Hi. I'm Emily Lenard. I'm a clinical research coordinator in the Healthy Mind Lab at Washington University School of Medicine. Today, we'll discuss treatment-resistant depression in older adults and present findings from the OPTIMUM study, the largest ever clinical trial for treatment-resistant depression in older adults. You'll be interested in this seminar if you see older adults with depression and want an update on antidepressant treatment. Our speakers today will be Jordan Karp, professor and chair of psychiatry at University of Arizona College of Medicine. Dr. Karp will present an up-to-date review of the literature on late-life treatment-resistant depression and the rationale for the OPTIMUM study. Helen Levretsky is professor of psychiatry at UCLA. Dr. Levretsky will speak next about the preliminary effectiveness outcomes from the OPTIMUM study. And she'll be followed by Eric Lenz, professor of psychiatry at Washington University School of Medicine. Dr. Lenz will present preliminary safety outcomes from the OPTIMUM study. After the three talks, I'll lead us in a question-and-answer period. Thanks, Emily, for that introduction. I'm Jordan Karp, professor and chair of the Department of Psychiatry at University of Arizona, and I'm going to provide an overview of the OPTIMUM project. So my disclosure slide. And I want to remind the audience about the consequences for late-life treatment-resistant depression. Late-life treatment-resistant depression is really the norm, not the exception for older adults, because over 50% of older adults don't respond to first- or second-line antidepressant pharmacotherapy. We know that persistent depression, probably more than any other condition in late life, decreases older adults' quality of life. And untreated or undertreated depression is a leading cause of disability, excess mortality, both all cause as well as related to suicide and cognitive decline. Indeed, both decreased physical activity or exercise and depression are the two biggest risk factors for cognitive decline and dementia in older adults. Well, I'd like to begin the talk by talking about some process challenges to the effective treatment of late-life treatment-resistant depression. Now, I think that the audience is probably aware of the limited geriatric mental health workforce in North America. There's less than 2,000 board-certified geriatric psychiatrists in the United States, less than 700 geropsychologists. Generalist mental health clinicians often have a limited understanding of how cognitive decline and medical illness mutually exacerbate depression and how they can effectively intervene in this vicious cycle. PCPs, who provide the majority of depression and anxiety care to older adults, often prioritize, quote, medical conditions like diabetes, high blood pressure, cancer, and prevention efforts over psychiatric conditions like depression. We need to also acknowledge that frequently clinicians experience clinical futility and therapeutic nihilism in caring for these patients, which can lead to some inertia in the clinical care because there may be thoughts that, well, these patients are just not going to get well. They're too difficult to take care of. Finally, we also need to acknowledge that primary care physicians often have discomfort with managing complex mental illness and may not know when to refer or how to get better care for their patients. And at the payer level, payers are still focused on risk, not patient and family-focused care that embraces value. These patients are often difficult to take care of. They require between-session attention, which is often not reimbursable. So that's another challenge to getting these patients effectively treated. Well, despite these process challenges, treatment-resistant depression and TRD in late life are ubiquitous. And this is work that was really led by Dr. Levretsky and published in 2016, describing the workforce or the TRD burden in psychiatric and primary care. And what she'd found was that of primary care physicians and geriatric psychiatrists and general psychiatrists who responded to surveys from Psychiatric Times and from the American Association of Geriatric Psychiatry, about 40% described that 26 to 50% of their practice was dedicated to taking care of treatment-resistant depression. And among their treatment-resistant depression practice, the amount of time that was dedicated to late-life treatment-resistant depression was for about 60% of the respondents described as about 51 to 75% of their practice. So quite a burden and quite a large amount of time is spent taking care of these patients. Well, let's move on to the state of the science and this question of when somebody has a difficult to treat depression, they haven't responded to two antidepressants, maybe there's partial response, maybe there isn't. Do you switch or do you augment? Do you add on another medication? And that's really the goal of what, one of the goals of that the Optimum Project is trying to address. But before we talk about switching versus augmentation, we need to review some principles of pharmacological treatment of late-life depression. Some pearls to remember are that late-life or late-onset depression often has a worse prognosis, a more chronic course, a higher relapse rate, certainly is associated with higher levels of medical comorbidity as well as cognitive impairment and mortality. Again, all cause mortality as well as suicide. We know this mantra, start low, go slow, but go all the way. And indeed, we need to recommend or encourage our colleagues to go all the way because this is a frequent contributor to pseudo-treatment resistance in late-life depression because physicians and psychiatrists are hesitant to push an antidepressant up to a therapeutic dose. We do require a greater degree of vigilance in taking care of the old, old, in particular those older than 75 because of concerns about falls or pharmacodynamic or pharmacokinetic interactions. It's really not related to the black box warning of an increased risk of suicidal ideation with antidepressants because if you read that black box warning more carefully, actually there's a reduction in suicidal ideation with antidepressants in older adults. But we do know that there are pharmacokinetic changes of aging that may make treatment with antidepressants more complex. So there's a decreased rate of absorption likely related to changes in gut motility. There's increase in bioavailability likely related to changes in first pass metabolism because of decreased blood flow and smaller liver volume. There's an increase in the half-life for lipid soluble drugs and the relative concentration for water soluble drugs and metabolites. We also, and I'm going to keep reiterating this, know that medical burden and polypharmacy increase the risk for pharmacokinetic and pharmacodynamic drug interactions. So both metabolism and effect. There seems to be some serious adverse events that are more common in late-life