E0513 - Outcomes for Patients on DOACs in Compensated and Decompensated Cirrhosis

Aysha Aslam, MD¹, Saurabh Gombar, MD, PhD¹, Paul Kwo, MD^2
1Stanford Health Care, Palo Alto, CA; ²Stanford University University School of Medicine, Redwood City, CA

Introduction: Cirrhotic patients are prone to thromboembolism as well as bleeding. Use of novel anticoagulants is well established in patients with compensated cirrhosis and CTP class A patients however data on the efficacy and safety profile of DOACs in decompensated cirrhotic patients is limited. Our AIM was to determine if the clinical outcomes differed in those with compensated cirrhosis receiving DOACS (apixaban, rivaroxaban, edoxaban, and dabigatran) compared to those with decompensated cirrhosis in terms of thromboembolic events, atrial fibrillation (a fib), bleeding events, and TPA administration

Methods: Single center review from 2015-present in all subjects with cirrhosis with/without decompensation at time of DOAC administration. Patients were required to have 90 days of history prior to the DOAC and at least 1 year of follow-up. Decompensation was identified based on ICD codes (ascites, jaundice, gastric and esophageal varices, portal hypertension, hepatorenal syndrome,hepatic encephalopathy). We performed sensitivity analysis for MELD < 11, 11-14 and >14. The cohorts were compared for new onset a fib, bleeding (GI bleed, hemorrhagic stroke, retroperitoneal bleed and other bleeding events), new onset thrombotic events (pulmonary embolism, DVT, occlusive stroke) and new administration of TPA before and after matching for all observable confounders using high dimensionality propensity score matching.

Results: 452 decompensated subjects were compared to 321 compensated subjects. The cohorts were similar in age and sex with decompensated cohort having fewer white patients (53.5% vs 64.8%) and higher baseline comorbidity score. While differences were noted in unmatched analyses, no significant differences in outcomes were noted after matching in the variables studied (Table). For each subgroup of MELD scores (MELD < 11, 11-14 and >14) the findings remained consistent

Discussion: In a single site observational analysis we identified no difference in thrombotic events, a fib, bleeding, or administration of TPA when comparing compensated and decompensated liver cirrhosis patients who were on a DOAC. We matched patients on all baseline observable confounders at the time they began DOAC administration. For sensitivity analysis we stratified presenting MELD into low, medium, high and for each subgroup the findings were consistent.



Disclosures:


Aysha Aslam, MD¹, Saurabh Gombar, MD, PhD¹, Paul Kwo, MD². E0513 - Outcomes for Patients on DOACs in Compensated and Decompensated Cirrhosis, ACG 2022 Annual Scientific Meeting Abstracts. Charlotte, NC: American College of Gastroenterology.