Drug-Gene Testing Opening
Andrea: Okay. So we’re going to talk about, if it were not for the great variability among individuals, medicine might as well be a science and not an art. We have looked at treating patients individually as a black box without really understanding what it is that we’re doing or why people respond the way they do.
I’m going to give you a little bit of a clinical scenario. My son was born in 1989 and I was given Percocet for an episiotomy. Interesting enough, it didn’t give me any pain relief at all, but I thought, well, you know, first child, whatever. In 2010, I underwent an urgent dental procedure and they gave me Percocet, and I took six Percocet. I took them two every two hours, opioid virgin, and had no relief. I didn’t get nauseated, I didn’t get dizzy, I didn’t get light-headed, and I, certainly, didn’t get any pain relief.
And I lecture on this stuff, and so I said… Well, I had to go back the following week for another procedure and I said, “Do me a favor, humor me, just give me a couple of Darvocet. Here’s the bottle of Percocet back because it didn’t do me any good.” And I took one Darvocet and I couldn’t get off the floor. So I got genetically tested, it showed that I was 2D6 deficient.
And so then, a little while later, my son, who was in Florida and now up in Alaska, was scheduled for wisdom teeth extraction. And so like any good pain doctor mother, I called the surgeon and said, “You know, what are you gonna do? What kind of anesthesia, whatever postoperative medicine?” And he said, “Well, we’re gonna do hydrocodone.” I said, “That’s fine, that’s fine. Just realize that I’m genetically proven 2D6 deficient. It’s conceivable my son is also 2D6 deficient and hydrocodone doesn’t work in people who are 2D6 deficient.”
And, of course, he blew me off. My son took 10 hydrocodone, got no relief, went back to the surgeon, and got thrown out for drug seeking. So I want you to keep this scenario in mind as we start moving forward and try and talk about some of the variability in analgesia.
Clinicians who treat pain have always known that the response to opioids varies widely. It has nothing to do… The little lady might need 15 pills, and the big guy needs 1, and he’s on the floor, and it has never really made any sense. It’s not per milligram, not milligram per kilogram, not sensitivity or not sensitivity.
We’ve said the [inaudible 00:02:39] was bioavailability. Remember, the FDA says that a generic medicine can be 20% more or 20% less and still be considered equivalent. Now, when you’re talking about a blood pressure medicine, you and I can’t…I mean, the patient can’t tell whether or not it’s made a difference. All you can tell is whether the blood pressure changes. But with pain medicines, patients can tell.
And we also, though, think that there have been some differences in pain stimuli. Well, you know, this doctor was very rough when he was doing his injection, this doctor was very gentle, and that’s why there was a difference. But we also are now recognizing the genetic makeup may be likely a strong factor.
Pharmacogenetics in Pain Management
So when we talk about pharmacogenetics in pain, there are a couple of different types of genetic issues. One of them is this concept that there’s a genetic contribution of a variety of different pain types. So if genetic basis underlying how pain is expressed, there’s varying mechanisms of nociceptive, neuropathic, and visceral pain, and then the potential exists for differences in response, and that can be genetic related.
But there’s also this genetic influence on drug effectiveness and safety. So when we look at how patients respond to a stimulus, there’s a whole spectrum of pain sensitivities. Given that this is the same stimulus, patient A and Patient B and Patient C may respond very differently, and that has a genetic makeup.
We know that the receptors for pain, we know can differ genetically, we know that the connected tissue between patients can be different. But what we’re really gonna spend our time on today is looking at this concept of pharmacokinetics, pharmacodynamics, and pharmacogenetics, and how it mix together.
So a drug action to a patient metabolic status or the ability to metabolize certain drugs affects the drug’s pharmacokinetics. And so a patient with impaired metabolism may be unable to activate a prodrug to its active form, like codeine going to morphine.
The patient’s ability to respond to a drug is what we consider pharmacodynamics. And so if you’ve got a receptor that doesn’t work, it doesn’t matter how much medicine you give that patient, they’re not going to have an effect or they’ll have an ultra-effect.
So when we look at the pharmacogenetics of analgesia, there are several ways that genetics could influence that analgesic response. One of them is the drug metabolism enzymes. So we’re gonna spend our greatest time on that issue. We’ve got drug transports, opioids or other pain receptors, and then structures that are involved in the perception of pain.
So we’re gonna spend probably the most amount of our time looking at how opioids are metabolized, and I will tell you that this can be fairly complicated, and I’m going to, at the end of my lecture, have my email address. Email me, I’ll be glad to give you a copy of the lecture, I’ll be glad to give you some of the resources that are in there as well. So when we look at the opioid primer…
Man: It’s not working?
Andrea: Not working. Not yet. So…
Man: [inaudible 00:06:09].
Andrea: Best laid plans. Well, good try.
Man: Good try, though. Good try.
Andrea: Good try. It was a good try. Okay. So the cytochrome P450 system has a whole variety of enzymes, some 50 isoenzymes, most of which, though, are listed here. And when we look at the distribution of them, we see that the majority of them are 3A4 and 3A5, 2D6 is in yellow, and that these can have what we call a polymorphous. Now, polymorphous means that there can be multiple genes that will change between patients and it will change the way these drugs are active.
And also, though, these drugs can be induced, they can be stimulated, you’ve got inflammation, age, sex, that can influence them, and… Thank you. I’ll try that. There we go, wonderful. And so the drug… This leads to a variety of drug interactions. So we have a couple of concepts that we have to take into mind.
What is a substrate? It’s any medication metabolized by that enzyme. There are the inhibitors, medicines that slow down the metabolism. And, for instance, if that happens, you could either have excessively high level, if it’s an enzyme that excretes a drug, or you can have a standard effect, or it can have a related to toxicity, so that you have more of a toxic effect. And then we have inducers, medicines that boost the metabolism, speed up the metabolism of it. And that can result in an accelerated excretion or an accelerated conversion to an active form.
Metabolism of most of the opioids
I love this slide. Look, this is the metabolism of most of the opioids. Now, it’s very intimidating when you first look at it, but I’m gonna break it down into little pieces for you and show you how it works. So codeine is metabolized by 2D6 to morphine. Codeine is a prodrug. It has no activity until it’s metabolized to morphine. So when you give somebody Tylenol No.3 and they are either 2D6 deficient or you’re giving them a medicine that inhibits 2D6, what you’re giving them is Tylenol because codeine has no activity.
