Why do primary anatomic total shoulder arthroplasties fail today? A systematic review and meta-analysis

Anatomic total shoulder arthroplasties (aTSAs) are increasingly used to treat glenohumeral osteoarthritis, offering favorable long-term outcomes. Despite these successes, contemporary failure rates remain significant, warranting a comprehensive analysis of the causes of primary aTSA failure to enhance patient selection criteria, refine surgical techniques, and inform improvements in implant design.

This PROSPERO-registered systematic review and meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. A literature search was performed on PubMed, Cumulative Index to Nursing and Allied Health Literature, Embase, and SPORTDiscus, covering publications from 2014 to 2024. Data extraction focused on failure modality, and subgroup analyses were performed for early and late-term follow-up, study type, and metal-backed or all-polyethylene glenoid implants. Proportional meta-analysis was conducted using random-effects models and presented with 95% confidence intervals (CIs).

A total of 44 studies involving 35,168 aTSA procedures and 2744 failures were included. Implant loosening was the most common failure mode (26.1%, 95% CI [18.9%-34.9%]), with 21.7% (95% CI [14.4%-31.3%]) of failures attributed to glenoid loosening. Rotator cuff insufficiency accounted for 17.3% (95% CI [10.9%-26.4%]) of failures, followed by infection (10.2%, 95% CI [7.6%-13.4%]) and instability (10.4%, 95% CI [7.0%-15.1%]). Early-term failures (<5 years) were predominantly due to implant loosening (30.1%, 95% CI [15.9%-46.3%]), rotator cuff insufficiency (21.6%, 95% CI [9.9%-40.8%]), and infection (12.6%, 95% CI [10.6%-15.0%]), whereas late-term failures (>5 years) were most often attributed to implant loosening (52.9%, 95% CI [23.8%-80.2%]), primarily of the glenoid component (46.6%, 95% CI [25.0%-69.6%]). Failures in metal-backed glenoid implants were predominantly attributed to implant loosening (all were glenoid loosening) (18.4%, 95% CI [6.3%-43.1%]), rotator cuff insufficiency (9.2%, 95% CI [2.9%-25.6%]), and instability (7.0%, 95% CI [1.7%-24.2%]). Failures in all-polyethylene glenoid implants were predominantly due to rotator cuff insufficiency (39.2%, 95% CI [16.0%-68.7%]), infection (12.5%, 95% CI [5.7%-25.2%]), and instability (12.5%, 95% CI [5.7%-25.2%]).

This study identified implant loosening (26.1%), particularly of the glenoid component, as the most common cause of contemporary aTSA failure, followed by rotator cuff insufficiency (17.3%), instability (10.4%), and infection (10.2%). These findings underscore the multifactorial etiology of aTSA failures and emphasize areas for improvement within aTSA survivorship.