Histopathological Challenges in Suspected Drug-induced Liver Injury

David E. Kleiner


Liver International. 2018;38(2):198-209. 

In This Article

Abstract and Introduction


When a patient with suspected drug-induced liver injury (DILI) undergoes liver biopsy, the pathologist is confronted with two major challenges. The first and most important is to establish the pattern(s) of injury which are present. Patterns of injury represent stereotypical responses of an organ to injury and relate to specific aetiologies of liver damage. The pattern of injury and the histological details of that injury can then be analysed with respect to the patient's intercurrent diseases and medication history. The specific expertise of the pathologist can be used to weigh the prospect of DILI against the likelihood of other explanations of injury. The second challenge is to characterize specific types of injury and the severity of injury, both of which may have importance for clinical decision-making and prognosis. The pathologist's report should convey both an accurate description of the pathology as well its interpretation.


Liver biopsy is not an obligate part of the evaluation of patients with suspected drug-induced liver injury (DILI), but it is a frequently performed procedure. In the U.S. Drug-Induced Liver Injury Network (DILIN), a multicenter prospective registry of suspected and confirmed cases of DILI, liver biopsy was performed in about 50% of patients.[1] Although the utility of liver biopsy in the evaluation of DILI has not been rigorously evaluated, as a single clinical test it has the potential to provide a wealth of direct information about the character of the liver injury. In the broadest sense, a liver biopsy provides information about both the severity of injury and the overall morphological changes, generally termed the pattern of injury (Table 1).

The pathologist faces several challenges in the interpretation of liver biopsies from patients with suspected DILI. First, DILI is uncommon—population-based studies have identified an incidence rate that varies from 8.1 per 100 000 in a 3-year study from the Nièvre district of France[2] to 19.1 per 100 000 identified in a 2-year survey in Iceland.[3] To put these numbers in perspective, the overall incidence of DILI is about 10 times that of autoimmune hepatitis (AIH) (1.9 per 100 000),[4] primary biliary cholangitis (0.2–2.4 per 100 000)[5,6] and primary biliary sclerosis (1.3 per 100 000),[4] three uncommon diseases that often enter into the differential diagnosis in suspected DILI. Of course, DILI is not a single entity, but many different drug-specific injuries, which will lower the incidence of any one type of drug injury to far less than 1 per 100 000. As the use of liver biopsy to evaluate these diseases is limited, the opportunity for any pathologist to gain wide experience in the histological evaluation of uncommon liver disease is further limited. The pathologist must develop a systematic approach to the pathology to overcome this limitation of practical experience.

Additional challenges are apparent in Table 1. Patients with known underlying liver disease—most often fatty liver disease or viral hepatitis—may develop an acute injury suspected to be DILI. The challenge to the pathologist involves distinguishing histological features that can be explained by the underlying disease and those that may be related to DILI. If the suspected drug causes an injury like that of the underlying disease (eg, suspected minocycline injury in a patient with chronic hepatitis C), it may be impossible to define a contribution to the injury by the drug. Nevertheless, there may be histological changes, like zonal necrosis or cholestasis, that would not be expected in pathology owing to the underlying disease, allowing the suggestion of superimposed DILI to be made. Examination of prior biopsies, if available, may become very important in the distinction of additional injury. Similarly, patients who are acutely ill, on multiple medications or in the intensive care unit present a different kind of challenge. Patients may be at risk for sepsis or hypoxic/ischaemic injury, which have their own distinctive patterns of injury. In cases of polypharmacy the pathologist may need to sort through the medication list to identify the most likely candidates for DILI. Finally, the pathologist may be asked for their opinion on whether the degree of injury would permit the continued careful use of a medication that cannot be replaced and is providing a clear, life-sustaining benefit. It is important that both the pathologist and the clinician maintain open communications during and after the biopsy evaluation. Severe or unusual injury may require immediate contact of the clinician. Information from additional clinical testing indicated by changes in the biopsy or from longer follow-up may provide the opportunity for deeper interpretation of the biopsy.