• Home
• Research & development
• Marathons magazine
• Issue 3
• Toxicogenomics: out from under glass

Toxicogenomics: out from under glass

Now and again a new technology will emerge in a blush of excitement, promising to transform drug development overnight, and yet for all the acclaim it must still demonstrate everyday utility.

“Technology under glass,” Patrick Wier calls it. For several years GSK scientists have been taking just such a specimen out of the display case and putting it to the demanding test of vetting potential medicines for side effects.

The technology under trial is toxicogenomics, or the investigation of changes in gene expression for signals that a drug may do harm.1

Wier, a scientist in Safety Assessment, explains: “We are putting toxicogenomics into a practical context. Now it’s part of the way we work. We’re actually applying it to our pipeline. It is producing information of value in assessing the total weight of the evidence on safety as we do drug development.”

Specifically, GSK is using toxicogenomics to help evaluate the risk that a drug may injure the liver—needless to say, a crucial concern in drug development.

The liver is the great clearing house of the body. It houses the metabolic machinery that breaks down chemicals in preparation for their disposal, not only medicines but also industrial toxins, medicinal herbs, alcohol, illicit drugs, and substances synthesized within the body. The trouble is that, while safeguarding other organs, the liver itself may falter from these toxic assaults. More than 900 chemicals including medicines have been reported to cause liver injury.2

A search for signals of hepatotoxicity is thus a crucial element of any drug-development effort. The initial safety studies done in animals will often yield findings that terminate an experimental drug before it ever reaches the clinic. Once studies in people do get under way, a constant watch continues for any signal of risk, such as elevated blood levels of enzymes released by injured liver cells. Typically, this gauntlet of testing either keeps hepatotoxic drugs from the market or characterizes their risk so that they can be used judiciously.

But no preclinical safety test is a perfect predictor. With one drug, a signal may emerge quickly but mildly and then grow in a manageable way with higher dosing. With another, there may be no signal, only a rare but sudden crisis of liver failure, and then only after a drug has passed through animal studies and early clinical trials with no hint of trouble. Age, sex, or race may play a role in some cases, and across any population, risk may be idiosyncratic owing to individual constitution.

Consequently, hepatotoxicity has led to more drug withdrawals over the past half century than any other side effect.3 And surely there has been another toll: the jettisoning of safe, effective, and needed drugs during development because benign or mild events were mis-read as danger signs.

Toxicogenomics, then, is an alluring tool. It is simple enough in concept: Determine which genes in the liver are expressed in association with liver injury after a drug has been administered.4 Changes in gene expression may amount to new signals, perhaps especially early signals.

Here simplicity stops, however. Reducing concept to routine practice demands years of fastidious analysis and adaptations along the way. The gains are incremental, and the demand for sustained research commitment substantial.

The toxicogenomics project at GSK started just over five years ago. The first step was to assess gene expression in response to diverse chemicals already known to injure the liver. Control chemicals were included as well, those known not to injure the liver. The testing was done in short-term studies in rats, the usual first step in assessing safety prior to clinical trials.
 
“We want to be able to predict safety issues earlier,” Wier continues. “This is the earliest study in animals, and toxicogenomics is a cutting-edge technology. We said, ‘Let’s give this a try.’ But we weren’t about to do it piecemeal. We wanted it to be done uniformly across the organization. It had to be practical, efficient, and easily integrated. We resisted the tendency to just do something to show we could do it.”

Changes in expression coincident with chemical exposure were evaluated among more than 5,000 genes. Genomic changes were evaluated by tracing their association with signs of liver injury which had been identified by more conventional methods, such as measurement of enzyme levels in the blood or microscopic examination of tissue damage. One complication was that gene expression can vary for reasons apart from chemical exposure and liver injury, most notably diet or feeding behaviour. So the effects of diet on gene expression were studied independently, to distinguish them from chemical effects.

