Transcranial magnetic stimulation (TMS) Therapy involves the use of repetitive electromagnetic pulses to non-invasively penetrate the skull and stimulate cortical neurons at the site of stimulation and more distal neurological pathways⁸. The use of repetitive transcranial stimulation in the treatment of disorders is called TMS Therapy.

TMS Therapy is the only brain stimulation device yet invented that can non-invasively, and relatively painlessly, focally stimulate the brain of awake individuals. Although direct transcutaneous applications of electricity can influence brain function, it is extremely painful and must be administered with anesthesia. Moreover, as the skull acts as a large resistor, it is difficult to focus the electricity to specific brain regions. TMS Therapy is able to focally and painlessly stimulate the cortex by using a time-varying magnetic field. This localized pulsed magnetic field over the surface of the head induces electrical currents in the neurons in the cerebral cortex of the brain, depolarizing the axonal membranes of these cells, causing them to fire.

The general technique of single pulse TMS has been and is currently being utilized to investigate neurologic function (attention, memory, movement, speech, vision, etc.) in the research setting and to evaluate disease processes in the central and peripheral nervous system in the clinical setting. Many investigators have used TMS Therapy to influence brain functions at this immediate or intermediate time domain to explore how the brain works. These studies have investigated movement ³⁶, visual perception⁴⁵, memory⁷⁶, attention, speech⁵³, neuroendocrine hormones ²⁶, ⁷⁰,¹⁷⁴,²⁰¹ and mood ⁷⁰.

Research on the use and effects of TMS on brain function is now widespread and expanding rapidly. There are at least four books on the technology ⁶³,⁶⁴,¹⁶⁰,¹⁶⁴, multiple reviews ⁶⁶,⁷³,⁷⁵,²¹⁵ and over a thousand published articles.

The amount of electricity needed to cause changes in the cortex varies from person to person and also from one brain region to another ²¹⁰. One commonly used method for standardizing and adjusting the amount of electricity induced by TMS across different individuals is to determine each person’s motor threshold (MT). The MT is commonly defined as the minimum amount of stimulation energy needed to produce movement in the contra-lateral thumb, when the coil is placed optimally over the primary motor cortex. MT can be determined either by using EMG recordings or by using visible movement ¹⁷⁷.

The use of TMS as a tool in brain mapping studies of the motor cortex and prefrontal regions led to the observation that TMS Therapy provided mood changes in patients receiving TMS Therapy. Several factors have further driven the investigation of TMS Therapy for the treatment of major depression. Early reports of changes in mood, the non-invasive nature of TMS, elimination of the need for anesthesia, the lack of side effects and stigma compared to ECT and the non-response of over 35% of patients to ECT have all played a role.

Five meta-analyses of existing TMS Therapy trials have each shown that out of all TMS Therapy trials conducted and available in the literature (¹⁶ open, ²⁴ controlled comparisons and 5 comparisons against ECT), TMS Therapy has a statistically significant antidepressant effect greater than sham ²²,⁸⁹,¹¹⁰,¹³⁰,¹³⁵.

The results of trials that compare TMS Therapy with ECT have shown TMS Therapy to be similar in efficacy to ECT for MDD patients without psychotic features ³⁴,⁸⁰,⁹¹,⁹²,¹⁷³.  The durability of the response to TMS Therapy, as assessed at 3 and 6 months after treatment, also is similar to the durability of antidepressant effects seen with ECT³⁴

In general, studies using higher power, low duty cycle, high frequency TMS Therapy applied to the left dorsal lateral prefrontal cortex appear to result in the most promising results. Increasing the number of stimulations per day and extending the number of treatment sessions for some patients also appear to show the most promising results. The cumulative nature of the effect over time is consistent with other anti-depressants commonly used today.

Single and repetitive TMS are generally regarded as safe and without lasting side effects ²⁰⁷. A TMS Therapy procedure is completely non-invasive and anesthesia is not required. Side effects are generally limited to discomfort at the site of stimulation due to depolarization of sensory and motor neurons in the scalp under the stimulation coil. A muscle tension headache may result in some patients and can persist for 1-2 hours post stimulation.

