Michalek, Joel

Abstract Introduction: Endocrine therapy (ET) is central to treating hormone receptor-positive breast cancer (BC), but its side effects impact adherence and quality of life, especially in patients with metabolic dysfunction. Sodium-glucose cotransporter-2 (SGLT-2) inhibitors offer cardiometabolic benefits, potentially improving inflammation, glycemic control, and physical well-being. However, data on SGLT-2 use in BC patients is scarce. This study is the first to evaluate the real-world effects of SGLT-2 inhibitors on all-cause mortality and ET tolerability in BC patients. Methodology: We conducted a real-world analysis using the TrinetX database, which includes data from 184 million patients in 156 healthcare organizations, to identify BC patients on ET (letrozole, anastrozole, exemestane, tamoxifen) who received SGLT-2 inhibitors vs. those who did not. Patients were grouped by age (≤50 and ≥51 years as a surrogate for menopausal status. Primary outcome: all-cause mortality from the start of the index event, defined as the first day meeting BC diagnosis and ET use, with or without SGLT-2. Secondary endpoint: ET tolerability. Propensity score matching completed by TriNetX adjusted for confounders. Multivariate analysis yielded odds ratios and risk differences with 95% CI. Results: Among 1,136 BC patients ≤50 years, 568 were assigned to each group (SGLT-2 vs non-SGLT-2). For patients ≥51 years (n=28,600), each group included 14,300. Demographics across cohorts were comparable. No mortality benefit was seen in premenopausal patients. In postmenopausal women, SGLT-2 use reduced all-cause mortality—647/13,676 events (SGLT-2) vs 953/13,517 (non-SGLT-2) [OR: 0.655 (0.591-0.726)]. In premenopausal patients, SGLT-2 reduced endometrial cancer risk [RD: -0.018 (-0.029, -0.007)] but increased diarrhea risk [OR: 2.305 (1.078, 4.930)]. No significant differences were found in hot flashes, PE, DVT, depression, osteoporosis, joint pain, fatigue, nausea/vomiting, abdominal pain, or dizziness. Among postmenopausal patients, SGLT-2 reduced PE [OR: 0.757 (0.605-0.948)] and osteoporosis [OR: 0.877 (0.784-0.981)], but increased risks of depression [OR: 1.153 (1.007-1.319)], nausea/vomiting [OR: 1.120 (1.001-1.253)], diarrhea [OR: 1.193 (1.058-1.345)], and dizziness [OR: 1.234 (1.103-1.380)]. No significant differences were found for hot flashes, DVT, EC, joint pain, fatigue, or abdominal pain. Conclusion: This analysis reveals novel evidence supporting the cardiometabolic potential of SGLT-2 inhibitors in hormone receptor-positive BC, particularly postmenopausal populations. These agents were linked to lower all-cause mortality and improved tolerance of specific ET-related effects. However, certain adverse risks require caution. Future prospective studies are essential to validate their role and safety in oncologic care. Citation Format: C. Jones, J. Michalek, E. Obomanu, Y. Arya, A. Syal, S. Haddad, V. Kaklamani, S. Shaikh. Evaluating all-cause mortality and endocrine therapy tolerability with sodium glucose cotransporter-2 inhibitors in hormone receptor-positive breast cancer: A real-world analysis [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 PS1-05-28.

