on the size of the lesion
(Figs. 2-A and 2-B). A 30°
arthroscope was introduced into the glenohumeral joint through the posterior
"soft-spot" portal. With use of this posterior portal (Portal A),
adequate visualization of tears involving the superior two-thirds of the
subscapularis tendon without retraction could be achieved
(Figs. 2-A and 2-B). However,
if visualization was limited with this portal, then we used a lateral portal
(Portal C) for the arthroscope within the subacromial space in order to
perform the tendon repair (Figs. 2-A and
2-B). Two anterior "working" portals (Portals D and E)
were used to perform the débridement, to place the anchors, and to pass
the sutures through the torn tendon. The anterosuperior portal lateral to the
coracoid (Portal E) was used to pass the sutures through the torn tendon edge,
and an anterolateral portal in the rotator interval (Portal D) was used to
release the subscapularis, to débride the subcoracoid space and the
lesser tuberosity, and to place the suture anchors. In patients with larger
and retracted tears, an anteroinferior portal (Portal F) was used to perform
the subscapularis release circumferentially.
In general, ruptures of the subscapularis were readily identified once a
careful diagnostic arthroscopy was performed. We employed a new classification
scheme in order to further characterize the subscapularis tears
arthroscopically and used this classification system to guide our operative
approach to the repair of these lesions. In the past, subscapularis tears have
been classified as either partial or full thickness and as either isolated or
combined with tears of the supraspinatus
tendon6,7,19.
We propose a new classification system that achieves two primary aims. First,
the classification system accounts for the observation that subscapularis
tears occur from the deep to superficial layers—a phenomenon that has
been recognized through diagnostic arthroscopy. Second, the classification
system guides therapy and differentiates tears that are repairable from those
that we believe are not repairable (full-thickness tears with static anterior
subluxation of the humeral head on the glenoid, resulting in subcoracoid
impingement). According to our classification system, subscapularis tears may
be divided into five types (Table
I). Type-I tears are localized to the superior third of the
subscapularis tendon and are partial tears of the deep fibers at the insertion
onto the lesser tuberosity. These tears never display tendon retraction
because the superficial fibers of the subscapularis remain intact. Two
patients in the present study had this tear pattern. Type-II tears are
complete ruptures that are limited to the superior third of the tendon
(including both the superficial and deep fibers). Four patients in the present
study had this tear pattern. Type-III tears are complete tears of the superior
two-thirds of the subscapularis tendon. The intact inferior one-third of the
tendon limits the degree of retraction that may occur with these lesions.
Seven patients in the present study had this tear pattern, and all seven of
these patients had an intermediate level of tendon retraction between the
lesser tuberosity and the glenoid rim with the arm held in neutral rotation.
Type-IV tears are complete tears of the entire subscapularis tendon from its
insertion, combined with retraction of the tendon edge to the level of the
glenoid rim without anterior eccentricity of the humeral head on the glenoid.
Four patients in the present study had this tear pattern. Type-V tears are
complete tears of the subscapularis with retraction and an eccentric humeral
head that is displaced anteriorly on the glenoid because of disruption of the
force-couple of the rotator
cuff13. None of the
patients in the present study had this type of tear because, in our practice,
such tears are not treated with arthroscopic repair.
Tears of the superior third of the subscapularis were often covered by a
synovial membrane or, less often, by an intact anterior pulley and could be
clearly identified after débridement and positioning of the arm in
flexion and internal
rotation20,21.
More extended ruptures with a retracted tendon were frequently combined with
detachment of the superior glenohumeral and coracohumeral ligaments from their
insertion on the humerus. These ligaments usually were still attached to the
superolateral border of the subscapularis tendon, creating the so-called comma
sign as described by Burkhart and
Tehrany6. Although
the comma sign can aid in localizing the superior border of the torn
subscapularis tendon, none of the patients in the present study had a
supraspinatus tear; therefore, débridement of the lesser tuberosity
from the subacromial view often was required to correctly identify the tendon
edge.