depression. So antidepressant related bone loss, serotonin syndrome, extrapyramidal side effects, and neuroleptic malignant syndrome. We also need to be attending to falls, hyponatremia, and gastrointestinal bleeds in older adults. Finally, there is the risk of QTC prolongation, in particular with citalopram. Meta-analysis suggests that late-life depression requires longer treatment trials. So what I want you to take home from this slide, there's a lot of information here, is that we can get people well, but it often takes longer in late life and late-life depression. Patients are often at greater risk of recurrence or relapse if the treatment is stopped. Well, there are expert guidelines and clinical algorithms for the treatment of late-life depression, but they're getting a little stale. So the United States published some guidelines in 2001. The Canadians published some guidelines in 2006. I like this updated algorithm that came from the Canadian Geriatric Psychiatry Association and published in 2014 that described some stepwise approach for initiating treatment and trying to help patients who have more difficult to treat depression. There's this question of for those with minimal or no response, they recommend switching to nortriptyline, maybe bupropion. If there's partial response, they recommend augmenting the antidepressant with lithium or an atypical antipsychotic, or using combination approaches by adding another antidepressant. But again, the precise recommendation is not clear. So they suggest six weeks of duration, but maybe four weeks, maybe eight weeks. It depends on treatment response. It really doesn't guide a precise approach for what kind of treatment, switching versus augmentation for what kind of patient for how long. So this brings us back to the question of switching versus augmentation, when and for whom. So the general accepted notion is that for non-responders to a SSRI or a serotonin norepinephrine reuptake inhibitor, from younger adults, we know that there's evidence that two psychotropics are superior to monotherapy, but the evidence is not as strong in late life and concerns of polypharmacy is greater. For partial responders to an SSRI or an SNRI, there is preference for augmentation. So adding a non-FDA approved antidepressant or combination, adding another FDA approved antidepressant may be consistent with the 2001 U.S. guidelines. And it's also consistent with the Canadian guidelines that we may lose any partial improvement by switching to something else. Well, this brings us to the Optimum Project. And Optimum is funded by PCORI, and part of PCORI's mission is that stakeholders are involved in the inception, design, implementation, and dissemination of results. So we asked our primary care physicians, psychiatrists, and geriatric psychiatrists, what sort of things they would like us to test and would such a clinical trial be useful? And again, this was also led, again, by Dr. Levretsky. So we asked, how helpful would you find the results of a large randomized study that compares the risks and benefits of augmentation to switching strategies for TRD in patients 60 and older? And this was unanimously thought to be helpful. We also asked, what treatment choices would you like to see randomized? So do you want to know if we should augment with aripiprazole, augment with bupropion, augment with lithium, or switch to nortriptyline, or switch to bupropion? And in general, these were the approaches that our psychiatrists and PCP colleagues wanted us to test. And we asked, do you think your practice will benefit from having evidence from a large RCT like this? And again, the answer was a unanimous yes. Well, we also asked depressed older adult stakeholders how we should design the study. And they recommended that they get their treatment from their physician of record, from their own psychiatrist or PCP, but that the research team, the expert geriatric psychiatrists provide decision support. They wanted us to assess psychological well-being as an important outcome in addition to depression measured with the MADRS, the Montgomery Asperg Depression Rate. And they wanted us to assess Montgomery Asperg Depression Rating Scale. Safety is important to older adults. And so they wanted us to assess fall risk as an important measure of tolerability and safety. As I mentioned earlier, depression is a risk for cognitive decline and dementia, and cognitive decline is a tremendous worry for older adults. So they wanted us to also assess cognition and mobility to see if this helps to predict outcomes. And they also wanted us to collect DNA to see if there were genotypes that may predict outcomes. So there are three aims to the Optimum Project. And I'll describe these as themes. The first are to assess antidepressants, benefits and risks in older adults. And I think that's something that makes Optimum really unique because not only are we assessing improvement in depression and well-being, but we're really focusing on safety and tolerability with a focus on falls. So we're trying to determine which antidepressant treatment, a switch versus augment strategy, leads to improvement in well-being and symptomatic remission, and which to medical complications and fall-related injuries. Theme two is how does aging change this balance of benefits versus risks? And these, so we are assessing what the effects of aging on brain and systemic health and how this interacts with treatment response. And we are stratifying on patients who are young, old, so those who are 75 and younger, and those who are 76 and above. The third theme is to maximize stakeholder engagement. And as I said, this is a requirement and part of PCORI's mission is to engage stakeholders. So from the inception, development, implementation, and now we are in the dissemination phase, we are giving voice to our stakeholder colleagues, including families, fellow physicians, primary care docs, and payers and policymakers. Let me show you what the optimum study design looks like. Our goal was to recruit 1500 people at five sites across North America. We will talk about what we, who, the number that we actually recruited. All patients had to be in a current episode of major depression and have tried and failed at least two antidepressants of adequate dose and duration. Now, some patients came to us, they were still depressed. They had tried at least one antidepressant, but they weren't currently on an antidepressant. So we would make a recommendation to their physician to start something, get them on it. And then we followed up with them in six weeks to see if they were still depressed. And if they were, then they could come into the study. We first randomized these participants to either augmentation with aripiprazole dosed at two up to 15 milligrams a day, augmentation with bupropion dosed at 150 up to 450 milligrams a day, or switched to bupropion at the same doses. And we treated them for 10 weeks. But we recommended treatments to their primary care