Can Deep Brain Stimulation Finally Deliver for Depression?

Deep brain stimulation (DBS) has transformed the treatment of Parkinson’s disease (PD), where stimulating a precise node in the brain’s motor circuitry can stop a tremor in seconds. The FDA approved it for advanced PD in 2002. Seven years later, the agency authorized DBS for obsessive-compulsive disorder (OCD) under an effective humanitarian device exemption.Researchers thought the next success story would come in treatment-resistant depression (TRD). But in 2017, when results from the BROADEN study — the largest randomized trial of DBS for TRD — were published in The Lancet Psychiatry, the much-anticipated findings were disappointing. Six months after implantation, there was no significant difference in symptom improvement between active and sham stimulation.Media coverage, and much of the field, suggested the results were proof that DBS did not work for depression.Paul E. Holtzheimer, MDBut the trial’s investigators didn’t accept that framing. While the trial missed its primary endpoint at 6 months, an open-label phase in which all participants received active stimulation showed response rates at 2 years far higher than the blinded phase had produced. The durability of the response — in patients who had failed every prior treatment — was difficult to dismiss.“The treatment didn’t fail. The study failed,” said Paul E. Holtzheimer, MD, the lead author of the Lancet Psychiatry paper and professor of psychiatry and surgery at the Geisel School of Medicine at Dartmouth in Hanover, New Hampshire.Now, investigators are trying again with the TRANSCEND trial, a multicenter, sham-controlled study of DBS in TRD. The first patient was implanted at Mount Sinai, New York City, in early 2025, and the trial is now recruiting across 25 US sites.Alik Widge, MD, PhDTRANSCEND reflects what researchers have learned since the first trial, but there is more riding on the findings this time, said Alik Widge, MD, PhD, psychiatrist and biomedical engineer at the University of Minnesota in Minneapolis and TRANSCEND site investigator.“If we miss the primary endpoint again,” he said, “we are going to mess up our field so badly.”A Pacemaker for the BrainIn DBS, a neurosurgeon threads thin electrodes, each roughly the diameter of a piece of spaghetti, deep into a target region in the brain, then connects them to a pulse generator implanted in the upper chest.The device delivers continuous electrical stimulation at a high frequency, typically 130 Hz, sending rapid pulses along the white-matter fiber tracts that connect one brain region to another. Everything sits beneath the skin.The same stimulation parameters are used across PD, OCD, and depression. “It’s almost like aspirin,” said Brian Kopell, MD, professor of neurosurgery at the Icahn School of Medicine at Mount Sinai and the national neurosurgical principal investigator for TRANSCEND.Brian Kopell, MDFor depression, DBS set at these frequencies sends a current through the white-matter fibers running through the target area in the brain, jamming connections that pass signals from one neuron to the next. The idea is to break up the faulty electrical activity that drives the disorder.In PD, the therapy transformed care because the science preceding it was airtight. Decades of research had mapped the motor circuit precisely enough to model aspects of the disease in monkeys, interrupt it with lesions, then substitute stimulation and watch the tremor stop. The readout was immediate and visible. “It literally was step by step: disease, pathophysiology, animal model that absolutely replicated, interrupt that circuit with lesions, substitute a lesion with stimulation, and then the rest has been refinement,” said Helen S. Mayberg, MD, founding director of the Nash Family Center for Advanced Circuit Therapeutics at the Icahn School of Medicine at Mount Sinai in New York City. In 2005, Mayberg published the first study of DBS targeting the subcallosal cingulate white matter in the brain for depression, a pilot in six patients. Those results, and the open-label studies that followed, led to BROADEN.Helen S. Mayberg, MDBut the solid foundation on which DBS for PD was built didn’t exist for depression. There is no blood test, brain scan, or biomarker to confirm a depression diagnosis. There are no animal models that replicate the full syndrome. There is no equivalent of the tremor seen in PD. Instead, researchers had to rely on subjective symptom questionnaires administered over weeks and months, knowing that mood fluctuates naturally and the placebo response to invasive procedures can be substantial and durable. Both of those factors could skew results.What Went Wrong in BROADEN?BROADEN began in 2008, with plans to enroll 201 patients with major depressive disorder whose symptoms didn’t improve after at least four antidepressant treatments. After screening, surgeons implanted electrodes in the subcallosal cingulate white matter in the brain. Patients were then randomly assigned in a 2:1 ratio to active or sham stimulation for 6 months. Following that double-blind phase, they completed a 6-month open-label extension, with all patients receiving active DBS and treatment continuing beyond that as