RESEARCH TRIANLE PARK, N.C.—A new term has entered the growing 'omics lexicon: It's called pharmaco-metabonomics, a method described in a recent study by researchers at the University of North Carolina, Chapel Hill and the Hamner-UNC Institute for Drug Safety Sciences (IDSS) that can be used to identify individuals who are at risk for developing drug-induced liver injury (DILI). According to the researchers, achieving this ability would represent a major advance in personalized medicine and alleviate the drug safety concerns that often create bottlenecks in drug discovery and delivery.
Noting that it is not currently possible to identify which patients are susceptible to DILI, and that DILI is the major adverse event leading to regulatory actions on drugs—including denial of marketing approval, restrictions with respect to clinical indications and withdrawal from the marketplace—the researchers suggest that urine tests might identify the susceptible individuals before the liver is injured. Their study, published in the July issue of Clinical Pharmacology and Therapeutics, builds on a 2006 study led by T. Andrew Clayton at Imperial College London that demonstrated that the pattern of endogenous metabolites in urine could predict susceptibility to acetaminophen-induced liver injury in rats.
In their study, Clayton, et al., proposed the term "pharmaco-metabonomics," which they defined as "the prediction of the outcome of a drug or xenobiotic intervention in an individual based on a mathematical model of pre-intervention metabolite signatures." Additional support for the pharmaco-metabonomic approach was reported in another rodent study, in which pre-dose urinary metabolomes could be used to predict which rats would develop diabetes after treatment with streptozotocin.
However, the Hamner-UNC team noted that the ability of the pharmaco-metabonomic approach to predict susceptibility to the toxic effects of drugs has not been tested in humans. Humans, they pointed out, differ from laboratory rodents in many ways, including greater genetic heterogeneity and wider variation in age, diet and lifestyle, and these have been shown to result in marked variation in the urinary metabolome. Thus, the researchers set out to demonstrate that urine tests may identify the susceptible individuals based on samples obtained from healthy adult volunteers who were given acetaminophen.
"Our goal is to find biomarkers in urine that can be universally used and not reliant on such pristine conditions," says an investigator in the study, Dr. Paul Watkins, a professor of medicine at UNC and the director of IDSS. "Genetics is only part of the story. As people get older, take different drugs, develop certain conditions, etc., environmental factors also become important. Our hypothesis was that the human urinary metabolome contains sufficient information to discriminate the responder phenotype from the nonresponder phenotype."
Watkins' team administered 4 grams of acetaminophen (two 500-mg tablets every six hours) daily for seven days to 71 healthy men and women aged 18 to 55 years. Urine samples obtained daily before and during dosing were analyzed by high-resolution magnetic resonance spectroscopy. After about one week of dosing, blood testing revealed that some of the volunteers developed mild and reversible liver injury. The researchers then found that within one or two days of treatment, specific patterns of urine metabolites could predict who would subsequently develop the mild liver injury.
"This distinction was not apparent when the urine was examined before the start of the acetaminophen treatment, but became evident once treatment began," the researchers concluded. "Because the pretreatment metabolome was not predictive, and the acetaminophen metabolite contribution was relatively minor, the ability of the post-treatment urinary metabolome to predict susceptibility was due chiefly to changes in the endogenous metabolome. These changes presumably reflect phenotypic changes evoked by acetaminophen treatment."
The ability to predict if a patient will develop these reactions has long eluded the research community, and the metabolomic approach "has the potential to make many drugs much safer," Watkins says.
"My major interest is the insight it gives us into the mechanisms involved in liver toxicity," he says. "In certain types of treatment, you could greatly shorten the length of treatment, or predict whether or not a person will get into trouble, without a lot of extra doctor visits and blood tests. I think this will move us toward personalized medicine."
Having served as a hepatologist, physician and researcher for more than 30 years, Watkins' research spans preclinical models to patients in clinical trials and utilizes current genomics technologies, including genetics, transcriptomics and metabolomics. Speaking from his experience, Watkins acknowledges shortcomings and waning interest in the use of metabolomics to predict who will develop DILI.
"Genetic testing has been somewhat disappointing, and there has been a sort of fall-off in enthusiasm for metabolomics," he says. "Many large pharma companies are getting out of it, because they were hoping there would be quick advances. As with many technological advances in drug development, there is a bit of overselling and hype, and the whole field falls back. There is a great enthusiasm to hop on, and then we get overwhelmed by our inability to handle the information that is created. But I believe the application of metabolomics to clinical trials where there is a potential signal of toxicity is a fertile area. It will be interesting to see how far it goes. We're hoping out paper will help bring back some of the interest into the use of metabolomics in drug discovery."
Watkins and his colleagues are now conducting a much larger study of 100 patients in China who are being actively treated for tuberculosis (TB)—"the poster child for drug-induced toxicity," Watkins says.
"The exciting thing about the study we just did was that it was a healthy volunteer study that looked at a trivial amount of liver injury. Now, we're conducting a much larger study of people being actively treated for TB," Watkins says. "About 15 percent of these people will have to stop treatment because of liver toxicity. In addition, these people have very little concomitant diseases, and are predominantly healthy people otherwise. They are also ethically similar, with 90 percent of them being Han Chinese."
