Abstract
BACKGROUND: Reverse total shoulder arthroplasty (rTSA) for proximal humerus fractures is becoming more common. The aim of this study was to investigate the radiographic outcomes of uncemented rTSA for proximal humerus fractures and compare these outcomes based on 3 different implants. METHODS: This was a retrospective review of patients treated with uncemented primary rTSA for acute proximal humerus fracture at 2 large healthcare institutions using one of 3 different humeral stem designs. Group A (n = 46) received a Zimmer TM stem, group B (n = 27) received a DJO Altivate stem, and group C (n = 18) received a Medacta stem for a total of 159 patients. Three independent reviewers assessed the most recent x-rays for 1) humeral subsidence, 2) humeral tilt, 3) humeral radiolucent lines, 4) humeral stem at risk for loosening, 5) osteolysis, 6) humeral stress shielding, 7) component failure, 8) spot welds, 9) scapular notching, and 10) tuberosity union. RESULTS: The mean follow-up was 21 months (range, 6 months to 10 years). The average age at the time of surgery was 71.0 years old. Radiographic analysis showed group C (67%) was associated with significantly higher rates of stress shielding compared to group B (19%) and group A (6.5%) (P < .001). Two group C patients (11%) and one group A patient (2.2%) met criteria for a humeral stem at risk of loosening, a finding not observed in group B (P > .9). A logistic regression model further confirmed that group C stems were independently associated with increased odds of stress shielding (odds ratio = 28.667; 95% confidence interval: 6.229-131.924; P < .0001) compared to group A, whereas group B stems were not significantly different than group A (odds ratio = 3.258; 95% confidence interval: 0.712-14.906; P = .128). There were 4 cases of postoperative instability, 3 of which involved humeral loosening and required revision surgery (2 from group B and 1 from group C). One patient experienced a periprosthetic humerus fracture 8 years post-op, and another patient underwent several revision surgeries for periprosthetic joint infection, both from group A. None of the 3 groups demonstrated significant subsidence or scapular notching. CONCLUSION: Overall, we found significant variability in radiographic performance, particularly regarding stress shielding and risk of stem loosening when evaluating radiographic outcomes of these 3 cementless humeral stem designs used in rTSA for proximal humerus fractures.