Narrator - Dr. Abel  00:00
Welcome to HelixTalk, an educational podcast for healthcare students and providers, covering real life clinical pearls, professional pharmacy topics and drug therapy discussions.

Narrator - ?  00:11
This podcast is provided by pharmacists and faculty members at Rosalind Franklin University, College of Pharmacy.

Narrator - Dr. Abel  00:17
This podcast contains general information for educational purposes only. This is not professional advice and should not be used in lieu of obtaining advice from a qualified health care provider.

Narrator - ?  00:27
And now on to the show.

Dr. Sean Kane  00:31
Welcome to HelixTalk. Episode 91 I'm your co host,

Speaker 1  00:34
Dr. Kane. I'm Dr. Schuman, and today for HelixTalk, Episode 91 our title is 23 reasons to be excited about personalized medicine. We're excited to have a special guest with us. Dr. Dyson Wake is Senior Clinical Specialist in Pharmacogenomics at NorthShore University HealthSystem's Center for Molecular Medicine here in Chicagoland. He's here to kind of help talk with us about current and future applications of pharmacogenomics to the area of personalized medicine, as well as maybe expose some misconceptions. Dr. Wake, we're really excited to have you here today.

Speaker 2  01:02
You here today. Hi, and I'm very excited to be here today. I love talking about pharmacogenomics, and I'm, uh, I've loved lecturing classes at Rosalind Franklin, so it's finally, it's good to be on the podcast, uh, getting the voice out, that's fantastic.

Dr. Sean Kane  01:13
So why don't you just tell us a little bit about your background and kind of where you're at now and how you got to where you are today. Sure.

Speaker 2  01:20
So I graduated with a pharmacy degree from Drake University over in Des Moines, but I also graduated with a bachelor's in computer science. So I was looking at ways that I could use some of this computer science information to kind of benefit people in the realm of maximizing their medication efficacy. I also did a PGY one where we had a strong informatics and medication safety focus over at South Bend Memorial Hospital South Bend. From there, I found out about pharmacogenomics and kind of the dividends it can really pay for systems that are able to utilize it to improve patient care. So I went and did a PGY-2 in pharmacogenomics at the University of Florida, down in Gainesville, and then I was lucky enough to find a great position over at North Shore Hospital in Evanston, Illinois, where we have a great pharmacogenomics program, and we're really trying to get these benefits out to patients, get this information out to providers about how they can best utilize this information.

Speaker 1  02:11
So how many residency programs are there right now in pharmacogenomics?

Speaker 2  02:14
So I believe there are currently three residency programs. It's a little difficult to say, because some of them aren't fully accredited. Some of them are in the process, but I believe we have two fully accredited programs in one that's pending accreditation right now.

Dr. Sean Kane  02:26
And needless to say, this is like a growing field, right? So it makes sense that we wouldn't have as many as ambulatory care, critical care, oncology, but that's okay, right? Because this is kind of this emerging field where that's how it kind of emerges.

Speaker 2  02:39
Yeah, exactly. And there are more fellowships available for pharmacogenomics, but those are going to differ slightly in the activities that are provided for the resident or the fellow, where those are going to be a little bit more focused on research versus clinical application. Those opportunities a little more numerous, but again, it depends on exactly what the resident or fellow is looking for in their education.

Speaker 1  02:58
So first, I think, just to go back to some of the basics, how would you define pharmacogenomics? Exactly?

Speaker 2  03:04
That's a great question to start off with, because some people may define it slightly differently, but I think pharmacogenomics at its core is using these genetic variations that we all know we have the same way, we have variations in our hair color, eye color, all those things. We also have variations that determine things like how well our liver and our drug receptors function. So pharmacogenomics is really just using that information to predict which patients are going to best respond to different medications, or which medications we expect to be somewhat problem, medications for those patients. And I think

Speaker 1  03:35
it's interesting, because this is something that I believe that's been around for a while. We're just now getting into. I always love history, and so I was like looking into the history of pharma of pharmacy. And so I believe that really was something they first noted in identical twin studies, looking and saying, Hey, there's some similarities in drug metabolism among identical twins versus fraternal twins. So there must be something to it that predetermines how these individuals dispose of certain medications.

Speaker 2  03:59
Yeah, exactly. So we we knew that there was this genetic component. But only recently, as the cost of genetic testing has become somewhat more affordable, more accessible for more patients, have we been getting into the realm of being able to test kind of the the more common, more routine patient for these types of variations.

