Abstinence Tied to Lower Mortality in Alcoholic Cirrhosis With Portal Hypertension

Alcohol abstinence was associated with improved prognosis across all stages of clinically significant portal hypertension in patients with alcohol-related cirrhosis, according to an observational study.

Among 320 such patients, multivariable analysis found a 61% lower risk of hepatic decompensation over 3 years of follow-up among those who abstained from alcohol (adjusted HR 0.39, 95% CI 0.28-0.56), reported Thomas Reiberger, MD, of the University of Vienna, and colleagues.

And abstinence was linked with a 57% lower risk of liver-related mortality (adjusted HR 0.43, 95% CI 0.26-0.70) and 55% lower all-cause mortality risk (adjusted HR 0.45, 95% CI 0.30-0.69), the group detailed in Clinical Gastroenterology and Hepatology.

Alcohol abstinence was also associated with a significant reduction in the cumulative incidence of decompensation among those with clinically significant portal hypertension involving hepatic venous pressure gradient (HVPG) levels of 10-19 mm Hg (P<0.001), and those who had already progressed to high-risk portal hypertension with HVPG levels of 20 mm Hg or greater (P=0.002).

At 3 years, the probability of developing decompensation was far lower for those who abstained from alcohol:

• HVPG levels 10-19 mm Hg: 32.4% vs 60.0% for active drinkers
• HVPG levels 20 mm Hg or above: 57.5% vs 82.6% for active drinkers

“Our study demonstrates the beneficial long-term impact of alcohol abstinence in a large-scale cohort of patients with alcohol-related cirrhosis who have already progressed to clinically significant or high-risk portal hypertension,” they noted. “Achieving and maintaining alcohol abstinence should be the key priority in the management of ALD [alcohol-related liver disease] patients, regardless of disease severity.”

Abstinence and age were the only independent predictors of liver-related and all-cause mortality, the authors reported. HVPG, model for end-stage liver disease (MELD), and abstinence were the only independent predictors of hepatic decompensation. Abstinence did not change the risk of developing hepatocellular carcinoma, which occurred at a rate of 6.6% in the abstinent patients and 3.8% in those who continued to drink (HR 1.75, 95% CI 0.51-6.03).

A leading cause of liver-related deaths is ALD, which accounts for half of all cirrhosis-related deaths, Reiberger’s group noted. Once diagnosed with alcohol-related cirrhosis, most of these patients already present with decompensated liver disease, but remain at high-risk for further decompensation and even death.

“Portal hypertension is the underlying driver of hepatic decompensation in these patients and develops as a result of hepatic architectural and functional changes following chronic alcohol consumption,” the group explained. “In addition to these chronic changes, alcohol ingestion acutely aggravates portal hypertension and increases blood flow through portosystemic collaterals, thus further increasing the risk of hepatic decompensation.”

While non-selective beta blockers may lower portal pressure and reduce systemic inflammation to treat ALD, alcohol abstinence is the principal therapeutic intervention for those with alcohol-related cirrhosis.

Reiberger and colleagues examined data on 320 patients with alcohol-related cirrhosis from the Medical University of Vienna from January 2004 to December 2020. Those included underwent a baseline HVPG measurement, which is the gold standard for assessing portal hypertension severity and were required to pause any non-selective beta blocker use for at least 5 days prior.

Included patients had also been diagnosed with clinically significant portal hypertension, defined by an HVPG level of 10 mm Hg or greater. Diagnoses were confirmed by lab tests, imaging, liver biopsy, and histology for most patients. The analysis adjusted for HVPG, MELD, and prior baseline decompensation.

Among the 241 patients who remained abstinent, 109 had HVPG levels of 10-19 mm Hg while 132 had levels of 20 mm Hg or greater. Of the 79 patients in the non-abstinent group, 39 had HVPG levels of 10-19 mm Hg while 40 had levels of 20 mm Hg or greater.

Initial or further decompensation occurred in 45%, with new onset/worsening of ascites being the most common event among the abstinent (37%) and non-abstinent (34%) groups.

The average patient age was 57, and 76% were men. The median HVPG level was 20 mm Hg, and 87.5% had decompensated disease. Most of those with HVPG levels of 20 mm Hg or higher had large esophageal varices, prior liver decompensation, and higher MELD/Child Pugh scores, as well as certain higher biomarker levels.

Over an average follow-up of 36 months, three-quarters remained abstinent while one-quarter still actively consumed alcohol. Overall, 37.5% died, including 81.7% from liver-related complications/hepatocellular carcinoma and 18.3% from non-hepatic causes. Only 13.3% received a liver transplant.

Those with active alcohol use were younger versus those who were abstinent (average age 55 vs 57), with higher gamma-glutamyltransferase and aspartate transaminase (AST) levels, but had significantly lower MELD and Child-Pugh scores.

Among those who were abstinent, the average abstinence duration prior to baseline was about 6 months.

The duration of abstinence was significantly tied to lower baseline MELD scores, Child-Pugh scores, C-reactive protein levels, liver stiffness scores, AST levels, and white blood cell counts.

No relationship was seen between the duration of abstinence and HVPG at baseline.

At 3 years, the cumulative survival probability was greater for those who abstained from alcohol:

• HVPG levels 10-19 mm Hg: 82.8% vs 72.1% for active drinkers
• HVPG levels 20 mm Hg or above: 70.5% vs 50.7% for active drinkers

“Our findings suggest that active alcohol consumption may be the most important driver of mortality in alcohol-related cirrhosis,” they noted.

Study limitations included the fact that alcohol abstinence was not assessed by gold-standard methods, such as ethyl glucuronide in hair, and not all patients had decompensated disease.