 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1999 The Journal of Bone and Joint Surgery, Inc. Transfer of the Latissimus Dorsi Muscle After Failed Repair of a Massive Tear of the Rotator Cuff. A two to five-year Review* , TORONTO , LONDON, ONTARIO, CANADAInvestigation performed at The Toronto Hospital, Western Division, University of Toronto, Toronto, and the London Health Sciences Centre, University of Western Ontario, London
Introduction: Seventeen patients with an average age of fifty-five years (range, thirty-two to seventy-seven years) who had ongoing pain and impaired function following failed operative treatment of a massive tear of the rotator cuff were managed with a transfer of the latissimus dorsi muscle as a salvage operation. Methods: The patients were examined at an average of fifty-one months (range, twenty-four to seventy-two months) after the operation. Pain, function, and satisfaction were assessed with use of a questionnaire, visual analog and ordinal scales, physical examination, and the University of California at Los Angeles shoulder score. Results: Fourteen of the seventeen patients were found to have significant relief of pain (p < 0.0001) and a significant improvement in function (p < 0.001 for all activities except lifting more than fifteen pounds [6.8 kilograms], for which the p value was <0.0036) and were satisfied with the result of the operative procedure. Fifteen patients stated that they would have the operative procedure again under similar circumstances. Seven of eight patients with a detached or nonfunctional anterior portion of the deltoid had substantial improvement. Three operations were classified as failures because the patients were not satisfied with the result and had ongoing pain and impaired function. All three failures were in patients who had a work-related injury. Overall, six patients had a work-related injury, and only three of them had a satisfactory result. There were three complications, all related to contracture of a hypertrophic axillary scar. Conclusions: The results in this series indicate that transfer of the latissimus dorsi muscle is a reasonable approach for salvage after failed operative treatment of a massive tear of the rotator cuff.
Tears of the rotator cuff are common orthopaedic problems that may cause pain and impaired function. Massive tears increase the complexity of the problem, and management of these lesions remains a somewhat controversial area. Many operative solutions have been proposed2,3,5-8,10-15,17-21,23-26. When these approaches fail, the situation is difficult, with few available options for treatment. Reasons for failure include inadequate decompression4,9,22, failure of repair4,9,16,22, and detachment of the deltoid with impaired function4,16,22. A reoperation after these failures does not necessarily relieve pain4,9,16,22 or improve function, especially if the deltoid is one of the causes of failure. For these reasons, we have searched for operative techniques that could potentially salvage those situations. One such repair is the transfer of the latissimus dorsi muscle, as described by Gerber et al.12. The objective of the present retrospective study was to analyze one surgeon's experience with transfer of the latissimus dorsi muscle to manage patients in whom operative treatment of a massive defect of the rotator cuff had failed. We evaluated the results at a minimum of two years after the operative procedure.
