External Validation of Risk Scores for Predicting Venous Thromboembolism in Ambulatory Patients with Lung Cancer.

Simple Summary: People living with cancer are at a higher risk of developing a venous thromboembolism (VTE). A VTE may interrupt anticancer therapy, which increases the risk of mortality. Cancer treatment guidelines recommend the use of preventive anticoagulants only in those at high risk of a VTE, identified by a valid risk assessment tool, such as the Khorana score. However, the Khorana score at its original cut-off value of 3 points has a low sensitivity and discriminatory capability in lung cancer patients, and the updated Khorana score with a 2-point cut-off value lacks validation in this population. Other risk scores, such as the PROTECHT, CONKO, and COMPASS-CAT scores, have not been validated in large lung cancer cohorts. The aim of this study was to evaluate these four existing risk scores in this patient population. We validated the Khorana score with a cut-off value of 2 points and the CONKO score with a cut-off value of 2 points. Background: The purpose of this study was to evaluate the discriminatory capability of the Khorana, PROTECHT, CONKO, and COMPASS-CAT scores in ambulatory patients with lung cancer. Methods: This retrospective cohort study included 591 patients with newly diagnosed lung cancer. A symptomatic or incidental VTE occurred in 108 patients. Results: The Khorana score at a 2-point threshold had a discriminatory capability with an odds ratio (OR) of 1.80 and an AUC of 0.57 for 6 months, and an OR of 1.51 and an AUC of 0.55 for 12 months. The CONKO score at a 2-point threshold had a stronger discriminatory capability for both 6 months and 12 months with ORs of 3.00 and 2.13, and AUCs of 0.63 and 0.59, respectively. Additionally, higher white blood cell counts, higher neutrophil counts, hypoalbuminaemia, and not undergoing lung surgery were related to VTE occurrence (p < 0.05). Conclusions: The Khorana score with the 2-point threshold was validated in ambulatory patients with lung cancer, with the results indicating a decline in its discriminatory capability over time (at 12 months vs. 6 months from diagnosis). The CONKO score at the original 2-point threshold showed a stronger discriminatory capability but further validation with a larger sample size is recommended. The identified predictors should be further investigated in future research. [ABSTRACT FROM AUTHOR]