Haemorrhagic risk in elderly patients with atrial fibrillation that develop low platelet count during direct oral anticoagulant treatment: insights from the ATHERO-AF study

Abstract Background Bleeding risk in atrial fibrillation (AF) patients is complex and dynamic. Among others, thrombocytopenia has been suggested as a bleeding risk factors. No data on thrombocytopenia occurring during direct oral anticoagulants (DOACs) therapy are available. Purpose We investigated the incidence rate of thrombocytopenia and major bleeding (MB) in AF patients on DOACs. We also investigated whether the addition of incident thrombocytopenia to the HAS-BLED score might increase its predictive value compared to baseline HAS-BLED calculation. Methods 955 AF patients from the prospective ongoing ATHERO-AF study followed for a mean of 38.4±26.6 months. All patients were on DOACs. Thrombocytopenia was defined by a platelet count <150 x109/L registered during follow-up visits. MB events were defined according to ISTH definition and were recorded at each follow-up visit. Patients with thrombocytopenia at baseline were excluded. Multivariable Cox proportional hazards regression analysis was used to calculate the hazard ratio (HR) with 95% Confidence interval (95%CI) for each factor in relation to bleeding risk. Results Mean age was 77.3±9.0 years and 40.1% were women. During a follow-up, 139 patients developed thrombocytopenia with an incidence rate of 0.8% per year. We recorded no difference between thrombin and factor Xa inhibitors. During follow-up, 179 bleedings occurred, of which 80 were major. Patients experiencing bleedings were more likely to suffer from arterial hypertension, heart failure, anaemia and had higher CHA2DS2-VASc and HAS-BLED scores. At multivariable Cox proportional hazards regression analysis, factors associated with MB were incident thrombocytopenia (HR 1.995, 95%CI 1.211-3.288), antiplatelet use (HR 2.716 95%CI 1.269-5.816) and age (HR 1.033, 95%CI 1.003-1.064). The addition of incident thrombocytopenia to the HAS-BLED score increased the predictive value of baseline HAS-BLED in patients without thrombocytopenia for MB events (AUC from 0.54 to 0.61, p=0.016). Conclusion Incident thrombocytopenia is associated with an increased risk of MB. The addition of incident thrombocytopenia to the baseline HAS-BLED score increased its predictive value, suggesting that continuous bleeding risk evaluation of AF patients may provide additional prognostic information compared to the baseline only.