All right. So, again, Samir Sheth. So the, you definitely appreciate the request from the scientific program committee to talk about this topic, trial design and functional neurosurgery. And, you know, I think it's certainly an appropriate issue to be discussing these days. You know, as we all know in our field, there's just more and more innovation, neuromodulation, DBS in particular, but, you know, all kinds of neuromodulation is certainly becoming increasingly utilized and new indications are being explored rapidly. You know, funding from Brain Initiative and NIH in general is really fueling this as well as just, you know, a lot of interest and obviously a lot of need from patients. So as we start exploring these, you know, indications, I think it's important to keep in mind, you know, how to design trials to show that something is effective. So what I'm going to do is, you know, it's just 15 minutes. I'm going to kind of focus on two areas. One is movement disorders and the other is psychiatric disorders. And they kind of, you know, it's an interesting point, counterpoint in terms of how these have developed. And it's a lot of DBS, but not entirely. So these are disclosures, not relevant for this talk, some funding sources. So, you know, the randomized controlled trial, of course, is sort of our workhorse and, you know, some may say gold standard for demonstrating that a therapy is effective. And, of course, in our field, medically, you know, medical treatments of a lot of the disorders that we treat, certainly the gold standard by which, you know, medicines are approved for movement disorders and epilepsy and whatnot. This is just kind of my, you know, abbreviated way of showing the design that I'll keep using it. So just to familiarize yourself with this, you know, the randomized trial, as shown here, is kind of comparing two different treatments. You know, there's some baseline period and then the randomization occurs. And you have some treatment over here compared to something else. And then an assessment that could be blinded at the end. So you could have a novel treatment against placebo to show that this treatment is, you know, better than nothing. Or, of course, if there is already an existing treatment, then it's a novel treatment compared to a standard treatment. So, you know, the standard design for these trials. You know, let's start off, because this is, you know, kind of functional neurosurgery in general, talking about, you know, pre-DBS era. And we'll just spend a few minutes talking about this. So lesions for movement and psychiatric disorders. The palatotomy, you know, of course, has been around for many decades and been very effective. But, you know, no controlled trials with randomization that have ever been done for palatotomy. There's a lot of open label series. And through that experience, of course, we learned that this really seems to be pretty effective. The closest we get is a couple prospective controlled trials, case controlled trials. This is one published a few, you know, a couple decades ago, where they took patients who were candidates for palatotomy that met the criteria. But some of them, for insurance reasons, you know, were not able to have the procedure done. So they just kind of followed these two cohorts. One that got the procedure and the other ones who were similarly matched, but of course not in a specifically controlled way, just to see how they did. And, you know, and this one, like the other open labeled series of palatotomy, you know, before them and among them, you know, show that there's a good effect from the procedure. With psychiatric disorders, you know, we have for OCD, for example, or depression, capsulotomy, and cingulotomy. This is just a review that we wrote a few years ago, which, you know, at the point of this writing a few years ago, there was no controlled trials or certainly randomized double-blinded trials for the psychiatric literature. A lot of open labeled series of capsulotomy and cingulotomy and whatnot. That's changed recently. There's a very nice randomized trial for gamma knife capsulotomy, which I'll get to in a bit. And so again, this is just my little kind of graphical nomenclature of, you know, it's a line with an assessment at the end. So now let's just get into DBS. This is where, you know, we spend a lot of time thinking about this. And so, of course, the opportunity with DBS is that you can turn a device off and potentially you can, you know, have a sham procedure or certainly you can implant it and just not have it on and randomize people to on or off. And so a lot more interesting trial designs have been done with DBS. And so again, starting with movement disorders, and obviously, you know, so in the states, of course, we have had the approval now for 30 years. Is that right? 20 years? 20 years. And so, 30 years. Yeah. Thank you. And, you know, we've had a lot of experience, a lot of time to, you know, to have an understanding of what the therapy is before we have instituted trials. So this is, of course, one that you're all familiar with, published in 2006. Dushal et al. And, you know, there's a baseline period, there's a randomization, which in this trial occurred in pairs that took patients two at a time and randomized them to DBS or best medical therapy. And the assessment at the end was unblinded. The early stim trial, similarly randomization to DBS versus meds, and there was a blinding at the end of this one so that the raters at the end did not know whether the patient was on DBS or on best medical therapy. And then, you know, many others similar to this where we're now kind of, you know, delving in deeper to look at differences between different targets, and STN versus GPI. And so this is one, again, that looked at STN versus GPI therapy, the blinded assessment, a single blinded assessment at the end. And so, you know, again, over the course of the decades, we've had a lot of experience understanding the therapy, DBS for Parkinson's disease. And, you know, these trials have certainly demonstrated, as we all know, a high level of efficacy. And so where are we with, you know, with these trials and with this therapy? The efficacy of DBS for motor symptoms and, you know, levodopa-induced dyskinesia is well established. And now we're looking into further details regarding STN versus GPI and other, you know, things like that. The