By Meredith W. Salisbury
Thousands of researchers have toiled for years (and show no signs of slowing down) to prove the promise of genomics and its related sciences, including proteomics and bioinformatics — but of course the hallmark of success will ultimately be reached when genomics makes its way into mainstream medicine.
And it’s getting there, albeit slowly. In the clinic, says Gregory Tsongalis, director of molecular pathology at Dartmouth College, “the genomics part of it’s already here. That’s just exploding.” Pharmaceutical companies and diagnostic labs have recognized the value of tracking down SNPs to identify responders or non-responders to a drug, for example, and are already taking advantage of that.
The real excitement, just like on the research side, lies in the systems biology aspect of personalized medicine. “The systems approach is absolutely essential,” says Harel Weinstein, director of the Institute for Computational Biomedicine at Weill Medical College of Cornell University, adding that “integration is a sine qua non.”
That integration needs to include as many components as it does in the basic research world, says Thomas Metcalfe, head of the biomarker program at Roche. Those he rattles off — genomics, proteomics, metabonomics, pharmacogenetics, and pharmacodiagnostics — could be just a part of the greater scheme if all goes as hoped.
But the big dreams of these and other scientists face some stiff competition from the challenges facing the advance of genomics into medicine. For starters, says Jeff Trent, president of the Translational Genomics Research Institute, “Surprisingly, I think that there’s still quite a bit of controversy around groups that are trying to implement [personalized medicine], rather than just trying to stick to the status quo.” With clinical utility still largely unproven, there are plenty of people in the medical community who remain unconvinced that systems biology will live up to its hype.
It’s still early days for systems-biology-style work pertinent to the medical field, but any number of places are trying to make their mark. One of the most prominent university initiatives is at Duke, which within its Institute for Genome Sciences and Policy recently started a division focused purely on genomic medicine. In January of this year, says Hunt Willard, director of the genome institute at the school, Duke launched “a personalized health pilot project working with Duke employees who are insured by the Duke Health System.” After gathering personal health information and performing formal risk assessments, patients were stratified into various groups based on health risks. “About 3,000 to 4,000 patients [assigned to] to severely high and high risk groups … are receiving fundamentally different healthcare than everyone else,” Willard says. That is most evident in the amount of supervision they get, he adds: patients are assigned a health coach who might contact them as often as every day to make sure they’re taking medications, exercising, and so forth. Duke’s early efforts are targeted at hypertension, diabetes, tobacco use, and obesity.
At the Translational Genomics Research Institute, the effort is taking shape in the form of merging genomics with clinical trial design to focus on patients who are running out of therapy options for cancer. Using microarray analysis for whole-genome studies, researchers at TGen and partner institutions such as the Mayo Clinic and the Moffitt Cancer Center are trying to understand which people respond best to which drugs, as well as which pathways are most modified in any particular form of cancer, Trent says.
Michael Liebman at the Windber Research Institute relies on ties with the Walter Reed Army Medical Center to track down highly qualified samples and study them with genomics and proteomics techniques — also tying in lifestyle aspects such as smoking or weight — to help understand breast cancer.
Meanwhile, at St. Jude’s Children’s Research Hospital in Memphis, scientists are using functional and association studies to link polymorphisms to phenotypes, pharmacodynamics, and pharmacokinetics to get a bead on patient response to various therapy options.
On Tap for Tomorrow
There are some major differences of opinion in the field on how rapidly advances in genomics will take hold in healthcare. Some, like Liebman, contend that “we don’t yet completely understand the fundamentals, so the ability to convert that into immediate returns … is going to be limited,” he says.
Others are more hopeful. Peter Ray at the Hospital for Sick Children in Toronto predicts that the $1,000 genome sequence, which could replace piecemeal genetic tests, will be a possibility “in the not too distant future.” He adds that the community will continue to see more and more genetic screening in the next five years, while population screening, a key part of systems medicine, may be mainstream within the next decade.
Of course, all of that depends heavily on continued financial support for the field. Harel Weinstein expresses concern that the hype surrounding this field may cause investors to sour and take their money elsewhere.
Ray, however, notes that even today, genomics is contributing to “one of the most rapidly expanding areas of medicine now. [It’s] making just huge advances.”
Plenty of Hurdles
No one’s saying it’s going to be a smooth ride from benchtop to bedside, but most people agree that it will eventually happen. When we asked scientists what’s standing in the way, they had no shortage of things to talk about. Here are some of the challenges they identified in the slow advance of pharmacogenomics and personalized medicine:
Ethical and legal sticking points
Handling genetic issues with insurance companies is a looming problem, experts say. For one thing, notes Peter Ray, a senior associate scientist at the Hospital for Sick Children in Toronto, “We need some legal way to prevent insurance companies from demanding [test results].” If one company starts to require genetic testing and then refuses to cover high-risk people, Ray adds, all insurance companies will be forced to follow suit to stay in business.
