Thank you all for joining this webinar on Improving Medications for Opioid Use Disorder Practices. It will be presented by Dr. Mark Fishman. Dr. Fishman is board certified in addiction psychiatry and addiction medicine, and is a member of psychiatry faculty of the John Hopkins University of School of Medicine. He's also medical director of Maryland Treatment Centers. Before I hand over the stage to Dr. Fishman, a couple of housekeeping details. By participating in this webinar, you will be able to earn one CME credit. I'll send more details about that after this webinar is over. During this presentation, if you have any questions, please type them in question and answer box. We are monitoring the chat box and we'll take those questions at the end of the presentation. If you are interested in receiving the research articles that Dr. Fishman is discussing in this webinar or his presentation, please send me an e-mail. I'm dropping my e-mail address in the chat box. I can send those documents via e-mail after this presentation is over. Having said that, Dr. Fishman, it's your show. Terrific. Well, thank you and hi, everyone. It's a pleasure and a treat to be here with you today. This is the 2023 version of a talk that I've been giving every year for umpteen years to the local Maryland DC Society of Addiction Medicine, where I pick a handful of self-selected journal articles from recent times and call them the research greatest hits of the year or the past couple of years, and cram them in a quick rapid-fire lecture to give a condensed whirlwind summary of some recent findings. I pick them for all sorts of idiosyncratic reasons, but mostly for their practical clinical impact. This year is no different, but this year they've clustered around the theme of approving MOUD practice. That's, of course, particularly relevant to this group. I'm going to show you today four articles. Often I do six or seven, so you get a reprieve. We can go just a teens slower. This year I asked the local addiction fellows to help, so I gave them a little reprieve to be able to go a little slower, but I'll be solo today. Here are my disclosures. The most relevant for today are that I've done a little work for both Alkermes and Indivior, who make some of the medicines that we'll be discussing. I don't think that biases me, but you guys will be the judge. Here's what we're going to be talking about today. Again, this is a mini journal club, but with too many cups of coffee, we'll be going quick. Then, of course, to the extent that we want to have a discussion or questions afterwards, we can do that if there's time. The first paper we'll talk about is about real-world retention in OUD medication treatment. Who stays in treatment or unfortunately, more germanely, who doesn't stay in treatment? Because unfortunately, as you guys know, dropout is alarmingly high. The second paper will be about the value and impact of initiating buprenorphine in carceral settings and some of the impacts in the emerging knowledge base of why that's so important and so helpful to patient outcomes. Third paper will be about the use of MOUD for OUD that co-occurs with other substance use disorders. Opioids plus cocaine, opioids plus alcohol, etc. Commenting both on how unfortunately, MOUDs are less used in that context, but that they remain safe and effective. Then lastly, some material on dosing strategies for both buprenorphine and methadone, addressing the question of what's the most effective and impactful way of titrating and adjusting dose. Let's get into it. First paper, real-world MOUD retention. This is a 2021 paper from the group that you see in New York and Baltimore, Robin Williams. One of the authors has been doing work in this field for a long time, some of the others as well. The background is that all of you know that MOUD is a great treatment. It's effective, it's life-saving, it's the standard of care, but retention leaves much room for improvement. Nowhere near as many people get onto treatment, but this paper is about that nowhere near as many people as ought to stay in treatment. These two images give you some of the background. On the left is from Robin Williams' seminal paper about the OUD cascade of care. Just to orient you, this is looking at different milestones or stations in the cascade from the number of people in the population estimated to have the disorder, to the number of people who get a diagnosis, to the number of people who are linked to care, to the number of people that's stage 5 who are initiated on standard of care MOUD medicines, to number 6, retention greater than the benchmark of six months, etc. This is a familiar public health tool. Some of you may recognize it from other chronic remitting relapsing illnesses, take HIV. One of the public health care benchmark standards is that from milestone to milestone in the cascade of care, if we're successful, if we're doing well, if we're making appropriate progress towards beating an epidemic, public health experts say that we want, of course, it's not going to be 100 percent continuity between each step, minimize drop-off step-to-step, and 90 percent is the benchmark. That is, we would hope for 90 percent of those that have the disorder get diagnosed with the disorder, and 90 percent of those that get diagnosed with the disorder get linked to care for the disorder, and 90 percent of those that get linked to care get initiated on medicines, and 90 percent of those that get initiated on medicine should be retained six months or more, etc. That's the top of the red bar, the aspirational 90 percent cascade. But unfortunately, it is the blue that represents the estimate of the