The retrospective study included seventeen patients who had been managed between January 1990 and October 1995 by one of us (A. M.). The patient's history; the results of clinical examination; preoperative radiographs; and subjective questionnaires documenting preoperative and postoperative pain, function, and satisfaction were analyzed. Twelve patients were male and five were female. Ten patients had involvement of the dominant extremity. Eleven patients had had one previous operative procedure, four had had two, and two had had at least three. The anterior portion of the deltoid was thought to be detached or nonfunctional in eight of the seventeen patients, and there was clinical evidence of disruption of the subscapularis in four of the seventeen patients. All patients had preoperative pain and an objective lack of active motion in forward elevation, abduction, and external rotation. All described an inability to use the upper extremity to perform activities of daily living and work. Clinical examination revealed positive impingement signs and weakness of external rotation and on supraspinatus testing in all patients. Six of the patients had a work-related injury and were collecting Workers' Compensation benefits. Preoperative radiographic examination demonstrated proximal humeral migration in all patients, mild degenerative glenohumeral arthritis with sclerosis of the glenoid and the greater tuberosity in seven, and joint-space narrowing with an osteophytic spur in one. Nine patients had no obvious evidence of radiographic osteoarthritis at the glenohumeral joint. Every patient had an initial exploration of the deltoid, the acromion, and the subacromial space. If the failure of the operation was believed to be due solely to inadequate decompression or failure of repair, or both, a repeat decompression or a repeat repair and decompression was performed and the patient did not have a latissimus dorsi transfer. The present series includes only patients who had previously had an adequate operation and did not have tissue available for a repeat repair of the cuff. The four patients with absence, attenuation, or a tear of the subscapularis all had a positive lift-off test. At the time of the index operation, two of these four patients had complete absence of the subscapularis with no identifiable tissue present. These patients had the transfer of the latissimus dorsi muscle at that time despite previous literature suggesting that the absence of the subscapularis is a relative contraindication to the procedure12. In these patients, the latissimus dorsi was anchored to bone troughs both laterally and anteriorly along the neck of the humerus and medially to any remnant of the rotator cuff that was available. The other two patients with a positive lift-off test did have some soft tissue remaining anteriorly in the position of the subscapularis. One patient had what we considered to be a partial tear of the superior portion of the subscapularis tendon. The tendon was thickened and frayed. The remaining tissue was adequate, and the latissimus dorsi could be transferred to it. The fourth patient had what appeared to be a thin atrophic subscapularis. We were not sure about its function, but it was adequate to hold the sutures for the muscle transfer. In eight of the seventeen patients, anterior deltoid weakness was evident on physical examination and was believed to be related to detachment or impaired function, or both. These patients also demonstrated severe anterior and superior migration of the humeral head, which was easily seen and palpated in the subacromial region. In each patient, the tissue attached to the acromion was found to be thin and atrophic at the time of the operation and attempts were made to advance healthier tissue. Function, pain, and satisfaction were evaluated clinically with use of both ordinal and visual analog scales. The overall status of each shoulder was assessed with use of the University of California at Los Angeles shoulder score1, which is also used to evaluate pain, function, range of motion, and strength. Function was evaluated by determining the degree of difficulty that the patient had with the performance of ten different activities (hair-combing, dressing, perineal care, sleeping on the shoulder, washing under the contralateral arm, pulling ten pounds [4.5 kilograms] or more, using the hand overhead or at shoulder level, working or performing daily activities, and lifting fifteen pounds [6.8 kilograms] or more). The degree of difficulty with these activities was rated on a 4-point scale, with 0 points indicating that the activity was not difficult; 1 point, that it was slightly difficult; 2 points, that it was moderately difficult; 3 points, that it was very difficult but not impossible; and 4 points, that it was impossible. Pain at work, at rest, and at night were also evaluated with use of a 4-point scale, with 0 points indicating no pain; 1 point, occasional mild pain; 2 points, frequent mild-to-moderate pain; 3 points, moderate-to-severe pain; and 4 points, severe pain. In addition, the patient indicated the degree of pain and function on a visual analog scale, which was an open-ended, unmarked 100-millimeter line. Finally, overall satisfaction with the shoulder and with the procedure was assessed with use of a questionnaire as well as a visual analog scale. Data were analyzed statistically with the Student t test for paired data and multiple analysis of variance as indicated. Analysis of variance was used to determine whether there was a significant difference between the group of patients who had weakness of the deltoid and the group who had a normal deltoid. The means and standard deviations for all of the measured parameters were analyzed.
Operative Procedure
The latissimus dorsi was dissected free, with the surgeon taking care not to injure the neurovascular structures on its undersurface, which come from proximal to distal and can be damaged. Division of these structures will devitalize the tendon transfer. The tendon was followed to its osseous attachment on the humerus. It is imperative that the tendon attachment to the humerus be visualized to reduce the risk of neurovascular injury and to improve the probability of obtaining tendon tissue of adequate width and length. Maximum internal rotation helps the surgeon to visualize the attachment of the tendon to the humerus. The surgeon must be careful when detaching the tendon. The radial nerve lies just distal to the tendon and should at least be palpated or visualized. If the surgeon is not satisfied with the latissimus dorsi tendon, the teres major can also be obtained at this point (Fig. 3). Both tendons and muscles were obtained in one patient in this series because we thought that the size of the defect and the latissimus tendon were mismatched.