Once the tendon edge was identified, a burr or shaver was used to
decorticate the lesser tuberosity in preparation for anchor placement and to
optimize the environment for tendon healing. In patients with retracted tears,
a circumferential release of the subscapularis tendon was performed with the
shaver and an electrothermal device (Vapor; Mitek, Raynham, Massachusetts) in
order to mobilize the tendon. It should be noted that the intra-articular
release of the subscapularis requires débridement of the middle
glenohumeral ligament from the posterior aspect of the subscapularis.
Superiorly and anteriorly, the subdeltoid and subcoracoid adhesions were
released. In order to improve visualization and to assist in mobilizing the
tendon, we also routinely released the origin of the coracohumeral ligament
from the coracoid process. A traction suture was often placed through the
anterolateral portal in order to facilitate the release of the subscapularis
tendon (Fig. 3). The anatomic
structure limiting the release of the subscapularis tendon in patients with
type-III tears was the axillary nerve. However, in order to mobilize some
type-III and IV subscapularis tears, adhesions between the brachial plexus and
the retracted tendon were released and careful dissection was performed in
order to avoid damaging the two branches of the subscapular nerve on the
anterior surface of the muscle belly. No coracoplasties were performed in this
series because patients with an eccentrically positioned humerus associated
with a retracted subscapularis tear were not included in the study.
Reinsertion of the tendon was performed with metallic anchors (G4 super
anchor; Mitek) loaded with #2 Ethibond sutures. These sutures were passed
through the subscapularis tendon with use of a shuttle-relay technique.
Reconstruction of the footprint always proceeded from the most inferior aspect
of the torn tendon, working proximally. Anchors always were placed along the
anterior border of the bicipital groove in order to achieve an anatomic
footprint repair. In patients with type-III and IV tears, a mattress suture
was used inferiorly and medially and a more simple suture was used laterally
over the footprint in an attempt to perform a double-row suture anchor repair.
We believe that this repair technique maximizes the contact area for tendon
healing and results in a more stable reconstruction.
At the time of surgery, two patients had a rupture of the long head of the
biceps tendon. Of the remaining patients, six had a normal biceps tendon, two
had mild fraying of the biceps tendon, and seven had a partial tear of the
biceps tendon. Nine of the fifteen unruptured biceps tendons with pathologic
changes underwent tenodesis with use of suture anchor fixation (G2; Mitek)
with a simple suture configuration of #2 Ethibond.
Evaluation of the Structural Integrity of the Repair
The structural integrity of the subscapularis repair was evaluated by
analyzing the reconstruction of the tendon on serial axial and sagittal cuts
of the computed tomographic arthrogram. The "water-tightness" of
the repair could not be used as a reliable indicator of the integrity of the
subscapularis reconstruction. The technique described above for release of the
subscapularis tendon for the treatment of full-thickness subscapularis tears
(performed for fifteen of the seventeen patients in this study) requires the
resection of part or all of the rotator interval in order to achieve a
tension-free repair. As such, extravasation of contrast medium is seen
routinely during postoperative computed tomographic arthrography
(Figs. 4-A and 4-B). The axial
and sagittal cuts of the computed tomographic arthrogram were used to
determine if there was any evidence of structural failure of the repairs in
the present study.
Postoperative Rehabilitation
The shoulder was immobilized in an abduction pillow for six weeks after
surgery in order to protect the rotator cuff repair. During this period, only
passive motion was allowed, with internal rotation to the abdomen but not
behind the back. Passive external rotation to neutral and forward flexion
within a range that was not painful for the patient were permitted under the
supervision of a physiotherapist during the first six weeks after surgery.
Unrestricted active-assisted forward flexion and rotation were allowed after
six weeks, and strengthening of the rotator cuff was permitted after three
months.
Statistical Analysis
Statistical analysis of preoperative and postoperative paired data were
done with the Wilcoxon test. The level of significance was set at p < 0.05.
Multiple-regression analysis was used to evaluate the relative influence of
clinical factors (including the age of the patient at the time of surgery, the
mechanism of injury to the subscapularis tendon, the duration of symptoms
before surgery, and the degree of fatty infiltration) on the clinical
outcome.
A total of 169 subscapularis tears (including ten type-I tears, ninety-five
type-II tears, and sixty-four type-III and IV tears) were identified at our
institution during the study period. Therefore, the prevalence of isolated
subscapularis tears in the present series was 10.1% (seventeen of 169).
