Prognostic factors affecting clinical outcomes after arthroscopic rotator cuff repair: importance of functional recovery by 3 months after surgery

Rotator cuff tear requires a large amount of treatment [1, 2]. Arthroscopic rotator cuff repair (ARCR) is a well-developed, modern procedure that produces acceptable postoperative outcomes [3]. Cole et al. [4] showed that ARCR resulted in significant pain relief and improvements in active range of motion, strength, and function. In patients with ARCR, American Shoulder and Elbow Surgeons shoulder (ASES) scores, range of motion, and strength are improved regardless of postoperative tendon healing status [5]. In a recent systematic review, clinically significant improvements in patient-reported outcomes, range of motion, and strength were observed up to 1 year after ARCR [6].

Kurowicki et al. [7] showed that all patients who underwent ARCR demonstrated significant incremental improvements in pain, function, and motion at 3-, 6-, and 12-month time points, suggesting that the plateau of maximum recovery after ARCR occurred at 1 year with high satisfaction.

Several studies [8‐10] reported preoperative prognostic factors associated with functional outcome after ARCR. Pécora et al. [10] showed that age is an independent predictive factor for obtaining good and excellent results after ARCR. Cho et al. [8] reported that the presence or absence of preoperative contractures affects internal and external rotation up to 6 months postoperatively, and contractures are also cited as prognostic factors. Ohzono et al. [9] reported that preoperative fatty degeneration of the subscapularis muscle and infraspinatus muscle affects clinical outcomes following the repair of large/extensive rotator cuff tears without re-tear.

In contrast, a few studies have focused on analyzing early postoperative prognostic factors associated with functional outcomes after ARCR. Kim et al. [11] demonstrated that stiffness of internal rotation at 3 months postoperatively affected the intensity pain pattern at the final follow-up. Tonotsuka et al. [12] reported that either an elevation less than 120° or external rotation less than 10° at 3 months postoperatively was associated with worse functional outcomes at 24 months postoperatively, suggesting the importance of functional results at the early phase of the postoperative course. In a recent systemic review [13], it was reported that postoperative re-tear tends to occur before 3 months postoperatively and significantly reduces both the University of California Los Angeles (UCLA) and ASES postoperative scores and strongly correlates with clinical outcomes. These reports consistently emphasized the importance of functional recovery at 3 months postoperatively and postoperative re-tear. Therefore, the purpose of the present study was to evaluate clinical results in patients with ARCR using the Japanese Orthopedic Association (JOA) score and to seek prognostic factors affecting clinical outcomes at 2 years postoperatively as evaluated by this score, extending to 3 months preoperatively to postoperatively.

Seventy-one patients were included in this study. The mean age at surgery was 63.4 ± 9.5 years. The mean symptom duration preoperatively was 10.1 ± 9.0 months.

The JOA score of all patients significantly improved from 63.7 ± 11.5 points preoperatively to 90.3 ± 9.6 points at 2 years, postoperatively (P < 0.05) (Fig. 1a).

The mean preoperative JOA scores were not significantly different between the two groups (Fig. 1b).

Postoperative outcomes of ARCR are generally favorable. However, the analysis of prognostic factors has included only preoperative factors, such as preoperative contracture, pain level, ROM, presence or absence of diabetes, thyroid disease [19], deep shoulder infection [20], and industrial accident [21‐23]. However, although there are a few reports on postoperative factors, it has been reported that ROM, pain level, and re-tear at 3 months postoperatively affected the postoperative outcomes of rotator cuff repair (RCR) [4, 24]. In the present study, we analyzed how factors from preoperative to 3 months postoperative and re-tear affect JOA outcomes 2 years after ARCR. As a result, we found that active elevation at 3 months postoperatively (≥ 110°) and VAS at 2 months postoperatively (< 5 points) were significant prognostic factors. However, postoperative re-tear was not found to be a prognostic factor based on the analysis performed in the present study. Thus, owing to our analysis of the range, reported to up to 3 months postoperatively, including preoperative factors, it was suggested that postoperative factors are more important than preoperative factors.

