Director of pharmacogenetics
Virginia Commonwealth University
Name: Bonny Bukaveckas
Position: Director of pharmacogenetics, assistant professor in pathology and pharmacy, Virginia Commonwealth University, 2004 — the present
Background: Clinical chemistry fellow, Pharmacogenetics diagnostics laboratory, University of Louisville, Ky., 2002 — 2004
Education: PhD in Environmental Biology, University of Louisville, Ky., 1998
Fellow, National Academy of Clinical Biochemistry, 2006
In light of the small number of commercial services devoted to interpreting genomic information for use by clinicians, and clinicians’ continued lack of understanding of pharmacogenomics [see 8/2/2006 PGx Reporter], Pharmacogenomics Reporter asked Bonny Bukaveckas to describe how she sees the field moving into the clinic.
Buckaveckas, director of pharmacogenomics at Virginia Commonwealth University, has been working on warfarin-dosing aids that incorporate genomics, as well as improving clinicians’ understanding of pharmacogenomics to speed its acceptance.
She chaired a session at the July meeting of the American Association for Clinical Chemistry entitled Perspective and Tools: How to Interpret PGx Testing, and she delivered a talk entitled It Takes a Village: Pharmacogenomics in an AcademicMedicalCenter.
Your talk at the American Association for Clinical Chemistry meeting about interpreting pharmacogenomic information was mostly intended for diagnosticians, who are generally further ahead of clinicians in understanding the field. How much further ahead are they, and how can they help bring pharmacogenomics into the clinic?
Well, we on the laboratory side of things are very good at interpreting something in terms of another laboratory test. We’re far ahead of, say, tying the pharmacogenetics test to something from therapeutic drug monitoring.
So those medications that require blood levels to be analyzed — something like phenytoin, or even the response to medications such as warfarin, where you measure INR [international normalized ratio] — we in the laboratory are very good about interpreting our genetic tests to go with another laboratory test.
Having said that, that’s not exactly what you need in order to actually use it in clinical practice. What you need to know is, ‘So what do I do with this patient? What alteration, if any, do I need to make in medications?’
That was the basis of the workshop, to try to give people some tools on the laboratory side to move in that direction: [to] get away from an interpretation that’s based in the laboratory and on laboratory results, and move it more into practice: ‘How do we do that?’; What tools do we need?’ — [to] take advantage of the medical record and all of the rich information that’s in there, in order to actually apply it to this patient in front of the physician.
Going even further, that sort of works when you’ve already had the physician order the test and they have the result sitting in front of them — what should it look like? And what change should be made in the clinical situation?
More importantly, I think, in order to integrate this into care, is to make it so easy to use as part of clinical care that’s it’s like the bar on a door that you exit an auditorium with — it’s obvious what the use is. It helps you to open the door. Our pharmacogenetic tests are nowhere near that easy to use. So, we have to integrate them into basic laboratory services, so that it’s easier to use them than to not use them.
But pharmacogenetic tests really can’t be used in isolation. You really have to take into account a lot of other factors in order to make a decision about drugs, other medications.
So, one of the things that we’re going up against is that many physicians … have a favorite dose of medication that they put everybody on. We don’t often optimize medications to the patient in front of them now, using the factors that are not genetic that we know about. And now we’re asking physicians to use genetics.
It’s sort of like asking them to take into account a further modifying factor, when they’re not even taking into account other factors. For example, we know that [in warfarin prescribing] age affects recommended dose, and yet many prescribers do not adjust for age. Now we know genetics is important— again, it’s not being taken into account.
What we’re doing here, and I think it’s being paralleled at other sites, is … ‘What is it going to take to actually make it easier to use this, when it should be used, than to not use it?’
So there are a lot of electronic tools that we’re working on [at Virginia Commonweath University]. We’re developing some PDA tools for physicians — specifically the first one we’re working on is for warfarin. It’s called SmartWarf. Pharmacists or physicians will basically be able to input the patient information — whatever’s important for the algorithm to calculate dose, including genetics.
