The Khorana score for prediction of venous thromboembolism in cancer patients: an individual patient data meta-analysis.

BACKGROUND: Oncology guidelines suggest using the Khorana score to select ambulatory cancer patients receiving chemotherapy for primary venous thromboembolism (VTE) prevention, but its performance in different cancers remains uncertain. OBJECTIVE: To examine the performance of the Khorana score in assessing 6-month VTE risk, and the efficacy and safety of LMWH among high-risk Khorana score patients. METHODS: This individual patient data meta-analysis evaluated (ultra)-low-molecular-weight heparin (LMWH) in patients with solid cancer using data from seven randomized controlled trials. RESULTS: A total of 3,293 patients from the control groups with an available Khorana score had lung (n=1,913; 58%), colorectal (n=452; 14%), pancreatic (n=264; 8%), gastric (n=201; 6%), ovarian (n=184; 56%), breast (n=164; 5%), brain (n=84; 3%), or bladder cancer (n=31; 1%). The 6-month VTE incidence was 9.8% among high-risk Khorana score patients and 6.4% among low-to-intermediate-risk patients (OR 1.6; 95%-CI, 1.1-2.2). The dichotomous Khorana score performed differently in lung cancer patients (OR 1.1; 95%-CI, 0.72-1.7) than in the group with other cancer types (OR 3.2; 95%-CI, 1.8-5.6; Pinteraction =0.002). Among high-risk patients, LMWH decreased the risk of VTE by 64% compared to controls (OR 0.36; 95%-CI, 0.22-0.58), without increasing the risk of major bleeding (OR 1.1; 95%-CI, 0.59-2.1). CONCLUSION: The Khorana score was unable to stratify patients with lung cancer based on their VTE risk. Among those with other cancer types, a high-risk score was associated with a 3-fold increased risk of VTE compared with a low-to-intermediate risk score. Thromboprophylaxis was effective and safe in patients with a high-risk Khorana score.