As the world’s leading drug regulatory agency, FDA has a large and growing repository of clinical trial data from NDA submissions on the safety and efficacy of drugs in many diseases and patient populations. FDA’s experience conducting meta-analyses (analyses of data aggregated from multiple clinical trials) has yielded important findings beyond those discovered in the individual studies. However, such analyses occur infrequently due to the lack of efficient and cost-effective mechanisms of aggregating and transforming multiple independent datasets. Furthermore, emerging data from sources such as electronic medical records, wearable technologies, biometric monitoring devices, and social media networks have created new opportunities for conducting novel analyses that are not currently performed in a systematic manner.
The Office of Hematology and Oncology Products (OHOP) at FDA, which is responsible for making safe and effective drugs for cancer and hematologic conditions available to the public, seeks to develop a big data environment for clinical trials and other emerging data sources for novel explorations and analyses. This initiative involves the following core components:
- Transformation of FDA’s existing clinical trial datasets into a common standard;
- Development of a big data environment for storage and mining of transformed datasets; and
- Incorporation of diverse pipelines of data (e.g. electronic medical records, wearable technologies, biometric monitoring devices, social media networks) into the big data environment
The EIR will serve as the catalyst for organizing efforts in support of the core components of the initiative. The EIR will lead the development of new methods to transform datasets to support current internal projects, such as the newly launched effort to develop predictive tools for serious adverse events of small molecule kinase inhibitors. The results of this work have the potential to significantly impact the way adverse events associated with new targeted cancer therapies are estimated, and will serve as a blueprint for future predictive modeling projects. The EIR will also initiate the development of a scalable big data environment through collaboration with FDA’s internal teams and building strategic partnerships (e.g., academic and public-private partnerships) in support of the initiative. The EIR will help create a sustainable institutional framework for continued growth of the initiative by working with FDA leadership to develop policies and procedures governing the use and maintenance of the big data environment.