Predicting the Incidence of Venous Thromboembolism Using the Khorana Score: A Literature Review

Venous thromboembolism (VTE) is the second most common cause of cancer-related deaths globally. The Khorana score, a VTE prediction model, is calculated using the site of cancer, white blood cell count, hemoglobin level, platelet count, and body mass index. This study aimed to investigate the usefulness of the Khorana score, using data available in the literature. On July 28, 2020, we collected papers using the following keywords: "cancer'', "venous thromboembolism'', "deep vein thrombosis'', "pulmonary embolism'', and "Khorana score'' on PubMed. Papers published after 2016 were eligible. The selection criteria were as follows: "English or Japanese'', "original paper'', "abstract and full text'', and "comply with the clinical question''. There were 131 papers that matched the keywords, and 15 of them complied with the selection criteria. In 15 papers, Khorana score was calculated in 8047 patients. In the low- and intermediate-risk groups, 532 of 6812 patients developed VTE [7.8%, 95%confidence intervals (CI) 7.2-8.5], whereas in the high-risk group, 127 of 1235 patients developed VTE (10.3%, 95% CI 8.7-12.1) [odds ratio (OR) 1.3, 95% CI 1.0.1.6] (I-2= 0%, tau(2)= 0, p= 0.50). Venous thromboembolism prediction using the Khorana score might be useful. However, most of the number of VTE patients are in the low- and intermediate-risk groups. Therefore, a comprehensive evaluation according to clinical conditions is required, regardless of the risk classification using the Khorana score.