Background: Patients with venous thromboembolism (VTE) and cancer are at higher risk of adverse outcomes (mortality, recurrent VTE etc.) vs patients with cancer alone; as such, clinical guidelines recommend anticoagulant treatment for patients with VTE and cancer. There is limited real-world data about how anticoagulant treatment and thromboembolic outcomes differ by tumor type in patients with VTE and cancer. Understanding such differences may help identify appropriate anticoagulant treatment for specific tumor types.
Purpose: To describe anticoagulant treatment patterns and thromboembolic outcomes by tumor type among patients with VTE and cancer.
Methods: Patients with VTE and cancer age ≥65 were identified from the Surveillance, Epidemiology and End Results (SEER) Medicare database from 1/1/2014-12/31/2019. Patients were required to be enrolled for ≥6-months prior to their first VTE diagnosis (index) and without evidence of other conditions requiring anticoagulant use (i.e., atrial fibrillation) prior to index. Cancer status and tumor type were identified from SEER or Medicare database in the 6 months prior through 30 days post VTE. This analysis focused on the following specific tumor types: high-risk (brain, pancreas, and stomach) and common tumor types (breast, and prostate). Patients treated with an anticoagulant within 30 days after index were included in the final population. Major bleeding (MB) and recurrent VTE events were measured during follow-up (index date through earliest of disenrollment, death or 12/31/2019).
Results: A total of 3,546 anticoagulated patients with VTE and cancer of interest met all study criteria (breast [n=1,197], prostate [n=849], pancreatic [n=995], brain [n=248] and stomach [n=257] cancer). Patient mean age ranged from 73 (brain) to 76 (stomach) at index. Anticoagulant treatment patterns varied by tumor type (figure 1). LMWH was more likely to be used in the 3 high risk tumor types whereas apixaban and rivaroxaban were more likely to be used in the 2 common tumor types. The incidence rate of recurrent VTE and major bleeding events also varied among different tumor types: ranging from 1.4 (breast) to 6.4 (pancreatic) per 100 person-years for recurrent VTE and from 4.3 (prostate) to 15.1 (pancreatic) per 100 person-years for major bleeding (figure 2).
Conclusion: There are notable variations in anticoagulant treatment patters and the risks of major bleeding and recurrent VTE events by tumor type among patients with VTE and cancer. Further research is needed to understand which anticoagulant treatment option is more appropriate for VTE patients with specific tumor type.