 |
 |
 |
|
 |
FDA Home Page | Search FDA Site | FDA A-Z Index | Contact FDA
This is the first in a series of interviews that explore some of the projects underway as part of the Critical Path Initiative. Today we interview four FDA staff who have been involved in the Critical Path project related to pharmacogenetics of warfarin therapy. This project will help improve patient safety while greatly reducing related healthcare costs.1
Randall Lutter is FDA's Acting Deputy Commissioner
for Policy
Clark Nardinelli is a senior member of FDA's economic
staff
Myong-Jin Kim is a Clinical Pharmacology Team Leader in
the Division of Clinical Pharmacology 3, Office
of Clinical Pharmacology, Office of Translational Sciences
Rafel (Dwaine) Rieves is an Acting Division Director in
the Division of Medical Imaging
and Hematology Products, Office of Oncology Drug Products
Q: What is FDA's Critical Path Initiative?
Lutter: For medical products, the "Critical Path" is the scientific process through which a new medical product is transformed from a discovery or "proof of concept" into a useful human drug or biological product or medical device. When a product is identified as a promising candidate in basic research, it then enters a product development process that consists of a series of scientific evaluations, often increasing in size and complexity, to predict whether the candidate will be safe and effective. These tests also guide a drug sponsor in choosing an appropriate dose and regimen.
Unfortunately, many of the scientific tools and tests used to evaluate the safety and effectiveness of a candidate product have not kept pace with the rapid evolution occurring in other arenas and do not allow us to capitalize on those advances in basic science. We must make a concerted effort to apply new scientific knowledge — in areas such as gene expression, analytic methods, and bioinformatics — to product development. The FDA's Critical Path Initiative seeks to harness these new sciences in the service of product development, so we can learn more about a product before it is marketed. With better information, sponsors will be able to make more informed decisions, for example, about which candidate products to move to the next phase of development, what doses will most likely prove to be safe and effective, and which products will likely fail. In short, these new sciences could provide information needed to help sponsors develop safer and more effective medical products.
Q: The working paper you posted to the AEI-Brookings Joint Center for Regulatory Studies Web site, "Health Care Savings from Personalizing Medicine Using Genetic Testing: The Case of Warfarin," evaluates the potential savings to the healthcare system that could accrue if we could develop a genotype-driven dosing regimen for the drug warfarin. Could you tell us a little about this drug?
Lutter: Warfarin is a commonly used anticoagulant (sometimes called a "blood thinner"); in the United States, roughly 2 million persons start taking warfarin each year. It is prescribed to patients who are at risk of developing blood clots, such as persons with atrial fibrillation (a type of abnormal heart beat), recurrent strokes, deep venous thrombosis, pulmonary embolism, or those who have received heart valve replacements. Warfarin is the second most common drug (after insulin) implicated in emergency room visits—causing more than 43,000 emergency cases per year.
Rieves: Unfortunately, warfarin is a difficult drug to prescribe. The optimal dose varies greatly from person to person, and the consequences of under- or over-dosing can be significant. Too little warfarin puts the person at increased risk of forming blood clots and having a stroke; too much warfarin puts the person at risk for a potentially devastating bleeding event—bleeds in the gastro-intestinal track and brain are the most common major events.
Q: How do doctors decide how much warfarin to give their patients?
Rieves: A variety of factors influence how much warfarin each patient may need, including sex, age, and weight. But, even taking these factors into account, the initial dose of warfarin chosen for any individual patient is an approximation. Doctors choose what seems to be the best dose, then test the patient for the next several weeks, adjusting the warfarin dose until the test results show that the patient's blood is properly anti-coagulated. It can take weeks to determine the right dose. During this time, the patient may be under- or over-anticoagulated, putting him or her at risk for the complications I just mentioned. In fact, the highest risk of warfarin complications occurs during the first 30 to 60 days after beginning warfarin therapy, during the period when the warfarin dose is being adjusted.
Q: Why is there such variation in how much warfarin is the right dose?
Kim: There are a lot of reasons. We are learning that some of the variation stems from individual differences in the enzymes that metabolize warfarin (enzymes that break warfarin down and destroy its anticoagulant activity). Some people have enzymes that break down warfarin more slowly than average—those people would need a lower dose of warfarin.
Q: And these genetic tests may help doctors determine who has the "slow metabolism" enzymes"?
Kim: Exactly. Scientists now know which genes code for these key warfarin-metabolizing enzymes and are developing accurate and relatively inexpensive tests for these genes (about $350), so we can identify persons who have the variations in these genes that cause slower metabolism of warfarin. If we can identify these patients in advance, they could be started on a lower warfarin dose than they otherwise would have received.
Q: How common are those genetic variations?
Kim: We have found that roughly one-third of the population carries a variation of a warfarin-metabolizing gene that results in slower warfarin metabolism. So, you see, the issue is quite significant from a population perspective.
Q: Designing a person's warfarin dose based in part on the person's genes sounds like "personalized healthcare."
Kim: That's right. A person who metabolizes warfarin slowly is at increased risk of a serious bleeding event on the average starting dose. But if we can identify those people in advance and start them on a lower dose, we can reduce the number who will have warfarin-related bleeds. We estimate that by identifying people who carry genes for the slower metabolizing enzymes, we could reduce serious bleeding events from 27.6% to 12.6% among these people. Similarly, we estimate that a genotype-driven dosing regimen could reduce warfarin-related strokes by half. Physicians may under-dose or hesitate to prescribe warfarin because of a bleeding risk. In this case, genotyping can help doctors to identify persons with normal warfarin metabolism and start them on warfarin without under-dosing and reduce stroke events.
