Risk of thromboembolic and haemorrhagic events associated with anticoagulation among 11,355 patients with non-valvular atrial fibrillation and thrombocytopenia

Abstract Background Oral anticoagulation (OAC) is the mainstay for stroke prophylaxis in patients with non-valvular atrial fibrillation (NVAF). Management is challenging among patients with NVAF complicated with thrombocytopenia. Purpose We compared risk of thromboembolic and hemorrhagic events between vitamin-k antagonist (VKA) and direct oral anticoagulants (DOAC) users. Methods We retrospectively analyzed patients with thrombocytopenia and NVAF prescribed VKA (N=5,251, 46.2%) and DOAC (N=6,104, 53.8%) between January 2012 and July 2022 in 16 public hospitals in Hong Kong. Mild to moderate thrombocytopenia was defined as mean platelet count (PLT) 100-150 × 10^9/L and 50-100 × 10^9/L respectively, during the year of OAC initiation. Propensity-matched cohorts of VKA and DOAC were generated on the basis of age, sex, baseline co-morbidities (components of CHA2DS2VASC and HAS-BLED scores) and medications. Time to intracranial hemorrhage (ICH), gastrointestinal bleeding (GIB) and ischemic stroke (IS) was analyzed using competing risk regression, accounting for death as competing risk. Subgroup analysis was performed to detect heterogeneity of effect across mild and moderate thrombocytopenia. Secondary outcome included (i) comparisons between 4 DOAC agents with full dose; (ii) Full-dose DOAC vs. reduced-dose DOAC. Results Of the whole cohort (N=11,355, mean age 75.8±10.5, female 38.6%), 10.3% (n=1,174) and 89.7% (n=10,181) had moderate and mild thrombocytopenia, respectively. After propensity-score matching, VKA and DOAC (n=4,302 each cohort) users were well balanced at baseline characteristics, with a median follow-up of 4.0 (IQR: 1.6-7.3) years. Overall, utilization of DOAC was associated with reduced risk of ICH [Hazards Ratio (HR) 0.56 (95% CI 0.45-0.70)] and GIB [HR 0.60 (95% CI 0.48-0.75)], compared with VKA. No difference in risk of IS was observed [HR 1.00 (95%CI 0.83-1.20)]. Dosing reduction of DOAC made no difference in all outcomes. Subgroup analysis demonstrated the benefit in decreasing GIB might be more marked in moderate thrombocytopenia [(HR 0.24 (95%CI 0.09-0.61)], than in mild thrombocytopenia [HR 0.64 (95%CI 0.51-0.81)]. Conclusion In patients with NVAF and thrombocytopenia, DOAC has similar efficacy but better safety profile than VKA. In terms of GIB, the superiority appeared to be more pronounced in patients with more significant thrombocytopenia.