Once the tendon had been detached it was retracted, the undersurface of the latissimus dorsi was visualized, and the neurovascular pedicle was identified and dissected free. The muscle was mobilized by operative dissection as close to the muscle origin as necessary to ensure adequate proximal migration. The tendon was prepared by using two number-1 nonabsorbable sutures of different colors, with one color running down each edge of the tendon. The use of different colors helped to ensure that the tendon was not twisted during passage. The interval between the posterior portion of the deltoid and the long head of the triceps was then identified and dissected. With the two muscles retracted, the posterior aspect of the acromion was palpated and a long curved clamp was passed from anterior to posterior under the acromion, exiting between the posterior portion of the deltoid and the long head of the triceps. Spreading the jaws of the clamp fashioned the passageway that allowed excursion of the tendon so that the tendon could be passed anteriorly. After passage of the tendon, its lateral edge was sutured through drill-holes into a bone trough in the greater tuberosity, in a manner similar to that described for standard repair of the rotator cuff. Medially, the tendon was attached to any remaining mobile cuff tissue. Anteriorly, the leading edge of the latissimus dorsi was sutured to the free edge of the subscapularis muscle (Fig. 4). Alternatively, when the subscapularis was totally absent, an anterior bone trough was used to anchor the leading edge of the tendon. The shoulder was moved through a full range of motion to ensure that there was not excessive tension on the repair. A suction drain was placed posteriorly to eliminate the dead space for twenty-four to forty-eight hours. The wound was closed, and bulky compressive dressings were applied.
Postoperative Protocol
The seventeen patients were evaluated at a minimum of two years (average, fifty-one months; range, twenty-four to seventy-two months) after the latissimus dorsi transfer was done following failure of treatment of a massive tear of the rotator cuff. The average age at the time of follow-up was fifty-five years (range, thirty-two to seventy-seven years). Three of the six patients who had a work-related injury had a failure, and these were the only failures in the series. Two of those patients stated that they would not have the procedure again under similar circumstances, but the third stated that he would still choose to have the operation. This patient noted some subjective improvement, but this did not translate into an improvement in the outcome measures that were used to evaluate the results in this study. Pain: As demonstrated on the ordinal scale, all but three patients had significant relief of pain at work, at rest, and at night (p < 0.0001 for all three) (Table I). The score according to the visual analog scale also improved significantly (p < 0.0001), from an average of 86.1 millimeters preoperatively to an average of 26.0 millimeters postoperatively. The operation was considered to have failed in the three patients who had little pain relief, as the primary purpose of the operation was to alleviate pain.
