In this article you will learn:
- What neuromodulation is and which device types are studied in trials
- How EU and US rules split oversight between device and medicine pathways
- Why device trials use a surgery or fitting phase, programming visits and a control group
- Why the control condition in device trials is hard to design, and why a sham device is used
- What happens if a medicine is tested together with a device (drug–device combination)
- What side effects and follow-up checks you should report during and after a trial
- Where to search for recruiting trials, including clinicaltrials.eu
What is neuromodulation and which devices are studied in clinical trials?
Neuromodulation is the delivery of electrical stimulation to nerve tissue in order to change the activity of the nervous system. The devices used for this range from implanted to non-invasive brain stimulating and recording technologies, and further to digital therapeutics delivered through smartphone applications[1]. Because that definition spans such a wide range of technologies, the regulatory route a given study must follow depends less on the word "neuromodulation" than on what the product actually is. An implanted pulse generator, an external stimulator, or a software-driven therapeutic each sits under a different set of rules[2].
In drug-resistant epilepsy, three stimulation techniques are used, separately or in combination: vagus nerve stimulation (VNS), responsive neurostimulation (RNS) and deep brain stimulation (DBS). Each of the three is an implanted system, which means that the trial schedule has to accommodate surgery, device programming and long-term follow-up rather than only the visit-based assessments familiar from medicine trials[3].
The non-invasive techniques most often examined in trials are transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS)[4]. These approaches avoid implantation, but they bring their own design challenges, most notably the difficulty of keeping participants and therapists unaware of whether active or inactive stimulation is being delivered[5].
How do the three implanted epilepsy systems differ?
- Vagus nerve stimulation (VNS) – received FDA approval in 1997 and is the oldest neuromodulation therapy for drug-resistant epilepsy; the system consists of an implantable pulse generator placed subcutaneously in the chest, a lead cuffed around the vagus nerve in the neck, and an external programming system[3].
- Responsive neurostimulation (RNS) – approved in 2013, works much like a heart pacemaker: the device is implanted in the skull and attached to one or two leads placed at seizure foci, and the closed-loop system senses abnormal electrical activity and triggers modulating impulses[3].
- Deep brain stimulation (DBS) – approved in 2018 for focal epilepsy, with electrodes typically placed in the anterior nucleus of the thalamus[3].
Who regulates neuromodulation trials: the European vs US picture?
In Europe, medical devices are regulated at EU Member State level, but the European Medicines Agency (EMA) is involved in a number of regulatory processes. That division of responsibilities matters for sponsors planning a neuromodulation trial, because the device component and any medicine component may be reviewed by different bodies under different procedures[2].
A manufacturer can place a CE (Conformité Européenne) mark on a device once it has passed a conformity assessment. This assessment usually involves an audit of the manufacturer's quality system. It can also include a review of technical documentation on the safety and performance of the device[2].
EU Member States designate accredited notified bodies to conduct those conformity assessments, and for certain high-risk devices notified bodies must consult expert panels before issuing a CE certificate. For an implanted neuromodulation system, this is the stage at which the technical documentation on safety and performance is examined before the device can carry a CE mark[2].
Regulators themselves describe the two systems as different paths. The FDA in the United States and the European Commission (EC) in Europe are each charged with regulating neuromodulatory devices. Their different paths to device approval have resulted in varying availability of these devices between countries[5].
The EC works through Competent Authorities designated in individual countries. These authorities have enforcement powers. They designate Notified Bodies to assess applications for CE marking. CE marking indicates that a product complies with EU regulations and may be marketed in that country[5].
The timing differences are documented. The Neuronetics NeuroStar system for the treatment of depression received FDA approval in 2008. CE Mark approval was not obtained until 2012. The Brainsway TMS system for depression received CE Mark approval several years before receiving FDA clearance in 2013. These examples show that a device available to patients in one region may still be under review in another. This is one reason trial availability can differ across countries[5].
For comparison, in the United States medical devices are subject to the regulatory controls in the Federal Food, Drug, and Cosmetic Act (FD&C Act). The Center for Devices and Radiological Health (CDRH) is responsible for the review and authorisation of devices under its purview[1].
How is a neuromodulation trial designed, and what does the participant do?