Speaker 1  04:17
And so I believe that one of the first things that was noticed in the 1950s and 1960s were regional differences in response to isoniazid related to variants in the gene coding for N‑acetyltransferase (NAT2). It evolved from there. The term "pharmacogenetics" was first used by Friedrich Vogel in 1959.

Speaker 2  04:39
Oh, and I think that's actually a good point, because we might actually say, on this podcast, pharmacogenomics, pharmacogenetics somewhat interspersed. And now there may be slight differences in those definitions, where pharmacogenomics is typically going to look at the whole genome, versus pharmacogenetics looking at specific genes. But in general, for most use, we're going to basically use those interchangeably. We may also at some point. Excellent. Let's say PGx, which is the abbreviation for those two

Dr. Sean Kane  05:03
and of course, like since 1959 our knowledge of the genome, our knowledge of how drugs are metabolized, everything has changed since kind of that first use of pharmacogenetics or pharmacogenomics. So where are we at today, in terms of, for the typical listener that isn't a pharmacogenomic specialist. What are some of the things that we're doing today that are relevant to what they might actually see in clinical practice?

Speaker 2  05:29
So some of the things we're doing today kind of a lay practitioner or more, the more common practitioner may see is these types of new tests looking at a lot of drug metabolizing enzymes that are breaking down some fairly common medications, and these may be impactful for whether those medications are going to show increased risks of adverse effects, decreased rates of efficacy, and whether those medications are going to be kind of a proper choice for that patient. However, there's not exactly a standard across all of these as to what information exactly they should be provided. So we're going to talk a little bit, I guess, about the specifics of those so that you can be a little more well informed before those kind of arrive at your doorstep.

Speaker 1  06:05
So I think the first thing to talk about is going to be this metabolic pathways to be phase one metabolism, so these enzymes that make molecules more polar, oxidation, reduction, hydrolysis. A lot of times we think about, you know, changing small pieces of the medication. So I believe one of the first ones to talk about is going to be CYP2C19. Is that correct?

Speaker 2  06:23
Yeah, and CYP2C19 is kind of a great place to start for someone who's interested in learning more about pharmacogenomics in general, because it has some somewhat profound effects without the huge degree of variation we see in some of the other genes we're going to talk about. CYP2C19, a drug-metabolizing enzyme, breaks down a fairly large number of common medications, and I believe the first one we should kind of talk about is clopidogrel. So when we look at CYP2C19 we can kind of divide patients into drug-metabolizing phenotypes. We can call them normal metabolizers with the expected amount of activity, intermediate metabolizers who have a little bit reduced enzymatic activity, or poor metabolizers who have almost minimal or absent enzymatic activity. We also have patients with increased activity for that enzyme. So those are going to be our rapid or ultra rapid metabolizers, who basically have more workers on that assembly line, more drug going through that process. This becomes important for our medications, because for breaking it down, or processing it more slowly, more quickly, we're going to be producing more or less of those active metabolites.

Dr. Sean Kane  07:24
So, Dr. Wake, I understand that we kind of give the phenotype to a patient, but on a kind of genetic test, we're going to get star then a number after it, right? Can you kind of orient the listeners to what the star means and how those come about?

Speaker 2  07:38
Yeah, exactly. So the first kind of area of confusion is the way these results are reported. Can be many different formats. Some of them are going to be at that phenotype level. Some are going to be genotype you may even get raw genetic data, which at that point you're going to need to ask someone for a little more help, because even I don't want to be looking at Raw genetic data. So the star alleles are kind of how we describe the variations that have been found. Typically, Star One is what we call wild type. It means we didn't find any variations. And importantly, it doesn't mean that there's not any variations. It means we just didn't find any and then we also may have things like star two, Star Three, Star 17. These are going to vary in what they mean by Gene, but for specifically, sip 2c, 19, our star two and our Star Three mean that that is a loss of function allele. Patients with one of those copies has less activity, and patients with a star 17, that's an increased function copy, so they're going to have more activity.

Dr. Sean Kane  08:29
And I noticed that you skipped from two to three to 17. So what happened to four through 16? Yeah.