“This is really turning the Queen Mary around in terms of the healthcare system and healthcare approach in this country,” says Hunt Willard, director of the Institute for Genome Sciences and Policy at Duke University. He notes that as more and more genetic tests hit the market, there will be an increasing need for a comprehensive analysis tool that would make sense of all the test readouts in relation to each other. Adds William Evans, scientific director at St. Jude’s Children’s Research Hospital, “Until we make it easier for [physicians] to interpret those readouts, they’re not likely to request” genetic testing.
“A lot of physicians that are in practice right now have never had molecular biology or genetics,” says Gregory Tsongalis, director of molecular pathology at Dartmouth College. “So that’s a huge effort to bring people up to speed.” And education is made more difficult because of how fast the field is advancing, says Ray at the Hospital for Sick Children. “You can’t teach … this in medical school because by the time they get to practice it’s out of date.”
Many experts agree that despite the promise of genomics and its descendant disciplines, hard-and-fast data proving clinical utility of this science is scarce. “We have all the sequences and the genes … [but] at the end of the day, is it going to matter?” asks Tsongalis. “We really need some clinical outcome studies.”
Technical hurdles abound as well, points out Mary Relling, chair of pharmaceutical sciences at St. Jude’s Children’s Research Hospital. “There are very few labs that have taken the trouble to make sure their genotypes can withstand scrutiny,” she says — especially for issues like avoiding false negatives or positives. “There aren’t really excellent standards in terms of clinical genetic tests.”
“From my experience in both academic and industry [settings],” says Michael Liebman, CSO of the Windber Research Institute, the primary challenge is “gaining access to highly controlled clinical samples.” Liebman says Windber, which is investigating breast cancer, has an advantage in this arena because of the highly qualified samples it receives through its relationship with Walter Reed Army Medical Center.
First up: diagnostics
Of the experts GT polled, most agree that diagnostics will be the first point of impact for genomics in mainstream medicine. People like Thomas Metcalfe at Roche note that genomic-based testing, such as his company’s cytochrome P450 AmpliChip currently available in Europe, is already a way of life. Peter Ray, a senior associate scientist at the Hospital for Sick Children in Toronto who heads up the molecular diagnostics lab there, says that in recent years most of the physicians who order genetic testing for patients are not geneticists — a marked contrast to how things were not too long ago, he says. “At the moment about 70 percent of our samples come from non-geneticists,” he says, adding that based on trends he expects that number to rise to 90 percent in five years.
There’s also a growing sort of grass-roots support for such diagnostics, says Kevin McKernan at Agencourt Bioscience. McKernan’s company receives a steady stream of biopsies and paraffin tissue samples, he says, from people who want to be tested for the Egfr mutation linked to response to the cancer drug Iressa. There’s no official diagnostic, and Agencourt, which is not yet CLIA approved, has to be careful to remind people that its services are purely research-side rather than diagnostic, McKernan says — but despite that, demand for the assay is growing.
That cancer will be the first disease area where genomics becomes truly prominent is not doubted by experts. “The big push is going to come using oncology,” says Gregory Tsongalis at Dartmouth. “Only about 20 to 30 percent of patients respond to cancer therapeutics — there’s huge room for improvement.”
What can I do?
Genome Technology asked people who are involved in or closely watching the approach of genomic sciences to clinical medicine whether there’s anything scientists back in the basic research camp can do to help push things along. The answer was resoundingly positive. Listed below are some of the opportunities these experts see for researchers to pitch in.
“Be aware of what kind of information is required for a company like LabCorp to bring out a test,” says Paul Billings, vice president of biotechnology and healthcare strategy at diagnostics giant LabCorp. “Most of the time, scientists focus on technical issues — [but] we need to be able to be paid for those tests.” Diagnostic tests have to fit into the standard healthcare cost coding system that allows companies such as LabCorp to be reimbursed for their services, and the argument made to payors for clinical utility has to be airtight.
“It’s the marriage between the informatics and the data sets,” says Hunt Willard, director of the Institute for Genome Sciences and Policy at Duke University, noting that the community needs solid statistical approaches to help researchers downstream work more effectively with the rapidly growing volume and breadth of data collected for pharmacogenomic applications.
Thomas Metcalfe, head of Roche’s biomarker program, says that scientists can really help in developing standards across tool platforms. “As you move out of the research setting we have to be able to use these sorts of tools or results which are produced by different investigators,” he says. “They really have to be made more comparable. [That] has to happen for the field to move forward.”
Flood the market
“The best thing everyone can be doing is more sequencing,” says Kevin McKernan, a CSO at sequencing company Agencourt Bioscience who admittedly has a bias in the field. The more utility that scientists can illustrate, McKernan figures, “the more hospitals and HMOs will be drawn to it. Until there’s more sequencing done, it’s not going to be a compelling enough argument for that transition to happen.”
Phone a physician
A good way to get involved is to “partner with the physicians at their centers who are developing clinical trials,” says Jeff Trent, president of the Translational Genomics Research Institute, “and work to get information out for validation from multiple centers.”