The sensitivities of the modified lift-off and belly-press tests used in
the present study were 91.7% and 70.6%, respectively. The combination of the
two examinations identified sixteen of the seventeen patients with a
subscapularis tear as four of the five patients who were unable to perform the
lift-off test were able to perform the belly-press test. The computed
tomographic arthrogram had a sensitivity of 94%, with only one patient having
a normal preoperative examination despite having a subscapularis tear that was
subsequently identified and repaired arthroscopically.
At the time of the most recent follow-up evaluation, performed at an
average of twenty-nine months (range, twenty-four to thirty-nine months) after
surgery, twelve patients were very satisfied with the final clinical result,
four patients were satisfied, and one was not satisfied.
The preoperative and postoperative Constant scores are shown in
Table II. The mean Constant
score improved from 52 points preoperatively to 84.9 points postoperatively (p
< 0.001). The mean age and gender-adjusted Constant score improved from 58%
(range, 19% to 80%) preoperatively to 96.4% postoperatively (p < 0.001).
The mean UCLA score also showed significant improvement, from 16.2 points
preoperatively to 32.1 points postoperatively (p < 0.001).
The patients also experienced significant overall improvement in terms of
pain, with the mean pain score improving from 5.9 points preoperatively to
13.5 points postoperatively (p < 0.001). Only one patient rated the
shoulder pain as problematic.
The patients also had marked improvement in terms of objective shoulder
function. The average passive external rotation decreased from 70.9°
preoperatively to 65.3° postoperatively. In our experience, active
external rotation is limited by pain in patients with subscapularis tears and
repair permits a return to full active external rotation. We hypothesize that
the arthroscopic repair of these lesions also results in substantially less
trauma to the surrounding soft-tissue envelope than is experienced after open
repair. As a result, an increase in active range of motion is observed. In the
present study, the average active forward flexion increased from 145.6°
preoperatively to 174.7° postoperatively (range, 150° to 180°) (p
= 0.005). The average active external rotation in adduction improved from
50° (range, 10° to 80°) preoperatively to 60.3° (range,
10° to 80°) postoperatively (p = 0.03). The mean active internal
rotation improved from the level of the sacrum preoperatively to the L1-L2
level postoperatively.
In order to assess subscapularis function, we compared the strength and
pain during the modified lift-off and belly-press tests before and after
surgery. The mean gain in force on the belly-press and lift-off tests
postoperatively was 2 points. We believe that the clinical evaluation can
demonstrate the structural integrity of the repair. Prior to arthroscopic
repair of the subscapularis, five of our patients were unable to perform the
lift-off test because of pain or stiffness that restricted internal rotation
of the shoulder. At the time of the most recent follow-up evaluation, all of
the patients were able to perform this test and to demonstrate improvement in
the force generated during this examination. In four patients, strength
continued to be classified as =3 on a 5-point scale. One of these patients
had a subluxated biceps tendon, and another had a re-rupture of the superior
two-thirds of the subscapularis repair. In the other two patients, the control
computed tomographic arthrogram did not provide evidence of any abnormality
that might explain the postoperative weakness.
The patients in the present series also demonstrated a marked increase on
shoulder strength testing after subscapularis reconstruction. The average
score for strength in abduction improved from 7.4 points (range, 0 to 12
points) before surgery to 15.6 points (range, 6 to 24 points) after
arthroscopic reconstruction (p < 0.001).