reality. We are woefully underperforming, and the red added up is the treatment gap, so plenty of room for improvement. On the right is the main outcome for the famous X-Spot trial or CTN-51. Most of you will know this famous paper, but just summary, this was a randomized controlled trial comparing extended release naltrexone versus sublingual buprenorphine for OUD. This was, of course, not the same public health broad real-world estimate that we see in the cascade of care on the left. It was a research effectiveness trial in specialty centers. But it did a number of things. First of all, it showed us that for those that got onto medicine, extended release naltrexone and buprenorphine turned out to be roughly equivalent in preventing relapse to regular opioid use. This is that relapse-free survival curve. But I show it here to remind folks that even in fancy pants boutique research, the best of the best with NIDA funding and caseloads of two and a half. I'm sure your caseloads are only two and a half. Even then, six-month relapse-free rates and dropout-free rates are in the high 40 percent. Roughly around the benchmark that we talk about as 50 percent. And although that's way better than without medicines, don't get me wrong, MOUD is still the standard of care. It is effective. It is lifesaving. There's plenty of room for improvement. So what is MOUD retention in the real world? We saw this in the projections on the cascade of care. We saw it in boutique research. What about in the real world? And what patient characteristics are associated with retention? The methods for this paper were they took the TEDS data set. Some of you will know that. Public sector programs across the country are supposed to report certain outcomes. It's a problematic data set. Reporting is not 100 percent. There are some difficulties in terms of the exact reliability of the data. But it's pretty darn good, and it's pretty darn big. This study took the entire TEDS data set across the nation for 2017. We're looking for those specifically with an OUD diagnosis. And in 2017, in the TEDS data set, there were 300,000 such episodes of outpatient specialty community SUD treatment episodes with primary diagnosis of OUD. Forty-three percent of them included MOUD. Wait, did I say that everybody in public sector treatment with OUD gets treated with MOUD? No, I did not. So point number one, only 43 percent of people received MOUD. Just initiating, never mind retaining, so we can do better than that. And now let's look at the retention. So here's real-world retention across these 300,000 treatment episodes. And we looked at the data from the research study, the XBOT study, at about six months and about 180 days, and we saw 45 or 50 percent retention or 45 or 50 percent non-relapse. Here in the real world, that number is even smaller. So 36 percent of these patients were retained at six months. And at a year, only 22 percent are retained. Again, way better than without medicines. Certainly, medicines are working. We wouldn't see anywhere near these numbers for patients who didn't get buprenorphine, methadone, or extended release naltrexone. But again, lots of room for improvement, right? And what about clinical correlates of these retention numbers? So in the paper, you can look at the details. There's a long list of 30 different things, but I just show you, you know, the top 10 or so just to give you a flavor of which are the factors that are most correlated with poorer retention. A caveat, this is not a research data set, so there are limited number of characteristics reported. So there are probably things that we would say, I wish I knew whether such and such were correlated with retention or not, and that wouldn't be in this rather clunky data set. But still, there's some pretty good and important and impactful things. Look at number one, methamphetamine use. The co-occurring use of methamphetamine with an opioid addiction leads the list with reducing the likelihood of six-month retention on MOUD by greater than one-half. A close second is the demographic of age, so developmental vulnerability conferred by being a young adult. By the way, these aren't new. You will have seen everything on this list, or almost everything on this list, in other settings. You will have seen them in your clinical work. Some of them will be intuitive to you, but it's interesting to see the data, right? A 50% reduction in the likelihood of being retained in MOUD because of being young. Criminal justice referrals. So again, not surprising, being involved in the criminal justice system, a marker for a certain kind of severity, for a certain kind of difficulty with self-regulation, perhaps difficulty with stigma and access to MOUD care. All of those things likely impactful. Being homeless, not surprising. And things like arrest. You see cocaine use. You see unemployed use. You see primary heroin compared to prescription drugs. If this is 2017, if this were 2023, we'd want to ask the question of heroin versus fentanyl, right? That wasn't so much a question in 2017, et cetera, et cetera. Interesting what to make of the health SUD referral. That's not as intuitive to me. I'm going to speculate that this had to do with requirements for participation in health settings and the idea that in some integrated specialty SUD and healthcare clinics, the time course of a treatment episode, A, might have been artificially limited. Like we have a 12-month IOP and then at discharge, you're no longer in the program and no longer retained. Or it might've been that if you don't come to counseling, you don't continue to get medicines and that might have artificially listed and delimited the impact of retention. Not sure. But the other things I think are intuitive. And another way they looked at it is by state. Looking at a very wide regional distribution in the proportion of treatment episodes that met the threshold of six months. And I underlined two here because as I told you, I originally gave this discussion to the Maryland DC Association of Addiction Medicine. So my home state of Maryland where I'm sitting now, we didn't do so well, right? 