In this painstaking manner, GSK scientists have so far derived 16 “thematic panels” comprising 137 genes. Each panel consists of a small set of genes whose expression in the liver has been found to change in response to chemical exposure. Some panels relate to specific types of liver injury, such as fibrosis, cell death, or an obstruction to bile flow within the liver. Others relate to the underlying modes of liver injury, such as changes in drug-metabolizing enzymes or the formation of toxic drug metabolites. Interpretive guidelines have been developed to ensure rigor and common practice in interpreting the data.
 
An early study of a cancer drug illustrates one way in which these data are being brought to bear on everyday questions of drug development. Studies in rats given the drug showed markedly elevated blood levels of bilirubin, a breakdown product of red blood cells. Generally, such elevation results from obstructed bile flow, a condition known as cholestasis; bilirubin, a component of bile, backs up and then spills into the blood. In this instance, however, the toxicogenomic data did not suggest cholestasis. Additional laboratory work identified the mechanism for the bilirubin increase to be inhibition of a transporter protein in liver cells which facilitates removal of bilirubin from blood.

Toxicogenomics thus eliminated concerns about cholestasis: The drug may have an effect on liver function, but it has not caused liver injury. This kind of knowledge can help to focus the chemistry of refashioning the molecule—to keep scientists on the right path to developing a safe medicine.

What more can be said of toxicogenomics at this date? It clearly provides insights into modes of liver injury which are not evident by traditional methods. As for predicting injury, it introduces a new line of evidence into the totality of the data that researchers consider while deciding whether to progress a drug to the next, longer-term stages of testing in rats. It may tip a close call. Once the province of 70 scientists in GSK who generated the early data, toxicogenomics is now broadly embedded in early drug development across all therapeutic areas. It has contributed to evaluations of more than 170 drug candidates.

To be sure, toxicogenomics in no way has replaced the mainstays of early safety assessment, those being histopathology (microscopic examination of tissues) and clinical pathology (measurement of biochemical and cellular changes in blood). Learning whether toxicogenomics can reliably predict safety issues that escape these longstanding methods awaits follow-up studies. Only now are data coming in from longer-term rat studies, and data from humans are still further off.

Says Wier, “It will take more experience for toxicogenomics to fully establish its place relative to the tools that have been used in toxicology for decades, more time to accrue the numbers we need to say whether the particular genes we’ve selected are the ones most pertinent to liver toxicity.

“But we’ve now established toxicogenomics as a core competency in early drug development that we can adapt to measure other gene-expression changes not only in the liver but also other organs as the state of the science evolves.

“It all had to start with the commitment to taking it out from under the glass.”

###

1 Gene expression is the cellular reading of the genetic code to produce the specific proteins that determine a cell’s functions. A gene that is expressed is often said to be “turned on” or upregulated; a gene that is not expressed, “turned off” or downregulated.  A gene may be expressed at varying levels; that is, it may be activated so as to produce varying amounts of the encoded protein.

2 Friedman SE, Grendell JH, McQuaid KR. (2003). Current diagnosis & treatment in gastroenterology. New York: Lang Medical Books/McGraw-Hill, p664-679.

3 U.S. Food and Drug Administration. Drug-Induced Liver Injury: Premarketing Clinical Evaluation (Draft guidance for industry). October 2007.

4 Gene expression is detected and measured by a laboratory technique called polymerase chain reaction, or PCR, which has revolutionized molecular biology over the past quarter century. PCR amplifies DNA derived from an expressed gene to enable detection and then measurement of that expression. Kary Mullis, who invented PCR, won the Nobel Prize for Chemistry in 1993. He has written that the technique “requires no more than a test tube, a few simple reagents, and a source of heat.” It is now highly automated, though, and it is put to many uses, including molecular finger printing and genetic studies of extinct species, as well as drug development. GSK has innovated its own use of PCR and called it HepatoTaq© after the name of an enzyme commonly used in PCR, Taq polymerase.

Back to top