Lessons Learned from the Recent Cell and Gene Therapy Product Development and Approval

Oregon Convention Center 

Room: CC-E145 

Novel cell-based therapies that use a person's cells are unlocking new potential in the fight against some of the most difficult-to-treat diseases such as cancers, while gene therapies modify a person's genes to treat or cure disease. Cell and gene therapy (CGT), which can consist of therapies made up of live cells or functional gene segments that are injected, implanted or grafted to treat disease, is an active research field to treat diseases across multiple therapeutic areas. Since "treatment" can be derived or modified from a donor's or patient's own cells/genes, there is inherent variability when evaluating the treatment effect. Variability may be attributed to many factors including differences between patients, starting material in cell expansion, and manufacturing process. In certain diseases, such as cancers, additional interventions are often needed after patient enrollment, leading to challenges in interpreting treatment effect. These unique features create opportunities for statisticians and data scientists to provide best available scientific evidence to support data-driven strategic and policy decisions.
Given the promising patient outcomes, the development of CGT has expanded exponentially in recent years. The discussion panel, comprised of members from the Biopharmaceutical Section CGT Scientific Working Group, aims to share learnings and strategies on CGT development, and to promote the adoption of sound statistical methods to avoid misinformation. The discussion will focus on the followings:

1. Dose Escalation & Dose Finding
• Traditional maximum tolerated dose may not be suitable for these therapies; approaches to address toxicity-efficacy trade-off during dose escalation can help optimize the recommended Ph 2 dose.
• Methods to leverage data extrapolation from prior products of identical construct with similar features to help accelerate dose escalation.
• Practical solutions to analyze data from subjects who receive non-conforming products.

2. Estimand
• Strategies and experiences in directing stakeholders to reach a clear definition of scientific questions and treatment of interest.
• Methods to handle intercurrent events occurred after surgery or leukapheresis for autologous CAR T cell therapy, such as the use of lymphodepleting therapies and/or bridging therapies.
• Methods to handle intercurrent events due to manufacturing failure or starting new therapies without progression.

3. Real World Data & Real-World Evidence
• Use cases of leveraging data sources from clinical sites, registries and research databases that are derived and set up to generate an external comparison arm for ancillary analysis of efficacy endpoints.
• Approaches for data integration and matching.

4. Big Data for Predictive Biomarkers
• Experience in building data warehouses for biomarker identification, including manufacturing data, pre- and post-treatment clinical and biomarker data that may be informative for predicting CGT product safety and efficacy.
• Examples of using artificial intelligence and machine learning approaches.

5. Master Protocol & Novel Trial Design
• Utilization of master protocol designs to conduct long-term follow-up studies (typically 5-15 years post treatment).
• Studying multiple versions of CGT product in early phase using umbrella trial designs and Bayesian information sharing/borrowing.
• Basket trial designs implemented in targets with lower prevalence cancer subtypes.

Yes

Biopharmaceutical Section

Ad Hoc Good Clinical Practices Committee

Health Policy Statistics Section