16% of these treatment episodes only were retained for six months or more. DC did better, but I'm concerned that there were so few episodes reporting from DC. So I don't know what to make of that. Tennessee has just happened to be here with a big number, only reported seven. So I wouldn't put too much in that. But again, the main take-home point is wide variations in practice, wide variations in outcomes. And that's not what we want, right? What we want is aspirationally to have a broad standard of care where patients, no matter where they are, can get universal access with high quality treatment and reduction in barriers to be able to stay in treatment. So just to conclude, real-world MOUD retention is alarmingly poor. Certain characteristics predict worse prognosis. And one of the reasons that's so important is not just to stage prognosis from the beginning to know who's likely to be more vulnerable and who's not, but some of those things are actionable, right? We can add those characteristics early on in recognition of knowing that those would be appropriate targets for adding to the treatment tool chest. If a person is involved in the criminal justice system, how do we interact with the criminal justice system? If a person is younger, how do we take into account developmental vulnerability, include parents or other kinds of things? If a person is homeless, how do we target making sure that we're working with housing services, et cetera? But in any case, we clearly need to do better. That should be at the very least a wake-up call. And these next studies will give us some ideas about ways in which we can do better. So paper number two, recidivism and mortality after in-jail buprenorphine treatment for opioid use disorder. And this is a topic that is certainly one that's very important and has been in the news a lot, has been on people's minds as there is increasing momentum across the country, not as much as there needs to be, but increasing momentum to involve the criminal justice system, to reach into the criminal justice system, which has traditionally had an antipathy towards MOUD, especially the agonist MOUDs, methadone and buprenorphine, and give people in carceral settings access to MOUD before they are released and to increase linkages to community care after they are released. Up to two-thirds of people who are incarcerated have a substance use disorder. Maybe half of those have specifically opioid use disorder. And overdose is the leading cause of death at the point of post-incarceration release, at the point of reentry. And there have been estimates from 10 to 50-fold increases in the background rate of opioid overdose deaths. People have lost their tolerance. They're back out into the community celebrating. They're typically not on MOUD. They're typically not linked to MOUD. Lots of work has explored the efficacy and effectiveness of linking people behind the walls to all three of the FDA-approved medicines. They've been done for methadone. They've been done for buprenorphine. They've been done for extended release naltrexone. But unfortunately, MOUD is not available in many and most carceral settings. Rhode Island has been a pace-setter and leader in the country, Massachusetts. Maryland is early on in mandating involvement and implementation is ongoing. But lots of places have not yet followed suit. One of the things to think about is what's the way that we're selling this? And most of the prior research has focused on SUD outcome benefits, not surprisingly. So looking at links to community treatment, looking at relapse rates, looking at rates of continuation of medication and the like. Reduced overdoses, although it's a low frequency event, has been demonstrated. And obviously, that's a big impact one. Reduced rates of risks for infectious diseases, HIV, HCV, others, a big one. But most of that work has showed beneficial outcomes that are salient to the care community through a health care lens. And you'd think that's mom and apple pie, but not everybody has that lens. So criminal justice recidivism, as a target, has additional importance because it's a critical downstream functional outcome which matters to a different subset of people making decisions, to lawmakers, to public safety officials. So if we could connect MOUD to decreased recidivism, health is good. Not having a fatal overdose, sure, that's good. But public safety goals may have salience to another group of folks to highlight the value of increasing the footprint of MOUD in correctional settings. So the question was, how does availability of buprenorphine during incarceration affect post-release recidivism outcomes after return to the community? Now, this was not a randomized prospective trial. This was a natural experiment done in Massachusetts where there was a time between implementation of jail MOUD between adjoining rural counties. So although it's not randomized, these were very close by in the same state, same demographics. Presumably, many of the same features involved in the populations. One started offering MOUD in 2016, and the other started offering MOUD in 2019. So there was this natural ability to look at outcomes when one had done it and the other had