Good morning. Welcome to the clinical trials. It's my great privilege and honor to introduce Dr. Maurizio Fava. Dr. Fava is the Chair of Psychiatry and Psychiatrist-in-Chief at the Massachusetts General Hospital and Executive Director of the Clinical Trials Network and Institute, and he is here to give a talk on pharmacological approaches to treatment-resistant depression. Thank you so much, Sophia, for inviting me to give this presentation. As a clinician, I have to confess I've been very interested in resistant depression for a long time, and it's one of those challenges that we all face, patients who don't respond to standard therapies. So let me first proceed with my list of disclosures. Well, this is lifetime, so throughout my career, but I do all my consulting through Mass General. I've been doing that for almost 20 years, which means that all the consultant's fees go to the hospital and not to me personally, directly or indirectly, to avoid any conflict of interest. I do work a fair amount with the industry in terms of research, so I conduct trials and I participate in trials of a number of compounds. You can get all my disclosures through the website. So I would like today to make it a little bit interactive in the sense that, you know, we have an hour and a half, and I don't want to bore you for the next 90 minutes. So if you have any questions, you can go to the mic, ask them. But the first question that I ask myself when I see someone that says, doctor, I tried many things and nothing has worked for me, the first thing I ask myself is, what is the diagnosis? And the reason is that, you know, over the years, I tried to make sure that all referrals to me were of unipolar depression. You know, and I would say, look, you know, this is what I'm doing my research on. You know, please, if you refer someone, please refer them who are unipolar. And 39 years later, half of my patients are bipolar. So despite the selection bias in referrals. So why is that? That's a common, you know, kind of clinical issue is that patients that look unipolar turn out to be bipolar at some point. But also, what are the psychiatric and medical comorbidities? What else does the patient suffer from? One of the things that we learn from STAR-D, for example, is that if you have obsessive compulsive disorder, even if you have obsessive compulsive traits, you're significantly less likely to respond standard therapies. Another important question is, what were the previous trials? How long were they? Did they take place? What was the duration of the trial? And what was the dose? Another question I ask myself is, for example, in the case of patients that report 12 weeks on a good dose of an antidepressant, they say, I didn't feel anything. And then I asked, did you have any side effects? And the patient says, no, not at all. I get blood levels of the antidepressants they're on, and they're often below the normal range. And by the way, the normal ranges for most labs are, let's say for searchling, is based on patients on 50 to 100 milligrams a day. So they're not really, they're fairly underestimating the true range. And they're still, the patient may be below that range, which means there may be ultrarapid metabolizers. In those cases, you may want to get a pharmacogenetic test so that you don't repeat the mistake of underdosing the patient. Because these patients may require 800 milligrams of searchling, 200 milligrams of fluoxetine, because otherwise they don't get adequate blood levels. But fundamentally, the question that we have to ask ourselves are, what are the contributing factors? I have already mentioned misdiagnosis. Substance abuse is a common, you know, there are many patients that will underestimate their degree of drinking or use of drugs. And medical comorbidity, of course, psychotic features. You know, sometimes patients are very reluctant to disclose that they're having paranoid delusions or that they're having hallucinations. But we know that those patients are not going to respond to antidepressants alone. And I've already mentioned the issue of the ultrarapid metabolizers. So in practice, we really have four strategies pharmacologically. I'm not going to cover ECT. I'm not going to cover TMS. I'm not going to cover psychotherapy. My brother would be very disappointed if I didn't say that, because he's a big fan of psychotherapy. And I'm not going to cover other, you know, non-pharmacological treatments. So I'm going to focus just on the psychopharmacology of resistant depression. So the options we have are switching, dose increase, augmentation, and combination is combining two antidepressants. So I'm going to ask, I'm going to kind of see, just do a little polling here. Let's say you see a patient that was treated with an SSRI for 12 weeks, with the dose progressively raised to the maximum tolerated dose. And the patient has not responded at all and has some bothersome side effects, but not too much. Who would favor switching? Please raise your hand. Okay. Who would favor pushing the dose even further? Okay. Who would augment, add something? All right. Who would combine two antidepressants? Okay. So the winner for this case is clearly switching. And so I'm going to cover switching first. So for those who are, it's not an adaptive slide, you know, show where depending on your result, I have different, but where do we typically switch? And I would say that the favor switching happens in the presence of no response, as in the case that we discussed. Sometimes it happens in partial response when the patient is bothered on some side effects. And of course, we switch when the patient just can't tolerate the antidepressant. So in the field, there have been typically two approaches to switching. Now, one approach is the switch within class. SSRI to SSRI, acenari to acenari, tricyclic to tricyclic, MAOI to MAOI. And people used to say, why doing that? Well, you know, you're switching from tranlucipramine to phenazine, you get a little bit more of a GABA modulation. If you're going from, let's say, citalopram to paroxetine, you may get more of an oradrenergic effect or fluoxetine. So there were rationales for it. If you're switching from amitriptyline to cumipramine, you get more of a serotonergic effect. But there were supporters of this view and people who are kind of dismissive of this approach. But typically, people favor the switch to different class because it implied you're changing the pharmacology, you're changing which neurotransmitter system or which receptors or which transporters you're affecting. And there is also data from suggesting that one subtype may be particularly responsive to another one. For example, the Columbia group with Don Kline and Fred Quitking and Jonathan Stewart and Pat McGrath, they had shown consistently that monoamine oxidase inhibitors would work better in major depressive disorder with the typical features and tricyclics. So the tendency was to favor switches to a different class. And so when we