[Predictive Khorana's model in patients with venous thromboembolic disease and cancer]

Background and objective: The predictive Khorana's model was developed to score the thromboembolic disease risk in cancer patients on chemotherapy and to identify which patients would benefit from thromboprophylaxis. We analized the results and applied the predictive Khorana's model in patients with cancer and who were diagnosed with deep vein thrombosis.

Material and methods: Retrospective analysis of prognostic characteristics of Khorana's model in 122 patients based on a prospective analysis.

Results: Seventy-nine percent of the total were in the low and intermediate risk category and 21% had high risk according to the Khorana's predictive model. This model had a sensitivity and prognostic precision of 20.8% (95% confidence interval [95% CI]: 14.6-28.7) and a false negatives proportion of 79.2% (95% CI: 1.3-85.4).

Conclusions: Application of this model in our patients would not be enough as the unique tool to identify cancer patients who should receive tromboprophylaxis. The use of both biomarkers and clinical models seems to be the best cost-effective strategy for this purpose. Future, randomized, prospective, placebo-controlled studies are needed for find better treatment strategies in cancer patients.