Personalized medicine: A biological approach to patient treatment
A conversation with Issam Zineh, Director of CDER’s Office of Clinical Pharmacology
How does CDER/FDA define personalized medicine?
Personalized medicine has a fairly broad definition, but, essentially, we're talking about using genetic or other biomarker information to make treatment decisions about patients. These could include decisions about who should get certain kinds of therapies or specific doses of a given therapy, or who should be monitored more carefully because they're predisposed to a particular safety issue. The terms genetics, pharmacogenetics, personalized medicine, and pharmacogenomics have been used interchangeably to mean the study of genetic variations and their influence on the way people respond to medications.
Describe CDER's involvement in this area.
Over the past decade, CDER has been proactive in thinking about personalized medicine in terms of drug development and regulatory decision making. Back in the early 2000s when the human genome project was coming to completion, we had a sense of the amount of genetic variability in the human genome. Senior leaders in the Office of Clinical Pharmacology and the Center articulated a vision for the integration of genomic sciences into regulatory review and drug development. We began publishing on this topic as early as 2001, and have continued over the years.
CDER has been developing infrastructure programs and review capacity to be on the leading edge of personalized medicine initiatives. Most recently, we have been working with other FDA centers to make sure that we develop informative and pragmatic guidance and policies on personalized medicine in a timely way.
How's that working?
I think it’s working well. The challenge has been that there are a lot of moving parts to personalized medicine.
Personalized medicine is generally comprised of two elements. One is the drug, biologic, or other therapeutic intervention, and second is the diagnostic test. We have needed to develop not just CDER’s policies, but multi-center policies on personalized medicine that involve the Center for Devices and Radiological Health (CDRH), the Center for Biologics Evaluation and Research, and other FDA centers. In my view, it was pretty challenging to coordinate that interaction from about 2008 to 2011, mainly because the centers were not physically located near each other.
Now that CDER and CDRH are on the same campus, there's been much closer interaction among the centers as far as joint guidance development and sharing of regulatory review experiences and processes. I see the last year or so as a success and, moving forward, it will be much easier to coordinate now that we are closer and realize more the need to work together to produce some of these important work products.
Have you had anyone say, "No, this is not the way to go," or had any disagreements about this area?
That's a great question. From an industry perspective, there has been a lot of promise in terms of personalized medicine. The idea that you could identify subsets of patients that are more likely to respond to a particular treatment has been very appealing. The fact that you can identify those patients with a biomarker has been something that many companies have looked for.
I think people had been feeling that there had not been many successes outside the field of oncology. However, a couple of recent drug approvals in other therapeutic areas have been notable for their kind of personalized medicine attributes. For example, we have had a recent approval of a new drug to treat cystic fibrosis and activity in several other therapeutic areas, signaling that the science might now be more broadly applied to areas outside of hematology and oncology.
People don't argue with the idea that you can refine your understanding of risk/benefit through personalized medicine. The main point of discussion is how you actually do it. What kinds of trial designs are required? What sort of evidence do you need? If you're studying one particular biomarker group, what kind of evidence do you need for the other group that may not be in that category (that is, “negative” for the biomarker)? So some questions are still up in the air. But no, I don't think there is opposition to the principal of personalized medicine.
You spoke earlier about industry -- what is industry’s overall reaction to the concept of personalized medicine?
If you talk to industry, there seems to be a lot of enthusiasm about where the science might be going. I think that's especially been true because of several guidances we've published in the area, or we have promised to develop under PDUFA reauthorizations.
Some companies have personalized medicine groups, though they may be referred to as “pharmacogenomics” groups or “stratified medicine” groups, which are obviously pretty enthusiastic about the science. However, a recent survey looked at the drug pipeline portfolios of about 20 major companies and showed that a very small minority included what could be called stratified or personalized medicine. When you ask people in industry about this, some feel it is because there's no clear regulatory pathway or guidelines.
We have been working to create that framework. Guidances have been, or are being, developed to help people incorporate principles of personalized medicine in early phases of drug development, and then leverage that information to make decisions about patient selection or clinical trial designs in later phases of drug development. I think that this guidance, along with our commitment to develop coordinated review processes between CDER, CDRH, and the other centers, will create more confidence that a regulatory framework to handle personalized medicines is in place.
What sort of timeframe are we talking about?