Abstract Introduction: Endocrine therapy (ET) improves survival in hormone receptor-positive breast cancer (BC) but has side effects that may affect compliance. GLP-1 receptor agonists (GLP-1 RA), used for diabetes and obesity, possess anti-inflammatory properties and may reduce cancer incidence. Their impact on all-cause mortality and ET tolerability remains underexplored. This real-world analysis is the first of its kind to examine the effects of GLP-1 RA on these outcomes. Methods: We used de-identified data across 130 million patients and 108 health systems from the TriNetX database to identify obese women (BMI ≥30) with BC receiving ET. Patients were grouped by GLP-1 RA exposure into age cohorts (≤50 and ≥51 years, as a surrogate for menopause). The primary endpoint was all-cause mortality; the secondary endpoint was ET tolerability. The index event was the first day meeting all criteria: BC and obesity diagnoses, and initiation of ET ± GLP-1 RA. Patients with outcomes before the study window were excluded. Propensity score matching performed by TriNetX adjusted for demographics, diabetes, comorbidities, cancer stage, and prior chemotherapy. Multivariate logistic regression assessed associations (risk ratios [RR], 95% CIs). Results: After matching, 2,574 patients were assigned to the age ≤50 group: Cohort A (GLP-1 RA exposure, n=1,287) and Cohort B (non-exposure, n=1,287). 13,012 patients were in the age ≥51 group: Cohort C (GLP-1 RA exposure, n=6,506) and Cohort D (non-exposure, n=6,506). In the ≤50 group, the mean age at index event was 42.3 ± 5.2, 99.5% female, 68.9% White, 16.9% Black, and a mean BMI of 37 in Cohort A, with similar demographics and BMI in Cohort B. In the ≥51 group, there was a mean age at index event of 61.8 ± 7.7, 99% female, 71.7% White, 17.4% Black, and a mean BMI of 37 in Cohort C with similar demographics and BMI in Cohort D. GLP-1 RA use significantly reduced all-cause mortality in all cohorts. In the age ≤50 group, there were 29/1,242 events in Cohort A and 71/1,258 in Cohort B [RR 0.414 (0.271-0.633)]. In the age ≥51 group, there were 240/6,287 events in Cohort C and 400/6,285 in Cohort D [RR 0.600 (0.513-0.701)]. GLP-1 RA use worsened ET side effects. In the age ≤50 group, GLP-1RA increased the risk of hot flashes [RR: 1.473 (1.235-1.756)], venous thromboembolism (VTE) [RR: 1.439 (1.090-1.902)], nausea/vomiting [RR: 1.460 (1.242-1.717)], abdominal pain [RR: 1.353 (1.184-1.545)], diarrhea [RR: 1.319 (1.096-1.589)], joint pain [RR: 1.382 (1.243-1.536)], and depression [RR:1.634 (1.382-1.932)]. No significant risk was found in endometrial cancer (EC) and osteoporosis. In the age ≥ 51 group, GLP-1 RA use increased hot flashes [RR: 0.1.239 (1.091-1.407)], [RR:0.591 (0.482-0.724)], nausea/vomiting [RR: 1.429 (1.324-1.542)], abdominal pain [RR: 1.310 (1.227—1.399)], diarrhea [RR: 1.446 (1.328-1.575)], joint pain [RR: 1.334 (1.270-1.400)], osteoporosis [RR: 1.110 (1.026-1.201)], and depression [RR: 1.536 (1.413-1.670)]. No significant differences were noted in VTE and EC risk. Conclusion: This real-world analysis demonstrates that GLP-1RA use is associated with a significant reduction in all-cause mortality among obese BC patients undergoing ET. However, GLP-1 RA exposure appears to exacerbate several ET-related adverse effects. Future prospective studies are needed to determine these associations as they may warrant clinical consideration. Citation Format: C. Jones, J. Michalek, E. Obomanu, Y. Arya, A. Syal, S. Haddad, V. Kaklamani, S. Shaikh. Real-world Analysis on the Impact of GLP-1 Receptor Agonists on All-cause Mortality and Endocrine Therapy-Tolerability in Obese Breast Cancer Patients [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 PD8-06.