physician or psychiatrist. We provided decision support. They remained with their physician of record. Patients who did not respond had the opportunity to move into step two, where they were randomized to augmentation with lithium. And we dosed that to plasma levels of 0.4 up to 0.6 nanograms per milliliter, or they were switched to nortriptyline with the plasma level goal being 50 to 150 nanogram per milliliter. We tried to make it even tighter with the target plasma level 80 to 120, if we could get there. Not just remitters, but all comers who were randomized, we then followed for one year to assess treatment durability, relapse recurrence, and our other important outcomes of safety, falls, and well-being. So as I said, there were five sites in North America, Los Angeles, Western Pennsylvania, and Pittsburgh, New York City, Columbia was the site, Toronto and rural Ontario, the University of Ontario and CAMH was that site, and Washington University in St. Louis. And this was led by Dr. Lenz, who was the coordinating site. So I want to come back to the medications that we used. Again, clinicians recommended testing nortriptyline, bupropion, lithium, or an atypical antipsychotic. This is somewhat aligned with the CANMAT 2016 clinical guidelines with special populations focused on the elderly. But still, there's some questions here. So here, CANMAT is saying second line is switched to nortriptyline. The other Canadian guidelines I showed you a little earlier said that this was for step three. And then for their next step was to switch to amitriptyline or imipramine. Actually, earlier they recommended phenylzine or catiopine. There's still a disconnect about what is the safest and most effective treatment for older adults with treatment-resistant depression, switching versus augmenting, and that's what we're trying to determine. So again, clinicians recommended these approaches, augmenting or switching aripiprazole, bupropion, lithium, or tryptophan, or switching to bupropion. We also decided to use aripiprazole because of work that we published in The Lancet in 2015. Again, Dr. Lenz was the leader of this project where we compared for older adults with treatment-resistant depression, is aripiprazole superior to placebo during 10 weeks of augmentation pharmacotherapy? Again, dose from 2 up to 15 milligrams a day. And we found that it was superior to placebo with a 44% versus 29% response rate and a number needed to treat of 6.6. It also seemed to be safe. So compared to placebo, there were higher rates of akathisia and Parkinsonism in those receiving aripiprazole, but there was no greater incidence of increase in fat, insulin or glucose changes, or QTC prolongation. So we wanted to know where does aripiprazole from the class of atypical antipsychotics fit into this switch versus augmentation strategy? So again, our goal is to clarify the use of these commonly used medications. Tricyclics, there is evidence for their efficacy and safety. We wanted to use nortriptyline because it's less likely than the tertiary amine tricyclics recommended by the CANMAT guidelines for treatment-resistant depression in older adults. Nortriptyline is less likely to cause orthostatic hypotension falls and anticholinergic side effects than imipramine or amitriptyline. We wanted to test bupropion, both the United States and the Canadian guidelines recommended as second-line treatment, but it has not been assessed in a published trial of late-life depression. And it's not supported by STAR-D in which adult patients, not so much older adults, switch from citalopram to bupropion, sertraline or venlafaxine did not differ in outcomes, tolerability or adverse events. So we need to know how this fits into the algorithm. And finally, lithium. We need to know how safe it is, especially in the old, old, and we wanted to replicate other studies, which of all these augmentation strategies, lithium has the most replicated data for any agent in late-life depression. So I think I've given you a pretty good oversight or overview of the project. I want to just share with you some next steps as I wrap up my talk. So Optimum has provided a really unique infrastructure for mechanistic and add-on clinical trials, and we've been successful in competing for funding to ask some really important and interesting questions to improve outcomes for older adults. So the first is the Optimum Neuro project, which is funded by NIMH. It's still ongoing and it's asking how does late-life depression accelerate cognitive decline and lead to dementia and does successful treatment mitigate that risk? And we're able to use this large platform of older adults with TRD to ask these questions. The second is the Optimum Bone Health Study, which is funded by NIAMS, and the goal is to look at the long-term effects of serotonergic medication exposure on bone health. The third is the Optimum Neurosuicide Supplement funded by NIMH, and we're able to study some cross-sectional and prospective associations between cognitive impairment, because we're doing the Optimum Neurostudy, and suicide risk in this population highly vulnerable for both dementia and late-life suicide. And finally, we're hoping and hopeful for the next clinical trial for older adults who have not responded to first or second-line pharmacotherapy to test how IV ketamine plays a role for both safety as well as efficacy. We submitted this once and we were scored and we plan to resubmit this again later this year. So in summary, we're hoping that there will be four main outcomes of the Optimum project. First, clinical. We're hoping that it will answer questions about the safest and most effective treatments for late-life TRD. Precision. We want to know what's the right treatment, switching versus augmentation in the right patient at the right time, and we're hoping that an important dissemination outcome will be guidelines for older adults with treatment-resistant depression. These ancillary studies are helping us to better understand mechanisms, so about late-life treatment-resistant depression, cognition, bone health, and suicide. And finally, dissemination, which is so important for PCORI-funded projects. Our dissemination plans focus on the scientific community, so peer-reviewed publications and presentations, stakeholders, so direct-to-consumer patients and families, health advocates, such as the American Association of Geriatric Psychiatry and the American Geriatric Society, and payers, such as CMS and other commercial insurers. This has been a large team effort, and I want to thank our collaborators at the other sites, as well as our PCP and psychiatrist partners, our patient participants, and our stakeholder advisory board. And I leave my email up here for a brief time, if people want to write this down. I'm happy to answer any questions. People can contact me for further information about the Optimum Project. Thank you. I'd like to now introduce my colleague, Helen Levretsky, who is the PI for the Los Angeles site. She