Morphine is metabolized by UGT2B7 to an M6G, which actually has analgesia with it, and M3G or M3-glucuronide that’s hyperalgic, and we’ll come back to that in a minute because this becomes very important when we talk about opioid hyperalgea.
Then we’ve got hydrocodone and that hydrocodone is also a prodrug. It has very little effect until it gets metabolized by 2D6 to its active form, which is hydromorphone or Dilaudid. Now, how many you ever heard a patient say, “The only medicine that works for me is that D word, Dilaudid.” And we said, “Drug seeker.” But if they’re 2D6 deficient, that’s actually true. A hydrocodone does not work for them.
Oxycodone, [inaudible 00:09:32] is also mostly a prodrug, though oxycodone itself does have some activity. It’s metabolized by 2D6 to oxymorphone, which is its active form. It’s also metabolized, though, by cytochrome 3A4 to noroxycodone, which is its inactive form. And so you get this competition between those two enzyme systems.
Tramadol, fascinating medicine, as you remember, it’s a weak opioid, but it’s also a serotonin-norepinephrine reuptake inhibitor, which gives a very good effect for neuropathic pain. It’s a prodrug, it’s metabolized by 2D6 to its active metabolite, O-desmethyl-tramadol, which may be up to 40 times more potent than the primary drug. It’s also excreted by 2D6.
So when you have somebody who is 2D6 deficient and you give them tramadol and it doesn’t work, what do they do? They take more. Well, as a 2D6 deficient person, they can’t get rid of it, and so those are the patients that are at a particular risk for toxicity and seizures because they’re taking more because it doesn’t work and they’re not getting rid of it.
Actually, the metabolism for tramadol is a little more complicated than that because what happens is the tramadol comes into the cell, it gets metabolized by 2D6 to that O-methyltransferase. Okay. That’s the O-methyl-tramadol. But it also gets metabolized by 2D6, and 3A4, and 2B6, and UGT, and ABC2 to active drugs as well. So it’s a wonderful medicine, one of my favorites, but it is one that does have multiple, multiple drug interactions.
Fentanyl is an active drug, but it is metabolized by 3A4 to its inactive metabolite. So anything that is inhibiting 3A4 will lead to an accumulation of fentanyl and potential overdoses.
And then methadone. Truly, if I only have one medicine to take on to a deserted island, it would be methadone. It works better than any other pain medicine I’ve ever seen on neuropathic pain. And my patients tell me that once they’re on a stable dose, they don’t feel like they’re taking a pain medicine. They just don’t hurt. However, it’s probably one of the most dangerous medicines we use because of the metabolism issues.
So methadone is metabolized to EDDP, which is its inactive compound, and it has a secondary metabolism by 2B6 and also by 2D6, and we’ll talk about those in just a minute. The methadone metabolism is always variable. My patients are… It stuns me how many people come to me on methadone with no understanding of drug-drug interactions. They’ve never been told by their physician to avoid 3A4 medicines because anything that inhibits the 3A4 enzyme will lead to a rise in methadone levels and potential toxicity and death.
And so this is… They’re also modified by 2D6, 2B6, the OPRM1 opioid receptor, which we’ll talk about in a minute, and the ABCB1 drug transport, which we won’t talk about. There’s only a limit to what we can do in an hour.
So why is this important? Well, it turns out that when you look at the metabolism of these medicines, the populations can be considered a extensive metabolizer, which is really a bad term. It’s really a normal metabolizer. So EMs are normal and they’re actually getting ready to change that to NMs because it just…EM doesn’t make any sense. But you have intermediate metabolizers where you’ve got one good copy of a gene and one bad copy of a gene, so it’s sort of works halfway. You’ve got poor metabolizers which have two poor copies of the gene. And then you’ve got the ultra-rapid metabolizers, which may have multiple copies of good genes, and so they metabolize much more quickly.
And if you look at this, you’ll see that the vast majority of people are the normals in this column, but you will see that 10% of Caucasians are poor 2D6 metabolizers. That’s not an insignificant amount. And these change by populations, by the racial makeup, and there can be significant differences with different populations. Someone who is Chinese and somebody who’s Ethiopian may have totally different responses to the same dose of medication.
And with this concept, comes the fact that if a standard dose was two pills, you would give two pills to your standard patient. However, if they’re an intermediate, they probably only need about a pill and a half. If they’re poor metabolizers, they probably only need one pill. And if there are extensive metabolizers, they may need three to get the same therapeutic response. And this becomes really important when you start to look at some of the deaths that we’ve seen with medicines, and this one was just recently.
A 5.5-year-old boy had his tonsils taken out, had 20 milligrams of tramadol. That’s a small amount, right? But he ended up being found comatose and it turns out he was an ultra-rapid metabolizer. He metabolized the tramadol extremely quickly to its active form, and so he had more of an effect. The consequence has been that people are now being told that they should be giving tramadol to kids instead of saying, “Let’s test the children before we give them the medicine to see whether it’s a problem.”
We do know that these drug interactions are tremendously… Patients who were taking 3 to 5 drugs had 29% risk of interactions. But I see patients taking 11 or 12 or 13 medicines. Every one of their doctors has given them a different one without looking at the list and EHRs have made that so much more of a problem, because you get this list like this and nobody goes through and march them out and says, “This is, you know, reconciliation.” Patients who were taking 11 or more had a 96% risk of interaction, and only 1% of the patients were aware of the potential for these drug-drug interactions.
So we can have a couple of possible issues. We’ve got drug-drug interactions, DDIs, drug-gene interactions, DGIs, and drug-drug-gene or DDGIs. It starts to sound like a epitaph of some kind at that point.
So how common are the drug and gene interactions? They looked at a prevalence of a sample of 1,143 patients, a total of 10,300…sorry, 1,053, potential interactions. The DDIs, the drug-drug interactions were about 60% of them. But when comparing those to the DDIs alone, the drug-genes and the drug-drug-genes increased the risk of clinically significant interactions by 50%. So you went from having 60% of them and then another 50% on top of that. So almost 90% of these patients had the potential for drug interactions.
So we’re gonna go over a couple of the specific genes. We’ve mentioned some of them already. Now, what I love about this is that I used to be the medical director of a urine and genetic testing company. And so at the Florida Society of Interventional Pain Physicians, one year, we tested 40 pain physicians, asymptomatic, we just did their genetic testing, and these are the results that we came up with.
