Confirmatory efficacy and safety trial of magnetic seizure therapy versus right unilateral ultra-brief electroconvulsive therapy in depression (CREST–MST): a randomised, double-blind, non-inferiority trial in Canada and the USA

Daniel M. Blumberger et al.

Lancet Psychiatry. 2026;13:376–386.

 

Research Question

Is magnetic seizure therapy (MST) non-inferior to right unilateral ultra-brief pulse-width electroconvulsive therapy (RUL-UB ECT) for the treatment of major depressive disorder (MDD), while causing fewer cognitive adverse effects?

 

Study Location

Multicentre randomised controlled trial conducted at three academic centres:

• Centre for Addiction and Mental Health (Toronto, Canada)
• University of Texas Southwestern Medical Center (Dallas, USA)
• University of California, San Diego (USA)

Study Sponsor

·      This study was funded by a grant from the NIMH (R01MH112815).

Inclusion Criteria

Participants:

• Aged 18 years or older
• Diagnosed with non-psychotic major depressive disorder
• Baseline Hamilton Rating Scale for Depression (HRSD-24) score ≥21
• Considered suitable candidates for convulsive therapy (ECT and MST)
• Able to maintain stable antidepressant treatment during the study period

 

Exclusion criteria

• Recent substance dependence in the last 3 months
• Significant neurological disease or dementia
• Major unstable medical illness
• Intracranial metal implants
• Requirement for anticonvulsants or high-dose benzodiazepines (>2 mg lorazepam equivalent daily)

 

How Many Participants

• 292 participants assessed for eligibility
• 239 participants randomised
• 236 participants received treatment
• 119 participants received RUL-UB ECT
• 117 participants received MST
• 219 participants completed a full course of treatment or achieved remission

 

Study Intervention

Participants were randomly assigned to receive either MST or RUL-UB ECT until remission, withdrawal, or completion of 21 treatments.

MST characteristics

• Twin magnetic coils placed over the frontal midline (Fz position)
• Stimulation delivered at 100 Hz
• Progressive escalation of train duration
• Mean number of treatments: 14.2

RUL-UB ECT characteristics

• Right unilateral ultra-brief pulse-width stimulation
• Treatment administered at six times seizure threshold
• Mean number of treatments: 11.4

Follow-Up

Participants were assessed:

• At baseline
• Every three or four treatments
• At treatment completion
• Six months after treatment completion

Endpoints

Coprimary outcomes

• Remission of depression (HRSD-24)
• Worsening of autobiographical memory (25% reduction from baseline on Autobiographical Memory Test [AMT])

Secondary outcomes

• Response rates
• Suicidal ideation remission
• Anxiety symptoms
• Quality of life
• Subjective adverse effects
• Neurocognitive performance
• Time to reorientation after treatment

 

Results

Depression outcomes

• Remission rates:
• RUL-UB ECT: 27.8%
• MST: 22.5%
• Absolute difference: 5.3% (non-inferiority margin was 15%)
• MST met criteria for non-inferiority.

Response rates

• RUL-UB ECT: 48%
• MST: 47%

Suicidal ideation remission

• RUL-UB ECT: 48%
• MST: 48%
• Non-inferiority established.

Cognitive outcomes

Autobiographical memory worsened (binary outcome defined as a worsening from baseline of at least 25% in the AMT total score) in:

• 17.3% of RUL-UB ECT patients
• 2.7% of MST patients (p = 0.0003)

Compared with ECT, MST was associated with better outcomes in:

• Global cognition (MoCA)
• Verbal learning and memory
• Attention
• Executive function
• Verbal fluency
• Time to reorientation after treatment

Adverse effects

Compared with MST, RUL-UB ECT was associated with higher rates of:

• Memory impairment
• Jaw pain
• Headache
• Confusion

However, MST was associated with more post-treatment breathing difficulties.

There were:

• 13 serious adverse events overall
• No treatment-related deaths
• No suicide-attempt-related outcomes

 

Other Relevant Information

• Approximately 82% of participants had failed at least two antidepressant trials.
• Around one-third had previously received repetitive transcranial magnetic stimulation (rTMS).
• Most participants were treated as outpatients.
• Participants with psychotic depression were excluded.

 

Limitations

• Target recruitment was not achieved.
• The COVID-19 pandemic disrupted recruitment and treatment delivery.
• The trial excluded psychotic depression.
• Participants were younger than those typically treated with ECT.
• Electrode placement was not changed to bilateral treatment in ECT non-responders, reducing generalisability to routine clinical practice.

 

Summary and Implications

This trial demonstrated that MST produced antidepressant effects that were statistically non-inferior to RUL-UB ECT while causing substantially fewer cognitive adverse effects. Although remission rates (approximately 20-30%) were lower than anticipated in both groups, MST appeared to better preserve autobiographical memory and was associated with reduced transient cognitive side-effects while maintaining efficacy.

The findings suggest that MST could become an important alternative for patients who might otherwise decline ECT because of concerns regarding memory impairment and other cognitive adverse effects. Wider implementation will depend upon regulatory approval, clinician training, and further studies examining long-term outcomes and optimisation of treatment protocols.

 

Comments

Magnetic seizure therapy (MST) has been in development for over two decades and this study was the largest randomised controlled trial to-date. Rather than passing electrical current diffusely through the brain, MST induces therapeutic seizures using strong focused magnetic stimulation, which results in highly focalised brain stimulation compared with ECT, which instead generally stimulates the whole brain. This approach theoretically can preserve efficacy while reducing cognitive adverse effects, particularly by minimising stimulation of memory related brain regions. The current results suggested broadly similar antidepressant effects of MST compared to right unilateral ultrabrief ECT (RUL UB ECT), the type of ECT which shows the slowest antidepressant effects, though has the best cognitive side-effect profile relative to other types of ECT.

There are several considerations related to this study and its primary findings, both clinically and from a research perspective.

• Both ECT and MST were well-tolerated by participants
• While the study results support the utility of MST as a potential viable alternative to RUL UB ECT, for example, in people who refuse ECT and are non-psychotic, it does not support the use of MST as an alternative to ECT as treatment is most typically given in clinical practice. This is due to both how ECT was administered (i.e., with no change of electrode placement allowed due to minimal response and treatment continuing to up to 21 treatments) and the study population (mostly outpatients, overall younger relative to a more typical ECT population and all with non-psychotic depression).
• The low remission rates with both treatments generally do not support the use of both treatment approaches in this sample with these clinical characteristics (see above).
• It is also important to note that while the overall remission rates were broadly similar at post treatment, MST overall required a higher number of required treatments (mean 14.2 MST vs 11.2 ECT treatments) for participants to either achieve remission or complete the entire treatment protocol of up to 21 treatments. Thus, when considering whether to refer patients for MST, clinicians and patients would need to weigh the necessity for speed of antidepressant effects against risk for adverse cognitive effects.
• Last, due to the choice of the primary cognitive outcome measure in the study, the Autobiographical Memory Test (AMT), unfortunately, the cognitive results from this study with MST cannot be directly compared with other seminal studies of ECT, which predominantly have used the Columbia Autobiographical Memory Questionnaire – Short Form (C-AMI-SF). Retrograde autobiographical memory side effects are the most persistent adverse cognitive effect of ECT, while other adverse cognitive side effects are mostly transient. To establish whether MST definitively has a more favourable cognitive profile compared to RUL-UB-ECT and other forms of ECT, longer term follow-up data using the C-AMI-SF would be required.