Clinical TMS: A Brief look back and then glimpses of the exciting future.

The following article was adapted from the Keynote Speech at the Clinical TMS Society Annual Meeting in New York, May 2014. Original speeach given by Dr. Mark George, M.D. It was adapted by Ted Williams.

This talk is called: “Clinical TMS: A Brief look back and then glimpses of the exciting future.” So I’m going to be part historian and part futurist. I’m from Charleston, South Carolina. It’s kind of a historical town. It was important in the revolution. The Civil War started there or as we say in Charleston, “The war of Northern Aggressions” started there. So, in Charleston, I’ve become more of a historian, and history comes naturally to me. But I’m also going to talk a bit about the future. Before I begin, as a sidenote, in order to try and be a good scientist and not let my checkbook dictate my scientific conclusions, I’ve decided to be economically independent the best I can in respect to these brain stimulation techniques. And specifically regarding TMS because I ran the NIH trial and now I co-lead a VA trial, I don’t take any money from anybody who makes a TMS device.

What about medications? They work, but the psychopharmacological revolution that was so important in psychiatry really forgot a very important fact. And the fact is that the brain is an electro-chemical organ. And that’s extremely important.

In fact, electricity is the currency of the brain. Every thought we have is actually electrical signals going through our brains. The electrical signal goes down through and then causes release of chemicals. So when we remember that the brain is an electro-chemical organ and that electricity is the currency of the brain, it makes absolute sense that with electricity we can get in and communicate with the brain in very powerful ways. That’s the fundamental new knowledge that’s driving the brain stimulation revolution. All the synaptic pharmacology is actually that which happens after electrical stimulation.

TMS was around even as far back as 1910 when Silvanus Thompson put his head between his big TMS devices and could see phosphenes. He could see flashes of light. He wore tails to the lab, top hats and tails. As everyone probably knows, in 1902, Pollacsek and Beer, working in Vienna, right around the corner from Sigmund Freud, filed a patent to use something that looked very much like TMS to treat depression.

Modern TMS started in 1985 by Tony Barker, who was actually trying to stimulate the spine and he realized that when he built the machine that with stimulation that was deep enough to stimulate the spine, he could stimulate the brain. But he was afraid because people were just starting to talk about the brain as a computer back then and he was afraid if he discharged the TMS coil on the brain he might erase the hard drive and cause somebody to be a vegetable. So he didn’t do it. Instead, he called up Merton and Morton who were working at Queen’s Square in London. They were doing electrical stimulation of the brain and they said “Bring it down.” So Tony did. He got on a train, went down to London, gave them the machine, they picked it up, put it on their motor area, just naturally, and their contralateral thumb popped. And then a line formed around the labs and that was that. Tony then collaborated with John Rothwell who was doing motor physiology. That was in 1986. I did my first fellowship in Queens Square in 1989 and I luckily had the office right below John Rothwell. I ran into somebody in an elevator one day who said “somebody but a magnet on my head and caused my thumb to move,” and I said “oh what floor was that?” he said “eight.” He got off, I punched “8” and I went to John Rothwell’s office and saw TMS for the first time myself. He was doing motor cortex work and I asked him then “What would happen if you moved that coil forward and put it over the prefrontal area?” I’ll never forget. He looked up and he said “I don’t know. Why would you ever want to?” And so much of my career and our work is “why would you want to?” and that’s the question.

So, why would you want to? One of my first careers was as a photographer, an imager. And we had PET scans and I put some healthy Washingtonians in the scanner and we had them remember the saddest events in their lives. And we gave them a script and we brought them to tears. We’d say “Remember that funeral, your father’s funeral. Remember how you felt. Remember the look in your mother’s eyes.” We’d get them very sad. They cried. And then we had a control condition. We looked at blood flow in the brain and we subtracted it. Lo and behold, there were areas that were uniquely activated during sadness. So we were able to get some of the first pictures of induced sadness in the brain. Front page of the New York Times, great, this is cool! The brain, and emotion, lives in circuits. If you notice, we got CG25 activation long before people started thinking about that for deep brain stimulation in that area. But what was unclear and no reporter asked me is, what is this actually doing? Is this causing the depression or the sadness? Or is it trying to undo the sadness? No reporter asked me the causal question, but I wondered whether TMS might be a way to actually ask that question: Can we stimulate these regions and cause sadness or undo sadness?

