Abstract Background: Alpelisib, a PI3Kα-specific inhibitor, is approved for use in hormone receptor-positive (HR+), HER2-negative advanced breast cancer with PIK3CA mutations. However, its known metabolic side effects, particularly hyperglycemia, raise concerns for patients with pre-existing Type 2 Diabetes Mellitus (T2DM). This study aimed to evaluate real-world treatment patterns and clinical outcomes in breast cancer patients treated with Alpelisib, stratified by diabetes status. Methods: We conducted a retrospective analysis using the TriNetX US Collaborative Network. Two cohorts were identified: (1) patients with breast cancer and T2DM treated with Alpelisib (n=655), and (2) patients without diabetes treated with Alpelisib (n=1,102). Treatment pathways were analyzed separately for each cohort. A comparative outcomes analysis was performed using 1:1 propensity score matching (503 patients per cohort) to balance baseline characteristics. Kaplan-Meier survival analysis was used to evaluate time-to-event outcomes. Patients with prior outcomes were excluded from each respective analysis. Results: After matching, each cohort included 503 patients. Mean follow-up was 562.2 days (SD 484.1) for the T2DM group and 525.1 days (SD 459.7) for the non-diabetic group. Among those without prior death records, 268 of 502 (53.4%) T2DM patients and 274 of 502 (54.6%) non-diabetic patients died during follow-up (median survival: 651 vs. 575 days; HR 0.948, 95% CI: 0.801-1.122, p=0.747). For hospitalization, 67 of 201 (33.3%) T2DM patients and 83 of 233 (35.6%) non-diabetics were hospitalized (median time: 1,178 vs. 1,009 days; HR 0.808, 95% CI: 0.584-1.116, p=0.631). Cancer progression occurred in 24 of 79 (30.4%) T2DM patients and 27 of 75 (36.0%) non-diabetics (median time: 799 vs. 765 days; HR 0.769, 95% CI: 0.443-1.337, p=0.381). Regarding adverse events, T2DM patients had a higher risk of hyperglycemia (95 of 294 32.3% vs. 116 of 463 25.1%; HR 1.336, 95% CI: 1.018-1.752, p=0.036). For hypokalemia, rates were similar (88 of 360 24.4% T2DM vs. 93 of 406 22.9% non-diabetics; HR 0.985, 95% CI: 0.735-1.318, p=0.444). Dermatitis was seen in 35 of 453 (7.7%) T2DM patients vs. 42 of 455 (9.2%) non-diabetics (HR 0.791, 95% CI: 0.505-1.239, p=0.249). Heart failure occurred in 37 of 441 (8.4%) T2DM vs. 34 of 474 (7.2%) non-diabetics (HR 1.120, 95% CI: 0.703-1.785, p = 0.282). Myocardial infarction rates were 19 of 494 (3.8%) T2DM and 17 of 484 (3.5%) non-diabetics (HR 1.055, 95% CI: 0.548-2.031, p=0.593). Diarrhea occurred in 62 of 324 (19.1%) T2DM and 87 of 386 (22.5%) non-diabetics (HR 0.831, 95% CI: 0.600-1.152, p=0.283). Pneumonitis developed in 20 of 468 (4.3%) T2DM vs. 15 of 467 (3.2%) non-diabetics (HR 1.284, 95% CI: 0.657-2.509, p=0.806). Treatment pathway analysis revealed that breast cancer patients with T2DM were seven times more likely to discontinue or not proceed to subsequent therapy after Alpelisib compared to non-diabetics, despite similar follow-up durations. Conclusions: In this large real-world cohort, breast cancer patients with T2DM treated with Alpelisib experienced comparable survival and cancer progression outcomes to those without diabetes. However, they were at significantly higher risk for hyperglycemia and had less complete treatment documentation, potentially reflecting clinical complexity or data capture limitations. These findings underscore the importance of proactive metabolic monitoring and individualized care strategies for patients with comorbid diabetes receiving Alpelisib. Further research is warranted to optimize treatment sequencing and mitigate toxicity in this high-risk population. Citation Format: F. Khan, M. Zafar, B. Singeltary. Comparative Outcomes in Breast Cancer Patients Treated with Alpelisib With and Without Type 2 Diabetes Mellitus: A Real-World Analysis from the TriNetX Network 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 PS1-03-13.
Khan et al. (Tue,) studied this question.