is a professor at the UCLA School of Medicine. My task today is to describe the initial analysis based on demographic and clinical characteristics and effectiveness outcomes. Next slide. So this is my disclosure slide, and I'll go through some of the recruitment lessons that we learned early on and present data on demographics, psychosocial, clinical, and some of the results on effectiveness from Phase I and Phase II. Next slide. So the sources for recruitment were primary care settings with our colleagues in primary care, from mental health services using Los Angeles City and County mental health resources, New York City and Metro, Pittsburgh and surrounding counties, St. Louis and rural Missouri, and Toronto and rural Ontario. And all representatives of the community dwelling North America, we were hoping to represent minorities, and we were partially successful in our numbers of recruiting minorities. Next slide. So the inclusion criteria for the study were age 60 and above, DSM-5 diagnosis of major depressive disorder, a PHQ-9 score of 10 and above. Previous treatments was at least two antidepressants with an adequate dosage for at least four weeks. And previous antidepressants needed to be different from the medications tested in the optimum protocol. The primary outcome was the symptoms of depression assessed by the Montgomery-Asberg Depression Ratings course of less than 10 after 10 weeks of treatment. Next slide. So the study protocol included randomization into augmentation with aripiprazole, bupropion, or switch to bupropion. And lasted for 10 weeks acute phase of treatment. If they responded, they would continue receiving the same medications, but if they didn't show full response or partial response, they would be randomized to augmentation with lithium or switching to nortriptyline for another 10 weeks. Next slide. So the hypothesis included effectiveness hypothesis. We were comparing not a simple antidepressant or individual antidepressants, but rather strategies. Augmentation arms will show greater improvement than monotherapy arms for effectiveness outcomes. And the improvement was expected to be in depression, but also in psychological well-being. A second hypothesis reflected safety. We were hoping to see that augmentation arms will show greater tolerability and safety concerns than monotherapy arms and measured by serious adverse events, false and fall-related injuries. And then we were going to examine moderators of effectiveness and safety differences between treatment arms that were used that included age demographic and psychosocial variables and also comorbid medical conditions. Next slide. So this is the analysis of baseline step one and step two cohorts by ethnicity, race, gender, age and marital status and education. So the majority were white, non-Hispanic, female in the younger group, 60 to 69 and mostly married or separated. And education, 15 years of education in both groups. Next slide. I summarized the table. So participants were predominantly female with high burden of medical comorbidities. And about one third had a history of falls preexisting prior to this study. And 8% have injured themselves from a fall. Baseline depression severity was moderate and the history of two or more failed antidepressant trials was consistent with our definition of treatment resistance. Most patients endorsed suicidal ideations at baseline and those were assessed as low risk. Next slide. So we screened over 6,000 patients, 1,045 were eligible for the study, 884 consented and 743 were randomized. Next slide. So how we supported antidepressant treatment, it was based on the model that included decision support calls. And we called study participants every two weeks, assess their depression symptoms, suicidality and safety, and this information and recommendations were shared with participants, physicians, either by emails or phone calls or personal contact. And if depressive symptoms were minimal or absent and the medication was well tolerated, no changes were recommended. If PHQ-9 score remained high and the medication was well tolerated, we suggested that dosage increase. When problematic side effects were encountered, it was led usually to dose reduction or treatment discontinuation. It was possible early progression to step one, step two, depending on where in the treatment timeline they were found at that time. During every decision support call, participants were also asked about their adherence and those who endorsed non-adherence or their physicians decided to change their medications. They were provided with brief counseling and also this decision changes were discussed with the physicians to help overcome barriers to antidepressant adherence and ultimately a sufficient response. Next slide. So I'll present just briefly the results of the effectiveness outcomes for both steps, phase one and phase two. Next slide. As I said previously, 621 subjects were randomized into phase one, which included augmentation with Aripiprazole 212, received that. Augmentation was Bupropion 206 and 203 switching to Bupropion. And we considered outcomes of symptomatic remission while being safety, false cognitive function and mobility. Next slide. So here you see that both Aripiprazole augmentation and Bupropion augmentation arms did pretty similarly and 68 and 66 subjects in each arm achieved remission. It was lower count in switching to Bupropion 46. Next slide. In the second step that also continued for two weeks, 127 were randomized to augmentation with Lithium 124 to switching to Nodriptyline. And those who completed both acute phases in step one and step two successfully were followed in the one year continuation later on. But here we're reviewing acute results here. Next slide, please. For the step two and Lithium augmentation resulted in 27 subjects responding and switching to Nodriptyline equally to 26 subjects responding. Next slide. So here what we observed remission outcomes favored augmentation arms. And interestingly, side effect profile also was not much different for both arms in step one. So overall the study really responds as Dr. Karp said to the needs of clinicians and patients to make decisions based on evidence and account for both benefits and risks of treatment options. Next slide. So that concludes my talk. Hi, I'm Eric Lenz. I'm the Renard Professor of Psychiatry at Washington University School of Medicine. I'm gonna present the safety data from the optimum study regarding the treatments in both step one and step two. Next slide, please. This is my disclosure. Next slide. Reminding you that hypothesis two regarded safety and our hypothesis was that the augmentation arms in both step one and step two would show greater tolerability and safety concerns than monotherapy arms, both in terms of serious adverse events and in falls and fall-related injuries. Next slide, please. So just reminding everyone what a serious adverse event is. Defined for optimum, a serious adverse event is