Industry Insight

The promise of digital therapeutics for CNS conditions has coincided with the increasing popularity of wellness apps. However, digital therapeutics are distinct from these apps, since they provide an evidence-based therapeutic intervention and must be clinically evaluated – in many cases, requiring a clinical trial to demonstrate safety and efficacy. There are a number of aspects to these clinical trials that differ from traditional therapeutics trials, of which – due to the relative newness of digital therapeutics – sponsors may not be aware.

The majority of digital therapeutics in the CNS space are not intended to be used as stand-alone treatments. Instead, they may aim to improve daily functioning for people suffering from a severe illness, or provide ongoing, additional support for patients who have experienced some improvement to a condition following treatment but continue to experience impairment. Examples of digital therapeutics in the psychiatric space include bringing cognitive behavioural therapy (CBT) modalities to patients — for example, exercise modules on cognitive restructuring — and contingency management, which recognises and reinforces individual positive behavioural change. In neurology, some available digital therapeutics have focused on managing mood and anxiety symptoms associated with a specific neurologic indication, or providing physical and occupational therapy components to improve daily functions such as walking or writing.

Here, we explore how the present regulatory and payer landscape for digital therapeutics in CNS treatments impacts clinical trial design and additional elements of clinical trial design for digital therapeutics, which sponsors may need to consider.

All digital therapeutic products must demonstrate clinical evidence of safety and efficacy and, therefore, demonstrate clinical value. However, a digital therapeutic’s classification within the regulatory landscape will guide the requirements for a clinical trial. Within the US Food and Drug Administration’s (FDA) Center for Devices and Radiological Health, digital therapeutics are classified under ‘software as a medical device.’ Under the FDA, digital therapeutics are then classified as Class I, II or III medical devices.

Since Class I medical devices pose minimal risk to patients, nearly all of these devices do not require clinical trials for safety, and are excused from the regulatory process.1 In behavioural health, digital therapeutics are most often categorised as Class II medical devices, which are described as having a moderate risk to user safety.2 Finally, Class III medical devices pose a high risk to patient safety, encapsulating medical technologies such as life-supporting devices and implants.1 Digital therapeutics may also be grouped into one of three categories that describe intended purpose: disease treatment, disease management or health function improvement.3 Digital therapeutics must always administer a therapeutic intervention and support the claims of a product using clinical endpoints, regardless of the intended purpose.3

Meanwhile, payer coverage of digital therapeutic interventions can pose unique challenges as a result of contrasting evidence standards between disparate regulatory bodies, such as the Centers for Medicare and Medicaid Services (CMS) and the FDA. For example, the CMS requires that intervention be ‘reasonable and necessary,’ whereas the FDA demands that products be ‘safe and effective.’ Oftentimes payers follow CMS decisions because of its exacting formulary guidelines. Clinical trials should be designed to maximise data collection to ensure payer coverage, while also striving to minimise the length of the trial to keep participant burden low.

When designing clinical studies, it is important to account for the type of evidence that should be generated to ensure CMS and payer coverage. This often requires more granular or specific information, such as pinpointing exactly which populations might benefit from the intervention and including evidence of the long-term effects of a therapeutic.

Taking into consideration the need for evidence generation for regulatory review and payer reimbursement decisions, the specific type of data used to quantify endpoints should be carefully selected. Primary, secondary and exploratory endpoints measure the positive or negative impacts that a therapeutic may have on a participant of a clinical trial. When assessing these endpoints in CNS clinical trials for digital therapeutics, it is important for healthcare professionals to consider the various types of reported outcomes – patient-reported outcomes, performance outcomes, clinician-reported outcomes, observer-reported outcomes and even ecological momentary assessments – and which ones are appropriate to use for evaluating the different levels of study endpoints. For example, clinician-reported outcomes should be used with primary endpoints, which are generally efficacy measures that address the main research question.

To get long-term efficacy and safety data, sponsors may benefit from using tokenisation. Participant tokenisation links different sources of patient-level, real-world evidence – while, at the same time, protecting patient privacy – thereby giving researchers invaluable insights into new data trends. Tokenisation may offer a deeper understanding of patient journeys, providing past and current data on a participant’s medical history, medical events, concomitant medications and ongoing endpoint evaluation. Furthermore, tokenisation can integrate wider populations into the data set, improving sample diversity and overall comparative accuracy.

Data strategies are vital when designing a clinical trial with digital devices. The use of digital therapeutics in decentralised clinical trials (DCTs) often produces larger data sets than traditional trials, requiring a data infrastructure that is robust, adaptable and scalable. For example, a phase 2 clinical trial for CNS disorders that uses a 50Hz mobile sensor for movement data from a few dozen patients can easily approach one billion data points per day. The complexity of data sets in CNS trials is multilayered, since data is often combined with electronic patient-reported outcome software. Therefore, data will need to be integrated from a variety of different sources, making it necessary for companies to implement a strong cybersecurity plan to ensure that data coming from disparate sources is equally protected.

Without a sufficient number of willing participants, a clinical trial cannot run. In fact, in phase 3 clinical trials, participant drop-out rates can exceed 30%.4 Digital therapeutics’ capabilities are f lexible and can be used in hybrid and fully decentralised trials. This has the potential to ease costs and enhance participant recruitment, retention and diversity with its remote options, introducing f lexibility in trial participation and removing location as a limiting factor. For example, a DCT on a drug for heart failure symptoms recruited 52% more women and 17% more non-Caucasian people when compared to traditional chronic heart failure trials.5

While hybrid or fully decentralised trials for digital therapeutics may increase participant engagement by reducing a patient’s travel burden, decentralised elements of a clinical trial can also present challenges for engagement or compliance if patients feel ill-prepared or confused. To ensure patient support is fully embedded in decentralised or hybrid trials, patients should have access to an intuitive, end-to-end digital platform that creates a unified interface between the participant and the clinical trial, including apps that consider user experience and have features that reinforce study compliance. Patients should also be supported by dedicated concierge services, which complement the digital platform and collectively reduce patient burden. For example, a digital platform can provide patients with reminders and monitor compliance and concierge services can directly reach out to patients via text message when non-compliance occurs.

Additionally, sponsors should aim to incorporate digital therapeutics that are intuitive, easy to use and require minimal actions from the participant, allowing for seamless data collection. Providing staff contact information, such as phone number or email, within the digital therapeutic product can also help ensure patients will reach out if they encounter an issue.

Of course, safety should be paramount when conducting any clinical trial. This is magnified in digital therapeutic clinical trials, especially decentralised ones, in which participants will have far more interaction with the therapeutic than with investigators and healthcare providers, imposing further safety considerations, such as data protection. As such, when designing a CNS trial using digital therapeutics, it is important to consider safety monitoring, which should inf luence the type of data collected, frequency of data collection and the review of this data.

Additionally, in these trials, symptom severity should be assessed, enabling appropriate monitoring for decompensation. Both criteria and steps for intervention should be included in the protocol, in the event that a participant’s symptoms worsen. For example, in a trial for patients with major depressive disorder, suicidality assessment should be assessed at baseline and regular intervals, giving investigators, monitors and sponsors insight into the participant's risk.

To ensure their widespread use moving forward, digital therapeutics will have to demonstrate their usefulness, namely through improvements to clinical outcomes and healthcare savings. Moreover, they must address underserved needs in healthcare, including insufficient access to care. This is of the utmost importance in the CNS space, in which it is often harder to find success in clinical trials.6 As such, establishing the utility of CNS digital therapeutics in clinical trials will prove critical to their long-term adoption in patient care.