This study aimed to investigate the retear rate after RC surgery at different time points, also evaluating both patient-related and not patient-related factors. The first ones concern preoperative patients’ characteristics, such as age, tear size, and fatty infiltration. The second ones are related to postoperative rehabilitation protocol, or intraoperative choices of surgical procedures, and RC repairs techniques. The present meta-analysis, including only level 1 and 2 evidence studies, reports data on over 2500 RC repairs.
Retear rate and patient-related risk factors
The advanced age of patients and larger tear sizes are predictors of RC retear, in agreement with previously published studies [
4,
6,
20,
90]. The negative influence of older age on the tendons healing process also depends on other concomitant factors age-related, such as lifestyle, bone mineral density, comorbidities. Some studies report that older age is not an independent predictor of RCR [
5]. Larger tear size has been associated with a higher retear rate also in previous systematic reviews [
2,
91]. The results of our investigation confirm this association, showing a strong statistical significance (
P < 0.0001). However, in the attempt to include as many as possible studies in the quantitative analysis, we identified two macro groups. The first group included studies reporting postoperative retear rate for patients with small and/or medium tear size, and the second one included studies enrolling only patients with large and/or massive tear size. For this reason, only 11 studies were analyzed to investigate the relationship between preoperative tear size and retear rate, excluding those studies that enrolled patients independently from the preoperative tear size. Even if 97% of the included studies reported data about patients’ age and preoperative tear size, no one has provided a direct association regarding the number of patients who experienced a retear and their presurgical features. Moreover, the collected information was not enough to perform a stratified analysis, so the definition of the independent effect of both patients’ age and preoperative tear size was not possible.
Current literature reports that RCs with higher muscle fatty infiltration have an increased likelihood of suffering RCR [
92]. Our results showed no statistically significant difference (
P = 0.7588). This deniable statement agrees with the results reported in a recent study [
91]. A plausible explanation is that there is no robust scientific evidence. Moreover, studies reporting data on preoperative fatty infiltration did not provide postoperative variations in fatty degeneration and any correlation with RC retear. Although fatty infiltration has been considered as one of the main factors influencing tendons healing after surgery [
20,
93], further clinical investigations should be performed to corroborate its impact with exhaustive evidence.
As reported in previous studies, additional patient-related factors that could negatively influence the healing process of the repaired tendons are smoking, diabetes, osteoporosis, hyperlipidemia [
5,
6]. In the present work, these risk factors were not analyzed, and further investigations are needed to provide more robust clinical evidence.
Retear rate and not patient-related risk factors
The biomechanics of the repaired tendons may also be affected by not patient-related factors. In the postoperative period, patients may experience limited functionalities of the affected arm and pain. In this regard, wide debates arise in the definitions of the best rehabilitation programs that should minimize the risk of healing failure and guarantee a successful return to activities of daily living [
1,
94]. In the current clinical practice, the postoperative management of patients’ underwent RC repair can be slightly different among studies in terms of time points in which start specific movements and physical exercises. Based on the available literature, the postoperative rehabilitation protocol could be split into four main phases [
95]. The first phase refers to the immediate postoperative period until the 6-week during which supervised passive ROM and active-assisted ROM are allowed; in the second phase (weeks 6–12), patients start to execute full-active ROM; in the third phase (months 3–4), stretching and strengthening exercises can be initiated and continued in the fourth phase (months 4–6) to completely restore full and pain-free active ROM and return as normally as possible to sports, activities of daily living and work. The timing of immobilization has been investigated in some recent randomized controlled trials [
65,
70,
77]. Longer periods of immobilization may result in shoulder stiffness, which negatively influences the healing process after surgery [
81]. Commonly, most patients are asked to wear an abduction pillow for 3 to 6 weeks, during which home postural exercises and assisted ROM during physical therapy are prescribed [
65,
84]. The recommended immobilization period may change based on the preoperative tear size. At the same time, the effect of early passive mobilization on the healing rate after RC surgery has been investigated [
64,
68,
84]. Compared with the delayed rehabilitation protocol, the early mobilization aims to avoid the likelihood that adhesions would give rise to shoulder stiffness; conversely, delayed rehabilitation protocol seeks to preserve the tendon-to-bone integrity, avoiding retear. One study compared the retear rate in two groups immobilized for four or eight weeks, avoiding any type of passive or active ROM exercises [
67]. At a mean of 6.8 months, MRI showed a retear rate of 12.5% for 4-weeks immobilization group and a retear rate of 8.3% for 8-weeks immobilization group. The same study proposed a subgroup analysis, including only patients without preoperative shoulder stiffness. Results showed that at 24-months postoperatively, the 8-week immobilization group had a higher percentage of patients with stiffness [
67]. Such findings suggest that the risk of shoulder stiffness might be avoided executing balanced and limited ROM in the first weeks after surgery. Our analysis showed no statistically significant differences for immobilization periods within 6 weeks or longer than 6 weeks postoperatively (OR, 0.4; 95% CI; 0.1 to 1.2;
P = 0.0912). Some recent meta-analysis investigated the outcomes of early versus delayed rehabilitation [
18,
24]. These studies report that early motion protocol corresponds to an increase of ROM after RC repair, but also the risk of retear increases. Based on our findings that larger RC tear size may experience a lower healing rate, we suggest that early motion could be recommended for smaller tear size, while the delayed motion for larger tear size. In the first phase of rehabilitation (within the 5 weeks), active-assisted ROM should be avoided, since a higher retear rate was found for active-assisted ROM starting before 5 weeks (OR, 0.5; 95% CI, 0.4 to 0.7;
P < 0.0001). In the second phase of rehabilitation (weeks 6–12), full active ROM can be recommended; in particular, our results suggest a higher healing rate if full active ROMs are started before the 8th week (OR, 2; 95% CI, 1.3 to 3.2;
P = 0.0028). Usually, strengthening exercises are recommended after the 12th week [
28,
57,
85], when full active ROM and dynamic shoulder stabilization should be reached [
95]. Our results suggest a higher healing rate if strengthening exercises are started before the 12th week, although this result showed no statistically significant difference (OR, 1.1; 95% CI, 0.8 to 1.5;
P = 0.4653).
