Daniel J Antoine
Department of Molecular and Clinical Pharmacology and MRC Centre for Drug Safety Science, University of Liverpool, UK.
The potential of mechanistic biomarkers to improve the prediction of drug-induced liver injury (DILI), a significant cause of patient morbidity, mortality and attrition within drug development is widely acknowledged. Prediction of clinical DILI remains difficult, particularly in cases characterized by marked inter-individual variation. A lack of sensitivity, specificity and an indirect mechanistic basis of currently used biomarkers of hepatic injury remains a factor for the delayed identification of DILI. Currently, ‘Hy’s law’ represents the regulatory endorsed model to predict serious DILI and is the standard for novel DILI biomarkers to surpass. Recently, the FDA and EMA issued letters of support to encourage the further development and qualification of new mechanistic DILI biomarkers. Clinical and pre-clinical DILI studies have shown the identification and development of circulating or imaging-based biomarkers that provide enhanced hepatic specificity (miR-122) and hepatic function (ICG clearance). These biomarkers can also inform on mechanistic events such as necrosis (keratin-18, HMGB1), apoptosis (caspase-cleaved keratin-18), mitochondrial dysfunction (glutamate dehydrogenase) and inflammation (acetyl-HMGB1, colony stimulating factor-1). Our experience of >1500 acetaminophen overdose patients and the application of these markers in Phase I liver safety studies has demonstrated that markers such as miR-122 provide a sensitive and early identification of liver injury that permits stratified treatment and that both acetyl-HMGB1 and colony stimulating factor-1 are highly prognostic. We have also investigated the impact of intra and inter-individual, as well as diurnal, variation in >200 healthy volunteers subjects, a critical aspect to help interpret novel biomarker data. Our research also adopts a ‘tissue to periphery’ based approach in which we link biomarker histological expression patterns to circulating values in preclinical studies and from specimens from DILI subjects. By using hepatocyte specific conditional knock out mice we have also demonstrated that HMGB1 plays an integral role in the mechanism of toxicity in experimental DILI by linking cell death to pro-inflammatory events. These data have also formed the basis of a programme of work that has led to the successful targeting of HMGB1 with monoclonal antibodies and inhibitory peptides. The integrated use of these biomarkers, therapeutic interventions and enhanced mechanistic understanding will be discussed in the context of drug safety science and understanding fundamental liver injury and repair.