Most importantly, it will allow them to see [if] genetics is even relevant before they order the test. Because if you’ve got somebody who’s 90 years old in front of you, there’s no reason to order genotyping, because it’s not going to change anything.
Building those kinds of tools, I think. We’re going to use that at the [Veterans Administration], actually.
Having said that, mostly what we’re dealing with are hospital patients, so in many ways, the environment is much more controlled.
Also, there are very low-tech tools, like having grand rounds that are very focused on a specific disease state or a specific medication — not the test. Nobody’s going to come to hear a seminar about a new lab assay — nobody cares. What they do care about is how to care for asthma patients better, or how to care for patients with [acute lymphoblastic leukemia].
So, instead of producing continuing medical education courses on pharmacogenomics, just incorporating pharmacogenomics into disease centered courses or sessions?
That’s a good summary. Approaching this from the lab side is not going to work. I think we’ve seen that — it hasn’t worked for the past four years, five years, where we’ve had good tests available. Because frankly, I don’t think physicians or ordering pharmacists care about most lab tests that come out — it’s just not the way they think. They care about the patient.
So, if we really focus on what can make the care of the patient in front of you better — what are we offering for that? — the lab tests are just part of that, in order to make a decision.
What do you think physicians need in a test’s output? So, in the case of warfarin, in order to give the right dose, are diagnosticians in a position to recommend a certain dose, given a certain height and weight?
That’s really where it’s going in the case of warfarin. Warfarin’s a little bit special and a little bit easier, because it’s really hard to dose — everybody needs titration or adjustment of the dose, and it’s well recognized. It’s also often handled within a central location within a health system (now, out in the community, it’s different) — but oftentimes there’s a warfarin clinic that people go to. So it’s kind of the ideal environment to try something like this.
It is also a little bit special, in that you can give the first dose just on the package insert recommendation, which is five milligrams for everybody. And you’ll be fine, because of the nature of the drug. You really don’t have to worry about modifying factors, such as age or genetics — that sort of thing — until it’s the third or fourth day, which really works out well in terms of using genetics along with the other modifying factors. Because by then, you’ve got the testing results back.
And so, the result — if it’s incorporated with the rest of the medical record — really will be in a milligrams-per-week answer. The genotype should be reported, but that’s not what the physician is looking at. He’s looking at the dose [and] the dose-change interval. So, if it’s not quite right the first time, we’re going to make an adjustment.
But the time that we wait to see if we need to make another adjustment [the adjustment interval] — that also changes as a result of genetics. And so, the useable result on the physician’s side is recommended dose and this dosing interval. And underlying that is a genotype result, but it’s also a height, a weight, a gender, an ethnicity, if applicable.
So that’s one thing we’ve really learned — we report genotypes but it’s not the primary focus of the result, it’s the actionable result that’s important.
So for other pharmacogenomic indications, do you expect them to be similar?
I think that in order for it to be useful, it’s going to have to be.
Nobody wants ‘star-this’ or ‘A-this’ or ‘T-that,’ it’s not useful — we know that — until it becomes so widespread and people see the results again and again, and they know exactly what it means. Sort of like the way that the HIV field has gone.
Physicians who are caring for HIV patients — because they see so many of them — can look at a genotyping result and they can look at the letters and numbers to see different polymorphisms of the virus. And they basically know what [is necessary] without seeing the red, yellow, green names of the drugs.
That’s a very small subset of physicians in the United States.
They’ve just gained a level of familiarity?
Exactly. And that will happen with pharmacogenetic tests as well -- although I would say that HIV genotyping is a pharmacogenetic test.
But the results reporting — one of the speakers in our workshop was actually talking about the parallels between where HIV was about 10 years ago, versus where we are with pharmacogenetics now. [They are] about comparable, in terms of what the reports look like and how useable they are for the physicians and who’s using them.
I hope it doesn’t take 10 years for us — I don’t think it will. I think there’s an acceleration in acceptance of these kinds of tests.