Long-term (minimum 10 years) survival and outcomes of pyrocarbon interposition shoulder arthroplasty

There are some major controversies surrounding the use and longevity of pyrocarbon interposition shoulder arthroplasty (PISA). The objective of this study was to investigate the long-term survival and outcomes (minimum 10-year) following PISA for osteoarthritis (OA) in young and active patients. This was a retrospective review of prospectively collected data of patients who underwent PISA (InSpyre; Tornier-Stryker) for OA between 2009 and 2012. Arthroplasty survival was known for 71 patients followed longitudinally for a minimum of 10 years. The clinical and radiologic outcomes were assessed in 62 patients (62 shoulders) reviewed with radiographs. The mean age at surgery was 60 years (range, 23-72 years), and 31 shoulders (50%) underwent prior surgery before PISA. The diagnosis was primary osteoarthritis (POA = 29), post-traumatic osteoarthritis (PTOA = 23), and postinstability osterarthritis (PIOA = 10). Clinical failure was defined as repeat surgical intervention involving prosthesis revision. Clinical outcomes were assessed with the Constant score (CS) and Subjective Shoulder Value (SSV). The mean duration of follow-up was 11 ± 0.6 years (range, 10-14 years). Overall, the survival rate was 90% (95% confidence interval [CI] 82.8-96.8) at 5 years and 87% (95% CI 79-94.8) at a 10-year follow-up. Survival was 100% in PTOA (type 1 fracture sequelae) and in PIOA as well as 95% in primary OA with type A glenoid. Revision surgery was significantly higher in biconcave (type B2) glenoid (44%) compared with concentric (type A) glenoid (2%), respectively (P = .002). Among the 7 patients who were revised to reverse shoulder arthroplasty, 5 had painful glenoid erosion and 2 had bipolar (glenoid and humeral) erosion with thinning and finally fracture of the greater tuberosity. Two shoulders with glenohumeral erosion were associated with secondary rotator cuff tears (1 supraspinatus and 1 subscapularis tear). The mean time to revision and revision was 4 ± 1.7 years. Glenoid wear was more often superior (81%) than central (19%), P