Richard J. Friedman, Bradley S. Schoch, Josef Karl Eichinger, Garrett B. Neel, Marissa L. Boettcher, Pierre-Henri Flurin, Thomas W. Wright, Joseph D. Zuckerman, Christopher Roche
5 min
This study aimed to compare the clinical and radiographic outcomes of two primary shoulder replacement procedures—anatomic total shoulder arthroplasty (aTSA) and reverse total shoulder arthroplasty (rTSA)—in patients diagnosed with osteoarthritis who had an intact rotator cuff and no history of prior shoulder surgery.
The researchers conducted a retrospective analysis of 740 patients (370 in each group) from an international, multi-institutional registry. Patients were matched for age, sex, and follow-up duration. All procedures utilized a single-platform shoulder system. The study evaluated patients using six standardized outcome scoring metrics, four active range of motion (ROM) measurements, and radiographic assessments at a minimum of 2 years post-surgery.
At a mean follow-up of 41 months, both groups demonstrated excellent pain relief and high patient satisfaction. While clinical outcomes and function were generally similar, aTSA patients showed greater external rotation. Notably, the aTSA group experienced a significantly higher rate of complications (4.9%) compared with the rTSA group (2.2%). Revision rates were similar between the two cohorts. The study suggests that rTSA is a viable, safe alternative to aTSA for patients with an intact rotator cuff, offering comparable functional results with a lower risk of adverse events.
Traditionally, aTSA has been the standard for patients with an intact rotator cuff, while rTSA was reserved for more complex cases like cuff tear arthropathy. This study provides evidence that rTSA can be used effectively in a broader range of patients, potentially reducing the complication burden associated with traditional anatomic replacements.
INTRODUCTION: This study's purpose is to compare clinical and radiographic outcomes of primary anatomic total shoulder arthroplasty (aTSA) and primary reverse total shoulder arthroplasty (rTSA) patients with osteoarthritis (OA) and an intact rotator cuff with no previous history of shoulder surgery using a single platform TSA system at a minimum follow-up of 2 years. METHODS: A total of 370 aTSA patients and 370 rTSA patients matched for age, sex, and length of follow-up from an international multi-institutional Western Institutional Review Board approved registry with a minimum 2-year follow-up were reviewed for this study. All patients had a diagnosis of OA, an intact rotator cuff, and no previous shoulder surgery. All patients were evaluated and scored preoperatively and at latest follow-up using six outcome scoring metrics and four active range of motion measurements. RESULTS: Mean follow-up was 41 months, and the mean age was 73 years. Preoperatively, the rTSA patients had lower outcome metrics and less motion. Postoperatively, aTSA and rTSA patients had similar clinical outcomes, motion, and function, with the only exception being greater external rotation in aTSA exceeding the minimal clinically important difference. Pain relief was excellent, and patient satisfaction was high in both groups. Humeral radiolucent lines were similar in both groups (8%). Complications were significantly higher with aTSA (aTSA = 4.9%; rTSA = 2.2%; P = 0.045), but revisions were similar (aTSA = 3.2%; rTSA = 1.4%; P = 0.086). CONCLUSION: At a mean of 41 month follow-up, primary aTSA and rTSA patients with OA and an intact rotator cuff with no previous history of shoulder surgery had similar clinical and radiographic outcomes. Greater external rotation was noted in aTSA patients at follow-up. However, aTSA patients had a significantly greater rate of complications compared with rTSA patients. rTSA is a viable treatment option in patients with an intact rotator cuff and no previous shoulder surgery, offering similar clinical outcomes with a lower complication rate. LEVEL OF EVIDENCE: Level III.
Sam: [probing] What were the specific failure modes on the anatomical side? Was it purely loosening? [[RP_SECTION:failure-modes-and-limitations|Failure Modes and Limitations]]
Alex: [analytical] Primarily glenoid and humeral loosening, plus some rotator cuff tears — which makes mechanistic sense, because the anatomical design places the glenoid component under shear stress that the reverse geometry avoids by shifting the center of rotation medially and inferiorly. The reverse group had a more diverse complication profile — infections, component dissociation — but at lower overall frequency.
Sam: Did the implant platform or surgeon variation factor in?
Alex: [honest] They used a single-platform system across all sites, which standardizes the hardware. But surgeon-specific rehab protocols and site-level variation remain uncontrolled — the authors flag this as a residual confound. It's the kind of thing a registry design can identify but can't fully resolve.
Sam: [pushing back] And the follow-up window — 41 months on average. That's under four years. Is that long enough to capture the true failure rate of a glenoid component?
Alex: [reflective] It's the primary limitation, and it's a meaningful one. Glenoid loosening is a long-term failure mode. At under four years, you're catching early mechanical failures, but late-stage degradation — the kind that shows up at seven or ten years — is almost certainly undercounted. The complication gap between the two constructs could narrow, widen, or invert at longer follow-up. We don't know yet.
Sam: [analytical] And even with propensity matching, there's a residual selection problem. Surgeons who chose the anatomical implant may have done so precisely because the patient had better bone stock — a variable that wasn't captured in the matching algorithm.
Alex: [acknowledging] The authors admit that. Glenoid morphology and bone density likely influenced the initial implant choice, and those variables weren't fully measured. So some of the reverse group's favorable safety signal could reflect unmeasured patient-level advantages rather than implant-level superiority. That's the honest read of what the data can and can't support.
Sam: So the load-bearing finding is that the reverse construct is competitive on function and safer on complications — but that conclusion is constrained by a short observation window and the limits of registry matching. [[RP_SECTION:future-clinical-directions|Future Clinical Directions]]
Alex: [quiet confidence] That's exactly right. The study's contribution is shifting the prior: the reverse construct is no longer a fallback. But the next methodological step is a predictive model that integrates preoperative imaging and bone density — moving beyond the current binary choice toward implant selection that's tailored to the individual patient's anatomy and long-term risk profile.
Sam: [thoughtful] The field is shifting from asking which procedure is better, to asking which procedure is better for this patient.
Alex: [settling the point] And this study gives surgeons the evidence to have that conversation seriously. The reverse construct has earned a place in primary planning — not as the default, but as a genuine option rather than a last resort. Thanks for listening to ResearchPod.