Treating Anxious, Depressed Patients to Wellness

Jeffrey E. Kelsey, MD, PhD
Mood and Anxiety Disorders Clinical Trials Program
Department of Psychiatry and Behavioral Sciences
Emory University School of Medicine
Atlanta, Georgia

Please use your browser's print button to print this section

Slide 1 Our first speaker is Dr. Jeff Kelsey. Dr. Kelsey is a Clinical Assistant Professor at Emory University at the Department of Psychiatry, and his major interests are in clinical trials in anxiety and depression as well as social phobia. Dr. Kelsey is going to speak to us today on treating anxious depressed patients to wellness. Well thank you very much Gene, and welcome. I was particularly happy to accept the invitation to come out here as I did my residency just down the road at Stanford. It's always nice to get a good excuse to come back to the West Coast and it looks like a beautiful day outside, so I hope you're not going to be stuck indoors the entire time that you're out here in San Francisco.

Slide 2 What I am going to focus on in my remarks is various strategies and why we think about trying to get our patients who have anxious depression, or just depression, well, as opposed to just better. We're going to toss some terms around. We're going to call it remission; we'll call it wellness, as opposed to just getting better, or just getting a response.

Slide 3 The bottom line if you take home nothing else other than just this one point, is what we are aiming for in this patient population is to achieve a state of remission. That's what we aim for. Sometimes it's easy, sometimes it's hard to get to, but that's where we need to push to. Now what we're going to talk about is kind of a combination of bad news and good news.

Slide 4 The bad news is that there's an awful lot of patients, for one reason or another, who get better and that's what they settle for. Either they are still depressed enough that they can't really remember what it felt like to feel well, or they're depressed enough that there are still some guilt components as well, they figure, "Well, you know, this is just as good as it gets for me." So they get stuck. In my practice I spend about half my time doing our clinical trials program, half the time in an outpatient practice. We get patients who come in and they come in because they have moved; they change their insurance; sometimes it's because their psychiatrist has died and that's a little creepy, but I see them anyway. And they tell my secretary all they want is a refill for their antidepressant medication. And Doreen is really good at drawing the line that says no, you have to come in for a whole evaluation. They come in, it's obvious that yes, they were depressed. And it's also very obvious that they are markedly improved over what they were. But we see that there is still some room for improvement. We kind of nudge and we urge and we get some leverage, and we get them to change, and about two to three months later they come back and say, "You know, I had forgotten what it was like to feel like this. But you know what the really scary thing was, doctor? All of those years I thought I was doing okay." And they were just settling for improvement. So that's the bad news.

Slide 5 The good news is we have an awful lot of different approaches that we can use to get our patients to that state of remission, and we'll be reviewing some of those as we get to the latter half of the talk. We can conceptualism the stages of treatment that people go through. There's an awful lot of people who are not diagnosed or misdiagnosed. They get the wrong treatment for one reason or another, and it's not going the way it needs to go. Over the last few years we've gotten significantly better, though, and we're getting quite good at getting people into treatment. And that's a necessary first step because if we don't start treatment, they're not going to see the improvement. And fortunately a sizable percentage of those individuals do in fact get better and that's clearly an improvement over where they were, but still it's not quite good enough.

Slide 6 Quite literally the goal of our treatment is to get our patients well. That's what we're aiming for but we donut want to stop short. I remember a few months ago I was doing a dinner presentation, and after the presentation across the table from me was a psychiatrist who looked about my age, and he says, "You know, Jeff, you talk about remission, you talk about wellness, I hardly ever get my patients to that point." So what we spend our time in therapy doing is discussing, you know, this is what you have to settle for "It's good enough; this is as good as it's going to get." I was so sorely tempted to say: what you really should be discussing in therapy is like this, "Here's a list of three physicians who will care about you more than I care about you, and will treat you aggressively enough to get well." Because when we stop short, that's the life sentence our patients end up with.

Slide 7 We can take a disease management perspective to see how we can optimize this treatment. The first step, of course, is to have a high index of suspicion. And we also need to remember that many of the signs and symptoms of depression are very interpersonally sensitive. We want our patients to feel like they can talk to us, they can trust us; we are going to be responsive to their concerns. We want to very aggressively monitor outcome and also want to make sure that we recognize depression is not just something that affects the individual. It is, if you will, a ripple effect disease. It affects family, friends, and loved ones. So we want to make sure that we're involving the family or the support group, letting them know what's going on, what the treatment is, why there is a reason this person has been withdrawn or isolative, or more irritable for these last several months, and they can't just make themselves feel better. You want to make sure that patients take the treatment we prescribe because when we see they're compliant, that goes so much farther in terms of reducing some of the negative outcomes that we can see with untreated depression. And for those who have recurrent major depressive disorder but are not yet in maintenance therapy, you want to make sure you identify those recurrences sooner rather than later. When we look at this list - where it's clinically that we see a need to shore up our activities a little bit - it's right at the point of monitoring outcome. This is where would want to make sure that we are providing the objective assessment and we are not stopping short until our patients do, in fact, get well.

