Abstract PS1-11-11: Outcomes in patients with PIK3CA-mutated breast cancer treated with metformin

Abstract Background: Phosphoinositide 3-kinase (PI3K) inhibitors, including alpelisib (ALP), are commonly used to treat metastatic hormone receptor-positive (HR+) breast cancer (BC) with PIK3CA mutations (PIK3CAmut). However, hyperglycemia (HG) is a frequent and challenging side-effect. Metformin (MET) is a first-line treatment (Tx) for HG and has anti-neoplastic properties. We retrospectively reviewed tumor samples and associated clinical outcomes from real-world patients (pts) with HR+/HER2- BC treated with ALP and MET, focusing on prior vs concurrent Tx. Methods: A retrospective review was performed using metastatic BC samples that underwent next-gen sequencing of DNA (592-gene/whole exome) and immunohistochemistry (IHC) at Caris Life Sciences (Phoenix, AZ). IHC (intensity/% cells stained) was used to identify HER2- (≤1+ or ≤10%, or 2+/>10% and CISH-null), HR+ (ER+ (≥1+/≥1%) or PR+ (≥1+/≥1%)) samples. Pathogenic/likely pathogenic PIK3CAmut were identified in 3922 samples, of which 315 pts received ALP but not MET (“ALP alone”) and 198 pts received ALP + MET in various sequences, including MET prior to (but not concurrent with) ALP (n = 41) and MET concurrent with (but not prior to) ALP (n = 86). Endpoints, inferred from claims data, include overall survival (OS, date from first ALP Tx to last contact) and time on treatment (ToT, first to last ALP Tx). Hazards ratios (HR) and p-values were calculated using Cox proportional hazards models and log-rank tests, respectively. ICD10 codes for HG and type 2 diabetes mellitus (T2DM) were used to identify incidence of co-morbidities. Outcomes are reported as: median months (m); HR [95% CI]; p-value. Results: Compared with ALP alone, MET concurrent with ALP was associated with both longer OS (25.1 vs 13.6 m; HR = 0.61 [0.44-0.84]; p = 0.003) and longer ToT (6.0 vs 3.0 m; HR = 0.69 [0.53-0.90]; p = 0.006). However, MET prior to ALP had similar OS and ToT as ALP alone. An intermediate ALP+MET subgroup, MET prior to and concurrent with ALP (n=71), was associated with longer OS compared to ALP alone (22.6 m; HR = 0.59 [0.40-0.87]; p = 0.008), but similar ToT. HG was reported for 94 (29.75%) ALP alone and 60 (69.77%) MET concurrent with ALP pts. Among these, MET concurrent with ALP had significantly longer OS compared to ALP alone (26.1 vs 12.9 m; HR = 0.51 [0.33-0.80]; p = 0.003), as well as longer but not statistically significant ToT (5.7 vs 3.1 m; HR = 0.71 [0.49-1.03]; p=0.071). T2DM was reported for 101 (31.96%) ALP alone and 55 (63.95%) MET concurrent with ALP pts. Among these pts, MET concurrent with ALP was again associated with longer OS compared to ALP alone (28.7 vs 13.5 m; HR = 0.58 [0.37-0.92]; p = 0.0196), as well as longer but not statistically significant ToT (6.0 vs 3.2 m; HR = 0.72 [0.50-1.06]; p = 0.096). Conversely, MET prior to ALP had similar OS as ALP alone (14.3 m; HR = 1.23 [0.74-2.07]; p = 0.424), along with shorter though not statistically significant ToT (1.6 m; HR = 1.37 [0.80-2.35]; p = 0.25). Conclusions: Our finding that MET, when given concurrently with but not prior to ALP, was associated with improved outcomes suggests this therapy combination could have clinical benefits. Further research is warranted to examine this impact as well as any adverse effect profiles. Citation Format: C. Travaline, R. Plagens, A. Elliott, G. Sledge Jr., J. Hundal, J. P. Leone, S. L. Graff, M. Lustberg, M. Vasekar. Outcomes in patients with PIK3CA-mutated breast cancer treated with metformin [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS1-11-11.