This transcript has been edited for clarity.
Madhukar Trivedi, MD: Hello. Welcome to the podcast. My name is Madhukar Trivedi. I’m professor of psychiatry at UT Southwestern and the founding director of the Center for Depression Research and Clinical Care. Today it gives me great pleasure to introduce my colleague and friend, Dr Jonathan Downar, who is the world’s leading expert in treatment for depression with transcranial magnetic stimulation (TMS) and many other treatments for treatment-resistant depression. He has been working with TMS for a long time, since the early 2000s, and is the associate professor of psychiatry at the University of Toronto. Welcome, Jonathan.
Jonathan Downar, MD, PhD: Hi, Madhukar. Thanks so much for inviting me. It’s great to be here.
Trivedi: We are going to address the issue of the role of TMS in major depressive disorder and also electroconvulsive therapy (ECT), but primarily TMS. What exactly does TMS do when you put the stimulator on the scalp?
Downar: That’s a great place to start. TMS was originally developed quite a while ago, back in the 1980s, originally as a tool for stimulating the brain noninvasively by neurophysiologists and neurologists who wanted to study the functions of the brain’s motor cortex. The old-fashioned way to do that was to put a couple of electrodes on the scalp and use electrical pulses, but that’s painful and not very efficient. In the mid-1980s, they were able to develop a system for using magnetic pulses to stimulate the brain much more safely, noninvasively, and nonpainfully. The initial technology was developed to use powerful, focused magnetic-field pulses that would pass harmlessly through the skull and encourage the neurons to activate. If you place the magnetic conductor over the area of the brain that moves the thumb or the toe or whatever, and you apply a couple of pulses, you’d actually see the person’s whole hand move or you’d see the person’s whole foot move. The original way that the technique was developed was simply to stimulate a brain area and noninvasively, just as a mechanism for studying the functions of these areas. Later, in the mid-1990s, it was realized that if you stimulated these areas over and over again, you could gradually strengthen their activity. As you fire the neurons over and over again, you could wire them together and strengthen their connections.
The idea emerged that maybe this would be useful as a treatment for depression. In the 1990s, it was discovered that if you place the simulator not over the motor cortex but over areas of the frontal lobes that perform functions like self-regulation of emotions and thoughts and behaviors, if you did that same function of stimulating day after day with hundreds of pulses, you could gradually strengthen the activity of these areas of the brain. If you succeeded in doing that, then the person would gain more control over their thoughts and their emotions, and there were benefits on things like depression and anxiety. Fundamentally, this is a technology that bypasses the chemistry of the brain and goes directly to the brain connections, and it can be used to strengthen the connections and boost the activity in any area of the brain that you target with the stimulator.
Trivedi: Wonderful. I’m glad you brought up the issue of chemistry because as we showed in the early 2000s with the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study and subsequent studies, we’ve spent the past 30 years with a lot of pharmacology looking at serotonin reuptake blockers and norepinephrine reuptake blockers, and we consistently showed two fundamental things: One is that with any given treatment, only about a third of the people get better in the short term, and that doesn’t persist. And the second thing is, after the first two treatment steps, with these pharmacologic treatments, you get a whole lot of people who don’t still benefit from it. There is where the excitement for TMS, ketamine, etc., is beginning to come in. Where does it fit right now with the given evidence? And what is the evidence for efficacy in general that we currently have?
Downar: Repetitive transcranial magnetic stimulation (rTMS) seems to be finding a home after one or two failed medication trials, but before you proceed to more invasive forms of brain stimulation like ECT and so on… When a person has tried their very first medication trial in your own STAR*D trial, you find that their success rate or the remission rate can be around maybe a third. The problem is that with each successive medication trial after that, the remission rate drops, so it goes down to 25% on the second trial. Eventually, by the time you’re on your third or fourth or fifth medication, the success rate can be down closer to around 10% or 15%. Now, in the past, a person can always go on to ECT, but that’s much more invasive. It’s not as acceptable to the patient because of the anesthesia, the seizures that are involved, and the cognitive side effects. So there was a need for something in the middle that people could go to that would maybe have a success rate higher than with medications, but less invasive and with fewer side effects than ECT. That’s where rTMS really seems to have found its home. Within the United States, the FDA approval is for treatment-resistant depression or at least depression that has been resistant to at least one failed trial of medication. If there’s anxiety comorbid with the depression, then that seems to work equally well. Also, with things like OCD, there’s been an approval in the past few years and there’s also approval for nicotine addiction. But by far and away, the most utilized application of rTMS in the United States today is for depression. That’s certainly where it fits in, after you’ve tried at least one medication without success but before you proceed on to things like ECT.