So here you see the poor metabolizer. He will not go named here, but you do know her. And then we compared to a rapid metabolizer here. So up to 25% of the commonly prescribed medicines are metabolized by 2D6. And that is antidepressants, antipsychotics, opioids, antiarrhythmics, and tamoxifen, and that will come up again. Poor 2D6 activity has poor conversion of prodrugs to their active or more active metabolites and ultra activity leads to overexpression and the potential for overdose.
And you can see it here, this was a study where they looked at hydromorphone concentrations in poor metabolizers. So this is the active drug, remember, compared to extensive metabolizers. So the poor metabolizers do not make the active drug over multiple timeframes and multiple doses.
And so when you look at this pain sensitivity and the clinical response to weak opioids, poor metabolizers may undergo no metabolic formation, leading to inadequate analgesia, which I can promise you happens clinically, and conversely, the ultra metabolizers may have quicker analgesic effects, but they may have higher mu-opioid related toxicity. Those are the ones who take a pill and they’re sick as a dog, they’re throwing up their toenails with just a simple amount of medicine.
And so then the literature suggests the potential usefulness of the determination is in elucidating among the adverse events and preventing subsequent inappropriate selection or dosing of medicines like codeine or tramadol.
It’s not just the opioids, though. This was a study looking at non-opioid metabolism with antidepressants. And on here, you’ll see, this is the drug dosage levels that were necessary to reach a therapy…the dose that was necessary to reach a therapeutic level in poor metabolizers, in intermediate metabolizers, in the normal metabolizers, and in the ultra metabolizers.
And so you can see that there can be a tremendous difference in the metabolism levels and the drug doses needed between some of these levels. And just to show it in a tabular form here, let’s just look at Doxepin. If you’ve got a poor metabolizer, they should only get about 36% of a standard dose compared to an ultra metabolizer that may need 172% of a standard dose, as opposed to sertraline, which is not 2D6 metabolized, where there is no change based on 2D6 stats.
And we’ve seen this. We’ve seen this in medicines like amitriptyline, where you’ll have…give somebody…1 person will get 10 milligrams and they can’t wake up, and you’ve given somebody else 150 milligrams, and they’re not getting any effect. So in psychiatry and geriatrics, 52% of psychiatric, 49% of psychogeriatric, and 46% of geriatric patients use at least 1 drug metabolized by 2D6. It’s a huge issue. And the 2D6 has a special relevance in psychiatry and in the treatment of depression and schizophrenia.
Other ones, dextroamphetamine or Adderall is completely dependent on the 2D6 system for excretion. So poor metabolizers are at risk for toxicity as opposed to Ritalin which doesn’t use 2D6 for its metabolism. So here, you’ve got a patient, it makes a big difference which drug you’re using, especially if you’ve got the auto substitutions with your formulas.
Tamoxifen, I told you we’d come back to this. It’s metabolized to its active form via 2D6. If somebody is a poor 2D6 metabolizer, they may not get adequate levels of active tamoxifen, and that puts them at risk of a recurrence of their breast cancer because they did not get adequately treated even though they were taking the medicine every day.
And this is not just people who are genetically involved because we’ve got lots of medicines that are 2D6 inhibitors. Turns out, oh, goodness, Paxil and Prozac are huge 2D6 inhibitors. Go ahead and eat. Eat it. But Lexapro and Celexa are not. So if you’ve got somebody who’s on Tamoxifen and they’re a normal metabolizer, and now you put them on Prozac because of the depression they have for their cancer, all the sudden, you potentially would put them at risk for recurrence of their breast disease.
All currently used 5-HT antagonists are metabolized by 2D6 and 3A4 except for granisetron, which is primarily metabolized by 3A4. So it turns out, patients who are 2D6 ultra metabolizers increase incidence of post-operative nausea and vomiting because they get those medicines metabolized too quickly and it doesn’t stay in their system. They need bigger doses or they need to be looking at changing the antinauseants.
So we look at 3A4 and 2B6. So, again, poor metabolizer 2B6. So it turns out 2B6 is responsible for the clinical response of metabolism and clearance of methadone. And that makes a big difference because the slow metabolizers, they looked at 23 methadone-related fatalities, there was a significant correlation between the high methadone blood concentrations and the slow 2B6 metabolism. And so 3A4 will cause a problem with metabolism, but so will 2B6.
And this is what you end up getting. So you have the methadone arrhythmias, you get an occasional PVC, and then the next thing you go you’ve got [inaudible 00:23:23]. And this is strongly associated with the 2B6 isoenzyme. So if you’ve got somebody on methadone, it might be good to know whether or not you need to watch them more carefully and monitor their QTC intervals much more closely.
So they looked at the drug interactions modulating 2D6 and 3A4 major effects on oxycodone as well. And so it turns out, 2D6 activity was correlated with oxycodone experimental pain assessment. In those patients who were rapid metabolizers, they had increased pharmacodynamic effects, whereas the cold pressor test and the pupil size were unchanged in the poor metabolizers.
In other words, if you were looking and giving somebody…sticking somebody’s hands and eyes, their pupils would change and they would say, “Oh, that hurts,” unless they were poor metabolizers. If they were rapid metabolizers, you saw a change, you saw the pupils constrict, you saw a reduction in the pain. If they were poor metabolizers, you didn’t see that.
And so the 2D6 blockade reduced subjective pain threshold for the oxycodone by 30% and the response was similar to placebo. So it was the equivalent to giving the patients a placebo, a sugar pill, which is what I found, clinically, when I took oxycodone, it didn’t do anything. And side effects were observed after 3A4 blockade or in the 2D6 ultra metabolizers.
So if you’re metabolizing very quickly, you get high levels of oxymorphone, which makes people sedated, nauseated, and all of those. So when we look at the metabolism of oxycodone, I told you is fairly complicated. You’ve got your oxycodone going to oxymorphone via 2D6 and noroxycodone via 3A4. If you’ve got a 3A4 inhibitor, you’ll raise the levels of oxycodone because you’re not metabolizing it, you’re not getting rid of it, so you’ll end up with more oxymorphone, more of an effect. If you inhibit the 2D6, you get less of effect and shifting more of it to the noro metabolite.
