Real world analysis of VTE incidence in lung cancer: A comprehensive assessment of the Khorana score and other clinical factors in predicting VTE incidence.

Abstract

e23290 Background: Patients with lung cancer have increased risk of venous thromboembolism (VTE). ASCO guidelines do not support routine VTE prophylaxis but encourage use of risk-stratification tools such as the Khorana Risk Score (KRS) to determine its need. We investigated the efficacy of the KRS and examined other clinical variables including demographics, actionable genomic alterations (AGA), and treatment regimens in predicting VTE risk in a real-world population of patients with advanced lung cancer. We also determined the impact of developing a VTE on healthcare utilization and overall survival (OS). Methods: We performed a retrospective analysis of patients with advanced lung cancer treated at our institution between 2015 and 2023. Our cohort included patients with extensive stage small cell lung cancer (ES-SCLC) and stage III and IV non-small cell lung cancer (NSCLC). Patient demographics, genomics, treatment information, and outcomes were obtained via review of the medical record. Chi square analysis was used to assess the correlation between KRS and incidence of VTE. Results: 552 patients with advanced lung cancer, including 67 patients with ES-SCLC and 485 patients with NSCLC were analyzed. We observed VTE events in 123 patients (22.3%) with a median follow up period of 18.6 months. VTE occurred in 8 patients with ES-SCLC (11.9%) and 115 patients with NSCLC (23.7%). 51 patients developed deep vein thrombosis (DVT, 41.5%), 50 developed pulmonary embolism (PE, 40.7%),13 developed concurrent DVT and PE (10.6%), and 9 developed other types of VTE (7.3%). Median time to VTE was 251 days (range 39 – 1,108 days). Age less than 66 years (p = 0.0001), black race (p = 0.025), NSCLC histology (p = 0.03), and the presence of either EGFR, ALK, RET, or MET mutations were independently associated with increased risk of VTE in our patient population. Chi square analysis of VTE incidence did not show a correlation between low (0-1) and high (≥2) KRS across all patients (Pearson correlation coefficient 0.200). VTE incidence in patients on anticoagulation prior to treatment initiation was 5.6% versus 24.8% for those not on anticoagulation (p = 0.001). Among patients with a VTE, 40% had an ED visit, of which 80% resulted in hospitalization. Five hospitalized patients (4.1%) experienced thrombosis related mortality. Overall survival was not impacted by VTE diagnosis. Conclusions: KRS assigns intermediate risk (1 point) to patients with lung cancer, however, there are many distinct types of lung cancer. In our study population, KRS did not correlate with risk of VTE. We observed higher rates of VTE in patients with NSCLC vs SCLC. In patients with NSCLC, the presence of an AGA further increased their risk. Our results indicate that risk models need to account for the histologic and biologic heterogeneity among lung cancers to better identify patients at highest risk for VTE.