The Evolution of Decentralized Clinical Trials: Blending Innovation and Regulation in the Digital Age
Nate Hughes
JLABS, Harvard Innovation Labs
Isaac R. Rodriguez-Chavez
4Biosolutions Consulting
D

ecentralized clinical trials (DCTs) are a paradigm shift from the more traditional site-centric models to more patient-centric approaches. A new development within DCTs is “smart” trial protocols that adapt dynamically to participant needs and data patterns. These protocols apply artificial intelligence toward real-time optimization of study design, thus allowing, in some cases, shorter trial durations, accelerated patient timelines, and generally better logistical outcomes. This is also where blockchain technology is increasingly coming to prominence, as it offers a secure method to manage participant consent and data sharing, long a bugbear for decentralized research in terms of data privacy and integrity.

The DCT Revolution: Technology to the Forefront

As defined by the US Food and Drug Administration (FDA), decentralized clinical trials (DCTs) “leverage technology to remotely collect and/or evaluate data from individuals participating in the trial.” The COVID-19 pandemic accelerated the evolution of DCTs, given the necessity for remote solutions that demonstrated new possibilities with decentralized methodologies. DCTs range on a continuum from hybrid models that incorporate traditional and remote elements to an entirely decentralized trial conducted wholly outside a traditional clinical setting.

The revolution of DCTs is very much enabled by a set of innovative technologies. Wearable devices and sensors that can be applied continuously provide real-time insights into health metrics from participants. Telehealth platforms enable consultations without the need to meet in person; in fact, this bridges the gap between researchers and participants. Digital patient-engagement tools, such as mobile apps, help improve communication and adherence. All the data from all the interactions are housed in a safe, secure environment, robust in storage and security protocols that guarantee the integrity and confidentiality of large volumes of digital data.

Probably the most critical DCT innovation path at the moment is the involvement of artificial intelligence (AI) and machine learning (ML) in patient enrollment and recruitment. New tools analyze large data sets for suitable candidates and even predict trial outcomes, targeting one of the most vexing challenges in clinical research: how to enroll patients efficiently. For instance, IBM Watson for Clinical Trial Matching at the Mayo Clinic successfully matched an additional 80% of patients in clinical trials for breast cancer over 11 months in 2018. Because of the technology, clinical trial coordinators were able to review more patients in a much shorter amount of time in identifying qualified candidates and cut down the time it took to screen a patient for eligibility in a clinical trial from one to two hours down to about 10 minutes. Similarly, Deep 6 AI was tasked to find qualified participants for a hard-to-find cardiac clinical trial. In 2016, Deep 6 AI unearthed it in a few minutes as part of a pilot study at Cedars-Sinai Medical Center, rummaging through 2.1 million patient records. This, if done by human reviewers, would take approximately six months. In contrast to the usual methods, it unearthed 16 times as many eligible patients.

A growing benefit of DCTs is the potential for “precision recruitment”: not just identifying eligible participants but, through advanced analytics and AI, pinpointing those most likely to benefit from a particular treatment due to unique genetic and environmental conditions. That, in turn, could result in more targeted, efficient trials and, ultimately, more personalized options for treatment. One industry example is Icarus Therapeutics, an early-stage TechBio company which uses a NextGen platform to remove the bottleneck of clinical trial enrollment for both principal investigators and patients by matching them within a defined zip code or remotely through ML and GenAI by reading inclusion/exclusion criteria in protocols.

A very new development within this field is the integration of digital twins into clinical trials. The virtual models of patients, which come through the amalgamation of real-world data with advanced simulations, can help predict which individual patients will respond to treatment, optimize dosing regimens, and even simulate the entire trial outcome well before the first physical patient enrollment is enrolled. This way, it can significantly reduce the time and hopefully costs conventionally associated with traditional trial design and execution. Digital twins can model multiple disease pathways, predict the efficacy of potential treatments, and forecast the transmission of diseases. One such example of digital twins in clinical trials is linked to the company Unlearn AI which uses TwinRCTs to shorten time to enrollment in late-stage studies by requiring fewer patients to achieve the same power as traditional clinical trial designs.