Trivedi: Can you give us a sense of the standard-of-care duration?
Downar: When a person signs up for rTMS, they’re signing up for a course. A typical course would be between 30 and 36 sessions of treatment — at least that is the initial course. And the standard way of delivering it as of today is to apply treatments once per day. If you come in to the clinic and you sit down in the chair, you’ll get a treatment session that can last anywhere between 3 minutes for the newer treatments up to between 19 and 38 minutes for the older treatments. And you’ll do that once a day, Monday through Friday, more or less continuously until you’ve completed 36 sessions. At that point, most people will have achieved the maximum gains, but there are some slow responders who will continue to benefit if you go beyond 36 sessions out to another with an additional 10 or 15 sessions. Typically that’s what you’d be signing up for.
Trivedi: Say a little more about that, because one of the major contributions that you’ve made over the past couple of years is to not stop at that point for people who are not benefiting, right?
Downar: One of the nice things about rTMS, as I mentioned before, is the success rate is higher. In clinical trials, we tend to see remission rates of 30%-35% at that 30-session mark. There’s evidence that if you’re willing to go up to a larger number of sessions, a person who is benefiting from treatment may have a higher success rate if they treat all the way to remission. If you are willing to come in and keep going beyond 36 sessions out to 50 sessions or, in a few cases, even beyond that, we tend to see remission rates that are above 50%, and maybe even above 60%, out in the community. So that part’s very encouraging. Aside from the success rate, the other big advantage is a very good safety and tolerability profile for rTMS. The technique is very safe. No one has ever died from rTMS. As far as we know, there’s under a 1-in-10,000 risk for seizures that occur with rTMS. So you’d have to be very unlucky. I’ve seen maybe one in my clinic over the course of a decade, and in terms of things like sexual side effects, headache, sleeplessness, insomnia, nausea, and weight changes — those standard side effects that we talk about with antidepressant medications — they really don’t apply with rTMS at all. As a result, when you look at the data from clinical trials, somewhere in the realm of about 25% of people, even in clinical trials, will tend to drop out early before completing the course of medication. In rTMS, the dropout rate is much lower. It’s around 5%-8% percent, and that tends to be more for logistical reasons. It’s less about side effects and more that the biggest downside of rTMS is having to schlep back and forth between your home and the clinic every day for these treatments. Convenience and the cost of rTMS seem to be the two major things that have been holding it back thus far.
Trivedi: I find it important for us to address this issue that you just alluded to, and that is while there is the potential risk for seizure, in your practice where you’ve done thousands, if not more, you’ve seen only one.
Downar: If you go and read about rTMS, we always talk about the common side effects. The serious side effects of these common side effects are that most people will find that it feels like being tapped on the forehead with a woodpecker over and over again. It’s sore, but that’s not a deal breaker for most people. In addition to that, there is this rare side effect of inducing of a seizure. The frequency is about 1 in 10,000 for that happening. For comparison, if a person goes on medication, the seizure frequency seems to be about 1 in 1000. If you’re on a medication, you’re already at higher seizure risk than you would be if you were just taking rTMS with the added benefit that, in the rare cases where it does happen, it happens during the treatment itself. It’s not going to happen while you’re out there doing something in the community or driving around or walking around. So overall, the safety profile for rTMS seems to be one of the major advantages over medications. And in terms of tolerability, the same thing — far fewer side effects. The big thing we have to work on now to get this out to as many people as possible is to improve things like the cost and the convenience, so that people don’t have to schlep, to drive to a clinic 50 times or 36 times, and that the cost can somehow become a little more comparable to what we have for a treatment like medication that doesn’t require somebody to be standing there with an expensive piece of equipment delivering the treatment to you.
Trivedi: You mentioned the difference between pharmacology, which we have used traditionally for a long time, and circuit-based changes. Addressing the circuits is a novelty and at least a departure from our SSRIs and SNRIs. I’m aware that you’ve done this work, so maybe give us a little bit of an idea of what happens to these circuits when you do TMS.