designed STAR-D, STAR-D was the largest clinical trial ever conducted in the field of depression with over 4,000 patients. And I had the opportunity of being one of the three PIs with John Rush and Madhuka Trivedi. And we were kind of designing the study. This is a long time ago because we're talking 25 years ago. And at that time, we had anticipated that the first SSRI to become generic would be citalobram. We were wrong, but so we put everybody on citalobram in level one. And then if the citaloreparanol responders would be randomized to one of three switch options, another SSRI, sertraline, and a SNRI, venlafaxine, and a non-serotonergic antidepressant, bupropion. And I remember that table. I was more on the side of, you know, switching within the class. I've seen it work. And other people at the table saying the switch to a different class will clearly win. So let me show you what happened, is that there was no difference. Pretty much one in four patients remitted. Now you see a little bit of a difference on the Hamilton depression rating scale because it was given only at time zero and at the end of the 12 weeks, whereas the quids SR was given at every visit. So it's a little bit more accurate trail. So it's working on this side, but not working on the other side. We had a few technical problems this morning, so I apologize. It's working now. All right. So as you can see, no difference fundamentally. One in four patients remitted. Whether it was a switch within the class to serotonin, or whether it was a switch to a different class, venlafaxine and bupropion. And that was in some ways challenged, kind of the view that switching within the class made no sense. Now, what if you keep switching? In STAR-D, we then gave the option of no responders to level two. So they had failed two antidepressant trials. They were given the option of switching again. And here again, I'm going by the quids SR, even though it was not the primary outcome measure because of a greater accuracy. Fundamentally, a third, you know, a second switch or a third antidepressant didn't really work. This was mirtazapine and nortriptyline. The efficacy of that was fairly, fairly low. So what that means is that although switching, in the case we discussed, made perfect sense, if I had told you the patient had failed 12 weeks of an SSRI and 12 weeks of an SNRI, your odds of success with a second switch would be fairly low. So how do we switch in practice? And the, you know, kind of the approach that I favor is the abrupt switch when you go within class. So SSRI to SSRI or SSRI to SNRI. But I'll tell you that, you know, a number of colleagues of mine always tease me because they said, oh, Maricel, I know you're an abrupt switch kind of guy, but I favor the cross taper. And so I, that's why I put the cross taper first, because I know it's more popular in practice than what I do. And you start low, go slow, and you switch from one agent to another. And sometimes though, when you do that, you have to consider potential drug-drug interactions. That is, first drug perhaps inhibiting the metabolism, the drug you're switching to. And now a question is, where do MOIs or TCAs fit in the treatment algorithm? I like tricyclics and MOIs. I confess that my favorite tricyclics are either very noradrenergic, like protryptony, or very serotonergic, like clomipramine. And for MOIs, I'm partial to phenazine. So you're going to ask me, so why did we use tranlcipramine in the level four of STAR-D? I was outvoted. So where do rapid acting treatments fit in the algorithm? What about a combination of extrametorphin and proprion, which shown effects perhaps a little bit more rapid. I think that it's often a function of your, of the insurance of the patient, how many prior authorizations you have to fill out, because many of the more rapid antidepressants are not just in the formulary and don't require a prior authorization. One thing, when I switched to an MOI, be very, very careful, because there have been cases of people dying when they switch from fluoxetine after only a two-week washout to the MOI. Because remember, fluoxetine, with or without fluoxetine, may take up to four weeks to be eliminated. And the drug-drug interaction, you can have a serotonergic syndrome when you combine MOI with an SSRI. So in general, two weeks going to an MOI and going off of the MOI, except for fluoxetine. And maybe, maybe protreptiline because of the longer half-life, although I'm less worried about potential drug that I do two weeks even with protreptiline. The next option in treatment-resistant depression is the dose increase. And the dose increase, I've been a big supporter of this approach. People, the Mass General, like Sophia, probably know that whenever, and I see other folks here, see Brian from Mass General, they're, you know, they know that if they are uncertain whether they can go higher on a dose, they typically ask me so they can document that I've done it before. And so, but I've been a fan of pushing the dose if tolerated. And why do I do that? To increase the chances of obtaining adequate blood levels in the ultra-rapid metabolizers. But also because as you go up on the dose, you may get other, you may recruit other pharmacological actions. So, if you push the dose on phenylzine, you get more of a GABAergic effect. If you push the dose on fluoxetine, you get more of a noradrenergic effect, and so forth. And of course, milifaxine, we all know that you have to get above 150 to get a noradrenergic effect. That was Sheldon Prescourt, a very informative tyramine challenge study. So, I'm a big fan of using all the parts of the buffalo, which means that when we were doing a study on the long-term treatment of depression with fluoxetine, this was in 1989. This was a randomized withdrawal study. And I said to myself, oh, we have to now, everybody who kind of doesn't respond is not going into the second phase of the trial. So, we lose information of what happens to them. So, I reached out to colleagues that were part of this four-side multi-center study, and I said, why don't we do a trial where we take the patients who didn't respond to fluoxetine, and instead of simply dropping out of the study because they wouldn't go into the randomized withdrawal phase, we offer a dose increase or an augmentation with the zipramine and lithium. And so, 25 to 50 milligrams of the zipramine. Fluoxetine is a cytochrome P450 to the 6 inhibitor. So, I didn't want to get blood levels and risk elevation of the levels of the zipramine. So, we used low-dose, the zipramine. And 300 to 600 milligrams of lithium. And by the way, I, this is one of the things, I then applied for an R01 with Jerry Rosenbaum. we got the R01 based on this design. And our hypothesis was that the zipramine and lithium augmentation would be more effective than just pushing the dose. So we got the funding, and right after we got the funding, we found out that increasing the dose was far more effective than, so we actually used the data of