So with those kinds of ideas, others and I started thinking about depression as a dysfunctional circuit in prefrontal cortex and limbic areas. In normal health, emotions are regulated by the cortex, reevaluating emotion and keeping it all in check. In depression, for some reason, that system is not working.

We have a lot of snakes around our cabin in North Carolina and while hiking, I run across them periodically. What I do is I look at it and I have a fear response. I get very scared and I want to flee. I have a cortex that says, “Wait a minute, I’m in the North Carolina mountains, this is a Timber Rattler. Timber Rattlers never bite anybody. You can pick them up and wrap them around you. They’re afraid of humans.” If the same snake were down in Charleston, it would be an Eastern Diamondback and it would kill me. But my cortex is able to come to the rescue, suppress my fear response, I’m able to stand there and watch the snake slither away. This is just an example of the cortical regulation of emotional life that I think happens not just in fear, but happens in normal emotion and this dysfunctional depression. So, with those ideas about functional circuits we thought maybe we could use TMS to get in there and reset depression.

Depression is such an interesting illness because people do come out of that state and they go back to normal. There’s no residual. We know that ECT works. Somehow ECT can reset that governance.

In 1993, a very important paper came along by Alexander, Delong, and Strick that postulated that there were cortical and subcortical governing loops. I’ll never forget drawing them on a napkin in the cafeteria at the NIH and showing it to my boss and saying that I thought we could use TMS cortically to get in there and change these loops. And he looked at me like I was talking nonsense, but to his credit he said, “Why not? Go ahead.” So we did.

So how did we use TMS? Well, we were interested in the prefrontal area because of the imaging studies, the cortical and subcortical governing loops. There was a very important paper by Harold Sackeim showing that, after ECT, the amount of shutdown in the prefrontal cortex actually correlates with the clinical response. To start, I chose higher frequency stimulation and my first studies were actually done with 20 hertz. It had been shown that 20 Hz, over the motor cortex, sometimes caused no response, but sometimes the stimulations caused a superbolic response. I had the idea that at 20 hertz we might get super volleys of cortical signal down deeper, and that’s what we wanted. So we started with 20 hertz. There’s no evidence that higher frequency in the 20 Hz, 10 Hz, 5 Hz range is unsafe, but we moved down to 10 Hz because it’s safer and easier. We started on the left prefrontal because I did an open study comparing left versus right and found that people got more anxious with stimulation on the right. Assured that daily dosing would not cause cognitive problems, we decided to dose daily, including weekends. Then my wife Eloise said, “If you continue to do that we won’t be married.” So I call not treating on weekends the Eloise rule, and I can say that we’re 28 years and still going. But we don’t really know whether this matters, the weekends off, or not. We determined the number of stimuli and the length of the session was at least enough, in healthy subjects, to cause a change in the hypodynamic cortical loop interaction.

The first patient I ever treated had Bipolar II disorder, very treatment resistant, and was treated with left prefrontal much like the strategy currently approved by the FDA. She had failed everything. With TMS, she got a very good response. She got into remission for the first time in five or six years. So this was really exciting – open label, the first person, showing that the crazy idea might work. We quickly set up a double blind study in out-patients and saw a very small, but statistically significant reduction after two weeks.

This was exciting, but it made me a little anxious. Here was something that radically different that might work to treat depression. What could I do to make TMS a viable, FDA approved, option? I was so worried that I would screw it up. Fortunately, those turned out to be unfounded worries. I actually still have that emotion all the time. I want the TMS Society to be huge and I’m worried that we’ll do some things that won’t make it reach its potential, but I can say that with great leadership it really will go great places.