Speaker 2  08:35
So unfortunately, these aren't necessarily named in the most convenient order, and there are star fours. There's actually a star for B, when we get into talking about some of the more complicated genes, there's, you know, Star four, A through K. And there's committees that determine these precise nomenclatures. What we need to focus on there is, these are interesting things. In my position, there might be interesting things for a clinical pharmacist, the provider at the therapeutic decision step, really can't be dealing with that level of information. We really need to assess those star alleles and get them to that phenotype so they can get to their recommendation. So we use these genotypes to make what we call a diplotype, which is going to show the two copies of that gene the patient has. So it may say star one, star one, which would be completely wild type. Maybe it says star one, star two, and that would be a normal function and a loss of function for CYP2C19, and then we would take that *1/*2 diplotype and assign a phenotype of intermediate metabolizer (reduced activity).

Speaker 1  09:29
So as another example, then let's say somebody had one copy of star 17, so increased function, and then one copy of star two or a loss of function. What would that combination do?

Speaker 2  09:38
And that's it's a great question, because this is an area where we're actually still doing kind of research and discussion today. And you may actually see some discrepancies if you look at the European versus American versus Dutch guidelines as to how exactly to interpret that. So we have an increased and a decreased function. So some may say that would result in a normal function. You're kind of averaging out in general for this. Studies we've seen so far, we see that the decrease in function has a more profound effect than the increase, so we still want to call that patient a an intermediate or slightly decreased function. Maybe that will change in coming years, maybe we'll get more information to clarify that. Maybe it differs between different medications. Those are things that we're going to look into more in the future, but current guidelines will help you with that information.

Speaker 1  10:22
I think that's an important again. Point that we'll make throughout this talk is the idea this is a dynamic field. This is a field that's continuing to evolve as more information comes out, and so it's important to kind of have have some baseline knowledge, but also to be up to date to understanding where these things may go in the next few years.

Dr. Sean Kane  10:38
Dr. Wake, you brought up a good point that I just wanted to clarify with the listeners. When I first learned about the star alleles with 'wild type' I assumed they had sequenced the entire CYP2C19 gene; that's not the case.

Speaker 2  11:02
Yeah, exactly. Dr. Kane, this is, this is actually a common misconception. That's something that we have to make sure we cover every time we talk about pharmacogenomics, is that star one is not actually a diagnosis. We'll say Star One is a an exclusion. We haven't basically means we didn't find anything. So then you're, by default, assigned star one. And this is important if you look at historical primary literature again, going back to CYP2C19, originally, those studies only looked for *2 and *3. They weren't looking for Star 17, and they reported their findings as patients with loss of function or patients without loss of function. Current studies look at additional alleles (eg. *4, *5, *6, *17). If older panels labelled a patient as *1, a modern panel might reclassify them — panels that don't test the relevant alleles can misclassify patients.

Dr. Sean Kane  11:58
This is actually really interesting, because I know with Warfarin pharmacogenomics, this actually happened in the sense that many of the early studies were done primarily in Caucasian patients, and then they ended up finding, wow, there's actually really relevant alleles in African Americans that just weren't well represented in the earlier studies, hinting that you probably really do have to have A really good base of patients that you're studying in terms of ethnicities to make sure that you're accounting for some of these alleles, right? Precisely.

Speaker 2  12:27
Dr. Kane, so some of the initial looks into warfarin were again focused on Caucasian patients. They are the largest body of almost all of our research just in terms of availability and willingness to participate in these studies. We have large studies looking at the Caucasian patient population. However, the most common variants for CYP2C9 in Caucasians are not the most common variants in African Americans. So if you do a panel that only looks at those Caucasian predominant variants, you will actually dose Warfarin less correctly using genetics than you would without using it. So the current recommendations are, if your panel doesn't cover these more wide variants, that you shouldn't even try and use the pharmacogenomics because you're going to more often harm the patient through inaccurate dosing than help them.

Dr. Sean Kane  13:12
So let's get back to CYP2C19. That one is relatively straightforward — patients can be ultrarapid, rapid, normal, intermediate or poor metabolizers depending on their alleles. And we've been talking about Plavix or clopidogrel. So where does that fit in, in terms of the knowledge of your phenotype as it relates to drug therapy?

Speaker 2  13:32
Yeah, so clopidogrel is a good place to start, because it has somewhat of a more straightforward reaction. Clopidogrel is a prodrug activated by CYP2C19. Patients with reduced CYP2C19 activity may have reduced activation of clopidogrel. So when looking at CYP2C19, we've had studies evaluating whether patients with loss‑of‑function alleles have worse outcomes. And there's been a recent study by it's called the Ignite network, implementing genomics. In practice, around 1800 patients had increased cardiovascular events in patients who were poor metabolizers or intermediate metabolizers given clopidogrel, versus if those patients were given different agents, because the pro drug is less activated by that, less enzymatic activity, then you get less active moiety less therapeutic benefit.

