Mark I. Avigan

Drug-induced liver injury (DILI) continues to be a significant challenge in drug development. Just 1 or 2 cases of severe hepatotoxicity in a clinical trial can be enough to raise drug approval concerns. Though great strides were made in DILI risk assessment and mitigation in clinical trials during the early 2000s, drugs continue to fail in development due to DILI, while severe cases still occur in late-phase clinical trials and post-market. Also, changes in drug development will challenge the US Food and Drug Administration’s (FDA) current paradigms for DILI risk assessment. Such changes include the rise in biologic agents, which may cause hepatotoxicity distinctly different from small-molecule drugs, the increasing reliance on smaller, underpowered clinical trials for rare diseases, and the growing interest in drugs for acute and chronic liver diseases. We highlight eight key topics that may benefit from review, clarification, research, and discussion between industry, academia, and regulators as the FDA works to address these issues.

Immune checkpoint inhibitors (ICIs) block CTLA‐4, PD‐1 and PD‐L1, or other molecules that control antitumour activities of lymphocytes. These products are associated with a broad array of immune‐related toxicities affecting a variety of organs, including the liver. ICI‐associated immune‐mediated hepatitis (IMH) ranges in severity between mild and life‐threatening and is marked by findings that bear both similarities as well as differences with idiopathic autoimmune hepatitis. Hepatotoxic events are often detected in clinical trials of ICIs that are powered for efficacy. Risk levels for ICI‐induced liver injury may be impacted by the specific checkpoint molecule targeted for treatment, the ICI dose levels, and the presence of a pre‐existing autoimmune diathesis, chronic infection or tumour cells which infiltrate the liver parenchyma. When patients develop liver injury during ICI treatment, a prompt assessment of the cause of injury, in conjunction with the application of measures to optimally manage the adverse event, should be made. Strategies to manage the risk of IMH include the performance of pretreatment liver tests with regular monitoring during and after ICI treatment and patient education. Using Common Terminology Criteria for Adverse Events developed at the National Cancer Institute to measure the severity level of liver injury, recommended actions may include continued ICI treatment with close patient monitoring, ICI treatment suspension or discontinuation and/or administration of corticosteroids or, when necessary, a non‐steroidal immunosuppressive agent. The elucidation of reliable predictors of tumour‐specific ICI treatment responses, as well as an increased susceptibility for clinically serious immune‐related adverse events, would help optimize treatment decisions for individual patients.

In the United States (US), the risk of hepatotoxicity linked to the widespread use of certain herbal products has gained increased attention among regulatory scientists. Based on current US law, all dietary supplements sold domestically, including botanical supplements, are regulated by the Food and Drug Administration (FDA) as a special category of foods. Under this designation, regulatory scientists do not routinely evaluate the efficacy of these products prior to their marketing, despite the content variability and phytochemical complexity that often characterizes them. Nonetheless, there has been notable progress in the development of advanced scientific methods to qualitatively and quantitatively measure ingredients and screen for contaminants and adulterants in botanical products when hepatotoxicity is recognized.

A workshop was convened to discuss best practices for the assessment of drug-induced liver injury (DILI) in clinical trials. In a breakout session, workshop attendees discussed necessary data elements and standards for the accurate measurement of DILI risk associated with new therapeutic agents in clinical trials. There was agreement that in order to achieve this goal the systematic acquisition of protocol-specified clinical measures and lab specimens from all study subjects is crucial. In addition, standard DILI terms that address the diverse clinical and pathologic signatures of DILI were considered essential. There was a strong consensus that clinical and lab analyses necessary for the evaluation of cases of acute liver injury should be consistent with the US Food and Drug Administration (FDA) guidance on pre-marketing risk assessment of DILI in clinical trials issued in 2009. A recommendation that liver injury case review and management be guided by clinicians with hepatologic expertise was made. Of note, there was agreement that emerging DILI signals should prompt the systematic collection of candidate pharmacogenomic, proteomic and/or metabonomic biomarkers from all study subjects. The use of emerging standardized clinical terminology, CRFs and graphic tools for data review to enable harmonization across clinical trials was strongly encouraged. Many of the recommendations made in the breakout session are in alignment with those made in the other parallel sessions on methodology to assess clinical liver safety data, causality assessment for suspected DILI, and liver safety assessment in special populations (hepatitis B, C, and oncology trials). Nonetheless, a few outstanding issues remain for future consideration.