defined as death, a life-threatening event, hospitalization, an event that required intervention to prevent impairment or a disability or incapacity. This does not include all adverse events then. Typically, these are only adverse events that led to a hospitalization. The numbers I'm going to provide are raw frequencies and rates and doesn't account for covariance associated with outcomes like medical severity or depression severity. And not necessarily attributed to study medication. And in fact, we know that if you do a clinical trial in older adults, many will have hospitalizations as a result of their pre-morbid medical conditions. And I'm only presenting the data as did Dr. Labretsky on the acute phases of step one and step two, not the 12-month follow-up. Next slide, please. Just reminding you, starting with step one, we randomized 621 people about equally to augmentation with aripiprazole, augmentation with bupropion, and a switch to bupropion for 10 weeks. Next slide, please. So what you see here are bar graphs showing the absolute numbers of serious adverse events associated with each of those randomization choices in step one. Starting on the far left, the gray bar shows that 7.9% of people randomized to aripiprazole augmentation had a serious adverse event compared to in the middle 7.7% of those randomized to bupropion augmentation. And then on the right, compared to 10.5% of those randomized to bupropion switch. Okay, next slide, please. Just taking a look at the various types of adverse events in each of the randomization arms, at the very top, psychiatric hospitalization was a relatively uncommon cause of serious adverse events with four psychiatric hospitalizations total during the acute phase out of those over 600 individuals. In contrast, non-psychiatric hospitalization was a much more common cause of a serious adverse event. And looking at the middle there that's outlined in red, we're looking at serious adverse events related to a fall. And there were actually three falls that led to mortality, one in each of the randomization arms in this study, which highlights the serious issue of falls in older adults. And in particular, depressed older adults were taking antidepressants. What you can see there is numerically, there were more adverse events, serious adverse events due to falls in the bupropion augmentation group compared to the other two arms. Next slide, please. Moving on to our falls analysis, what you can see on the top right of this slide, the pie graph is that within the acute phase of step one, one quarter of individuals had a fall. So 155 people reported one or more falls, 466 reported no falls during those 10 week acute step. And what you can see in that histogram on the left is the fall frequency. You can see again that most people did not report a fall, 75% did not report a fall, 15% reported exactly one fall, 10% reported two or more falls to a maximum of six in that step. Next slide, please. With respect to fall related injuries, 82% of the sample had no injurious falls. Another way to think about it is that 70% of the people who fell did report an injury of some type, although only 5% of the people who fell were considered a serious adverse event. And you can see in the histogram there, the distribution of fall related injuries. Next slide, please. Here we see the falls, the number and rates of falls by randomization arm. Starting again on the left, the erypiprazole augmentation arm, you can see that 25% of those individuals reported a fall, 17% reported an injurious fall, 9% reported multiple falls. In the middle, you can see the bupropion augmentation group, slightly higher percentage, 29% reported a fall, 21% reported an injurious fall, 15% reported multiple falls. And finally, in the bupropion switch group on the right, 26% reported a fall, 18% reported injurious, and 8% reported multiple falls. Next slide, please. Now I'm gonna move on to step two, which again, to remind you, that randomized 127 people to augmentation with lithium and 124 people do a switch to nortriptyline in that 10 week step. And again, we looked at serious adverse events, falls and fall related injuries. Next slide, please. What you can see here is the same rate of serious adverse events in both of these randomization arms. So of those randomized to lithium augmentation, 9.8% reported a serious adverse event versus 9.2% of those randomized to a switch to nortriptyline. Once again, I wanna point out these serious adverse events aren't necessarily due to the medication, but are just reported in people randomized to that treatment arm. Next slide. Here's another breakdown of serious adverse events now in step two. At the top, again, you can see that psychiatric hospitalizations are a relatively infrequent cause of serious adverse events compared to non-psychiatric hospitalizations. And in the middle outlined in red, you can see that serious adverse events due to falls are approximately equal one versus two in this phase, in this step. Next slide, please. What you see here, starting in the pie graph on the right is the percentage of people who reported one or more falls, which was a 21% of the sample. And then in the histogram on the right, you can see the distribution of falls. 79% reporting no falls, 14% reporting one fall, and 7% reporting two or more falls. Let's go to the next slide, please. And then with respect to fall-related injuries, overall, 85% had no injurious falls, but 70% of those who fell reported an injury, 6% of those who had a fall considered it a severe adverse event. And the histogram here, again, shows the distribution overall of fall-related injuries in that step, step two. Next slide, please. And here we see the rate of falls by randomization arm. And you can see that they're essentially equal in the lithium augmentation arm on the left, 22% had reported a fall and 17% reported an injurious fall, 7% reported multiple falls. In the nortriptyline switch arm, 20% reported a fall, 13% reported an injurious fall, 8% reported multiple falls. Next slide, please. And this is my final slide here in the final slide of this talk before we go to question and answer. From the safety data, we can make some conclusions. First of all, looking at the overall serious adverse event rates, they're comparable to other studies. In fact, the most comparable or similar study to Optimum is something called the VAST-D study, a large comparative effectiveness trial done in largely middle-aged veterans by the VA system. And they reported 10 to 11% severe adverse event rates for the same treatment arms as we studied, bupropion and aripiprazole. Interestingly, there was no difference in severe adverse event rates by medication, which was contrary to our hypothesis. Further, it did not appear that fall rates differed by medication as well, although there may have been a slightly higher fall rate in the bupropion augmentation arm. So what this suggests to us is that augmentation, which Dr. Carp had earlier said, is