Dr. Sean Kane  14:23
And usually we think about metabolism as getting rid of the drug. In this case, as you said, it's a pro drug, where we're actually making the drug active, and if you don't do that as well, clearly the drug isn't going to work as

Speaker 2  14:33
well, right? Exactly. And that is a somewhat of an interesting point in that the clopidogrel and things like some of the pain medications we'll talk about in a second, are pro drugs. So whenever we're talking about decreased versus increased metabolism, the it's basically going to be a switch of what the assumed effect is. And some reports might not make that quite clear to the provider that decreased activity for CYP2C19 doesn't mean increased drug remains — for prodrugs it can mean reduced activation, the opposite effect.

Dr. Sean Kane  15:00
Sherman, I know in your neck of the woods there's some genomic things related to the drugs that you use, right, right?

Speaker 1  15:05
So one interesting one that I know there's a fair amount of information on both with CYP2C19 and CYP2D6 is our tricyclic antidepressants. And so this would be a more straightforward one. The medication itself is active. And so, for example, amitriptyline is methylated by CYP2C19 and converted to nortriptyline, which is also an antidepressant, but has a totally different profile to where it's less anticholinergic and generally more preferred in the elderly. And so we have the ability to say, if you have a certain amount of CYP2C19 metabolism, you'll convert more to nortriptyline more rapidly, versus shunting via other pathways.

Speaker 2  15:42
And I think that's a thank you, Dr. Schuman, that's a great point, because CYP2C19 and CYP2D6 for the tricyclic antidepressants are more straightforward and that it's back to our normal understanding of metabolism, but it actually introduces another problem we see sometimes with pharmacogenomics: the concept of multiple genes all impacting the same medication. So instead of looking at just clopidogrel and CYP2C19, it's kind of a one-to-one, this drug-this gene — well, now we have two genes that could be potentially impacting the same medication. So you start getting into these graphs, these charts, of how is the overall effect if we have an increase and a decrease, where does that come down? And that can be somewhat complicated again, for the provider to interpret, so it's best to try and present that information in the most convenient way for them.

Speaker 1  16:29
Yeah, I think that's something we can get into. We'll get to the end is how is this presented? But first I'd like to talk a little bit more about CYP2D6. So what makes CYP2D6 different in terms of how you evaluate an individual's phenotype compared to, say, CYP2C19?

Speaker 2  16:43
Sure — CYP2D6, I will say, is somewhat of a beast on its own because of the amount of research and the number of named star alleles. There's over 100 and a dozen or so named star alleles, and each one of those can have a few variants. There's dozens and dozens and dozens of Rs, IDs or single nucleotide polymorphisms, and all these things increase the complexity. You're not going to have a panel that necessarily looks at all of those things, but they're out there, and they're possible. But the important thing for CYP2D6 is that you can also frequently have gene duplications and numerous variant alleles; this increases complexity beyond what we see with CYP2C19. For CYP2D6 we calculate an activity score: loss‑of‑function alleles score 0, decreased‑function alleles score 0.5, normal alleles score 1, and duplicated alleles increase the total — then you add the haplotypes to get the overall activity score.

Speaker 1  17:43
score, right? So for example, I believe that if an activity score is greater than two, that would be considered Ultra rapid. Am I correct Exactly?

Speaker 2  17:51
So greater than two? Basically, normal is ~1 per allele, so you'd exceed AS>2 only if you have gene duplications — for CYP2D6 that means extra copies of the gene in some individuals. Again, these vary by patient population anywhere between two to five, up to maybe 10% based on a few studies, have had increases in copy number here. So maybe they have three or four copies of this gene. And if you have three or four recipes all making the gene at the same time, you're gonna end up with more copies, more activity.

Dr. Sean Kane  18:21
So just to clarify: for CYP2C19 (Plavix) you have two alleles (one from each parent). For CYP2D6 you may have extra copies (gene duplications) that increase the total number of functional gene copies.