not. So it's going to be a comparison between county one, Franklin County, that had been using during those years MOUD in the jails versus the second county, Hampshire County, that had not. And the primary outcome was recidivism, as defined as any incarceration, any probation violation, or arraignment after the index jail release. Again, these are jails, not prisons. And let's look at the results. So here are the different ways of looking at that outcome. Any incarceration, being in the county that adopted in-jail BUIP, drastically reduced all of these recidivism outcomes, including any incarceration and any recidivism. Markedly lower rates. So 42% I'm sorry, 21% rate of subsequent incarceration with BUIP, almost twice that, 39%, with no BUIP. And you see the same kind of big change between any recidivism. And in the county that did adopt BUIP versus the county that had not yet adopted BUIP in the jails. Here are the recidivism survival curves. So blue is the county that had early adoption of buprenorphine in the jail. Red is the county that had not yet done it. And you could see at all time points, there was less recidivism when patients had buprenorphine made available to them pre-release. Now another, and by the way, the cumulative hazard ratio was pretty substantial, 0.71. Now one of the things that's also interesting though is that the gap narrows over time. And that won't shock you because this is a year follow-up. And so the idea of, this is longer than a year, but the idea that medication initiated at time point A will not necessarily be as impactful over time. Now the interesting thing is it is impactful, but it's less impactful over time. And that likely relates to the point we were making before that retention is alarmingly low. But anyway, bottom line, this natural experiment across two similar rural jails in Massachusetts found that incarcerated adults with OUD offering buprenorphine in jail had reduced risks of recidivism. And after the logistic regression was controlled for baseline, that was a 32% reduction, pretty substantial. So evidence that agonist MOUD reduces recidivism. And this is, as I said, a critical public safety outcome, which is going to give a salient impact, both in terms of financial and societal costs, to an audience that may not be as swayed by the health outcomes. And hopefully, that will support the increasing momentum to do more of this across the country. And the take-home message is that this is something we should be doing. We should be offering MOUD in jails. Those of you who aren't working in jails should be working to coordinate with jails and receive the pass, if you will, in linkage to reentry continuation MOUD treatment and form those relationships for warm handoffs, because that makes a difference. And as I said, we've got older evidence that it makes a difference in terms of care, reduced overdose, reduced transmission of infectious disease, reduced relapse. But here's evidence that it also improves public safety. All right, next paper. Comparative effectiveness associated with buprenorphine and naltrexone in opioid use disorder and co-occurring polysubstance use. And the idea here is, what happens with our patients who have not only OUD, but also disorders of other substances, co-occurring alcohol, co-occurring stimulants, co-occurring sedatives. And the reason that's so important is because our opioid overdose is increasingly an epidemic, not just of opioid use, not just a fentanyl use, although fentanyl clearly drives the fatal overdoses, but specifically also an epidemic of polysubstance use. And it is understudied and undertreated. These two graphs give you two snapshots into the problem of overdoses with methamphetamine on the left and cocaine on the right. And in both instances, the top line with the steepest increase is in this combination, right, of stimulant plus opioids. And the combination is quite problematic, quite lethal. And unfortunately, that has deterred many practitioners from using MOUD or being more restrictive in their use of MOUD based on concerns both about safety, if I give MOUD to patients that are not just using opioids, but using other things, benzos, alcohol, stimulants, will that pose a safety problem? And is it effective? This paper attempts to answer the question, what is the comparative impact of MOUD, buprenorphine and naltrexone, on OUD with and without concurrent other SUDs, both from a safety and an effectiveness point of view. So here are the methods. This was a retrospective review of a big claims data set. It combined some commercial and some Medicaid claims data, big data set of almost 200,000 claims across this period, 2011, 2016. And the inclusion criteria were looking for claims of folks 12 to 64, whose primary claim diagnosis was OUD, who weren't treated with methadone, who had a claim for buprenorphine or naltrexone or psychosocial treatment without medicines. And the primary outcome was a non-fatal drug-related poisoning. That's the way that that's listed as an outcome in these data sets as an overdose. So non-fatal drug-related overdose or poisoning that resulted in a claim seen in an ED visit or a hospitalization. And they looked at the subset of these patients, these claims who received any of these two MOUDs and had at least one poisoning event. Now that substantially narrowed the data set. So from 200,000 down to 12,000, but still a big number and looking for the association between the number of days receiving MOUD, a year before and a year after the index poisoning episode. So a wide view of the lives of these patients comparing those with and without additional co-occurring SUDs, right? So we know that the number of days exposed to MOUD, old news, reduces the risk of having such a poisoning but how did that compare if you had only OUD versus OUD plus? So just some characteristics here and some results. Of the patients with OUD in this claims data set, 43% of patients got medications, same number or similar number as we said before. Again, woefully inadequate, less than half of the patients with OUD are getting MOUD. 57% of them got psychosocial treatment only. Now, of course, this doesn't say people who got no treatment because they wouldn't be in a claims data set. There wouldn't be a claim. Of the people that got medicines, you're not surprised. Most of them were bup and many fewer were extended release naltrexone or oral naltrexone. 