Much-Needed Guidance Arrives on Drug-Induced Liver Injury

Nancy S. Reau, MD


November 21, 2022

It is estimated that in the past month, nearly half of Americans used at least one prescription drug and one quarter used three or more, which doesn't account for the untold numbers of over-the-counter herbal and dietary supplements consumed. Many of these prescriptions and supplements carry a small but important risk for drug-induced liver injury (DILI), which is broadly classified as either direct (ie, dose-dependent, intrinsic, and predictable) or idiosyncratic (ie, largely dose-independent and unpredictable).

Although the consequences of DILI are considerable and potentially life-threatening, diagnosing it is no easy task for clinicians, who must exclude competing causes of injury and weigh several clinical symptoms, all without the aid of a validated diagnostic biomarker. With these challenges in mind, the American Association for the Study of Liver Diseases (AASLD) convened a group of experts who recently published a guidance document on various aspects of diagnosing and managing DILI.

Robert Fontana, MD

To hear more about the AASLD committee's key recommendations, Medscape contributor Nancy S. Reau, MD, chief of the hepatology section at Rush University Medical Center in Chicago, spoke with Robert Fontana, MD, the guidance's co–lead author and professor of medicine and director of the Transplant Hepatology Fellowship Program at the University of Michigan, Ann Arbor.

Why Diagnosing DILI Needs Expert Guidance

You and your co-authors note that this is a guidance, and not a guideline. Can you define what makes for that distinction?

Although it was commissioned by the AASLD Practice Guidelines Committee, it is a guidance as opposed to a guideline because the evidence that we had to review was primarily limited to case series, observational studies, and a fair amount of expert opinion. There really weren't the randomized controlled trials in this space that lead to the higher level of evidence to produce guidelines.

That being said, we thought it was important and timely to get this out for the medical community, to give them a framework and help organize their thinking. I think the AASLD commissioned it because, as the leading liver society, it needs to be in the forefront of the diagnosis and treatment of liver injury.

Given that DILI can present in several ways, how does this guidance help a clinician recognize it?

I think the most useful element in the document is the proposed diagnostic algorithm for patients with suspected DILI. We laid out everything to think about when seeing a patient with possible DILI.

For example, we remind the clinician of the importance of considering the temporal association between when someone took the drug and when they got sick. We also talk about the injury pattern as defined by the R value. An R value > 5 identifies hepatocellular injury, < 2 cholestatic liver injury, and between 2 and 5 mixed liver injury pattern. This helps remind the clinician of all the things you need to exclude while thinking it could be the drug.

Because we've seen in the Drug-Induced Liver Injury Network (DILIN) study and other cohorts that acute hepatitis C keeps happening, unfortunately, with the opioid epidemic in the United States, it's important to test everyone with suspected DILI for this using HCV RNA testing in the acute setting.

Some people might be unfamiliar with the R value or how to use it. Can you walk me through this?

The R value is defined as the ratio of serum alanine transaminase (ALT) divided by serum alkaline phosphatase (ALP), using upper limit of normal (ULN) for both values.

Basically, if you have a predominance of serum aminotransferase elevations, your R value will be high, whereas if you have a predominance of serum ALP elevations, your R value will be lower.

When a clinician is first approaching a patient, it helps remind you of the differential diagnosis you must go through when trying to identify DILI. Acute viral hepatitis is going to be predominantly hepatocellular, whereas things like pancreaticobiliary disease, such as gallstones and cancer, will be more cholestatic.

Underlying Risk Factors and Certain Drugs

Individuals with preexisting liver disease are a heterogenous group. You've highlighted that there is an increased risk for liver injury with some agents. However, do you have recommendations how to identify people with an increased risk and then how to monitor them for DILI?

As we know, patients with obesity and diabetes frequently have underlying fat in their livers. For those with presumed underlying fatty liver, I'd say there's only two drugs that we confidently know will increase the risk for progressive DILI: tamoxifen and methotrexate.

Tamoxifen is used to prevent breast cancer recurrence, but it can also cause hepatic steatosis. It can worsen preexisting fatty liver disease or steatohepatitis. That's been prospectively shown in women with breast cancer in large studies.

Methotrexate is an effective immunomodulator for rheumatoid arthritis and psoriatic arthritis. But it can facilitate or accelerate liver fibrosis progression in those with preexisting liver disease.

Idiosyncratic Liver Injury

You've outlined several agents and risk profiles that are more likely to lead to idiosyncratic DILI. Would you suggest heightened monitoring if a patient fell into this profile, especially those with obesity and diabetes?

For about 20 years, we haven't found an overrepresentation of people with diabetes or people with obesity among those who get clinically apparent idiosyncratic DILI from other agents, so it doesn't appear to be an overt risk factor beyond methotrexate.

That's a good thing, because 20%-25% of the population has diabetes and/or fatty liver. So, I don't think that's much of a differentiator and certainly not a bona fide susceptibility or risk factor with the vast majority of prescription and over-the-counter drugs.

In fact, there's not a lot of clinical risk factors for idiosyncratic DILI. It's not surprising; by its definition, it's unpredictable. Idiosyncratic DILI can happen in nearly any person, young or old, male or female. There's no gender, age, or race who has an increased susceptibility for most drugs.

Supplements: An Unregulated Headache for Clinicians

I want to turn to one of my least favorite topics to discuss with patients: herbal and dietary supplements. These supplements are often deemed "safe" by consumers, but that's not true.

Herbal and dietary supplements are a big umbrella. There are approximately 80,000-100,000 products, accounting for something like $11 billion dollars in annual sales in the United States alone.