It Takes A Village:

Of course, TMS is the result of the work and ideas of many people. Tony Barker built the first equipment. There were, actually, two case reports in Austria doing single pulse TMS over the vertex, just the idea in open label. Nimrod Grisaru and Bob Belmaker in Israel had the same kind of field, they did maybe twenty pulses and published it with the idea. Saxby Pridmore, a great psychiatrist who does ECT in Tasmania, the very small island off of Australia, became interested. Just when the internet was starting, he got a TMS machine in his practice and he started treating patients clinically and asking “Does TMS work?” “How does it work compared to ECT?” Saxby, a good clinician, was seeing results that mirrored what we were finding in research. This was very important because, as a scientist, I want to be objective and try to discover the world but I’m always worried that my thumb might be on the scale and that I’ve somehow fooled myself. And so Saxby’s data was extremely important to me early on.

Bob Post was the one who allowed me to do this crazy work and was and is a great scientist and warm mentor. Then I befriended Harold Sackeim, who is a great scientist. He’s a great scientist because he allows his view of the world to change with new data. He did not worship at the church of the seizure. When I said that we don’t have to have a seizure to get people better with the device, most of the ECT community said that I was crazy. They actually asked me to leave the ECT meeting when I said that. But Harold had shown that good seizures alone did not guarantee clinical efficacy, and he’d started to disentangle seizures from efficacy, and realized efficacy is about the circuits and the charge. Harold and I had some good early conversations and he thought we should use TMS to cause seizures, that’s called Magnetic Seizure Therapy. I didn’t think we needed seizures at all. And it’s turned out that way, right?

Leon Grunhaus did a comparative study to ECT and found near similar efficacy. I still think that what we do now with TMS is not as powerful as what we do now with ECT, in terms of efficacy, but in terms of side effects, TMS is superior. ECT still has better efficacy, I think, but at least TMS is in the same ballpark and Grunhaus’ study was a very important one early on. Saxby has confirmed his initial findings. He did comparative studies to ECT showing that TMS was nearly as good, that the two technologies could be used together, doing ECT twice a week, and TMS in between. He showed that TMS can be done after ECT to keep people from relapsing, and he showed it all in just a couple of years of hard work in his clinic. He showed that TMS normalizes the dexamethasone suppression test, so it really showed it was getting into the brain. He hinted that it could treat pain, and work as well as medications.

Eric Wassermann is my policeman and colleague. When I went to Mark Hallett, who had the TMS machine at the NIH, and said I wanted to use it to treat depression, he called in Eric and said, “Eric meet Mark. Eric, I want you to make sure that Mark doesn’t hurt anybody. And I want you to always be in the lab when he’s there.” That began a great friendship, and Eric was really helpful, especially in respect to early studies on safety. We formed a society, the ISTS and we had several meetings, one in Europe and one in Philadelphia. There are many others – the late Marty Szuba who was doing TMS at UPenn and has passed away. Others: Thomas Schlaepfer, Leon, and Stan Miller was the very beginning of the industry collaboration. Paul Holtzheimer and many others, have worked hard. It takes a village. Some have looked at safety. Alvaro Pascual-Leone, published a wonderful second round of safety which allows us to move forward in a safe way.

Without Chip Epstein, quite a pioneer, I don’t think we would even be here. Chip is a tinkerer and he is the one who patented the solid core coil that later turned into Neuronetics. And without that patent we wouldn’t be here because that allowed industry to get together and do the critical trials for the FDA. With DARPA money we were asked to build a portable TMS device. We built one that can run on a battery, like a trolling battery for a boat. It weighs twenty pounds, could fit in a kit, and the parts for it cost 3,000 bucks.

We were able to do TMS inside the MRI scanner and I thought that that was going to be the clue, actually finding out about biomarkers and hyperspacing it towards improvement. It’s been helpful, but it really hasn’t been satisfying in terms of telling us clinically how we should go. But it did show us that when we do a TMS pulse in the brain, there is not much difference in terms of blood flow than when we move the thumb volitionally. It showed that TMS is really just using normal physiology.

So we have been able to show some very important relationships. For example, we were able to show that about 60% of the variance between individuals, in terms of motor threshold, is due to the distance between the skull and brain. Initially, it was thought that TMS would not work on anybody over the age of 55 because, as people get older, there is some prefrontal atrophy. We were concerned that treating people at over 100% might put them at risk for seizures, but were getting a lot of non-responders in people over 55.

