By Malorye Branca
May 7, 2002 | A series of study results and a new stance by the Food and Drug Administration may provide badly needed fuel for pharmacogenomics.
The study of why individuals respond differently to drugs — pharmacogenomics — is one of the most potentially exciting applications of the Human Genome Project. In the not-too-distant future, doctors will be able to use patients’ unique genetic profiles to guide treatment decisions, and medicines will be tailored to genetic makeup.
But despite widespread anticipation of the promise of pharmacogenomics, surprisingly few findings to date have supported this view. “There was a big wave of investment in this field awhile ago, and people found it was harder to get results than they thought,” says Dennis Henner of MPM Capital.
As a result, the field is losing steam. Some of the smaller pioneering pharmacogenomics companies are anxious to reverse that trend.
Researchers at Genaissance Pharmaceuticals, based in New Haven, Conn., recently reported “statistically significant” relationships between genetic markers and patient responses to cholesterol-lowering drugs, called statins. If confirmed, these relationships could lead to pharmacogenetic tests to determine which drug will work best in a particular patient.
“Number one, it [the study] validates what we are doing, and that we can identify these variations that predict response to therapeutic agents,” says Carol Reed, executive director of medical affairs at Genaissance. “Secondly, the associations are significant.”
“Genaissance does not sell any of these drugs, and they are the most successful category of drugs in pharmaceutical history,” says Peter Tollman of the Boston Consulting Group. “If these associations are significant enough, this could create a very interesting dynamic. We will see a whole new set of competitors going after the Big Pharma drugs, and a need to respond by Big Pharma.”
These results come from an initial analysis of the company’s STRENGTH I (Statin Response Examined by Genetic HAP Markers) study, which involved simvastatin (Merck & Co.’s Zocor); atorvastatin (Pfizer’s Lipitor); and pravastatin (Bristol-Myers Squibb’s Pravachol). The study originally also involved cerivastatin (Bayer’s Baycol), but Genaissance discontinued the Baycol arm of the study when that drug was voluntarily withdrawn from the market last summer after some patients taking it developed potentially fatal side effects. A second study, STRENGTH II, is also nearing completion. It includes the three drugs left in STRENGTH I, as well as lovastatin (Merck & Co’s Mevacor).
The statins represent a huge market. According to IMS Health, Lipitor and Zocor were the second and third top-selling drugs in the world in 2000. IMS recorded almost $16 billion in worldwide sales for cholesterol- and triglyceride-lowering drugs that year.
Statin therapy aims to reduce blood levels of LDL (bad) cholesterol and to improve the ratio of HDL (good) cholesterol to bad. Genaissance has been looking for links between genetic markers and changes in the levels of cholesterol and triglycerides in patients taking these drugs.
Genaissance uses sequencing to find genetic variations called single nucleotide polymorphisms (SNPs) in human DNA. This requires many samples, from the right mix of people, to get a good representation of a target population. The company applies proprietary algorithms to find sets of SNPs, called haplotypes, which are inherited. The goal is to find haplotypes that are “red flags” of disease susceptibility or drug response.
Studying haplotypes, rather than individual SNPs, reduces the number of patients needed for these types of trials because combinations of markers provide a stronger signature than single markers. Genaissance has compiled a database of more than 5,000 such haplotypes, which they call HAP markers. The company has also identified 100 genes that may be linked to patient response to statins.
If careful review and further testing confirm the study results, it could lead to products for Genaissance and to better treatment for patients. Tollman, however, advises caution. “The study is small and has other limitations,” he says.
In STRENGTH I about 150 patients received each particular type of statin. Overall, more than 500 patients were enrolled. Analysts and researchers are waiting for more details before deciding the study’s importance.
“We did the press release to let the clinical world and investor world know that the study is complete, and that we are analyzing the data, and that there is interesting data there,” says Reed, adding that the company plans to publish and present findings from the study throughout the year. (The first presentation was planned for late April, after this issue went to press.) “The study was big enough for us to derive statistically relevant correlations. We expect to derive commercial value from these associations.”
FDA’s Positive Position
Perhaps the biggest hurdle to pharmacogenomics has been the perception that the FDA would resist the introduction of these tests. “The FDA never actually said this, but many people believed that the FDA would never allow these types of tests to be approved,” says Allen Roses, senior vice president of genetic research at GlaxoSmithKline (GSK), which is a leading player in this field.
