Research, Policy Advances Needed to Fulfill Promise of Genetic Medicine, Experts Say at AAAS

Dr. Mark B. McClellan

Personalized medicine has been touted as the next revolution in health care, but progress in bringing the right treatment to the right patient at the right time has been limited and will require some policy changes to blossom, according to Dr. Mark B. McClellan, former head of the U.S. Food and Drug Administration.

McClellan spoke at a 20 June conference at AAAS where experts and stakeholders with a wide range of interests discussed both the promise and challenges of translating some of the recent advances in genetics into real-world applications that could benefit patients. The conference was co-sponsored and co-organized by the Food and Drug Law Institute and the AAAS Scientific Freedom, Responsibility and Law Program.

Personalized medicine is not a new concept, of course. Doctors have always considered factors such as family history and lifestyle when recommending treatment. But the decoding of the human genome has raised the possibility that researchers can tailor diagnostic tests and treatments to a person's individual genetic profile.

With more sophisticated information to work with and a better understanding of how specific diseases arise, McClellan said, researchers could help patients avoid health risks at an affordable cost. There have been some successes, including the finding that two genes can affect how patients respond to warfarin, a widely used blood-thinning drug. The FDA recently changed the drug's label to assist physicians in prescribing the safest dose for patients.

Maryellen de Mars

Maryellen de Mars, director of the clinical biomarkers program for the nonprofit Critical Path Institute in Tucson, Ariz., noted several other promising developments, including use of the gene-targeted drug Gleevec for treatment of chronic myeloid leukemia; use of the drug Herceptin against breast cancers with a specific protein biomarker called Her-2; and use of a new class of drugs that target certain lung cancer cells. But she said that even the success stories have been mixed. Gene mutations have been producing resistance to Gleevec in some patients; about 15,000 patients a year are tested incorrectly with the breast cancer biomarker test; and the targeted lung cancer drugs are effective in only 10 to 20% of cases.

McClellan, de Mars and others at the conference said that much more remains to be done, not only in a better understanding of the underlying science of personalized medicine but also in regulatory and insurance reforms to ease the transition to the new era. Significant privacy and consumer protection issues also have been raised by the genomics revolution and personalized DNA testing.

"The potential is there for a lot more tailored and targeted therapies," McClellan said. "A lot of investment is going on in this area." But the potential remains unfulfilled, he said, both because of the continuing scientific challenges and because of the impact of government policies, such as Medicare reimbursement guidelines, that can affect the adoption of new tests and procedures. Moreover, the discussion about the future of personalized medicine is playing out against a more general debate on how to provide universal, cost-effective access to health care.

McClellan, now director of the Engelberg Center for Health Care Reform at the Brookings Institution, brings an unusual perspective to the discussion of personalized medicine. In addition to his stint as head of FDA, he also served in the Bush administration as head of the government agency that oversees the federal Medicare and Medicaid insurance programs. In the era of personalized medicine, he said, it is going to be increasingly difficult for insurers to come up with rules of coverage, given the complexity of patient preferences, histories, test results and other factors that will be built into a system of more personalized medicine.

"It is a very challenging environment," McClellan said, "even though the potential payoff is tremendous." He spoke of regulatory reforms to support a more targeted development process, including steps to pull together information more effectively about the impact of tests and treatments on individual patients. It is essential, he said, to know whether a particular product actually works for a specific patient. That will require better tracking of adverse drug effects after the product is approved, he said. There are steps underway to do active surveillance of drugs and other medical products with the help of better electronic records keeping. Under a life-cycle approach, he said, the initial approval of a drug "would not be the end of the line."

McClellan also emphasized the need for reimbursement reform. Already in Medicare and private plans, he said, there has been a trend away from payments based on the number of products being sold and toward payment based on the value of a drug for the patient. He predicted even more use of tiered payment systems that favor treatments with demonstrated benefit for subsets of patients.

Despite the complex challenges, McClellan described himself as a cautious optimist regarding the future of personalized medicine. Changes in regulations and reimbursement guidelines could provide a more viable pathway to new, targeted therapies, he said. But as a veteran of Washington's bureaucratic wars, he also noted, "It's not going to be easy. It's not going to happen right away."

Susan Dentzer

In a morning roundtable at the conference, panelists grappled with a hypothetical case study that demonstrated some of the challenges of personalized medicine. Moderator Susan Dentzer, editor of the journal Health Affairs, presented background on a hypothetical biomarker called Luc-3, which preliminary studies showed might be the most accurate predictor of whether a breast cancer will return.

Under the scenario, 95% of women with advanced cancer that had spread to the lymph nodes would have the Luc-3 marker. But the scenario also postulates that some doctors would advocate testing all women with breast cancer for the marker, even those with small tumors that have not spread to the lymph nodes. Any woman who tested positive for the marker would receive chemotherapy—with its adverse, sometimes serious side effects—even if the risk of tumor recurrence was low and chemotherapy would not have been recommended under standard treatment protocols. Should women in the low-risk group go ahead with the treatment anyway? Dentzer asked.