CNS and, you know, ASFN and AANS joint guidelines group just published this article just recently establishing, you know, some guidelines for STN versus GPI DBS. And so now, you know, this, of course, like I said, has all been done after a worldwide experience of thousands of procedures. Once the therapy is understood, it's just a lot more effective, perhaps, to design a trial that can really test it against different therapy-like medications. Now, with psychiatry, things are very different. You know, the first report of DBS for OCD was from the Belgian group, Bart Nutan, in 1999, four patients. And then, you know, there's a few open-label series. One of the larger ones and the one that is very relevant for the approval here in the U.S. was Ben Greenberg and that group, which has sites in the U.S. as well as in Belgium, 26 patients and, you know, 58% response. It's an open-label series. It's just, you know, the sites more or less did similar things, but there were even differences between the sites. Open-label, and they followed them out, and even kind of changed some of the, you know, ways that they were targeting over the course of the many years of the study. And so, the experience, you know, at this point, you know, is dozens of patients, really, and certainly a lot less than we had with Parkinson's, you know, 10 years into DBS for Parkinson's. So, the difference in the story here with the psychiatric disorders, though, of course, is that, you know, we started jumping the boat, I think, and kind of, you know, moving towards these trials a lot sooner than we had in Parkinson's disease. So, this is a study from the French group where they did DBS targeting the ventromedial portion of the STN as opposed to the typical dorsolateral for movement disorders and targeted this area for OCD. 16 patients is a very interesting, you know, and well-designed study. There's implantation, and there was a blinded crossover phase. So, on versus sham in a double-blinded fashion. And so, the graph, the figure shown here in the off state, they're not doing as well. This is the Y axis here is the Y box, you know, symptom score, and smaller is better. And so, in the on phases, you know, the patients do well. And so, this, you know, this showed efficacy of STN DBS. It hasn't really been reproduced, you know, or taken much further. And the cutoff for response rate was a 25% reduction as opposed to the typical 35%. So, there's a couple asterisks, but in general, this is really the only class one evidence that we have until recently of DBS for OCD. This was the level of evidence paper, again, by CNS, double NS, and ASFN. I'm saying, again, class one evidence for STN DBS, really just class two evidence for the target that we have the hematurin device exemption approval for here in the U.S., the VCBS target. Again, so all this background led to an NIMH-sponsored randomized trial led by the Brown Group, which has finished enrollment and is not yet published. And, you know, the analysis is going on right now. So, we'll see what that, you know, double-blinded randomized trial looks like. This is really comparing active to sham, and is the only such trial that, you know, we really have here in the U.S. So, I won't say much more about that because we don't really know what those data look like. If we look at lesions, again, this is worth showing because this is the, really the only existing blinded sham-controlled radiosurgical, you know, lesion trial for psychiatric disorders. This is from the San Paolo Group in Brazil, and they took 16 patients and randomized them to active versus sham, and they even, you know, built this sham chamber so that, you know, the patient really didn't know whether they were getting treatment or not. And the assessments were double-blinded. I won't, you know, spend too much time on the results, but they missed showing a significant difference in the primary outcome, which is the number of responders, but just by a very little amount. And it's an interesting couple of details that, you know, there are two out of eight versus zero out of eight responders at the 12-month blinded phase, after which they unblinded the patient. But just a few months later, you know, at months 14 and 18, two of those non-responders became responders. So, speaking to the fact of, you know, the amount of information going into the trial is really going to have an important impact on, you know, whether we see, you know, whether a therapy is successful or not. And certainly you could ask the question that, well, if the blinded period had been, you know, 16 months or something like that, you know, would this have been a very different type of outcome? But despite that, you know, other metrics, secondary metrics, as in Y-box reduction certainly was a lot more prominent and significantly different in the active versus the sham group. Switching gears to depression, again, several open-label trials. You all know the history, but, you know, the VCVS target came out of the finding in the OCD studies that comorbid depression improved. And so this target has been explored for depression as a primary indication in the subgenual cingulate target. Initially started by Dr. Lozano in Mainburg in Toronto and then continued largely at Emory. Again, you know, a lot of open-label series that kind of went into both of these targets that then fueled two large randomized trials, which, you know, were conducted a few years ago and results published in the past couple of years. And they're very instructive, so it's worth spending a couple of minutes on them. The reclaimed trial was the VCVS target. Planned 208 enrollment, but, you know, 30 were enrolled at the time when the sponsor decided to have a look. This was the design, you know, four weeks after implantation they were randomized to active versus sham for 16 weeks and then unblinded and then followed open-label after that. And as we know, essentially there was no difference. There's no significant difference between sham, 14 percent response rate versus active. And so, you know, this goes to show, you know, the placebo effect that occurs in these types of studies is something very important and we have to really consider that in our design. Even the open-label kind of follow-up after that, you know, not really blow you away type of responses. And so there's, you know, several ideas about why this differs so