Prediction of primary reparability of massive tears of the rotator cuff on preoperative magnetic resonance imaging

We studied magnetic resonance images of rotator cuff tears to determine whether it was possible to establish preoperatively the feasibility or infeasibility of primary repair. The study comprised 27 shoulders in 26 patients who underwent magnetic resonance imaging before surgery because of complete tears of-the rotator cuff and who were treated with primary repair or by the patch graft technique because primary repair was not feasible. The length and width of each tear, the thickness of the supraspinatus muscle at the superior margin of the glenoid, and the presence or absence of a high signal intensity in the infraspinatus muscle were statistically analyzed. Primary repair was often not feasible when both the length and width of the tear exceeded 40 mm on a preoperative magnetic resonance image, when the supraspinatus muscle was thin at the superior margin of the glenoid, and when a high signal intensity was observed in the infrospinatus muscle.