the R01 to buff up our results with 142 patients, so a single site study, and in addition to the data pooled from the pilot study, but fundamentally, going up on the dose was more effective than lithium or the zipramine augmentation. Now, when we published the original pooled data, Larry Price and Craig Nelson wrote a letter saying, ah, study's completely flawed, he used low doses and so forth. So in the subsequent trial, we actually looked at those with adequate blood levels for lithium and adequate blood levels of the zipramine because we didn't have enough power, and it turns out that even among those with adequate blood levels, going up on the dose was more effective. So how do we increase the dose? You know, if the patient is tolerating, I usually double the dose. Now, again, I say 50 to 100% increase because people say, Mauricio, you're a little bit too cavalier. I don't feel comfortable doubling the dose, but if the patient is tolerating, that's what I do, and I wait at least four weeks to see if the strategy helps, and if that doesn't help, I'll go higher again if the patient has no side effects. Again, blood levels or if you have pharmacogenetic data, you can feel comfortable pushing the dose very high. Now, any questions about switching or pushing the dose thus far? You can use the, what is it, keep going? Okay, go ahead. You want to? No, because if we're going to have questions, we should use the mic because of the virtual audience, so thank you. How often do you, when you're going above the threshold of, yeah, I don't know, 40 Prozac, 200, do you almost always get genomic testing or sometimes? No, so the, it's a great question. 40 milligram dose to me, I did a fluoxetine, I don't consider it high, so I usually go without testing to 60 or 80. Now, if the patient at 80 milligrams is still having no side effects and no response, I may get a blood level or genetic testing. Thank you. Sorry if I'm jumping ahead, but can you talk about switching from one MAOI to another? Yes, so switching from one MAOI to another requires basically two weeks of, now, you're going to say why, why? Because you're, these are covalent binding of the, it's irreversible, why should it make a difference? Well, it turns out that the tranalsipramine may have some, it's debatable, but may have some fetamine-like properties and therefore, there is a risk of elevation in blood pressure going to tranalsipramine. The question is, and so there have been reports of problems switching from one MAOI to another. Going from tranalsipramine to phenolazine may be safer, but I still, since we don't have any data, I still do two weeks. Go ahead. My question is, when you are increasing the dose of an SSRI of a patient who's been on this medication for a long time, will you expect to see side effects, for example, like Zoloft, GI side effects, or let's say also anticholinergic side effects from other medications? Sure, you know, great question. So first of all, I try not to keep people for a long time on one antidepressant. So I may do the dose increase after four weeks, after six weeks, and so forth, but the dose increase is easier if there are no side effects because the moment you're starting to see side effects, let's say in SSRI, sexual dysfunction, or, you know, the anticholinergic side effects, you see them more with the SNRIs, and they're not really anticholinergic. They're really, in fact, you can treat them with drugs like Cardura, you know, antagonists of the alpha receptors. But I would say I try to do the dose increase quickly, not wait a long time, so, thank you. We'll go ahead and take one more, and then go back to the presentation. You sure? Sure, sure, sure, absolutely. I asked you to be interactive. You can talk to the questions. Okay, sure. Yes, go ahead, go ahead. Perhaps I'm missing some giant available data set that everyone else has, but appropriate serum levels for most of the antidepressants outside of tricyclics, do we have a well-established standard? I remember years ago going through a very painful exercise, one patient, to try to figure out about whether her citalopram levels changed, and when it got to the point of having to convert data from milligrams per deciliter to molars, I almost committed suicide. Is there an established standard for most of our currently used drugs? Thank God, yes, and the, again, I have no conflicts of interest with the Mayo Clinic. In fact, I would say Mass General and Mayo Clinic, they don't have a little rivalry there, but they do have very reliable measurements of blood levels, so our lab sends it to the Mayo Clinic. They have fairly reliable for most agents. Yes, but the standard, what would be, but therapeutic goals, what the therapeutic range is, we have data on that? Yeah, so I would say it's limited because often it's based on 20 or 30 patients, and so with standard doses, so I would say that it underestimates the true range in the higher level because, again, for sertraline, it's 20 or 30 patients of 50 to 100 milligrams. Well, that, to me, is not a normal dosing range. For sertraline, it would be 50 to 200, so it's conservative. So if your blood level comes back and it's above that range but not a lot, it's probably normal. Well, I'll just make a pitch. If there's anyone in the audience with the capacity to start doing those studies that is looking at responders and remitters and what their serum levels are, give us some guidance. I'd be very happy to see it. It's a great idea. I think it's well overdue, in my mind, since those tests at Mayo Clinic are expensive. I think it may be 400, I don't wanna be quoted for that, but they're expensive. If I were Mayo Clinic, I would do the study on their healthcare system and look at EHR data on how patients are doing, what levels they have. They could do it internally, so you may wanna reach out to the Mayo Clinic because they could do it at no cost to them. And just final comment, though, but pharmacogenomic testing, which may be less reliable because it's not the end result, is gonna be equally expensive. The pharmacogenetic, yes, but it will potentially inform you as you go from one agent to another. Thank you. I'm gonna ask a question from the online audience. So this question states, after what time do you prefer to change the type of antidepressant? We know that an antidepressant does not work immediately and we need to know the wait time for it to start working. But I guess the question is, what is an adequate trial for you? Right, I mean, our group at Mass General has studied this extensively. I would say that if you haven't responded after eight weeks, the odds of response are very small unless your dose increase was delayed. So I typically, after four weeks, increase the dose, double it. And then if still well-tolerated after two weeks, I may go to an either higher dose. At that point, if after eight or 10 weeks, you don't see much of a response, it makes sense to switch. Now, other people say, oh, you just have to wait 12 weeks, 14 