Clinical outcomes of anatomic vs. reverse total shoulder arthroplasty in primary osteoarthritis with preoperative rotational stiffness and an intact rotator cuff: a case control study.

Abstract

Reverse total shoulder arthroplasty (rTSA) has begun to challenge the place of anatomic total shoulder arthroplasty (aTSA) as a primary procedure for certain indications. One purported benefit of aTSA is improved postoperative range of motion (ROM) compared to rTSA especially in internal rotation; however, it is unclear whether aTSA can provide patients with significant preoperative stiffness superior ROM compared to rTSA. Our purpose was to compare clinical outcomes of aTSA and rTSA performed in stiff vs. non-stiff shoulders for rotator cuff intact (RCI) glenohumeral osteoarthritis (GHOA). A retrospective review of an international shoulder arthroplasty database identified 1608 aTSAs and 600 rTSAs performed for RCI GHOA with minimum 2-year follow-up. Defining preoperative stiffness as ≤ 0° of passive external rotation (ER), we matched: (1) stiff aTSAs (n=257) 1:3 to non-stiff aTSAs, (2) stiff rTSAs (n=87) 1:3 to non-stiff rTSAs, and (3) stiff rTSAs (n=87) 1:1 to stiff aTSAs. We compared ROM, outcome scores, and the rate of complications and revision surgery at latest follow-up. Despite stiff aTSAs having poorer preoperative ROM and functional outcome scores for all measures assessed (P<.001 for all), only poorer postoperative active abduction (113±27° vs. 128±35°; P<.001), active ER (39±18° vs. 50±20°; P<.001), and passive ER (45±17° vs. 56±18°; P<.001) persisted postoperatively compared to the non-stiff cohort. Similarly, stiff rTSAs had poorer preoperative ROM and functional outcome scores for all measures assessed compared to non-stiff rTSAs (P≤.044), but only poorer active abduction (108±24° vs. 128±29°, P<.001), active ER (28±17° vs. 42±17°, P<.001), and passive ER (36±15° vs. 48±17°, P<.001) persisted. When comparing stiff rTSAs to matched stiff aTSAs, no significant differences in preoperative ROM or functional outcome scores were found. However, stiff aTSAs had greater postoperative active internal rotation score (4.8±1.5 vs. 4.2±1.7, P=.022), active ER (40±19° vs. 28±17°, P<.001), and passive ER (46±18° vs. 36±15°, P=.001). Postoperative outcome scores were similar across all matched cohort comparisons despite motion differences. The rate of complications and need for revision surgery did not differ between any group comparisons. Patients with RCI GHOA who have preoperative rotational stiffness have poorer postoperative ROM compared with non-stiff patients following both aTSA and rTSA, but similar functional outcome scores. Notably, preoperative limitations in passive ER do not appear to be a limitation to utilizing aTSA. Indeed, patients with limited preoperative ER treated with aTSA had greater postoperative internal rotation and ER compared to those treated with rTSA.