Function: Interestingly, function improved for all of the activities rated with the ordinal scale (p < 0.001 for all except lifting fifteen pounds [6.8 kilograms] or more, for which the p value was <0.0036) (Table I). The patients' subjective assessment of function on the visual analog scale improved significantly (p < 0.0001), from an average of 13.8 millimeters preoperatively to an average of 56.1 millimeters after the operation (Table I). In addition, both the active and the passive range of motion improved in forward elevation and in internal and external rotation (p < 0.0001) (Table I). Overall shoulder score: The University of California at Los Angeles shoulder score1, which was used to provide a general overall assessment of the shoulder, also improved significantly (p < 0.0001), from an average of 6.8 points preoperatively to an average of 16.4 points postoperatively (Table I). Patient satisfaction: Overall, fourteen of the seventeen patients were satisfied with the improvement following the operative procedure, and this was reflected by the improvement (from an average of 10.9 millimeters preoperatively to an average of 68.5 millimeters postoperatively; p < 0.0001) in the overall satisfaction with the shoulder as indicated on the visual analog scale. All fourteen of the patients who were satisfied with the improvement of the shoulder stated unequivocally that they would have the procedure again under similar circumstances, although a few expressed concern over the duration of recovery and the time needed for rehabilitation. Success of the procedure: On the basis of these assessments, fourteen of the seventeen patients were deemed to have had a successful operative procedure, whereas the procedure was considered to have failed in three. Comparison of results for patients who had an intact deltoid with those for patients in whom the deltoid was not intact. The patients were divided into two groups according to whether or not the deltoid was intact. Eight patients were believed to have either a detached or an attenuated deltoid, and nine had an intact deltoid. With the numbers available, we could not detect any significant differences, either preoperatively or post-operatively, between the groups with regard to pain, function, range of motion, University of California at Los Angeles shoulder score, or the overall satisfaction with the shoulder. This finding suggested to us that the patients had similar amounts of disability before the operation and had similar levels of success after the procedure. Complications: No operative infections or neurovascular injuries were encountered in this series. Three patients had contracture of the incision in the axilla with hypertrophic scarring. All three were treated with injections of corticosteroids, and one patient had limitation of abduction and forward elevation related to the contracture. The scar for this patient was subsequently revised with a z-plasty in the axilla, which allowed the patient more motion. The three patients who had hypertrophic scarring were among the first six to be managed, and because of this complication the remaining eleven patients had a modified z-plasty-type incision through the axilla. Hypertrophic axillary scarring has not been a problem since.
The literature dealing with revision after failed attempts at rotator cuff repair is sparse. Various authors4,9,22 have suggested that the cause of failure must be determined before revision is attempted. The usual causes of failure include inadequate decompression, failure of repair, and detachment of the deltoid with impaired function. Nevertheless, reoperations to treat these problems do not necessarily guarantee success. Both Bigliani et al.4 and Neviaser and Neviaser22 noted some relief of pain, not necessarily associated with improved function, following repeat attempts at repair. DeOrio and Cofield9 reported that, of twenty-four patients who were available for follow-up after management with a second attempt at repair of a failed repair, fourteen (58 percent) had a poor result. Patients who have a detached deltoid and an irreparable defect probably have an even poorer prognosis. The present series represents a subset of patients in whom no local tissue remained and previous attempts at operative treatment had failed to provide pain relief or improved function. The transfer of the latissimus dorsi muscle in these patients was effective. Pain relief, the primary goal of this operative procedure, was achieved in fourteen of the seventeen patients. The range of motion and function also improved, a finding that had not been totally expected. Despite the significant improvement in function, none of these patients would be considered as having normal or nearly normal function. The patients' subjective rating of function on the visual analog scale, although significantly improved from the preoperative rating, was only 56.1 millimeters after the operation. The University of California at Los Angeles shoulder score1, which was used in this series to provide an overall impression of the status of the shoulder, improved to an average of only 16.4 points, which would, at best, be considered fair. This reflects the fact that the patients were still moderately disabled. The most important finding, however, was that there was a significant overall improvement between the preoperative and postoperative University of California at Los Angeles shoulder scores. The subset of eight patients in whom the anterior portion of the deltoid was detached or nonfunctional had substantial improvement. Seven of the eight patients had relief of pain, had improvement with regard to function and the range of motion, and were quite satisfied with the result of the procedure. Our subjective examination revealed that the patients eventually were able to raise the arm, and, although superior subluxation was not totally eliminated, the patients seemed to be able to control the humerus from subluxating superiorly when the arm was in forward elevation. Preoperatively, our subjective impression was that the patients with detachment of the deltoid had more difficulty with activities of daily living and had a smaller active range of motion than the patients in whom the deltoid was intact. However, this impression could not be demonstrated statistically. Detachment of the deltoid as well as a massive irreparable defect in the rotator cuff did not seem to preclude the use of the latissimus dorsi transfer, as both groups (those with and those without detachment of the deltoid) seemed to have marked improvement after the operation (Table II). As previous reports4, 9 have failed to show improvement following a repeat operative attempt at repair of the rotator cuff in patients with an abnormal deltoid, we were encouraged by our results.
