A modified multiparametric assay using HepaRG cells for predicting the degree of drug‐induced liver injury risk

The approach for predicting the degree of drug‐induced liver injury (DILI) risk was investigated quantitatively in a modified multiparametric assay using HepaRG cells. Thirty‐eight drugs were classified by DILI risk into five categories based on drug labels approved by the Food and Drug Administration (FDA) as follows: withdrawn (WDN), boxed warning (BW), warnings and precautions (WP), adverse reactions (AR), and no match (NM). Also, WP was classified into two categories: high and low concern. Differentiated HepaRG cells were treated with drugs for 24 h. The maximum concentration was set at 100‐fold the therapeutic maximum plasma concentration (C_max). After treatment with drugs, the cell viability, glutathione content, caspase 3/7 activity, lactate dehydrogenase leakage and albumin secretion were measured. As modified cut‐off values of each parameter, the TC₅₀ (toxic concentration that decreased the response by 50%) and EC₂₀₀ (effective concentration giving a response equal to 200% of controls) were calculated. In addition, the toxicity score (total sum score of the cytotoxic level of each parameter) was calculated. This modified multiparametric assay showed an 87% sensitivity and 87% specificity for predicting the DILI risk. The toxicity score showed a good predictive performance for WDN, BW and WP (high concern) categories [cut‐off: score ≥ 1; area under a receiver operating characteristic curve (ROC‐AUC): 0.88], and for WDN and BW categories (cut‐off: score ≥ 3; ROC‐AUC: 0.88). This study newly indicated that the degree of DILI risk might be predictable quantitatively by assessing the toxicity score in the modified multiparametric assay using HepaRG cells. Copyright © 2016 John Wiley & Sons, Ltd.