2C19, here’s a rapid metabolizer and a poor metabolizer. So, again, these were in asymptomatic, not pain patients. These were physicians who were at risk for having problems. 2C19 is responsible for metabolizing clopidogrel, the proton pump inhibitors, and some of the antidepressants, imipramine and sertraline.
Diazepam has a four-fold longer half-life in people who are homozygous variants. So if you are a poor metabolizer, the diazepam, which is, oh, by the way, metabolized into two active metabolites can last even longer.
And 2C19 loss of function alleles impair the formation of active metabolites, resulting in reduced platelet inhibition. So if you’ve got a poor 2C19 and you’re put on an antiplatelet medicine because of your heart, what happens? It doesn’t do you any good. And so what they showed was that the loss of function alleles impaired the formation of active metabolites, reduced platelet inhibition, and increased the risk for serious adverse cardiovascular deaths among those patients treated with acute coronary syndromes.
And then 2C9, here’s a poor metabolizer. It’s responsible for metabolizing those vitamin K antagonists, Coumadin, warfarin, [inaudible 00:27:13]. Sort of something you’d like to know. And then the OPRM1 is the actual receptor. We know now that there are multiple opioid receptors. You’ve got your mu receptors, your kappa, delta, and sigma, and each of those has a different site in the body and a different activity. And, in fact, many of their actions are antagonistic.
You’ve got the decreased respiratory drive with the mu receptor, an increase with the sigma. You’ve got decreased dieresis with the mu receptor, increased with the kappa. And so the effect of an opioid, not only from its metabolism, but also how it interacts at the opioid receptor, and each of these opioids has a different effect at different receptors.
Turns out oxycodone has a lot of kappa agonists. That’s that perky Percocet. That’s why people have problems sleeping when they’re taking Percocet. It’s because it’s a kappa agonist. It increases that alertness. And so when you have somebody who is a poor 2D6 metabolizer, you’re raising the oxycodone levels and they won’t sleep very well and they have those sorts of problems. They’re now stimulating a different receptor.
So there are OPRM…sorry, OPRK1, OPRD1, so there are different receptors for each of those and each of those are genetically controlled as well. So OPRM1 is the primary side of most of the opioids. Ultra function leads to reduced potency of opioids, and it’s potential tool to predict adequate opioid dosage. And I’ve had patients where we’ve tried multiple opioids, nothing worked, did the genetic testing and sure enough, their OPRM1 didn’t work.
No opioid is gonna do them any good. They didn’t have an active opioid receptor and so all they got were the side effects of the medicines because it didn’t do them any good. And so that patient that’s sent to me for an intrathecal pump, well, we’re not putting morphine in there, we’re not putting… So we looked at…
Man: [inaudible 00:29:26].
Andrea: [inaudible 00:29:26]. Exactly. And so the pharmacogenetics of this is that it’s been the target of interest in large number of pharmacogenetic studies as well because of the structural variations as well as the rule of opioid receptors in a variety of disorders. They naturally regulates the analgesic response, but also controls the reward effects of many drugs of abuse including opioids, nicotine, and alcohol.
So think about it a second. If you’ve ever been to an AA meeting, what are they all doing? They’re all smoking, aren’t they? And they’re all that… We’ve always worried about that co-dependence of alcohol and nicotine while they work at the same…I mean, alcohol and opioids while all three of those are working at the same mechanism, which is one of the reasons why I am a rabid anti-smoker advocate because these are stimulating the same receptors and I think it puts patients at risk for relapse in their disease if they’re continuing to smoke.
We also have COMT, which is Catecholamine-O-Methyltransferase. It’s the enzyme that creates dopamine, norepinephrine, and epinephrine. Now, this one takes a little bit of thinking about it because it’s currently the most studied gene in relation to pain modulation. It’s been implicated in the pathogenesis of migraines and anxiety disorders as well as a whole variety of cardiovascular diseases.
So we can look at two different types. We’ve got the low COMT, the Met/Met is the wild type. That’s the standard. High COMT is the variant. The high COMTs, they’re metabolizing their increased breakdown of neurotransmitters. I see these as the adrenaline junkies. They can’t keep enough adrenaline in their system. They’re the high-risk behavior sorts of things. They have a higher pain threshold.
Think about a rugby player. In fact, if you take a rugby player and if his son plays rugby, that grandson will have a demonstrably high pain threshold. You’ve selected for that. Nobody plays rugby if they hurt a lot when you hit them.
There’s a decreased opioid receptor density and an increased opioid requirement. It just takes a lot more opioid to get through the system. The low COMTs have a lower pain threshold. They’re not warriors, they’re worriers, and increased opioid density and they have a decreased opioid requirement. So the Met/Met is the normal, the Val/Val needs more morphine than the Met/Met or the Val/Met. And so we know low levels were significantly more frequent in patients with fibromyalgia compared to healthy controls, then the Met/Met genotype was genetic risk for gambling and drinking problems.
So then we’ve got the MC1R, which is the melanocortin-1 receptor. These genes have evidence potential for analgesia based on sex. Up until now, everything we’ve talked about seems to be pretty evenly distributed between the sexes, but this one is different. There is evidence that women, more than men, respond to kappa-induced analgesia. Remember we talked about the mu receptor versus the kappa receptor. So they did… Women carrying two nonfunctional alleles had higher analgesic response to pentazocine, which is a kappa agonist. And I would expect with buprenorphine as well that that hasn’t been done.
Man: It only works on women. We’ve seen that for years.
Andrea: And now it explains it. Now it explains it.
Man: It only works [inaudible 00:33:11], I never understood why.
Andrea: Yeah. But what’s particularly interesting, 75% of individuals who have red hair and pale skin end up having two or more inactive variants in the MC1R. So how many of you are anesthesiologists? Okay. What if we say redheads were harder to anesthetize, they bled more? And it’s not a wives tale, there is a genetic basis for this.
One of the last ones we’ll talk about is UGT and I’m not even gonna try to do that. It influences the levels… Remember, I told you we’ll get back to the M3G. So M3G is hyperalgic. It causes pain. And when you have a shift in the metabolism of morphine from morphine to M6G, which is analgesic, to M3G, which is hyperalgic, you start ending up with a situation where you’re giving more medicines and it’s causing more pain.
