Symptomatic aseptic loosening of a short humeral stem following anatomic total shoulder arthroplasty

The rate of aseptic humeral component loosening with standard-length stems is historically so low (0.3%) that presence of humeral loosening was given substantial weight in defining periprosthetic shoulder infection by the International Consensus Meeting (ICM) in 2018. This study aims to confirm that the historically low rate of humeral stem aseptic loosening is not affected by the adoption of a novel short humeral stem.

Following institutional review board approval, this retrospective multi-institutional study was undertaken. A review of anatomic total shoulder arthroplasty (TSA) cases at 3 institutions provided 184 shoulders that received TSA with a grit-blasted, rectangular short humeral stem without ingrowth coating (Univers Apex; Arthrex). The average patient age was 62.1 years (range: 30-84), and 57.1% (105/184) were male. One-year radiographic follow-up was achieved in 64.7%. Patient clinical charts were reviewed for onset of newly painful shoulders, radiographic evidence of loosening, and return to the operating room for any reason. All patients with concern for clinically significant humeral loosening underwent workup for periprosthetic shoulder infection. Immediate postoperative films were reviewed to identify any differences in prosthetic canal fit. The definition of periprosthetic shoulder infection used was provided by the recent ICM shoulder guidelines.

Twenty-three (12.5%) patients presented with a painful shoulder and radiographic concern for potential humeral loosening at a mean follow-up of 1.5 years (range: 1.5 months-3.4 years). Thirteen (7.1%) of these underwent revision shoulder arthroplasty where a loose stem was confirmed. All revisions underwent tissue culture, and 3 cases were consistent with probable or possible periprosthetic infection at the time of revision arthroplasty. The rate of symptomatic aseptic humeral loosening in this series was 10.9% (20/184), with 5.4% undergoing revision surgery. Patients with symptomatic aseptic humeral loosening were more likely to be male (90.5%) than those patients without symptoms (52.8%, P < .001). There were no differences in canal fit between patients with concern for symptomatic loosening and those with pain-free, stable implants.

In this multisurgeon, multicenter study, the early humeral loosening rate for this stem design far outpaces previously reported rates, and this study likely under-reports the true incidence of clinically significant loosening as it only contains limited short-term follow-up. The perception that humeral loosening is nearly pathognomonic for periprosthetic shoulder infection should be reconsidered for certain short humeral stem designs.