The success of procedures such as arthroscopic rotator cuff repairs may be thought of as a race between tendon-to-bone healing and the disruptive forces of load and motion. Iannotti et al. report on the timing of failure of arthroscopic rotator cuff repair as demonstrated by magnetic resonance imaging (MRI) over a one-year follow-up period in a prospective, multicenter cohort. The most important novel finding in this study is that half of the retears occurred between three and six months after surgery, later than has been reported in earlier studies. Retears were not correlated with clinical outcomes but were correlated with shoulder strength at the one-year follow-up evaluation.
This study is an important contribution as it is the first to analyze time to failure with use of serial MRI after arthroscopic rotator cuff repair. While the risk of a retear after arthroscopic repair of the rotator cuff has been well documented1, data on when retears occur are relatively limited. A couple of recent studies have used ultrasound to evaluate survivorship of rotator cuff repairs2,3. In a prospective study of twenty-two rotator cuff tears larger than 3 cm that were fixed arthroscopically, Miller at al.3 reported a retear rate of 41%. Retears occurred early, with seven of nine found before three months. Kluger et al.2 evaluated a cohort of 107 patients with rotator cuff tears (no size reported) fixed with an arthroscopy-assisted mini-open technique. Using serial ultrasound examination, they found a failure rate of 33%, with the majority (74%) of the retears occurring within the first three months. The findings in the study by Iannotti et al. suggest that tendon-to-bone healing after arthroscopic rotator cuff repair may be a “longer race” than previously thought.
In the current study, 113 patients aged twelve to seventy-five years who were to undergo arthroscopic repair of a rotator cuff tear (with subscapularis tears and patients with grade-III or IV fatty infiltration excluded) 1 to 4 cm in size as seen on preoperative MRI were enrolled in the nonrandomized study at thirteen sites. It is important to recognize that the study population was relatively limited in terms of the degree of pathology. Furthermore, there is a possibility of selection bias in this cohort as the percentage of potential patients enrolled at each site over the study period was not reported, and the demographics of the study patients were not compared with those of patients who did not enroll. Thus, the generalizability of the study is limited.
A transosseous equivalent surgical technique was used with an average of two medial and two lateral anchors, and at least one lateral anchor was used in all patients. It is unclear whether a double-row transosseous equivalent is necessary for tears of this size as recent reviews have not shown a consistent benefit to double-row fixation4,5. The impact, if any, on healing is unclear. Shoulders were immobilized for six weeks, and strengthening did not start before twelve weeks. Serial MRIs were obtained at two, six, twelve, sixteen, twenty-six, and fifty-two weeks after surgery. The treating surgeon reviewed the MRI at each time point and diagnosed a retear based on fluid signal density in the tendon indicating a full-thickness gap. Nineteen patients (17%) were diagnosed with a retear, at a mean of nineteen weeks, with only one retear after twenty-six weeks (this occurred at fifty-two weeks). There was no correlation between retears and clinical outcomes at the one-year follow-up evaluation, but the mean ratio of scapular abduction strength of the affected shoulder to that of the normal, contralateral shoulder was lower in patients with a retear (75%) than in those with an intact repair (92%) (p = 0.0026).
A major limitation of this study is the technique used to diagnose a retear. Although MRI has been shown to have moderate reliability for diagnosing full-thickness retears6, the validity and reliability of having the treating surgeon read the MRIs are unknown. Scans were performed with 1.5 and 3.0-T machines, which could have affected the relative accuracy of the imaging. Finally, it was not clear whether the patient and/or evaluator was made aware of the MRI findings, which could have biased strength testing at the time of final follow-up.
However, the authors should be commended for performing a prospective collaborative study in an attempt to answer an important clinical question on the timing of retears after rotator cuff repair.
Their data demonstrate that patients who undergo arthroscopic repair of a rotator cuff tear are at risk for retear up to six months postoperatively, with a potential peak in the risk at three to six months. Although it may not be a marathon, the race to healing after arthroscopic rotator cuff repair certainly appears to be longer than a sprint. While the authors suggest protecting the repair from loading for up to six months and tissue-engineering strategies to augment healing as potential ways to reduce the rate of retears in this patient population, how to optimize the rotator cuff repair surgical technique, postoperative rehabilitation, and return to activity to improve the odds of winning the race is yet to be determined.