part of a long-term, naturalistic study.Four years later, as investigators prepared to conduct an FDA-required futility analysis, 90 patients at 13 sites across North America had joined the trial. Researchers had expected a 40% response rate in the active arm after 6 months of treatment. But only 20% of participants met the primary outcome, compared to 17% in the sham arm — a difference that was not statistically significant. The futility analysis estimated a 17% probability of success if the study continued, and the sponsor, St. Jude Medical (now Abbott), decided to stop the trial.Among those who completed the study, response rates at that time were 29% at 12 months, 53% at 18 months, and 49% at 24 months. Remission rates were 14%, 18%, and 26% at those timepoints, respectively.When the paper was finally published in The Lancet Psychiatry in 2017, the open-label response rates were reported in the body text but not in the abstract.“When that paper came out, everybody said, ‘Oh, it failed,’” Mayberg said. “It’s not a failed trial; it’s a halted trial.”Did the treatment fail, or did the execution falter? Before BROADEN, smaller open-label studies at single centers had consistently shown response rates to DBS of 50% or higher in TRD. Something in the transition to a large, multisite trial had diluted the effect, but researchers didn’t know what. And why did some patients respond to treatment while others didn’t?Mayberg’s team at Emory University, Atlanta, began looking for answers. One of the key issues was the problem with electrode placement. When viewed with conventional scans, it appeared that the electrodes were placed in the same location within the subcallosal cingulate in every patient. But a closer look with a specialized MRI technique called diffusion tractography told a different story. Introduced at the National Institutes of Health in 1994, tractography traces structural pathways in the brain by measuring how water moves through tissue. In white matter, water diffuses along axons rather than across them, so the direction of diffusion reveals the orientation of the fibers.Researchers found that in those patients who responded to therapy, the electrodes sat at a crossroads where four white-matter bundles intersect. These bundles are called the forceps minor, cingulum bundle, uncinate fasciculus, and frontostriatal fibers. If a surgeon placed electrodes that missed even one of these bundles, the patient didn’t respond. The antidepressant effect depended on stimulating the precise location where the bundles converged.“Not being on target is like giving the wrong dose in a drug trial,” Kopell said. “You would come up with garbage.”To test the bundle convergence hypothesis, Mayberg’s team conducted a prospective study in a new set of 11 patients, using tractography to place each electrode at the bundle convergence from the outset.The response rate exceeded 70% at 6 months after the surgery and climbed to 82% at 1 year, more than three times what the same team had achieved with the anatomic targeting used in BROADEN.But targeting wasn’t the only problem in BROADEN. During the 6-month blinded phase, the protocol prohibited patients from starting new psychotherapy or medications. TRANSCEND allows patients to continue psychotherapy during the trial.“We have learned that recovery is more than a stimulator,” Mayberg said.Reducing the NoiseIf BROADEN’s active arm underperformed because of misplaced electrodes, the challenge became how to ensure that the correct brain target is hit every time.When investigators designed the TRANSCEND trial, they added a centralized review process to ensure correct electrode placement. Kopell likened the system to the protocol for launching a nuclear weapon.“There have to be two keys turned. The surgeon implants and turns one key. The centralized system turns the other key. Then the patient gets randomized,” he said.TRANSCEND also narrowed the patient population compared to the previous trial, which enrolled anyone with TRD who had failed four or more medications, regardless of how their depression actually presented. The new trial uses a centralized psychiatry review group to evaluate video interviews of each prospective patient and determine whether the presentation fits the phenotype the treatment targets.“We’re trying to have better rules about the phenotype — how to avoid people who don’t have what we’re treating,” Mayberg said. “DBS does not treat anxiety. Anxiety is on the checklist for depression, but it’s actually a different circuit.”The blinded phase was also extended from 6 to 12 months, reflecting the accumulating evidence that DBS for depression works slowly and that some patients may require longer periods of chronic stimulation to achieve remission.Recent research may explain why. A study co-authored by Mayberg showed that chronic subcallosal cingulate stimulation in monkeys thickened the myelin sheath on nerve fibers in the cingulum bundle, one of the four tracts DBS must target to achieve results. The finding has since been observed on repeat scanning in human patients.