Regulatory Landscape: FDA’s Evolving Stance

The FDA has become increasingly supportive of DCTs, recognizing the potential of these trials to improve efficiency and diversify patient pools. This was corroborated when the agency issued a draft guidance in May 2023 on how to conduct a DCT, with recommendations on ensuring patient safety, maintaining data integrity, and considering technological issues. In fact, the FDA’s commitment to understand and regulate emerging technologies being used within clinical trials has been further underlined by establishing the Digital Health Advisory Committee.

The FDA has recently initiated a pilot program for “Real-Time Oncology Review” in DCTs. This program might allow the most important efficacy and safety data to be reviewed side by side with the FDA staff just when such data is generated, thereby accelerating the approval process of promising oncology treatments. If successful, this approach could be extended to other therapeutic areas, significantly speeding up drug development and approval processes going forward.

Benefits and Challenges of DCTs

The advantages of DCTs include increased patient recruitment and retention, more diversified study populations, real-time data capture and monitoring, reduced costs, and accelerated timelines. However, they face a number of challenges, like ensuring quality and integrity in remotely collected data, the problem that some participants may find with technological access, how to maintain regulatory compliance across multiple jurisdictions, and how to strike a balance in remote monitoring without compromising the safety of the patients.

Ensuring Fairness and Representation

While it is evident that the DCT can serve underrepresented populations better, the issue of the digital divide must be addressed with multilingual support and cultural barriers to participation. The integration of DCT with real-world data sources such as electronic health records and patient-generated health data would round out the patient health and treatment effects profile.

The development of “community-based trial hubs” represents a new frontier in diversifying populations within DCTs. Such local centers, often in underserved regions, combine the strength of decentralized technology with hands-on support and thus are poised to serve as a bridge for communities that have often been excluded from clinical research. This could be one way to increase representation in trials without losing efficiency. One such community-based trial hub is SiteBridge Health, which enables marginalized and underserved communities to take part in clinical research by decreasing administrative burdens, providing transportation to sites, language interpretation, and more efficient compensation.

The Future of Clinical Trials: Adaptive and Patient-Centric

This flexibility of DCTs is making adaptive designs of trials possible, whereby study parameters can be changed in response to interim data analyses. This could result in trials that are more efficient and informative. As the field evolves, we will likely see further DCT hybridization of trial models, more integration of AI and ML throughout the trial life cycle, global development of standards for implementing DCTs, and a stronger focus on patient-reported outcomes and real-world evidence.

With DCTs, one of the exciting frontiers is the idea of “living protocols”: trial designs that continue to evolve in response to real-time data and participant feedback (see box on “Hypothetical Examples”). This may enable it to quickly adapt to new scientific insights or changing patient needs and could revolutionize how we do these long studies, mostly for chronic conditions or rare diseases.

Conclusion: A New Era of Medical Research

DCTs are the future of medical research, enabling unparalleled opportunities to accelerate drug development, improve the patient experience, and diversify trial populations. As technology continues to evolve and regulatory frameworks change concomitantly, clinical trials will indeed become even more decentralized, data-driven, and patient-centric. Adopting these innovations means battling very real challenges. If this can be achieved, a new era of more efficient, inclusive, and impactful medical research will have arrived. As regulators (such as the FDA and others) continue to improve their guidances, they will play the critical role in ensuring that innovation in clinical trials truly translates into better health outcomes for all.

The next step for DCTs might be their integration with advanced biotechnology, such as organs-on-a-chip models and in silico trials. This could further reduce the need for human subjects during early-stage trials by accelerating the process of drug development at much less risk for patients. Moving forward, this synergy between decentralized approaches and cutting-edge biotechnology will be monumental in bringing a shift into the landscape of clinical research, potentially ensuring quicker and more targeted therapies for a wide range of conditions across a larger subset of patient populations.