Predicting Drug Action
Abstract
It is a given in clinical medicine that patients vary in their responses to drug therapy: lack of efficacy is common and serious adverse drug reactions (ADRs) remain a leading cause of morbidity and mortality. Several decades of work in pharmacogenomics have illuminated critical genetic pathways contributing to this variability.
Prominent successes in the field have identified single common variants with large effect sizes for important drug outcomes, and some of these are now being implemented clinically at our center and elsewhere to increase the likelihood of efficacy or to identify patients at high risk for serious ADRs.
However, despite initial application of candidate and unbiased genomic methodologies to large populations, major gaps remain in our understanding of mechanisms of individual variability in drug action. These gaps, in turn, highlight the major unmet needs of modern pharmacogenomic science: to develop robust predictors of drug response in an individual patient prior to embarking on treatment, to develop highly efficient screens to detect ADR liabilities in new drug candidates prior to investment in their development and patent harm, and to exploit an understanding of the molecular basis of drug action to rationally target available drugs.
Prominent successes in the field have identified single common variants with large effect sizes for important drug outcomes, and some of these are now being implemented clinically at our center and elsewhere to increase the likelihood of efficacy or to identify patients at high risk for serious ADRs.
However, despite initial application of candidate and unbiased genomic methodologies to large populations, major gaps remain in our understanding of mechanisms of individual variability in drug action. These gaps, in turn, highlight the major unmet needs of modern pharmacogenomic science: to develop robust predictors of drug response in an individual patient prior to embarking on treatment, to develop highly efficient screens to detect ADR liabilities in new drug candidates prior to investment in their development and patent harm, and to exploit an understanding of the molecular basis of drug action to rationally target available drugs.
Projects
Project 1
|
|
The drug-induced long QT syndrome (diLQTS) continues to be a problem for clinicians balancing risk and benefits across multiple therapeutic areas, and a fundamental issue that this Project will address is the extent to which diLQTS risk is predictable in an individual patient.
QT interval prolongation and arrhythmias have been a major cause for drug relabeling and withdrawals. However, only a small minority of patients exposed to culprit drugs develops the ADR, and the fundamental determinants of this individual sensitivity remain unexplained. In this project we will derive cardiomyocytes from individual patients whose drug-response phenotypes we have established. We will compare cells from those with drug-induced long QT syndrome (diLQTS) to those who have been drug-tolerant both at baseline and with exposure to drugs known to elicit this ADR, including HERG blockers as well as those inhibiting PI3-kinase, a new pathway to diLQTS we have recently delineated.
QT interval prolongation and arrhythmias have been a major cause for drug relabeling and withdrawals. However, only a small minority of patients exposed to culprit drugs develops the ADR, and the fundamental determinants of this individual sensitivity remain unexplained. In this project we will derive cardiomyocytes from individual patients whose drug-response phenotypes we have established. We will compare cells from those with drug-induced long QT syndrome (diLQTS) to those who have been drug-tolerant both at baseline and with exposure to drugs known to elicit this ADR, including HERG blockers as well as those inhibiting PI3-kinase, a new pathway to diLQTS we have recently delineated.
Dan Roden
Principle Investigator
|
Charles Hong
Co-Investigator
|
Bjorn Knollmann
Co-Investigator
|
Tao Yang
Co-Investigator
|
Project 2
|
|
Immunologically mediated adverse drug reactions (IM-ADRs) contribute disproportionately to drug-related morbidity and mortality and the cost and uncertainty of drug development, and results of these studies will inform strategies for the prediction of severe HLA-associated IM-ADRs and guide drug development and design.
Our work with abacavir stands as a landmark in translational success having identified a critical ADR risk allele that is now routinely tested in practice to prevent abacavir hypersensitivity. A major gap in our knowledge and hurdle to translation for many other HLA-mediated ADRs is that only a minority of patients carrying a risk allele develop the reaction. Using stored cells and DNA from patients with specific drug tolerant and hypersensitivity phenotypes we will define associations between HLA class I and II alleles, specific T-cell receptor (TCR) usage and severe T-cell mediated drug hypersensitivity. These resources will be further used to test a new heterologous immunity model of drug hypersensitivity that proposes that effector memory T cells that have been generated during an earlier infection cross recognize a second unrelated antigen, in this case a neo-antigen created only in the presence of drug.
Our work with abacavir stands as a landmark in translational success having identified a critical ADR risk allele that is now routinely tested in practice to prevent abacavir hypersensitivity. A major gap in our knowledge and hurdle to translation for many other HLA-mediated ADRs is that only a minority of patients carrying a risk allele develop the reaction. Using stored cells and DNA from patients with specific drug tolerant and hypersensitivity phenotypes we will define associations between HLA class I and II alleles, specific T-cell receptor (TCR) usage and severe T-cell mediated drug hypersensitivity. These resources will be further used to test a new heterologous immunity model of drug hypersensitivity that proposes that effector memory T cells that have been generated during an earlier infection cross recognize a second unrelated antigen, in this case a neo-antigen created only in the presence of drug.
Elizabeth Phillips
Principal Investigator
|
Simon Mallal
Co-Investigator
|
David Koelle
Co-Investigator
|
Project 3
|
|
The promise of genomic medicine is the personalization of therapeutics based on one’s genetic makeup, and the results of this study will be to dramatically increase the catalog of genetic predictors of drug response and to create a library of potential repurposing for nearly all medications.
We have used BioVU, our repository that links DNA samples from almost 220,000 subjects to deidentified electronic health records, to develop the new analytical methodology of the Phenome-Wide Association Study (PheWAS). We have enhanced the PheWAS methods and have shown that we can identify new targets (for efficacy and for ADRs) for existing drugs. We will apply enhanced PheWAS to probe pleiotropy in known and newly-discovered genomic predictors of ADRs, including those targeted in Projects 1 and 2, and to identify new indications and on-target side effects of existing drugs.
Joshua Denny
Principal Investigator
|
Wei-Qi Wei
Co-Investigator
|
Cosmin Bejan
Co-Investigator
|
Robert Carroll
Co-Investigator
|
Core
Analytical and Administrative Support Core
The Core is charged with administrative management and with providing centralized data management and analysis functions, and bioinformatics and biostatistics support to the program in order to facilitate progress toward achieving the Center’s short and long term goals.
Dan Roden
Co-Principal Investigator
|
Joshua Denny
Co-Principal Investigator
|
Elizabeth Phillipps
Co-Principal Investigator
|
Nancy Cox
Co-Investigator
|
Todd Edwards
Co-Investigator
|
Simon Mallal
Co-Investigator
|
Digna Velez-Edwards
Co-Investigator
|
Yaomin Xu
Co-Investigator
|