Nardinelli: These are substantial numbers. Overall, if use of genetic testing to determine initial warfarin dose becomes standard practice and is as effective as some research has suggested, American warfarin users could avoid 85,000 serious bleeds and 17,000 strokes annually. Of course, such gains are still speculative and must await validation from data on clinical outcomes.
Q: The public health benefits of enhancing the safety of this drug are clear. Where does your economic analysis fit in?
Nardinelli: Strokes and bleeds are expensive to treat. On average, the direct healthcare costs of treating a bleeding event is $13,500, and the cost of treating a stroke is $39,000. Reducing the number of these serious adverse events will result in sizable savings to the healthcare delivery system. We have estimated that the annual net healthcare savings of integrating genetic testing into the decision-making process for administering warfarin therapy could be as much as $1 billion per year.
Q: That sounds like a lot of money.
Nardinelli: In some ways, these figures are conservative. We only looked at one genetic source of variations in warfarin metabolism. There is a second gene that also appears to play a role. In addition, our analysis substantially understates the full social benefits of using genetic testing to guide warfarin dosing decisions, because it accounts only for direct change in healthcare delivery costs and does not include changes in the costs associated with reduced functional status and quality of life for warfarin users.
Q: But simply knowing that a person metabolizes warfarin too quickly or too slowly does not tell us exactly how much warfarin to prescribe. How will doctors know how much warfarin to prescribe?
Rieves: Research is now underway to determine what dose to prescribe to a patient, based on his or her genotype.
Lutter: You know, these investigations are a great example of Critical Path research—evaluating how to apply the latest in genetic science in a clinical trial or a treatment setting.
Q: This doesn't sound like product development, it sounds like healthcare delivery. What does this have to do with the Critical Path Initiative?
Lutter: They are interrelated. The sciences that help us to develop medical products are the same as the sciences that help us know how to best use those products in the clinic. In both cases, it's about developing the knowledge with as much depth and precision as possible. If we can learn more about products before they are approved, we will be able to use them more intelligently after they are approved.
Warfarin has been around for many many years, but if the sponsor of warfarin had known about the genetic sources of variation in warfarin metabolism and had a genetic test for those variations when warfarin was being developed, the sponsor could have designed the clinical trials differently. The trials could have been designed to identify modified dosing regimens, and the trial subjects would have had fewer side effects. The trials could have been more efficient and more informative. This would have made warfarin a safer drug postapproval because clinicians would have had the information they need for safe prescribing. Drugs such as warfarin entering a product development process today would benefit from our growing understanding of differences in drug metabolism — possibly also from a genetic test for metabolism differences.
Nardinelli: In fact, many Critical Path tools would help individualize therapy, by identifying who is likely to respond well to a treatment, who may need a modified regimen, and who should avoid the product entirely. Warfarin dosing is a great example. The paper we wrote provides one concrete example of how this new revolution in medical care, commonly termed "personalized medicine," can be good for patients—in this case, by reducing the devastating complications that may result from under- or overdosing warfarin, even as it saves significant healthcare dollars. These results illustrate why the FDA remains committed to leading this revolution in healthcare, through its Critical Path Initiative.
Q: It is easy to see how better information about how drugs are metabolized can improve how we use them. Are the economic benefits of the Critical Path Initiative limited to improving drug regimens?
Nardinelli: No. For example, we currently are looking at the potential economic impact of having a rapid test for diagnosing and identifying tuberculosis that is resistant to standard tuberculosis therapy. Today it can take weeks or months to learn whether a person's TB can be treated with standard drugs, or should be treated more aggressively – or even determine with certainty that the patient has TB. A rapid test that could give us this information in just hours or days would be invaluable to researchers and companies trying to develop new therapies for these patients—and just imagine how it could revolutionize care delivery for TB as well as other infectious diseases, especially in developing countries. We also are studying the potential economic impact of a test that could rapidly distinguish influenza from the common cold. We plan to apply our economic analyses to many other developments in the areas of personalized medicine and Critical Path.
Q: Why is the FDA taking leadership in this effort? Isn't product development industry's job?
Lutter: FDA is uniquely suited to serve as a hub for this effort, because it oversees evaluation of all U.S. medical products and can identify the key scientific hurdles that commonly cause setbacks in product development for companies industry-wide. For example, in March 2006 we published the Critical Path Opportunities List and Report, which describes how more modern science could revolutionize medical product development. Based on FDA experience and the input of many stakeholders, this report also describes 76 concrete research projects that would further that goal.
Rieves: Because FDA sets the scientific standards for product development, we are uniquely situated to encourage the use of new Critical Path tools by accepting the results of the new tools as important considerations in product review. Also, we approve the product label, and labels influence how doctors and patients use drugs.
Lutter: But you're right, even though we have helped initiate more than 30 projects and collaborations to help modernize the Critical Path, we can only do our small part. Industry, academic researchers, patient advocacy groups, and others have to help do the research and consensus development that will show how to use the new science in product development work. Many of the Opportunities on the national Critical Path Challenges List will remain unmet opportunities unless we can succeed in getting folks to collaborate on the hard work of developing new tools for product development.
Thank you for your time.
1 FDA's economics staff have been exploring some of the potential reduction of healthcare costs and recently posted a paper to the AEI-Brookings Joint Center for Regulatory Studies Web site entitled Health Care Savings from Personalizing Medicine Using Genetic Testing: The Case of Warfarin. See http://www.aei.brookings.org/publications/abstract.php?pid=1127.