Does daily TMS work for treatment-resistant depression?
This is the randomized, masked, sham-controlled acute phase of the NIH-sponsored OPT-TMS trial, built to test daily left prefrontal repetitive TMS under tighter conditions than the earlier literature allowed. Before it ran, most reviewers had concluded that daily left prefrontal rTMS has antidepressant properties, and most meta-analyses reported a large effect size for symptom change versus sham. That conclusion was contested on quality grounds: the validity of the sham controls was questioned, and critics argued the effects were not robust enough to be clinically meaningful. Many early trials used small daily doses and stopped at 2 weeks.
TMS is a noninvasive brain intervention that modulates activity in discrete cortical regions and their associated circuits by inducing intracerebral currents. Repetitive TMS means the pulses are applied repeatedly within a session. Here the target was the left prefrontal cortex, located by moving the coil 5 cm anterior to the motor threshold site along a left superior oblique plane.
What the trial did:
- Design: 190 patients in the intention-to-treat sample, randomized 1:1 to active or sham rTMS at 4 US sites
- Population: Antidepressant medication-free outpatients aged 18 to 70 with DSM-IV major depressive disorder and a 24-item Hamilton score of 20 or higher
- Dose: 3000 pulses per session at 120% of resting motor threshold, 10 pulses per second in 4-second trains with a 26-second intertrain interval
- Course: Daily weekday sessions, 15 in a fixed 3-week phase, with a variable extension of up to 3 more weeks for patients who improved
- Masking: An active sham that mimicked the somatosensory experience of rTMS, with patients and administrators masked to the acoustic signals of stimulation
The design goals were explicit: optimize the treatment parameters to give robust antidepressant effects the best chance of appearing, fix the methodological weak points of earlier trials, and show consistency across research sites.
Dr. Kumar’s Take
The methods are the reason I take this trial seriously. Sham control has been the soft spot in the TMS literature for years, and the investigators went after it directly with an active sham that reproduced the scalp sensation and with masking of the sound the coil makes. They also required every outcome evaluator to pass a competency certification, checked rater reliability against a masked external expert, and kept assessing whether the masking was holding. On top of that, each patient got a head MRI with vitamin E capsules marking the intended coil site, and in 33.2% of patients the mark landed over premotor cortex rather than prefrontal, so the coil was moved 1 cm anterior. That single detail tells me how often the standard 5 cm rule misses its target, and it is a practical point for anyone delivering this treatment.
The dose here was also deliberately intense compared with the early trials: 120% of motor threshold and 3000 pulses per session, not the small doses given for 2 weeks that produced much of the earlier controversy.
Study Snapshot
This randomized, masked, sham-controlled trial enrolled antidepressant medication-free outpatients aged 18 to 70 with DSM-IV major depressive disorder, single episode or recurrent, a current episode lasting 5 years or less, and a 24-item Hamilton score of 20 or higher. Entry required a moderate level of treatment resistance by Antidepressant Treatment History Form criteria: insufficient benefit from 1 to 4 adequate medication trials, or intolerance to 3 or more. Patients had to remain stable through a 2-week medication-free lead-in, and stayed free of antidepressant, antipsychotic, and anticonvulsant drugs throughout active treatment.
Stimulation was standardized at 120% of the patient’s resting motor threshold, 10 pulses per second for 4 seconds with a 26-second intertrain interval, giving 75 trains and 3000 pulses across 37.5 minutes of stimulation in a visit lasting about 50 minutes. Intensity could drop to 110% during the first week for tolerability and had to return to 120% from week 2. Treatment ran daily in a 5-day weekday sequence for 15 sessions over a fixed 3-week phase, followed by a variable extension of up to 3 additional weeks for patients who improved, so the acute course spanned 3 to 6 weeks.
The primary outcome was remission, defined as a Hamilton score of 3 or less or two consecutive scores below 10. Secondary outcomes included response, defined as a 50% or greater decrease in Hamilton score from baseline at the final visit, plus Montgomery-Asberg scores, Clinical Global Impression severity scores, and patient-reported Inventory of Depressive Symptoms scores. A certified, masked rater who gave no TMS assessed patients weekly.
Results in Real Numbers
Roughly 860 patients were screened, each by a psychiatrist, to randomize 199. Seven patients took part during the first year while the sham method was still being developed and were left out of the intention-to-treat analysis by a blinded decision of the executive committee, and 2 more exited before receiving any treatment, leaving 190 in the intention-to-treat sample. Enrollment ran from October 15, 2004, through March 31, 2009, at the Medical University of South Carolina, Columbia University and the New York State Psychiatric Institute, the University of Washington, and Emory University.
Demographic and clinical features did not differ statistically between the two arms. Treatment resistance in the current episode was balanced, averaging 1.5 failed research-quality adequate trials, which translates to roughly 3 to 6 clinical antidepressant trials. Across their lifetimes, patients had failed an average of 3.3 research-adequate trials, about 9 clinical attempts. The group as a whole was moderately treatment resistant.
The duration-adaptive design set a clear threshold for continuing. Patients whose Hamilton score dropped less than 30% from baseline by the end of the fixed 3 weeks left phase 1 and crossed over to open treatment without being unmasked. Those with a 30% or greater reduction continued up to 3 more weeks with twice-weekly Hamilton assessments, and improvers who had not remitted stayed on treatment only if they kept improving, defined as at least a 2-point Hamilton reduction at every other rating. Once stable remission criteria were met, the acute trial ended, rTMS was tapered over 3 weeks, and an antidepressant was started.
