By: Susan Lubejko PhD and Abigail Copeland PhD, Regulatory Affairs Specialists at Mittal Consulting

For both drugs and devices, personalized therapies that target specific patient biomarkers and tailor treatments to individual biological needs are the next big thing. Historically, the FDA has opted for device clearances with narrow permutations that don’t differ much between patients. Given the promise of these paradigms and the potential mismatch with the typical regulator stance, case studies that demonstrate success, especially those that use unconventional submission strategies, are important for device manufacturers looking to innovate in this space.

Wave Neuroscience’s recent 510(k) clearance for the MeRT System (K260402) provides an intriguing view into potential regulatory pathways for introducing individualized therapy into well-established neurotech fields. The Magnetic EEG/ECG Resonance Therapy (MeRT) System’s proprietary machine-learning driven algorithm (MRT-005) uses patients’ EEG/ECG recordings measuring brain activity at rest to predict the ideal, personalized transcranial magnetic stimulation (TMS) treatment parameters to program into compatible TMS hardware to deliver therapy to patients with post-traumatic stress disorder (PTSD).

MeRT is the first to bring both treatment for PTSD and use of individualized therapy to the established TMS space. Existing FDA-cleared TMS treatments are only “individualized” based on a field-standard procedure to determine each patient’s intensity threshold, not neural signatures. Stimulation location, frequency, pulse pattern, and intensity in relation to that threshold are typically locked in premarket submissions. Wave Neuro received clearance for a range of some of these parameters, allowing for personalization for each patient.

While the novelty of Wave Neuro’s innovation lies in its proprietary algorithm component MRT-005, the company successfully secured their 510(k) using an established MagVenture TMS hardware system indicated for obsessive compulsive disorder (OCD) as the predicate. Because 510(k) submissions require demonstration of equivalence in intended use and technology to an FDA-cleared device, using a hardware predicate for a software algorithm is usually unsuccessful. How was Wave Neuro able to introduce therapy individualization to the established TMS field using this unconventional argument?

Wave Neuro used three important strategies that all likely contributed to this success.

1. First, building upon an established partnership with MagVenture, Wave Neuro posed their device as a two-component system: their MRT-005 algorithm and another manufacturer’s compatible hardware system. Wave Neuro’s 510(k) submission focuses heavily on TMS technology itself, with less discussion of the treatment-defining algorithm.
2. Second, MRT-005 can only recommend treatment parameters that already fit within the existing operating specifications of the MagVenture hardware (e.g. MRT-005 may recommend stimulation frequency between 8 and 13 Hz, which is well within the 0.1 – 30 Hz range that the hardware can already accommodate). These first two points plausibly allowed FDA to conclude that there are no new patient risks associated with keeping the system within its existing limits.
3. Third, Wave Neuro conducted a clinical study to support the shift from OCD to PTSD and demonstrate that personalization of the TMS parameters is at least as effective as standard, non-personalized TMS therapy. This follows a common thread in the TMS space to use clinical data to validate treatment of new disease states and further allowed Wave Neuro to emphasize that the personalized stimulation is just as safe as the established methods FDA is used to reviewing.

While the particulars of this predicate argument are nontraditional from a regulatory perspective, the MeRT clearance demonstrates that a move toward individualized neurological therapy is underway and becoming accepted by US regulators. For submissions, FDA’s neurology branch appears to be open to introduction of novel software algorithms for this purpose based on the demonstration that the outputs work within the specifications of cleared hardware. Manufacturers of software innovations may consider similar partnership with hardware manufacturers, or at least the argument of matching specifications, in discussions with FDA review teams to further decrease the regulatory burden of new software-driven innovations.