Two patients had a structural failure of the subscapularis repair as
documented with computed tomographic arthrography. The initial lesion in one
patient was a rupture of the superior two-thirds of the subscapularis (a
type-III tear according to our classification system), which was retracted to
the glenoid. We suspect that this failure resulted from a technical error as
only one suture anchor was used to reconstruct the subscapularis tendon. The
lift-off test was mildly positive at the time of the most recent follow-up
evaluation as the patient was able to lift the hand off of the back, but there
was decreased strength against resistance as compared with the contralateral
side. The belly-press test was negative, with normal strength as compared with
the contralateral side and no pain. The overall clinical result for this
patient was adequate, with an absolute Constant score of 83 points, a relative
Constant score of 93%, and a UCLA score of 30 points. The computed tomographic
arthrogram showed a rupture of the superior two-thirds of the repair. The
initial lesion in the second patient who had a re-ruptured tendon was a
type-IV subscapularis tear with complete detachment from the insertion,
retraction to the glenoid, and Goutallier stage-2 fatty infiltration. The
postoperative belly-press and lift-off tests were positive. The postoperative
computed tomographic arthrogram demonstrated a retear of the superior
two-thirds of the tendon. At the time of the most recent follow-up evaluation
for this patient, the absolute Constant score was 64 points, the relative
Constant score was 71%, and the UCLA score was 26 points. In the case of one
other patient, the computed tomographic arthrogram demonstrated cleavage of
the superior third of the subscapularis tendon. The cleavage was not
classified as a re-rupture, and the patient had an excellent clinical
result.
There were two failures of biceps tenodesis in our series. One patient who
had a normal biceps at the time of the subscapularis repair subsequently had
development of subluxation of the long head of the biceps tendon without
re-rupture of the subscapularis. We suspect that this may explain the
unsatisfactory clinical result in this patient as reflected in a pain score of
7 points, an absolute Constant score of 54 points, a relative Constant score
of 67.5%, and a UCLA score of 16. There were no infections. Two patients had
development of a complex regional pain syndrome that resolved with medications
and did not result in limited shoulder function at the time of latest
follow-up.
Analysis of the plain radiographs that were made before and after surgery
did not show any progression of osteoarthritis, and there was no evidence of
hardware failure. Critical assessment of the computed tomographic arthrograms
did not show progression of fatty infiltration in any of our patients. In
addition, analysis of the subscapularis reconstruction on computed tomographic
arthrograms demonstrated an intact repair of the tendon in fifteen patients
(Figs. 4-A and 4-B). This
result has been reported to be consistent with complete tendon healing
22.
With the numbers studied, multiple-regression analysis of our data did not
reveal any relationship between age at the time of surgery, the duration of
symptoms prior to surgery, the follow-up period, and the ultimate clinical
outcome. In addition, with the numbers available, Workers' Compensation
status, the etiology of the tear (degenerative or traumatic), the size of the
rupture, and the preoperative status of the biceps tendon did not have
significant influence on the ultimate outcome in our study population. The
impact of re-rupture and persistent fatty degeneration could not be determined
on the basis of the data from our study.
To our knowledge, the present study represents the largest reported series
of patients who have been managed with arthroscopic repair of an isolated
subscapularis tear. It is also the only study in which the structural
integrity of the arthroscopic repair of the subscapularis tendon has been
evaluated on the basis of postoperative computed tomographic arthrograms.
Although several authors have reported on the clinical results of isolated
subscapularis repair, the large variations in terms of the sample size, the
follow-up period, the size of the tendon tear, and the degree of retraction
have made direct comparisons difficult. In our review of the literature, we
found only two studies evaluating the arthroscopic repair of subscapularis
tears. Bennett reported on a series of eight patients with isolated
subscapularis repairs who were followed for a minimum of two
years7. The average
postoperative Constant score was 74 points. However, the clinical evaluation
of subscapularis function with use of either the lift-off or belly-press test
was not reported. In addition, Bennett did not report on the structural
integrity of the repairs postoperatively. As previously reported by Edwards et
al., the strength component of the Constant score does not accurately reflect
subscapularis
dysfunction2.
Rather, it is a measure of overall shoulder strength as a function of the
supraspinatus and deltoid muscles. Burkhart and Tehrany also reported on a
series of eight patients who were managed with arthroscopic repair of an
isolated subscapularis
tear6. Although the
early results were excellent in most of the patients, the short duration of
follow-up (minimum, three months) did not permit definitive conclusions to be
made.