We realized, “wait, we might not actually be even getting into the brain of people over 55 with 100% of motor threshold.” So then we went and did some more imaging and calculated what is now known as the 120% rule, that will actually get you into the brain of most people up to age 70.

But, I am reminding you of this, so if you have a patient who’s 70 who’s not getting better you should think that you might not actually be getting into the brain and might want to do a scan to see if they do have a lot of atrophy and then gently go up higher. But anyway so we discovered some of these rules along the way that are so important.

We did a study with SPECT and showed that stimulating the dorsolateral prefrontal cortex causes immediate drops in the subgenual cingulate. So TMS is a push-pull of circuits, stimulating and causing a reduction in the cingulate. With improved techniques combining TMS and MRI, Xingbao Li showed that stimulating over dorsal lateral prefrontal cortex clearly shows changes in insula and other mood areas.

A study by Andy Spear clearly demonstrated that high frequency stimulation does different things than 1 Hz. He showed, in glucose studies, that 20 Hz stimulation excites the brain for about 30 minutes, but 1 Hz decreases glucose utilization. It is now clear that the same stimulation, but in different frequencies, produces different results and that TMS is a powerful tool. So this is success! TMS is one of the most successful clinical brain stimulation methods. We started in 1993, and now we know that ECT works but that a seizure may not be needed for efficacy. With some first approximations and circuit models, we now have FDA approval of two different devices for treating depression: Neuronetics and Brainsway. We have several other devices that are in stages of testing and FDA clearance. I don’t know how many machines are out there, possibly 700 in the US. Now we have a clinical society. I don’t know how many patients have been treated, maybe 13,000 to 15,000. I calculate that there are about 7-10 remitters every day, even weekends in the US now because of all your hard work, so that if I decide to stay on the beach and chill, the seven people that are getting better that day as a result of this body of work is really cool. And together with you, who are treating patients, I can enjoy the satisfaction. There is growing acceptance, and new algorithms, and insurance reimbursement. TMS is a success. In Japan there is an exceptional clinic with 60-70 machines. However, there are challenges. We still use the initial approximations that we came up with in the beginning, with marginal refinements. We don’t have the same remission rate as with ECT. We don’t know the upper safety dose. We could be massively undertreating by a factor of ten or one hundred – we don’t know. It is amazing that we don’t know that. And we don’t understand the translational neurobiology. What is it that we’re actually doing in the brain, and specifically, how are we actually getting people undepressed.

Looking ahead with TMS

So that’s our history, told from my perspective. I’m going to turn into a futurist now. The technology is always exploding. I think the technology is always going to be a few steps ahead of us as clinicians. There are always going to be new tools and we’re going to struggle with how to actually use them appropriately.

What we really need are short-term biomarkers of change.

If we could find short-term changes that would predict individual response we could titrate, move to each person and really improve our ability to give clinical care. Also, I think we’re learning a lesson Bob Post tried to tell me the very first day I did my first patient. I didn’t hear it at the time. I was getting ready to do the first patient in that American Journal Study and I said, “Bob I’m doing my first patient tomorrow.” He looked at me and said, “Well, what are you going to have them do while you’re treating them?” And I said, “What do you mean have them do? I don’t know. I’m going have them read the paper. I don’t know. I’m too busy.” He said, “I think it matters.” And so Bob said that actually, it probably matters what is going on in the brain while we’re stimulating. I ignored him. And most of our field has ignored him until the very recently. Some of the OCD , PTSD and craving studies are starting to incorporate challenges. With stroke we’re always doing it.

We’re using these brain stimulation tools and we’re now re-understanding that like Hebb showed, we need to activate the circuit in order to get plasticity. What should patients be doing before or during a treatment? I think that’s one of the important challenges of our field and I hope this society can systematically look at what we should be doing while we’re giving a treatment. I think it probably matters.

The society could figure that out because you’ve got the patients out there, and there are ideas out there about society-driven research.

Then there are the technologies. There are the Neuronetics and Brainsway devices. There is a new device by NeoSync that I don’t even know can be called TMS because it’s not able to make your thumb move. But it is a magnet and it is outside of the brain, transcranial, so it does fit under the definition.