Slide 8 What are the special considerations we need to take from this paradigm when we're dealing with the geriatric patient? Well, we often times get a different symptom profile. The geriatric patient is much more likely, when they are depressed, to generate physical and/or somatic complaints as opposed to psychological complaints, or maybe what we're seeing is depression reflected as a loss of independence, or an increase in irritability. Now, let's face it, it's often times not the patients who complain of the increased irritability, it's often times the people who have to deal with them, but this could be another sign of depression that's lurking out there. There is a certain stigma among some generations that depression is just not part of their vocabulary. Depression is what happens when you're lying in bed all day because you don't have the energy to get up and kill yourself, and, "No, doctor, that's not how I feel. I still get up and make it to through the day." So we need to kind of normalize or reframe it for them, and let them know it doesn't need to be that severe to be depression. Clearly the older patients are going to tend to have more medical concerns that will take time and visits, and we really want to be careful we don't get sucked into the trap of saying, "Mr. Jones, good news. I know why you're depressed. You're depressed because your spouse died; you had a heart attack; you had a stroke; your children moved away; whatever..." Explaining why someone is depressed never works as a long-lasting therapeutic intervention and in fact when we recognize the stressors that we think might be contributing to the depression, that's a really good reason to make sure we are ever so much more aggressive in our treatment. The person needs all of their coping skills to get by. What is this anxious depression that were talking about? Well, we can conceptualize it is going to be major depressive disorder with some anxiety symptoms. Not enough anxiety symptoms to really be an anxiety disorder necessarily but it's going to be depression with anxiety. When we look at this clinical entity we come up with some very interesting facts and figures.

Slide 9 What is this anxious depression that were talking about? Well, we can conceptualize it is going to be major depressive disorder with some anxiety symptoms. Not enough anxiety symptoms to really be an anxiety disorder necessarily but it's going to be depression with anxiety. When we look at this clinical entity we come up with some very interesting facts and figures.

Slide 10 Clearly, anxious depression is more the rule rather than the exception. Patients with anxious depression tend to stay sicker for a longer period of time. They tend to be our somewhat more challenging patients to treat as well, and due to the anxiety component that is going to increase the risk of suicide. You got depression, you add anxiety; the risk of suicide goes up. You got depression, you add panic disorder, even more anxiety; the risk of suicide goes up even higher. We also know that due to the anxiety component there is going to be a certain increase in the sensitivity to side effects of medication. One of the things we find in our clinical trials programs is often times people take placebo for the first week. We know it, they don't know. And the more anxious participants are always the ones with more side effects to placebo. I get these messages on my voice mail at 3 o'clock in the morning: "Dr. Kelsey, I took first dose of the study drug. I haven't slept a wink all night. I'm sorry, I need to drop out of the study. If I don't sleep, I'm going to lose my job." The more anxious someone is, the more powerful the experience of swallowing the pill. We also know that there's a time course consideration here as well, and that is that the depressive symptoms often times improve before the anxiety symptoms. So let's keep this in mind as we look at these next two slides.

Slide 11 This is from an outpatient study looking at patients who have major depressive disorder with high levels of anxiety - anxiety consistent with what we might see in a GAD, or generalized anxiety disorder population. This looks at improvement in depression, placebo in red, active antidepressants in yellow and green. After one week of treatment both antidepressants have separated from the placebo and maintained that separation throughout.

Slide 12 Keep the time course in mind as we look at this next slide. This is now looking at improvement in anxiety. We get all the way out to week 8 before we start to see any consistent separation. Almost always depressive symptoms get better first, then the anxiety. What do we tell our patients? We tell them they need to be patient. This is not to be confused with being a patient but we tell them to be patient. Now sometimes people will say they're depressed, they're anxious. Right. I've got a great idea. I'm going to give them an antidepressant but I'll also start them on a benzodiazepine. It will make the antidepressants easier to take, it will cut down the anxiety, and when the antidepressant kicks in then I will taper them off the benzodiazepine. What a wonderful theory. You can probably guess were I am headed with this one, because when we follow these patients up months down the road, if we're lucky they are still on the antidepressants, but you can bet they are still on the benzodiazepine. This is a very hard patient population to get off the benzodiazepine, so what we really need is more a picture of time.

