Glimpsing the Future of Regulatory Science
Use "averted vision" to see possibilities in the dim light of the future.
Astronomers are familiar with employing "averted vision" as the best way to see faint objects in the darkness. By looking just off to the side of the target, the image falls on the part of the retina with rods that are better at perceiving black-and-white images in dim light.
I have been working with regulators and academics in sub-Saharan Africa to learn about the regulatory review process in low- and middle-income countries (LMIC) and to develop ideas for enhancing the process. This work has given me an appreciation for the great strides that African scientists and regulators have made in strengthening their capacity and expertise.
By looking just to the side of the target of strengthening scientific and regulatory capacity in LMIC, you can bring into view a possible future with implications for the global regulatory community.
Regulatory Reliance
Regulatory agencies in Africa have been encouraged to use regulatory reliance, a process that enables an authority in one country to leverage the regulatory work performed by a competent and trusted regulatory authority or institution in another country to support its decision-making. In practice, regulatory reliance entails receiving completed or nearly completed reviews by regulatory authorities, such as the US FDA or the European Medicines Agency, with little opportunity for LMIC regulators to actively participate in the process.
As scientific and regulatory capacities are continually strengthened in sub-Saharan Africa, it is possible to see how reliance could evolve into the sharing of novel and currently untapped insights from LMIC. This would lead to exciting possibilities for the global regulatory community.
Leapfrogging to the future
Poor-resourced countries have an imperative to utilize available resources efficiently and effectively. In response, African regulators have formed cooperative transnational networks that permit each nation to leverage its strengths and compensate for its deficiencies by collaborating to accelerate access to quality-assured, effective, and safe medical products.
Although still under construction, the hard work of building and sustaining these networks gives insight into what might be possible if these agencies can harness their work to leapfrog their well-resourced cousins.
Leapfrogging is the idea that LMIC can move forward rapidly by adopting modern systems without going through intermediary steps. Skipping landline telephony and leapfrogging to mobile is the best example of this in action.
What would it look like to leapfrog into a future that takes full advantage of the imperatives to function efficiently and effectively, along with ongoing capacity building and emerging technology to reshape the global regulatory process?
Framing the Future of Regulatory Science
My work with regulators and academics from sub-Saharan Africa exposed me to several developments that could put these agencies in a position to shape the future of regulatory science.
First, the regulatory review of a new drug requires a close collaboration that underpins the trust and transparency necessary for success.
We think of technology as the hard part required for crafting the future. But the hard part is getting the people in institutions, government, and bureaucracies to collaborate. Reviewing a new drug dossier by a regulatory network requires a close collaboration that, over time, spawns a deep level of trust that is essential for a team to operate successfully.
This collaboration enables the network to draw on the expertise of clinicians and regulators working in very different contexts. Although randomized clinical trials of new drugs are the mainstay of regulatory review, the context of drug use reflects the differing characteristics that affect a drug's safety and efficacy profile after marketing. Insights into the context of differing target populations are critical to successfully adapting product labeling with appropriate instructions for dosing adjustments, safety monitoring, alerts for drug-drug interactions, and so forth.
Second, a growing body of knowledge from global health research programs is being developed in Africa.
There has been a marked expansion in the scope of philanthropic-funded global health research that can provide novel insights into the context of patient populations and their attendant healthcare issues in LMIC. For example, the One Health approach has been proposed to tackle complex global public health threats at the interface of humans, animals, and the environment. The initiative includes establishing an academic community committed to providing robust scientific evidence, addressing technical development gaps, and synthesizing empirical experiences and practice to tailor strategies for specific social-ecological settings.
In addition, international donors have funded the DELTAS African Initiative to provide support for research consortia controlled by the African Academy of Sciences Alliance for Accelerating Excellence in Science in Africa (AESA) and supported by the African Union Development Agency (AUDA-NEPAD). Evidence suggests that the DELTAS Africa initiative significantly impacts African research production, the numbers and quality of African trainees, and the strengthening of African institutions, particularly concerning knowledge translation and community and public engagement.
Third, confidentiality necessitates a communication infrastructure that underpins the exchange of information essential to the regulatory review process.
A critical component of this infrastructure is the capability to capture and share the key ideas, hypotheses, intuition, and knowledge generated by the various disciplines involved in the review. The information can be represented in a searchable graph database that distills and interrelates the conceptual network of all factors and relationships across the different biological, social, geographical, and economic contexts.
As the database generated by the review process grows, it will contain extensive information at multiple levels of granularity regarding drug classes and specific compounds that the regulatory network has scrutinized. This information can include drug pharmacology, disease pathophysiology, known and suspected risk factors, propensity for drug interactions, demographic characteristics for target populations, and curated feedback regarding the questions raised during the review.
Fourth, there is the opportunity to design the review workflow to include artificial intelligence-enhanced support.
This database, combined with data and analyses from global health research and pharmaceutical research development projects across contexts, drugs, and drug categories, would be continually updated as new factors and insights emerge and shared with the global regulatory community as a public good.
Consultation with the database for potentially salient insights could become a standard component of an AI-enhanced review process at all key milestones to increase the confidence of regulatory agencies and all stakeholders in the risk-to-benefit calculations that support the decision-making process.
The process of querying the database with the support of Large Language Models (LLMs) provides the opportunity to curate the feedback to the LLM on the salience, accuracy, and completeness of its answers to prompts from various disciplines. This would improve the quality of the answers in specific problem domains, such as regulatory review.
Fifth, implications for systems-level changes to enhance regulatory science
The activities described above suggest intriguing possibilities with implementing new tools and processes for utilizing the information generated during the R&D and review life cycle.
Recognition of the value of the questions and perspectives of regulators across the global community would prompt input from these perspectives early in the design of the clinical development plan rather than late in the process under the current reliance process. In addition, strengthening the ethics committees in LMIC provides an early view of the rationale and design of clinical trials and another opportunity to become prospectively engaged in the R&D process.
The ongoing review and curation of information in the AI-enhanced database would permit ongoing interrogations by regulatory agencies and industry scientists. Continued accretion of high-quality data from multiple and varied sources over multiple R&D and review life cycles would provide an unprecedented resource for enhancing regulatory science.
Conclusions
Implementing regulatory reliance has fostered innovative and effective forms of collaboration that best use the available resources and expertise while avoiding duplication. This allows agencies in sub-Saharan Africa to concentrate their regulatory efforts and resources where they are most needed while forging new approaches to the challenge of regulatory review.
The enhancement of reliance with sharing across the regulatory community provides a framing for the future of a global regulatory review community to work towards an innovative process that does more than make minor adjustments to the existing process. It requires putting our imaginations to work in conjuring up a future of innovative collaboration, communication, and novel perspectives that go well beyond what is possible in the existing process.
There are reasons to fear the future - it is not difficult to find endless streams of frightening news to induce a case of chronophobia in the most optimistic personality. At the same time, there is optimism in imagining an attractive future when accelerated access to new medicines, new technology, and new ways of living can address critical public health needs in the global community.