FDA Grant Awards for Projects Supporting the Use of Real-World Data to Generate Real-World Evidence in Regulatory Decision Making
As part of the agency’s real-world evidence (RWE) efforts, the U.S. Food and Drug Administration announced four additional U01 grant awards in 2023 (RFA-FD-23-025) to examine the use of real-world data (RWD) to generate RWE in regulatory decision-making. These awards follow four U01 grants awarded (RFA-FD-20-030) in 2020. Through this award program, the agency seeks to encourage innovative approaches to further support the use of RWE while ensuring that scientific evidence supporting marketing approvals meet FDA’s evidentiary standards. These selected projects enhance the agency’s already diverse portfolio of RWE demonstration projects that broaden our understanding on the potential use of RWD and RWE to support the approval of new drug indications or to satisfy post-approval study requirements for approved drugs. For information on other ongoing and completed demonstration projects, visit our RWD/ RWE Demonstration Projects webpage.
2023 Grant Awards
Methods to Improve Efficiency and Robustness of Clinical Trials Using Information from Real-World Data with Hidden Bias
This project, led by Xiaofei Wang, Ph.D., at Duke University and Shu Yang, Ph.D., at North Carolina State University (NCSU), will develop innovative statistical methods to address hidden biases when integrating RWD to improve the efficiency and robustness of clinical trials. External controls (ECs) from RWD may be used to construct the comparator arm in randomized controlled trials (RCTs) but concerns regarding the comparability of RWD and RCTs have limited their use in a broader context thus far. Specifically, the project team will develop a) a novel sensitivity analysis framework for the use of ECs from RWD sources to assess the robustness of results to hidden biases, and b) efficient analytic methods that selectively borrow and adjust for data discrepancies to mitigate the impact of hidden biases. The findings will be disseminated to researchers from industry, academia, and regulatory agencies through representative applications, software, statistical analysis plan templates, a website, and workshops and tutorial sessions.
Generating Reproducible Real-World Evidence with Multi-Source Data to Capture Unstructured Clinical Endpoints for Chronic Diseases
This project, led by Tianxi Cai, Sc.D., and Florence Bourgeois, M.D, M.P.H., at the Harvard-MIT Center for Regulatory Science and Harvard Medical School, will develop novel approaches to generate RWE from data in electronic health records (EHRs) for assessing efficacy and safety of disease modifying treatments (DMTs) used in chronic diseases. Availability of RWE for DMTs has been limited by the lack of reliable computable information on disease progression measures, which are typically captured in unstructured EHR text. To address this unmet need, and focusing on rheumatoid arthritis and multiple sclerosis, this project will develop strategies for identifying scalable disease-progression endpoints and adverse events from EHR data by linking EHRs to registry data, using data from multiple healthcare systems for each condition. The corresponding work products, including open-source computing algorithms and software, will result in new capabilities for the application of real-world clinical data in regulatory decision making.
Real-World Data to Generate Real-World Evidence in Regulatory Decision-Making
The project, led by Peter O'Dwyer, MD at the ECOG-ACRIN Medical Research Foundation, is designed to foster approaches to capture, organize, and analyze RWD to produce RWE. Currently, RWD from electronic health records may be limited by problems including difficulty in extracting reliable data from pathology and radiology reports, lack of detailed data on social determinants of health, and inaccuracy in the dates of major events such as progression. This project proposes the development of a longitudinal approach to real-world data acquisition starting during Phase III trials and feeding into Phase IV. These combined data will be used to generate robust RWE for efficacy and safety of treatments in rare and less common tumors. Toward this goal, ECOG-ACRIN has formed ongoing collaboration with industry partners to facilitate identification of study patients, and build relevant study populations to interpret community-wide treatment results. By developing approaches to rigorous data collection, the project will provide a robust database for comparative analyses.
Development of Novel Methods to Enable Robust Comparison of Real-World Progression Free Survival (rwPFS) and Clinical Trial PFS in Multiple Myeloma
This project, led by Khaled Sarsour, Ph.D., Ashita Batavia, M.D., M.S., and Jennifer Hayden, M.S. at Janssen Research & Development, LLC, aims to develop methods for robustly comparing real-world progression free survival (rwPFS) to PFS determined in clinical trials among patients with multiple myeloma (MM). Previous efforts to use real-world evidence for regulatory decision-making in MM were limited by misclassification bias and surveillance bias. This project aims to develop methods to correct for misclassification bias related to differences in applying International Myeloma Working Group response criteria, and also aims to correct for surveillance bias. Additionally, this project will explore if methods for aligning rwPFS and PFS in patients historically under-represented in clinical trials are needed. The bias correction methods developed will be evaluated using control arms from Janssen MM trials. The team will work closely with academic methods experts, multiple myeloma researchers, & Flatiron Health, and the study will leverage Flatiron Health’s deidentified database derived from electronic health records. At the conclusion of the work, the team plans to make these novel methods available through peer-reviewed publications.