27% of these patients had a co-occurring SUD. Alcohol was the most frequent, stimulants right there behind and sedatives, most of those would have been benzos a third but lots of them, right? Not the majority, but a very substantial minority had co-occurring SUD. So it's an important and impactful target population. So the first thing is what did this mean in terms of treatment receipt? So the first thing is what did this mean in the target population of those MOUD and another SUD? They were less likely to get MOUD. Of those that had the co-occurring other SUDs, 30% got MOUD as opposed to 47% of those who did not have a co-occurring other SUD. So practitioners are less likely to use MOUD. The MOUD was less than 44%. It held for extended release naltrexone. Oddly, I'm sorry, they were more likely to get either extended release naltrexone or oral naltrexone. These are small numbers, but I've got to presume even for that small minority that it reflected a idea that buprenorphine was less safe, perhaps because of the potential combination of depression, synergy of alcohol or benzos. I don't know, but again, hard to put too much stake in those small numbers, but big differences with less likely to get buprenorphine. Now, what about the impact of the use of MOUD? Many did not get it, but what about those that did get it? How did the protective impact of MOUD differ between those who had a co-occurring SUD other than opioids versus those that did not? So this is 9% of the total MOUD population did have at least one poisoning. Remember, we were funneling it down to the 12,000 that had at least one poisoning. That was 9% of the total MOUD population. Most of those had more than one such poisoning, right? What's the best predictor of a future overdose? It's a past overdose. So what's the difference between having a co-occurring SUD and not having a co-occurring SUD in addition to opioid addiction in terms of the impact of MOUD? And the answer is hardly any, right? Here are the odds ratios reflecting the impact of days of exposure on preventing these poisonings. And you can see that the little squares are roughly the same for buprenorphine and extended release naltrexone with a co-occurring SUD, alcohol, opioids, stimulants versus no co-occurring SUD opioids alone. So that means two things. It means that it wasn't a safety problem. The MOUD didn't cause a poisoning, right? Worried specifically, for example, about, well, is alcohol, is Xanax gonna contribute to the possible respiratory depressive effects either of the medicines or of the illicit street opioids, heroin, fentanyl, others? Looks like it did not on average. So it's relatively safe, good news. And secondly, is it still effective in preventing overdose? Not pictured here is efficacy or effectiveness in perhaps to opioid use or criminal recidivism to name an outcome that we talked about in the previous paper or other functional outcomes like employment, all of which we know MOUD does impact positively. But the ones studied here are these overdose poisonings. And the medicines were just as effective in preventing overdose poisonings whether or not there was co-occurring alcohol sedative stimulants. Good news. And let's look at the individual substances and not that much difference, right? This is the same control we saw before, no co-occurring SUD. We see buprenorphine and extended release naltrexone, forget oral naltrexone, right? We all know that that's not very effective for anything. But here, I mean, maybe it is for alcohol, but it's certainly not for opioids. People don't take it. But if you look at alcohol, you look at stimulants, you look at sedatives, maybe it's a little weaker of an effect for stimulants, but overall pretty good for each of these potential co-occurring SUDs with an opioid SUD. So safe and effective. So conclusions and summary, fewer patients with co-occurring SUDs received MOUD compared to those without these co-occurring SUDs, but MOUD, that is buprenorphine and extended release naltrexone, was equally protective for those with and without co-occurring SUDs. So the take-home here, the practical impact for your practice at home is you should not be hesitant to use MOUD for these patients with co-occurring SUDs. It is safe, it is effective, it is lifesaving. Maybe they don't work quite as well. Maybe the co-occurring SUDs are not targeted well, right? Buprenorphine is not especially good treatment for Xanax addiction, of course, but for their primary indication and effectiveness and safety for the MOUD, they are good to go. Use them, do not hold back, do not withhold these medications from patients while you are also addressing the co-occurring methamphetamine, their co-occurring cocaine, their co-occurring benzos. They should still be treated full throttle with MOUD. All right, homestretch, last paper. Buprenorphine and methadone dosing strategies. This paper was called Buprenorphine and methadone dosing