Speaker 2  18:43
So for CYP2C19, we really don't see this duplication. Typically, you're going to have your one copy from mom, one copy from dad. But for CYP2D6, you could have potentially three to four copies on one chromosome and maybe one copy on the other. It actually can get a little bit complicated, because one of the star alleles we look at (*5) is a deletion of CYP2D6. So now we're looking at anywhere between the possibility of zero copies and multiple copies. And you can have instances where the patient has two copies on one chromosome and zero copies on the other. So this is why we do this activity score, just because it gets very complicated very quickly to figure out how much activity are we actually ending up with.

Speaker 1  19:19
So I think one example would be to say that an individual notes a CYP2D6 diplotype, consisting of one copy of *2 and one copy of *5. In that case *5 indicates no function and *2 indicates normal function, giving an activity score consistent with a normal metabolizer.

Speaker 2  19:36
So that's a kind of a variant of normal function. So we it's slightly different, but it has the same amount of activity, and that star five again, being that gene deletion really isn't bringing any thing to the party. So that'd be an activity score of one, and currently we'd call that patient still a normal metabolizer. One important thing to talk about is we did mention that these things are always changing, and there's been discussions and literature in the work as to does this activity. Score calculation currently best capture the final amount of activity in the next one to two years. The exact calculation may change. Some of the variance may change in activity score. So this might not be the gospel truth as it stands, but currently, the idea of this calculation is the most important part double check if there's been any changes in the activity score calculation at the time you're making that therapeutic decision, but the process and the idea should be the same.

Dr. Sean Kane  20:25
I think this is a really good example of the importance of both the clinical pharmacist with knowledge of this, but also the electronic clinical decision support tools that are used to help even the clinical pharmacist interpret what could be, you know, multiple different versions of that gene and how complex that can get. It's not as simple as a lab test to check 'Do you have X' and then you're done. A lot is going on with CYP2D6, right? Exactly.

Speaker 2  20:51
Dr. Kane, so especially for CYP2D6, we're really looking at a need for clinical decision support. And clinical decision support are tools in the electronic health record, or in the prescribing system, or really anywhere that you can get someone to open up something on a computer that allows them to access this information in a more provider or clinician friendly manner. So we don't want to be showing providers raw rsIDs. We want to give them final recommendations — tools that interpret CYP2D6 activity scores and their clinical effect will become increasingly important as more information emerges.

Speaker 1  21:28
So we kind of already talked about tricyclic antidepressants. To recap: these can be metabolized by CYP2C19 as well as CYP2D6. If we have different activity in those pathways (poor metabolism in one, extensive in the other), the metabolite ratios and clinical effects can change substantially.

Speaker 2  21:50
And I think one thing that I mentioned there is when we talk about extensive metabolism, extensive metabolism is the old terminology we used for normal metabolizer. And the reason we've somewhat changed, it was because there was at some points confusion over whether extensive was the same as rapid or was the same as normal. When we talk about extensive, we may be using that interchangeably with the new normal metabolizer discussion.

Dr. Sean Kane  22:13
So of course, when we have a patient, let's say that is this really, really poor metabolizer, we would assume that maybe they would have increased TC exposure, therefore high risk of side effects. And if they're more this ultra rapid metabolizer, maybe they won't have much drug concentration and they won't get a good therapeutic effect. Is that kind of the rationale for

Speaker 2  22:32
even the testing? Exactly. So thinking that if we have decreased metabolism, we're basically going to have a buildup of medication, increased blood level of that medication, and with that increased blood level, we're expecting increased side effects the same way you would expect increased side effects if you doubled the dose in the same vein, we're expecting decreased efficacy if we're metabolizing that medication too quickly, because the patient's going to have a lower effective dose of the medication. So in general, those are the expectations whenever we're changing these metabolism activity scores. And I

Speaker 1  23:01
Think to flip it on its head: codeine is a prodrug and must be activated to morphine — and that's mediated by CYP2D6, correct?

Speaker 2  23:15
Precisely — CYP2D6 metabolizes codeine into morphine. This is important because ultra‑rapid CYP2D6 activity can produce excessive morphine concentrations. Rapid metabolizers. There's been case reports of both children and breastfeeding mothers rapidly metabolizing this codeine into morphine, having a very high effective dose of the active moiety, again increased therapeutic efficacy at the cost of very increased adverse effects. And there's been case reports of significant morbidity and actually mortality in these patients who didn't know that they were effectively taking a increased dose for themselves.

