Should massive rotator cuff tears be reconstructed even when only partially repairable?

PURPOSE: (1) To evaluate midterm functional outcomes of arthroscopic repair of massive rotator cuff tears and (2) to determine the prognostic factors that could influence outcome. The hypothesis was that both partial and complete repairs would result in equivalent improvement of clinical score.

METHODS

From a prospective series of 525 rotator cuff repairs, we analysed records of the 73 patients who were treated for massive tears. The median follow-up was 41 months (range 29-55), and functional outcome was evaluated using the Constant score, shoulder strength, and subjective shoulder value.

RESULTS

The median CS improved from 34 points to 81 points (p < 0.001). The scores were better for both types of two-tendon tears, posterosuperior (83, n = 33) and anterosuperior (85, n = 13) (n.s.), than for three-tendon tears (74, n = 27) (p < 0.001). The scores were also better when fatty infiltration was of stage I (84, n = 28) than of stage II (78, n = 34) (p < 0.001) or stage III (74, n = 11) (p = 0.04). The scores were only slightly higher for completely reparable tears (81.5, n = 50) than for partially reparable tears (79, n = 23) (n.s.). Ultrasonic examination revealed incomplete healing, in 10 of the 50 completely repaired tears, and in 11 of the 23 partially repaired tears.

CONCLUSIONS

The results of the present study compare favourably with those in recent literature and confirm the hypothesis that both partial and complete repairs of massive rotator cuff tears produce equivalent improvements of Constant scores. The clinical relevance of these observations is that even if repairs of two-tendon tears result in superior functional outcomes, repairs of three-tendon tears produce equivalent 'relative' improvement that grants sufficient patient satisfaction and autonomy.

LEVEL OF EVIDENCE: Comparative case series, Level IV.