Drug-induced hepatotoxicity, defined as liver injury caused by exposure to a medication, is the most frequent reason for withdrawal of marketed medications and one of the most common reasons for termination of otherwise promising therapeutic agents during pre-clinical studies. It is also the leading cause of acute liver failure among patients referredfor liver transplantation in both the U.S. and Europe. The determination of the comparative safety of medications is an area of major importance in Comparative Effectiveness Research. Given the clinical and public health impact of drug-induced hepatotoxicity, the development of methods to predict the likelihood of liver failure in the setting of hepatotoxicity and determination of the comparative risk of liver failure for medications within particular drug classes are crucial to ensuring the comparative safety and effectiveness of medications. To date, no studies have yet yielded a validated method to predict the potential for a medication to lead to liver failure, and no data have compared the risk of liver failure associated with different medications within drug classes. To address these issues, this proposal first seeks to develop and validate a predictive index to classify patients with hepatotoxicity by their risk of progression to liver failure (Aim 1). Using data from Kaiser Permanente Northern California (KPNC), we will start by evaluating the performance of the well known but unproven "Hy's Law", and then seek to improve upon it, by modifying cut-off points for liver aminotransferases and total bilirubin, examining their rate of rise, and evaluating whether additional commonly available biomarkers improve predictive ability to identify outcomes of acute liver failure. Once the parameters that best predict liver failure are determined, internal validation will be performed. In the second phase of the study, external validation of the predictive index will be conducted using data from the National Veterans Affairs (VA) Health Information System (Aim 2), which will provide information on its generalizability, particularly in AHRQ priority populations. We will then compare the risk of liver failure associated with different medications used for the treatment of priority conditions of importance to the Medicare and Medicaid programs using data from both KPNC and the VA (Aim 3). We will first examine classes that include a medication with known hepatotoxicity, to confirm our approach, and then compare medications within other important classes. Finally, using the index developed in Aim 1, we will determine the risk of severe hepatotoxicity for medications in the drug classes evaluated in Aim 3, to provide further evidence on the comparative hepatic safety of these medications (Aim 4). Thus, the overall goal of this series of studies is to produce new methods and valuable data that will enhance substantially the comparative safety and effectiveness of drug therapies for a variety of priority clinical conditions, with a particular focus on their comparative hepatic safety.

Public Health Relevance

Given the clinical and public health significance of drug-related hepatotoxicity, the development of a validated index to identify accurately signals that predict the potential for severe hepatic injury would enable differentiation of drug therapies with little likelihood for severe liver injury from those with increased potential for liver failure. Comparing within drug class the association between the use of different medications and liver failure is necessary to provide evidence on the comparative hepatic safety of these medications, informing decision-making about the appropriate treatments for various conditions and settings.

Agency
National Institute of Health (NIH)
Type
Research Project (R01)
Project #
5R01HS018372-05
Application #
8705390
Study Section
Health Care Technology and Decision Science (HTDS)
Program Officer
Perfetto, Deborah
Project Start
Project End
Budget Start
Budget End
Support Year
5
Fiscal Year
2014
Total Cost
Indirect Cost
Name
University of Pennsylvania
Department
Biostatistics & Other Math Sci
Type
Schools of Medicine
DUNS #
City
Philadelphia
State
PA
Country
United States
Zip Code
19104