Two-year outcomes of the reverse humeral reconstruction prosthesis.

Background: Extensive proximal humeral bone loss in the setting of shoulder arthroplasty represents a difficult challenge for the treating surgeon. Achieving adequate fixation with standard humeral prostheses can be problematic. Allograft-prosthetic composites are a viable solution for this problem; however, high rates of complications have been reported. Modular proximal humeral replacement systems are another potential solution, but there is a paucity of outcome data on these implants. This study reports the 2-year minimum follow-up outcomes and complications of a single system reverse proximal humeral reconstruction prosthesis (RHRP) for patients with extensive proximal humeral bone loss.

Methods: We retrospectively reviewed all patients with minimum 2-year follow-up who underwent implantation of an RHRP for either (1) failed shoulder arthroplasty or (2) proximal humerus fracture with severe bone loss (Pharos 2 and 3) and/or sequelae thereof. Forty-four patients met inclusion criteria (average age 68.3 ± 13.1 years). The average follow-up was 36.2 ± 12.4 months. Demographic information, operative data, and complications were recorded. Pre- and postoperative range of motion (ROM), pain, and outcome scores were assessed and compared to the minimal clinically important difference (MCID) and substantial clinical benefit (SCB) for primary rTSA where available.

Results: Of the 44 RHRPs evaluated, 93% (n = 39) had undergone prior surgery and 70% (n = 30) were performed for failed arthroplasty. ROM improved significantly in abduction by 22° (P = .006) and forward elevation by 28° (P = .003). Average pain on a daily basis and pain at worst improved significantly, by 2.0 points (P < .001) and 2.7 points (P < .001), respectively. Mean Simple Shoulder Test score improved by 3.2 (P < .001), Constant score by 10.9 (P = .030), American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form (ASES) score by 29.7 (P < .001), University of California, Los Angeles (UCLA), score by 10.6 (P < .001), and Shoulder Pain and Disability Index score by 37.4 (P < .001). A majority of patients achieved the MCID for all outcome measures assessed (56%-81%). The SCB was exceeded by half of patients for forward elevation and the Constant score (50%), and exceeded by the majority of patients for the ASES score (58%) and UCLA score (58%). The complication rate was 28%; the most common complication was dislocation requiring closed reduction. Notably, there were no occurrences of humeral loosening requiring revision surgery.

Discussion: These data demonstrate that the RHRP resulted in significant improvements in ROM, pain, and patient-reported outcome measures, without the risk of early humeral component loosening. RHRP represents another potential solution for shoulder arthroplasty surgeons when addressing extensive proximal humerus bone loss.