(B-159) Do GLP-1 RA medications change the rate of total knee and hip replacement surgeries? Comparing different GLP-1 RA medications as a proxy of their effect

Background: Total knee and hip arthroplasty (TKA and THA) are among the most frequently performed elective surgeries in the US. Obesity is a risk factor for osteoarthritis (OA) and pain, both increase the probability of TKA or THA. FDA approved several GLP-1 receptor agonists (GLP-1 RAs) for weight loss indication. Clinical trials and real world data analyses suggest the strongest weight loss effect of tirzepatide, followed by semaglutide, then liraglutide.

Objectives: We evaluated the association between tirzepatide, semaglutide or liraglutide use with TKA and THA rates, comparing different GLP-1 RAs as a proxy of their effect to minimize bias when comparing GLP-1 RA users to nonusers.

Methods: We created retrospective cohorts using HealthVerity TaXonomy closed claims which capture patient data linked across 225+ public and commercial payers, thus providing a comprehensive view from GLP-1 RA initiation to TKA/THA.
We included patients who used tirzepatide, semaglutide or liraglutide from Q3 2016 onward. Eligible patients had continuous medical and pharmacy insurance coverage 6 months prior to the first GLP-1 RA fill, through 1 year after. Patients who had TKA or THA surgery 6 months prior to GLP-1 RA fill were excluded (no longer at risk of outcomes). We stratified the patients by presence of Type II diabetes (T2DM) and the presence of OA diagnosis. We measured the incidence of TKA or THA after 90 days through 1 year post GLP-1 RA initiation and compared it by type of GLP-1 RA. A washout period of 90 days reduced reverse causality by excluding GLP-1 RA prescriptions intended as pre-op optimization for planned surgeries.

Results: Cohort size varied across GLP-1 RAs. Tirzepatide as the newest approved drug with weight loss indication had the smallest sample (122,545) compared to semaglutide (932,714) and liraglutide (335,213 patients). The rates of TKA or THA 90 day to 1 year post GLP-1 RA fill were differently associated with the strength of the GLP-1 RA, by T2DM. TKA/THA rates were 0.84%, 0.74% and 0.63% for tirzepatide, semaglutide and liraglutide patients with T2DM, respectively. TKA/THA rates were 0.55%, 0.70% and 0.70% for tirzepatide, semaglutide and liraglutide patients without T2DM, respectively. The different association patterns by T2DM, particularly for tirzepatide, was also seen among OA sub-cohort.

Conclusions: We found an increased rate of TKA/THA by GLP-1 RA strength among patients with T2DM. Among patients without T2DM, tirzepatide (with the strongest strength) showed the lowest rate of TKA/THA. Further research should confirm the association in matched cohorts.