Prediction of venous thromboembolism in ambulatory patients with cancer receiving chemotherapy: An expanded thromboembolic risk score model.

Abstract

e19551 Background: Cancer and chemotherapy are well recognized risk factors for thromboembolic events (TEs). Nadroparin has been demonstrated to reduce the incidence of TEs by about 50% in cancer outpatients receiving chemotherapy. Patients receiving regimens containing cisplatin, carboplatin or gemcitabine and their combination have the highest risk of TEs. Khorana and colleagues have validated a score (KS) to identify the risk of TEs in cancer patients, stratifying them according to five variables (site of cancer, platelet count, haemoglobin level, leukocyte count and body mass index). Patients with a score ≥ 3 are at high risk to develop TEs. Methods: In our analysis we have expanded the KS by incorporating the chemotherapy variable: for patients receiving chemotherapy regimens containing cisplatin or carboplatin or gemcitabine 1 point has been added to the score; while regimen containing platinum compound plus gemcitabine 2 points. The aim of this post-hoc analysis was to evaluate the ability of this expanded Khorana risk score (ExKS) to identify high risk patients among the PROTECHT trial (NCT 00951574) population. A receiver operating characteristic (ROC) curve has been used to assess the accuracy of KS and ExKS. Results: Among the 1150 evaluable patients: 381 were allocated in the placebo arm and for 378 of them an evaluation of KS and ExKS has been performed: 11.1% (42/378) and 32.8% (124/378) were at high risk of TEs in the KS and ExKS, respectively. In the study population a total of 15 TEs has been reported. 33.3% (5/15) and 66.7% (10/15) of TEs occurred in the high risk patients according to KS and ExKS, respectively. A ROC curve analysis has been performed to evaluate the predictability of the ExKS respect to KS. The Area under the ROC Curve, which reflects the predictive value of the scores, was larger with ExKS in comparison with KS (0.70 and 0.65 respectively) Conclusions: ExKS could have a higher predictability to identify high risk patients for TEs compared to KS in the PROTECHT study. A prospective study is warranted to validate the score