So my best example was a kid on the cancer ward with testicular cancer, on 1,000 milligrams of morphine a day just screaming, screaming. We weaned him down to 30 milligrams and he got good relief because at those higher doses, the metabolism was shifting to M3G. And so it turns out that tamoxifen, diclofenac, naloxone, carbamazepine, tricyclics, and benzos are all inhibitors of this enzyme, all potentially shifting the patients to the hyperalgic state, one of, again, the many good reasons to get your patient off their benzos because of its interaction with the opioids.
There are some environmental effects on the genes as well. We know that smoking is a potential, a very potent inducer of 3A4. So what happens is you’re smoking and it decreases caffeine levels. So you drink your two cups of coffee, no big deal, you stop smoking, and what happens? Caffeine levels rise, they get all jittery and anxious. I just tell them to cut back on the coffee. It’s amazing. They’re what they need for [inaudible 00:35:24], but it was the recognition of this.
But you’ll also have decreased theophylline levels, so you have to be careful. If they’re on theophylline and they stopped smoking, they may get toxic. And here is a marked increase in inhaled insulin levels. So if any of you are having those patients who are having the inhaled insulin, it’s actually a contraindication in smokers and those who have quit less than six months.
But there are other influences on smoking, and one that I’ve seen a lot has been the recognition smokers have higher pain scores and required more hydrocodone than nonsmokers. This study actually looked at two groups of people, gave each of them 4 hydrocodone 10, and the smokers had higher pain scores and less hydromorphone in their system. And I’m convinced that smoking is a 3A4 inducer, shifting the medicine to its inactive norhydrocodone form. And so they had significant lower serums of hydrocodone than the nonsmokers.
And, in fact, you need to be aware that smokers may have indetectable serum levels of hydrocodone and they get what? They get thrown out of the clinic, right? Because their drug levels aren’t there, but it may be because they’re smoking. And I actually had patients who take the urine toxicology because we can show them that they’re high levels of nor metabolizer. [inaudible 00:36:48] And I’ve had them put that on their refrigerator as a motivator and, sure enough, and as they stop smoking, they say over and over again the pain medicines are working better.
And then you have to recognize about the herbals. More and more people are adding herbals to their diet and to their metabolism and what ends up happening is there can be significant interactions. One of the most common is St. John’s Wort, which is a very potent 2D6 inhibitor. But some of them are much more subtle.
So just recently, I went into a patient’s room, she was on a stable dose of oxycodone, and she said, “Doc, the oxycodone is not working.” “Okay, so what did you add? Did you start any new medicines?” “No.” “You sure?” “No, nothing.” “Did you start any supplements?” “No.” “Did you start any new vitamins?” “No.” “Okay, something’s going on here. Did you start anything new?” “No. Well, I mean, my dad convinced me to start taking cinnamon to try to and lower my blood sugar. “Sure enough, [inaudible 00:37:52] a 3A4 inducer and that was causing her oxycodone to go to the noroxycodone level and not to the oxymorphone. We got her off the cinnamon and her levels went back up and she got analgesic again. So there are many things that can interact with these genes.
So why should you consider genetic testing? Well, one of it is drugs are metabolized slowly in individuals carrying a genetic polymorphism that causes an absent or decreased enzyme activity and these individuals are in increased risk for adverse drug effects. And when we talk about opioids, that adverse drug effect can be death, Okay? So we’re talking about serious issues.
The drug therapy could be ineffective if the drug is metabolized too quickly. And so knowledge of these polymorphisms before beginning drug therapy can help in choosing the right agent. Somebody who is 2D6 deficient should not be on oxycodone, codeine, hydrocodone, or tramadol. They will have bad effects from that. So you shift them over to new center, morphine, something else, and you don’t waste your time trying medicines that didn’t work or can’t work.
And so what’s the evidence of this? Well, there are some. There are some clinically relevant genetic biomarkers from the brain in alcoholism, and what they were able to show is high resolution. Looking at these genetic issues, gave the counselors a convenient visual image of the distribution of alcohol-related polymorphisms. And it said that it may facilitate a gene-based personalized counseling of alcoholism.
We looked at the utilization of pharmacogenetics and therapeutic drug monitoring for opioid pain management…I’m sorry, pain management. These suggest that patient care may be improved by genotyping as well as following therapeutic drug levels, which we can do fairly easily with urine and drug testing.
The impact of genetic variability on sensitivity to opioids, a systematic review showed that if they had the allele for the 118G allele for the 3A4s, it showed that you had less nausea and vomiting and higher pain ratings. And in their analysis, they found that the 3A4G carriers consumed less opioids than the homozygous-1 patients.
So, again, when you’re looking at… When patients say, “I need more medicine,” or “I needed less medicine, I’m very sensitive,” we said, “Oh, you,” we patted them on the head, “Yes, isn’t that sweet,” and ignored them completely. But you need to listen to these patients.
This was the promoter variant in… Remember we said this was the metabolism of morphine, evidence may be important clinical implications for morphine therapy in patients with sickle cell disease. That high frequency of the allele and its detection of significant effects on morphine metabolism suggests the need for future studies.
They were able to predict which of the sickle cell patients truly were having trouble with their opioids and not drug seeking. So here was a whole class of people that have been disabused as being simply drug seekers because they had sickle cell and they were able to… Eat, eat. They were able to… Eat it [inaudible 00:41:30]. They were able to identify that these patients truly had pathology in their metabolism of the opioids.
And then the severity of side effect of opioids. The occurrence of central side effects of opioids can be caused by the COMT gene and the sedation death is two times more in the genes of the AA and the GA carriers in comparison to the GG carrier. The GI side effects may be connected with genes of the steps of the OPRM1 and COMT. Vomiting is 2.8 times more in the OPMR1 gene carriers.
And then the severity of cutaneous side effects, the itching, the histamine release was made to be dependent on the COMT variables in genotype GG for cutaneous pharyngitis.
And the one I really… This one was really fun, though. Amitriptyline or not, that’s the question, where they looked at 2D6 and 2C19 and they could predict those patients who were at high risk or low risk for side effects with amitriptyline. Now, we got a ton of patients have been shifted to amitriptyline because it’s cheap. Really important, but it’s a dirty drug and it has multiple potential devastating side effects like the orthostatic hypotension, urinary retention, all of those, and so it would be good to know whether or not your patients could safely take a cheap medicine or whether they need to be taking a more expensive one.
