Alcoholic hepatitis is an inflammation of the liver caused by excessive and typically long term alcohol use. In severe cases, it can require hospitalization and even lead to severe organ damage and death. Currently, avoiding alcohol is the only intervention that improves long term survival.

To determine the severity of alcohol hepatitis, we calculate the Maddrey Discriminant function. You take the patient's prothrombin time minus the normal prothrombin time, multiply this by 4.6 and add the total bilirubin in milligrams per deciliter to find the patient's Madrid discriminant function.

This formula can be found on MDCalc. A Maddrey Discriminant function score of greater than 32 indicates severe alcoholic hepatitis. Previous studies have shown survival benefit in using steroids like prednisolone in acute severe alcoholic hepatitis.

This is why we use steroids to this day in the management of severe alcoholic hepatitis. The reason we use prednisolone as a steroid of choice is that prednisone is a prodrug that needs to be converted to prednisolone by the liver. Prior studies have shown a lower level of steroid in the blood when prednisone is used versus prednisolone in alcohol hepatitis, likely as a result of liver inflammation leading to poor liver function.

Unfortunately, more recent data has placed the benefit of steroids in improving long term survival into doubt. The STOPAH trial, also known as the Steroids or Pentoxyphylin for alcoholic Hepatitis trial, looked at the benefits of steroids for alcoholic hepatitis. It was a multicenter, randomized trial of over 1000 patients.

Pentoxyphylin did not improve survival in patients with alcoholic hepatitis. When comparing patients who received prednisolone versus those that did not, a trend toward reduced 28 day mortality in the steroid group was seen, but it did not reach statistical significance. In addition, the survival benefit was no longer seen at 90 days.

Interestingly, the rate of infection in the steroid group was twice as high as the placebo group, but mortality rates remain the same. Though the standard of care is still to use steroids, we do calculate a Lillie score to see if an individual has responded to steroids and whether continuation would be beneficial versus those who do not have improvement and where continuing steroids would have no benefit due to the lack of response in their predicted survival. A Lillie score greater than 00:45 indicates a poor long term survival.

In addition, it indicates a lack of response to steroids and the steroids should be discontinued to prevent potential complications like infection. The Lillie score is typically calculated at day seven of admission, but recent data has shown that calculating at day four can help identify steroid nonresponders earlier. Another topic of controversy is the use of antibiotics in alcoholic hepatitis.

When a patient presents with alcoholic hepatitis. They commonly present with jaundice, fever, abdominal pain, weakness, and or confusion. They may have a leukocytosis or elevated white blood cell count.

When someone has an elevated white blood cell count, we naturally think of infection in alcoholic hepatitis. Liver injury can also lead to the elevation in white blood cells. As a result, it's not uncommon for patients with alcohol hepatitis to be placed on antibiotics due to concern that there may be an underlying infection.

Now, animal models have shown that antibiotics can help alcohol induce liver injury. But does this translate to humans? Do antibiotics help in alcoholic hepatitis? Well, those are the questions. The JAMA article titled Effect of Prophylactic Antibiotics on Mortality in Severe Alcoholic Related Hepatitis aimed to answer this article Reviews the AntibioCor Trial. This is a two group, randomized, placebo controlled trial conducted at 25 centers across France and Belgium.

Patients were 18 to 75 years old and were eligible if they were heavy drinkers who were clinically diagnosed with alcoholic hepatitis and had a Maddrey score of 32 or higher. All the patients received 40 milligrams per day of prednisolone for 30 days. The participants were randomized to either amoxicillin clavulanate, 1 gram/ 125 milligrams three times daily or to placebo for 30 days.

The main outcome of the study was to evaluate 60 day all cause mortality. The result of the study showed that those who received antibiotics did not have a significant difference in 60 day all cause mortality compared to those who received placebo. What is interesting is that there was a significant difference in the rates of infection between the two groups.

In the antibiotic group, infections were seen in 17.3% versus 41.5% in those receiving placebo.

However, the difference in rates of infection did not translate to a difference in 60 day all cause mortality. In addition, the seven day lillie score and the MELD score at 60 days were also not different between these two groups. As a result, we now know that prophylactic antibiotics do not improve survival in severe alcohol hepatitis, even though it does decrease the rate of infection.

It is important to understand there are limitations to this study. First, the population studied were from centers in France and Belgium. Extrapolating this information to the United States or other countries based on cultural and ethnic differences would be premature.

Second, typically we assess a patient with a lillie score seven days after being on prednisolone to determine if the steroids should be continued. In this study, instructions were not provided whether to discontinue steroids or not, so this can confound the results. Despite these limitations, I think the study highlights three important points.

First, alcohol cessation is still the most important factor in improving long term survival for severe alcoholic hepatitis. Resources should be focused on providing support and counseling patients on how to abstain from alcohol and addressing underlying mental illness that can contribute to addiction. Second, antibiotics should not be considered a mainstay of treatment and should be weaned off if no active infection has been identified.

Third, despite this information, we should still treat with antibiotics when there is concern for infection. And we know that infections in this population contributes to a higher chance of morbidity and mortality.