|The following article features coverage from the European Hematology Association 2020 virtual meeting. Click here to read more of Hematology Advisor’s conference coverage.|
The recently developed prognostic tool VTE-BLEED failed to predict bleeding complications in patients with venous thromboembolism (VTE) undergoing conventional anticoagulation regimens, according to findings presented at the Virtual Edition of the 25th European Hematology Association (EHA) Annual Congress.
“Recently, the VTE-BLEED score has emerged, in an attempt to improve the predictive capacity of bleeding risk in patients with long-term anticoagulated VTE,” according to a presentation led by Marta Hidalgo Soto of Hospital 12 Octubre in Spain.
The aim of the study was to validate the VTE-BLEED score in the Spanish population using cohorts from the RIETE Registry, an international database on patients with VTE.
The authors stratified the incidence of bleeding events in patients with VTE according to VTE-BLEED to validate its predictive accuracy. In addition, the investigators prospectively compared the predictive value of HAS-BLED and RIETE scores to those of VTE-BLEED.
In total, 759 adult patients who presented at the hospital in Madrid with VTE were included in the study. All patients had complete 1-year follow-up data. The mean patient age was 68 years. Overall, 13% of patients experienced at least 1 episode of bleeding.
According to the VTE-BLEED risk stratification, 58.8% of the patients were considered low risk, while 41.2% were high risk. Of the 102 total bleeding events, approximately half of the events occurred in each group (P =.086). Among low-risk and high-risk patients, major bleeding was reported in 8 and 6 patients, respectively (P =.32). Death due to bleeding was also similar between the groups: there were 5 deaths in low-risk patients and 3 deaths in high-risk patients (P =.83).
According to the HAS-BLED risk stratification, 89.2% of the patients were considered low risk, while 10.8% were high risk. Nearly all of the 102 bleeding events (~90%) occurred in low-risk patients (P =.73). The occurrence of major bleeding (P =.30) and death due to bleeding (P =.32) were not significantly different between low- and high-risk groups.
Finally, according to the RIETE risk stratification, 8.3%, 88.8%, and 2.9% of the patients were considered low risk, intermediate risk, and high risk, respectively. Of the 102 total bleeding events, approximately 4%, 95%, and 1% occurred in the low-risk, intermediate-risk, and high-risk group, respectively (P =.093). The occurrence of major bleeding differed between the groups, with 1 and 13 events in the low-risk and intermediate-risk group, respectively, and no events where reported in the high-risk group (P =.025). Death due to bleeding among the 3 groups was not significantly different (P =.82).
The sensitivity and specificity were 49.02% and 59.97% for VTE-BLEED, 9.8% and 89.04% for HAS-BLED, and 87.5% and 8.26% for RIETE, respectively. The positive and negative predictive values (likelihood ratios) were 0.83 and 1.11 for VTE-BLEED, 1.05 and 0.43 for HAS-BLED, and 0.97 and 1.02 for RIETE.
The discriminative power was not significantly different between the 3 scores. The area under the curve ranged from 0.56 for the VTE-BLEED score to 0.50 for the HAS-BLED score.
“These scores cannot differentiate patients with higher from those with lower risk of bleeding,” said Dr Hidalgo Soto.
The study findings should be interpreted with caution, and larger studies of VTE-BLEED are needed to improve the predictive capacity of the score to facilitate decisions regarding anticoagulation in this patient population.
Hidalgo Soto M, Zamanillo Herreros I, Diaz Pedroche C, et al. VTE-BLEED score to predict bleeding complications in patients with venous thromboembolism: from the RIETE Registry. Presented at: Virtual Edition of the 25th European Hematology Association (EHA) Annual Congress; June 2020. Abstract S324.