weeks. From my perspective, when patients are struggling, I really feel bad to ask them to wait longer when the odds are very small. Okay. Is there evidence for raising the dose in elderly people when they have not too much side effects? Sorry, can you repeat the question? Is there evidence for raising the dose in elderly people when they are up to 65? Oh, I see, I see. Yeah, unfortunately, most of the studies were done with populations 65 and below. And in elderly patients, there is greater chances of side effects, so you wanna be more careful in those cases. But it's not uncommon to see elderly patients who can tolerate higher doses of SSRIs. Okay, thank you. So, my name's Noah Friedman. I'm a psychiatrist in Queens. Just a question, about 80 milligrams of Prozac. Sorry, we're talking about 80 milligrams of Prozac. At that point, you get a level or pharmacogenomics testing, and then if it's clear, you go right to 160, because you're talking about doubling it. And do you do this in patients who are anxious as well as depressed, or just pure depression? So, someone with 80 milligrams on fluoxetine that is not responding and has no side effects, I may get a blood level or pharmacogenetic testing. At that point, let's say that the blood level is below the normal range, I may go to 120 or 140 or even 160. And then you can recheck the blood level and see, are you in the range? Because those patients may require very high doses of the antidepressant. And if they're an anxious patient as well? Well, even if they're anxious patients, the SSRIs have anti-anxiety effects as well. So, I'm not worried about the dose increase because anxiety is present. Anxiety can be a side effect when you start them on an SSRI, not necessarily when you push the dose if the dose is low, I mean, if the blood level is low. Thank you. You're welcome. All right, good. See, I like the fact that we had a little, you know, a series of questions. So, let me move on to augmentation. The augmentation, the definition of augmentation is the use of a psychotropic agent without per se an indication for depression to enhance the effect of an antidepressant. Why do we do that? To get a different neurochemical effect, to affect a different neurotransmitter system, to broaden the therapeutic effect, let's say mixing something that is more anxiolytic, or to combine agents with very different mechanism of action, receptor blockade with inhibition of transport. So, let's start with lithium. And, you know, lithium, you know, it's an interesting area because I see, you know, Dr. Coase is here from Cornell, who's been, you know, he and I go way back, and, you know, when we were practicing the 1990s, lithium augmentation was very popular. It was definitely the augmentation de jour. But, if you're asking people, actually, I'm gonna ask you, how many of you use lithium augmentation in unipolar depression in someone who has not responded to an SSRI for 12 weeks, adequate doses? So, still some people continue to use it. Okay. So, I stand corrected. Now, what are the disadvantages of lithium augmentation? Risk of toxicity, you know, the renal toxicity, the thyroid toxicity, the need for blood monitoring. Now, the advantage is that the meta-analysis make lithium augmentation look very good. You know, Bowers and Dorfman had the meta-analysis showing fairly robust effects. And Craig Nelson, who's always been a big fan of lithium augmentation, also showing robust effects of lithium augmentation. But then, you know, our group conducted a study, this was many years ago, with Andy Nierenberg, where we augmented nortriptyline with lithium. It was a placebo-controlled study, and we showed no effect. With adequate blood levels of lithium, I was the one, you know, the unblinded doctor checking the blood levels. Adequate blood levels, no effect. So, when we designed STAR-D, again, we're talking 1999, so 25 years ago, I was among those who thought lithium, it's not gonna make it. But I was surrounded by people who really liked lithium augmentation. So, I'll show you the data on lithium, but I was also, you know, I'm persuaded by lithium augmentation because of the trial we had done, where going up on the dose was better than adding lithium or desipramine. So, subsequent to STAR-D, a study looked at citalopram added to imipramine versus lithium added to imipramine. And lithium did not work as well as adding citalopram to, or combining, if you wish, with imipramine. So, what about thyroid? Okay, how many of you use thyroid augmentation in someone who didn't respond to an SSRI for 12 weeks with adequate doses? How many people? Okay, a little fewer than lithium. You know, one of the, I think, common reasons when I ask colleagues, you know, how come you're not using T3? And they say, well, because the primary care doctor gives me a hard time, scolds me every time I put someone on T3. But the reality is that 25 to 50 micrograms of T3 don't really, shouldn't really cause significant changes in the thyroid function because of the feedback loop. So, you don't really, in theory, need to get blood levels of your thyroid function with this. And T3, ARPRANG, really one of the pioneer in the use, at UNC, of the use of T3, showed T3 is probably better than T4. Meta-analysis showed an effect, but overall, perhaps not as robust as lithium. So, when we did this, started the trial, my colleagues, the same ones that outvoted me on the phenosine, were saying lithium is gonna be the winner. And I said, you know what, I think thyroid is gonna be the winner. And they kind of made fun of me, but here are the data, and thyroid, augmentation did not significantly better, but you know, a 10% difference in remission rate is not trivial, you know, a number needed to treat of 10. So, T3 worked better. Despite the publication, this was, you know, we published this in 2006 or 18, you know, we published this in 2006 or 18 years ago, I have not seen any uptake of T3 in practice. In fact, as you see, and so, for whatever reason, it's not a popular augmentation. Now, buspirin augmentation, far more widely used, far more widely used, I think, well, let me test. How many of you will use buspirin first line after, you know, 12 weeks on the SSRI? Buspirin, not too many, not too many. Well, you know, my predecessor, the former chair of psychiatry, my general, Jerry Rosenblum, is known to have said, buspirin, everything you want to see in an anti-anxiety medicine, in terms of safety, except efficacy, so he was not a big fan of buspirin. In fact, the only buspirin augmentation trial, actually, there are two, were both negative, suggesting that it doesn't really work. However, post hoc analysis looking at more severe depression showed some benefit of buspirin augmentation. The reason why I like buspirin augmentation is because I don't use it alone. Why not? Turns out that our group was involved in the development of a combination of buspiron and melatonin. The reason why we developed that, by the way, there are some seats, because I see