Speaker 1  23:47
So again, in our current climate of concern about, you know, opioid overdoses and misuse, to know that there are some individuals that truly would not be able to activate codeine to morphine, who may not get an analgesic effect, and who may say morphine works better for me than codeine. And on the flip side, individuals may be at risk of overdose if they rapidly convert codeine to morphine.

Dr. Sean Kane  24:11
It's important to note that this is actually a boxed warning for codeine products. The FDA recognizes CYP2D6 polymorphisms and the potential harm that they could cause. And codeine products have that boxed warning for that reason.

Speaker 2  24:24
And we're seeing more and more either boxed warnings or just FDA labeling about pharmacogenomics as companies start doing these types of tests, it is important to read the exact wording, because the wording of the recommendation may differ slightly, but it initially appears based on how the FDA label is presented, but it's nice that we're seeing that information be now pushed all the way into the kind of the package

Dr. Sean Kane  24:45
insert, absolutely.

Speaker 1  24:46
All right, so again, we kind of talked that. So in general, Star One is referring to that the wild type, which is going to be the most common, generally, for for individuals in a population. Is there ever a case where star one is not the most. Common Yeah,

Speaker 2  25:00
Thank you, Dr. Schuman — that's actually a great question, because one of the most common questions I get from patients that come into the clinic and talk about their pharmacogenomics reports is, why does it say I'm a poor metabolizer for this gene? And why aren't we talking about that? That gene is actually CYP3A5. CYP3A5 affects tacrolimus metabolism; in many populations most people are poor expressors of CYP3A5. So the majority of the population breaks it down more slowly than average. And that doesn't quite make sense when we talk about it that way. But based on the way the studies are done, based on the way the actual gene is historically been studied, most patients are going to be classified as a poor metabolizer. The important thing there is not to modify the dose of the medication because it's already dosed, assuming that they're going to have that decreased metabolism. In that case, a normal metabolizer effectively has higher activity, and we would want to increase that dose.

Speaker 1  25:56
So I think another topic worth covering is HLA‑associated adverse effects. Dr. Wake, what can you tell us about HLAs and how they impact medication use?

Speaker 2  26:12
Thank you. Dr. Schuman, so the HLAs may be one that the more common patient may actually experience most frequently out of these because some medications actually have labeling to require testing for these before the medication is prescribed. The reason for that is because these genetic variants are actually associated with some of the more profound, more significant adverse effects. So the human leukocyte antigens or HLAs, are immune function genes that can potentially, if you have the right variant and the right medication increase the risk of these severe cutaneous adverse reactions, anything up to Stevens‑Johnson syndrome, a very profound, serious reaction, and

Dr. Sean Kane  26:51
typically we're talking like ridiculous rash, right? So not just a normal rash, but a life threatening rash for a patient, right? Exactly.

Speaker 2  26:58
So a life threatening, almost full body immune mediated burn. And these are very profound, very scary, and it's very low percentage of the population. But if you have that variant, the increase in risk is significant. And this is this is definitely somewhere where we have to look at matching the test to the patient population, because the rate and the proportion of these variants varies hugely by ancestry and by patient population — some are very rare in Caucasians (near 0%) but can be up to ~5% in Southeast Asian patients. That's a very significant change when we're looking at the risk of the event. Maybe you could give

Dr. Sean Kane  27:35
us an example of some typical medications that would benefit from HLA testing. Sure.

Speaker 2  27:40
So abacavir may be one of the most common because it actually has labeling requiring testing before the medication is initiated. Some other common ones that this may be encountered for are things like allopurinol as well as carbamazepine, oxcarbazepine and phenytoin. Phenytoin is both broken down by CYP2C9 but also has potential to interact with one of these HLA variants. Now, the variants themselves have somewhat complicated nomenclature. The important thing to know is that these reactions are strong enough that if it's suspected, you might want to choose another medication just for the possibility of avoiding that reaction.

Dr. Sean Kane  28:15
And what we're actually testing here, as you mentioned, was a variant of the HLA does that mean that the patient does not have that human leukocyte antigen or they do have it? Is this a different one? Or what is that test actually telling us?

Speaker 2  28:30
Yeah, and the results for those tests, as opposed to being reported as metabolizer status or as these diplotypes, you may actually see these reported just as positive or negative, because the important thing there is whether or not the patient really has that variant we're really testing for. Is this there? Because if it is there, it is a big flag saying the reaction may happen. We don't necessarily need to know the variance and activity. It's really just a yes or no, a positive or negative,

Dr. Sean Kane  28:54
which is great, because that makes it a little bit simpler to understand that result versus CYP2D6 — which, as we mentioned, is more complex.

