Idiosyncratic drug induced liver injury (DILI) is a rare adverse drug reaction that poses major challenges to healthcare practitioners and regulatory agencies. We aimed to update the clinical characteristics and outcomes of DILI patients, and the drugs frequently implicated in hepatotoxicity in Spain. We analyzed 915 DILI cases (842 single episodes, 55 re-challenges, 18 double-episodes due to different drugs) in 857 patients included in the Spanish DILI Registry from 1994 to 2015. Cases were adjudicated using expert clinical judgment/RUCAM scale and compared according to pattern of liver damage (hepatocellular, HC; cholestatic, Chol or mixed, Mix). The cohort median age was 57 years (range: 11-90 y) with a mean body mass index of 25.8 ± 3.8 kg/m2. Male gender predominated (52%). HC, Chol and Mix patterns of liver damage were identified in 65%, 18% and 17% of cases, respectively. More than half of the cases were of moderate severity (58%). Patients with Chol and Mix pattern were older (median 64 y and 62 y, respectively) than HC patients (median 52 y), p<0.001. Anti-infectives, central nervous system, cardiovascular and anti-inflammatory agents were the most commonly implicated therapeutic classes accounting for 37%, 14%, 11% and 9% of cases, respectively. Amoxicillin/clavulanate remains the agent responsible for the highest number of DILI (21% of cases). Substantial increase in anabolic androgenic steroid-induced hepatotoxicity was observed in recent years. A cluster of DILI cases reported to the registry (i.e. ebrotidine, tetrabamate, nimesulide, amoxicillin-clavulanate, Exolise®, Epistane®) contributed to adoption of regulatory measures. The pioneering prospective Spanish DILI Registry proved to be very valuable for in-depth clinical phenotyping of hepatotoxicity, providing consistent figures in clinical characteristic outcomes and implicated drugs. It also constitutes an important tool for public health promotion in postmarketing drug surveillance.

Most drug-induced liver injury (DILI) patients develop symptoms while on drug treatments; however in some instances DILI develops after drug cessation. The reasons behind this delayed onset are currently unknown. In this study we aimed to determine drug-properties and host-factors potentially involved in DILI with delayed onset. 413 DILI cases from the Spanish DILI Registry were classified according to time of first symptom appearance, during (no delay onset, NDO) or after (delay onset, DO) causative agent treatment. The drugs were divided into two groups: drugs with (WP) or without potential (WOP) to induce delayed onset. Only oral drugs with at least three cases in the registry were selected. Amoxicillin-clavulanate (AC) cases were analyzed independently. A total of 11 WP and 33 WOP drugs were found, corresponding to 204 and 209 cases, respectively. Ninety-four (46%) cases, of which 83% were due to AC, presented DO between 2-52 days after treatment cessation. The WP drug cases, omitting AC cases, showed shorter duration of treatment (11 vs 54 days, p=0.0001), time to onset (14 vs 40 days, p=0.0001) and higher mean dosage (744 vs 374 mg, p=0.0001) than the cases induced by WOP drugs. The DO cases showed lower incidence of comorbidities (p=0.0001). In terms of drug properties, the WP drugs showed lower ratio of hepatic metabolism (p=0.0190) and higher proportion of parental drug excretion (p=0.0015). The interactions analyzed showed that drugs with mitochondrial liability have stronger effect on delayed onset in younger patients, females and absence of cardiac diseases. Delayed onset potential appears associated with higher drug dose, shorter treatment and less comorbidities. The fact that drugs with low hepatic metabolism are more frequently associated with delayed onset suggests that additional mechanism(s) outside hepatic metabolism influence the probability of delayed onset.