“We are repairing the actual broken highway that connects the front of the brain to the back of the brain,” Mayberg said. Years of depression, she said, may wear ruts into that highway. A course of DBS therapy appears to help repave it.Beyond DepressionOCD is the only psychiatric indication for which DBS is authorized in the US, and it is only allowed under a humanitarian exemption, a route reserved for rare conditions that requires evidence of probable benefit rather than proof of efficacy.The evidence has been remarkably consistent: a meta-analysis of 34 studies involving 352 patients showed a 47% reduction in OCD symptom scores and a 66% response rate. The treatment has shown benefit at centers around the world, across decades and with varying surgical techniques. Yet fewer than 600 patients have ever received it.“The number of centers in the US that do more than five a year — there are barely five,” said Sameer Sheth, MD, PhD, professor of neurosurgery at Baylor College of Medicine in Houston and a site investigator in TRANSCEND.Sameer Sheth, MD, PhDSheth identified three obstacles: lack of awareness among referring psychiatrists, the lingering association with lobotomy, and reimbursement barriers. Even in PD, where the evidence is unambiguous, only about 10% of eligible patients undergo DBS.Widge, who performs DBS for OCD under the humanitarian exemption at his center in Minnesota, has never received a single referral from another psychiatrist.“Every patient finds me on their own,” he said, either through social media, online forums, or word of mouth. “I will know things have really changed when I start getting referrals from other psychiatrists.”Yet researchers aren’t deterred, as they’re also exploring DBS for Tourette’s syndrome, where studies have shown reductions in tic severity exceeding 50% in many patients. A completed trial at University College London, London, England, Op-TICS, studied DBS in 20 patients using a double-blinded crossover design, and a small trial conducted at the University of Florida, Gainesville, found a closed-loop approach, being stimulated only when tic-related brain signals were detected rather than continuously, to be feasible and safe.But a competing pharmacologic approach may narrow the eligible population: ecopipam, a first-in-class dopamine D1 receptor antagonist, significantly reduced tic relapse in a phase 3 trial and is awaiting FDA review.What Comes NextDBS is not the only neuromodulation option for patients with TRD. Accelerated transcranial magnetic stimulation protocols have achieved high remission rates in small trials over 5 days with no incision, though durability beyond 6 months remains a challenge. Focused ultrasound can create targeted brain lesions without opening the skull, though the lesions are permanent.DBS is the most precise and adjustable of the three, and the only one that can be turned off. It is also the only one that requires a craniotomy and lifelong hardware management.Emerging biomarkers currently under study may eventually help clinicians verify that stimulation is engaging the right circuit, but Widge said they aren’t likely to make a self-adjusting device. In PD, a biomarker can signal the need to alter stimulation frequency in near real-time. But biomarkers for OCD and other behavior disorders shift weeks or months before symptoms reflect the change. So those models will look less like a self-adjusting pacemaker and more like checking a drug blood level. “When we see these biomarkers, what they tell us is: ‘I have dosed the stimulation correctly for your brain,’” Widge said. “I’m going to let it be and let it do its work.”If TRANSCEND succeeds, the next hurdle will be scaling the therapy. There are only a handful of centers with deep experience in DBS for depression. Expanding access would mean opening centers without that background. This happened with DBS in PD, Holtzheimer said. When the procedure was concentrated at high-volume centers with expertise across neurology, neurosurgery, and psychiatry, outcomes were strong. As it expanded to hospitals with less concentrated expertise, response rates fell and complications rose.For now, the field waits on TRANSCEND. The trial is enrolling at up to 25 US sites, though manufacturer Abbott has not released how many patients have been implanted or when results are expected. With a 12-month blinded phase followed by 2 years of open-label follow-up, definitive data are still years away.A lot is riding on the success of the trial, Widge said. “If DBS for depression doesn’t survive this trial, I don’t know what brings it back.”Mayberg reported consulting and licensing of intellectual property to Abbott Labs. The intellectual property was licensed in 2003. She is an ongoing consultant to the Abbott-sponsored TRANSCEND trial. Widge reported equity in Resilient Neurotherapeutics, a DBS company. He is a site investigator and design committee member for the TRANSCEND trial. Sheth reported co-founding Motif Neurotech (formerly Leaf), a company developing an implantable TMS device. He is a site investigator in the TRANSCEND trial. Holtzheimer reported receiving royalties from Oxford University Press for co-editing a book on transcranial magnetic stimulation and from UpToDate for co-writing several articles on brain stimulation in depression. Kopell has consulted on DBS technology for Abbott Neuromodulation.