strategies to reduce risk of relapse in the treatment of opioid use disorder. And the background is that in general, there's kind of some fair consensus that high doses are needed, that there is some high enough dose of buprenorphine and or methadone to reduce relapse to opioid use. The old idea that we should be very sparing. We have emerging evidence first with methadone and later with buprenorphine that more is better, that we've been underdosing, but it's unclear what high enough is, right? And it's unclear how you get there. Current guidelines recommend that bup should be at least 60 milligrams. Methadone should be at least 60 milligrams, although certainly 80 to 120 is perhaps more typical. Both of those numbers probably should be higher now in the era of fentanyl, that's not as well studied. But again, not clear consensus on what is high enough and how to get there. So that means that the decisions about our robust or maybe less robust response to clinical course and adjusting dose in accordance with the patient's outcome in real time doesn't have really great evidence-based guidance. So this paper tries to add to the evidence to answer the question, to what degree do particular MOU deducing strategies affect the risk of treatment drugs? It was a complicated methods. We're not going to spend a lot of time on it, but it was secondary analysis of pooled data from three CT clinical trials network randomized comparative effectiveness trials, which in and one of them was the X-Spot paper we talked about before. And all told, there were 1,800 participants. And in all three of these trials, they were not strict pharmacological efficacy trials. They were effectiveness trials in which the investigators were given leeway to do dose adjustment based on clinical judgment without very clear guidelines other than do your best, use your best judgment. And so there was heterogeneity, heterogeneity in dose within a range, heterogeneity in speed of titration of the dose within a range, and heterogeneity in terms of what principles or approaches were used by individual practitioners. So that left some variance that could be studied. The way this was, the outcome was time to opioid relapse starting at about three weeks, giving people time to settle in and monkey around. And so looking at relapses, time to relapse from three weeks to 12 weeks, the analytic method was a thing that I don't fully understand. And I think you need a PhD in mathematics to fully understand. I will quote to you, sequentially doubly robust estimator of longitudinal effects, which is a complicated method having to do with repeatedly doing machine learning algorithms, repeating various scenarios for each patient, and creating estimates of what would have been the outcome given the comparison to participants or patients in the studies who did get a particular method, and using mathematical modeling to estimate what would have happened to a different patient had that method been used. And the four methods or the four strategies that emerged from the data that were then compared with each case in this machine learning algorithm were number one, a dynamic dose increase. That is, each time a patient had trouble, had opioid use, had a lapse, the dose was increased. A second method was, a second strategy was weekly dose increase to a target threshold. And that emerged to be a 16 milligram target threshold for buprenorphine, and 100 milligram target threshold for methadone. A third strategy that emerged was a hybrid one, combining the first two, in which weekly doses were increased to the target, and dynamic dose increasing beyond that continued in response to opioid use. And a fourth treatment strategy was no change in dose, constant dose after two weeks of treatment. You get to where you get, and you stay there. And looking at cumulative risk of relapse, this graphic compares these four different dosing strategies, labeled here as constant, dynamic, target, and hybrid. And this is buprenorphine. You won't be shocked that the three strategies that involved increasing the dose did better than the strategy that kept the dose constant after two weeks. So dynamic, target, and hybrid, you see here, all have lower relapse rates throughout and at the end than the constant dosing strategy. And the hybrid strategy was better than either the target or the dynamic strategy alone. So getting up to a dose quickly and increasing thereafter in response to how the patient in front of you is doing did best. When you say it that way, it sounds pretty intuitive. But not everybody did that. Not everybody does that. So it's nice to have this evidence. Here's the methadone, in which, again, all the strategies that involved dose increases did better than the constant strategy. And the gap is even bigger here for methadone. But there was no difference between the dynamic, target, and hybrid. They all did just about the same. So conclusions, for both of these medicines, we want to increase the dose, we want to push the dose, and we want to do it in response to how people are doing. For bup, the dose should be rapidly increased to a target threshold of at least 16 milligrams and increased dynamically based on response. And remember, these studies were done in the heroin era. And so that target threshold is probably considerably higher in the era of fentanyl. For methadone, the dose should be increased either to a target threshold of 100 milligrams and or increased dynamically based on response. Both of those strategies were good. And again, that threshold of 100 milligrams is from the era of heroin and almost certainly is going to need to be higher now in the era of fentanyl. So