They're not adequately overseen by regulatory agencies like the US Food and Drug Administration. The reason for that is a law called the Dietary Supplement Health and Education Act of 1994. It basically says that the supplements you can buy at, for example, a drug store or supermarket, are presumably safe and should be regulated like foods. Therefore, there are no necessary manufacturing standards or testing in humans before they're marketed. This is quite amazing, and unlike every medication we've ever prescribed.

It's a real problem, particularly with herbals from plant-based materials, in which the actual ingredients can change over time. There can be batch effects and changes over time in the active ingredients in any given natural product. A consumer doesn't know that.

This is extremely confusing to patients who go to pharmacies and see rows of these products in a medical environment and think that if they're being sold in a pharmacy or retail store, they must be safe.

How do you approach a patient who likely has DILI from these supplements?

It's very complex, but go back to the history and identify what they were taking. When did they start it? Do they have a picture of the product label that you as a physician can look at? Even then, the labeling might be misleading.

DILIN did a mass spectrometry analysis of actual products taken by patients and compared the results with the printed product labels. We identified a 50% discrepancy between the actual constituents and ingredients listed on the label, meaning that there were ingredients listed on the labels they couldn't find, and there were also ingredients present that weren't on the label.

It makes a causality assessment of whether a particular product did or didn't cause liver injury a lot more difficult to establish.

Do you ever have patients bring in the agent so that you can send it off to have it tested?

Absolutely. I'm very fortunate that, as part of DILIN, we have a collaboration with a top-notch analytical chemist who can tell us exactly what ingredients are in every tablet that these patients are taking.

Unfortunately, most clinicians do not have access to that, and we're only able to do this under a research protocol. But we're looking to continuously publish what we find and make people aware if we see recurring ingredients in products associated with liver injury that might eventually lead to some regulatory change.

But it's a bit of a moving target, because frequently the manufacturers of these products change, either intentionally or unintentionally, the components of these multi-ingredient products. You're always playing catch up when you're trying to do this kind of detective work.

The Challenges of Patient Management

I think we struggle when we must stop these agents, and we always want to do something else. After drug withdrawal, is there ever a time when you recommend steroids, N-acetylcysteine (NAC, or ursodeoxycholic acid or alternative options?

The sicker the patient, the more tempted you are to intervene. But the most important thing is to sit on your hands initially. You have to go back and look at the evidence.

The use of NAC was studied in a randomized control trial of acetaminophen-induced acute liver failure published in 2009 in Gastroenterology. It looked at approximately 170 patients with non-acetaminophen acute liver failure; about 50 of them had severe idiosyncratic DILI. They all had acute liver failure and were all hospitalized, so quite sick and many heading toward liver transplant. It showed that the transplant-free survival significantly improved in those who received NAC with early-stage encephalopathy compared with those who received placebo. So, I do believe that there's a role for NAC for 72 hours in adults with severe DILI.

There was a similar study done in children with non-acetaminophen acute liver failure, and NAC did not help them. In fact, there was a signal of potential harm.

So, if you have someone under the age of 18 with acute liver failure from a drug, I would not give NAC. But for adults, you could certainly strongly consider it. Yet, that applies to less than 1% of patients with DILI, because most don't end up with acute liver failure.

Fortunately, we also have steroids to consider in highly selected patients, although I'm very cautious with them. They've never done a randomized controlled trial, but there's some evidence that compares steroids to just waiting to see if the liver injury would go away. But those are commonly used in medical practice, oftentimes prescribed by the oncologist for patients with moderate to severe checkpoint inhibitor hepatitis.

The only other group with reasonably good evidence for steroid use is in patients with drug reaction with eosinophilia and systemic symptoms (DRESS). With DRESS, the liver is oftentimes not the main organ, but rather the skin and other solid organs. There's prospective studies in patients with DRESS who are quite ill showing that high-dose, oral steroids for around 2-3 months can be beneficial.

Beyond that, you're a little bit in no-man's land on pulling the trigger on steroids.

I would strongly encourage following patients closely. If they're heading toward liver failure in the hospital, getting them to a liver transplant center is your next move, because the outcome of acute liver failure from idiosyncratic DILI is not good. It's only about a 30% spontaneous transplant-free survival.

Resources for Identifying DILI

I think the LiverTox website is an incredible resource that's evolved from what it once was. Now, it's kind of like you get a book.

There's over 1000 chapters now written on the website. Nearly all of the medications that we know cause idiosyncratic DILI are in there. It's basically an encyclopedia that is updated periodically. There are also now 60 chapters on dietary supplements.

I encourage you to look at that content when you're encountering one of those more difficult cases. There is a committee appointed through the National Library of Medicine who reviews all the content and updates it regularly. So, these are very carefully vetted chapters and updates.

We would encourage anyone in the United States treating a patient whom they believe has idiosyncratic DILI from any drug, or herbal and dietary supplement, to contact one of the six clinical sites in DILIN, because we would love to enroll them into our studies. That's how we accelerate our learning on things like drug-specific genetic susceptibility factors, natural history, biomarkers, and so forth.

The DILIN website itself is also very useful. We have almost 90 papers that give a great deal of information about DILI due to specific agents. That's a handy, quick reference, and it's free.

Nancy S. Reau, MD, is chief of the hepatology section at Rush University Medical Center in Chicago and a regular contributor to Medscape. She serves as editor of Clinical Liver Disease, a multimedia review journal, and recently as a member of, a web-based resource from the American Association for the Study of Liver Diseases (AASLD) and the Infectious Diseases Society of America, as well as educational chair of the AASLD hepatitis C special interest group. She continues to have an active role in the hepatology interest group of the World Gastroenterology Organisation and the American Liver Foundation at the regional and national levels.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.