people standing, there are some seats in the front, because I hate for you guys to stand throughout, so there are some seats here and there, okay. That company called Brain Cells found that the combination of buspiron and melatonin was a powerful kind of inducer of neurogenesis, both in vitro and in vivo. Yeah, there are, okay, good. Could you raise your hand if you have a seat next to you so that other people can sit down? Yeah, we have plenty of seats. So they had found this stimulation of neurogenesis in vitro and in vivo, and so we ended up helping them conduct a trial in depression of buspiron, 50 milligrams a day, and melatonin, three milligrams a day, both at night, and we compared that to buspiron alone, or placebo, and the study showed an effect of this combination. Now, the company kind of failed, so no longer exists, because the commercial, commercialization of a combination of buspiron and melatonin, both generic, was, you know, was not really feasible, so the company no longer exists. However, we learned from that trial that we published, and in my practice, if I'm using buspiron for augmentation, I also give three milligrams of melatonin. Now, all this, of course, it's off-label, because like lithium or thyroid, these are all off-label uses, but with this published data. Another off-label use is stimulants. You know, when I trained at Mass General, my goodness, particularly in the consultation liaison service, stimulants were very popular, but unfortunately, the evidence for their efficacy in treatment of depression is lacking, so this is a study of, at least this amphetamine emesulate in the two phase three studies, showing no benefit whatsoever. Now, what about modafinil? Modafinil's another, it's not a psychostimulant, per se. You know, Nora Volkow had shown that it's a dopaminergic inhibitor, fairly powerful. It's, you know, it's used for narcolepsy and so forth, but off-label, it's been used in depression, and we've done a study, pulled data from two trials, and in SSRI, partial responders with depression, major depressive disorder, and persistent fatigue and sleepiness, and in fact, modafinil was better than placebo. How much, you know, people are using, I think, again, because of prioritization, it's limited in its use, but I'm curious, how many of you are using modafinil augmentation? Okay, so that's not like lithium. Again, off-label. Another off-label use is dopamine agonists. You know, there have been reports, open case reports with ropinerol, with primopexil. Our group, Chris Cousine and Roy Perlis, did a trial, 1.5 milligrams twice a day of primopexil. As you can see, the sample size was small, so only 60 patients, but the difference in response rate is more than 10%, so suggesting that there is an effect of this augmentation, again, off-label. Now, we can switch to on-label augmentations because pretty much most atypical antipsychotics have an FDA indication for augmentation. Now, I expect a lot of hands, but I could be wrong. How many of you use augmentation with atypical antipsychotics? Yeah, so it's the most popular by far. RAP prozole, usually 5 to 15 milligrams of RAP prozole. There have been three placebo-controlled studies showing robust effects. Similarly, we've seen that augmenting with RAP prozole was more effective than switching, so reinforcing the equitypine, again, 150 to 300 milligrams a day of quetiapine showed an effect. Expiprozole, again, showing efficacy. And now, zeprazidone doesn't have an indication, but this was a study from our group showing its benefit added to acetylopram. And cariprazine has an indication from the FDA showing benefits of this agent at doses of 1.5 or three milligrams a day. Now, we because atypical antipsychotics seem so effective as augmentation, we collaborated with Acadia Pharmaceutical to look at pimavanserin. And we had a very positive phase two study with pimavanserin. It's an antipsychotic drug that is approved for Parkinson's psychosis. So we then went into a phase three, and the study showed a very, very small, non-significant effect. So Acadia did not pursue further studies. And so it is used off-label. Again, prior authorization can be an obstacle. Now, let's move on to an area that I've been interested in for a long time. As a former endocrinologist, I always loved things that affect, in some ways, multiple systems and neuromodulation. So compounds that affect that one carbon cycle, methylation, have many, many effects, increased membrane fluidity, and so forth. So we did a study of acetylsylmethionine, or SEMI, 1,600 milligrams a day. And this was a study done at Mass General. First author was George Papakostas. And we found that SEMI was significantly more effective than placebo. We also did a multicenter study, this one using methylfolate. And methylfolate, also part of the one carbon cycle, like SEMI, also involved in methylation. 15 milligrams a day. While 7.5 milligrams did not work, 15 milligrams a day worked better than placebo. What about omega-3 fatty acids? This was a study, Mark Rappaport, Jurczyk, and others, showing some benefit of omega-3s after eight weeks. Those that they used, 1.2 grams, turns out Mark Rappaport, David Michelon, Odi Danlop, and I did a trial, subsequently, showing that four grams a day of omega-3 are significantly better for depression than one milligram. So, one milligram was probably underdosed. The issue is, when you go to four grams a day, is the risk of atrial fibrillation and bleeding. So, you have to be careful, and not treat subjects at risk for atrial fibrillation. For atrial fibrillation, but if you're using omega-3s, you want to use a higher dose. Again, all these things off-label, because they're not approved by the FDA. Another off-label use is amantadine. Amantadine, very rarely used in practice. Actually, I didn't ask you. How many people use SEMI in your practice? Okay, and how many use methylfolate? Some more, and omega-3 for depression. So, I would say methylfolate wins over SEMI and omega-3. Okay, amantadine, I don't even ask, because I'm telling you, this is kind of an old augmentation strategy. People have been doing this since the 1980s. This was a study by Rogers and colleagues, showing amantadine is a dopaminergic agent, but also an MDA antagonist. It has antiviral properties. And interestingly, that despite the popularity of ketamine, dextromethorphan, people don't use amantadine in practice. But this study of 150 milligrams a day shows some benefit in the men, compared to, and less so in the women. But the big augmentation these days is ketamine. Ketamine was discovered by a group of investigators, Rob Berman and colleagues at Yale, discovered by serendipity that ketamine was effective. And it was interesting, because ketamine has shown to be efficacious in animal models of depression, like the Forssman test. And it was considered the reason why the Forssman test was not a good test, because people say this preclinical test doesn't work at all. Even ketamine works in this model. And