There is a great paper in the journal, Brain Stimulation, where a group from Duke tried to think of every possible type of TMS coil one could imagine, all of them. There are round coils, figure eight coils, and one I call the Mickey Mouse coil. And then they simply tried to show the change in magnetic field in a simple sphere. What they came out with is that the Deep TMS Coil really does have a greater magnetic field that goes deeper into the brain. However, this group said that there’s an overall rule about the world with conventional TMS coils: the deeper you go in the brain, the less focal you are going to be. So the deeper you go, and you can go deep, but you can’t go focal. The Holy Grail is to actually go deep and focal. This was disappointing. It said we can go deep, but we’re going to have to hit everything on the way in and the way out. However, what about multiple coils? There is now a multi-channel stimulator. Multiple coils can stimulate one place at high frequency, and another at low frequency, even simultaneously. It’s just a whole new world. And we have to figure out as clinicians, how’s that going to help? What do we even do with them? But the technology is just marching ahead in a wonderful way.

There is a technology in which the stimulation is actually synced with the person’s own brain alpha frequency, getting harmonics into the brain. A small multi-site, randomized, sham-controlled trial found a statistically significant effect. Such a device could never cause a seizure and could be sent home. We could have home TMS, maybe by next year. The technology is just flying. With conventional TMS, though, I think we still need to increase the dose. I did a graph where I summed the amount of stimulation in a session, the amount in a week, and the amount in the whole treatment. In the early days, back in 1997, we were worried about the FDA, we gave very lose doses. Now, we give in a single session what we used to give in an entire course of two weeks. Then again, every time we’ve had the opportunity we’ve increased the dose in terms of the number of stimuli.

So other people have asked whether we have to have a single treatment every day for 4-6 weeks. Can we bake the cake faster? Can we get somebody undepressed quicker? Are we bound to the weekday treatment with the Eloise rule as a kicker? Is that the way the world is, or can we do it better? Paul Holtzheimer and Chip Epstein did an open label study where they gave 15 sessions over two days, inpatient, and they found fast effects.

With Geoff Grammar, we just completed a study looking at patients who were admitted for suicidality. These are patients that are rarely studied. They’ve just come in off the bridge or they’ve just come in from the ICU after an overdose and we gave them 18,000 stimuli per day or 54,000 stimuli in three days. So it’s the highest dose known to man in terms of stimuli over three days.

We did it as an inpatient, because the last thing we wanted to do is to have somebody go jump off a bridge immediately after getting TMS. So, what did we find? Safe, no side effects, tolerable, they loved it! We had, in the active group, a statistically significant reduction in one day in terms of the suicide scale. Luckily this was inpatient, so they could make changes and talk in therapy. Everybody in the study got better. That’s why we hospitalize people when they are suicidal. But the acute change at one day is what I think shows, and what I hope will be, the next round of work. We asked them, with the visual analog scale, “How bothered are you?” Those people who got active treatment were much less bothered, even from the first treatment, all the way through the rest of their hospital stay. In research, I can’t make students depressed. We can’t take people and make them depressed, so how can we get more information? I like to study pain because I can make people hurt. I can take them in the lab, and induce pain. Pain has a lot of overlap with depression. We know, for instance, that as people go through a course of TMS their painfulness from the procedure drops off. In practice, we have to hold their hand the first two days, and after that, they are sleeping in the chair by the third week. How is that happening? What is going on? My colleague Jeff Borckardt, who is a genius, not just for this work, but for other things, decided to actually do a study on patients who had just had gastric bypass. He gave them a single 20 minute session of TMS in the step down ICU. Right after they came out of surgery, he swooped in, while they were still kind of groggy (half of them don’t ever remember getting the TMS), and gave them a single treatment. He looked at how much morphine they gave themselves with the PCA pump. That single session, with what is now standard treatment used for depression, resulted in a 40% drop in morphine over the next day and a half. Really remarkable. So does TMS have an anti-pain effect in the dose and location used for depression? Unbelievably, the effect is probably bigger than the anti-depressant effect. He has now replicated this in a larger study that was funded by the NIH. We can do all this in the lab as well. Joe Taylor, a precocious PhD student in my lab, found that the analgesic effect of TMS could be blocked by pre-treating with naloxone. Sure enough, when patients are pretreated with naloxone, there is no analgesic effect of prefrontal TMS. Stimulating the cortex with TMS obviously works through endogenous pharmacology. I love that study because it shows that the mechanism is not stimulation or pharmacology. They are the same thing. We’re just applying pharmacology focally with our technology.