Applications in the clinical setting. This was looking at a variety of the genes that will give information on what it is the effect on opioid addiction. So we’re beginning to drill down more and more about what is leading to this addiction. So, again, if you knew going in that your patients had a risk of opioid addiction, then you would monitor them more closely, you might avoid opioids more readily. You have a reason to avoid opioids more readily, and you would try and keep them on as close a leash as possible looking at alternative, perhaps alternative delivery systems, making sure that you are looking at opioids that are less likely to be liked by that particular genetic issue.
And then the… We already talked about a little bit the modification of antidepressants. That you ought to be dosing the antidepressants, not by what the PDR says, but by what their genetic makeup is for that particular medication and that particular 2D6 metabolism.
There are some ethics, though, involved in doing this testing and the genetics don’t change, so your testing really can only be done once. And, unfortunately, there are very extensive tests and there are very limited tests. And so you have to be aware that if you’re using limited testing, that may be the only chance your patient has to get that testing. And the limited studies may prevent reimbursement for future testing, so you wanna know that your patient is getting the most reliable information.
We can get some clues from the urine and I strongly encourage you to look at your urine quantitatively if you’ve got a patient who you’re having problems trying to figure out what’s going on. So in this particular situation, hydrocodone, norhydrocodone, hydromorphone, relatively even equal amounts. Patient is getting good relief, this is fine. But then you look at this patient and what do you notice?
Man: No hydromorphone.
Andrea: No hydromorphone. There’s no… Lot’s of hydromorphone. This is either a 2D6 inhibitor or a 3A4 induced patient who is… When they say they’re not getting relief from their medicine, they’re not getting relief from their medicine. Okay? And so you see it again. Here, you see the oxymorphone, high levels of noroxycodone, and very low levels of oxymorphone. With methadone, low levels of methadone, more than three times the level of the EDDP.
Man: Rapid metabolizer.
Andrea: Rapid metabolizer. And so when the patient says, “I’m getting [inaudible 00:45:51] halfway through the next dose,” that may be it. One of the things that we viewed, this may be a smoker too, though, that induction of the 3A4 that I think goes on with smoking. And what we’ve done is we’ve very gingerly added 3A4 inhibitors, so well and so here is what I tell patients.
I gave them an example of how important it is for them not to start or stop any medicines for any reason without discussing with us first. When you look at the patients, I tell them, okay, you have cimetidine, ranitidine, Tagamet, Zantac, over the counter stomach medicine. They sit on the same shelf, they’re first cousins, they are equivalent, right? Wrong. Ranitidine is not a 3A4 inhibitor, has no effect on methadone. Cimetidine will double the blood levels and they’ll turn blue and die.
And I tell them, “I will not have my name on a bottle next to your cold dead hand. The first time you accept a medicine from anyone, another physician, over the counter medicine, without discussing with us first, you will be taken off the methadone. End the story.” I say, “Look, it doesn’t do any good to call me after you’ve gotten the prescription because if I say, no, now what are you gonna do? I’m not gonna write it for you. If you’re getting an antibiotic, you need to, before somebody hands you that prescription, you say, ‘I’m on methadone.'” Do not rely on them having seen it on the list, do not rely on them remembering. You say, “I am on methadone.”
So you can get some interesting clues from urine. This was the drug testing that I originally did and I could get four different tests. This is what we have available now and that, I think, gives you much more information about what might be useful for your patient. So you need to know what your panel looks like and what it is that you’re doing.
The other ethics… There are some really positive things about this, about the genetic testing. It’s a validation of what the patient has been saying.
When the patient says, “This doesn’t work,” or “I’ve been too sensitive,” or “My mother had a terrible time with medicine X and I’ve had a terrible time with medicine X, ” that should really tell you there’s likely to be a genetic problem there.
It’s avoidance of possible side effects and it’s a choice of potentially more effective medicines. I promise you, you are a hero if you’re not doing the same thing everybody else is doing, and we sort of feel as though we have to go down this love dreamless. Well, if you already know that these aren’t going to work or these are dangerous in this patient, the patients love the idea that you’re actually treating them and not just a case.
But you do have some downsides, and the biggest one that we worry about are these research fees. Again, we’ve been seeing, recently, where physicians are being given money to do genetic testing. If a lab gives you money for doing anything, you need to be concerned about it, I’m afraid.
But there’s also then a lot of refusal of insurance coverage and so we end up with a situation where we’ve got a tool that works really well and we’ve had a lot of problems getting the insurance companies to pay for it. And so what has happened is the companies have figured out, and you’ve gotta admire them, that they will do all comers with the thought that the ones that do get paid subsidize the ones that don’t get paid. And as long as that works, that’s fine, but you need to make sure that you have in your mind the medical necessity of doing these tests any more than you wouldn’t do an ECRP in somebody who didn’t have symptoms. I mean, you wouldn’t a CT on somebody who wasn’t having problems.
So it needs to be a medical necessity for doing this. Though, I think it’s relatively easy to come up with those because our patients are the ones that haven’t done well. They don’t get to pain management because they do well with opioids and they’ve got a simple problem that’s going away. These are the most of the complicated of the patients and many of them have already failed multiple treatments, and so I think that validates the necessity for most of our patients.
Centers for Medicare Medicaid will cover the C29 or the VKRC alleles. If they’ve not been previously tested, they’ve had fewer than five days of warfarin and are enroled in a prospect and randomized controlled clinical style study. But they won’t do it, they’re not covered for a variety of other anti-inflammatories. Same sort of thing, the non-genetic testing for 2C19 is considered investigational and also the 2D6 is considered medically necessary if there is amitriptyline or nortriptyline and Tetrabenazine greater than 50 milligrams per day or re-initiation of therapy with doses greater than 50 milligrams. But it’s not covered for the vast majority of other things at this point.
So what I do is I look…I’ve got a lot of medical necessity that multiple studies have shown. I have the patient has undergone these recommended procedures, the patient has had genetic testing ordered because, and the physician signature that this is a medical necessity of doing this particular test.
Economics, well, we’ve got more than 100,000 deaths per year. It’s not a trivial problem due to the adverse drug events. 2.2 million serious ADDs every year, $1.4 to $4 billion of direct hospital costs due to adverse drug reactions. The ADR is twice as frequent poor metabolizers, 44% compared to 21%. And the cost of ADR for an ultra metabolizer or a poor metabolizer is 4,000 to 6,000 higher per year than it is if you look at the intermediate or normal metabolizers, and their hospitalization is longer.