Slide 13 Is there a price to pay if we fall just a little bit short of achieving our mission? The data that we have so far suggests that yes, there is a very significant price to pay. We see a couple of things.

Slide 14 In the clinical trial world there is this category called partial remission. And when we do the math, these are people who are better but they're not yet well. And this is what happens to them: they're going to have a higher relapse rate. One study, looking at 64 patients followed for just over 1 year, found that of those who got well right to here, one in four relapsed. Of those who got better but not quite well, three in four relapsed. This is leverage we use when we are talking to patients about changing strategies, and I say, you know: You tell me which odds you want - the one in four chance in having another episode, or do you want the three in four chance? And every now and then someone says they want the three in four chance and that's when we think, "Now who was that psychoanalyst that I referred my patients to?" We also look at functional impairment - objective impairment. There is a 25 to 45 percent chance that they're not doing well at work, as rated by others. Here's a novel idea; let's ask the patient how they think they're doing. And there are close to _ who think that they are not doing very well at work. We also worry about the possibility of suicide attempts.

Slide 15 This is from a very nice large-scale study by Jan Fawcett, looking at close to 1000 patients treated for major depressive disorder. Each one of these is an independent risk factor for a future suicide attempt. Which of these people are better, but not yet well? They probably still have some anxiety; maybe their concentration is okay, maybe it's not; and perhaps they still have some insomnia and maybe they don't. But again, another compelling data set that points the way toward wellness or remission is what we really need to aim for. Testing whether laughter is the best medicine... It's always good to have some experimental framework from which to work. And just in case you are curious, when the blind was broken, this gentleman received the sham laughter condition, which is why it is without apparent effect.

Slide 16 But when we're treating our anxious depressed patients we want to think: what is the pathophysiology? The two major neurotransmitter systems in the brain, the noradrenergic and the serotonergic systems, are involved in an awful lot of common disorders that we treat: depression, anxiety, and chronic pain. And the evidence for all three disorders is fairly similar if we look, for example, to chemical neuroanatomy. Let's say we take a brain that someone is finished with. It's really important that they're not going to use it anymore. Slice it open and we stain the cell bodies and the pathways for norepinephrine and serotonin. We are tremendously impressed by the cross talk between these two systems. It's not: one goes here, one goes there; they communicate back and forth with each other. Now we're going to focus on depression and anxiety, but we also need to understand the limitation of - what do the medications do that we currently have available? One can make a very cogent argument that, well, you know, I bet a corticotrophin-releasing factor antagonist would work, or a substance-P antagonist would work for my depressed patient. Yet they probably would, but what do you think the chances are you can write a prescription and actually have someone fill the prescription for the patient? So we need to work with what we have currently available. Many of our antidepressants work by inhibiting reuptake of neurotransmitters.

Slide 17 This is from rat synaptosomal data and is plotted by the log arithmetic ratio of serotonin to norepinephrine reuptake potency. Longer bars are more serotonergic, shorter bars are more noradrenergic. And we see that the available antidepressants span a number of orders of magnitude. Had we a drug that was a one-to-one blocker of both norepinephrine and serotonin, it would fall right along this line here, and we see there's only a handful of drugs available in this country today that fall within an order of magnitude of that relationship. Imipramine, Tofranil^®, amitriptyline or Elavil^® would fall right about there. Both are wonderfully effective antidepressants if, and it's a big if, we can get our patients up to a high enough dose. Clomipramine or Anafranil^®, venlafaxine or Effexor^®, so what we have is primarily noradrenergic drugs, serotonergic drugs and a group that sit in the middle that are mixed reuptake inhibitors.

Slide 18 So you say, what difference does it make? Conventional wisdom says all antidepressants work equally well, isn't that what you shrinks have been saying all of these years? Well it's a kind of, "it depends."

Slide 19 And it depends on where we look, so let's just define a couple of terms so we're all on the same page here. If we look at response, which is the typical endpoint of a clinical trial, you've got a Hamilton Depression Score - it is knocked down by 50 percent, you're a responder. These are people who are better. It's treating away from the disease state. In contrast, as clinicians, what we would really rather see is remission - because remission is defined as an endpoint. These are patients who are well, who have got minimal if any symptoms. When we're looking for a response, we are pushing people away from the disease state. When we're looking for remission, we're pulling them towards wellness. And we would much rather our patients get well then just get 50 percent better.