Speaker 2  29:04
So depending on the gene and the the reaction we're expecting, the the way the results are reported can be fairly different, and those might be easier for providers to understand initially. And maybe it's the best place to start if you're looking at starting pharmacogenomics, just because it is such an easy yes or no question.

Speaker 1  29:19
So we've really talked about a number of these different metabolic pathways and human leukocyte antigens. So how do clinicians, how do we get it into the hands of providers? Is there a governing body who writes the guidelines? Is it really up to the individual organizations, such as, you know, American College of Cardiology, American Psychiatric Association, who really makes these determinations?

Speaker 2  29:38
So currently, I believe the best guidelines to go to right now are going to be the Clinical Pharmacogenetics Implementation Consortium, or CPIC (cpicpgx.org). Those provide both a great background on the expected reaction, a good evidence base for why the reaction is suspected, and also the therapeutic recommendations based on the diplotypes and phenotypes. And these guidelines aren't in the 60 to 70 page we might expect for some of those ACCP and other guidelines, these are typically in the six to seven page range. So a pretty quick read if you're interested. Now, the one caveat there is we have a lot of suspected pharmacogenomic interactions. We only have a limited number of these guidelines because they are based on the most well validated evidence. So if the reaction you're looking for is potentially newly discovered, it might not have these guidelines currently available. So I think

Speaker 1  30:27
Let's walk through a couple of examples. I think we already talked about the 2016 guidance on adjusting tricyclics based on CYP2D6 and CYP2C19 phenotypes. What about statins? One transporter of interest is SLCO1B1, which can alter simvastatin exposure and predisposition to myalgia — what do the guidelines say?

Speaker 2  30:55
Yeah, and that's a good point, because we've been talking a lot about drug-metabolizing enzymes, but we also have variants associated with transporters and receptors as well. For SLCO1B1, patients with decreased function of the transporter have higher simvastatin exposure and higher risk of myopathy. In such cases consider an alternative statin (for example, atorvastatin or rosuvastatin) or a lower simvastatin dose.

Speaker 1  31:35
And what I found interesting in the guidelines is they provide numbers: for each non‑functioning SLCO1B1 allele the risk of myopathy with simvastatin increases (≈2.6× at 40 mg). At higher doses the risk rises further — a clear, clinically meaningful effect.

Dr. Sean Kane  32:02
data and bit of extra data. And I think it kind of makes sense that we have an organization that's publishing these guidelines that is specific to pharmacogenomics, as opposed to the American College of Cardiology, because we're really at this not only like novel area where it's kind of emerging data, but also you have to have an expertise in pharmacogenomics to just understand some of these nuances. What is an allele? Things like that. So I think it makes sense that one organization is encompassing many different practice types, not just a cardiology Association. With that said, though some of the guidelines are moving towards a mention of using pharmacogenomic testing, typically saying, do it or don't do it, as opposed to, if they have this and do this, they have this other thing, then do this other thing, which makes sense too. Yeah.

Speaker 2  32:45
And I think one of the interesting points there is that the CPIC guidelines are great at providing therapeutic recommendations — they translate diplotype to phenotype and say what should be done. But they actually don't make recommendations in general as to who should be tested. So the opportunity for a liaison between traditional specialty guidelines and CPIC is deciding which patients should be tested. Because that's still somewhat of a missing piece for some of these conditions.

Speaker 1  33:17
So one of the things again we look at, and we've talked about, is where we have the guideline based approach, but only also, and how you're doing your testing your clinic. But also we can't an elephant in the room, the idea that that direct consumer products now available for individual consumers to be able to go home and do pharmacogenomic testing on their own. And I think this is one of the there's, there's great power and benefit in that. But the same time, I'm imagining there's some concerns with, with that more indirect approach. What would be some concerns with, with having just some patient I just go ahead and get the testing done on your own, and we'll we'll deal with the results later. Yeah.

