This study reports that 12.9 years after primary open labral repair and capsulorrhaphy for post-traumatic anterior shoulder instability, patients presenting with specific histopathologic changes within the subscapularis muscle at the time of surgery had a significantly increased incidence of recurrence compared to patients with normal histology (33.3% vs. 4.3%, respectively). Except for external rotation, the presence of these tissue alterations did not affect other clinical and radiographic outcomes. Indeed, ER1 was significantly decreased on the operated compared to the contralateral side in patients with normal histology (-11.9°), but not in patients with abnormal histology (-6.1°).
This study sample closely relates to the population undergoing shoulder stabilization for post-traumatic anterior shoulder instability, with patients’ demographic characteristics (sex ratio, mean age) matching those recently reported in a dedicated multicenter register .
The overall incidence of recurrence in our cohort (14.3%) is within the range reported by other authors with a follow-up >10 years after the same procedure and including painful subluxation in the definition of recurrence (4.5% to 16.1%) [3–6]. An association was found between histopathologic anomalies of the subscapularis muscle (type 1 fiber predominance, interstitial fibrosis and focal atrophy) and recurrence of shoulder instability. To our knowledge, this association has never been reported yet. A possible explanation could be that the anterior stabilizing function of the subscapularis muscle is adversely affected by a scarring process associated with disuse atrophy within its substance. Decreased muscular tone resulting in failure of dynamic shoulder stabilization and some degree of laxity could lead to an increased risk of recurrent shoulder instability . These histopathologic anomalies might correspond to the macroscopic lesions of the same muscle (muscular tears seen intraoperatively) described by Rowe et al. in 7% of 158 unstable shoulders . When compared to the 40% prevalence of microscopic lesions observed in our previous report, this lower prevalence of macroscopic tears leads us to hypothesize that all these lesions may belong to a unique pattern of progressive muscle injury, with macroscopic tears representing a more advanced stage . These lesions could be related to the initial traumatic dislocation and recurrent episodes, to the difficulty in reduction, and to the healing process within the injured muscle. However, the presence of histopathologic changes in the subscapularis muscle was not associated with an increased number of dislocations before surgical stabilization. This suggests that the initial traumatic dislocation and difficulty in reduction may be more important than the number of dislocation episodes before stabilization in the development of these histopathologic changes.
Active ROM at follow-up revealed a significant decrease of mean ER1 (-10.0°) and ER2 (-12.7°) of the operated shoulder for the whole study cohort. Other studies with a follow-up >10 years after the same procedure have shown similar results (reported decreases ranging from 6° to 33.6° for ER1 and from 8.7° to 24.4° for ER2) [3–6, 26, 33–35]. The presence of histopathologic lesions affected differently the decrease of ER1 and ER2 with no significant decrease of ER1 on the operated side among patients with histopathologic changes. Open labral repair and capsulorrhaphy is a procedure restricting external rotation due to tightening of anterior soft tissues [6, 23, 36]. The absence of a significant postoperative decrease of ER1 in patients with subscapularis muscle histopathologic lesions could be linked to reduced passive resistance and attenuated dynamic shoulder stabilization activity of the pathologic muscle.
Functional follow-up scores confirmed the good long-term results already reported in the literature after the same procedure [3–6, 26, 27, 34]. Although patients with subscapularis muscle histopathologic anomalies had more recurrence in this study, their functional scores at follow-up were not significantly different when compared to patients with normal muscle histology. However, they presented with a trend towards a lower Rowe shoulder score. This is mainly explained by the fact that one-third presented recurrence of shoulder instability and half of this score assesses stability.
At follow-up, degenerative changes were absent in nearly half of all patients, mild in one third, moderate in one tenth, and severe in none. The severity of these changes was not correlated to the number of preoperative dislocation episodes. Results from other studies with follow-up >10 years suggest that primary open labral repair and capsulorrhaphy for post-traumatic anterior shoulder instability are associated with progressive shoulder osteoarthritis developing over the years [3–6, 26, 27, 34]. This process could possibly be caused by the physiologic interference of the surgical repair. Osteoarthritic changes have been associated with loss of external rotation and overtightening of anterior soft tissues around the glenohumeral joint following stabilization [5, 6, 36, 37]. The present study showed that patients with subscapularis muscle histopathologic changes had less loss of external rotation than patients without these anomalies. Consequently, we would have expected them to have less degenerative changes. However, there was no statistically significant difference regarding degenerative changes between both patients' groups. This may be because of a too small sample size with insufficient statistical power to obtain significance, a possible issue with the validity of the Samilson and Prieto classification not currently documented, or because degenerative changes are primarily due to the traumatism of the initial dislocation [38, 39].
This study has several limitations: 1) the bone loss on the glenoid rim or the humeral head (an important factor for recurrence after shoulder stabilization), which could not be evaluated as a confounding factor because it was neither consistently recorded on the operative notes nor clearly identifiable on available preoperative radiographs. Of note, we were able to review anteroposterior plain radiograph in neutral rotation for each patient, which was the routine in-hospital preoperative workout in our institution at the time surgeries were performed; however, additional views (anteroposterior with internal and external rotation, axial, or CT-scan in some instances), which were performed before hospitalization as part of a complete work-out, were no more available for a large number of patients, because they had been lost by the patients, not kept by the archive department of our institution, or of a too bad quality to allow for an accurate evaluation of the bone stock. 2) The relatively high rate of patients lost to follow-up (32.7%), which may be explained by their young age (mean age at surgery, 27.2 years) and the long delay since surgery (mean follow-up, 12.9 years); 3) the biopsy specimen harvested in one limited area, which may not imply that a significant portion of the subscapularis muscle presents with the same histopathologic features; 4) the number of preoperative shoulder dislocations, which may have been imprecisely recorded (relying on history taking); 5) the retrospective assessment of shoulder dislocation recurrence; and 6) the small number of events with respect to our primary outcome, leading to large confidence intervals.