Slide 20 So, if we activate multiple neurotransmitter systems, does it make a difference? Again, it all depends where we look. In many inpatient studies, yes, there's a significant difference in advantage going to mixed reuptake inhibition. In most outpatient studies that use response as an endpoint, no, it doesn't make any difference whatsoever. And this is the data where people say all antidepressants work equally well. But if we look at remission as an endpoint, we see that yes, often times there are studies that suggest mixed reuptake activation does confer an advantage.

Slide 21 Mrs. Hoskins is here to match wits with you regarding her symptoms. When we find ourselves in this sort of physician/patient interaction it often times suggests that things are not going the way we had expected or at least hoped they would. And when we're treating major depressive disorder, we need to keep in mind what is the goal of treatment. It's not decreased symptoms, it's not increased productivity at work, it's not decreased irritability where they're living, the goal to treating major depressive disorder is to achieve a state of complete euthymia. Remission as opposed to response, well as opposed to better, call it what you will, this is what we're aiming for. Now what I do in my outpatient practice, after someone has been in treatment for a while I pose them the following question - based on a scale from 1 to 10, 1 is the absolute pits - as bad as it gets; 10 is just a whole bunch of wonderful days in a row -where do you think you were when we started treatment? And most of my outpatients endorse a level of distress somewhere in the neighborhood of 2, 3, or 4. That's what brings them in for treatment. Then I ask: where do you think you are at now? And what I'm really hoping to hear is around a seven and a half or an eight. The person who says, "Well, Dr. Kelsey, it's a 5. I guess on a really good day it is a five and a half, but basically it's a 5," they are better but they are not yet well. And if we do nothing different when we have the information it's just like walking in to see our patient with hypertension and pronouncing: good news - your blood pressure is down to 150/100 and I'm satisfied with that. Of course we wouldn't be. Now in contrast, when I posed the question and the patient says, "Well gosh, Dr. Kelsey. Everyday is a 10 day," it's time to grab for the lithium, because we have clearly overshot the mark just a touch. At the risk of sobering up the mood for just a moment, let me propose a somewhat scary thought to you. Pretend, if you will, we all work for the Food and Drug Administration. We are not out here in San Francisco, we're back on the East Coast in some little cubical looking at all of these antidepressant trials that come across our desk. The FDA considers someone a responder to antidepressants if, as we discussed, there's a 50 percent or greater decrease in their Hamilton Depression Score. A moderate depression HAM-D of 20 goes down to 10, they are a responder. HAM-D of 10, pretty close to being OK. A somewhat more severe depression, HAM-D of 36, goes down to an 18. Technically, yes, they are a responder, yet a HAM-D of 18 still qualifies them for most outpatient clinical studies done today. Better, but not yet well. And sadly the vast majority of clinical trial reports focuses only on response.

Slide 22 Let's look at the left-hand side of this graph to start with. This is a compilation of four different outpatient antidepressant to studies. Looking at placebo in the blue, venlafaxine XR, fluoxetine in red, this looks at response as the endpoint. Placebo response rate, pretty typical for an outpatient trial - 37 percent to 38 percent. Both active antidepressants work better than placebo. Neither separates one from the other. Again, this is why people say all antidepressants work equally well. This, however, is response. This is people getting better. It didn't say better in that bullseye earlier on, did it? It said well, and that's looking at remission, and here's where the pattern changes just a little bit: placebo remission rate less than 20 percent. Both active antidepressants still work better than placebo, but now what we see is what I eluded to before, it's that mixed noradrenergic-serotonergic activation that seems to push a certain percentage of people farther along the road to getting well.

Slide 23 Are there consequences if we don't get it right the first time? Well, yes there are, unfortunately. Most patients don't generate the same enthusiasm for antidepressant trial No. 3 that they did for trial No. 1. We also worry about the consequences of untreated depression. The consequences are not static; they keep going up. You have a geriatric patient population that's got more medical disorders - the longer they stay depressed the worse off their medical condition is going to be. When people are depressed and they end up in the hospital in a medical surgical unit, they stay in the hospital longer than those who were not depressed. People are depressed after an MI; they die off at a faster rate than people who aren't depressed. There is no up side to staying depressed for a prolonged period of time. We also know that it becomes a more expensive illness to treat - not only the direct costs and expenses, but the indirect cost as well, and are very hard pressed to put a price tag on quality of life.

