Statistical Innovation Roundtable at RISW 2024

Justine Rochon (Boehringer Ingelheim)

Among the many intriguing topics offered for roundtable discussions at this year's ASA Biopharmaceutical Section Regulatory Industry Workshop 2024, I chose to attend the unique session on "Statistical Innovation in drug and device development and regulation – how can we maximize its impact?" I thoroughly enjoyed it. Under the leadership of Marc Vandemeulebroecke (UCB), I joined colleagues George Chu (Edwards Lifesciences), Jiangeng (Victor) Huang (AbbVie), Natalia Muehlemann (Cytel), and Sofia Villar (MRC Biostatistics Unit) for a highly insightful dialogue.

My understanding of the difference between invention and innovation truly deepened during my visit to the MIT in January 2023. There, I learned about the 24 steps of Disciplined Entrepreneurship from the phenomenal teacher, Bill Aulet, and his team. I learned that invention alone is not sufficient. Innovation is a more comprehensive concept: Innovation = Invention x Commercialization. Without commercialization, an invention does not become an innovation. A year later, Kaspar Rufibach and his colleagues (2024) published an insightful paper on this topic. In this paper, the authors applied Bill Aulet’s formula to statistical innovation within a pharmaceutical company. They emphasized that innovation, defined as the successful large-scale implementation of a new invention, is crucial for the continued success of drug development enterprises. They discussed both components of innovation for statistical methods – invention and large-scale implementation, typically referred to as commercialization using three examples of successful commercialization of innovations at Roche: implementation of open-source software, namely rpact, for all clinical trial design activities at Roche, implementation of estimands, and adoption of reference-based mean imputation (rbmi) for longitudinal endpoints. The last example is noteworthy because two large Phase 3 Alzheimer disease trials were re-designed using this new method, which had been developed by Roche statisticians in collaboration with academic partners and implemented in a high-quality R package in-house. The redesign was even more remarkable as rbmi typically delivers smaller (but less biased) effect estimates compared to the traditional methods, but the statisticians succeeded to convince internal and external stakeholders (especially clinicians) that redesigning the trials was the right thing to do.

Highly inspired by this paper and our own professional experiences, we focused our roundtable discussion on the tangible impact of statistical innovation in the pharmaceutical and MedTech industry. We unanimously agreed that such innovation should aim to minimize patients’ exposure to futile interventions and streamline the processes of drug and device development, approval, and health technology assessment. Our ultimate goal is to expedite patient access to advancements in medicine and technology.

The discussion on the role of academia in innovation led us to examine the ideas recently presented by Heinze et al. (2024). We agreed that this role should extend beyond merely generating ideas and should involve the development of robust methodologies, contribute to the understanding of method limitations, and foster capacity building through training and software development.

Our roundtable yielded several key insights. We underscored the necessity of fully comprehending the problem at hand and the stakeholders’ needs before proposing disruptive innovations. We also highlighted the crucial role of a product owner for ensuring clarity and accountability, and the significance of continuous education at all levels in fostering innovation, including entrepreneurship. These elements are vital in the journey from invention to innovation and in maintaining its momentum in the long-term. Furthermore, we delved into the optimal environment for statistical innovation and the importance of creating a safe place for regulatory-industry-academia collaboration.

The innovation roundtable at RISW2024 was a great opportunity to share experiences and learn from each other. I hope to maintain these connections and anticipate witnessing the impactful outcomes of our future interactions.

References

Aulet, B. (2024). Disciplined Entrepreneurship: 24 Steps to a Successful Startup, Expanded & Updated. John Wiley & Sons.

Heinze, G., Boulesteix, A. L., Kammer, M., Morris, T. P., White, I. R., & Simulation Panel of the STRATOS Initiative. (2024). Phases of methodological research in biostatistics—building the evidence base for new methods. Biometrical Journal, 66(1), 2200222.

Rufibach, K., Wolbers, M., Devenport, J., Yung, G., Harbron, C., Bedding, A., Huang, Z., Lin, R., Pang, H., Sabanés Bové, D., & Wang, J. (2024). Implementation of statistical innovation in a pharmaceutical company. Statistics in Biopharmaceutical Research, 1–12. https://doi.org/10.1080/19466315.2024.2327291