In a retrospective study, 33% of patients with severe side effects were poor metabolizers, and in another prospective study, all the patients experiencing toxicity, confusion, sedation, or orthostatic hypotension after desipramine were poor metabolizers. And then guidelines have been published, codeine and 2D6. So if you’re looking at codeine as a medication, testing for 2D6 has been part of the guidelines. If you’re looking at Plavix, it’s part of the guidelines now and if you’re looking at warfarin, it’s part of the guidelines.
And then you’ve got the Pharmacogenetics Working Group of the Royal Dutch Pharmacists and they recommended several TCAs, SSRIs, and NSRIs, and established those recommendations based on the individual 2D6 genetic type, so it was a very… They now have actually laid out exactly what the dosage should be if you’ve got a poor metabolizer, if you’ve got this allele group or that allele group. Tremendous amount of work. So for, example, imipramine dose should be reduced by 50% to 70% in poor metabolizers, plasma concentration should be monitored. In an ultra metabolizer, an alternative drug may be considered, plasma concentration monitored or an increase in imipramine dose by 70%.
And then the AMA has gotten into it, looking at pharmacogenetics, increasing the safety and effectiveness of drug therapy. So they look at a variety of medications using pharmacogenetics to predict and prevent adverse drug reactions. Plavix is an example, warfarin as an example. And then this was…
I’m gonna end with this study, which was a very interesting one. And this was done in…just published last month. And what they did is they had a patient, 48-year-old guy with a past medical history, that had attention deficit disorder, obstructive sleep apnea, polymyalgia, post-traumatic stress disorder, and chronic low back pain. Sounds like a lot of our patients, doesn’t it? So he noted the extreme sensitivity to antidepressants and insufficient response to his ADH medicines for depression and for ADH.
So here we’ve got a situation where the patient has said, “I have bad effects from these medicines,” and, of course, we all say, “Yeah, yeah, yeah, yeah.” And they actually tried him on an NSRI, had terrible headaches and problems. So they said, “Okay, let’s genetically test him.” And so they found he had reduced activity in COMT… We really talk about the…we’ll talk about the [inaudible 00:55:30]. But his regular ones were pretty normal. So he’s really not the CYP system that we’ve talked about up until now.
So low COMT, we talked about being it corresponds to increased concentrations of dopamine and norepinephrine. You’re not getting rid of these medicines. The Catecholamine-O-Methyltransferase breaks down serotonin and norepinephrine and epinephrine. So if you’re a low COMT, you should have increased levels of this. And he’s also on the methylphenidate, which blocks the reuptake of norepinephrine and dopamine, resulting in increased dopaminergic and noradrenergic activity in the prefrontal cortex.
So he should have had plenty of neuropathy around, but he wasn’t getting relief from his ADHD and he wasn’t getting relief from his depression. Okay? So what’s going on? Well, he’s also has reduced MTHFR activity. This is the enzyme which is responsible for converting methylenetetrahydrofolate to methyltetrahydrofolate, which is the predominant form of circulating folate. So this means that with low levels, he has low levels of folate. Okay?
Well, it turns out, this poor activity is associated with poor or limited response to antidepressants, pain medications, and [inaudible 00:57:01]. So they proposed to increase his folate levels using either methylfolate, which crosses the blood-brain barrier, or… So they started him on leucovorin at 10 milligrams each morning. This was in November of 2016. It’s the active form of folate, doesn’t require this enzyme for activation. Possible side effects included zinc depletion, so they gave him zinc sulfate every evening.
Pain scores decreased from 9 to 10 to 2 to 3 within 1 week and he continues…the last time they saw him eight months later, he still had excellent relief from his pain, his depression, and his ADH. Never would have thought of that without having done the genetic testing. It gave a clue that led to the answer in this particular patient. Isn’t that cool?
Man: [inaudible 00:57:56].
Andrea: No. No, because he had full… The T3T4 is like the testosterone free and total testosterone. No, he didn’t have activated folic levels.
Man: [inaudible 00:58:15].
Andrea: I don’t know that.
Man: Jeff [inaudible 00:58:19] is there in that study. I know Jeff. He was with me at [inaudible 00:58:22].
Andrea: Yeah. That could well be. Yeah. It’s a… I don’t know. But I thought it was absolutely fascinating. It just came out last…I just saw it a couple days ago, so I haven’t had a chance to talk to him about that. But in summary, pharmacogenetics hold great promise to answer and predict the response to a given medicine and to decrease the potential for side effects. But currently, it’s primarily being used retrospectively on a limited basis.
I’m fascinated by the idea that you all are in a situation to actually be able to prospectively look at your patients, predict what might be a problem, and use a whole new tool that you haven’t been able to use up until now. At the very least, patients with poor tolerance to medications in the past may benefit from genetic testing.
I’ve got a couple of resources for you. This is a review of the role of genetic testing in pain medicine. This was from Pain Physician which is Open Access. Genetic testing for opioid pain management, a primer. This is also an Open Access journal. And then genetics and implications of perioperative analgesia for those of you who still do some OR work. And actually, when we’re taking our patients to the OR, we may need to warn the anesthesiologist, this patient is at risk because we know their genetics and the anesthesiologist doesn’t. And this is also Open Access. There are a variety of websites that are… As I said, I’ll be glad to give you this lecture if you want or the copies of those articles. Just email me and I’ll be glad to send you that information. Thank you very much.
Man: The FDA issued a box label warning for kids that die following [inaudible 01:00:03].
Andrea: Yeah. And they did not bother to recognize that the problem was the genetics. So it’s the typical FDA response, which is paint everything with the same brush, instead of identifying those kids who are at risk and documented risk, they should not be getting tramadol. But they probably shouldn’t be getting codeine either and they probably know. So they probably shouldn’t be getting hydrocodone, they shouldn’t be getting oxycodone. Really, what they need to be getting is morphine, but nobody wants to do that in a kid. Instead, we’re gonna give them Tylenol. Yeah. And so how many kids are gonna suffer from that? We’ve already them off…
We took Vioxx off the market. Unfortunately, we also took Bextra off the market with no evidence. And for those of you… Again, here’s a genetic issue. Vioxx was metabolized by cytochrome [inaudible 01:01:05] enzyme. Turns out, one of the other a few other things that were metabolized by [inaudible 01:01:11] is aldosterone. So if you’re not breaking… This is competitive inhibition. So if the Vioxx is blocking the metabolism of aldosterone, what happens? You get higher levels of aldosterone, which leads to fluid retention, hypertension, and heart attacks and strokes, you know. As opposed to Bextra which was metabolized by 2C19, has nothing to do with the same metabolism system. But, you know, we’re not gonna give the [inaudible 01:01:47] because they’re gonna bleed. So just we’re…
Andrea: You know, we’re putting ourselves between a rock and a hard place.