So, how big will clinical TMS get? Many people think that the field of non-invasive brain stimulation, like TMS or whatever the next generation of TMS is, will be as big as psychopharmacology; but five years ago, not many thought so. I personally don’t think it will get quite that big, but, who knows who knows, it may.

So, what are the next steps? I think we need to find other diseases that respond to TMA, and that’s going to happen. There is data in pain, in migraines, and FDA approval for migraines. There’s been work on addictions, stroke recovery, OCD, epilepsy, Parkinson’s, and others. All of these disease states have circuits, and we can get in there with these tools. We just have to be clever enough to see if we can manipulate them in therapeutic ways. I have to apologize to Tarique, because I was initially a part of this society and then I got cold feet. I worried about whether this group would try and oversell TMS, and then I dropped out, but I came back. I think we can’t afford to oversell TMS. The data are good enough in themselves, but we just have to be very honest about what we can and can’t do. There is a real problem if we try to oversell. We don’t have to say that TMS is better than another treatment or TMS is better than ECT. It is good enough, and we have to resist false claims. We shouldn’t say more than we can do, because what we can do is pretty darn good. It’s revolutionary, actually.

Remember, the scientific method is our friend. It's how we got here. Not “post hoc ergo propter hoc.”

Does anyone know what that means? This is why athletes wear stinky clothes. So they go, they’re in the playoffs, they win the game, they say why did we win that game? It’s because I had these magic socks on. And the next day they wear those same stinky socks, and they win again. And they keep wearing those same stinky socks through the whole playoffs. They falsely think that just because something happened after that it was causal. We make this mistake in medicine all the time. Athletes do it, and I think we members in the society do it. We’re treating somebody, and they’re not getting better after three weeks. We’re compassionate. We want them to get better, so we do something different: we change the lighting in the room, we move the coil somewhere else, we do something… and they get better. And then we get a belief: oh, they got better because they were wearing the stinky socks. Unless we subject our observations to a randomization, we really don’t know whether it was that thing that we did, or was it just more time in the chair? As clinicians, we’re always experimenting. Every patient is an experiment. We’re trying to get ideas. A lot of our best ideas come from serendipity, what we find in the clinic. But we have to subject what we think we’re seeing to the scientific method, and I think that the Clinical TMS Society really has to remember post hoc ergo procter hoc. I think we have to be agnostic about the methods. I’m not wedded to left, ten hertz. I don’t care. I want what actually works in the right patients, and I think this society has to have this kind of open-minded approach.

Remember Clever Hans the math horse. In the 1800s this guy in Germany had the idea that animals are just as smart as people, but they don’t have the language capabilities that we do. He thought that if he could just figure out how to communicate, he could prove it. So he took a cat, a horse, and a bear and tried to teach them math. Anybody who owns a cat knows that the cat is just going to walk away. Anybody who knows a bear knows the bear is going to eat you. But he thought he was actually able to teach his horse to do math. He would say to the horse, “two plus two,” and Hans the horse would stomp his hoof four times. He actually believed that he had discovered something about the world. However, other people, who were open-minded, but skeptics, said, “Let’s think about this.” They found that the correct answers declined the further the horse was away from the crowd. If the handler didn’t know the answer, Hans wasn’t that good, and if a curtain was put between the horse and the crowd, Hans was no better than chance. So it turned out that Hans was actually getting subtle clues from the handlers and the crowd about the right answer. He was looking around at faces, body language and he was able to figure that out. With TMS we are right there beside every patient. If we have a false belief about whatever, we communicate that to our patients. That is why, with TMS, we have to do very rigid sham control trials where everybody, including the treater, the patient, and the nurses, are blinded. We have to have sham control in a very rigorous way because of the amount of interaction we have with patients.