Downar: That’s one of the really interesting developments that’s happened in the past 25 years, thanks to 25 or 30 years of brain imaging studies since the early 1990s. We have fairly detailed maps of at least the major circuits that seem to be most involved in depression and the beginnings of a picture of what it might look like when a person falls into it. There are three circuits that are particularly important. Most of us have heard that the brain has a pleasure center or reward center, or, more accurately, reward circuit, whose job it is to motivate us to seek out opportunities and things we ought to be doing, and then to motivate us to pursue those goals. The problem with just having a reward circuit is that it doesn’t motivate you to deal with all the setbacks and the pitfalls of life. You also need a second circuit called the nonreward circuit. And that one’s less well known. But every brain also has one of those, and its job is to motivate you to try to spot, ahead of time, things that might go wrong or things that have gone wrong in the past — to motivate you to avoid the same pitfalls. It’s a problem if your reward circuit gets stuck in a loop and you have a craving for something. But it’s also a problem if your nonreward circuit gets stuck in a loop and you start being obsessed with all the things that could go wrong, or you’re locked into a cycle of self-criticism, or locked into a cycle of anxiety and worrying that’s unproductive. The reward circuit and the nonreward circuit are both necessary for our function, but if either one of them gets stuck in a loop, we’re in trouble.
We think there’s a third network in the brain, which at the moment is called the salience network, and its job seems to be to allow for regulation or self-control of our thoughts and behaviors and emotions. In the event that our reward circuit or our nonreward circuit gets stuck in a loop, it’s the salience network that can activate if we snap out of it, either through realizing that we have something more pressing to do or, in therapeutic settings, by doing things like mindfulness or meditation. We’ve got these three circuits: the reward, the nonreward, and this control there called the salience network. These are the various circuits that we can target if we’re trying to help somebody with depression or anxiety or other things. There are invasive ways of targeting these. We can use things like deep brain stimulators. Twenty years ago, some headlines were made by Helen Mayberg and her team about implanting the same sort of pacemaker electrodes, called deep brain simulators, that we use for Parkinson’s disease and trying to implant them in these circuits for reward and nonreward, and showing that in some cases there were dramatic and very rapid improvements in the person’s mood — if you could just get it in the right place and stimulate it. Now there’s a shortage of people whose depression is severe enough that they’re willing to have an electronic pacemaker drilled into their skull and implanted in their brain surgically. So we’d really like to be able to get to those same circuits noninvasively. That seems to be the core benefit of rTMS: that, using a magnetic stimulator, we can place the stimulator over these target regions of the brain that are part of these different circuits, and we can either boost up their activity or try to reset their activity, depending on the type of treatment we’re doing.
Trivedi: These two things are coming down the pike, and so we need to make sure the audience recognizes that these are now the next steps for TMS. But I’d like us to address two components in the remaining time. One is what you started talking about: figuring out a way to reduce barriers to access. Then we’ll get to the precision part, as in identifying where to target this.
Downar: That’s exactly right. In a perfect world, what we would like from our noninvasive brain stimulation techniques, like rTMS, is for them to work for as many people as possible. We’d like them to work as rapidly as possible and we’d like them to be as affordable and accessible as possible. In terms of the rapidity of the effect, people have been experimenting since about 2010, when doing more than the standard one session per day. In the first studies, we used one session per day, I guess because something had to be chosen; but it was quite a while before people started systematically checking. Should it really be two sessions a day? Should it be four sessions a day? Should it be 10 sessions a day? People have experimented with trying to accelerate the completion of those 36 sessions of treatment or more by doing multiple sessions per day. And it looks like you can’t necessarily do them all immediately back-to-back. If you go into the brain and do a stimulation treatment every 5 minutes or so, then you’re not necessarily going to get ahead. It’s a cycle; these circuits seem to have to reset themselves.
We’re familiar with this by, I guess by analogy. Every one of us who has a flush toilet in our bathroom knows that if you just trigger the system every 5 seconds, you’re only going to get one effect. Whereas if you trigger it and then wait for a while and then trigger it again, you can get more than one effect. It may be a similar situation we have here, for as long as you do the treatments about an hour apart. It may be possible to do three a day, five a day. Most recently, my colleague, Nolan Williams over at Stanford, has made a lot of headlines by developing a treatment that is using 10-times-a-day simulation. With 10-times-a-day stimulation, you’re getting 50 sessions done between Monday and Friday, and they’ve been reporting two really interesting things there. One is that they’re seeing some people responding in as little as 2 or 3 days, and most of the people are getting full effect in just 5 days. That’s a lot faster than most medications, so that is really quite a major benefit. It may be especially attractive in things like emergency settings or hospital settings, where you don’t have weeks to treat the person. You really need to see some effect within a couple of days. That’s very encouraging.