Regulators point to three questions that are both important and difficult in trials of neuromodulatory devices: the choice of control or comparison condition, determining dose–effect relationships, and patient heterogeneity. Each of these questions shapes what a participant will actually be asked to do during a study, from the type of comparison arm they may be assigned to, to how often stimulation settings are adjusted[5].
For sponsors seeking regulatory approval of devices in the United States, the FDA has encouraged the use of new tools, including innovative clinical trial designs such as adaptive trials as well as modelling and simulation[5].
In stimulation trials the "dose" is a set of device parameters, so programming visits are part of the study schedule. In VNS the therapy delivers continuous, intermittent electrical stimulation to the vagus nerve, with settings that can be customised for individual patients. Because those settings are individualised, programming sessions become a recurring point of contact between participant and study team, and changes made there are part of the data collected[3].
Why is blinding difficult in stimulation trials?
Many TMS studies use sham controls, but because TMS is intrinsically multisensory, it is difficult to blind the patient and therapist. The sensations and sounds produced by the coil give cues that a sham condition cannot fully mask, which is why blinding is treated as a recognised design problem rather than a routine procedural detail[5].
A sham device looks and is used like the active one but does not deliver the intended stimulation. In one randomised, controlled, double-blind trial in multiple sclerosis, 20 patients with gait deficits took part: 10 used the active device and 10 used a sham control device that did not deliver stimulation[6].
A separate study assessed the effects of the same device versus sham, combined with cognitive and physical rehabilitation, in 14 patients with multiple sclerosis, 7 in each group[6]. The small group sizes in both studies illustrate a point made in the wider literature: many stimulation trials remain small feasibility studies rather than large confirmatory trials[4].
What did a long-term device trial show?
According to initial and long-term clinical studies, the median reduction in seizure frequency with responsive neurostimulation was 44% at 1 year and 53% at 2 years, increasing to 75% by year 9. The progressive improvement over years of follow-up is the kind of outcome that requires long-term study schedules and sustained participant commitment[3].
In the same epilepsy programme, about 40–50% of patients with VNS have a seizure reduction of 50% or more, but less than 5% become seizure-free. This distinction between reduction in seizure frequency and freedom from seizures is important when reading trial results, because the two are not interchangeable outcomes[3].
What changes when a medicine is tested together with a device?
Many medicines need a medical device to be used, and with ongoing advances in technology the number of new therapies that combine medicines and devices is increasing. For a participant, this means a single study may involve both a medicine assessed under medicines rules and a device assessed under device rules[2].
EMA assesses the safety and effectiveness of medicines used in combination with a medical device within the centralised procedure application for the medicinal product[2].
Where a device contains an ancillary medicinal substance, the notified body must seek an EMA scientific opinion. The opinion covers the quality, safety and usefulness of that substance. This applies when it is derived from human blood or plasma. It also applies when it has been previously evaluated by EMA. This is a distinct route from the assessment of a medicine used with a device. It applies to the substance built into the device itself[2].
A current example is a combined drug–device programme: on 7 May the FDA approved the launch of a phase 2/3 clinical trial of NRX-101, a fixed-dose combination of D-cycloserine-lurasidone, when added to transcranial magnetic stimulation for treatment-resistant depression and suicidality[7].
The randomised, placebo-controlled MIND1 study will include approximately 240 participants at multiple sites, with additional participants recruited from military treatment facilities. In a programme like this, the medicine and the stimulation are both part of the intervention, so the trial design has to account for two active components at once[7].
How does a device or combination therapy reach the market?
| Route | What is assessed | Who assesses it | Documented example |
|---|---|---|---|
| Device conformity assessment (EU) | Manufacturer's quality system; technical documentation on safety and performance, depending on device type | Accredited notified bodies designated by EU Member States[2] | CE mark placed after a passed conformity assessment[2] |
| High-risk device (EU) | Conformity assessment with consultation of an expert panel before the CE certificate is issued | Notified body plus expert panel, with EMA technical and scientific support[2] | Expert panels benefit from EMA's technical and scientific support[2] |
| Medicine used with a device (EU) | Safety and effectiveness of the medicine | EMA, within the centralised procedure application for the medicinal product[2] | – |
| Device with ancillary medicinal substance (EU) | Quality, safety and usefulness of the ancillary substance | EMA scientific opinion requested by the notified body[2] | Applies where the substance is derived from human blood or plasma or has been previously evaluated by EMA[2] |
| New low- to moderate-risk device (US, for comparison) | New device not substantially equivalent to an already marketed device | FDA, de novo pathway, Class II[5] | Cerena TMS system for migraine preceded by aura, approved in 2013[5] |
The FDA awarded de novo approval on 26 May to a wearable neuromodulation device for PTSD treatment. The manufacturer reported it as the first neuromodulation device approved for PTSD symptoms under the de novo pathway[7].