2020 Grant Awards
Enhancing evidence generation by linking randomized clinical trials (RCTs) to real-world data (RWD)
This project, led by Mehdi Najafzadeh, Ph.D., M.A., M.Sc., at Brigham and Women’s Hospital and Harvard Medical School, will study the benefits of making RCTs linkable to RWD. The study aims to demonstrate how linking RCTs with RWD can enhance trials by extending patients’ follow-up time beyond trial completion, capturing additional effectiveness and safety outcomes, employing methods to minimize missing data, and generalizing RCT results to real-world target populations. Linked RCTs with RWD will also improve understanding about the underlying reasons for potential discrepancies between RCTs and non-randomized studies. The study team, which includes trialists and RWD experts in collaboration with FDA, proposes to show that linkage to RWD is a low-cost, high-yield strategy that should be adopted in future RCTs.
Applying novel statistical approaches to develop a decision framework for hybrid randomized controlled trial designs which combine internal control arms with patients' data from real-world data source
This project, led by Michael Kosorok, Ph.D., and Lisa LaVange, Ph.D., at the University of North Carolina, and Herbert Pang, Ph.D., Jiawen Zhu, Ph.D., and Gracie Lieberman, M.S., at Genentech, will explore and develop recommendations for designing studies that can reliably and rigorously combine data from different sources (e.g., an RCT and RWD) and generate evidence that could be used to support regulatory decision-making. To evaluate these study designs, the team will use simulation studies, data from completed clinical trials, and RWD sources. The team will also develop and make R-packages publicly available that can be used to evaluate specific hybrid studies. Throughout the project, the team will convene meetings of experts to share progress and obtain feedback. At the conclusion of the study, the team will generate training materials and offer to conduct trainings at designated pre-conference workshops and at FDA.
Advancing standards and methodologies to generate real-world evidence from real-world data through a neonatal pilot project
This project, led by Klaus Romero, M.D., chief science officer at the Critical Path Institute (C-Path), and Jonathan Davis, M.D., professor of pediatrics at Tufts Medical Center and U.S. academic director of the International Neonatal Consortium (INC), will support the collection of neonatal intensive care unit (NICU) data from many key stakeholders worldwide. The data will then be deposited into a Real-World Data and Analytics Platform (RW-DAP).
Although many comprehensive datasets exist based on clinical care delivered in NICUs, a lack of systematic integration, data sharing, and data standards has greatly limited neonatal drug development. In this project, data will be used to define actionable reference ranges of commonly used laboratory values in neonates. In addition, a natural history model of bronchopulmonary dysplasia (a chronic lung disease common in preterm neonates) will be created. The INC and its members will partner with C-Path’s Quantitative Medicine Program and Data Collaboration Center on this project.
The electronic medical records data collected in this project will facilitate the design and conduct of clinical trials in neonates. This collaborative effort with C-Path and INC partners will help address the fact that neonates have relatively few FDA-approved therapeutic options for various medical problems.
Transforming Real-world evidence with Unstructured and Structured data to advance Tailored therapy (TRUST)
This three-year program, led by Dan Riskin, M.D., FACS, at Verantos, seeks to understand the impact of underlying data quality on RWE study results.
As RWE is used increasingly to make clinical assertions, rigorous methodological approaches are required to maintain evidentiary standards. This study compares traditional RWE approaches (using claims or health record structured data) to more advanced RWE approaches (including deep phenotyping and data linkage) on the same patient population in a real-world clinical study. Obtaining different results when using the two approaches would suggest that advanced techniques with high-data validity could be required to achieve credible RWE results. Findings can also inform future study design and definitions of fit-for-purpose data.
The study includes innovations in deep phenotyping, data linkage, and phenotype accuracy. By studying data quality and demonstrating rigorous approaches to RWE, confidence can be increased in implementing RWE within regulatory and clinical pathways.