The primary safety concern with TMS Therapy has been the risk of seizure induction. Eight seizures have been reported secondary to TMS Therapy ²⁰⁸,²¹³. These have occurred in a sample size estimated to be over several thousand TMS treatment sessions. Following the adoption and widespread use of guidelines prescribing a safe interval between pulse trains ²⁴ and the safety guidelines from a National Institute of Neurological Disorders and Stroke (NINDS) workshop on TMS Therapy²⁰⁸, no seizures have been reported since 1997 with the exception of one “pseudo-seizure” as reported by Conca et al (2000) ²⁹.

There have been no negative cognitive ⁷⁸,¹²³,¹⁹⁴,²⁰⁵ cardiovascular ⁵⁴, or direct brain ¹⁴⁵ sequelae reported as a result of TMS Therapy. Immediately following a TMS session, subjects have been tested and have shown no significant neurocognitive side effects. They are thus free to drive themselves home or return to work after a TMS stimulation session. Two reports found evidence of short- term hearing loss in subjects who had been exposed to TMS ¹²⁷,¹⁶¹ (although this finding has not been replicated ¹⁵⁹). However, in general, subjects in TMS studies wear earplugs to minimize risk of changes to auditory function.

In summary, many studies have established that TMS Therapy is well tolerated by patients with major depressive disorder. These studies, although limited in size, also lend support to the potential effectiveness of TMS Therapy in the treatment of MDD.

back to top

TMS Therapy stimulates cortical neurons by creating a time-varying magnetic field generated by brief but powerful electrical currents ¹⁸⁴.  Electrical current is rapidly turned on and off in the electromagnetic coil through the discharge of a capacitor. The end result of TMS Therapy is stimulation of the brain through the use of magnetic fields.

The electrical energy stored in a capacitor is discharged into the Therapy Coil with a peak current of about 3,000 Amps. This creates a powerful magnetic field pulse with a strength at the surface of the Therapy Coil on the order of 1.5 Tesla. This rapidly changing magnetic field (typically ~12-15kT/s) penetrates the scalp and skull and induces an electric field within the brain over a localized region near the cortical surface. The electric field induces current to flow in the brain in a closed loop in this localized region, which creates a transmembrane potential that depolarizes neurons.

TMS Therapy differs from techniques where direct electrical energy is applied to the brain, such as ECT, because the electric field is magnetically derived and is more directly focused to the target site. TMS Therapy also differs from the use of low-level static magnetic fields because they do not induce electrical currents in the cortex.

It is generally assumed that TMS Therapy produces its behavioral effects through the production of electrical current in the cortex of the brain and an induced stimulation of the mood circuits (limbic system). The magnetic field induced by TMS Therapy declines rapidly with distance away from the coil. Thus, TMS Therapy coils directly stimulate only a small region in the cortex, and do not produce direct electrical stimulation deep in the brain. Deeper brain structures can be influenced by cortical TMS Therapy, however, due to the cortex’s massive interconnections and redundant cortical-subcortical loops.

Cascading Neurologic Events Induced by TMS Therapy

Current in the coil generates a magnetic field B that induces an electric field E. The lines of B go through the coil; the lines of E form closed circles. The upper-right drawing illustrates schematically a lateral view of the precentral gyrus in the right hemisphere. Two pyramidal axons are shown, together with a typical orientation of the intracranial E-field. The electric field affects the transmembrane potential, which may lead to local membrane depolarization and firing of the neuron. Pyramidal axons are likely stimulated near bends, as illustrated, but other mechanisms exist and other neurons may be stimulated. Macroscopic responses can be detected with functional imaging tools, with surface EEG or as behavioral changes¹⁸⁴.

The scientific literature documents the numerous effects of TMS Therapy on the neurobiology and physiology of the brain. Human and animal studies have shown TMS Therapy to have similar effects to other known antidepressants ¹¹⁶ (Table 1).

As the exact cause or causes of depression have not yet been fully determined, it is difficult to pinpoint which one or combination of the effects of TMS Therapy on brain neurobiology contributes most significantly to its antidepressant effect. However, several of these changes have been linked to anti- depressant effects through one or more of the theories for causes of depression. Left dorso-lateral pre-frontal high frequency TMS Therapy appears to exhibit the greatest antidepressant effects and is consistent with the current thinking that many of the root causes of depression are located in brain circuits in this region, or circuits (Limbic System structures) connected to this region’s circuitry.