the take-home is MOUD agonist doses should be adjusted upwards rapidly and in response to clinical course to maximize effectiveness. Sounds like standard medical care, huh? What a surprise. Interesting. We'll wait to see what we're going to learn about dosing strategies now in the era of fentanyl. That work is not yet out. There is a study in the field now, clinical trials network study, CTN-100, which is comparing for buprenorphine, not for methadone, extended release buprenorphine, which wasn't around when the study I just shown you was done. Extended release buprenorphine versus 16 milligrams of buprenorphine versus a higher dose of 32 milligrams of buprenorphine. And so it'll be very interesting to see what those results are. So invite me back in five years and I'll tell you. Anyway, to pull all this together, here are the four summary punchlines. We should address major baseline statistics that we know are associated with poor retention, outcomes, stimulant use, younger age, criminal justice involvement, homelessness. Those things can be taken into account. Additional treatment tools can be incorporated. These are actionable things. We should offer MOUD in jails with linkage to reentry for continuing MOUD care, whether we're doing it by throwing the past from our work in jails or whether we're doing it by receiving the past from our work in the community with warm handoffs. We should use MOUD vigorously despite the presence of other co-occurring SUDs. We should certainly address those. They won't magically melt away, but we should not withhold effective and life-saving MOUD because a person hasn't yet resolved their stimulant, alcohol, cannabis, sedative co-occurring use. And for the agonist MOUDs, we should push the dose higher quickly and use dynamic dosing. This study didn't look at it, but we wonder if the same is true for antagonist treatments. Unfortunately, extended release naltrexone is now only available in the one dose, but is that enough for the era of fentanyl? Probably not. All right, time for a quiz. Get your pencils ready. No, just kidding. Instead, do we have time for discussion and questions? Anybody got comments? Dr. Schwartz, Dr. Vaci, Dr. Hayes, Dr. Hanna. I thought that was a wonderful and focused presentation. We really appreciate your knowledge and experience in the literature, doing your own research and your clinical practice, practical experience. I was wondering if you have any comments about treatment of co-occurring stimulant disorder. I know you have a lot of cocaine use in Maryland. What are your thoughts briefly for clinicians about addressing cocaine use in treatment programs and practice? Important question. I showed everybody the overdose data, and that's only the tip of the iceberg in terms of the way that cocaine and methamphetamine ruins lives in a variety of ways. Unfortunately, we don't have any specific FDA approved medicines that target stimulant use directly, but a number of things I think, nevertheless, should be incorporated into practice. The first thing is what I already said today. Use MOUD despite stimulant use disorders when people have OUD, both because it's lifesaving and effective for the OUD, but also for a substantial subset of people as one addiction is a sleeper effect in which other addictions may also improve. Now that doesn't happen for everybody, and some people have the paradoxical opposite effect that they whack-a-mole style increase addiction two when they improve addiction one, but lots of people, maybe the majority of people get better for better. Second thing is, even though there are no FDA approved medications, and most of the pharmacological efficacy work to date has been disappointing, that's not universally true, and there are some hints at some medicines that even if they're not as strong, as effective, they're not the home runs that MOUD is for OUD, they still ain't half bad and are worth trying. So for cocaine, well, the first point is that a lot of this work has been done separately for either cocaine or methamphetamine, not for both. There are biological differences, but I think in general, the consensus of experts in this field is that there is sufficient overlap in the biology of these two major addictive stimuli that if it works for cocaine, it probably is worth trying for methamphetamine and vice versa. So I tend in my mind, simplistic as that is, to combine the cumulative research on both of these medicines into one pool of knowledge, use them interchangeably. So for cocaine, there has been some demonstrated efficacy of topiramate, and even more so the combination of topiramate plus prescription stimulants. Adderall is what is studied, but there's no reason why it shouldn't be the same for methylphenidate or desven, not desvenlifaxine, desvenvivance. I always forget the generic name. And so, although there are concerns about misuse diversion of prescription stimulants, most of the evidence suggests that combination with topiramate, that's a potentially good strategy. There was a good paper in the last couple of years in the New England Journal of Medicine using the combination of extended release naltrexone plus the bupropion for methamphetamine. So that's certainly worth trying. And there's no reason why it couldn't also be bupropion plus oral naltrexone, although again, there's probably an adherence improvement with the extended release, but it may be easier to get coverage and affordability for the oral. So, and there are other medicines. Some people have thought that mirtazapine is not bad. Some people have thought that bupropion is not bad. And you can sometimes think, well, if you're gonna use an antidepressant, maybe