Revision of total shoulder arthroplasty to hemiarthroplasty: results at mean 5-year follow-up

Abstract

Background: Although there is a trend to manage failed anatomic total shoulder arthroplasties (aTSA) with revision to a reverse total shoulder arthroplasty, such revisions can be complicated by difficulties in baseplate fixation, instability, and acromial stress fractures. Some cases of failed aTSA may be safely revised to a hemiarthroplasty (HA). The objectives of this study were to report patient-reported outcomes after conversion from aTSA to HA and assess patient and shoulder characteristics associated with a successful outcome.

Methods: Patients who underwent revision from aTSA to HA between 2009 and 2018 were contacted. Patient demographics, surgical history, intraoperative findings, and microbiology results were collected. Patient-reported outcomes were collected with minimum 2-year follow-up. Preoperative radiographic characteristics were reviewed for component positioning and component loosening. Patients with a clinically significant improvement exceeding the minimal clinically important difference (MCID) of the Simple Shoulder Test (SST) were compared with those patients who did not improve past the MCID.

Results: Twenty-nine patients underwent conversion from aTSA to HA with a mean follow-up of 4.5 ± 1.8 years. Intraoperative glenoid or humeral component loosening was found in all 29 patients. Pain improved in 25 of 30 patients (87%), and mean pain scores improved from 6.2 ± 2.3 to 3.1 ± 2.4 (P < .001). SST scores improved from 4.1 ± 3.1 to 7.3 ± 3.2 (P < .001), and 18 of 29 patients (62%) had improvement above the SST MCID threshold of 2.4. The mean American Shoulder and Elbow Surgeons score at the latest follow-up was 64 ± 19, and the Single Assessment Numeric Evaluation score was 65 ± 23. Twenty-two of 29 (76%) patients were satisfied with the procedure. Four patients (14%) required conversion to total shoulder arthroplasty-2 to anatomic and 2 to reverse. An additional 3 patients (10%) had a revision HA performed. No significant differences in patient or shoulder characteristics were found in those patients who improved greater than the MCID of the SST compared patients who improved less than the MCID of the SST. Fifty-nine percent of patients had ≥2 positive cultures with the same bacteria, and 82% of these were with Cutibacterium. Seven of 8 patients (88%) with a loose humeral component had ≥2 positive cultures with the same bacteria.

Discussion: Component loosening is a common failure mode after aTSA. Revision to HA can improve pain and patient-reported outcomes in most patients.

Keywords: Total shoulder arthroplasty; anatomic arthroplasty; hemiarthroplasty; revision shoulder arthroplasty.