Digestive Diseases News
Winter 2009

NIDDK Hosts Workshop on Drug-induced Liver Injury

An international group of scientists met December 1–2, 2008, at the National Institutes of Health in Bethesda, MD, to help develop a standardized approach to diagnosing and determining the severity of drug-induced liver injury (DILI). The workshop, organized by the Liver Disease Research Branch of the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), was designed to foster discussion among DILI experts and stimulate international collaboration.

Recognizing DILI is a challenge, as it mimics all forms of acute and chronic liver disease, according to Neil Kaplowitz, M.D., director of the Research Center for Liver Diseases at the University of Southern California, Los Angeles. DILI-specific biomarkers have yet to be found. DILI is often reversible, but failure to recognize it can quickly lead to serious health complications, even death. “Predicting those at risk for drug-induced liver injury is a critical issue,” Kaplowitz said.

DILI is the leading cause of acute liver failure in the United States and is becoming an increasingly important health concern as more people take prescription, over-the-counter, and herbal medications. Liver injury is also the major reason drugs fail to win approval from the U.S. Food and Drug Administration.

Although the incidence of DILI is unclear—relying on spontaneous reporting from health care professionals—serious adverse drug reactions account for about 4.7 percent of hospital admissions in the United States. While reactions to specific drugs are thought to be rare, according to Einar Bjornsson, M.D., Ph.D., professor of gastroenterology and hepatology at the Sahlgrenska Hospital, Gothenburg, Sweden, “there is a huge underreporting of adverse reactions in general,” suggesting cases of DILI may also be much higher. Collectively, DILI cases represent a significant health problem, much of which is preventable.

Standardizing DILI Assessment

The Roussel Uclaf Causality Assessment Method (RUCAM) is the most widely known instrument for assessing DILI. Using RUCAM, a health care provider derives a score indicating the likelihood that liver injury is due to a specific medication, based on a patient’s clinical, biological, serological, and radiologic features. RUCAM, however, has not been widely adopted in the clinic.

“Components of RUCAM are broken,” said Paul Watkins, M.D., director of the Hamner-UNC Center for Drug Safety Sciences, Research Triangle Park, NC, who co-chaired a session about issues related to causality assessment.

RUCAM’s shortcomings include ambiguous definitions, lack of reproducibility, and components that are not supported by data, according to Maribel Lucena, M.D., professor of clinical pharmacology at the University of Malaga in Spain. Above all, RUCAM’s complexity limits its practical use. A show of hands at the workshop revealed only about 10 percent of participants use RUCAM.

Despite its shortcomings, many at the meeting believe it is better to fix RUCAM than start from scratch. “What we’d like to have is a computer [program] that you could put the data into to calculate the RUCAM score,” said Jay Hoofnagle, M.D., director of the NIDDK’s Liver Disease Research Branch. Such a program, according to Hoofnagle, would facilitate testing and deciding which components to include in a revised RUCAM.

The DILI Network

The impetus behind the workshop was the NIDDK’s DILI Network (DILIN), initiated in 2003 to advance understanding and research into DILI. The DILIN, said Leonard Seeff, M.D., senior scientist at the NIDDK’s Liver Disease Research Branch, provides a stimulus and resource for research and clinical investigation of all forms of DILI.

The DILIN, a collaboration among investigators at nine clinical study centers plus one data coordination center and the NIDDK, is sponsoring retrospective and prospective DILI studies with corresponding patient registries. Both studies are enabling investigators to examine clinical, genetic, immunological, and environmental risk factors in slightly different ways.

The retrospective study is unraveling why certain drugs elicit DILI in only a slim minority of people. The study is examining four specific drugs: isoniazid (INH), phenytoin (Dilantin), clavulanic acid/amoxicillin (Augmentin), and valproic acid (Depakote). DILI signs and symptoms resulting from these agents are relatively well-defined, providing the opportunity to explore common environmental and genetic contributors. The retrospective registry houses clinical details about DILI episodes along with DNA and blood samples.

The more ambitious prospective study is allowing scientists to study idiosyncratic DILI during the initial phase: from first presentation onward. Included are cases of DILI caused by all prescription and over-the-counter medications, herbals, and nutritional supplements. Once enrolled, patients are followed for a minimum of 6 months but possibly as long as 20 years. Controls—volunteers with similar drug exposure but no DILI—will be enrolled for comparison.

Improving instruments for causality assessment is one of the primary objectives of the prospective study. Three DILIN investigators review the clinical, diagnostic, and laboratory data for each suspected DILI case and then assign causality, first by expert opinion and then by RUCAM.

The DILIN worked out a scoring system similar to RUCAM. Like RUCAM, the expert opinion method derives a score indicating the likelihood that DILI is due to a specific medication. Scores range from 1 (definite) to 5 (highly unlikely). To put the scores into context, the DILIN assigned descriptive terms and percentages to the scores. For example, a score of 1 is defined by the DILIN as “Liver injury is typical for the drug or herbal product. The evidence for causality is beyond reasonable doubt,” with a 95 percent likelihood that the drug in question is responsible.

A comparison of the two causality assessment methods showed the DILIN expert opinion method had less variability and that RUCAM tended to underestimate drug injury association, according to Don Rockey, M.D., professor and chief of the division of digestive and liver diseases at the University of Texas Southwestern Medical Center. He said that although the expert opinion method cannot be applied broadly, as too few hepatologists exist relative to the number of suspected cases, he believes study data will help develop a more generalizable and quantitative assessment method.

For information about the NIDDK’s Liver Disease Research Branch, go to www2.niddk.nih.gov/Research/ScientificAreas/Liver/DILD.htm.

Member Organizations of the Drug-Induced Liver Injury Network

• Duke University, Durham, NC (Data Coordinating Center)
• Indiana University, Indianapolis
• Mayo Clinic, Rochester, MN
• Thomas Jefferson University, Philadelphia
• University of California, San Francisco
• University of Michigan, Ann Arbor
• University of North Carolina, Chapel Hill
• University of Pennsylvania, Philadelphia
• University of Southern California, Los Angeles
• University of Texas Southwestern, Dallas

For more information about the DILIN, go to http://dilin.dcri.duke.edu.

[Previous Article] [Top] [Next Article]

NIH Publication No. 09–4552
March 2009