Question and answer from the audience

Audience: Early in your lecture, you mentioned that you didn’t know why it was depression that was chosen as a start. Any speculation to that? Do you think that may have been industry driven? The large depression population? In your opinion, why depression?

Mark George: When I go to my colleagues, John Rothwell, he would look at me and say, “Why are you doing this with depression? We don’t know the neuroanatomy of depression. I don’t understand it, you should do the first studies with TMS with something that has a well-defined circuit, and that you know which way you want to drive it – like tinnitus. We’ve been pretty darn lucky with depression, and I think it’s because depression is a brain state that we can actually switch into, we’re not having to actually fundamentally rewire the brain, as with stroke. We’re not having to fundamentally cause a plastic change in a plastic way. Depression, whatever it is, is a state that people switch into, and quickly out of it. We found an efficacy early on. And I think that’s what it is.

Audience: If a patient has gotten well, how do you know when they might need a booster?

Mark George: So this is a tricky question, about when to bring them in if they start to relapse. What do you do? Here’s what we do in Charleston: I try to make people not have to come see me. I’m in an academic setting so it’s easy for me to do that. After treatment, we tell people to go out, and tell them, “We’re here, here’s our number,” so we don’t necessarily give maintenance unless people have shown that they absolutely need it. We tell them, “If you have a bad day, don’t call us. If you have two bad days make sure you know where our number is, but you have a bad week, call us.” So if they really are relapsing (and people who have been really depressed know their illness and know sadness versus illness), then they call us and we bring them back in, and we’ll start doing another full course. Often they won’t need a full course, they won’t have to go to four weeks. We can get them remitted with a shorter amount. We treat to remission, but don’t know exactly how much that would be for each individual. We’re prepared to go for 4-6 weeks, but we just do kind of a weekly remission rating. Then, in somebody who’s done that, we’d think about doing a maintenance schedule. Don’t ask me what the right maintenance is, I’m sure you guys have better experience. There’s really very little data about that.

Audience: When we first got the machine, it was the standard six week course, but part of me didn’t believe that because of my experience with psychopharm, where I saw an exponential difference of variance between different patients. Some might need 10 milligrams and somebody else may need 100 or 200. What we’ve seen clinically is that yes, many, most of our patients need six weeks. Some need fewer it appears, but some need more than six weeks. I wanted to get your experience on that and your thoughts about it.

Mark George: Clearly there are patients that don’t need that much, especially if they’re bipolar. What about longer? In the OPT-TMS trial we didn’t go into the data but we had a randomized phase and an open label phase and then we actually had an add-on at the end of that where patients could get four more weeks. We clearly showed, in that situation, that there are people who will ultimately be remitters, but who actually hadn’t changed at all after six weeks. So in clinical studies like that where we can sequentially go, there clearly are often those people who have been ill for a very long time, they often have anti-depressant treatment history form ratings of five or six. So I think that the length of illness and treatment resistance may actually push you into needing a longer course of treatment. But we need a short term biomarker. We need to be able to say to somebody based on what’s happening now in your brain or in your rating scale or whatever, we can predict that if you just stick it out we can get you well. But we don’t have that now.

Audience: We’ve had several patients who have gone through the six weeks without a buzz or we give up and then we have the luck of hearing from them a month or six weeks later, to my astonishment they seem to have gotten better belatedly.

Mark George: Yes, a delayed response. I’ve seen that a lot, yes. You never know. I always tell that to people who leave without a response.

Audience: What are your observations on the value of the adjunctive treatments like CBT and so on in terms of making a measurable difference in response and duration?

Mark George: So is Bob Post right? Should we be doing things in the chair? Will that change efficacy? I don’t know. We have different treaters in our place, some who do aggressive supportive therapy, and some do insight oriented therapy and others who do not interact with the patient. I think my patients do better when they have therapy as opposed to the treater who is just reading the paper, but I don’t know. It needs to be systematically examined and should be. And I hope that the society will help us figure this out. I think it clearly matters in OCD. No one is doing stroke recovery where they don’t actually integrate the stimulation with some sort of physical or behavioral therapy. I just think it’s an important idea. I can’t imagine it wouldn’t matter from all that we know.