Safety and efficacy of anticoagulation for atrial fibrillation in cirrhotic patients

Abstract

Abstract Introduction Atrial fibrillation (AF) is the most prevalent sustained arrhythmia and increases thromboembolism risk. AF is more prevalent in patients with cirrhosis and cirrhotics are at higher risk for ischemic stroke from thromboembolic mechanisms such as AF. Cirrhotics have complex pathophysiology of coagulation which can lead to bleeding diathesis or hypercoagulability. Cirrhotics have been excluded from major trials of anticoagulants (AC), thus there are currently no guidelines for AC in cirrhotics with AF. We aim to investigate how the incidence of ischemic stroke, major bleeding, hospitalizations, and mortality vary in patients with AF and cirrhosis between AC choices including no AC. Methods De-identified electronic health records from a U.S. health system between March 2013 and December 2022 were reviewed to identify patients with International Classification of Diseases (ICD) diagnoses of both AF and cirrhosis and who also had Creatinine, Bilirubin, and INR values within 6 months of their second diagnosis (either AF or cirrhosis). Model for End-stage Liver Disease (MELD) scores were calculated from these values and patients were followed from the first date of their second diagnosis. Incidences of ICD-coded ischemic stroke, ICD-coded bleeds, Current Procedural Terminology (CPT) coded transfusions, all-cause mortality, and hospitalizations were recorded. Propensity score weighting between AC and non-AC groups balanced covariates (Table 1). Bivariate analyses for outcomes were conducted between AC and non-AC groups and among warfarin, DOAC, and injectable AC subgroups. Results 1,205 patients with AF and cirrhosis were included. 605 patients were prescribed AC (50.2%), 384 were prescribed DOACs (63.5%), 275 were prescribed Warfarin (45.4%), and 131 (21.7%) were prescribed injectable AC. After propensity weighting, patients prescribed AC had similar rates of bleed (p=0.607), transfusion (p=0.225), and death (p=0.779) but decreased rates of ischemic stroke (p=0.001) and number of hospitalizations (p<0.001) compared to non-AC patients. Patients prescribed DOACs had similar rates of bleed (p=0.529), ischemic stroke (p=0.108), number of hospitalizations (p=0.870), and death (p=0.230) but decreased rates of transfusion (p<0.001) compared to patients on other forms of AC. Patients prescribed warfarin had similar rates of death (p=0.230) and decreased rates of bleed (p=0.013) but increased rates of transfusion (p<0.001), ischemic stroke (p<0.001), and number of hospitalizations (p = 0.024) (Table 2). Conclusions Patients with AF and cirrhosis receiving AC experienced decreased rates of ischemic stroke and number of hospitalizations compared to patients who did not receive AC. Among AC subgroups, patients prescribed DOACs had lower rates of transfusion compared to other AC therapies while patients prescribed warfarin had increased rates of transfusion, ischemic stroke, and number of hospitalizations compared to other AC therapies.