Lisanby and Belmaker ¹¹⁶ reviewed the current body of knowledge with respect to TMS Therapy methods of action as compared to ECT and other anti- depressants. The authors stated that “differences in the physical properties of magnetic and electrical stimulation result in marked disparities in the amount and distribution of electrical current induced in the brain; nevertheless, TMS Therapy shares many of the behavioral and biochemical actions of electroconvulsive shock and other antidepressant treatments.”

back to top

Combining TMS Therapy with functional and structural brain imaging is evolving as an important neuroscience tool for researching brain connectivity ⁶⁴,⁹⁴,¹¹²,¹⁵¹,¹⁵²,¹⁶⁵,¹⁹²,¹⁹³. This is reviewed in several recent publications ⁶⁴,¹⁶⁷. Combining TMS Therapy with functional imaging can also elucidate the mechanism of TMS Therapy as an antidepressant. In contrast to imaging studies with ECT which have found that ECT decreases global and regional activity ¹⁵¹, most studies using serial scans in depressed patients undergoing TMS Therapy have demonstrated increased activity in the cingulate and other limbic regions ¹⁹¹,²⁰².  The technique of interleaving TMS Therapy with blood oxygen level dependent fMRI, has allowed direct imaging of TMS Therapy effects with high spatial (1-2 mm) and temporal (2-3 secs) resolution ¹⁰,¹⁴,¹⁵,¹⁶,¹⁷,¹⁹³.  This technology has shown that prefrontal TMS Therapy at 80% motor threshold (MT) produces much less local and remote blood flow changes than does 120% MT TMS Therapy ¹⁴⁷.   Strafella and Paus ²⁰⁰ used PET to show that prefrontal cortex TMS Therapy causes dopamine release in the caudate nucleus and has reciprocal activity with the anterior cingulate gyrus ¹⁶⁶. Similarly, lateral prefrontal TMS Therapy can cause changes in the anterior cingulate gyrus and other limbic regions in depressed patients ³⁸,¹⁹¹,²⁰².  Teneback, et al has shown that left prefrontal TMS Therapy produces immediate blood flow increases in orbitofrontal cortex, hippocampus, and left prefrontal cortex ²⁰². The brain imaging studies to date thus strongly suggest that TMS Therapy delivered over the prefrontal cortex has immediate effects in important subcortical limbic regions which are involved in mood and anxiety regulation.

back to top

Numerous animal studies have been important in determining the modes of action of TMS Therapy. TMS Therapy studies with intracranial electrodes in rhesus monkeys have provided information about the nature and spatial extent of the TMS Therapy-induced electric field ¹¹⁹. Corticospinal tract development, aspects of motor control, and medication effects on corticospinal excitability have been studied fairly extensively in non-human primates using single pulse TMS Therapy ⁶,⁷, ⁴¹,⁷⁴,¹¹⁵,¹⁹⁹.  Such work has yielded information about TMS Therapy neurophysiological effects, such as the observation that TMS Therapy evoked motor responses result from direct excitation of corticospinal neurons at or close to the axon hillock ⁶,⁷.

Rodent TMS Therapy studies have reported antidepressant-like behavioral and neurochemical effects.  In particular, TMS Therapy enhances apomorphine- induced stereotypy and reduces immobility in the Porsolt swim test ⁵⁰.  TMS Therapy has been reported to induce electroconvulsive shock (ECS)-like changes in rodent brain monoamines, beta-adrenergic receptor binding, and immediate early gene induction ¹².  The effects of TMS Therapy on seizure threshold are variable and may depend upon the parameters and chronicity of stimulation ⁹³.  Recently, Post and Keck have completed a series of studies using more focal TMS Therapy in rat models ¹⁷⁰. They have largely replicated earlier TMS Therapy animal studies using less-focal coils.   Prefrontal TMS Therapy in the rat recently has been shown to induce increased levels of dopamine and glutamate in the nucleus accumbens ²¹⁸.