often preferred by physicians compared to a switch because of possibly greater effectiveness. Augmentation is not less safe, at least not in the short run, than switch in older adults. Let's go to the next and last slide. We're now going into question and answer discussion, and I'm gonna hand things back over to Emily Lenard, who will lead us. Thank you for those presentations, and I'm excited to simulate a Q&A session with our speakers. Our first question is directed to Dr. Lenz. So based on your new findings, what is your recommended antidepressant treatment algorithm for depression in older adults? Well, I'm gonna just start by considering a person who's completely treatment naive. They've not been tried on the antidepressant before, so they're not treatment resistant. I would start typically with an SSRI. They're safe, well-tolerated, easy to use, and we have a lot of data on them in older adults. If an SSRI doesn't work, then typically my second choice is an SNRI, serotonin norepinephrine reuptake inhibitor, medication like venlafaxine. I would also consider the drugs velazodone and vortioxetine as good second-line treatments, but do note that because they are, at this point in 2021, still patent-protected, they're rather expensive compared to the generic drugs like duloxetine and venlafaxine and desvenlafaxine. Then after that, if someone has failed both of those treatments, based on our data, I would probably look at erythropoizole augmentation as a likely next choice. Now, there are some times that you wouldn't do that. For example, if someone had Parkinsonism, but based on the greater effectiveness and good safety record, that would be my next choice. I think considerations of bupropion augmentation, switch to bupropion if you're finding tolerability issues, and mirtazapine augmentation, and as well as quetiapine augmentation are very reasonable as well. Brexpiprazole is another augmentation strategy, although again, because it's patent-protected, it tends to be expensive. Thanks, Dr. Lenz. Dr. Karp, anything to add? No, I think that that's a really useful algorithm to follow. I'd add a couple of points. So Eric, you said that you would start with an SSRI. I want to remind our reader or our viewers about the FDA advisory about using citalopram at doses greater than 20 milligrams in older adults. And so if you were going to use citalopram in somebody who was an older adult, I would get an EKG, but I would be more likely to consider an alternative. So likely Lexapro or escitalopram or sertraline. I keep my formulary pretty tight. I pick a couple of medicines from each class, and so I'm very familiar with how to improve tolerability and get patients to an effective therapeutic dose. I try not to give up at a lower dose. I may prolong the titration, but I want to get patients to an effective therapeutic dose as high as I can go till I see an effect or they can't tolerate it and have them stay on that for six to eight weeks before we throw in the towel. Thanks for that, Dr. Karp and Dr. Lavretsky, any more wisdom to share? Yes, it's important to understand other factors that might be contributing to treatment resistance such as comorbid medical conditions that may be undertreated like thyroid disease or any inflammatory diseases and try to optimize treatment for underlying medical conditions. Also, some of the medications prescribed for medical conditions can be perpetuating depressive states, so it's important to look at the overall picture behind treatment resistance. Also, frequently it may lead to cognitive disorders and this is something that we followed and diagnosed patients in our study whether or not they had cognitive problems and those problems like mild cognitive impairment or even early stages of dementia had to be addressed as well as potential contributors to treatment resistance. Also, behavioral activation is very important in patients who are very depressed for a long period of time, they tend to be inactive and I do a lot of behavioral activation, I commit them to lifestyle changes, sleep adjustment and other problems that might be chronic pain for instance and I use a lot of augmentation or additional strategies with lifestyle interventions or stress reduction techniques like meditation mind-body therapist. Although there are no studies for treatment resistance depression but this changes tend to be proactive for the patients and they feel like they are participating in their own wellbeing and those could be very helpful techniques for promoting depression remission. Thanks for- Can I follow up on one thing that Helen said, Emily? One of the unique things about geriatric psychiatrists is that we attend to even subtle cognitive changes in our patients and medical comorbidities in particular cardiovascular diseases and neurologic diseases. So we can still get people well who have medical comorbidity but medical comorbidity and comorbid anxiety and cognitive decline are all possible predictors of non-response and negative moderators of response. So we need to attend really carefully to these medical problems to make sure that we can get the mood and anxiety problems that we're trying to treat as improved as possible and that involves frequent collaboration with our primary care colleagues and working with our patients to prevent cardiovascular and neurologic illness. Very good. Thank you all. So I think that's a good lead into the next question directed first at Dr. Levretsky but we'll get everybody a chance to jump in. What features of aging are related to depression treatment outcomes? So for example, patients with late onset depression may have a vascular component to their depression making treatment more difficult. Right, so when we talk about geriatric depression, aging, underlying aging of the body and the brain are the factors that promote the risk for depression and chronic depression, treatment resistant depression. And those include neurodegenerative and vascular changes in the brain and our own studies and others have shown that these are contributors like for example, the amount of white matter hyperintensity so lacunar infarction, microvascular changes can lead to such features of geriatric depression such as executive dysfunction, apathy or anhedonia that are very resistant to treatment and require different methods of treatment. For instance, we tried augmentation with memantine and metalphenidate in our prior studies that was more effective in this difficult to treat patients with a lot of executive dysfunction and vascular changes in the brain. Also inflammatory changes and inflammatory conditions. Inflammation generally is a common denominator of all diseases of aging, including Alzheimer's disease, depression, arthritis, heart disease, cancer. And so paying attention to the amount of inflammatory conditions in the body is important and actually antidepressant treatment can reduce some inflammation in the body and it could be anti-inflammatory as has been shown by some