Background Hepatosplenic T-Cell lymphoma (HSTCL) is an aggressive type of lymphoma with extremely low incidence. It has a high fatality rate with a 5-year overall survival (OS) of <10% (Am J Surg Pathol, PMID: 26872013). Although there is no therapy that has proven to produce sustained remission, recent studies suggest improvement in outcomes with allogeneic stem cell transplant (allo-SCT) (Leukemia, PMID: 25234166). Clinical characteristics and therapeutic outcomes have been previously reported; however, documentation of outcomes in minorities such as Hispanics (HI) are lacking. Given the growth of the HI population in the US it is imperative to understand the difference in outcomes and patterns of care (CA Cancer J Clin, PMID: 30285281). The present study aims to fill this gap in the literature; therefore, the present population-based study evaluates the impact of ethnicity in clinical outcomes for HSTCL. Material and Methods We retrospectively searched for cases of HSTCL recorded in the Texas Cancer Registry. Inclusion criteria included and established pathologic diagnosis of HSTCL using the International Classification of Diseases for Oncology Third Edition (ICD-O-3) code list. Patients were divided between HI and non-Hispanics (NH). Demographic, socioeconomic, clinical and survival outcomes were recorded. Statistical analysis included Fisher's Exact test, Pearson's Chi-square test, T-test or Wilcoxon test. Survival analysis was calculation in years from date of primary diagnosis to date of death or last date of follow up. Survival distributions were described with Kaplan-Meier curves, and long rank testing was used to assess significance of variation in median survival with ethnicity. All statistical testing was two-sided with a significance level of 5%. The R language [R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria] was used throughout. Results From 2006-2016, a total of 25 patients were diagnosed with HSTCL; 23 (92%) were NH and 2 (7%) were HI [<0.0001]. The median age at diagnosis was 46 for HI and 48 for NH. It was predominantly seen in males for both HI and NH. The patients mostly self-identified as white (HI: n=2, Nh: n=16), and there were 7 (30%) NH patients that self-identified as black. For HI, one patient was in the bracket of poverty indicator of 10-19.9% and the other one was in the 20-100%; for the NH patients the bracket that prevailed was 10-19.9% [p-value 0.56]. Both HI patients had private insurance (PI); insurance status on NH patients included 7 (33.3%) with PI, 5 (23.8%) with Medicare, and 7 (33.3%) was unknown [p-value 0.64]. The majority of HI and NH patients were located in the metropolitan, non-border area. Stage at diagnosis for NH was mostly III-IV; one HI had a stage III-IV HSTCL and the other was unstaged [p-value 1.0]. From a therapeutic standpoint, the majority of HI and NH patients received chemotherapy, with 13 patients (HI: n=2, NH: n=11) receiving multiple agents as first-line therapy [p-value 0.84]. From the two groups, only 3 (13%) NH patients underwent allo-SCT [p-value 0.014]. The majority of HI and NH did not undergo radiation (Figure 1A). The median overall survival was 0.5 years in HI and 0.6 in NH. The survival probability at 2 years for HI and NH was 0.5 (CI=0.16, 1) and 0.29 (CI=0.15, 0.56) respectively; at 5 years for HI was 0.5 (CI 0.16, 1) and for NH was 0.23 (CI=0.10, 0.51) (Figure 1B). The overall survival probability at 10 years was not statistically different for HI vs NH [p-value 0.53] (Figure 1C). Conclusions Survival analysis shows that HSTCL is usually fatal with a median overall survival of less than a year, and no difference in survival probability at 10 years for both HI and NH. The majority of patients in our study were non-Hispanic white males. A significant finding was that NH patients had access to allo-SCT as opposed to the HI patients. No statistically significant difference in survival was noted depending on the insurance or poverty levels. No other demographics or sociocultural variables seemed to have an impact in outcomes. The main limitation of our study is our sample size that limits generalizability and power of our statistical analysis. Further studies using multiple cancer registries are warranted to assess the clinical characteristics and survival outcomes in HI with HSTCL. Disclosures Diaz Duque: ADCT Therapeutics: Research Funding; Molecular Templates: Research Funding; AstraZeneca: Research Funding; Hutchinson Pharmaceuticals: Research Funding; Seattle Genetics: Speakers Bureau; Verastem: Speakers Bureau; AbbVie: Speakers Bureau.