SAFETY OF LONG-TERM THERAPY WITH WARFARIN: HEMORRHAGE FREQUENCY AND CLINICAL PREDICTORS (results of a prospective 15-year follow-up)

Abstract

Despite the long-term clinical experience, bleeding remains the greatest danger for patients taking vitamin K antagonists. Moreover, even the presence of risk factors for hemorrhagic complications in a patient should not be the only reason for cancelling anticoagulant therapy. The challenge in assessing the frequency of hemorrhage is primarily explained by different approaches to hemorrhage classification. For patients receiving long-term anticoagulant therapy, along with fatal and life threatening hemorrhages, minor and clinically significant small hemorrhages should be taken into account to assess the safety of treatment. The 15-year prospective follow-up of patients receiving warfarin found that the frequency of all hemorrhagic complications was 8.66/100 patient-years, major hemorrhage — 2.98/100 patient-years, and clinically significant hemorrhage — 2.62/100 patient-years. According to the results of the discriminant analysis, the predictors for the development of major and clinically significant hemorrhagic complications in patients receiving long-term therapy with warfarin are administration of proton pump inhibitors, labile INR during therapy, recurrent minor hemorrhage in history and amiodarone therapy. The 15-year prospective follow-up of patients receiving warfarin demonstrated that minor hemorrhagic complications account for one third of all bleedings and are not associated with the subsequent development of clinically significant and major bleeding in patients. The prospective follow-up also found an increase in the frequency of hemorrhagic complications if one or two antiplatelet agents were added to warfarin therapy. The incidence of major and minor clinically significant hemorrhagic complications during triple antithrombotic therapy was 24/100 patient-years, which was higher than in the warfarin monotherapy subgroups and in the case of combining warfarin with an antiaggregant (4.4 and 8.47%, respectively). The localization of bleeding was analyzed to reveal that in triple antithrombotic therapy the greatest number of bleedings occurred in the upper or lower gastrointestinal tract. The relative risk was calculated to confirm that triple antithrombotic therapy significantly increased the risk of gastrointestinal bleeding by 3.02 times.