In summary, recent pilot human brain imaging and animal data provide strong support that TMS Therapy has neurobiological effects similar to other somatic and pharmacological antidepressant treatments. Although the exact mechanisms of action by which TMS Therapy improves mood are unknown, evidence to date shows that TMS Therapy has the ability to affect most brain regions and neurotransmitter systems involved in regulating mood.

back to top

Theory of TMS Therapy as an Antidepressant

The notion to use TMS Therapy therapeutically for the treatment of depression has grown in part from observations that patients receiving TMS Therapy for brain mapping and other neurologic studies experienced mood changes. The non-invasive nature of TMS Therapy, the relative lack of side effects compared to other treatment modalities, and the non-response of many patients to current treatment options are all also likely to have played a role in applying TMS Therapy to the treatment of depression. In some of the earliest trials with TMS Therapy, two open studies in Europe in the early 1990’s used round coils to deliver TMS Therapy over the vertex to potentially treat depression ⁷⁹,¹⁰⁶.  Results from these trials suggested TMS Therapy may have anti-depressant effects but the response was not robust. George, et al, reasoned that prefrontal and limbic regions were more important for mood regulation than the brain regions near the vertex and that ECT works only when applied over the prefrontal regions⁷² and in 1995, performed the first open trial of prefrontal TMS Therapy as an antidepressant⁷¹.  This trial was followed immediately by a cross-over masked study⁶⁹.  The theory behind this work was that chronic, frequent, sub-convulsive stimulation of the prefrontal cortex over several weeks might initiate a therapeutic cascade of events both in the prefrontal cortex and in connected limbic regions. Thus, beginning with these prefrontal studies, modern TMS Therapy was specifically designed as a focal, non-convulsive, circuit-based approach to therapy.

Clinical Studies of TMS Therapy

To date, a large number of trials have been conducted to evaluate TMS Therapy in the treatment of MDD; 16 open studies, 24 controlled studies and 5 studies comparing TMS Therapy and ECT. Tables 1-6 in Appendix 3, list the studies, patient and treatment parameters of these trials and clinical efficacy and safety outcomes. Most of these studies show antidepressant effects and for many controlled trials, these effects are significantly greater than sham treatment. However, these studies evaluated only small to moderate patient numbers and treatment parameters vary considerably from study to study.

One way of evaluating the state of the art of TMS Therapy as an antidepressant is to perform meta-analyses on the data contained in the published trials. There have now been five published independent meta-analyses of the published or public TMS Therapy antidepressant literature, each differing in the articles included and the statistics used²²,⁸⁹,¹¹⁰,¹³⁰,¹³⁵.  By and large, most depression patients in these trials have received no benefit from medication trials prior to trying TMS Therapy. They represent a more treatment-resistant cohort than in comparable studies of new antidepressant medication trials. The five meta-analysis results are the same; daily prefrontal TMS Therapy delivered over several weeks has demonstrated antidepressant effects greater than sham treatment. For example, Burt and colleagues examined 23 published comparisons for controlled TMS Therapy prefrontal antidepressant trials and found that TMS Therapy had a combined effect size of 0.67, indicating a moderate to large antidepressant effect ²².  A sub-analysis was done on those studies directly comparing TMS Therapy to ECT. The effect size for TMS Therapy in these studies was greater than in the studies comparing TMS Therapy to sham, perhaps reflecting subject selection bias. The authors suggested that perhaps TMS Therapy works best in patients who are also clinical candidates for ECT.

The meta-analysis conducted by Kozel and George was confined to published double-masked studies with individual data using TMS Therapy over the left prefrontal cortex ¹¹⁰. The summary analysis using all 10 studies that met criteria revealed a cumulative effect size of 0.53 (Cohen’s d) (0.31 to 0.97) with the total number of patients studied being 230. A funnel plot technique assessed whether there is a publication bias in the literature to date, and whether this bias might affect the results of the meta-analysis. The funnel plot indicated that a publication bias is likely and that there are more positive small sample studies in the TMS Therapy antidepressant literature than should occur by chance. These authors then employed techniques to determine how large this publication bias would have to be in order to change the results of the meta-analysis. The fail-safe results indicated that there would have to be 56 non-significant unpublished studies of approximately the same average sample size as the published studies, in order to change the cumulative meta-analysis effect to a non-significant result (56 studies with Rosenthal’s method, 22 with Orwin’s method).

