The performance of Khorana risk score for prediction of venous thromboembolism in patients with lung cancer: A retrospective cohort study.

Introduction: Cancer-related venous thromboembolism is one of the leading causes of mortality and morbidity in cancer patients. Lung cancer is the second most common cancer in the world and is closely related to venous thromboembolism. Venous thromboembolism affects survival in patients with cancer and it is important to be able to predict the possibility of thrombosis in patients with cancer. It was aimed to evaluate the predictive performance of the Khorana risk score in patients with lung cancer.

Materials and methods: The medical data of the patients followed up with lung cancer were analyzed retrospectively. Venous thromboembolism events in lung cancer patients were described. The relationship between the Khorana risk score and the risk of venous thromboembolism was investigated using the cumulative incidence function with compared risk models.

Result: Eight hundred fourteen lung cancer patients were included in the study. Venous thromboembolism was detected in 79 (9.7%) of the patients. Sixty one (77.2%) of the patients had pulmonary embolism, 15 (19%) had peripheral deep vein thrombosis and three (3.8%) had venous thrombosis of other sites. The cumulative incidences of venous thromboembolism for high and intermediate Khorana risk scores were 10.1% and 9.7%, respectively (p= 0.09). The cumulative incidences of venous thromboembolism at 3, 6, 12, and 24 months were 4.7%, 5.8%, 6.4%, and 9.6% for the high-grade Khorana risk score; 4.6%, 5.7%, 6.3% and 7.8% for the intermediate Khorana risk score (p= 0.11).

Conclusions: The Khorana risk score was not found useful in the risk stratification of venous thromboembolism (intermediate or high risk) in patients with lung cancer. New scoring systems are needed to calculate the risk of venous thromboembolism in patients with lung cancer.