Comparison of risk prediction scores for venous thromboembolism in lung cancer patients: a retrospective cohort study

Introduction: Systematic thromboprophylaxis to prevent venous thromboembolism (VTE) is not recommended for ambulatory patients with lung cancer. Several risk prediction scores to identify cancer patients at high risk ofVTE have been proposed, but their clinical utility remains controversial. Objectives: Assess and compare the predictive value of the Khorana, PROTECHT, and CONKO scores in patients with lung cancer. Patients and Methods: Retrospective, noninterventional cohort study including patients with lung cancer initiating a new cancer therapy between January 2016 and December 2018. Clinical data were collected until 6-month follow-up. The Khorana, PROTECHT, and CONKO scores were calculated and their performance was compared. Results: One hundred and twenty patients were enrolled initially, of whom 41 were dropped out (13 deaths and 28 loss to follow-up). Mean age of patients was 64 years [40-88 years] with male predominance (sex-ratio=6.9). Twenty four (30.4%) developed VTE during the 6-month follow-up period. Mean calculated Khorana, PROTECHT, and CONKO scores were respectively: 1.85, 3.04, and 1.89. At the conventional positivity threshold of 3 points, these scores classified 24.1% to 63.3% of patients as high risk. Area under the curve of the scores ranged from 0.44 to 0.55. The 6-month VTE incidence in these patients ranged from 26.3% (95% CI, 3.5-124) for the Khorana score to 28% (95% CI, 7.9-107) for the PROTECHT score and to 30% (95% CI, 1,08-119) for the CONKO score. Conclusion: The prediction scores performed poorly in predicting VTE in lung cancer patients. More studies are needed to prior considering introduction to clinical practice.