The trial was powered for 240 randomized patients to give 80% power to detect an odds ratio of at least 2, assuming a 10% sham remission rate and a 20% overall dropout rate.
Who Benefits Most
The patients studied here were antidepressant medication-free outpatients aged 18 to 70 with major depressive disorder, a current episode of 5 years or less, and moderate treatment resistance: insufficient benefit from 1 to 4 adequate medication trials, or intolerance to 3 or more. That maps onto a real clinical group, people who have already been through several drug trials without enough return.
The FDA’s approval of rTMS for unipolar major depressive disorder in adults covers those who have not responded to a single antidepressant in the current episode, and it rested primarily on one industry-sponsored trial in medication-free adults. The population enrolled here was more treatment resistant than that threshold.
Safety, Limits, and Caveats
Safety was assessed at every treatment visit through spontaneous adverse event reports, with auditory thresholds and a neuropsychological battery measured at baseline, at the end of the active phase, and at 6-month follow-up. An independent data and safety monitoring board reviewed participant safety and study progress, and three planned interim analyses for harm with respect to depression severity were run when 25%, 50%, and 75% of planned participants had completed phase 1.
The screening burden is real. Patients were excluded for a personal or close family history of seizure disorder, neurologic disorders, ferromagnetic material in the body or near the head, pregnancy, medications known to lower the seizure threshold, other current Axis I disorders apart from simple phobia and nicotine addiction, past failure to respond to adequate electroconvulsive therapy, and previous TMS or vagus nerve stimulation. Baseline testing included urine toxicology and electrocardiography, and patients positive for cocaine, marijuana, PCP, or opiates were excluded.
The time commitment is the other limit. This is daily clinic attendance, Monday through Friday, for 3 to 6 weeks, with each visit running about 50 minutes.
Practical Takeaways
- Confirm the patient can attend daily weekday sessions of about 50 minutes for 3 to 6 weeks
- Expect an antidepressant-free window if you are replicating this protocol: 2 weeks off antidepressants, antipsychotics, and anticonvulsants before baseline, and 5 weeks for fluoxetine
- Screen for seizure history, neurologic disease, ferromagnetic material near the head, pregnancy, and drugs that lower the seizure threshold before starting
- Do not assume the 5 cm rule lands on prefrontal cortex; MRI confirmation moved the coil 1 cm anterior in 33.2% of patients here
- Track the Hamilton score on a set schedule, since the decision to extend treatment in this trial hinged on a 30% reduction at the end of week 3
Related Studies and Research
Episode 31: Depression Explained, The Biology Behind the Darkness
Episode 32: Depression Recovery Roadmap: A Step-by-Step, Evidence-Based Plan
Accelerated TMS - moving quickly into the future of depression treatment
Default Mode Network Mechanisms of Transcranial Magnetic Stimulation in Depression
Psychiatric Applications of Repetitive Transcranial Magnetic Stimulation
Stanford Accelerated Intelligent Neuromodulation Therapy for Treatment-Resistant Depression
FAQs
Where does TMS sit relative to antidepressant medication?
Antidepressants are the most commonly prescribed class of medication across all of medicine, yet after acute-phase pharmacotherapy, psychotherapy, or both, most depressed patients either do not improve or improve only partly, and drug and other somatic treatments often carry treatment-limiting adverse effects such as sexual dysfunction. This trial compared active rTMS with sham in patients who were medication free, not against a drug.
What does a typical TMS treatment schedule look like?
In this protocol, sessions ran daily in a 5-day weekday sequence, 15 sessions across a fixed 3-week phase, with 37.5 minutes of stimulation inside a visit of about 50 minutes. Patients who improved enough, meaning a 30% or greater drop in Hamilton score, continued for up to 3 additional weeks.
How was safety tracked in this trial?
Adverse events were collected at every treatment visit through spontaneous reports. Auditory thresholds and a neuropsychological battery were measured at baseline, at the end of the active phase, and at 6-month follow-up, and an independent data and safety monitoring board reviewed safety alongside three planned interim analyses for harm.
Who is not a candidate for TMS therapy?
Exclusions in this trial included a personal or close family history of seizure disorder, neurologic disorders, ferromagnetic material in the body or near the head, pregnancy, medications known to lower the seizure threshold such as theophylline, other current Axis I disorders apart from simple phobia and nicotine addiction, past failure to respond to adequate electroconvulsive therapy, and previous treatment with TMS or vagus nerve stimulation.
Where was the coil actually placed?
Over the left prefrontal cortex, found by moving the coil 5 cm anterior to the motor threshold location along a left superior oblique plane, with the rotation point at the tip of the nose. Coil position was recorded with a mechanical positioning system so it could be reproduced, and head MRI with vitamin E fiducials was used to check placement, prompting a 1 cm anterior shift in 33.2% of patients, spread evenly across the 4 sites.
Bottom Line
This trial was the attempt to settle a question the earlier TMS literature could not: whether daily left prefrontal rTMS beats sham when the dose is intense, the sham is convincing, the raters are certified and masked, and the coil placement is checked against MRI. It randomized 199 medication-free patients with moderate treatment resistance at 4 US sites, 190 of whom made up the intention-to-treat sample, to a 3-week fixed course of 15 daily sessions at 3000 pulses and 120% of motor threshold, with up to 3 more weeks for those who improved. The methodological care built into it is why I still point to this protocol when discussing how TMS should be delivered.
