ata transparency, more collaboration, and technology will lead to faster, more thorough, data-driven decisions and speed up medicine development.

Unlimited Complexity Versus Unlimited Unmet Needs: Practical Solutions Required

Each of us is built from approximately 30 trillion cells, each of them containing hundreds of thousands of moving parts, constituting an almost infinitely complex feedback loop. Each of us is different, not only from any other human who ever existed, but also from ourselves at any other moment in time. Against this background, the fact that we can create safe and efficacious treatments is almost miraculous.

But even though we live in the age of speed, where everything needs to be fast or faster, these miracles are, tragically, not happening fast enough. In fact, they are becoming ever harder to come by, a phenomenon known as Eroom’s law. Furthermore, despite streamlining and efficiency efforts from regulators, sponsors, industry, and patients, regulatory review times of new drugs in the US have barely moved in the past quarter-century.

One could argue that regulatory review and decisions take but a fraction of the time spent on drug discovery and clinical development. Why not address the other challenges first? There is a practical answer: Unlike the complexity of biology, or the constant speed of light, there is no law of nature that sets a lower limit to review times. More importantly, the regulatory framework forms the foundation—the de facto “laws of nature”—that drive decisions and timelines across drug development. The opportunity to explore reimagining this foundation will have catalytic effects and a disproportionately large impact on speeding up patients’ access to life-saving therapies.

Biology, Resources, and Risk

We live in the age of speed. Why not instantaneous approvals? Clearly, the trivial solution to this question, i.e., that every new drug is immediately reviewed and instantaneously approved, is terrible. We fully support a thorough review—enabled by the technology of the 21st century.

The difficulty is fundamentally framed by biology, resources, and risk. The science, targets, candidates, and biology seem infinitely complex … because they are. Resources, including money and time, are often limited. The risk to patients, as well as commercial and political risk, is almost never zero.

Even unlimited resources will not fully minimize these risks, given the limited duration and limited patient population exposed to treatments during clinical development. Conversely, extending or unlimiting clinical trial duration actually increases the risk that patients will not receive beneficial treatments because they will not have access to new medicines in time to alter the course of their disease and improve quality of life. And speaking of resources, sponsors will not be able to finance decades-long development cycles.

The COVID pandemic response is often cited as an example of how we could accelerate the drug development and regulatory review and approval process. However, the pandemic may be more like a stress test of the human physiological limit (analogous to being chased by a lion): possible in short spurts, but not sustainable or realistic for the long haul. The pandemic will likely stand as the high-water mark of what is achievable with nearly unlimited focus and incremental efficiency gains but also with an unsustainable amount of work.

Efficiency and raw speed will only partially get us there. Faster transportation can get us to a meeting in less time, but a $250,000 Ferrari still gets stuck in the same traffic as a $25,000 Ford. Some meetings can be organized quickly and successfully conducted in a virtual mode, using today’s technology-enabled tools. Is there a Zoom solution in our space?

New Industry and Regulatory Data Framework

We, the authors, invite you to explore ways to re-imagine the entire framework of regulatory decision-making based on a radical transparency of data, better collaborations, and incentive (re)alignment.

Sponsors would shift to a continuous and instantaneous flow of data to regulators, as it is being generated. Later, these data could be shared more broadly with the healthcare community using common standardized content and consistent requirements across the industry, with particular attention to protecting patient privacy and intellectual property.

This would usher in a new era of data, not documents. It is not a shortcut, but an enabler for more thorough reviews using modern technology: Electronic versions of paper submission documents are a grossly inefficient, error-prone way to communicate, analyze, and review information, if only due to the time it takes to encode or generate the documents and then decode the information the documents contain.

In this new data-not-documents framework, regulators could dynamically review data as they become available, with synchronous communications with sponsors and with fellow regulators. This could even lead to approvals before a traditional, official submission or before last patient visit, or to stopping trials early for futility or safety concerns. Imagine a continuously running data monitoring committee taking into consideration a continuous stream of efficacy and safety data.

Following these two approaches, instead of spending time re-analyzing clear winners, we can devote more time to trial results that align with Isaac Asimov’s quote: “The most exciting phrase to hear in science is not ‘Eureka!’ but ‘That’s funny…’”

Ethics and Economics

These ideas are not as radical as they sound and are supported by both ethical and economic incentives.

Ethical considerations suggest that patients expect that the data that they generate are put to good use. They have generously agreed to participate in clinical research, despite the risks of experimental treatment plans and knowing they may not receive direct personal benefits. They expect other patients to receive help in the future and that their data will be used across clinical trials to help find a cure. This could help establish consistency in reporting data to participants and in creating standards-based mechanisms for allowing patients to further share their data with advocacy groups.

Sharing data will also provide great commercial incentives recently enabled by advancements in technology. Many believe that deep learning will accelerate drug development faster than anything before. However, these techniques are notoriously data-hungry, with few, if any, companies being able to provide enough “raw materials” to train them to their full potential. Pooling data across sponsors will result in orders of magnitude larger, more granular data sets—especially including negative and phase 1 trial results—which will be key to training better models. These in turn will enable smaller and safer trials, better science, and faster discovery and development cycles. The question is how do we facilitate this sharing of information in a way that catalyzes such a “win/win”? Other domains, such as space exploration, energy production, and air travel can show us the way.

Dynamic Data Review

Dynamically reviewing data as they are being generated requires technology, people, process, policy, and a good dose of calculated risk.

Technology includes cloud submission platforms that are secure, compliant, scalable, interoperable, efficient, and transparent, as well as better analytics. AI could initiate a first review to preempt basic issues and facilitate more effective subsequent review by humans. This could be especially useful if the industry and regulators standardize data requirements.

Process and policy could focus on a “one world – one health – one dossier – (one species)” approach. Some additional practical steps could include implementing continuous reviews and shifting most of nonclinical and CMC package review to stages earlier than the New Drug Application (NDA) or Biologics License Application (BLA) submission. Most importantly, we must proactively work to further minimize the risk to patients (for example, with digital biomarkers) while absorbing calculated business risk for sponsors and political risk for regulators.

Contribute to This Conversation

We strongly believe in exploring this new framework and invite you to contribute to this discussion: Please plan to attend DIA’s Biostatistics and Clinical Innovation Industry and Regulator Forum or send your contributions to Publications@DIAglobal.org.