The impact of glenoid polyethylene type, humeral head size and composition on revision risk following primary anatomic total shoulder arthroplasty for osteoarthritis.

Abstract

Glenoid loosening and polyethylene wear remain leading causes of failure after anatomic total shoulder arthroplasty (aTSA). The introduction of highly cross-linked polyethylene (XLPE) aims to reduce wear and improve implant longevity. Humeral head size and its composition (metal or ceramic) may influence glenoid wear patterns, yet its interaction with polyethylene type (XLPE vs non-XLPE) on revision risk has not been well defined. The study aim was to determine the effect of humeral head size and composition on revision risk in primary aTSA and to assess whether this relationship differs by polyethylene type. A retrospective cohort study using data from a national arthroplasty registry between 16 January 2004 and 31 December 2024 included primary aTSA for osteoarthritis using modern prostheses (available in 2024), and excluded total resurfacing anatomic arthroplasty. The cumulative percent revision (CPR) was determined using Kaplan-Meier estimates of survivorship and hazard ratios (HR) from Cox proportional hazard models adjusted for age, sex, humeral component type (stemmed/stemless), glenoid component type (non-modular metal backed glenoid, cemented polyethylene glenoid, hybrid metal/polyethylene glenoid and polyethylene glenoid with modified central peg). The primary analysis compared XLPE and non-XLPE glenoids, stratified by humeral head size. Secondary analyses examined humeral head material (metal vs ceramic). Of 10,431 aTSA procedures, the CPR at 15 years for all XLPE glenoids (n=4,797) was 6.0% (95% confidence interval (CI) 4.4, 8.3), and 11.4% (95% CI 8.6, 15.0) for non-XLPE (n=5,634). XLPE had a lower rate of revision (non-XLPE vs XLPE HR=1.45, (95% CI 1.19, 1.76), p<0.001) adjusted by age and sex, a modest but clinically meaningful increase in revision risk associated with non-XLPE glenoids. After 3.5 years, non-XLPE glenoid components combined with 44-50mm humeral heads had a higher revision rate than the equivalent XLPE combination (p=0.018). However, there was no difference between polyethylene types when the smallest(<44mm) or largest (>50mm) head sizes were used. Similarly, revision rates of XLPE versus non-XLPE aTSA did not differ with a ceramic humeral head when stratified by head size. In contrast, metal/ XLPE bearings showed a significantly higher revision risk than metal/non-XLPE with <44mm head sizes (p=0.018), all with multivariable adjustments. XLPE glenoid components had improved survivorship following primary anatomic total shoulder arthroplasty for osteoarthritis. The observed benefit was modest overall but appeared to be influenced by humeral head size and composition, suggesting that interactions between bearing surface characteristics may contribute to long-term implant performance. III, Retrospective Cohort Comparison using Large Database, Prognosis Study.