We published the draft guidance on pharmacogenetics in early development last February, and got comments back from the public. We are now in the process of modifying the guidance based on these comments. Another relevant guidance is on enrichment strategies in later phases of drug development, and includes a genomics component as one type of enrichment. This guidance was promised under the last Prescription Drug User Fee Act (PDUFA IV), and is currently being worked on. We also released a draft policy and definitional guidance on companion diagnostics that is now being finalized. One more guidance, this one on co‑development, is in the early stages.
We are also working on a variety of policies and procedures to coordinate the review of drugs and diagnostics simultaneously, once submissions come in under an investigational new drug application, a new drug application or a biologics license application. So, as far as timeframe, the guidance and other documents range from about to be released in final form to early stages of development.
How has the reauthorization of PDUFA affected the area of personalized medicine, if at all?
One of the enhancement proposals included in the recent PDUFA V reauthorization is related to personalized medicine. In developing the proposal, we made a compelling scientific and business case for further expanding the infrastructure that supports the review process. We provided several good examples of how review has contributed to advancing personalized medicine, and public stakeholders and the industry agreed.
As a result, this enhancement proposal allows for increased staff at CDER for review and guidance development in the area of personalized medicine. The proposal also includes a provision for holding a public workshop to explore how FDA, industry, external scientists, and consortia can interact in a more flexible way to advance biomarker science, pharmacogenomics, and other sub-disciplines in personalized medicine.
So the new proposal will give us more staff and also more opportunity for dialogue. Those two things, together with our advancements in policy development and review, mean we'll be in very good shape for continued progress over the next five years.
And finally, how will personalized medicine impact patients?
Well, that's the key issue. To understand the promise of pharmacogenomics, let’s look at the way we currently treat patients. If you see your doctor for high blood pressure for example, he or she is going to give you any of a number of blood pressure medications, based on very little information about what's going to work for you. In fact, there may not be any rationale for why you're getting that particular medication other than that your insurance covers it or the doctor has samples available. If the medication doesn't work in four to six weeks, you may be switched to another medicine, or get another medication added on top of the first.
We’ve learned that this trial-and-error approach leads to patient dissatisfaction, poor clinical outcomes, and greater expense, especially for chronic diseases. But in many situations this empirical approach is the best approach we have. Personalized medicine aims to streamline clinical decision making by using biological information available through a genetic test or biomarker, and then saying, "based on this profile, I think you're more likely to respond to Drug A or Drug B, or less likely to have an adverse reaction with Drug C.” The idea is to get patients on the right medication and to get them on it sooner.
If we can incorporate personalized medicine concepts into regulatory review and drug development earlier on, we could develop drugs with a higher likelihood for success in subpopulations. The patients who are most likely to show you that the drug works, if the drug, in fact, does work, could be identified up front as part of drug development. This would translate ideally to more effective medications and a greater understanding of the populations in which these medications work. It would also give us more knowledge about safety profiles, a better idea of who is likely to experience serious or life-threatening adverse events, and additional information about dosing. If you put all that together, you have a much more refined approach to treating patients, in terms of drug selection, drug dosing, and patient monitoring.
Issam Zineh is Director, Office of Clinical Pharmacology (OCP), and Co-Director of the Biomarker Qualification Program, Office of Translational Sciences, CDER/FDA. From 2008-2012, Dr. Zineh was the Associate Director for Genomics in OCP. He is an experienced clinical pharmacist who was formerly on the faculty of the University of Florida (UF) Colleges of Pharmacy and Medicine and Associate Director of the UF Center for Pharmacogenomics. Dr. Zineh received his PharmD from Northeastern University and completed his residency at Duke University Medical Center. He did a fellowship in cardiovascular pharmacogenomics at UF where he also obtained his MPH in Health Policy and Management. He is a recognized expert in the field of clinical pharmacology, pharmacotherapy, and pharmacogenomics. Dr. Zineh is chair of the Coriell Personalized Medicine Collaborative Pharmacogenomics Advisory Group, and sits on the Centers for Disease Control and Prevention's EGAPP steering committee. As Director of OCP, Dr. Zineh is a member of the CDER Senior Management Team and leads a staff of nearly 150 regulatory scientists in FDA's efforts to enhance drug development and promote regulatory innovation through applied clinical pharmacology.