so for years, this positive preclinical data were ignored, until by serendipity ketamine was found to have a rapid effect, IV at 1.5 milligram per kilogram. Carlos Zarate replicated the findings at the intramural program. This is his paper. Six years later, which again, suggests that sometimes, as a field, we're not quick to follow up on an observation. So it took six years for a replication. But after that, we started to see a number of ketamine clinics emerging, IV ketamine. And we did a study supported by the National Institute of Mental Health, with a number of sites. And we looked at which dose of IV ketamine works best. 0.1, 0.2 milligram per kilogram, 0.5 or one milligram per kilogram. And interestingly, the most effective was the same dose that by serendipity worked, 0.5 milligram per kilogram. Just seeing, went on to then in a study that I was involved with, test the benefit of repeated administration of IV ketamine, showing that twice a week was as effective as three times a week. And then, you know, last year, Amit Anand from our department and from the Brigham, has a joint appointment between the two institutions. Amit published probably the most impactful paper in my mind, in treatment of depression, where he published the results of his PCORI study comparing ECT to IV ketamine, and showing equal efficacy at three weeks. Now, why is it important? Because ECT was discovered in 1938 by Tirlitin being in Italy, was adopted quickly in the US. Stanley Cobb, first chair of Mass General, wrote a paper on immunosuppressant medicine in 1938 about ECT and quickly spread out throughout the US. And for 90 years, people always thought ECT is the most effective treatment for resistant depression. Until this paper, which in some ways challenged that statement, as you can see, even IV ketamine does a little bit better. Now, I have to say, though, that my colleagues who are ECT experts, they say, you know, it's easy to beat ECT at three weeks, give ECT more time, and ECT may win. So, you know, it's true that there's more of a cumulative effect with ECT, whereas ketamine is very rapid, and you gain less by extending the treatment. So the next trial, in my mind, should be really 12 weeks of ECT and 12 weeks of IV ketamine, and let's see who wins. But at three weeks, as you can see, very dramatic effects, slightly favoring IV ketamine. IV ketamine remains off-label, so despite the fact that we now have meta-analysis, significant, you know, it still doesn't have an FDA indication. Now, you can say, why? Because ketamine is generic. No company would spend the money, you know, 400, $500 million to get an indication when, you know, even if they market a branded ketamine, they have protection for five years, and anybody can use any generic ketamine. So we'll never see, in my mind, or it's unlikely that we'll see ketamine FDA approved for that reason. But widely used, lots of ketamine clinics. We have a large ketamine clinic at Mass General, which is run by Chris Cousine, and we just expanded the space because we have hundreds of people on the waiting list to try to get IV ketamine. J&J then did the thing that, in intellectual property, protection makes more sense, which is esketamine, which is the enantiomeric ketamine, is not FDA approved, was not FDA approved. It was only approved in Europe. So they decided to take esketamine, develop it intranasally, and this is their phase two study, and they got an indication they were approved as provato by the FDA for the treatment of resistant depression. Our group has done a meta-analysis, so the esketamine augmentation studies, and they show that there is a clear benefit of this augmentation. A study published in the Journal of Medicine last year showed a slight benefit of intranasal esketamine over quetiapine augmentation. Other glutamatergic compounds, so I want to leave time for questions, so I'll try to wrap it up. Regabalin, only an open-label study from our group. Mementine, one positive study. Oral, this is an MDA receptor antagonist, but several negatives, so I don't know if, again, it's off-label. Now, this is an FDA drug approval. That's the combination of bupropion and dextromethorphan. This is a study that Harriet Templeton published in the American Journal of Psychiatry showing that bupropion, which is a 2-to-6 inhibitor, and is used to primarily change the metabolism, dextromethorphan, combined with dextromethorphan, so 105 milligrams to 145 milligrams of dextromethorphan, one pill a day for three days, and two, and one pill twice a day, was more effective than bupropion alone, got an indication from the FDA for the treatment of depression. And now, in practice, actually, how many of you use ovality augmentation in practice? Okay, not too many, but... Oral is methadone, is under development, and other MDA receptor antagonist, and the phase two study was clearly positive, and phase three was negative, but the development is ongoing. Scopolamine, which Ron Duman has showed having effects on the mTOR pathway similar to ketamine, showed a benefit in a crossover trial by Maura Fury and Wayne Dravitz, and an oral scopolamine did the same. Now, seranol, it's a GABA, positive allostatic modulator of GABA, it showed an effect in depression, and this was a study by Kundus Bruce, Major of Medicine, but did not get an indication in depression. So, it is approved for postpartum depression, seranol, but not for regular MDD. Creatine, it's a brain energy metabolism booster, five grams a day, showed some benefit in women. Minocycline, neuroprotective and anti-inflammatory, showed some benefit in one study, and adjunctive NAC, two grams a day, did not really show any benefit. Now psilocybin, psychedelics, definitely we're seeing significant benefits. This was the original Harris study against the citalopram showing greater benefit. And the classic paper by Guy Goodwin of psilocybin showing 25 milligrams being more effective than one, 25 milligrams, more effective than one milligram. Ayahuasca also positive. And there have been positive studies also by a compound derivative of deuterated form of psilocybin developed by Saibin. And an endocannabinoid dracol palmitoyl etanolamide showed some benefit in one trial, but it's been unreplicated. And MU, this is a functional kappa antagonist that was developed by alchemists. Never got indication even after submitting the NDA to the FDA, but there were positive results both in phase two and phase three. And this is basically buprenorphine combined with MU antagonist, samidorphine. So buprenorphine is a partial MU agonist and a kappa antagonist. And so when you combine it with a MU antagonist, you have a functional kappa antagonist, which then explains, while atycoprine is under development, this was a study by Andy Crystal that showed some benefit with anhedonia. And actually, I couldn't add the slide because the paper just came out. But the paper just came out on atycoprine trial in phase two. Low-dose testosterone, there was no benefit in an