AASLD2020 | Prof. Bjornsson E talking about the diagnosis and treatment of DILI——"Further exploration of strategies for the treatment of medically induced liver injury is very necessary"

 Editor's note: 《Hepatology Digest》 had the honor to interview Prof. Einar Bjornsson, from the National University of Iceland, who was one of the authors of the “European Society of Hepatology Clinical Practice Guide: Drug-induced Liver Injury”, to comment on the hot issues in the diagnosis and treatment of drug-induced liver injury.

Hepatology Digest: Drug-induced liver injury is estimated to be 14 to 19 cases per 100,000 persons. There are 3 types of drug-induced liver injury: direct hepatoxicity, idiosyncratic hepatotoxicity and indirect hepatotoxicity. What are the characteristics and differences among the 3 types of drug-induced liver injury?

Prof. Einar Björnsson: Before last year, drug-induced liver injury was divided into only two types, direct hepatotoxicity and idiosyncratic hepatotoxicity. But in a paper in the new England journal of medicine that was published in 2019, this 3rd type indirect hepatotoxicity was introduced. 

The characteristics of the direct hepatotoxicity is that it's dose-related. The higher the dose of the agent, the higher risk of liver injury. The phenotype is cluttered by acute hepatic necrosis in several cases, this is usually related to high exposure and high doses of the drug. Most people are familiar with acetaminophen, as is called in the US but paracetamol mostly in Europe. But other type of drugs leading to direct hepatotoxicity is cancer chemotherapy of the old type and methotrexate, which is used in some type of cancer therapy, for example, in children and in other conditions. And this can also be related to amiodarone, which is an anti-arrhythmic drug given intravenously. All these drugs can cause direct hepatotoxicity which is dose-related as I mentioned earlier. 

However, the idiosyncratic type is rare. It's a rare adverse effect of a drug that leads to liver injury when most people tolerate the drug in the same dose or the same duration. This is usually not related to the dose, although recently there's been shown that this is not without dose relationship. For some drugs, the higher the dose, the more likely, although within the therapeutic range, that they can develop idiosyncratic injury, which is thought to be related to the metabolic or immunological reaction probably in many cases. for example, antibiotics such as amoxicillin and clavulanic acid, and isoniazid used for tuberculosis are typical drugs that can lead to idiosyncratic drug-induced liver injury. Many other drugs are of this type. One of the most common biochemical and histological phenotype is cholestatic hepatitis. Idiosyncratic type of reaction can also be hepatocellular with dominant AST and ALT elevation, or a cholestatic, or pure cholestatic reaction. A good example is a bland cholestasis related to the use of anabolic steroids. 

The 3rd type indirect hepatotoxicity is mostly caused by the action of the drug rather than its toxicity. In the idiosyncratic type, the hypothesis is that metabolic reaction can trigger the immune system in a certain amount, which leads to either direct toxicity in the susceptible individual, or immunological activation that can lead to injury. But this indirect hepatotoxicity is related mostly to what the drug does rather than what it is in the case of idiosyncratic reaction that is leading to direct toxicity. A good example is infliximab-induced hepatitis. It's not thought to be related to the toxicity of infliximab, which is not even metabolized within the liver, but is related to the imbalance in the immune system. This can also be seen with the increasingly used checkpoint inhibitors which lead to imbalance in the immune system. When the drug is interfering with the immune system, this can lead to all auto-immune phenomenon in different parts of the body and also within in the liver. To summarize the 3rd part, it's an indirect action of the agent on the liver or the immune system that leads to this reaction. 

Hepatology Digest: With few exceptions, there are no specific diagnostic markers for drug-induced liver injury. The traditional serologic markers are ALT/AST, ALP, GGT, albumin and TGF-β1. New serum markers include PON-1, miRNA (miR-122, miR-129), GLDH, GSTs and so on. In your opinion, which one is the most promising marker in terms of sensitivity, specificity and predictivity? What is the development prospect of diagnostic markers?

