Spontaneous Recovery and Mortality in Patients With ALD Referred for Early Liver Transplantation.

INTRODUCTION

Alcohol-associated liver disease (ALD), including alcohol-associated hepatitis, is the leading cause of liver transplant (LT) in the United States, and acceptance of LT without a mandatory period of abstinence is increasing. However, some patients may achieve spontaneous recovery (SR) without LT. We evaluated predictors of mortality and SR among patients with ALD referred for early LT.

METHODS

Recovery, survival, and outcomes-ALD is a retrospective cohort of adults with ALD and <6 months of sobriety referred for early LT between 2018 and 2021 at 6 transplant centers across the United States. Fine-Gray competing risk regression evaluated predictors of 12-month mortality, with LT considered as a competing risk. Multivariable logistic regression identified predictors of 12-month SR, defined by the expanded Baveno VII criteria.

RESULTS

Of 607 patients, the median model for end-stage liver disease 3.0 was 29 (interquartile range [IQR] 22-36); the median time from last drink was 2.6 months (IQR 1.1-4.1). Independent predictors of 12-month mortality included higher international normalized ratio (INR) (sHR 1.49, 95% confidence interval [CI] 1.23-1.79) and higher blood urea nitrogen (sHR 1.16 per 10 mg/dL, 95% CI 1.07-1.25). After 12 months, 18.5% achieved SR without LT. Predictors of SR at 12 months were lower INR (odds ratio 0.52, 95%CI 0.31-0.83) and higher platelet count (odds ratio 1.04 per 10 × 10 3 /µL, 95%CI 1.01-1.08).

DISCUSSION: Among patients with ALD and short-term abstinence, blood urea nitrogen and platelets are novel predictors of outcomes, along with INR. Using these predictors-or recalibrating existing model for end-stage liver disease coefficients-in this unique patient population may help refine medical selection criteria and improve utility and equity in access to early LT.