RSA can be an effective surgical option to restore function and relieve pain in the appropriate patient population. Most published RSA outcome studies focus on symptomatic rotator cuff-deficient patients for whom traditional non-constrained shoulder arthroplasty is not the best surgical option [3, 4, 13], and there are few reports that specifically analyze the outcomes of RSA for this massive rotator cuff tears without glenohumeral arthritis.

Wall et al. [3] provided one of the first reports of RSA outcomes according to etiology in 2007. In this review of 232 patients with a mean age of 72.7 years that had undergone RSA, 17.1 % were performed for a massive rotator cuff tear without glenohumeral arthritis. At a mean follow-up of 34 months, significant improvements were observed in all domains of the Constant score, 56 % improvement in pain, and elevation improved from 94° preoperatively to 143° postoperatively. With the Grammont-style prosthesis utilized, postoperative external rotation decreased by 6°. There was a 19.1 % prevalence of complications within the study population, but the complication rate was not subanalyzed in the massive rotator cuff tear without glenohumeral arthritis patient cohort.

In a retrospective multicenter study of 42 shoulders that had undergone an RSA after failed rotator cuff repair surgery, Boileau et al. [13] analyzed patients with and without glenohumeral arthritis. There was no significant difference between cohorts in regard to pain, Constant score, or forward elevation at a mean follow-up of 50 months. The mean postoperative forward elevation was 123° in both groups. 15.7 % of the study population that underwent RSA had rotator cuff deficiency without glenohumeral arthritis. These patients had 49 % improvement in their pain rating and a 40.9 % improvement in their adjusted Constant score. 7.5 % of the patients within the study developed a complication, but complications were not specifically analyzed in patients with massive rotator cuff tears without glenohumeral arthritis.

Mulieri et al. [14] published their results of RSA for the treatment of massive rotator cuff tears specifically without glenohumeral arthritis. They reviewed the prospectively collected data of 60 shoulders in 58 patients with an irreparable rotator cuff tear without glenohumeral arthritis. Mean statistically significant improvements were observed in the study population in regard to American Shoulder and Elbow Surgeons (ASES) score, simple shoulder test (SST), visual analog scale (VAS) pain and functional score, Short Form-36 (SF-36) score, and range of motion. 44 % improvement in pain and 82 % improvement in function were observed on the VAS. There was a 20 % complication rate. Forward elevation improved from 53° to 134°, and external rotation improved from 27° to 51°. In contrast to the studies by Wall and Boileau [3, 13], a non-Grammont-style reverse prosthesis was utilized that lateralizes the center of rotation from the native glenoid surface and utilized a more anatomical head–neck angle opposed to the valgus 155° angle in traditional Grammont-style prostheses.

When aggregating the results of these studies on RSA for massive rotator cuff without glenohumeral arthritis, there were 111 patients with a mean follow-up of 46.2 months [3, 13, 14] (Table 13.1). There was a 48 % improvement from the patient’s preoperative to postoperative subjective pain rating. Forward elevation improved by a mean of 64.3°, with a mean final forward elevation of 135.1° compared to 70.1° preoperatively. Different functional scores were utilized between the studies which precluded further analysis. Complications specific to this patient population were only available in the report by Mulieri et al. [14], which was 20 %, with 7 % of these complications due to baseplate failure that was attributed to a design flaw that has subsequently been addressed. The published complication rates of RSA for any indication are variable ranging from 0 to 68 % depending on the indication and definition of a complication [15–19]. The 20 % complication rate reported by Mulieri et al. in 2010 approximates the mean aggregate rate of clinically significant complications reported in the literature [14, 17]. There is not sufficient evidence to suggest that patients with rotator cuff tears without arthritis are at more or less risk of developing complications following RSA than patients with other indications.

As the indications for RSA have expanded beyond Hamada stage 4 and 5 cuff tear arthropathy over the last 15 years, it is essential to critically evaluate patient outcomes. Wall et al. [3] demonstrated statistically significant improvements in all domains of the Constant score when subanalyzed by indication including rotator cuff tears with and without arthritis, primary osteoarthritis, posttraumatic arthritis, and revision arthroplasty. However, equivalent outcomes cannot be expected for every patient treated with an RSA for a given diagnosis. This is especially true for patients with massive rotator cuff tears without arthritis because their preoperative function can be variable, which often affects the ultimate result. One of the most important preoperative factors to consider in these patients is the presence of pseudoparalysis, which is defined as less than 90° of active forward elevation in the presence of full passive elevation [20].

Boileau et al. [13] highlighted this concept when they analyzed the results of RSA for patients after failed rotator cuff repair surgery. They compared the results of patients with and without 90° of active forward elevation. In patients with pseudoparalysis, improvements were observed in all domains of the Constant score and active forward elevation improved from 56° to 123°. However, patients without pseudoparalysis demonstrated improvements in their pain rating and Constant score, but a mean decrease in active forward elevation from 156° to 122°. The authors concluded that patients with forward elevation above 90° with pain after failed rotator cuff repair surgery can expect moderate improvements in their pain rating and a potential decrease in function and therefore may be better treated with a procedure other than RSA. However, the authors did not subanalyze the effect of pseudoparalysis on patients specifically without glenohumeral arthritis.

Mulieri et al. [14] further analyzed the outcomes of patients with massive rotator cuff tears specifically without arthritis treated with RSA. The authors found that 7 % of the 60 shoulders included in the study had greater than 90° of forward elevation. While pain and function improved in all of these patients with maintained preoperative forward elevation, the authors also found a mean improvement in forward elevation from 126° to 155°. There was also a 50 % complication rate in this cohort of 4 patients. The improvement in forward elevation in these patients was in contrast to the report by Boileau, who reported a mean decrease in forward elevation as noted above [13].

Prior surgery is common in patients with massive rotator cuff tears without arthritis that are being considered for RSA. In the report by Mulieri, 43 % of these patients had prior surgery [14]. Multiple studies have investigated the influence of prior surgery on the outcome of patients undergoing RSA [13, 14, 20, 21]. While significant improvements were often observed in regard to pain, function, and motion, improvement was slightly less in patient with prior surgery [13, 20]. There has been only 1 study to directly compare the outcomes of patients with and without prior surgery treated with RSA for massive rotator cuff tears without arthritis, and no significant differences were observed [14].