The second thing that Nolan’s group reported, which has been very interesting, is a surprisingly high remission rate — depending on the study, somewhere between 65% and 90% remission. The question emerges as to, first of all, whether that’s something we’re going to see when the study is replicated in lots of other centers. Also, there’s the question of whether simply doing more sessions, and more often like this, can boost the remission rate all by itself or whether additional measures are needed. By additional measures, I mean that in the Stanford group, they have started to be very clever about doing a session of brain mapping, using functional MRI to try to identify and personalize the exact circuits, which are in slightly different places on each person’s frontal lobe, just like every one of us has a face. But our eyes and nose and ears are there in slightly different positions for each person, so we have a similar situation. What they have attempted to develop is an algorithm that can be used to analyze the activity of the brain over time and pinpoint these circuits so that each person can be stimulated at a personalized target. The potential thing here, if this can be replicated, is that if we are able to do that for the people who wouldn’t normally be responding to a center target, if you can go and personalize the simulation site by using a brain image, then you may be able to boost the remission rate above the usual level just by making sure that the simulator is on exactly the right spot for that person.
Trivedi: Obviously, like with all chronic brain disease and chronic medical illnesses, TMS is not going to serve well for everyone. So who should not get it, and once TMS doesn’t work, are other options available?
Downar: That’s a great question. I would say that in terms of if you’re a person who’s facing depression or if you’re a family member of someone who’s facing depression, the numbers favor trying rTMS for anyone who has tried at least one medication without success, or has at least attempted to tolerate a medication and a second medication and hasn’t been able to tolerate them. Most people do not have any contraindications to rTMS and most people who are facing depression and anxiety, if they haven’t done well on at least one medication — I would say everybody that I can think of among family and friends who doesn’t have a contraindication should go and take a look at it. Most people are suitable for rTMS, and I think it’s a treatment option. Everyone should certainly ask their provider about it if they’re not doing well on medication.
Now, even with the highest success rates that have been reported, up in the 65% to 80% or 90% range, that still leaves a certain percentage of people who are not responding to rTMS or who respond and then relapse rapidly and so on. For those people, there are more invasive options to take a look at on the brain stimulation side. One of those would be ECT or electroconvulsive treatments, which are the old standard. They have a high success rate. They do require you to go into a hospital setting and have anesthesia, so they are more invasive in that way. My colleagues, Dan Blumberger, Fidel Vila-Rodriguez, and Jeff Daskalakis, as well as Shawn McClintock over at UT Southwestern, are working on a magnetic version of ECT called magnetic seizure therapy. That’s a large rTMS device which can also induce a seizure in the brain but seems to be avoiding the memory complications because it’s using a magnetic rather than electrical field. So it’s not penetrating as deeply to disrupt the memory structures of the brain, called the hippocampus, which are fairly deep. So magnetic seizure therapy is a good alternative. That’s easy for a person who wants to try something like ECT but would prefer to do it under a magnetic approach.
Lastly, there are also ways of stimulating the brain using surgical and implanted devices. I mentioned deep brain stimulators. Those seem to have good success, even in some patients who have not done well on ECT. There were some very interesting results last summer in patients who have failed many different types of treatments. Once again, by taking recordings from circuits all over the brain and identifying a personalized circuit for the person, which seems to be associated with the depression, they were able to program a stimulator that could detect patterns of electrical activity that indicated that the person’s brain was about to go down into the deep, dark hole of depressive thoughts and could just send a couple of pulses of stimulation to disrupt that. And this seemed to be quite successful in keeping the person out of depression, even when all the treatments have failed.
So the take-home message here is that in addition to medication, we’re now seeing a whole spectrum of a new generation of treatments that use brain stimulation to directly target the circuits that are involved in depression rather than just using medications. On the noninvasive side, we have FDA-approved treatments like rTMS, and then if rTMS isn’t working, we have magnetic seizure therapy and ECT, which are still noninvasive. And then beyond that, we have invasive approaches like deep brain stimulation. What this means is that as we continue to develop these over time, there’s a lot of hope for people who are facing depression — that they’re not going to be stuck with the same treatments we had 20 or 30 years ago. There is a new generation of treatments coming, and because they’re interventions rather than medications, then just like any other form of technology, like our phones or voice recognition, we can keep tweaking them; we can keep playing with them and we can make them better over time. The optimistic message of the future is that these brain stimulation treatments are going to continue to improve just the way that our phones, our computers improve. And we can anticipate that even if people are having trouble with depression and anxiety, we’re going to have lots of new options for them on the table in the years to come.
Trivedi: Thank you very much. This was a wonderful discussion. It was a pleasure having you.
Downar: It’s always a pleasure to talk with you. Have a great afternoon.