Kim et al. [11] evaluated postoperative ROM at 6 weeks and 3, 6, and 12 months postoperatively using univariate and multivariate analyses and found that internal rotation at 3 months postoperatively affected the pain level at all time points (final observation at 12 months). Tonotsuka et al. [12] examined the association of ROM between 3 months and 2 years after surgery and reported that if an elevation of 120° or external rotation of 10° is not achieved at 3 months, then ROM is also decreased at 2 years postoperatively. In the present study, active elevation at 3 months postoperatively (≥ 110°) was identified as a significant factor in both univariate and multivariate analyses, which, in part, supports the report by Tonotsuka et al. [12].

Amako et al. [25] conducted a postoperative evaluation of rotator cuff tears in groups with and without preoperative pain and found a significant difference in the JOA scores and DASH scores up to 2 years postoperatively. They reported that persistent postoperative pain was a factor that reduces postoperative functional outcomes, patient level of satisfaction, and quality of life. They also reported that pre- and postoperative pains were closely related. In the present study, pre- and postoperative pains were included in our analyses. As a result, only postoperative pain was identified as a significant factor. Tonotsuka et al. [26] reported that it is important to control preoperative inflammatory pain to obtain better postoperative outcomes following rotator cuff repair. The results of the present study suggest that it is important to control postoperative pain. However, further examination is needed regarding the relationship between pre- and postoperative pain.

Postoperative re-tear is said to affect postoperative clinical outcomes. Chung et al. [27] reported that the acromiohumeral interval was a significant finding upon performing ARCR for extensive rotator cuff tears and performing evaluations based on the ASES score of 80 points. Furthermore, upon examining re-tear and clinical function in a recent meta-analysis, it was found that re-tear significantly reduced UCLA/ASES/Constant scores and that re-tear was a prognostic factor for clinical outcomes [13]. On the other hand, in the present study, re-tear did not affect postoperative outcomes. This should be further examined considering that the small sample size could have affected the results.

A significant difference in the number of partial repairs was observed based on the univariate analysis in the present study. If partial repair was considered a “structural failure” in re-tear cases, it was not a significant factor after multivariate analysis, although it could have fundamentally affected our results.

In the present study, multivariate analysis revealed that the pain level at 2 months postoperatively (VAS > 5) and the elevation of ROM at 3 months postoperatively (> 110°) affected the JOA score at 2 years postoperatively. This suggests that pain level and early ROM postoperatively are important factors to predict postoperative outcomes and simultaneously signifies that postoperative pain management, including attainment of early postoperative ROM, is important.

Shin et al. [28] reported that postoperative injection of steroids is useful in controlling pain for postoperative management. Cho et al. [29] administered postoperative local anesthetic transvenously or via the subacromial bursa and reported that it was useful in controlling pain. Nakagawa et al. [30] reported that postoperative cervical continuous epidural anesthesia for 2 weeks postoperatively significantly improved JOA score and pain evaluation at 3 months postoperatively. In our report, all patients received an anesthetic block only at the time of surgery, and postoperatively, they were only administered nonsteroidal anti-inflammatory drugs. Based on our results, we believe that postoperative pain management is important and better pain control is needed early postoperatively.

Shoulder-function rating systems for ARCR are used for the evaluation of clinical outcomes: UCLA score [31‐33], Constant score [34], and ASES score [7, 31, 33, 35, 36]. Our recent study [18] shows that the cutoff value of 28 points in the UCLA score corresponded to 83 points in JOA score in patients with ARCR. Since there is growing evidence in the English literature about patients with ARCR evaluated by JOA score [32, 37‐39], the present study focused on seeking prognostic factors affecting clinical outcomes after ARCR, exclusively using this score.

There are several limitations in the present study. First, this was a retrospective, small study. Second, evaluations solely comprised the JOA scores and did not include the patients’ perspectives; therefore, the data may have varied on the basis of the set standards. On the other hand, as a result of comprehensively analyzing preoperative factors to early postoperative factors reported to date, we conclude that pain level at 2 months postoperatively and elevation ROM at 3 months postoperatively have a greater effect than preoperative factors and that early postoperative intervention is important.

To obtain a JOA score of ≥ 83 points at 24 months postoperatively, a postoperative VAS of < 5 points at 2 months and a postoperative elevation of ≥ 110° at 3 months should be achieved following ARCR.