much from what we've seen in the open-label series. Some of the, you know, kind of take-home points from this trial, certainly at 30 it was not powered to see a difference, but that's the point when the sponsor said, look, we don't like how this is going, so the trial stops. Possibilities, I mean, 16 weeks is not a long period of time, especially for OCD, which we know now really takes a while. It can take months for patients to get better. An important one here is also parameter exploration during these double-blinded phases oftentimes is limited so that, you know, there's some standardization. And so when we're using movement disorder type parameters to treat OCD, you know, are those really the right stimulation frequencies and pulse widths that we should be using? Certainly don't know that yet and if you're really limiting, you know, what you're able to explore during this blinded phase, you may not, you know, reach that even local minimum, forget a global minimum in terms of your improvement. You know, placebo response rates as expected, so these have to be considered, but the active response rate, of course, was less. So, you know, a longer double- blinded phase. Other trial designs, which I'll talk to in a minute, and other targeting strategies. Broaden was published a few months ago, several months ago. Again, a planned interim analysis of 90 patients out of 201 that were, I mean, planned at 90 out of 201 that were planned to be implanted. Similar, you know, overall design, implantation followed by randomization for six months, you know, active versus sham. And again, strict programming restrictions during the double-blinded phase. And again, as we know, no difference between the sham and active groups, percentages shown here. The difference here is the open-label follow-up certainly had patients, you know, up to half the patients who did get better, but, you know, this is the double-blinded phase, again, no difference. And so, you know, similar sorts of lessons, so we need time to understand the therapy, you know, before we institute a very, you know, kind of strictly regulated double-blind randomized trial with, you know, active and sham arms. You know, if you guys were here in the plenary session this morning, the intrepid study of DBS for Parkinson's was, you know, the data were announced, and really that's the first time even in DBS that we have, you know, active versus sham comparison. And, you know, we're already trying this in psychiatric diseases without really understanding the therapy nearly as much in the parameter space. You know, how long does it take for the effect to mature so that we see a response? I mean, this takes kind of longer in many psychiatric diseases than in movement disorders. The parameter space, like I said, you know, using Parkinson's parameters to, you know, program ferocity and depression may not be the right thing, and accounting for, like I said, placebo and other effects. The last two studies I'll mention are just different trial designs, which, you know, are kind of interesting food for thought regarding should we be doing this maybe in a different way for certain disorders. This is one that came out of the, again, Bart Newten's group in Belgium. OCD implantation followed by, instead of an immediate randomization, followed by an open-label optimization period of many months where they had open-label, you know, access to adjusting parameters and seeing what could maximize therapy, you know, for many months and up to a couple of years. And then after they had established a certain level of therapy, then they were randomized in a double-line fashion to an on-off crossover with a follow-up after that. And to summarize the results, they found a very significant difference between on versus off. And so, you know, very similar therapy here, targeting differences, et cetera, but a different trial design and we see a very different outcome. And again, for depression, here's a study out of the Amsterdam group. Same sort of thing. Implant, a year of open-label optimization followed by the crossover. So these, you know, and the results here, again, show a significant difference, again, as opposed to the two trials in the U.S. So brings up the question of, you know, these alternate trial designs are something to consider. They, of course, have their own limitations. You know, randomizing someone to a therapy at the beginning is not the same as giving them therapy for a long time and then taking it away. But, you know, it's an interesting discussion. So final thoughts, you know, again, 30-plus years of DBS and, you know, 100,000-plus patients before, you know, we see the first double-blind active versus sham RCT. On the other hand, you know, with psychiatric disorders, it's been a much shorter timeline. And then, you know, this design of an upfront active versus sham comparison may not always be feasible, may not always be the right thing, depending upon the indication we're trying to treat. And, of course, if we are thinking about other designs, we're going to have to work with the FDA and IRBs, et cetera, to see what their thoughts are regarding, you know, the possibilities of whether that's going to be accepted as high-enough level evidence. So thanks very much. And I would take questions if we have a couple minutes. Thanks very much. I think we have time for one question. I mean it's some design where you don't have any sham arm or you don't yeah Yeah, well, there's other designs where you don't I mean there are you know the Florida group published their their experience with a Blinded staggered onset so you know everybody gets the therapy They just get at different times, and you don't know exactly when that is so I mean There is a period of time when patients not on therapy So it's it is off for that amount of time, but it's not a you know lengthy kind of period of time later So there's that's another design, so there are designs that don't Have sham even these I mean I maybe I've missed you know saying it But of course I mean with the sham arm at the end of that of course they receive Therapy you know in the open label after that, but there is a period of time So yeah, there's other designs where you don't necessarily have as lengthy of a period of time sham All right, thanks very much Samir

trial design

functional neurosurgery

DBS

movement disorders

psychiatric disorders