Purpose of the study: Arthroscopic repair of rotator cuff tendons is a well-described technique. Clinical outcome is good. The purpose of this work was to assess tendon healing as observed with the arthroscan after arthroscopic repair of rotator cuff tears. We search for epidemiological, anatomic, and technical factors predictive of healing.

Material and methods: This was a prospective consecutive non-randomized series of 167 shoulders. All patients were assessed preoperatively with the Constant clinical score and had a standard imaging protocol for assessing the status of the rotator cuff. Arthroscopic repair was used in all cases. Outcome was assessed with the Constant score and an arthroscan was available for 148 patients.

Results: Mean patient age was 59 years, 49% male and 77% dominant side. Mean duration of symptoms before repair was nine months. The tear resulted from trauma in 28% and was a work accident in 9%. The preoperative functional assessment was 52.48. An isolated supraspinatus tear was observed in 68% of the shoulders. Frontal retraction of the supraspinatus was distal in 74%. For 29 shoulders, reduction was difficult and the quality of the tendon was considered normal in 56. Non-anatomic repair was required in six cases. The Constant score at last follow-up (19 months) was 80. An arthroscan was available for 148 shoulders: healing was anatomic in 69, defective healing was observed in 27 and repeated tears were noted in 52. Factors predictive of healing were: demographic (lesion duration less than six months, non-dominant side, young patient, female gender, sedentary lifestyle); anatomic (isolated lesion of the supraspinatus with little extension and no retraction); technical (normal easily reduced tendon, good quality bone).

Discussion: The time before surgical management of rotator cuff tears was an important factor in this study. Tendon and muscle alterations after rotator cuff tears could explain, at least in part, healing failures.

Conclusion: This study confirmed the good functional and anatomic results obtained with arthroscopic repair of rotator cuff tears.

Purpose: The purpose of this study was to compare meniscal lesions observed arthroscopically with magnetic resonance imaging (MRI) descriptions.

Material and methods: MRI were read by one operator blinded to the arthroscopic findings. The operator noted the presence or absence of lesions, the topography of the lesion, and the Trillat classification. 188 menisci were studied in 94 patients.

Results: Arthroscopically there were 90 healthy menisci and MRI described 89 healthy menisci, giving an arthroscopic specificity of 98.9%. There were 98 meniscal lesions arthroscopically and 95 at MRI, giving 96.6% sensitivity. The medial meniscus was damaged in 78 cases and the lateral meniscus in 20. Arthroscopy and MRI found the same type of lesion in 62 of 98 cases, identified on the medial meniscus in 47 out of 78 and the lateral meniscus for 15 out of 20. MRI recognised 25/29 horizontal lesions but only one radial lesion of the medial meniscus among the seven cases identified arthroscopically. The topography was concordant in 18/98 cases. MRI oversized the lesion in 54/98 cases.

Discussion: MRI remains the gold standard complementary examination for the diagnosis of meniscal lesions, with high sensitivity and specificity. The precise anatomic description of the lesion provided by the MRI corresponds less well with the arthroscopic findings. Certain lesions appear to be easier to identify, particularly horizontal lesions, while others, particularly radial lesions, are more difficult to demonstrate.