Dr. Howard Levy

Dr. Howard Levy, an assistant professor of genetic medicine at Johns Hopkins University, warned against "the fallacy of genetic determinism." Except in rare cases, a genetic test alone is not going to give the final answer, Levy said. Patients must weigh risks and benefits in making their decisions, much as they do for any medical procedure. And Levy noted that medicine, in general, has become more patient-centered in recent years, a trend that will accelerate as treatments become more tailored to individual patients. "I would see my role," he said, "as just an adviser."

Annette Bar-Cohen, director of programs for the National Breast Cancer Coalition, said it would be important to learn as much as possible about the reliability of the hypothetical Luc-3 biomarker as a treatment guide. To that end, she said, doctors must be encouraged to get patients into clinical trials to learn about recurrence rates and long-term survival outcomes. "At this stage, we really don't have the solid evidence that we need" to prove the utility of the biomarker, Bar-Cohen said.

Charles Rotimi

Charles Rotimi, director of the NIH Intramural Center for Genomics and Health Disparities, agreed about the need for clinical trials and said that, even in a time of constrained budgets, money must be found to undertake the needed clinical evaluations.

Greg Downing

But Greg Downing, director of the personalized health care program for the U.S. Department of Health and Human Services, said clinical trials may not be the answer in all cases. "Everything won't be able to be tested through clinical trials," Downing said. He called for improved gathering and sharing of information about new developments in evidence-based medicine. He also stressed the need for better education and training programs on advances in personalized medicine. Such basic steps should not be delayed, he said. "It's about the future of medicine," Downing said. "If we don't make the investments now, we'll be struggling with the same situation 30 years from now."

Dr. Kavita Patel

Dr. Kavita Patel, deputy staff director of the Senate Health, Education, Labor and Pensions Committee, said that the funding outlook for new initiatives is not good. After a period in which the budget for the National Institutes of Health doubled, "we're just trying to keep the funding level flat across the Institutes," she said. Asked what she would recommend to her boss, Sen. Edward Kennedy (D-Mass.)—who is himself battling cancer—Patel agreed about the need for programs that do a better job of disseminating information to doctors on the latest developments in personalized medicine.

At the same time, Patel said, legislators must walk a fine line. "A little bit of knowledge can be a very dangerous thing for policy makers," she said. They must guard against the impulse to try to micromanage the scientific research process and instead find ways to help deliver the fruits of medical research to the public.

Patel also said that access to adequate health insurance will be essential in an era of personalized medicine. Consumers must not feel pressured to accept one treatment option over another because of constraints on their health coverage, she said.

Dr. Joanne Armstrong, a senior medical director for Aetna Inc., outlined some of the guidelines her company and others now use in deciding whether a medical service or test is covered. Insurers look favorably on tests that have been described in the peer-reviewed scientific literature and have what she called "analytic validity," or results that can be demonstrated repeatedly in the lab. She also cited a need for clinical validity: "Does the marker measure what you say it will measure?" And perhaps most importantly, Armstrong said, does it have clinical utility—is there evidence that reliance on the marker will actually improve the patient's prospects?

"The more innovative a product is, the more walls it runs into" in the approval process, said Dr. Bruce Quinn, a senior health policy specialist with Foley Hoag LLP who spoke during the afternoon session. He said 1980s conventions on reimbursement "are colliding with 2020 medicine." Reimbursement codes may not be flexible enough, he said, when there many distinctions between lab tests and multiple screening tests available for a particular disease. Steve McPhail, president of Expression Analysis Inc., a North Carolina supplier of microarray genomics tests, said there remain "very, very few examples of the cost benefit of pharmacogenetics" for personalized medicine. Companies like his need "to generate the data needed to support reimbursement decisions in clinical trials," McPhail said.

Given the complexities of developing diagnostic tests with sufficient sensitivity and specificity, panelists said doctors and patients alike will be entering some new territory. Carole Isaacson Barash, an ethicist with Genetics, Ethics & Policy Consulting, Inc., said they will face questions about what constitutes good clinical medicine "in the face of imperfect, insufficient, preliminary information." She also said the medical community has a firm obligation to protect that information (such as the finding that an individual has an elevated risk of disease) from unauthorized disclosure.

At the same time, there is a growing concern that consumers may decide to take matters into their own hands when it comes to learning more about their genetic predispositions. Already there are concerns about direct-to-consumer advertising by companies seeking to sell the newest genetic tests, whether they have demonstrated scientific utility or not.

Grail Sipes, an attorney with Covington & Burling, said that a few states have laws requiring that patients be counseled on the risks and benefits of tests being sold directly to consumers. "That's far from the case everywhere," she said. Consumers are asking for tests, she said, for which they are ill-prepared to make health care decisions. California recently told 13 companies to stop offering genetic tests directly to its residents after complaints about the cost and accuracy of the tests. Physicians have expressed concern, for example, about tests that may be based on mutations that only slightly raise the risk of disease.

The AAAS-hosted conference was the first in a planned series related to personalized medicine, according to Mark Frankel, director of the AAAS Scientific Freedom, Responsibility and Law Program. He said AAAS and the Food and Drug Law Institute would like to do several in-depth colloquia on the topic during the next year or two.

Earl Lane

30 June 2008