of our collaborators. Thanks so much, Dr. Labretsky. And Eric, do you have anything to add? Yeah, I think we have to remember that many older adults with depression are also gonna have co-occurring cognitive impairment. And that leads to a few issues. First and foremost, adherence. If an older adult has even a little bit of memory problems, they're gonna have an awful lot of difficulty adhering to taking medication. Particularly when they're on complex or polypharmacy regimens that we often see our patients taking. A second point is that these older adults with cognitive impairment are likely to have more tolerability issues with antidepressants or certain augmentation strategies maybe because they have some underlying Parkinsonism that makes antipsychotics harder to tolerate or because they have an underlying frailty phenotype that may make them more likely to fall. Finally, I think when older adults have co-occurring cognitive impairment, they may simply have a different phenotype. Colleagues have raised the possibility or the concept of mild behavioral impairment akin to mild cognitive impairment, suggesting that in many cases, late onset depression or anxiety or any other behavioral phenotype may simply be a feature of the neurodegenerative syndrome. It's funny that still in 2021, we do not have any uniquely geriatric or aging related treatments for depression. I think the closest to that has been Dr. Levretsky's trials in memantine and with methylphenidate, which are promising but require larger studies. And Dr. Karp? I'll just add on to something that Eric just mentioned about adherence and how that may be negatively affected by cognitive decline. So engaging family members as much as possible in treatment decision-making and coming up with efficient ways to remind older adults to take their medications on a regular basis is critical. I mean, we're also still in the middle of a pandemic and there are psychosocial contributors to treatment resistance that we haven't talked about, but we are well aware of and realize are part of a solid treatment planning, but there is an epidemic of loneliness and social isolation. And while medications certainly help this, we need to address these other contributors to depression and low wellbeing. Thank you. So of course we know that suicide is the gravest risk of depression in all populations. And in this study in particular with older adults, we have a very high risk sample. So were there any suicides during the study and how did the study team keep participants safe during their participation? And we'll go first back to Dr. Karp. Sure. So we expected that there would be suicides in this high risk sample of patient participants and we did not have one that we know of. We were pretty meticulous in documenting suicidal ideation, history and behaviors, documenting risk factors and protective factors and providing ongoing education to all of our clinicians and raters. We had a suicide risk management protocol that we followed. And I have implemented this in my own clinic with my MAs and nurses who I work with. So if patients endorse suicidal ideation between visits or when clinicians were assessing the patient participant, that would trigger a further suicidal risk evaluation. And then according to contingency planning and we published this protocol, happy to share it. They would either activate EMS or contact the psychiatrist and family members to really reduce the risk. Great, thank you. And Dr. Lavretsky, anything to add? Yes, I would like to also comment on qualitative analysis that were performed in the optimum study throughout COVID and that addressed the impact of the pandemic on suicidality and depression. And interestingly, our patients who were chronically depressed and older were dealing with this quite well, coping quite well through about June. And that's when we evaluated them. But then the defenses and the coping mechanism started disintegrating because of the chronic nature of the stress. And by November, they were not doing as well as they did. The reason they did well though was sort of a resilience training provided by chronic depression that put them in a better position of being prepared to deal with chronic stress by first of all, having their treatment and their providers already set up, psychotherapists and psychiatrists. And our program was quite helpful. And although many of them suffered from loneliness and social isolation, they had tools that they learned along the way of being chronically depressed. But this defenses started breaking up by November of 2020 when they were getting more depressed and suffering consequences of more severe stress of the pandemic. So I think it's important to, when you're evaluating chronic suicidality, chronic depression, to look at this factors of psychosocial contributors and underlying stress, psychosocial stress. And those factors would also need to be addressed for prevention of suicidal behaviors. I'll just add to Jordan and Helen that I'm still astonished, like Jordan had mentioned, that in this study, a very large study that followed a large number of older adults with treatment-resistant depression for a long period of time, both a 10-week acute period and then a one-year continuation period, did not see a single completed suicide. Now, of course, we'd like to chalk that up to the management that participants received during the study. But of course, we don't know why we didn't see any. The fact is that studying suicides and suicide prevention in a highly rigorous way is extremely challenging. It's extremely difficult to conduct, for example, a prevention trial for suicide, which is, I think, why we still don't have a lot of really firm answers on suicide prevention. But I just, I wanted to remind our viewers that I don't think the suicide story is over yet. And if we think back to what happened in Canada after SARS in 2003, there was a delay in a bump in suicide after that epidemic ended about nine or 10 months later. So we really need to be mindful and vigilant to assessing and preventing suicide best we can in all of our patients. We hope that they have good supports in place and really want to be working to enforce those, absolutely. And perhaps the kind of hope that participating in a treatment trial gives to patients has been a protective factor for them. What about those participants who didn't get better after their step two treatment trials of either nortriptyline or lithium? What happened with them and did the study make recommendations for them? Dr. Karp, going to you first. Sure, so this was a pragmatic trial and our goal was to test the switch versus augmentation strategies in two steps. We wanted to follow these participants for another 12 months. We were providing decision support to their prescribing physician. We didn't drop these participants, we cared about them. We wanted to make recommendations to their docs and we wanted them to stay in contact with us. So we would make recommendations