But a timely new announcement from the FDA may clear up that perception. “Now the FDA has come out and said, quite dramatically, ‘Yes, we will look at this,’” says Roses, referring to a recent article in The Pharmacogenomics Journal by Lawrence J. Lesko and Janet Woodcock, two FDA officials. Lesko and Woodcock reported that the FDA has received more than 15 recent applications involving pharmacogenomic tests, and that it is taking steps “to assure that, as a regulatory agency, FDA is prepared to deal with the future influx of [pharmacogenomic data].”
Genaissance’s Reed agrees with Roses about the impact of the article. “This could really change things,” she says. With the FDA’s go-ahead, more companies may enter the field, and smaller companies may have less difficulty convincing partners and investors that they can succeed.
Big Pharma’s Low Profile
With the exception of GSK, leading pharmaceutical companies have been slow in carrying out or funding pharmacogenetic studies of drugs that are already on the market. Last year Jurgen Drews, president of Genaissance and a former president of global research at Roche Pharmaceuticals, was quoted in the Wall Street Journal as saying that pharmaceutical companies were afraid of discovering “the limitations of their drugs.”
If pharmacogenetic tests can cause doctors to change their prescribing practices, it could erode some drugs’ market shares. But it is hard to tell what is keeping Big Pharma away. “I don’t think big companies are trying to delay the development of these tests,” says Henner. “The technology just hasn’t yet moved to a point where they see the value of it.”
Large pharmas and genomics firms have shown greater interest in applying pharmacogenomics to early-stage drug candidates. By finding out early on that a potential drug has serious side effects in some patients, or by identifying therapies that only work well in a subgroup of people, companies can save on drug development costs.
Later, the companies may market the therapies and the tests used with them together. Genaissance aims to do such studies as well, but because of the length of the drug discovery and development process, they may take even longer than studies of established drugs.
In the meantime, investors may have already grown weary of waiting for results. “Before pharmacogenomics can come into widespread use, these companies will have to show that the data they provide will make a difference to patients, and that doctors can actually use it,” say Henner. “Meanwhile, there are a lot of other biotech products that are going to happen much sooner, so that’s where the money is going.”
Over the last few months, there has been a small, but noticeable trickle of results from pharmacogenetic studies, some in major publications (see table). The next step must be to turn these results into actual products.
Early pharmacogenetic tests “looked at extreme responses,” says Tony Frudakis of pharmacogenomics startup DNAPrint Genomics of Sarasota, Fla. “They have looked at too few genes and too few sites in those genes to be useful for more than a very few patients.” To become clinically accepted, tests need to work across a wide range of patients. “We are close to this, and Genaissance is close,” he says.
DNAPrint recently announced finding genetic markers linked to patient response to paclitaxel (Bristol-Myers Squibb’s Taxol). The company uses an approach similar to that of Genaissance. It hopes to develop a test that will flag those patients who will not have a good response to the drug in order to redirect them to other therapies.
The sequencing of the human genome has led to the discovery of millions of SNPs, increasing the power of the studies by giving researchers more data from which to work. “The answer now relies on how good your informatics are,” says Frudakis. “We’re doing complex genetics now, and there are no algorithms out there to do this. You have to develop them yourself.”
Even with powerful algorithms, clinical trials cost money. “The big companies should be funding these kinds of studies and making better tools out of these drugs,” says Frudakis. “But addressing the current drugs falls on the shoulders of companies like us and Genaissance.” Still, Frudakis remains optimistic. “We are writing an NIH grant right now,” he says. “If we can get funding, we might have a product on the market in two to three years.”
However, GSK’s Roses is skeptical about whether smaller companies can pay for effective pharmacogenomic studies. “We have made a sizable investment in informatics for this, and we are already doing many different clinical studies all the time,” says Roses.
GSK, Roses says, embraced the field early. “We could see that what is good for the patient is good for us,” he says. Other large pharmaceutical companies may have been slower to climb onboard, he says, because “big pharmas are not mainly R&D organizations; they are sales and marketing organizations, so the people at the top may not fully understand the potential here.”
In the long run, Roses says, pharmacogenomics will make economic sense as well. If patients and physicians are more satisfied with the drugs, they’ll be more likely to use them.
According to the FDA’s Lesko and Woodcock, “The central issue is not whether [pharmacogenomics]-guided drug prescriptions will happen, but when and how.”
For the small firms trying to become big players in this field, success may come down to having the money to stay in the game. “Genaissance and the others will ultimately pay off, as long as they are doing the right kind of studies,” says Henner. “I just think it will take longer to see it.”