Does Stanford’s accelerated TMS work for severe depression?
In this open-label pilot study, 19 of 21 patients (90.5%) who completed Stanford’s accelerated TMS protocol met remission criteria for treatment-resistant depression. Counting everyone who enrolled, including the one participant who withdrew, 19 of 22 (86.4%) met remission criteria. This was an open trial with no sham comparison group, and the authors state that double-blinded sham-controlled trials are needed to confirm the remission rate.
Stanford Accelerated Intelligent Neuromodulation Therapy, or SAINT, combines three changes to standard intermittent theta-burst stimulation: multiple sessions per day at spaced intervals, a higher overall pulse dose, and precision targeting guided by resting-state functional connectivity MRI. The fcMRI scan identifies the region of the left dorsolateral prefrontal cortex most anticorrelated with the subgenual anterior cingulate cortex in each individual patient, so the target is personalized rather than anatomically standard.
What the data show:
- Remission, per protocol: 19 of 21 completers (90.5%) met remission criteria, defined as a Montgomery-Asberg Depression Rating Scale score below 11
- Remission, intent to treat: 19 of 22 enrolled participants (86.4%) met the same criteria
- Treatment schedule: 50 iTBS sessions over 5 consecutive days, delivered as 10 daily sessions with a 50-minute intersession interval
- Dose and intensity: 1,800 pulses per session at 90% of resting motor threshold, adjusted for cortical depth
- Cognition and tolerability: Neuropsychological testing before and after showed no negative cognitive side effects, and the protocol was described as well tolerated and safe
This study, published in the American Journal of Psychiatry in 2020, is a feasibility, tolerability, and preliminary efficacy trial. It is the first step in evaluating this protocol, not a confirmatory one.
Dr. Kumar’s Take
I read this as a promising open-label signal, not a settled result. The remission numbers in this small sample are high for a treatment-resistant population, and the design choices behind them make mechanistic sense to me: individualized fcMRI targeting of the dorsolateral prefrontal to subgenual cingulate circuit, a much larger pulse dose, and sessions spaced through the day rather than spread over weeks. But 22 people, no sham arm, and everyone knowing they got active stimulation is exactly the setup where expectation effects run highest, and the authors say plainly that sham-controlled trials are needed. The clean neuropsychological testing is the part I weigh most confidently, because an open design does not inflate a cognitive battery the way it can inflate a mood rating. I would tell a patient this is worth watching and worth asking about at a center running it, while being honest that the confirmatory evidence is not in yet.
Study Snapshot
This was an open-label study of 22 participants with treatment-resistant depression. All of them received active SAINT. There was no randomization, no blinding, and no sham arm. One participant withdrew, leaving 21 completers.
Resting-state functional connectivity MRI was used in each participant to individually target the region of the left dorsolateral prefrontal cortex most anticorrelated with the subgenual anterior cingulate cortex. Fifty intermittent theta-burst sessions were then delivered as 10 daily sessions over 5 consecutive days, at 1,800 pulses per session, with a 50-minute interval between sessions, at 90% of resting motor threshold adjusted for cortical depth. Neuropsychological testing was conducted before and after the treatment course.
Results in Real Numbers
Nineteen of the 21 participants who completed the protocol, 90.5%, met remission criteria, defined as a score below 11 on the Montgomery-Asberg Depression Rating Scale. In the intent-to-treat analysis, which counts all 22 enrolled participants, 19 of 22 met remission criteria, or 86.4%.
Because every participant received active treatment, these figures have no sham comparator to be measured against. The authors’ own conclusion is that double-blinded sham-controlled trials are needed to confirm the remission rate observed in this initial study.
Neuropsychological testing performed before and after the 5-day course demonstrated no negative cognitive side effects. The authors concluded that the protocol was well tolerated and safe.
Who Benefits Most
The study enrolled adults with treatment-resistant depression, so that is the population the findings speak to. Nothing here establishes benefit in milder or first-episode depression, and nothing here tests the protocol against an alternative.
Practically, this is an intensive commitment: 10 sessions a day for 5 consecutive days, at a center with the functional MRI capability to derive an individualized target. Patients who cannot commit to that schedule, or who cannot reach a center offering fcMRI-guided targeting, are not candidates for this specific protocol.
Safety, Limits, and Caveats
On safety, the study reported that SAINT was well tolerated and safe, with neuropsychological testing showing no negative cognitive side effects.
The limits matter as much as the results. This was an open-label study in 22 participants, and one withdrew. There was no control condition, so the remission rate cannot be separated from placebo response, natural course, or the effect of intensive daily contact with a clinical team. The authors frame the work as an examination of feasibility, tolerability, and preliminary efficacy, and call for double-blinded sham-controlled trials to confirm what they saw. The protocol also depends on resting-state fcMRI to define the target, which restricts it to centers with that capability.
Practical Takeaways
- Read the 90.5% remission figure as an open-label, uncontrolled result in 21 completers, not as a controlled effect size
- Note the intent-to-treat figure of 86.4% in 22 enrolled participants alongside it
- The protocol is specific: 50 iTBS sessions over 5 days, 1,800 pulses per session, 50 minutes between sessions, 90% of resting motor threshold adjusted for cortical depth
- Individualized fcMRI targeting of the left DLPFC region most anticorrelated with the sgACC is a defining feature, not an optional add-on
- Neuropsychological testing before and after showed no negative cognitive side effects
- Wait for double-blinded sham-controlled data before treating the remission rate as established
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
FAQs
How is Stanford’s protocol different from regular TMS?
Three things. It uses resting-state fcMRI to individually target the part of the left dorsolateral prefrontal cortex most anticorrelated with the subgenual anterior cingulate cortex. It delivers a higher overall pulse dose, 1,800 pulses per session. And it compresses treatment into 10 sessions a day for 5 consecutive days, with 50 minutes between sessions, rather than spreading single sessions across a longer course.
Was this study controlled?
No. It was open-label: all 22 participants received active SAINT, with no sham arm and no randomization. The authors state that double-blinded sham-controlled trials are needed to confirm the remission rate.
How was remission defined?
As a score below 11 on the Montgomery-Asberg Depression Rating Scale.
Did the intensive schedule affect cognition?
Neuropsychological testing conducted before and after the protocol demonstrated no negative cognitive side effects.
Bottom Line
In an open-label study of 22 patients with treatment-resistant depression, 19 of 21 completers (90.5%) and 19 of 22 by intent to treat (86.4%) met remission criteria after 5 days of fcMRI-guided, high-dose intermittent theta-burst stimulation, with no negative cognitive side effects on neuropsychological testing. That is a strong preliminary signal from a small, uncontrolled trial. The authors’ own bottom line is the right one: double-blinded sham-controlled trials are needed to confirm it.
