Luxation récidivante antérieure de l’épaule et rupture de la coiffe des rotateurs : résultats du traitement chirurgical

Abstract

Recurrent anterior dislocations associated with full thickness rotator cuff tear (RCT) carry a difficult therapeutic problem: should we treat instability and rotator cuff tear at the same time or only one of both pathologies? The goal of this study was to analyse a retrospective series of patients operated on to try to answer this question. Twenty-eight shoulders (27 patients) were operated on between 1988 and 2002. The mean age at first dislocation was 47 years (16-65), the average delay between first dislocation and operation was 6.1 years. Twenty-four shoulders presented with recurrent dislocations and four shoulders with recurrent subluxations; the average number of dislocations was 2.6 (1-20). Preoperatively, Hill-Sachs lesion was present in 96%, anterior glenoid rim fracture in 53.5% and glenohumeral osteoarthritis was observed in 37.5%. All the cases had full thickness rotator cuff tears: isolated supraspinatus in 43%,, Supra- plus infraspinatus in 35%, supraspinatus plus subscapularis in 4% and rupture of the three tendons in 18%. An isolated open stabilization with the technique of Trillat was performed in 19 cases when the cuff was not repairable or when the patient was not willing to accept rotator cuff (RC) repair (age and motivation); the mean age of the patients was 59.3 years in this group. Whereas an open anterior stabilization (Latarjet procedure) associated with RC repair was done in nine cases (average age at operation: 40 years). All the patients were followed up and had clinical-radiographic examinations more than two years after the operation. With a mean follow-up of 73.5 months (24-178), the average Constant score progressed from 63.1 to 78.1 points (p<0.05). Three patients who had isolated anterior stabilization had recurrence of instability (16%) whereas none of the patients with both anterior stabilization and RC repair had recurrence. Subjectively, 96% of the patients were satisfied with their operation. Postoperatively, the rate of osteoarthritis progressed to 64.3%. The decision not to repair the RCT in 19 cases was justified by the size of the tear, the muscular fatty infiltration of the RC muscles and the age-motivation of the patients. This decision lead to a greater rate of recurrence (16%) and less satisfactory functional results but the age at FU was 20 years higher in this group than in the group with cuff repair. No patient had an isolated RC repair because 92.5% of the patients in this series had either a bony Bankart (53.5%) or a Bankart type lesion (39%). The recurrent instability in this series was clearly under the dependence of the "anterior mechanism" and not under the dependence of the "posterior mechanism". Therefore, isolated repair of the cuff has never been performed because of the fear of higher rate of postoperative instability leading to RC re-tear. In case of recurrent dislocations associated with rotator cuff tear, treatment of instability should be proposed whereas the concomitant repair of the cuff depends upon the possibility to perform it: size of the rupture, fatty infiltration, age and motivation of the patients.