AUTHORS
Colton Jones, MD, Hematology, Oncologyfellow, Jacob Schmelz, MD2, Vanessa Velazquez, MD3, Dominic Earl Mitchell Jarrett, MD4, Usha S Perepu, MD5
AFFILIATIONS
1 University of Texas Health Science Center San Antonio, San Antonio, United States
2 University of Texas Health Science Center San Antonio, San Antonio, United States
3 University of Texas Health Science Center San Antonio, San Antonio, United States
4 University of Miami, Miami, United States
5 University of Texas Health Science Center San Antonio, San Antonio, United States
ABSTRACT
Background
Cancer-associated thrombosis is the second leading cause of death in patients with cancer. Obesity further amplifies venous thromboembolism (VTE) risk and is etiologically linked to 13 obesity-associated cancers (OACs), creating a population with a synergistic thrombotic burden. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have demonstrated antiplatelet and antithrombotic properties in preclinical studies, and observational data suggest their use is associated with reduced OAC incidence and lower VTE risk in patients with type 2 diabetes (T2DM). However, there is limited data on whether GLP-1RAs mitigate VTE risk specifically among patients with OACs. This is the first target trial emulation to evaluate the association between GLP-1RA use and incident VTE across the 13 OACs.
Methodology
De-identified data was used from the TriNetX Network, which encompasses over 140 million patients across 114 HCOs. Patients with one or more OACs between January 1, 2010-January 1, 2025, were identified using ICD-10 codes and stratified by GLP-1RA exposure (Cohort A) versus non-exposure (Cohort B). GLP-1RA initiation was required to occur after both cancer diagnosis and a documented diagnosis of obesity or T2DM. Propensity score matching (PSM) (1:1) adjusted for demographics, comorbidities, baseline VTE risk factors, BMI, HbA1c, and cancer stage. The primary endpoint was VTE incidence: deep vein thrombosis (DVT) and pulmonary embolism (PE). Cox proportional hazards regression estimated hazard ratios with 95% CI for incident DVT and PE. A 6-month landmark analysis was applied to mitigate immortal time bias, and sensitivity and subgroup analyses were conducted to assess robustness across cancer types, T2DM status, obesity status, and individual GLP-1RA agents.
Results
After PSM, 208,552 cancer patients were included (n=104,276 per cohort) with a median follow-up of 1,100 days. Cohort demographics were well-balanced: mean age 58.5 years, 79.2% female, 69.0% White, 17.2% Black, and 3.4% Asian. GLP-1 users had a BMI of 36.1 vs 33.6 for controls. Mean A1c was 6.9% for GLP-1RA users vs 6.4% for controls. In multivariable analysis, GLP-1RA use was associated with a significantly reduced risk of PE in patients with colon cancer (HR 0.80; 95% CI 0.64–0.99), gastric cancer (HR 0.47; 95% CI 0.24–0.93), and ovarian cancer (HR 0.54; 95% CI 0.39–0.77). This protective association was also observed among patients with T2DM (HR 0.81; 95% CI 0.74–0.89) and non-obese patients (HR 0.62; 95% CI 0.53–0.73). No significant association with PE risk was observed for breast cancer, endometrial cancer, renal cell carcinoma (RCC), hepatocellular carcinoma (HCC), pancreatic cancer, thyroid cancer, meningioma, or multiple myeloma (MM). No PE events occurred in the esophageal or gallbladder cancer subgroups. For DVT, GLP-1RA use was associated with a significant reduction in patients with ovarian cancer (HR 0.70; 95% CI 0.53–0.93) and pancreatic cancer (HR 0.65; 95% CI 0.46–0.92), as well as in patients with T2DM (HR 0.88; 95% CI 0.82–0.94) and non-obese participants (HR 0.69; 95% CI 0.61–0.77). Notably, GLP-1RA use was associated with a significantly increased risk of DVT in patients with thyroid cancer (HR 1.52; 95% CI 1.12–2.05). No significant associations with DVT were observed for esophageal cancer, colon cancer, breast cancer, endometrial cancer, gastric cancer, RCC, HCC, meningioma, or MM. No DVT events occurred in the gallbladder cancer subgroup. In agent-specific analyses, semaglutide was the only GLP-1RA associated with reductions in DVT (HR 0.82; 95% CI 0.73–0.91), though the reduction in PE did not reach statistical significance (HR 0.87; 95% CI 0.74–1.01). Numerically favorable trends for both DVT and PE were observed with tirzepatide and dulaglutide, though results did not reach statistical significance.
Conclusion
In this clinical trial emulation study, GLP-1RA use was associated with significant reductions in PE and DVT risk in select malignancies, most notably ovarian, gastric, colon, and pancreatic cancers, with semaglutide emerging as the only agent significantly associated with DVT reduction. Limitations include residual confounding, healthy-user bias, reliance on ICD-10 codes for VTE ascertainment without imaging adjudication, and inability to assess GLP-1RA dose-response relationships. These findings support further prospective investigation of GLP-1RAs as potential adjunctive agents for VTE risk mitigation in cancer patients.
DISCLOSURES
Colton Jones, MD, Hematology, Oncologyfellow: No relevant conflicts of interest to declare.