Timing and prognostic impact of thromboembolic events in lung cancer patients treated with immune checkpoint inhibitors: A time-dependent real-world analysis.

Abstract

11124 Background: Thromboembolic events (TEEs) are complications associated with malignancy, and emerging evidence suggests immune checkpoint inhibitors may further increase thrombotic risk; however, the prognostic impact of thromboembolic events in the era of immune checkpoint inhibitors, where inflammation and vascular toxicity may be amplified, remains unclear. We analyzed the incidence, timing, and impact on survival of thromboembolic complications in lung cancer patients treated with immunotherapy at a single tertiary academic institution. Methods: We conducted a retrospective study of adult lung cancer patients treated with immunotherapy at a single rural academic center (2015–2025). Thromboembolic events were classified as arterial (myocardial infarction, stroke) or venous (deep vein thrombosis, pulmonary embolism). Clinical characteristics and VAN Index scores were compared by TEE status, and overall survival was assessed using Kaplan-Meier and time-dependent Cox regression to address immortal time bias. Results: Altogether, 505 lung cancer patients received immunotherapy during the study period; patients were predominantly White (76.6%) and male (56.4%), with 21.4% identifying as Black. 22.8% experienced at least one TEE. Histologic subtypes included 59% with non-small cell lung cancer and 41% with small cell lung cancer. The mean age at diagnosis was 65 years. Among patients with TEE, 63.5% experienced arterial and 40.0% venous events. Some patients experienced both venous and arterial events during follow-up. The mean time to first TEE was 404 days, with venous events occurring earlier in the disease course. Patients with TEE had significantly higher comorbidity burden (9.28 vs 6.51) and thrombotic risk by VAN Index (31.99 vs 23.69; both p < 0.001). While unadjusted Kaplan-Meier analysis showed no survival difference (22 vs 19 months; p = 0.126) and multivariable Cox models treating TEE as a baseline exposure showed no independent association with survival (aHR 1.05, 95% CI 0.80-1.38), time-dependent Cox modeling demonstrated a significantly increased mortality risk following TEE (aHR 2.45, 95% CI 1.85-3.26; p < 0.001). Conclusions: Thromboembolic events occurred in nearly one-quarter of lung cancer patients treated with immunotherapy and were associated with higher comorbidity and thrombotic risk. Although standard analyses showed no survival difference, time-dependent modeling demonstrated significantly increased mortality following TEE. These findings demonstrate that failure to account for TEE as a time-varying exposure may substantially underestimate its prognostic impact, underscoring the need for prospective dynamic risk assessment and thromboprophylaxis strategies during immunotherapy.