The most critical meta-analysis of the TMS Therapy antidepressant field was recently conducted using the guidelines put forth in the Cochrane library ¹³⁰.  However, even this stringent meta-analysis included 14 trials suitable for their analysis and found that left prefrontal TMS Therapy at two weeks produced significantly greater improvements in the Hamilton Rating Depression Scale (HRDS) than did sham. In summary, all five meta-analyses of the TMS Therapy published literature concur that repeated daily prefrontal TMS Therapy for at least two weeks has antidepressant effects greater than sham.

Direct Comparisons of TMS Therapy to ECT

Although there is general consensus that TMS Therapy has statistically-significant antidepressant effects, an important question is whether these effects are clinically significant. The meta-analyses have on average, an effect size of Cohen’s D of about 0.65, which is a moderate effect, in the same range as the effects of antidepressant medications. For example, small to medium effect sizes (0.31-0.40) are common in randomized controlled trials of novel antidepressants. Thus, with respect to whether or not TMS Therapy has clinical significance, an important clinical issue is whether TMS Therapy would be clinically effective in patients referred for ECT.

This question has been addressed in a series of studies in which ECT referrals were randomized to receive either ECT or TMS Therapy ³⁴,⁸⁰,⁸¹,⁹¹,⁹²,¹⁷³.  Study design and outcomes for these studies are listed in Tables 5 and 6 of Appendix 3. In an initial study, Grunhaus et al compared 40 patients who presented for ECT treatment and were randomized to receive either ECT or TMS Therapy ⁸⁰,⁸¹.  ECT was superior to TMS Therapy in patients with psychotic depression, but the two treatments were not statistically different in patients without psychotic depression (although ECT arm patients were allowed to remain on psychotropic medications during the trial, whereas TMS Therapy arm patients were not). This same group recently replicated this finding in a larger and independent cohort with an improved design. Janicak and colleagues reported a similar small series, finding near equivalence between TMS Therapy and ECT, with a remission rate of 46% with TMS Therapy, ⁹¹,⁹².  It is important to note that these relatively small sample studies have not established clinical equivalence between TMS Therapy and ECT, merely similar effect sizes. The major differences between these studies and the rest of the controlled studies of TMS Therapy efficacy are the patient selection (suitable for ECT), the length of treatment (3-4 weeks), the lack of a mask, and the lack of a sham control.

Durability of TMS Therapy Antidepressant Effect

Few TMS Therapy studies have included a follow-up period after treatment to examine for continued efficacy. Many of the published TMS Therapy reports include anecdotal follow-up observations of beneficial effects of TMS Therapy lasting from as short as 1 week to as long as 6 months.

Pascual-Leone et al (1996) using 90%MT, 10 Hz, 2,000 pulses, 5 TMS Therapy sessions monthly/5 months, reported relapses in all 17 patients treated with TMS Therapy within 3 weeks after the last treatment ¹⁶².  Similarly, Conca et al.(1996) using 90% MT, 0.17Hz, 40 pulses, 10-14 TMS Therapy sessions, found the beneficial effects to decline over the 14 days following TMS Therapy²⁸.  Avery et al. (1999) using 80% MT, 10 Hz, 1000 pulses, 10 TMS Therapy sessions, reported sustained recovery for up to 4 months in some patients⁴.  Loo et al (1999) utilized 110% MT, 10 Hz, 1500 pulses, 10 TMS Therapy sessions and reported recovery to be maintained for approximately 6 months in four out of five treatment-resistant patients¹²⁵.

Triggs, et al (1999) in a small 2-week open trial of TMS Therapy delivered at 80%MT, 20 Hz, 2000 pulses, 10 TMS Therapy sessions, showed significant improvement (40% of patients with 50% reduction in HRDS) in mood that was retained at 1 and 3 months following treatment with concomitant pharmacotherapy ²⁰⁵.  Greene, et al similarly demonstrated a statistically significant improvement in mood in 40% of patients after 2 weeks of TMS Therapy treatment at 110%MT, 10 Hz, 1000 pulses,10 TMS Therapy sessions, with 44% and 32% response rates at 2 and 4 weeks post-treatment, respectively⁷⁸.  Patients did not receive concomitant pharmacotherapy during the follow-up phase, however 36% were placed on pharmacotherapy at the end of the study by their treating physicians. Dannon et al. (2002) showed durability of TMS Therapy given at 90%MT, 10 Hz, 1200 pulses, 20 sessions, was equivalent to durability of ECT at 3 and 6 months post-treatment with concomitant pharmacotherapy³⁴.

