Background Patients with lung cancer are at high risk of developing venous thromboembolism (VTE). Although several risk assessment models (RAMs) for cancer-associated VTE exist, their performance in advanced lung cancer is uncertain. This study investigated VTE incidence and risk prediction in patients with advanced lung cancer.

Methods In a subinvestigation of the Vienna Cancer and Thrombosis Study (CATS), a prospective cohort study investigating risk factors for cancer-associated VTE, data from patients with advanced lung cancer initiating chemotherapy were analyzed.

Results Of 277 patients (median age 61 years [IQR 56–67], 37% female; 89% received palliative treatment), 25 developed VTE during the 2-year observation period; with 24 events occurring in the first year. The cumulative VTE incidence was 6.6% (95% confidence interval [CI] 4.0–10.0), 9.3% (6.2–12.3), and 9.8% (6.5–13.8) at 6, 12, and 24 months, respectively. While several RAM (Khorana score, CONKO, PROTECHT, m-LCPI, HYPERCAN model) did not predict risk of VTE in our cohort, the CATScore identified patients at high risk of VTE (subdistribution hazard ratio [SHR, 95% CI] for CATScore ≥8% risk vs. <8% risk: 3.43 [1.11–10.55], p = 0.032). Laboratory parameters associated with risk of VTE included higher D-dimer, higher CRP, and lower platelet counts. Based on these biomarkers, we exploratively derived a simplified biomarker-based risk model.

Conclusions In this cohort of patients with advanced lung cancer, most established RAM showed limited discriminative ability. The CATScore and a newly derived biomarker-based model demonstrated signals of risk stratification. These findings require confirmation in larger independent cohorts.

Received: 26 February 2026

Accepted after revision: 21 May 2026

Article published online:
09 June 2026

© 2026. Thieme. All rights reserved.

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany