Revision reverse shoulder arthroplasty after failed shoulder arthroplasty: long-term survivorship and risk factors for failure

Reverse shoulder arthroplasty (rTSA) is increasingly used as a revision option, but long-term survivorship and patient-reported outcomes (PROs) after revision rTSA remain poorly defined. We hypothesized that revision rTSA for failed rTSA would have inferior survivorship and PROs when compared to revision rTSA for failed anatomic total shoulder arthroplasty (aTSA) or failed hemiarthroplasty (HA). Moreover, we hypothesized that shoulders undergoing secondary revision would have a higher risk of failure when compared to those undergoing first-time revision surgery. This study is a retrospective review of a prospectively collected, single-surgeon institutional database from 2002-2022. Patients undergoing first-time revision rTSA for aseptic mechanical failure of a prior arthroplasty were included if they had &#x2265;12 months of follow-up or required subsequent revision within 12 months. Survivorship was assessed using Kaplan-Meier analysis. PROs (American Shoulder and Elbow Surgeons [ASES] scores and satisfaction) were collected at one year, with subgroup analysis by index failed arthroplasty type: failed HA, failed aTSA, and failed rTSA. Modes of failure after secondary revision were also examined. Statistical significance was set at P < .05. Of 464 first-time revision rTSA for mechanical failure from 2002-2022, 320 revision rTSAs had appropriate follow-up and met inclusion criteria, with mean follow-up of 50.3 &#xb1; 45.7 months. Mean age was 67.4 &#xb1; 10.0 years; 54.1% were female. Overall survivorship of revision rTSA was 86.4% at 2 years and 71.4% at 10 years. In terms of survivorship stratified by index arthroplasty, the survivorship of failed HA, failed aTSA, and failed rTSA was the following: failed HA: 2-year 93.4% and 10-year 74.2%; failed aTSA: 2-year 88.6% and 10-year 82.1%; failed rTSA: 2-year 74.6% and 10-year 59.5%. Overall survivorship was inferior in shoulders revised from failed rTSA compared with failed HA or failed aTSA (P < .001). At one year, the overall cohort reported mean ASES of 61.3 &#xb1; 25.0, with 48.8% achieving Patient Acceptable Symptom State. Shoulders revised from failed rTSA had lower ASES scores (53.3 &#xb1; 26.2) than those revised from failed HA (63.3 &#xb1; 23.1) or failed aTSA (64.3 &#xb1; 25.0) (P = .008). Of the 320 revision rTSAs, 58 (18.1%) required secondary revision; mechanisms of failure included instability (5.0%), humeral loosening (4.1%), infection (4.1%), glenoid failure (3.4%), and periprosthetic fracture (1.6%). Of these, 17 (29.3%) underwent tertiary revision. Secondary revisions due to humeral loosening showed the highest hazard for tertiary failure. Revision rTSA for mechanical failure demonstrates modest long-term durability and functional improvement, with more than 70% survivorship at 10 years. However, less than half of patients achieved acceptable symptom states at short-term follow-up. Specifically, revision from failed rTSA yields inferior survivorship and PROs when compared to failed aTSA or failed HA. Secondary revision procedures carried greater risk of failure, with humeral loosening remaining particularly difficult to manage. These findings highlight the need for careful patient counseling and ongoing investigation into strategies to improve survivorship after revision rTSA.