for their doctors to consider, but that wasn't an outcome then. We tested the two steps. If they didn't respond, we made a recommendation for them and their doctor to consider. And then we followed up at the time points over the next year and just documented this as part of our assessments. Dr. Lavretsky, did you have anything to add? No. Or Dr. Linz? No. All right, thanks for that. So another question is about psychotherapy and other therapies. And this question is for Dr. Lavretsky. So I think we'd like to know more about were participants in the trial undergoing psychotherapy and then just clinically, do you have recommendations for behavioral and alternative therapies for patients? Yes, some of the participants were on a stable psychotherapy regimen. They were seeing their private or group psychotherapist for at least four months prior to starting the study. And they continued. And we do have this information in order to analyze it, whether those who were receiving psychotherapy did better or not than those who didn't. However, in practice, for those who prefer having psychotherapy, it's a good alternative. I'm not aware of any studies that were targeting treatment-resistant depression per se, but there are certainly known trends for psychotherapy for geriatric depression. And the best types of psychotherapy are problem-solving psychotherapy, interpersonal psychotherapy, and brief cognitive behavioral psychotherapy. And that's what I typically recommend for my patients with different preferences, especially if they have anxiety disorders and could benefit from learning some tools through cognitive behavioral therapy for anxiety and depression. And CBT also exists for chronic pain and insomnia. And those are typical conditions when I recommend having CBT training and psychotherapy. And it's a time-limited therapy and also skill training for patients that they would be able to use in the future. Great, thank you. Dr. Lin. And I'll just add to what Helen said, which I totally agree with, that many of these older adults will have a lot of co-occurring anxiety, whether it be worry, perhaps frank panic attacks or a phobic-like behavior. And thinking of adding psychotherapy in those cases is really important. Relaxation training, one of the simplest things that a patient and a therapist can do is actually quite effective for anxiety in older adults. The one thing I would also add is that, unfortunately, on the other hand, is that many older adults, again with depression, have cognitive impairment. And at least intuitively, that makes them a less good candidate for psychotherapy. So that's something to consider as well. And Dr. Karp, anything to add? I don't know if it makes them not a good candidate for psychotherapy, those with some mild cognitive impairment, but the psychotherapy needs to be adjusted. So modified to maybe be slowed down, simplified if there's some behavioral recommendations, or approaching psychotherapy as a sort of dyadic or team approach. So you engage a spouse or a caregiver and it's sort of a train-the-trainer model. So I think that there can be benefits from using psychotherapy across the lifespan. Great recommendations. Thanks for that. And I think we just have one concluding question, really kind of looking at next steps for treatment and research as well. So when should a clinician really just say that enough is enough with oral antidepressants and consider other agents like escadamine or neurostimulation? And Dr. Lenz, if you'd like to get us started on that. Yeah, this is a great question. Of course, the optimum study was focused on comparative effectiveness of antidepressants, but there are now several proven non-medication options and non-oral medication options for older adults. There's intranasal escadamine, which has some evidence in older adults from a clinical trial that was conducted. There's IV ketamine, which of course is not FDA approved for depression, but is commonly available in metropolitan areas, major metropolitan areas, although it does not have a whole lot of data in older adults to date. Then of course there's ECT, which is highly effective for older adults with depression and TMS, transcranial magnetic stimulation. So when would you say enough is enough? Well, some patients seem to not tolerate many antidepressants very well. So I would quickly consider perhaps TMS for them. And then after, I would say somewhere around three or four trials of oral antidepressant strategies, oral antidepressant strategies, it's time to move on to one of these other. Dr. Karp, anything to add? No, I agree with everything that Eric said. We haven't really talked about substance abuse, occult substance abuse, and the need to take a step back and really identify are there other behavioral or chemical contributors to non-response? I think that, again, everything that Eric said is spot on. Part of our work though, is to continue to instill hope in our patients and families and to not give up and to keep them in treatment because we assessed psychological wellbeing in addition to depression in optimum. So maybe there's other benefits that have a positive effect on mood and psychological wellbeing that staying in treatment and oral or other antidepressants can provide that may not be measured with something like the Madras or the PHQ-9. I have a question for my panel partners. What do you think about using psilocybin or other hallucinogens for treatment of treatment-assisted depression? It's a great question. I guess I don't think of psilocybin as a hallucinogen per se. I think it certainly does create a spiritual or other type of experience. I think medically it's a pretty safe thing to do. I'm not aware, however, aside from perhaps in cancer, I'm not aware of data in older adults. So I would describe that as an area where I'd like to really see some data first before I would even think to prescribe it or I should say recommend it in an older adult. Great question though. Yeah. Have you recommended it for any older patients? I had older patients come in and ask for it. You know, and that's, I live in Los Angeles. That might be the reason why. But, and I'm frequently interviewed about it by newspapers because they, first of all, the psychiatric use is on the rise and also the communities are more receptive to that. And with marijuana and MDMA being, approved by the FDA for treatment of PTSD, people are just requesting those treatments. Something else that we need to study. Something else we need to study, yes. I think we have lots of directions for future research and always good to have more work to do. Anyone have any closing thoughts? Think that about does it? You want to close us up, Emily? Thank you everyone for this enlightening discussion. Thank you. Thank you. Enjoy the meeting. Thank you.

OPTIMUM study

treatment-resistant depression

older adults

antidepressant strategies

aripiprazole

bupropion

lithium

nortriptyline

psychotherapy

medical comorbidities

cognitive impairment