Risk factors for greater tuberosity nonunion following reverse total shoulder arthroplasty for proximal humerus fractures: a systematic review and meta-analysis

Background: Greater tuberosity nonunion following reverse total shoulder arthroplasty (rTSA) for proximal humerus fracture is known to be associated with poor outcomes. However, it is unknown which factors place patients at risk for not healing their tuberosity repair after rTSA. The purpose of this review is to identify and characterize the predictors of greater tuberosity nonunion (GTNU) following rTSA for proximal humerus fractures. Methods: A systematic search was conducted using key terms “tuberosity healing reverse” and “tuberosity nonunion arthroplasty” in databases of English language articles published before April 2024. Human studies describing risk factors associated with development of GTNU following rTSA for fracture in either retrospective or prospective studies were included. Articles were excluded if they were not able to assess risk factors for GTNU. Data synthesis summarized outcome measures and study designs appropriately in the results. Using SPSS meta-analysis binary and continuous model function, the mean effect size estimate (MESE) with 95% confidence interval were reported for significant factors. Results: This search yielded a total of 129 studies and, after exclusion criteria were assessed, 21 articles comprising 1164 patients with proximal humerus fractures were included in the review. These studies included 429 fractures developing tuberosity nonunions after rTSA. Methodological quality varied from low to high. There were 23 factors assessed in the included studies. There were no statistically significant patient-specific risk factors for GTNU, including age, gender, dominant side and Type 2 diabetes (all P > .3). A 4-part fracture was found to be a risk factor (MESE = 1.24, [0.09-2.39]; P = .03) compared to the 3-part, while all other injury-specific risk factors were found to be nonsignificant. The use of autograft (MESE = −1.49, [{−2.57}-{−0.42}]; P = .006) was a significantly protective surgical-specific factor, while the use of a fracture stem vs. a conventional stem and the glenosphere size were not deemed significant risk factors (both P > .2). The majority of studies were of low-quality, with only 10.0% (3/30) of the included studies conveying Level I or Level II evidence. Conclusions: The use of autograft resected from the fractured humeral head and incidence of a 4-part fracture were the most significant factors involved in greater tuberosity nonhealing, while certain demographic and injury-specific factors were not found to be statistically significant. Future studies are warranted to use a prospective study design, identify greater tuberosity nonunion-specific risk factors, and implement evidence-based prevention strategies.