Man: [inaudible 01:01:59] pretty complicated stuff.
Andrea: It can be.
Man: We have prodrugs that require metabolism, we have drugs that [inaudible 01:02:09]. And then we have drugs that induce enzymes, drugs that inhibit them. We have patients on multiple drugs. And so when we get a genetic test, we get levels of activity like the chest arrhythmia. But the ideal [inaudible 01:02:31] kind of like a computer programme, [inaudible 01:02:35] genetic testing like in the medications that they’re taking. Now, the computer say, okay, well, this drug is interacting with this and the consequences and stuff. So, I mean, if you have a patient on tramadol and morphine, whatever it may be, and they’re also on some drugs that inhibit or induce enzymes, to sort that out all out on the screen. [inaudible 01:03:02] could do that. Mine, [inaudible 01:03:05].
Andrea: Well, so there is a programme that does that. Generex [SP] has a programme that does that. And so you log in, you either send them the list of the meds and they plug it in or you plug you plug it in, and then they combine that with the genetic testing and come up with a list of medicines that should not be used.
Man: That would be awesome.
Man: [inaudible 01:03:27] So, you know, thank you for [inaudible 01:03:31]. We are a big believer of this. We wanna be ahead of this. We believe that if we have this test, one, I’ll test these patients so that we can give them a better treatment. Now, our test that we [inaudible 01:03:47] a comprehensive [inaudible 01:03:49].
Andrea: The company [inaudible 01:03:59] that I used to work, was the medical director for, we had a one-page list of what was involved, but then we would have a multi-page that would list a variety of medicines and would say, for this particular patient, this is [inaudible 01:04:14]. Avoid these, these are okay.
What I tend to do is I look at my patient’s list and I’ll go down…there a couple of them that are just really bad actors. If somebody is on Prozac, then you know they’ve got a 2D6 problem. And so first thing I do is I try and clean up their meds. I put them on…if they’re on 2D6… We know 2D6 affects most of the opioids. Look at a couple of… If they’re on Paxil or Prozac, if they’re on cimetidine for 3A4. So there are a couple of them that I just look down the list, is that there? Okay. But, yeah, there are now fairly sophisticated programmes that are available.
Man: [crosstalk 01:05:00] widely or…
Andrea: Yeah, they’re widely available.
Man: I have an app. [inaudible 01:05:06].
Man: So [inaudible 01:05:08] the results for our patients.
Andrea: Oh, they love it.
Man: [inaudible 01:05:13] comprehensive. [inaudible 01:05:15] all the medications [inaudible 01:05:18] makes it really easy if you look at the [inaudible 01:05:21] red, the yellow, and the green. So the red are the ones like, you know, be careful here. So most of my patients go, “Yeah, I’m taking that.” So you might wanna bring this to your cardiologist. [inaudible 01:05:34]. Oh, my God. Thank you. And they really like it.
Andrea: They really like it.
Man: [inaudible 01:05:40] the results with our programme kind of says all the possible consequences are already there. So you can use the programme or you can just look at what the gene [inaudible 01:05:53], something happens, it tells you all the possible complications that might occur. It’s really part of the [inaudible 01:05:57].
Man: I guess what I’m getting at [inaudible 01:06:05]. But, I mean, even for me, if they’re on multiple drugs, over the counter drugs, I think I’d love to see that happen.
Andrea: Well, one other thing that I also do is tell them, “Look, you need to understand that if you’re on more than six medicines, your risk of having drug-drug interactions is 94%.” I show them that statistic. And so then we go, “Look are there… Let’s go back over this list. What is it that you actually have to have? What is so… So you’re on a medicine for constipation. Why aren’t you just on fiber? And you’re on a medicine for blood pressure, if you lost 10 pounds, you wouldn’t need to be on the blood pressure meds. If you have blood sugar, you’re on this dose of medicines for…” It’s a huge motivator for them to see that they have a chance to make a difference, rather than just somebody is telling them, “Oh, well, you know, lose 10 pounds.”
And with the smoking, to be able to show them the drug level in their urine because of their smoking and the ineffectiveness of their opioids. And I say, “Look, I could care less whether you have heart disease or lung disease, I’m telling you that you’re not gonna get pain relief from this medicine as long as you’re smoking. And I’m not gonna put my license at risk to increase the medicines as long as you’re smoking. So until you stop smoking, I’m not gonna increase those medicines. And this is why. Look at this.”
And so then, every time they come in, it’s, “How much are you still smoking?” Then we start pulling lists. “Okay, you were smoking a pack and look what it was here. You went down to half a pack and look at this and look at your pain scores.” “Doc, you’re right. Wow.”
Man: So, [inaudible 01:07:58], some of our… We have used multiple labs for [inaudible 01:08:05]. Well, what we do, so you know, so we have affiliate labs for us and even if it’s not covered, we will send our patients to collect. So we’ll do more and they won’t be send. A couple of things I [inaudible 01:08:27].
Andrea: For example, [inaudible 01:08:37].
Man: They ask $20 to your office visit. [crosstalk 01:08:40]. When you talk to your patients about all these things, the patients like them. In our times, this is the [inaudible 01:08:53].
Andrea: [inaudible 01:09:08].
Man: But the labs [inaudible 01:09:12] all the patients regardless and they will never be sent for [inaudible 01:09:16].
Man: Okay, when you say we…
Man: We means our affiliate labs.
Man: [inaudible 01:09:23].
Andrea: No. And again, if there’s a lab, they’ll take all comers and you get a consistency in the results that you get and you can start to actually interpret them and see what’s on going on.
Man: [inaudible 01:09:50].
Man: Thank you.
Andrea: You’re welcome.
Man: [inaudible 01:10:20] send me those slides. [inaudible 01:10:21].
Andrea: Yeah, [inaudible 01:10:23].
Man: I know the lab may wanna use your medical necessity.
Andrea: That’s fine. Welcome too. You’re welcome too.