Validation of risk assessment models for venous thromboembolism in patients with cancer receiving systemic therapies
Open Access
- 8 July 2025
- journal article
- Published by American Society of Hematology in Blood Advances
- Vol. 9 (13) , 3340-3349
Abstract
Cancer increases the risk of venous thromboembolism (VTE). To identify patients with cancer at high VTE risk who might benefit from primary thromboprophylaxis, several risk assessment models (RAMs) have been developed. The evolution of anticancer treatment, including the implementation of targeted and immunotherapeutic agents, might affect VTE risk and risk prediction. Therefore, we aimed to evaluate the performance of 6 externally validated RAMs (Khorana score, PROTECHT, CONKO, COMPASS-CAT, CATScore, and EHR-CAT) in a prospective observational cohort study of patients with cancer initiating contemporary systemic anticancer therapies. Eight hundred six patients (49.5% female) with a median age of 61 years (interquartile range, 53-69) were included. The most common cancer types were lung (21.8%), breast (10.8%), and pancreatic (10.3%). Anticancer therapies initiated at study inclusion included chemotherapy (48.3%), a combination of chemotherapy and immune checkpoint inhibitor (ICI; 16.6%), ICI monotherapy (15.4%), and targeted agents (19.7%). During an observation period of 6 months, 91 patients experienced a VTE (cumulative incidence, 11.2%; 95% confidence interval [CI], 9.0-13.3). The discriminatory performance of the RAMs varied, with the best c-statistic seen with the CAT Score, whereas the COMPASS-CAT score showed the lowest area under the curve value (c-statistics: Khorana score, 0.53 [95% CI, 0.50-0.56]; PROTECHT, 0.58 [95% CI, 0.56-0.61]; CONKO, 0.54 [95% CI, 0.51-0.57]; COMPASS-CAT, 0.50 [95% CI, 0.47-0.53]; CAT Score, 0.65 [95% CI, 0.62-0.67]; EHR-CAT, 0.55 [95% CI, 0.52-0.57]). Overall, we observed a poor to modest discriminatory performance of the RAMs in our contemporary cohort of patients with cancer, with the CAT Score performing best among all evaluated scores.Keywords
This publication has 51 references indexed in Scilit:
- A Predictive Score for Thrombosis Associated with Breast, Colorectal, Lung, or Ovarian Cancer: The Prospective COMPASS–Cancer-Associated Thrombosis StudyThe Oncologist, 2017
- Opportunities and challenges in developing risk prediction models with electronic health records data: a systematic reviewJournal of the American Medical Informatics Association, 2016
- Learning About Missing Data Mechanisms in Electronic Health Records-based ResearchEpidemiology, 2016
- Estimating risk of venous thromboembolism in patients with cancer in the presence of competing mortalityJournal of Thrombosis and Haemostasis, 2015
- Primäre medikamentöse Thromboembolieprophylaxe bei ambulanten Patienten mit fortgeschrittenem Pankreaskarzinom unter Chemotherapie?DMW - Deutsche Medizinische Wochenschrift, 2013
- A modified Khorana risk assessment score for venous thromboembolism in cancer patients receiving chemotherapy: the Protecht scoreInternal and Emergency Medicine, 2012
- Venous thromboembolism and prognosis in cancerThrombosis Research, 2010
- Development and validation of a predictive model for chemotherapy-associated thrombosisBlood, 2008
- Chemotherapy-induced thrombosisThrombosis Research, 2006
- Comparing the Areas under Two or More Correlated Receiver Operating Characteristic Curves: A Nonparametric ApproachBiometrics, 1988