In 2021 the FDA approved a portable neuromodulation stimulator. It is for the short-term treatment of walking impairments resulting from mild to moderate symptoms of multiple sclerosis. It is a prescription-only device. It is intended to be used together with a supervised therapeutic exercise programme in patients aged 22 years and older[6].
The same device had breakthrough device designation and was approved under the de novo premarket review pathway for new low- to moderate-risk devices[6]. The two examples together show that the de novo pathway has been used both for external stimulators and for wearable neuromodulation devices intended for psychiatric indications[7].
How mature is the evidence from stimulation trials?
A bibliometric review examined the 100 most-cited non-invasive neuromodulation clinical trials. The total citation count for these trials was 13,204. The articles were published between 2008 and 2014 in 50 different journals. The median impact factor was 6.52[4].
Almost half of those top-cited papers investigated mechanisms of action in healthy subjects, most studies were feasibility trials, and only 5 were pivotal trials, including the ones used for recent FDA approval[4].
The authors concluded that most of the cited clinical trials are still small feasibility studies, reinforcing the need for more robust clinical trials with larger sample sizes and effect sizes to better define clinical effectiveness. For anyone reading the results of a single neuromodulation study, this is a reminder that a frequently cited paper is not necessarily a large confirmatory trial[4].
What safety information and follow-up checks matter to a participant?
Safety information in device trials is collected during the studies and afterwards. In the multiple sclerosis studies of the portable stimulator, no serious safety adverse events were reported in clinical studies or in separate real-world data provided by the company[6].
The same authorisation listed groups who should not use the device, including patients with penetrating brain injuries, neurodegenerative diseases, oral health problems, chronic infectious diseases, unmanaged hypertension or diabetes, pacemakers, and/or a history of seizures. Knowing this list in advance allows a prospective participant to raise relevant history with the study team before enrolment[6].
Because that device delivers electrical stimulation directly to the surface of the tongue, precautions for its use are similar to those for transcutaneous electrical nerve stimulation. Electrical stimulation should not be used in patients with an active or suspected malignant tumour. It should not be used in areas of recent bleeding or open wounds, or in areas that lack normal sensation[6].
That device is not indicated in people younger than 22 years or in pregnant women, and people with a sensitivity to nickel, gold or copper should not use it[6].
For an implanted system, the settings that shape your exposure are adjusted during programming, so changes in stimulation settings are part of what you report on. VNS delivers continuous, intermittent electrical stimulation with settings that can be customised for individual patients[3].
How can you find a neuromodulation trial, and what are your rights as a participant?
Before agreeing to take part, ask the study team which control condition the trial uses, because the choice of control or comparison condition is one of the recognised difficult design questions in trials of neuromodulatory devices[5].
Ask whether the comparison is a sham device, and whether you or the therapist will know which one you receive, since blinding can be difficult in stimulation studies[5].
For a combination therapy, ask separately which part is the medicine and which part is the device. In the EU, the medicine is assessed by EMA within the centralised procedure. The device component follows conformity assessment requirements[2].
Recruiting studies can be searched in public registries; trials recruiting in Europe can also be searched on clinicaltrials.eu.
Summary
Neuromodulation trials test electrical stimulation delivered by implanted or wearable devices, and in the EU the device side follows conformity assessment by notified bodies while a medicine used with a device is assessed by EMA within the centralised procedure. Regulators name the control condition, dose–effect relationships and patient heterogeneity as the hard design questions, which is why programming visits, a sham arm and long follow-up appear in the schedule[2].
Before signing up, ask which control condition is used. Ask whether blinding is possible in the study. Ask which part of a combination is the medicine and which is the device. Note the listed contraindications for the specific device so they can be raised with the study team[5].
❓ What is neuromodulation?
Neuromodulation is electrical stimulation of nerve tissue intended to change nervous system activity. The devices studied include implantable and non-invasive brain stimulating and recording technologies as well as digital therapeutics delivered through smartphone applications. In drug-resistant epilepsy the techniques used are vagus nerve stimulation (VNS), responsive neurostimulation (RNS) and deep brain stimulation (DBS), used separately or in combination. In research settings, the most frequently examined non-invasive techniques are transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS).
