Failure analysis of rotator cuff repair: a comparison of three double-row techniques.

Background: The use of suture anchors has made arthroscopic repair of the torn rotator cuff possible. However, objective evaluations have demonstrated high failure rates. The goal of this study was to compare the modes and rates of failure of two double-row arthroscopic repair techniques and the mini-open double-row technique.

Methods: Thirty pairs of fresh-frozen human shoulders were used in this study. The specimens were prepared to simulate a cuff defect, which was then repaired. The repairs were done with three different lateral row techniques (Mason-Allen sutures passed through transosseous tunnels, the knotless anchor method, and the corkscrew suture anchor method) with the same medial row technique (corkscrew suture anchors). Cyclic tests were conducted at 33 mm/s with a cyclic force of 10 to 180 N. Specimens were cycled to 5000 cycles or to failure as defined as formation of a 10-mm gap at the repair. Failure rates and failure modes of the suture, tendon, and bone-anchor interface were compared for the medial and lateral rows and among the three techniques.

Results: Fourteen of the twenty repairs made with the transosseous technique, fifteen of the twenty repairs made with the knotless anchor technique, and ten of the twenty repairs made with the corkscrew anchor technique survived 5000 cycles. The failure rates for the medial row were not significantly different among the three repair techniques. For the lateral row, there was a significant difference (p < 0.01) in the rate of failure among individual transosseous tunnel-suture complexes (32%), knotless anchor-suture complexes (48%), and corkscrew anchor-suture complexes (75%), with a similar suture-tendon failure rate for all three techniques. The tendon and repair complexes with corkscrew suture anchors had the smallest displacement both at the first and the 5000th cycle.

Conclusions: Although repairs made with the anchor techniques had higher individual failure rates, the survival rates for the anchor techniques at the 5000th cycle were similar to that for the transosseous technique during cyclic tests. Suture failure was the main failure mode for the transosseous technique, whereas failure at the anchor-bone interface was the main failure mode for the anchor techniques.