Background: Most rotator cuff surgery is performed on chronic tears. As there is no animal model in which to examine the physiology of muscle and tendon injury and repair in this setting, we developed a chronic rotator cuff injury model in sheep.

Methods: The infraspinatus tendon was released in thirty-six female sheep. Biopsy specimens were obtained from the muscle and were analyzed for fat content. The force generated by the muscle with supramaximal stimulation was recorded intraoperatively. A control group (twelve sheep) underwent an immediate tendon repair. In the remaining twenty-four sheep, the tendon was wrapped in a dura substitute to prevent scarring and was repaired at six weeks (eight sheep) and eighteen weeks (sixteen sheep) after release. In the immediate repair group, four animals were killed at six weeks; four, at twelve weeks; and four, at twenty weeks. In the six-week delayed repair group, four animals were killed at twelve weeks and four were killed at twenty weeks after the repair. In the eighteen-week delayed repair group, eight animals were killed at twelve weeks; four, at twenty weeks; and four, at thirty weeks after the repair. Muscle biopsies and testing were repeated prior to killing of the animals.

Results: The average force of muscle contraction decreased 3.6 lb (1.6 kg) by six weeks after the injury and 3.9 lb (1.8 kg) by eighteen weeks after the injury. After the repair, the force of contraction in the six-week group improved by 0.8 lb (0.4 kg) at twelve weeks postoperatively and by 1.3 lb (0.6 kg) at twenty weeks postoperatively. In contrast, no improvement occurred in the eighteen-week group until thirty weeks after the repair, at which time a 0.9-lb (0.4-kg) improvement was noted. There was a twelvefold increase in intramuscular fat concentration; this lipid infiltration was partially reversed after the tendon repair. Isolated tendon samples demonstrated an increase in the modulus of elasticity after chronic detachment that partially corrected after the tendon repair in the earlier (six-week) repair group.

Conclusions: We found that earlier repair of the tendon results in a more rapid recovery of both muscle function and tendon elasticity compared with a more delayed repair. We concluded that there may be a "point of no return" in rotator cuff injury after which the elasticity of the muscle-tendon unit does not return to normal.

Clinical Relevance: Injury and subsequent healing of the rotator cuff is a complex process that is poorly understood. We established a chronic rotator cuff injury and repair model in sheep. The model is reproducible and can be used to analyze both a direct repair of tendon to bone (chronic group) and an indirect repair requiring augmentation with a degradable patch (irreparable group).