At A Glance
Name: Jeffrey Moe
Title: Executive in residence and senior director of business development for the health sector management program at the Duke University Fuqua School of Business
Background: Director of business development for the neurology division of GlaxoSmithKline
Education: PhD in organizational behavior from University of North Carolina, Chapel Hill
As the number and diversity of pharmacogenomics studies continue to grow, technology makers and service providers increasingly feel comfortable to market their products to physicians or even patients — even when data supporting the clinical utility of their products or services is scarce.
Cytochrome P450 tests, which are used more and more in the clinical setting each year — and have even made their way into the package inserts for widely prescribed drugs — are frequently cited as examples of good pharmacogenomics technologies gone bad: Eager to pocket a return on their investment, technology providers, pharma companies, and clinical reference labs flog these tests downstream with abandon.
But might there be a downside? Might the widely accepted fact that these products require additional research inadvertently hurt the health-care system — or even patients? Jeffrey Moe, senior director of business development for the Health Sector Management program at Duke University's Fuqua School of Business thinks so. SNPtech Pharmacogenomics Reporter caught up with Moe following his presentation at a two-day genomics conference held at Louisiana State University last week.
Let’s start with something you said last week at the LSU conference — that pharmacogenomics is exacerbating all of the problems in health care today. Pretty powerful comment. Care to explain?
One, when we talk about patient safety, which seems to be the promise of pharmacogenomics, there are currently big challenges as described in a series of reports that have come out of the Institute of Medicine [of the National Academies] and a variety of research reports that have been published in the New England Journal of Medicine. The notion that somehow we will be able to overcome those problems by adding pharmacogenomics, I would say it is just the opposite: That the promise of pharmacogenomcis may in a sense be swamped by the inherent difficulties we have with simpler technologies providing better patient safety.
For example, it is now estimated to take around 17 years from the publication of something that creates a new standard of care [to where] its adoption is widespread. So I would suggest there is a long period here before we can expect to have some modest gains in patient safety as a result of pharmacogenomics — not because there’s anything wrong with the science, but because it’s difficult in these human institutions; physicians changing practice and changing their own habits. That’s an enormous hurdle that must be vaulted before you can get [pharmacogenomics] there.
Is this a short term problem?
No, I think it’s a long-term problem. I think it’s endemic to health care; it’s endemic to the way physicians are trained. I don’t think you have to look too much further than TPMT [thiopurine methyltransferase] ... to say that physicians were the primary obstacle to asking that there be a label for mercaptopurine mandating that the test [for TPMT] be done. (The major dose-limiting effect of the mercaptopurine drug class is myelosuppression, and patients with an inherited deficiency of the TPMT enzyme are at risk of developing rapid bone marrow suppression. Because of the novelty of the TMPT test, and because of limited research available, some physicians are not willing to test patients for TPMT before prescribing the widely used drugs. —Ed.)
I’m not so sure that a regulatory requirement is the right way to go forward; one can debate that. And I think there are market forces: For instance, when third-party insurers say they won’t pay for 6-MP [6-mercaptopurine] for patients who have been diagnosed with acute lymphoblastic leukemia, this in effect makes the test mandatory with regard to those patients. ...
It’s clear that the physicians themselves had other markers that they could observe, and that they weren’t sure that one needed a diagnostic marker like the TPMT. That suggests to me that another barrier may be there beyond the fact that there are challenges within the whole realm of patient safety and risk management regarding drugs. I think there is also within the physician community some reluctance to have a mandatory test or something that would require them to take an [extra] step that would force them to think that there are other alternatives they have.
We’ve actually heard Janet Woodcock [head of the US Food and Drug Administration’s Center for Drug Evaluation and Research] say something similar, that she predicts — and indeed has already noticed — that one of the biggest challenges of pharmacogenomics is that of getting through to physicians, who are taught to evolve very slowly in the way they diagnose and treat disease.
If a physician is offered a CYP450 test versus a low dose of [blood-thinning drug] Coumadin, which is often cited with adverse events, far and away they would prefer to use Coumadin. Why? Because it doesn’t require a significant change in the flow of their practice or the way they do medicine, and they’re not having to interpret difficult pharmacogenomic data that they may not be quite clear exactly what it means.
We haven’t even gotten into the economics of these kinds of tests. Now, I'm making this up, but let’s say that the cost of a P450 test is $300. Over time, if I’m only looking for ... less than 20 percent of the population that has this defect, is that cost effective? I think a physician would say, ‘I don't think it is. It’s easier and safer, if you will, for me to just use low-dose Warfarin.’
Although studies have not been done to establish cost effectiveness, I think there is some significant instinct among physicians to say, ‘This is going to be too complex for me, and I think simpler, cheaper solutions exist without having to do pharmacogenomics.’
Now, I’m pro-pharmacogenomics, and I think there are some opportunities to reduce some risks for some adverse drug reactions — but I think they’re small compared to the size of the opportunity that many people inside the industry see.
Who do you think should take the lead in reconciling this? Clearly not the regulatory agencies or government, as you suggested earlier …
I think the opportunity is probably in a small subspecialty — and oncology comes to mind. They’re part of a smaller number of treaters, they are more familiar with dealing with high-risk therapies and procedures. I think there is a greater inclination among them to try new things — they’re innovators themselves — and their patients are more willing to be innovative because of their health status.
In the larger scheme of things, people have written that there’s not a day spent in medical school training on pharmacogenomics. But even if that were to occur, I think there are other barriers in the real-life circumstance — like the fact that patients generally don’t present with a single malady. They present with a variety of concurrent conditions and there are poly-pharmacy issues. And because the science is so young, we really don’t have a lot of experience, even in the pure vanilla case, exactly how to use this [pharmacogenomics technology] or how to interpret it. How does one titrate the dosage up or down for slow metabolizers or slow acetylators? We don’t know. Now that’s not to say that physicians don’t do that all the time; they do. But they like to have a foundation of information, and we just don’t have that rich source at this point.
Do you have a timeline where you think these issues might be resolved?
No. I don’t have any data — this is just a hunch — but I think that an event is going to occur that is going to bring some attention to [pharmacogenomics], and it is going to cause people to pay a bit more attention, if you will. I think it’s easier right now to be somewhat indifferent to pharmacogenomics in its clinical application. The science and the idea and prospect and promise of it is enormous. But its actual use an application in any real and significant way is much lower.
I think it will be interesting to see where something like TPMT goes. I certainly read between the lines that the [FDA] would like to revisit that decision that was made [not to change labeling for mercaptopurine drugs to require TPMY testing] to see if there can be some requirement. But I’m not arguing that making it a regulation is the right way to go.
But is certainly will have an impact. I think the fear among the providers is that if you make [the test] mandatory, that there may be some physicians who may say, ‘I don’t even want to use [mercaptopurine] therapy-that there will be an unintended consequence of dampening their enthusiasm or their use of mercaptopurine. But I do think that if the agency stepped in and took that step, it would certainly bring a lot of attention [to pharmacogenomics].
In think the market forces are already having an impact on it — frankly that strikes me as more efficient in the long run. Physicians are sensitive that patients want to be reimbursed, and if they looked at this and realized that the insurer is going to require that they use the TPMT diagnostic before the can be reimbursed for using it, that to me is a powerful incentive.