Occult Infection Risk with Stem Retention in Revision Anatomic to Reverse Shoulder Arthroplasty

In revising failed anatomic shoulder arthroplasty (aTSA) with a convertible humeral component to reverse shoulder arthroplasty (rTSA), surgeons must decide whether to retain or exchange the humeral stem. While retention is less morbid, it may risk missing occult infection in certain cases. The purpose of this study was to describe the risk of occult infection, microbiological findings, and clinical outcomes among patients undergoing conversion of aTSA to rTSA with stem retention using a convertible humeral component.

A single institution retrospective case series of patients undergoing conversion from aTSA to rTSA between 12/1/2013 and 7/1/2025 with minimum >1 year follow up was performed. All patients had a convertible humeral stem in which the stem was retained during the procedure. Patients with a clear preoperative diagnosis of infection (pus noted intra-operatively or a sinus tract) were excluded. Variables of interest included 2025 preoperative ICM score, complications, reoperations, postoperative infections, and patient-reported outcomes (VAS, SANE, ASES).

A total of 65 patients met inclusion criteria. The mean preoperative ICM score was 0.43±1.17, with all patients classified in the "Unlikely" PJI risk tier (<10 points). Intraoperative cultures were identified in 14% of the cases, 89% of which were Cutibacterium acnes. Overall, 6 patients (9%) received postoperative oral or IV antibiotics. However, only 2 patients (3.1%) underwent reoperation for infection, neither of which were found to have positive intra-operative cultures at the time of their index revision. Unplanned re-operations occurred in 12 (18.5%) of the cohort. Preoperative ICM scores were not associated with reoperation for infection (p=0.5). At a mean clinical follow up of 34 months, the mean SANE and ASES scores were 60.6± 29.6 and 61.1± 27.4, respectively.

Among patients with low preoperative concern for infection, retaining a well-fixed humeral stem in revision aTSA to rTSA results in a low rate of PJI. Despite all revisions being characterized as "unlikely" based on the ICM criteria, approximately 1 in 7 patients were found to have unexpected positive intraoperative cultures. While this did not appear to affect postoperative infection rates, high vigilance for infection, even in cases with low preoperative concern, should be considered.