Speaker 2  33:52
So there, it's a great opportunity. I love that there are so many patients who are out there who are interested. There's clearly a market share of these patients who want more genetic information. The concern we have, and the concern we have when we do any kind of direct to consumer testing or marketing, is that patients may look at the results in a vacuum, and they may start making changes that may appear well validated by the results, but potentially might not be the optimal therapeutic response. I believe the FDA announcement actually included special wording that patients should not use the test results to stop or change any medications, and instead, they should use the test results to kind of begin a dialog with their provider. So that is one concern we have, that patients may just make these changes in a vacuum. The other concern is that there's different panels available depending on the patient populations, and the 23andme panel may be good for a certain population, but it might not be capturing the variants that we need for a specific patient population. Maybe you're an African American patient, but the test is geared more towards Caucasians, and those types of nuances are difficult to assess for kind of a product that's being dispensed nationally.

Dr. Sean Kane  34:56
I think to highlight on that not using it in a vacuum for the patient. Also, the clinician also should not be using it in a vacuum, for example, the concept of a drug interaction, right? So even though you might be a CYP2C19 ultrarapid metabolizer, if a drug interaction inhibits that enzyme the situation changes. And you have to account for that, among many other variables, cost, side effects, of the patient, things like that. Again, focusing that this is one data point and many other data points that the clinician should be considering.

Speaker 1  35:29
Yeah, there's, I mean, again, we think about our basic assessment tools of pharmacists looking for hepatic function, a renal function. If somebody's kidneys are shot, if you know even it says their normal metabolizer, we know their kidneys can't filter the drugs. We've got to we've got to take that data and still use that

Speaker 2  35:44
too precisely. And I actually was just listening to your great episode on SSRIs in renal failure, and it's really the same type of discussion we'd have there. Because pharmacogenomics is a it's a new It's a fancy tool, but it's really just a tool. Pharmacists and clinicians should use it the same way they use hepatic and renal function. If the patient's been on the medication for five years and the result says they have an increased risk of side effects, well, they've been on it for five years and they haven't been complaining about side effects. Similarly, if the results come back and say that these medications are bad for the patient, these medications are good for the patient, but the patient doesn't have a desire to change them, or the patient's already tried it in the past and was allergic, or numerous different variables. We can't just take the results as gospel and move forward from there.

Dr. Sean Kane  36:33
So Dr. Wake, I really appreciate your time coming on today. I know that this is a hot topic that many students are really excited about, given it's a growing field and job opportunities may open up in the future. So what are some key concepts we want to leave our listeners with about where the field of pharmacogenomics is now and how to apply it in everyday practice today?

Speaker 2  37:02
So I think one thing to think about is that pharmacogenomics is going to continue to expand and grow, and what it is today shouldn't be what it is in five years. So the positions available in five years may be completely different from what they are right now. And like I said a little bit earlier, the goal is for pharmacogenomics to as cool as it is, it should become a mundane tool that pharmacists and providers use. So the eventual goal is that we don't need dedicated specialists dispersed everywhere. We have this type of training potentially available to everyone. So positions that help that happen, positions that improve those clinical decision support tools, are going to be paramount to kind of realize that future.

Speaker 1  37:41
And I think a second point is to look at the overview — by no means exhaustive — of topics we covered: Phase I metabolizing enzymes (CYP2D6, CYP2C9, CYP2C19) and how genetic variation alters drug response (including prodrugs). We also covered drug transporters (e.g., SLCO1B1) and HLA‑associated immunogenic reactions.

Dr. Sean Kane  38:14
And then for me, I think again, it's so important that clinicians appreciate that this is a data point in many data points, and you can't take off your clinical pharmacist hat or your clinician hat and just say, well, that's what the genetic test said, and that's it. You have to consider many other patient specific factors, as we would normally do when considering drug therapy for a patient. Again, this is that one data point among many other data points to consider. So I think that wraps up today's episode quite nicely. If you want to see show notes, including links to some of the resources that we've talked about, they're available at HelixTalk.com (Episode 91). We're also on Twitter at @HelixTalk. Keep those five‑star reviews coming in iTunes — they help. Dr. Wake, I really appreciate your time again with that. I'm Dr. Kane, I'm Dr.

Unknown Speaker  39:02
Schuman, and I'm Dr. Wake. I had a great time. Thank you. In the

Unknown Speaker  39:04
absence of Dr. Patel, everyone out there study hard.

Narrator - Dr. Abel  39:09
If you enjoyed the show, please help us climb the iTunes rankings for medical podcasts by giving us a five star review in the iTunes Store. Search for HelixTalk and place your review there

Narrator - ?  39:20
to suggest an episode or contact us. We're online at HelixTalk.com thank you for listening to this episode of HelixTalk. This is an educational production copyright Rosalind Franklin University of Medicine and Science.