The Watkins, et al., study, "Use of Pharmaco-Metabonomics for Early Prediction of Acetaminophen-Induced Hepatotoxicity in Humans," was funded in part by grants from the National Institutes of Health. Watkins' co-authors were J.H. Winnike, Z. Li, F.A. Wright, J.M. Macdonald and T.M. O'Connell.
Noting that it is not currently possible to identify which patients are susceptible to DILI, and that DILI is the major adverse event leading to regulatory actions on drugs—including denial of marketing approval, restrictions with respect to clinical indications and withdrawal from the marketplace—the researchers suggest that urine tests might identify the susceptible individuals before the liver is injured. Their study, published in the July issue of Clinical Pharmacology and Therapeutics, builds on a 2006 study led by T. Andrew Clayton at Imperial College London that demonstrated that the pattern of endogenous metabolites in urine could predict susceptibility to acetaminophen-induced liver injury in rats.
In their study, Clayton, et al., proposed the term "pharmaco-metabonomics," which they defined as "the prediction of the outcome of a drug or xenobiotic intervention in an individual based on a mathematical model of pre-intervention metabolite signatures." Additional support for the pharmaco-metabonomic approach was reported in another rodent study, in which pre-dose urinary metabolomes could be used to predict which rats would develop diabetes after treatment with streptozotocin.
However, the Hamner-UNC team noted that the ability of the pharmaco-metabonomic approach to predict susceptibility to the toxic effects of drugs has not been tested in humans. Humans, they pointed out, differ from laboratory rodents in many ways, including greater genetic heterogeneity and wider variation in age, diet and lifestyle, and these have been shown to result in marked variation in the urinary metabolome. Thus, the researchers set out to demonstrate that urine tests may identify the susceptible individuals based on samples obtained from healthy adult volunteers who were given acetaminophen.
"Our goal is to find biomarkers in urine that can be universally used and not reliant on such pristine conditions," says an investigator in the study, Dr. Paul Watkins, a professor of medicine at UNC and the director of IDSS. "Genetics is only part of the story. As people get older, take different drugs, develop certain conditions, etc., environmental factors also become important. Our hypothesis was that the human urinary metabolome contains sufficient information to discriminate the responder phenotype from the nonresponder phenotype."
Watkins' team administered 4 grams of acetaminophen (two 500-mg tablets every six hours) daily for seven days to 71 healthy men and women aged 18 to 55 years. Urine samples obtained daily before and during dosing were analyzed by high-resolution magnetic resonance spectroscopy. After about one week of dosing, blood testing revealed that some of the volunteers developed mild and reversible liver injury. The researchers then found that within one or two days of treatment, specific patterns of urine metabolites could predict who would subsequently develop the mild liver injury.
"This distinction was not apparent when the urine was examined before the start of the acetaminophen treatment, but became evident once treatment began," the researchers concluded. "Because the pretreatment metabolome was not predictive, and the acetaminophen metabolite contribution was relatively minor, the ability of the post-treatment urinary metabolome to predict susceptibility was due chiefly to changes in the endogenous metabolome. These changes presumably reflect phenotypic changes evoked by acetaminophen treatment."
The ability to predict if a patient will develop these reactions has long eluded the research community, and the metabolomic approach "has the potential to make many drugs much safer," Watkins says.
"My major interest is the insight it gives us into the mechanisms involved in liver toxicity," he says. "In certain types of treatment, you could greatly shorten the length of treatment, or predict whether or not a person will get into trouble, without a lot of extra doctor visits and blood tests. I think this will move us toward personalized medicine."
Having served as a hepatologist, physician and researcher for more than 30 years, Watkins' research spans preclinical models to patients in clinical trials and utilizes current genomics technologies, including genetics, transcriptomics and metabolomics. Speaking from his experience, Watkins acknowledges shortcomings and waning interest in the use of metabolomics to predict who will develop DILI.
"Genetic testing has been somewhat disappointing, and there has been a sort of fall-off in enthusiasm for metabolomics," he says. "Many large pharma companies are getting out of it, because they were hoping there would be quick advances. As with many technological advances in drug development, there is a bit of overselling and hype, and the whole field falls back. There is a great enthusiasm to hop on, and then we get overwhelmed by our inability to handle the information that is created. But I believe the application of metabolomics to clinical trials where there is a potential signal of toxicity is a fertile area. It will be interesting to see how far it goes. We're hoping out paper will help bring back some of the interest into the use of metabolomics in drug discovery."
Watkins and his colleagues are now conducting a much larger study of 100 patients in China who are being actively treated for tuberculosis (TB)—"the poster child for drug-induced toxicity," Watkins says.
"The exciting thing about the study we just did was that it was a healthy volunteer study that looked at a trivial amount of liver injury. Now, we're conducting a much larger study of people being actively treated for TB," Watkins says. "About 15 percent of these people will have to stop treatment because of liver toxicity. In addition, these people have very little concomitant diseases, and are predominantly healthy people otherwise. They are also ethically similar, with 90 percent of them being Han Chinese."
The Watkins, et al., study, "Use of Pharmaco-Metabonomics for Early Prediction of Acetaminophen-Induced Hepatotoxicity in Humans," was funded in part by grants from the National Institutes of Health. Watkins' co-authors were J.H. Winnike, Z. Li, F.A. Wright, J.M. Macdonald and T.M. O'Connell.
