From the AGA Journals

Model identified heavy drinkers at highest risk of ALD progression

View on the News

Real knowledge may bring behavioral change

In the life of a hepatologist few things are as gratifying as when a patient with alcohol-related liver disease (ALD) quits drinking. Though we wish this were the norm, ALD is both increasingly common and morbid. Tools to connect with and empower real change in our patients with ALD are urgently needed. Unfortunately, our toolbox is somewhat bare.

Dr. Elliot B. Tapper is an assistant professor in gastroenterology and internal medicine at Michigan Medicine in Ann Arbor

Dr. Elliot Tapper

To improve, we must become accustomed to (and partner with experts in) the care of substance use disorder. We must learn to maximize the impact of our counseling on our patients. Behavioral interventions for ALD require goal-setting and self-regulation and both depend on the patient’s outcome expectations. All would be immeasurably strengthened with concrete prognostic data.

This is why the Delacôte study is important. The authors create a multistate model with inputs from cohorts of patients with biopsy-proven and staged ALD. The result is a specific 5-year risk of cirrhotic decompensation or hepatocellular carcinoma tailored to the patient’s age, sex, body mass index, alcohol use duration, and liver histology. Although this model’s estimates have confidence intervals and their generalizability would be improved if histology were replaced with noninvasive indices, these data are amongst the most tangible illustrations of risk available for patient-doctor deliberations.

Knowledge, when communicated effectively, is the cornerstone of behavioral change. Translating the abstract concept of progressive ALD into personalized, modifiable risks is a leap forward. We have a new tool, let’s use it.

Elliot B. Tapper, MD, is an assistant professor in gastroenterology and internal medicine at Michigan Medicine, Ann Arbor. He has no conflicts of interest.



In heavy drinkers with alcohol-related liver disease, a Markov model based on age, sex, body mass index, and duration and extent of alcohol use predicted risk for disease progression, researchers reported in Clinical Gastroenterology and Hepatology.

The study included 2,334 hospitalized adults with consistently abnormal liver test results who had consumed at least 50 grams of alcohol (about 3.5-4 drinks) per day for the previous 5 years. The model was developed using data from 1,599 individuals with baseline liver biopsies and validated in 735 individuals with no baseline liver biopsies but available data on the presence or absence of hepatic decompensation.

For a 40-year-old man with F0-F2 fibrosis who had been drinking alcohol for 15 years, who drank 150 grams of alcohol daily, and who had a body mass index (BMI) of 22 kg/m2, the model predicted a 31.8% likelihood of having a normal liver at baseline, a 61.5% probability of baseline steatosis, and a 6.7% probability of baseline steatohepatitis. In women with the same baseline variables, respective probabilities were 25.1%, 66.5%, and 8.4%. Based on these findings, the 5-year weighted risk for liver complications ranged from 0.2% for men with normal initial liver findings to 10.3% for men with baseline steatohepatitis. Among women, the corresponding risk estimates ranged from 0.5% to 14.7%, wrote PhD student Claire Delacôte of Centre Hospitalier Universitaire de Lille (France), and associates.

“This tool might be used by general practitioners or hepatologists to identify heavy drinkers at high risk for alcohol-related liver disease progression,” the investigators added. “This model might be used to adapt patient care pathways.”

The patients in this study were admitted to the hepatogastroenterology unit of a French hospital between 1982 and 1997. The Markov model incorporated seven stages of alcohol-related liver disease: normal liver (no fibrosis or steatosis), steatosis and F0-F2 fibrosis, alcohol-induced steatohepatitis and F0-F2 fibrosis, steatosis and F3-F4 fibrosis, alcohol-induced steatohepatitis and F3-F4 fibrosis, liver complications without steatohepatitis, and liver complications with alcohol-induced steatohepatitis. Liver complications were defined as hepatocellular carcinoma or liver decompensation (bilirubin >50 mmol/L, gastrointestinal hemorrhage, or ascites). Risk for progressing to liver complications was based on METAVIR score and onset of alcohol-induced steatohepatitis.

The researchers also looked specifically at F3-F4 (severe) fibrosis because of its clinical significance and common use as a study endpoint. Among 40-year-olds with a 15-year history of heavy drinking, the estimated prevalence of alcohol-induced steatohepatitis was 30.0% for men and 33.3% for women. The 5-year risk for liver complications was higher in women (30.1%) than men (24.5%) and was highest among women with baseline alcohol-induced steatohepatitis (41.0%). Overall, women had a 24.8% greater risk for disease progression than men (hazard ratio, 1.248).

Risk for liver complications also increased with age, and each 1-year increase in age at the beginning of heavy drinking heightened the risk for disease progression by 3.8%, regardless of stage of liver disease. “Based on these predictions, 50-year-old women are a high-risk subgroup of [alcohol-related liver] disease progression and should receive close follow-up,” the researchers wrote.

In addition, obese individuals (BMI, 30) had an 11.8% greater risk for progression of alcohol-related liver disease, compared with those with a BMI of 22. Consuming an additional 10 grams of alcohol per day had less impact on risk, the researchers noted.

“If patients are identified as being heavy drinkers by the general practitioner with no evaluation of fibrosis, these patients should be referred to a hepatologist. Nevertheless, we think that the threshold defining the high-risk population, which has been arbitrarily fixed at 5%, should be discussed by experts because it affects the patient’s care pathway. An online application is being developed to help clinicians and general practitioners in their daily practice,” they wrote.

No funding sources were reported. Ms. Delacôte reported having no conflicts of interest. Three coinvestigators disclosed ties to AbbVie, Bayer Healthcare, Eisai, Gilead, MSD, Novartis, Sanofi, and Servier. The others reported having no conflicts.

SOURCE: Delacôte C et al. Clin Gastroenterol Hepatol. 2020 Jan 11. doi: 10.1016/j.cgh.2019.12.041.

Next Article: