Concurrent GLP1R-agonist use with chemoimmunotherapy for early-stage triple-negative breast cancer.

Abstract

1115 Background: Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have emerged as a key class of drugs for treating type 2 diabetes mellitus (DM2) and obesity. GLP-1 is rapidly degraded by DPP4, which led to the development of DPP4 inhibitors (DPP4i). Prior work has shown GLP-1R in tumor cells activates key growth signaling and GLP-1RA likely dampen inflammation. This suggests that GLP-1R activation may influence response rates to chemoimmunotherapy. This study aims to investigate the impact of GLP-1RAs and DPP4i (GLP1 drugs) exposure on pathological complete response (pCR) rates for patients with early-stage triple negative breast cancer (TNBC) receiving neoadjuvant chemoimmunotherapy. Methods: Patients with early-stage TNBC diagnosed between July 1, 2021, and December 31, 2023, who received the KEYNOTE-522 regimen were identified at three institutions. Patients using GLP-1RAs and DPP4i at breast cancer diagnosis and throughout the neoadjuvant period, alone or with other diabetes medications, were included. Those who started or discontinued GLP-1 drugs during chemoimmunotherapy were excluded. Group comparisons were made using Chi-square and two-sample t-tests. Human TNBCs were analyzed by IHC and CosMx 6000-plex spatial transcriptomics. Results: Among 343 patients, 7.5% were using GLP-1 drugs. The pCR rate among patients exposed to GLP-1 drugs was 30.8% compared to 64.4% in those not exposed (p = 0.001). For patients using other classes of DM2 medications (n = 46), the pCR rate was 65.2%, while for those not taking any DM2 medications (n = 271), the pCR rate was 64.2%. In univariate analysis, patients exposed to GLP-1 drugs were significantly older than non-exposed (median age: 60 vs. 51 years; p = 0.009), had a higher BMI (35.0 vs. 28.9 kg/m²; p = 0.002), and had higher rates of DM2, hypertension, and hyperlipidemia. In multivariate analysis, only age was associated with pCR (OR: 0.97, 95% CI: 0.96-0.99, p = 0.007). When comparing patients taking GLP-1 drugs with those using other DM2 medications, no significant differences were observed regarding age, BMI, or clinical T or N stage. To evaluate tumor-intrinsic factors that may influence treatment response, we examined TNBC specimens (n = 84) and identified GLP-1 receptor expression in tumor cells in 35.7% of cases and in the tumor microenvironment in 60.7% of cases. A spatial transcriptomics atlas of GLP-1 drug-exposed tumors (469,029 cells) provides evidence of GLP-1 pathway activity in both malignant and non-malignant cells of the tumor microenvironment. Conclusions: We observed significantly lower pCR rate among patients taking GLP1 drugs during neoadjuvant chemotherapy for TNBC. These effects were not observed with other diabetic medications. Detection of GLP1R expression in TNBC specimens indicates there may be direct and indirect effects of agonists to the GLP1 pathway on chemoimmunotherapy response rates.