❓ Why do neuromodulation trials use a sham device?
A sham device is used to create a comparison group that does not receive the intended stimulation. In one randomised, controlled, double-blind trial in multiple sclerosis, 20 patients with gait deficits took part: 10 used the active device and 10 used a sham control device that did not deliver stimulation. Sham controls are also used in TMS studies, although because TMS is intrinsically multisensory it is difficult to blind both the patient and the therapist. Ask the study team which control condition your trial uses.
❓ What is a drug–device combination therapy in a trial?
Many medicines need a medical device to be used, and the number of new therapies combining medicines and devices is increasing. In the EU, EMA assesses the safety and effectiveness of a medicine used in combination with a medical device within the centralised procedure application for the medicinal product. A current example is the phase 2/3 MIND1 study of NRX-101, a fixed-dose combination of D-cycloserine-lurasidone, added to transcranial magnetic stimulation in treatment-resistant depression and suicidality, planned for approximately 240 participants at multiple sites.
❓ How is a medical device authorised in Europe?
Medical devices are regulated at EU Member State level, with EMA involved in a number of regulatory processes. A manufacturer can place a CE mark on a device once it has passed a conformity assessment, which usually involves an audit of the manufacturer's quality system and can include a review of technical documentation on safety and performance. EU Member States designate accredited notified bodies to conduct these assessments, and for certain high-risk devices the notified body must consult an expert panel before issuing a CE certificate.
❓ What results have stimulation trials shown?
Results vary by device and condition. According to initial and long-term clinical studies, the median reduction in seizure frequency with responsive neurostimulation was 44% at 1 year and 53% at 2 years, increasing to 75% by year 9. In the same epilepsy programme, about 40–50% of patients with vagus nerve stimulation have a seizure reduction of 50% or more, but less than 5% become seizure-free. A bibliometric review of the 100 most-cited non-invasive neuromodulation trials found that most were small feasibility studies and only 5 were pivotal trials.
❓ Which side effects and restrictions should I know about?
Restrictions are specific to the device. For the portable tongue stimulator approved in 2021 for walking impairment in multiple sclerosis, the listed groups who should not use it include patients with penetrating brain injuries, neurodegenerative diseases, oral health problems, chronic infectious diseases, unmanaged hypertension or diabetes, pacemakers, and/or a history of seizures. It is not indicated in people younger than 22 years or in pregnant women, and people with sensitivity to nickel, gold or copper should not use it. Report any new symptom to the study team.
❓ Where can I search for a trial?
Recruiting studies are listed in public registries, and trials recruiting in Europe can also be searched on clinicaltrials.eu. When reviewing a listing, check which control condition the study uses, because the choice of control or comparison condition is one of the recognised difficult design questions for neuromodulatory device trials. Check also whether the study is randomised and blinded, and whether it involves a surgical implant or a non-invasive technique, since these determine the visits and recovery time you will need to plan for.
- [1] Navigating the FDA regulatory landscape - PMC — PubMed Central (accessed 20.10.2026) — https://pmc.ncbi.nlm.nih.gov/articles/PMC10700528/
- [2] Medical devices (accessed 02 June 2026) — https://www.ema.europa.eu/en/human-regulatory-overview/medical-devices
- [3] Neuromodulation therapy offers epileptic seizure relief (accessed 02 June 2026) — https://consultqd.clevelandclinic.org/neuromodulation-therapy-offers-epileptic-seizure-relief
- [4] Top 100 cited noninvasive neuromodulation clinical trials - PMC — PubMed Central (accessed 20.10.2026) — https://pmc.ncbi.nlm.nih.gov/articles/PMC6565369/
- [5] Regulatory issues - Non-invasive neuromodulation of the central nervous system (accessed 02 June 2026) — https://www.ncbi.nlm.nih.gov/sites/books/n/nap21767/ch8/
- [6] FDA OKs neuromodulation device to improve gait in MS (accessed 02 June 2026) — https://www.medscape.com/viewarticle/948215
- [7] FDA psychiatry roundup: May 2026 (accessed 02 June 2026) — https://www.medscape.com/viewarticle/fda-psychiatry-roundup-may-2026-2026a1000iop