Slide 24 We talked about trying to achieve this mixed noradrenergic, serotonergic activation. The good news is, there's lots of different ways to do that. If you want to give a single drug, give monoamine oxidase inhibitor. We give clomipramine as kind of the prototypical tricyclic. Neither of these are particularly high up on my list of things that I want to give geriatric patients, however, we give mirtazapine, or Remeron^®; we can give venlafaxine XR at doses of 150 mg a day or above. It's really that 150 plus that gives it that mixed activation. We can do combinations of medications. Take an SSRI or a mixed drug, you can add a tricyclic. We have to check a baseline EKG and we'd need to follow serum tricyclic levels. We can add Wellbutrin^®; we can add stimulants; we can add mirtazapine; if and when reboxetine, a noradrenergic selective reuptake inhibitor, becomes available, we can add that. We have lots and lots of ways to get here. Now for those of you psychiatrists in the crowd, I know you're at this thinking: you know what, this is what we do with our refractory patients anyway, and what we have discovered is a trickle-down effect. What works well to budge the refractory patient works wonderfully well to top off the response that we want to get from the less refractory patient.

Slide 25 What does the APA recommend that we do when we're getting stuck? They offer up the three broad strategies for the nonresponder. The first is an alternative non-MAOI with a different biochemical profile. Keyword here is different. Just like treating hypertension, if a patient doesn't respond to the first beta-blocker, don't give them another beta-blocker, go to a different class of drugs. We don't want to switch. We can keep the same antidepressants and we can add lithium - it's moderately effective; it's not the easiest medication to take. We can add thyroid hormone - now contrary to what some may believe, we as psychiatrists don't use T3 or Cytomel^®; we don't because T3 works better as an augmentation agent than T4 does. We're not treating hypothyroidism, we're treating the affected disorder - or we can add a second antidepressant. What all the strategies have in common, of course, is the old "do something pharmacologically distinct." My personal preference: I would much rather see a good aggressive dose of one drug work for my patients. It's simpler. It's cheaper. And as Jim Roerig will tell us in just a moment, it's going to minimize the chances of drug/drug interactions, which are going to be especially problematic for this population of patients anyway.

Slide 26 So what do we think about when choosing antidepressants? We don't think we've got a good shot at remission; it's just a dead-end path; have they or a biological relative responded to something in the past?; drug/drug interactions (to take just a little bit Jim's thunder, cleaner drugs are always better than dirtier drugs); the affordability; and very importantly, mechanism of action; pathophysiology of the disease state; how does the drug work?; have they failed something else?; what are we going to do that's different?

Slide 27 Let's see how we put all this together into an approach to treatment. It's just like everything else we do in medicine. First we make the diagnosis, then we initiate treatment, and then we assess response. And based on our assessment of the response, we will either keep on doing what we're doing, add to what we're doing, or change what we're doing. And we go back and forth, back and forth until we get the patient well. When we're getting stuck, we think: OK, what's the pathophysiology of the disease state? Am I getting noradrenergic, serotonergic, or dopaminergic activation? For anxiety it's noradrenergic, serotonergic, GABA-ergic activation. Then we remember, no, life isn't all chemistry. How about psychotherapy? Particularly for the chronically depressed, the chronically anxious patient, I am convinced we need to have some component of psychotherapy to really get the best treatment for them. When we're really stuck we take what we would call the blank slate approach. We step back, we reevaluate, and we make sure that yes, in fact, we are treating what the patient has. It doesn't matter how many times we write the diagnosis on the chart - if it's not the right diagnosis, our treatment isn't going to be as good as it could be.

Slide 28 So to summarize, what we find is that when we're treating depression, remission or wellness has to be what we insist on as the outcome for treatment. My hunch is that for everybody in this room - if you had a family member being treated for depression and your choice was do you want them better, or do you want them well? Assuming a relatively functional family, everybody would, of course, choose well as the desired endpoint. When we've got these anxious depressed patients they are harder than the patients who are depressed without anxiety - they're going to have more somatic complaints; they're going to have a greater sense of urgency; they're going to challenge us a bit more. But we, as healthcare providers, we're the ones that have to provide the objective measurement. Are we making all the progress that we need to be making, and are we making it in a timely fashion? And then when we're thinking about the medications that we're going to use, we think about the pathophysiology of the disease state and what drugs we're going to pick. The final point I'd like to leave you with is it is a genuine tragedy that there are so many people in this country who have chronic anxiety, chronic depression. They don't know what they have; they don't know it's treatable; all they know is that they are miserable. Just because you are used to something certainly does not mean you need to continue to tolerate it. Thank you.
Please use your browser's print button to print this section Return to Agenda