Other investigators have reported satisfactory clinical outcomes after open
repair of isolated subscapularis tears. The largest series, recently described
by Edwards et al., was a multicenter study of eighty-four shoulders that had a
mean postoperative Constant score of 79.5
points2. The authors
found that advanced fatty degeneration negatively influenced subscapularis
function after repair. They also reported that concomitant biceps tenodesis,
regardless of the presence of intraoperatively identified pathological
changes, was associated with improved function and decreased pain after
subscapularis repair. Gerber et al. reported on a series of sixteen patients
who were followed for a minimum of two years after open subscapularis
repair19. The
authors found that eleven patients who had a positive preoperative lift-off
test had normalization of the lift-off test after subscapularis repair. They
also found that a delay in treatment resulted in poorer clinical outcomes. The
average functional score for the shoulder was 82% of the average age and
gender-matched normal value. Interestingly, 18.7% of patients had a deficit of
external rotation of at least 30° and required arthroscopic release in
order to restore function. Deutsch et al. noted similar improvement on the
lift-off test after subscapularis repair as each of the thirteen patients in
that study had a normalized lift-off test at the time of the most recent
follow-up
evaluation3.
One of the primary objectives of the present study was to evaluate the
durability of anatomic reconstruction of the subscapularis. Several studies
evaluating rotator cuff function after open and arthroscopic repair have
demonstrated improved rotator cuff function in patients with intact rotator
cuff
repairs22-25.
Those studies involved repairs of the superior and posterosuperior aspects of
the rotator cuff. However, we have no reason to believe that the functional
implications of an intact rotator cuff repair in patients with subscapularis
tears will not be similar. In fact, the present study supports the hypothesis
that intact repairs will result in improved function since fifteen of the
seventeen patients had an intact repair on the basis of the analysis of
postoperative computed tomographic arthrograms as well as the overall
significant improvements in subscapularis strength and function. This result
compares favorably with that reported by Edwards et al., who reported that
eleven of eighty-four shoulders had clinical or radiographic evidence of a
failed repair2.
Prior to arthroscopic repair, five patients in our series were unable to
perform the lift-off test because of pain or stiffness. Postoperatively, all
patients were able to perform the lift-off test and significant increases in
strength on the modified lift-off and belly-press tests were realized. In
addition, our analysis of the belly-press and lift-off tests supports the
observations made by other authors with regard to the usefulness of these
tests for detecting lesions of the subscapularis clinically as sixteen of the
seventeen tears in the present series were accurately diagnosed with the
combination of these two
maneuvers2,3,6,19.
Unlike the findings of some other studies evaluating subscapularis
reconstruction with use of open techniques, none of our patients experienced
postoperative stiffness after the repair. In fact, all of the patients in the
present study had significant increases in the active ranges of flexion,
external rotation, and internal rotation. In the study by Gerber and
colleagues, three of sixteen patients required arthroscopic release for the
treatment of stiffness after subscapularis
repair19.
Similarly, in the study by Edwards et al., four of eighty-four shoulders had
stiffness after open subscapularis repair, although all four shoulders were
successfully treated with physical
therapy2. We suspect
that the minimally invasive nature of arthroscopic subscapularis repair may
result in less overall trauma to the soft-tissue envelope about the shoulder,
resulting in fewer adhesions and a lower prevalence of postoperative
stiffness.
One of the main weaknesses of the present study relates to the small sample
size of patients. This fact might also explain why significant relationships
between associated conditions we evaluated with multiple regression analysis
were not identified.
Our classification system for subscapularis tears has been designed to be
useful as a tool for diagnosis and as a guide for surgical treatment. As most
previous studies and techniques for evaluating subscapularis tears were based
on principles correlating with open tendon reconstruction, we believe that the
prevalence of partial tears of the subscapularis tendon is underreported. In
addition, current classification schemes do not differentiate between complete
retracted tears of the subscapularis tendon that result in an eccentrically
positioned humeral head and those with a completely congruent glenohumeral
joint1,7.
In our experience, this distinction is important because the outcomes between
these two patient populations differ markedly. The patients with a retracted
subscapularis tendon who have also had development of static anterior
subluxation of the humerus on the glenoid are not candidates for subscapularis
repair and in our opinion should be managed with tendon transfer or
arthroplasty. In conclusion, the results of the present study indicate that
arthroscopic subscapularis repair can result in a durable rotator cuff repair
with clinical results that are at least comparable with those of open repair
techniques. ?