Updated as of 16 September 2011

Background and Rationale

Liver injury due to prescription and non-prescription medication use is a medical, scientific, and public health problem of increasing frequency and importance in the United States. Indeed, drug-induced liver injury (DILI) is the most common reason for nonapproval, withdrawal, limitation in use, and clinical monitoring by the Food and Drug Administration (FDA). However, detection of signals for liver injury frequently relies upon the reporting of cases by practitioners to health authorities in post-marketing surveillance. Underreporting of cases, lack of mandatory reporting systems, and difficulties in establishing a diagnosis make the current system sub-optimal. Moreover, with the growing use of complementary and alternative medications (CAM), there have also been increasing reports of liver toxicity due to various non-prescription herbal, dietary, and food additive supplements. Because the manufacturing, dispensing, and testing of these products is not regulated, the hepatotoxic potential of these formulations is poorly characterized or completely unknown.

As a result, there is a great need to develop an improved means of detecting, defining, and studying DILI in the United States. The DILIN Prospective Study is a multi-center study designed to gather clinical information and biological specimens on cases of suspected liver injury due to drugs and CAM. The goals of this study include the earlier recognition of DILI, especially due to newer drugs, development of standardized instruments and terminology to help identify cases of DILI, investigating clinical and genetic risk factors that predict DILI, and performing a careful longitudinal follow-up of DILI subjects. The biological samples collected will be used in future studies of the mechanisms and genetics of DILI.

Specific Aims and Objectives

The primary objective of this study is to prospectively identify bona fide cases of liver injury due to drugs and CAM within 6 months of presentation. Secondary objectives include collecting clinical data and biological specimens including blood, DNA, urine, and liver tissue from affected patients for future mechanistic and genetic studies. We will also investigate the clinical, immunological, and environmental risk factors of drug-mediated hepatotoxicity by comparing DILI cases to matched controls with a similar drug exposure history but no evidence of clinically significant liver injury. The natural history of drug- and CAM-induced DILI will be tracked for at least 6 months following enrollment, with longer follow-up for those in whom there is evidence of chronic liver injury at 6 months. We will also develop and test causality assessment instruments for drug and CAM-induced liver injury that are sensitive, specific, and reproducible.

Basic Study Design

The DILIN Prospective Study is a multi-center, prospective, epidemiological study. Patients who are referred to one of the DILIN clinical sites and who, in the opinion of a gastroenterologist / hepatologist, experienced a drug-induced liver injury will be enrolled. Detailed clinical data and biological specimens will be collected. Clinical data will be reviewed by the DILIN Causality Committee, which will make the final determination of whether the subject qualifies as a bona fide DILI case. DILI cases will be followed for at least 6 months to derive the longitudinal profile of drug- and CAM-induced liver injury. Detailed clinical data and biological specimens will be collected at this time point. Patients who satisfy the definition of chronic DILI will be evaluated at 12 months and yearly thereafter.

Study Populations

Consecutive patients who are referred to one of the DILIN clinical sites and appear to have suffered a drug-induced liver injury will be considered for inclusion in the study. Subjects must be > 2 years at the time of enrollment; have evidence of liver injury that is known or suspected to be related to consumption of a drug or CAM product in the 6-month period prior to enrollment; and have documented clinically important DILI defined in terms of serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (Alk Phos). Subjects will be excluded if there is acetaminophen hepatotoxicity, a competing cause of acute liver injury, or liver transplant prior to the development of drug- or CAM-induced liver injury.

Data Collection Protocol

The “baseline” visit is the date of the first in-person visit by the participant to the DILIN clinical site. A number of evaluations will be performed at that time, including a detailed medication history of the implicated DILI drug plus any drugs from the same class of medications; the dose, duration, and indication for all medications including prescription, OTC, and herbal medications taken within 8 weeks of presentation; medical history; family history of drug allergies/ hepatotoxicity to the implicated drug or its class of drugs; and so on. Serological tests will be performed to exclude competing causes of acute liver disease; additionally, a complete blood count, kidney and liver function tests, total cholesterol, triglycerides, and a urinalysis will be performed. A blood sample will be drawn for DNA isolation, plasma, PBMC cryopreservation, and serum storage. A voided urine sample will be collected, aliquoted into cryovials at the clinical site, and frozen.

If the subject has not been previously imaged, a screening liver ultrasound will be obtained. A liver biopsy may be obtained for clinical purposes in cases of diagnostic uncertainty or for prognostic purposes. For example, if a patient fails to have improvement in liver biochemistries within 1 month of stopping the suspect drug or if there is a clinical suspicion of autoimmune hepatitis, a liver biopsy may be recommended for diagnostic purposes.

Follow-up evaluations will only be undertaken for DILI cases, and a subset of the tests will be repeated then. In patients with suspected chronic DILI (based on laboratory, clinical, or imaging criteria), a liver biopsy will be recommended for clinical purposes at the 12-month visit according to the local standard of care. A schedule of evaluations is provided.

Statistical Considerations

The primary goals of this study are to create a clinical database consisting of individuals who have experienced a drug-induced liver injury, to create a bank of biological specimens obtained from the-se individuals so that genomic DNA as well as EBV-transformed, immortalized lymphoblasts are available for continuing functional or genetic studies, and to follow prospectively DILI cases to ob-serve the natural history of their disease. The DILIN Prospective Study is designed as a multi-center, prospective, cohort study. That is, DILI cases will be identified and enrolled by specialists at each participating clinical site. Detailed exposure information prior to the DILI event will be collected using a telephone / personal interview with the participants as well as from patient charts and medical records. Moreover, DILI cases will also be followed prospectively so that the course of the illness, including results of follow-up physical examinations, liver tests, quality of life, other clinical, laboratory, and imaging features following the DILI event, can be observed.

Although we would like to include many thousands of DILI cases in the study, the study is limited by the network’s capacity to find and enroll DILI cases over the timeframe of this study. Genetic studies are one of the main focuses in DILIN. For genetics analyses, drug matched controls may not be necessary. This is because DILI is extremely rare (ranging from 1:500 to fewer than 1:10,000 prescriptions, depending on the precipitant drug) so there is a very low probability of misclassification, i.e., DILI cases contaminating the control series. Without drug-matched controls, we plan to use population controls to carry out genetic analyses to determine DILI-associated genetic risk factors. Sample size and power calculations are carried out based on the proposed genetic analyses using existing population control subjects.

Genetic studies will utilize existing population control subjects with available genomic data in case-control analyses. For genome-wide association studies (GWAS), the control cohort includes approximately 6,000 unrelated individuals of primarily European ancestry enrolled in the 1958 British Birth Cohort or the United Kingdom National Blood Service control cohort and genotyped as part of the Wellcome Trust Case Control Consortium (http://www.wtccc.org.uk/ccc2/). Additional controls for GWAS and whole-genome or whole-exome sequencing have been made available by the Duke Center for Human Genome Variation (serving as the genomic center for DILIN).

Study Administration

The administrative and funding mechanism used to undertake this project is an NIH “Cooperative Agreement,” which is an assistance mechanism. Under the cooperative agreement, the NIDDK assists, supports, and/or stimulates and is substantially involved with investigators in conducting the study by facilitating performance of the effort in a “partner” role. The Steering Committee is the main governing body of the project. It is composed of the principal investigators of the clinical centers, the principal investigator of the data coordinating center, and the NIDDK project scientist. The clinical centers, the data coordinating center and NIDDK each have 1 vote on the Steering Committee. All decisions are determined by majority vote. In addition, a number of subcommittees have been established and report to the main Steering Committee.