Background Early prediction and accurate characterization of risk for serious liver injury associated with newly marketed drugs remains an important challenge for clinicians, the pharmaceutical industry, and regulators. To date, a biomarker that specifically indicates exposure to a drug as the etiologic cause of liver injury has not been identified. Objectives Using cumulative proportional reporting ratios (PRRs), we investigated ‘real-time’ profiles of a set of pharmaceuticals, over the first 3 years of US marketing, for the signaling of clinically serious drug-induced liver injury (DILI) in a large spontaneous-reporting database. Methods Using report counts of hepatic failure or clinically serious liver injury obtained from the FDA Adverse Events Reporting System (FAERS) database, PRRs of adverse drug event terms were calculated by division of counts of domestic reports of these events by counts of all serious adverse events for each of 13 selected drugs associated with a broad range of hepatotoxic risk (including three linked to only rare instances of clinically apparent liver injury) with reference to all other drugs in the database. Drug-specific cumulative PRRs were measured at successive intervals (calendar quarters) using cumulative tallies of FAERS reports to generate time-based profiles over the initial 3 years of US marketing. Results In the set of drugs analyzed, those with no known hepatotoxic risk demonstrated time-based cumulative PRR profiles that approximate the background rates of hepatic failure and serious liver injury reported in the entire FAERS database. In contrast, those that were removed from marketing or subjected to marketing restrictions due to their potential to cause liver injury were associated with profiles of rapidly rising cumulative PRRs that were greater than 5 within the first 10 million domestic prescriptions or the first four quarters of US marketing. The systematic tracking and identification of rising PRRs for DILI associated with newly marketed pharmaceutical and biological agents is a valuable tool for identification of safety signals within the FAERS database. Limitations Disproportionality profiling of spontaneous reports in FAERS (e.g., cumulative PRR measurements), which signals an association between a recently marketed drug and liver injury, is not a method to quantitatively measure drug-related risk. Regulatory actions in response to emerging drug safety concerns often depend on an accurate assessment of risks using multiple sources of data and the consideration of overall benefits and risks of the agent. Causality must be determined through analysis of individual cases to exclude other etiologies of liver injury. Conclusion The FAERS database can be used to advance empiric hepatotoxicity time-trending reporting levels for newly marketed agents in order to rapidly identify recently launched potential hepatotoxic agents and initiate further evaluation.

Marketed pharmaceuticals are evaluated for safety by the US Food and Drug Administration (FDA) throughout the life cycle of the products. The FDA uses data from controlled clinical trials, from postmarketing case reports reported to the FDA's Adverse Event Reporting System, from epidemiological studies, and from registries to evaluate the safety of approved products. For some products, including some products used in dermatologic medicine, risks become apparent during the postmarketing period that require additional measures beyond product labeling and routine pharmacovigilance. The FDA continues to seek additional tools to assess risk, including pharmacogenomic biomarkers for adverse drug reactions and the use of large medical record and epidemiological databases for the systematic detection and characterization of drug-associated safety outcomes.

Two overarching goals tied to a futuristic vision of personalized medicine are to maximize individually derived benefits from available therapeutic drugs or biological agents and identify a reliable set of predictors of toxic outcomes to optimally manage a patient's treatment-related risk. This article will primarily address the second of these goals with a specific focus on challenges faced in the development of pharmacogenomic markers that would be used by practitioners to predict individual susceptibility to serious adverse drug reactions. The potential complexity of polygenic modifiers of risk for certain adverse drug reactions and demographic diversity in the prevalence of genomic variants pose significant challenges in the development of useful predictive biomarkers. Implementing rigorous scientific and logistical strategies to address these will be crucial in order to achieve meaningful success.