Abstract

Background/Aims
Oral anticoagulants may reduce risk of hepatic decompensation in patients with compensated cirrhosis, but well-powered randomized trials are missing. We aimed to estimate the effect of oral anticoagulants on risk of hepatic decompensation and major bleeding in patients with compensated cirrhosis and atrial fibrillation.
Methods
Observational data from Swedish healthcare registers 2011-2022 were used to emulate a target trial of oral anticoagulants in patients with compensated cirrhosis and newly diagnosed atrial fibrillation. Inverse-probability weighted marginal structural models were used to compare 5-year risks of hepatic decompensation and non-portal hypertension-related major bleeding in initiators versus non-initiators of oral anticoagulants.
Results
The study included 1,160 patients (715 men [61.6%]; median [p25-p75] age of 73 years [67-79]). The 5-year risk of hepatic decompensation was 10.4% (33/383) in initiators and 16.6% (112/777) in non-initiators (risk ratio [RR]=0.62, 95% confidence interval [CI]=0.33-0.92), corresponding to a number needed to treat of 17 (95%CI=9-112). The risk reduction was primarily driven by a reduced risk of ascites (RR=0.58, 95%CI=0.26-0.90). The risk of major bleeding was 19.0% (63/383) in initiators and 19.8% (149/777) in non-initiators (RR=0.96, 95%CI=0.64-1.28). Risks were similar between treatment groups regarding fatal, intracranial, gastrointestinal, and other bleedings.
Conclusions
In this nationwide observational study, patients with compensated cirrhosis and atrial fibrillation who initiated oral anticoagulants had lower risk of hepatic decompensation, and similar risk of major bleeding compared to non-initiators. The results suggest oral anticoagulants are safe in patients with compensated cirrhosis and may improve prognosis. Randomized trials are warranted to confirm these results.

• Keywords: emulated target trial, marginal structural model, liver disease, anticoagulation