use either mirtazapine or bupropion because maybe they have an anti-stimulant effect. Again, not home runs, but certainly worth trying. And there are others that we could talk about, but we don't have the time. A third principle is that both of those drugs, cocaine and methamphetamine, are associated with very high rates of severe psychiatric comorbidity, depression and psychosis. And although we wouldn't expect the addiction to just melt away with treatment of a co-occurring psychiatric disorder, they are reciprocally reinforcing and mutually sustaining. So a thing that I recommend is be aggressive about evaluation and treatment of co-occurring psychiatric disorders. And you can do that despite ongoing use. You don't have to wait for abstinence from methamphetamine, abstinence from cocaine to use an antipsychotic, to use an antidepressant and help relieve distress and improve their engagement while you're working on their primary substance use disorder. Contingency management is also worth mentioning. Various research data to support its effectiveness. Unfortunately, it's not broadly adopted into real world treatment for a number of reasons. One of them being who's gonna pay for it. The second being that CMS has had some skepticism about whether it constitutes fraud, and hopefully we're making some headway there and finding sustainable models of real world adoption of contingency management, but that should certainly be incorporated where it's practical. Mark? Yes. I'm struck by the number of patients that are offered XRNTX who turn it down. Do you have any ideas for increasing the desirability or the acceptability of the antagonist medication to opioid dependent patients? Yeah, it's a great question. It remains a small minority as you point out, but when we study it, if you get on the medicine, it looks just as good to any other medicine, but fewer people get onto it. Part of that is because agonists have been around a lot longer and have a bigger proven print and track record are better studied and are more widely dispersed among a broader set of practitioners. For example, we've got an entire industry of methadone treating OTPs across the country with a whole infrastructure and regulatory structure, et cetera. So I think historical adoption issues, but I also think that naltrexone is clunky to get onto. So the so-called induction hurdle, the difficulty of initiating buprenorphine remains a major barrier. Although that be quite as true in the age of fentanyl, in the age of heroin, just to compare initiating these three medicines, if you present at a clinic with active heroin use, I can give you methadone today, right now, even if you're methadone intoxicated or methadone withdrawing. I mean, if you're heroin intoxicated or heroin withdrawing, I can give you methadone now and I can give you buprenorphine by this afternoon. Now, in the age of fentanyl, maybe it's not quite as soon. Maybe it's not this afternoon. Maybe it's tomorrow morning. We're struggling with difficulties in initiating buprenorphine or delays in initiating of precipitated withdrawal because of the enduring effect in fat stores of persistent fentanyl, but it's still pretty darn quick. Naltrexone, on the other hand, you come to me, heroin or fentanyl this morning, and when can I give you Naltrexone? Well, I've got to detox you. And then after that, you've got to be opioid free for seven to 10 days. And by the time you do the arithmetic, that's two weeks plus to get to extended release Naltrexone. That's a significant barrier for a lot of people. It's particularly hard to do in an outpatient setting where you're still exposed to the community and access to illicit opioids, but you're not covered, but you're still craving, but you're still in subacute withdrawal. And I still can't give you the Naltrexone for another two weeks. It's easier to do in bed-based care where access to illicit opioids is restricted by confinement and other support, but lots of people vote with their feet and leave before they get a dose. So I think that's a major barrier. Now, there is ongoing work looking at addressing that. There is research work that has looked at outpatient initiation and in boutique research settings, it works pretty well, but hasn't been done under real world settings much. There's work that is looking at reducing time to initiation. So other means of initiating people onto Naltrexone by sparing opioids. So not using buprenorphine or other opioids for withdrawal management, either using maximizing doses of what we used to call supportive medicines or comfort medicines like Clonidine and other things, using sedative hypnotics, whether it's clonazepam or pregabalin or gabapentin. There's work that's been done in using low dose Naltrexone. When I say low dose, I mean like starting at a half a milligram and then a half a milligram BID. And using that as a way of accelerating tolerability of Naltrexone without precipitating withdrawal. And some of those studies have gotten us down from the traditional two weeks plus to maybe six days. So that's a lot better, but six days is still a lot longer than this afternoon. So I think that remains an issue. Okay, thank you. Other questions, comments? I think we're out of time. So we'll end there. I hope that was helpful. Thank you for your attention, tidbits to take home. And as we heard, if you want to dig into the papers themselves, you can email and we'll send them out to you and you can dig in for bedtime reading.

improving medications

opioid use disorder practices

Dr. Mark Fishman

stimulant use

younger age

criminal justice involvement

homelessness

medication for opioid use disorder

MOUD