Prof. Einar Björnsson: This is a very good but also a very difficult question. Unfortunately, I cannot state that I'm an expert in biomarkers but this is obviously a very important question. There is a big medical need to find the biological marker for DILI. As we know, we don't have a biological marker. When patients are suspected of having hepatitis C we can measure antibodies against hepatitis C, and also the virus itself in the blood with appropriate molecular diagnostic methods. But we lack up specific marker for DILI. We rarely have a complete proof of the role of the drug and relies on many other things and exclusion of other etiologies. This issue has been intensively studied in recent years. One of the studies that have tried to validate the biomarkers with DILI is so-called SAFE-T which was an innovative medicine initiative, Safer and Faster Evidence-based Translation consortium in Europe. There were several markers that mentioned here, for example, the miRNA. The SAFE-T study was ongoing and dealing with many different drugs, different adverse effects, also analyzing kidney injury, vascular injury and central nervous system injury. In terms of the DILI, the collection of cases is ongoing to evaluate these new biomarkers. These biomarkers are measured in patients with suspected DILI as both a mechanistic assessment for the severity of reaction and also for early diagnosis. Obviously it's a big medical need and would be very appropriate if we could have a market. In some way it can distinguish DILI from other causes of acute liver injury. In this TransBioLine study, cases are recruited in the PRO prospective European DILI registry, Pro-Euro DILI within Europe. Some of those patients who are suspected of DILI turn out to have other diagnoses such as virus hepatitis and cases who have an alternative diagnosis were used as a control group. This is valuable because one hypothesis is that some of these biomarkers can be used for early diagnosis. Hopefully they can be used to distinguish between DILI itself and other diagnoses. These biomarkers are for mechanistic assessment, early diagnosis, and also for predicting the course of reaction and prognosis. Some patients who develop DILI, will also particularly develop jaundice or acute liver failure, either die from liver failure or need a liver transplantation. But so far, I don't think we have very good evidence that these biomarkers are better than the others. Other biomarkers for example microRNA are tested. I cannot say which is the most promising biomarkers. I think we have to wait for the results of this TransBioLine effort.

Hepatology Digest: If there is a drug-induced liver injury treatment IND (Investigational New Drug) approved for clinical trials, the population are all liver injury patient. We know that the all the new drugs have the probability of liver injury. So how to balance the benefits and risks in the trials? How to reduce the risk of liver function deterioration in the trials?

Prof. Einar Björnsson：Thank you for this important question. One of the goals of the clinical therapeutic trials is to alter the natural history of patients with DILI. This will be only for a specific part of those with DILI because most patients who have mild and moderate DILI don't require any treatment at all because they will recover when the implicated agents are discontinued. But a certain group of DILI, for example patients with anti-tuberculosis regiments can die of liver failure. There is a medical need to try to develop a therapeutic strategy that can save these patients with severe DILI. 

I think this is a very good and important question because we don't want to further worsen this liver injury by giving them another potentially hepatotoxic drug. But I'm I don't really completely agree with this question that all new drugs have the probability of liver injury. When you look at drugs on the market, for example, drugs that are registered in the United States, only approximately 50% of these drugs have a potential of DILI when scrutinized. For example, a drug that was given in a US RCT study to treat acetaminophen-induced liver injury, showed that DILI can be therapeutic. Therapeutic advances have been shown in early phases. 

I'm quite sure that not all drugs on the market have the probability of liver injury. In clinical trials, other therapeutic agents that have been used for prevention and treatment of DILI have not shown convincingly to have effect on DILI. So if people would like to do a therapeutic trial on DILI, I think placebo would be comparative to our intervention. It is complicated to design a therapeutic trial on DILI and it's conceivable that at least in some patients an RCT system could be tried to as a treatment for advanced DILI. This design and endpoints need to be really analyzed before we can decide on a treatment for DILI. But obviously, I agree completely with that the people should be very cautious when starting a treatment. The treatment should not have a worsen effect. 

I think we need to come up with a therapeutic strategy in patients with very advanced DILI within the scientific community and the medical field because at this moment, we don't have any means other than a liver transplantation which unfortunately cannot be achieved by all patients.