eight-week trial. And inhaled nitrous oxide showed some benefit in the trial. Combination, combining it to antidepressants, clearly a big fan of these Cragnelson, tricyclics and SSRIs. However, in our hands, not so much. Atomoxetine did not work combined with SSRIs. And INSTAR-D combining bupropion with citalopram worked as well as buspiron, which you could argue maybe they didn't work either one. But I think that the combining bupropion is quite common. And that's why I now like to combine, to add bupropion and dextromethorphan to SSRI or SNRI. Mirtazapine augmentation is a trial by Pierre Blier showing the benefit. And we did use vilafax and mirtazapine. It shows similar efficacy to an MEOI. Triazodone, again, some benefit, but not well-established. So in conclusion, so I'm gonna leave time for some questions. Treatment resistance is common. Many strategies are available. And clearly, I would say augmentation, combination are becoming increasingly more popular. Thank you very much. So while folks are lining up, I'll go ahead and ask some of the questions that we have online. So one states that pharmacogenetic testing seems to point to vortioxetine, vilazodone, desolidinophaxine for patients who are slow metabolizers and have side effects with most agents. What is your experience with these agents? Sure, I mean, clearly, you may be guided by those testings towards drugs of that type. I see a long line, so let's try to be very quick. Go ahead. Great, thank you. So you had mentioned early in the talk that essentially it's established that you need to add an antipsychotic to an antidepressant in somebody with major depressive disorder with psychotic features. I was wondering how much evidence there is. I remember seeing a paper years ago that tried to look at a bunch of data of patients that have been treated over, I think, six months. And the evidence was inconclusive. So I was wondering what the evidence is to support that, rather than just waiting for a period to see if the depression improves without the antipsychotic. There was one trial many years ago that showed that the combination was better than antidepressant alone. But since then, nobody else has really tried to replicate those findings. So it's a good question. But I would say the standard practice is to combine an antipsychotic with an antidepressant. Thank you. From online, just from a comment of, how do you basically handle the idea of tachyphylaxis with a patient who initially seems to benefit from a particular treatment and then starts to lose that response? Do you increase? Do you switch? The first thing I do for tachyphylaxis is to increase the dose. And then if that doesn't work, I'll switch. Next. Dennis Helmus, psychiatrist from Ohio. If a patient is having a partial response to bupropion, can you augment with generic dextromethorphan rather than using the brand name product? The answer is yes, you can use dextromethorphan. The one drawback is that once the patient figures out that they can get lots of dextromethorphan, there is a greater potential of abuse with dextromethorphan. So yes, you can do it. Or you can then advocate with the insurance to cover a validity by authorization. Because at least with a validity, the patients know they cannot take more than three a day without lowering the seizure threshold. Thank you. And is there any data on the effects of hormone treatment in peri or postmenopausal women with treatment-resistant depression? Any data on what? Sorry, hormonal treatment. Hormonal treatment. Yeah, there is some data on that, but almost no data on treatment-resistant depression. Yes, Jim? My name is Jim Kosas. I'm here to revive your enthusiasm for lithium augmentation. All right. I know it, Jim. In my clinical experience, I treat a lot of patients with chronic depression, and I ask them about what I call hypo-hypomania, history of hypo-hypomania. Have they had periods in their lifetime where they have had a mild high without meeting the criteria for clinical bipolar? The other question I ask is about family history of bipolar disorder. And in those kinds of patients, I think that they have a much greater chance of responding to lithium augmentation. And it can be very rapid, and it can be very dramatic. So it's definitely something worth asking about. And in those kinds of patients, you know, using lithium augmentation, one of the things that STAR-D failed to do is that they failed to look at clinical characteristics that might, you know, favor one switch strategy or one augmentation strategy over another. Yeah, we did publish studies on comparing the outcome on different subtypes, anxious or atypical or melancholic. But we did not find substantial differences. Particularly in the case of lithium, for example, the remission rate was fairly low, close to 10%. But, you know, what you are alluding to is the kind of soft bipolar patient. I agree, lithium is a very effective treatment for bipolar disorder, and therefore, it's a very good augmentation for soft bipolars. But since this talk was on unipolars, you know, I was, that's why I remain unconvinced about MDD. Thank you. I wanna ask a very naive question. Looking at the very pretty pictures that you showed us of study results, it was impressive to me that it looked like some of the effect sizes were much greater in some of the trials than others. And I know, I'm no statistician, but I know there's a big problem in comparing studies and looking at how the results are depicted. Nevertheless, are you impressed with difference in effect sizes in the different approaches that you talked to us about? Well, it's actually a very sophisticated question, which basically, can you really compare effect sizes? And the answer is no. The reason you can't is because effect sizes are a function of the placebo response. A study from our group, from Nadia Iovino and George Papakostas, has shown that if your placebo response in depression is 40% or greater for approved antidepressants, the drug placebo difference is 10%. Yeah. If the placebo response is 30 to 40, it's 15%. If it's 20 to 30, it's 25%. So tell me what placebo effect you got in your trial, I'll tell you what effect size you're gonna get. So that's why the critical thing in trials is to manage effect size. In fact, in the Zurano trials, they did a terrible job managing the placebo response. Placebo response was extraordinarily high. When you look, there's a tiny difference, and that's why, in my opinion, they didn't get the indication. Thank you. Next. In a patient who is already on Synthroid and has a mid-range, has a TSH of four or five, would you consider thyroid augmentation, and how would you do it? Well, a TSH of four or five, it means that it's not well treated with the Synthroid, so thyroid augmentation could actually help the patient, but I don't consider four or five a good number for TSH. Okay, so you would push it? Well, I would talk to the endocrinologist and see if to optimize, and Arthur Prang would say maximize it with a