It is difficult to determine the true effect of the transferred latissimus dorsi tendon. Some might argue that soft-tissue interposition may have played a role in the reduction of the pain. Although this may be true, the tendon itself is quite thin, and it is difficult to imagine that soft-tissue interposition was the only reason for the improvement. It may be that some of the positive effect of the transfer was either due to active muscle contraction or simply due to a tethering or tenodesis effect of the transferred tendon. Clinically, it appeared that the latissimus dorsi was contracting when it was palpated during elevation of the shoulder. Whether the tendon acts as a passive restraint or an active muscle transfer, the humeral head is centered better in the glenoid and is maintained there, allowing the deltoid better mechanical advantage for elevation of the arm. Four of the patients in the present series had absence, attenuation, or a tear of the subscapularis. It is not possible to draw significant conclusions on the basis of four patients; however, we can report that, although technically the repair was difficult, the latissimus dorsi transfer was possible as the tendon was anchored anteriorly onto the humerus or into local surrounding tissue, which provided good fixation of the tendon. Our impression is that even patients in whom the subscapularis is absent can benefit from this procedure and, therefore, we do not consider absence of the subscapularis to be a contraindication. One of the intriguing findings in this study was the trend of poor results for patients receiving Workers' Compensation benefits. All three of the failures were in this subgroup: three of the six patients who were receiving such compensation had an unsatisfactory outcome. This finding, as always, is difficult to explain as there was documentable evidence of a serious lesion and it was made very clear before the procedure that we did not expect the patients to obtain enough improvement to be able to return to work. Whatever the reason, this group posed the greatest challenge, and there were no apparent distinguishing features to explain the failures. Fortunately, complications were not common, although infection or neurovascular injury could be a problem with this procedure because of the extent of the operative dissection and the proximity of vital structures to the latissimus dorsi tendon. The one problem that was noted was related to the skin incision, which we initially made across the posterior axillary fold. This resulted in three hypertrophic scar contractures at the axilla in the first six patients managed with this type of incision. One patient needed a revision of the operative scar to correct the problem as it caused discomfort and limited motion. Subsequently, the incision was modified to a z-type axillary incision, and to date this has eliminated the problem. There were weaknesses in this study. The group was not totally homogeneous, we performed a retrospective review, and some minor variations in the operative procedure were made; however, the group consisted of patients for whom, it was believed, there were no other treatment options and who represented what we thought to be the worst-case scenario. All patients were operated on and followed by one surgeon. We believe that the outcomes, although only fair, were quite favorable when compared with the preoperative status. We did not include in the present study any primary repairs of massive defects of the rotator cuff, a procedure that was suggested by Gerber et al.12. As a result of the favorable outcomes in this series, we now also consider using the latissimus dorsi transfer as the primary procedure for the treatment of an irreparable massive defect. Of course, that situation is rare, and decompression and repair of the rotator cuff with local tissue remains our first choice for the treatment of massive defects. In conclusion, transfer of the latissimus dorsi muscle for salvage after a failed attempt at operative repair of a massive defect of the rotator cuff is an effective procedure. Our patients reported marked relief of pain and improvement of function, although the shoulders were by no means normal or nearly normal. The procedure seems possible even when the subscapularis is weak or can no longer be identified. In addition, patients with the combined problem of a massive defect of the rotator cuff with anterosuperior subluxation and a detached anterior portion of the deltoid have a favorable outcome after this procedure. We suggest that transfer of the latissimus dorsi be used as an alternative operation in difficult situations in which a patient with a massive defect of the rotator cuff has had a failure of a previous procedure.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |