By Turna Ray
This article was originally posted on Oct. 29.
For comparative effectiveness research and genomically guided personalized medicine to converge, the US needs to invest in its research infrastructure to conduct studies in the real-world setting and not just in academic medical centers, a top official from the US Food and Drug Administration said at a conference in Washington, DC, this week.
"One of the principles of comparative effectiveness is generalizeable healthcare. And to me, it needs to be done in the community where healthcare is delivered. Not in academic medical centers strictly," Janet Woodcock, director of the US Food and Drug Administration's Center for Drug Evaluation and Research, said at a meeting on comparative effectiveness hosted by the Personalized Medicine Coalition.
However, "we don't have the kind of research infrastructure that links to communities," to bring personalized medicine principles into comparative effectiveness research, Woodcock said. "Patients want to be treated by their doctors. They don't want to be sent somewhere to have some comparative effectiveness or even personalized medicine studies done."
At the same meeting where Woodcock made her remarks, the PMC and the Lewin Group released a report recommending areas of alignment for CER and personalized medicine efforts — such as product labeling, payment policies, and utilization management — or risk the acceptance and advancement of "inadequate or misleading" efficacy and safety data about medical interventions [see PGx Reporter 10-28-2009].
As Woodcock made her suggestion for increased focus on community-based research in CER and personalized medicine, NIH Director Francis Collins warned personalized medicine and comparative effectiveness research were on a potential collision course if heterogeneity of treatment populations is not factored into CER. Collins made his comments at a separate meeting hosted by the American Association for the Advancement of Science this week [see PGx Reporter 10-28-2009].
In Woodcock's view, while CER should be generalizeable and personalized medicine characterizes the differences in specific sub-populations, the two disciplines can be mutually inclusive concepts if more research is conducted in local communities.
While "comparative effectiveness really needs to compare treatments in these optimized populations to the extent we can define them, … personalized medicine's job is to define these optimal populations," Woodcock said, adding that the research infrastructure necessary for aligning these two concepts does not exist in the US right now.
"We need to translate and investigate personalized medicine interventions and we need to perform comparative effectiveness evaluations on a pretty large scale," she said. "There is a huge amount of research that has to be done that is certainly going to dwarf the amount of money that is currently available just for comparative effectiveness research.
[ pagebreak ]
"So the number of questions that need to be answered about healthcare and healthcare delivery, as well as what treatment works best and for whom — the personalized medicine piece — far exceeds the current research capacity in the US."
One thing that unites CER and personalized medicine is that in the current healthcare debate both are highly contentious topics that are perceived as being disruptive to the current healthcare system and requiring significant monetary investments.
However, Woodcock maintained that this is a necessary and worthwhile investment.
"I believe the US needs to build a research infrastructure to support academic and community doctors, and other healthcare professionals, in performing this kind of research," continued Woodcock, who has been an outspoken proponent of personalized medicine at the FDA.
"It's a very unsexy topic but we aren't going to get from point A to where we want to be without making this investment," she continued. "If we don't involve the people in healthcare delivery in this research and have them be enlisted and invested, we're going to have trouble convincing people to change their ways."
As part of the American Recovery and Reinvestment Act, the US Congress granted a total of $1.1 billion for comparative effectiveness research: $400 million to the NIH, $300 million to Agency for Healthcare Research & Quality, and $400 million to the Office of the HHS Secretary to create the Federal Coordinating Council for Comparative Effectiveness Research.
Since the Federal Coordinating Council issued its priorities to HHS for CER research earlier this year, the PMC has been urging for greater focus on genomic strategies in CER [see PGx Reporter 07-08-2009]. In its report, the federal council mentioned the need to look beyond randomized-controlled studies to advance personalized medicine, but most of the priority areas for funding described traditional clinical trials for interventions for the average population.
NIH is supporting several CER projects in genomics with ARRA funds. Earlier this month, the National Cancer Institute awarded $4 million under the federal stimulus package to the Fred Hutchinson Cancer Research Center for the development of Comparative Effectiveness Research in Cancer Genomics, or CancerGen, which will focus on assessing the predictive and prognostic value of genomic tools in treating cancer. The first study conducted at CancerGen will investigate the predictive ability of Genomic Health's Oncotype DX test to determine which node-positive breast cancer patients will benefit from a tamoxifen-based chemotherapy regimen [see PGx Reporter 10-21-2009].
The NIH has also awarded researchers at the Kaiser Foundation Research Institute around $2 million to study comparative effectiveness in genomic and personalized medicine in colon cancer and nearly $2 million to the University of Pennsylvania for comparative effectiveness studies in genomic medicine.
Within AHRQ there will be an effort to analyze how readily its CER is considering effectiveness of treatments in specific subpopulations and determine if the agency's CER methods need to be updated.
Around the same time the Federal Coordinating Council released its report, the Institute of Medicine released its 100 top CER priorities, which included only a few proposals for comparing the effectiveness of treatments or clinical management strategies using genomics.
Plavix: An Example of Convergence
During her talk, Woodcock cited the example of the anti-platelet drug Plavix to illustrate an example of how comparative effectiveness research and personalized medicine can help inform healthcare providers of more effective treatments for specific subpopulations.
The labeling for Bristol-Myers Squibb/Sanofi-Aventis' best-selling drug Plavix was recently updated by the FDA to inform doctors and patients that poor CYP2C19 metabolizers have diminished response to the drug and increased risk of heart attack [see PGx Reporter 06-17-2009].
"The better way to do this is to not expose people to clopidogrel if they don't have a chance of benefit," Woodcock said at the meeting. "So a better comparative effectiveness trial would be to compare genotyped people who have a benefit on clopidogrel against prasugrel [Eli-Lilly/Daiichi Sankyo's Effient]. I'm not picking on anybody's drug, but this is obvious I think."
[ pagebreak ]
Woodcock added that "one of the reasons [Effient] was developed is that we don't have the capacity to do rapid genotyping in acute coronary problems," she added.
Effient was approved by the FDA in July and is indicated as a treatment to reduce the risk of blood clots in patients who undergo angioplasty. In a head-to-head study, although Effient has been shown to be more efficacious than Plavix in the general population, some patients on Effient displayed a higher risk of fatal bleeding.
Following FDA's update to Plavix's label, the pharmacy-benefit manager Medco announced the launch of the Genotype-Guided Comparison of Clopidogrel and Prasugrel Outcomes Study. The comparative effectiveness trial will enroll more than 14,000 acute coronary syndrome patients who have been newly prescribed either Effient or Plavix. The non-randomized controlled study will compare the efficacy of the two drugs by measuring the rate of cardiovascular deaths, nonfatal heart attacks, and nonfatal strokes. Patients in the clopidogrel arm of the study will be required to give a saliva sample so their CYP2C19 status can be established through genetic testing [see PGx Reporter 10-21-2009].
The structure of this and other studies at Medco may come close to the community-based studies Woodcock was referring to.
"We're focusing on conducting clinical studies that we can conduct using the real-world setting. We're not predefining academic or clinical centers from which we are recruiting our patients," Felix Frueh, Medco's vice president of research and development in personalized medicine, told Pharmacogenomics Reporter this week.
"For example, if you're running a randomized-controlled trial, like the NHLBI [Clarification of Optimal Anticoagulation through Genetics] trial for [the anticoagulant] warfarin, where you have 13 or 14 fairly highly skilled academic centers that are enrolling the patients, you are getting a different view than the studies that we have been doing with the Mayo Clinic, where we are enrolling patients from across the US," Frueh said. "We don't predefine sites from where we are recruiting patients."
The controversy surrounding the clinical value of PGx-guided warfarin dosing has sparked debate on whether personalized medicine treatments should be evaluated based on the evidentiary gold standard, randomized controlled trials, or if there are alternative methods by which to assess such treatments. Medco is trying to show that good evidence can be drawn from studies other than RCTs. Its clinical utility and cost-effectiveness studies of phamacogenomics treatments, under its personalized medicine program, are of varying design — observational, randomized-controlled, non-randomized controlled, etc. — and recruit patients from its diverse customer base of 60 million people.
Frueh noted that in a real-world setting the enrollment criteria aren't as stringent as studies done at academic medical centers. "So, again, the view that you get from the real world is a much broader and generally applicable viewpoint that reflects what clinical practice is going to look like," he said.
However, Frueh pointed out that Medco is doing personalized medicine research in areas where the gene associations are well-established in medical literature. The US research infrastructure needs to be improved for conducting "in-depth clinical trials in situations where the knowledge isn't as widely established," Frueh said.
Improvements in health information technology and greater adoption of electronic health records might facilitate more research in community-based settings.
The Rapid Learning Project, based at George Washington University and funded by the Robert Wood Johnson Foundation, is an effort to develop a national electronic system that links research databases and EHRs from private and public healthcare providers. The aim of the effort is to learn from real-world patient experience about treatments and adverse reactions, and to apply this knowledge to policy decisions and medical practice.
ARRA provided approximately $19 billion for Medicare and Medicaid Health IT incentives over five years and offered incentives to healthcare providers for moving to an EHR system.
In the PMC/Lewin Group report, implementation of health IT and EHR was identified as being instrumental to ensuring that "evidence pertaining to [personalized medicine] is present and actionable at the point of healthcare decisions."
Medco is "working on developing a more in-depth infrastructure with others," Frueh said. "If you talk to academicians, it is also within their interests to get access into that real-world setting."