combination of T4 and T3. Right, okay. Next. So, sorry, from online. Where in your hierarchy does an MAOI trial go? Two failures, three failures? Well, you know, the MAOIs have the problem of the diet. You know, I love cheese. I love some smoked salmon, so you're killing me. When you take those away, you know, you can get cottage cheese on a MAOI, but, you know, again, I would inflict a diet only after two failures, and not just one. Oh, so do you have, what is your first choice with depression and insomnia? Or if you have a worsening of insomnia and many awakening with trial with SSRI, do you switch or you add something? So, I would say in the US, the tendency for depression with insomnia is to use mirtazapine because of the anti-stimulant effects. On the other hand, you're from Europe, or? Oh, Canada. Canada, okay. Well, in Canada, I don't know, but in Europe, it's, I would say people use tricyclics. And so, drugs like clomipramine, very good for depression with insomnia, and may work at even lower doses. If you have a worsening of the insomnia and a trial with SSRI, do you switch or you add something? If you have a worsening of the insomnia, I would probably switch. And if the patient is having some benefit from it, I would switch to clomipramine because it's an SSRI with tricyclic structure and anti-histamine effects. Yes. Hi, my name is David Garion at Acacia Clinics. We do a lot of fly-in, five-day, st. TMS treatments, but also psychopharm. And I guess my question is, how do you now think about how to do interventional psychiatry? When to refer, how many medications and such? So the question is, interventional psychiatry is becoming very, very popular because when we're gonna see psychedelics, people are gonna have psychedelics, TMS, TDCS, ECT, ketamine. It's gonna be, at the end of the day, only trials that will compare one strategy to another will guide treatment decisions. Otherwise, it'll become more your own idiosyncratic preference. I'm gonna go with this over that. Thank you. Do you have any concern with adding bupropion dextromethorphan combination to SSRI given the 2D inhibitor improperties with regard to side effects? Not really, because you already, even if you combine 2D6 inhibitor, bupropion is a 2D6 inhibitor, so you would argue that it can be a very similar effect. Next. Hello, my name is Belinda. I'm from Australia. And I just had a question and a comment. So first question, any role or evidence for lamotrigine in treatment of unipolar depression? Lamotrigine, there is one study in resistant depression that was negative, showed no benefit. However, it's a little bit like a story that Dr. Koss mentioned. If there is a soft bipolar type, I will use lamotrigine because it is clearly effective in bipolar depression. And just a quick comment. I've got a couple of patients with ADHD. When they get an episodic depression, they actually do better on a slightly lower dose of stimulant for the duration of the depression because they sort of benefit from being a bit more hyperactive. And also, when they're depressed, the stimulant just makes them better at focusing on their depressed thoughts. So that's just a quick comment as well. Thank you. Next. Hi, thank you so much. My name is Rahab Khalifa. I'm a psychiatrist from England. And I just wondered if you can expand a little bit on the mirtazapine. We use it a lot. It's our monthly guidelines, first line of combination and augmentation. Sure. And also, I wondered your thoughts on the concept of broad spectrum antidepressant efficacy, if you can mention that quickly. Sure. Mirtazapine, it's kind of what I consider a dirty drug in the sense that it doesn't have just one effect. It's not just an inhibitor. Alpha-2 has a 5HC2 antagonism, 5HC3 antagonism. So if you look at pharmacology, dirtier drugs, meaning specific, sometimes have a broader spectrum of efficacy. Look at clozapine and psychosis. You look at anaphrenil. Clomipramine is a very, very broad spectrum efficacy. Now, anaphrenil is not approved for depression in the U.S. So it's used off-label, but it's approved in Canada. It's approved in Europe, et cetera. And it's a good antidepressant and good anti-anxiety drug. Thank you. From online, are there any studies or data concerning the use of alveolity in combination with ketamine or escetamine, giving their similar mechanisms? None that I'm aware of. Go ahead. I think I missed it, but the selegiline patch. What's that? Selegiline patch. I didn't hear you talk about that, if insurance will cover it. And you can eat cheese. It's true, at the low dose, yes. Selegiline, yeah, selegiline, it's a good, you know, monominoxidase inhibitor. And it's available in the patch in the U.S. I don't believe it's available in Europe. So I know there are some European colleagues here. But it's a good alternative, good point. Just to finish. Next. Going back to thyroid, sometimes it's not augmentation, but we're treating a mild type of thyroidism. I think there's a lot of controversy about what would be a cutoff for a normal TSH at baseline, I think a lot of reasons. What would you consider, at what point, if you see a baseline TSH of 3.5, at what point would you consider that a possible avenue of direct treatment? Well, if you're using a low dose of T3, there is really no cutoff because you're using it as an augmentation period. If you're using it to address a subclinical hypothyroidism, traditional view would be level of four or 3.5, may be suggestive. But at the end of the day, if you're treating subclinical hypothyroidism, I typically will consult with an endocrinologist. Thank you. And I think we have time for the last two questions for the folks in line. Thanks. So patient on ECT and lithium, what do you do with lithium? Okay, patient not responding to ECT and lithium. You get the patient, I do ECT, that's why I'm asking this question, that are on lithium, help with suicidality. Yes. I know your result was not very in favor of lithium. I was paying attention. I found it very good in people with some touch of bipolarity or hypomanic. What do you do with them? Because lithium and ECT. Yeah, someone who benefited from ECT and lithium, you wanna kind of add something, you know, the typical antipsychotics. So you switch? You could switch from lithium to the typical antipsychotic, yeah. Thank you. Thank you. Next, last question. Do you use much valproate for augmentation? Because I find the side effect profile better than lithium and you don't need the blood tests. Success with it? Great question. Valproate stimulates neurogenesis, you would argue should have antidepressant effects. We don't have much data on its efficacy and resistant depression. On the other hand, if it's a soft bipolar case, then it may be something to consider. Thanks everybody. Thank you.

treatment-resistant depression

Dr. Maurizio Fava

unipolar depression

bipolar disorder

pharmacogenetic testing

antidepressants

SSRIs

SNRIs

augmentation strategies

lithium

thyroid hormones

personalized treatment