In a recent study by Schule, et al. (2003), responders to TMS Therapy given at 100%MT, 10 Hz, 1500 pulses, 10-13 TMS Therapy sessions, (39% with 50% reduction in HRSD, 20 % remitters [HRSD ? 9]), depressive symptoms were shown to increase over time and the degree of deterioration correlated with interval to start of oral antidepressant use¹⁸⁹.  This effect was reversible in 9/10 patients where they stabilized or showed additional improvement.

In summary, information on the durability of the antidepressant response to TMS Therapy is limited and studies differ in results ranging from 1 week to 6 months after the last TMS Therapy treatment. The dose of TMS Therapy delivered may account for some of these differences with higher stimulations and/or longer treatment times giving a trend towards a longer duration of effectiveness. The use of concomitant antidepressant pharmacotherapy after TMS Therapy treatment may further improve continuation of the clinical response to TMS Therapy and may help to avoid deterioration of symptoms.

Long-term Administration of TMS Therapy

Studies cited above indicate that treatment 5x/week with TMS Therapy over 4 weeks do not present any additional safety concerns beyond those cited over shorter treatment times. Similarly, additional treatments of TMS Therapy over several months do not appear to alter the adverse event profile (Pascual-Leone et al (1996)(162).

George, et al (2003) described a study of seven patients with treatment-resistant bipolar disorder who had responded in an active TMS Therapy trial and were offered participation in a 1-year maintenance therapy of weekly TMS Therapy⁶².  TMS Therapy was performed 1 day per week at 110% MT and 5 Hz for 8 seconds for 40 trains (1600 pulses/ session). During the follow-up period, 4 subjects dropped out and were considered non-responders after an average of 25 weeks of TMS Therapy. Three subjects safely completed 1 full year of weekly TMS Therapy without a depression relapse.

These studies, albeit with small numbers of patients, suggest that long-term TMS Therapy is safely tolerated. They also suggest that repeat treatments with TMS Therapy may provide a means to maintain the antidepressant response in some patients.

back to top

The known most critical safety concern with TMS Therapy is the generation of inadvertently-induced true seizures ²⁰⁸. A seizure is induced during electrical stimulation of the brain via ECT and, given that TMS Therapy stimulates the brain by induction of electrical currents in the cortex, seizure could occur with TMS Therapy at high intensity. Interestingly, attempts to induce seizure with high power TMS Therapy have not always been successful. For example, in a study exploring TMS Therapy as a method to induce therapeutic seizures, stimulation parameters far above the published safety thresholds and beyond the range of any standard TMS Therapy system had to be used to induce seizures, and even so, not always successfully ²⁰⁴,¹²¹.  The clinical experience with TMS Therapy in several thousand patients has also shown that there is minimal risk of seizure, particularly when TMS Therapy is used within published safety guidelines.

Eight seizures have been reported secondary to repetitive TMS Therapy as described in more detail in Wassermann, 1998. These eight seizures have been reported with use of TMS Therapy in a sample size that is estimated to be over several thousand at a frequency well below 1%. Most of these seizures occurred in early studies designed to evaluate the safety of TMS Therapy. Review of these cases and further study of the role of the interval between pulse trains in seizure induction(24) led to recommendations for stimulation parameters and inter-train intervals to allow safe clinical use of TMS Therapy. Following the adoption and widespread use of these recommendations, there were no other reported inadvertently-induced true seizures in TMS Therapy antidepressant trials that were conducted within safety guidelines ²⁰⁸.

Additional recommendations on managing the risks of TMS Therapy (ISTS Consensus Statement) and a safety screening questionnaire for patients who are considering TMS Therapy clinical use have also been developed and published based on the 1998 recommendations ²⁰⁸.

The most common adverse effects of TMS Therapy are a muscle tension type headache and discomfort at the site of stimulation. These are mild, transient and generally respond to analgesics.

back to top