During these periods, patients’ compliance with immobilization and prescribed movements should be analyzed [
65,
96]. Monitoring patients using wearable technologies could be a plausible alternative if compared to a questionnaire-based investigation [
1,
97]. As highlighted previously, tendons healing is strictly associated with factors that surgeons could not handle totally because of dependence from patients’ biological characteristics, as age, tear size, muscles fatty degeneration, and atrophy. Due to the heterogeneity of the included studies and insufficient available data, a stratified analysis for the determination of the independent effect of each factor was not possible to carry out.
In the last decades, arthroscopic RC repair supplanted previous techniques thanks to progress in surgical and technological instrumentations [
98]. Faster recovery and better cosmetic results have been the main reasons to prefer the arthroscopic approach. Further studies also supported good clinical outcomes and a low retear rate [
99]. The present investigation did not show statistically significant results comparing the retear rate associated with arthroscopic or open and mini-open RC repair (OR, 1; 95% CI, 0.7 to 1.7;
P = 0.8524).
Our data indicated that double-row techniques yield a lower retear rate than suture bridge/transosseous (OR, 0.5; 95% CI, 0.3 to 0.7; P = 0.0001) and single-row RC repair, although there was no statistically significant difference for the latter (OR, 1.3; 95% CI, 0.9 to 1.9; P = 0.2036). The present study showed that single-row RC repair is associated with a lower retear rate compared to suture bridge/transosseous RC repair (OR, 0.6; 95% CI, 0.4 to 0.8; P = 0.0005).
Double-row RC repair has been described as biomechanically superior compared with single-row [
100]. According to our study, numerous systematic reviews and meta-analyses have shown lower postoperative retear rate after double-row repair.
Our findings partially agree with those of Hein et al. that found that double-row had significantly lower retear rate compared with single-row [
16]. However, they did not find any significant difference between double-row and suture bridge/transosseous and significantly lower retear rate with suture bridge/transosseous than single-row. Possible reasons for dissimilar results could be patient population, follow-up time, methods for retear diagnosis, and sample size.
As this study focuses on the retear rate based on imaging-classification, no conclusions regarding clinical outcomes as a function of repair technique can be made. Yang et al. demonstrated that postoperative RCR alters clinical outcomes [
101]. Future studies should compare differences in the effect based on repair types.
PRP is a promising treatment for some musculoskeletal diseases; however, evidence of its efficacy in the treatment of RC pathologies is still insufficient. Several studies focused on PRP injection for RC tendinopathy, showing benefits over sham injection, no injection, or physiotherapy alone in reducing pain at long-term follow-up [
102]. In this review, we analyzed evidence of PRP in arthroscopic repair of RC tears compared with conventional surgery. Lower retear rate was observed with the use of PRP, but several aspects need to be further focused. Many of the included studies specifically looked at the use of platelet-rich fibrin matrix for augmentation (PRFM), while others injected PRP directly into the repair site, including leukocyte-rich PRP (LR-PRP) and leukocyte-poor PRP (LP-PRP). The differences in concentration, content, preparation method and delivery technique do not allow to derive definitive conclusions. Moreover, patients were not stratified based on concomitant factors that can affect the retear rate, such as size, chronicity, atrophy, fatty infiltration, patients’ age, use of tobacco products, diabetes, and other patient-related factors.
Previous studies confirm that PRP has effects on RC structural integrity, promoting tendon healing to the bone, but no effects on clinical outcomes were observed after RC repair. As mentioned above, the present review focused on the retear rate based on imaging-classification, and no conclusions regarding clinical outcomes as a function of repair technique can be made. The efficacy of PRP in arthroscopic repair of RC tears remains under investigation.
The results of this review show that augmented RC repair has a lower retear rate. Structural integrity in postoperative imaging has been documented, but literature is still insufficient.