The pharmaceutical industry is currently navigating a significant shift toward open-source programming, with R becoming an increasingly vital tool for regulatory submissions. In a recent R Consortium webinar, Johnson & Johnson (J&J) shared a detailed case study of their successful “Hybrid R/SAS” submission strategy to the FDA.
Presented by Alicia Humphreys, Steven Haesendonckx, and Linshan Yuan, the presentation outlines how J&J leveraged the strengths of both languages to achieve FDA approval, offering a roadmap for other organizations looking to modernize their submission pipelines.
The Hybrid Strategy: Best of Both Worlds
Moving entirely to a new language can be daunting. J&J adopted a pragmatic “Hybrid R/SAS” approach to balance innovation with stability:
- SAS for Data Derivation: ADaM datasets were created using SAS to leverage existing institutional knowledge.
- R for Analysis and Visualization: All Tables, Figures, and Listings (TFLs) were generated using R.
This approach allowed the team to utilize R’s flexible visualization framework and cutting-edge statistical methods while seamlessly integrating into workflows built around legacy tools.
A Proven Success
The presentation highlighted two major milestones regarding this strategy:
- Submission 1: A pilot trial utilizing this hybrid strategy was submitted and approved by the FDA within the designed timeline.
- Submission 2: A second submission is currently ongoing with positive agency response.
Ensure Reproducibility
One of the biggest hurdles in R submissions is replicating the sponsor’s environment (often Linux-based) on the reviewer’s system (often Windows).
J&J navigated this by adapting the practices used in R consortium pilot submissions. Specifically, in Analysis Data Reviewer’s Guide (ADRG) they included a specific Appendix titled “Set-up R environment and Analysis” which guided the reviewer through:
- Installation: Using a script to install required packages from CRAN or GitHub.
- Locking the Environment: Creating an
renv.lockfile after installation to snapshot the ecosystem locally, rather than trying to transport a potentially non-portable lockfile between operating systems.
- Execution: Scripts were converted to
.txtfiles (e.g.,tsidem01-r.txt) to meet eCTD requirements while maintaining traceability.
R consortium’s Note on eCTD Requirements: While the J&J team converted scripts to .txt files for their submissions, please note that regulatory standards have recently progressed. As a result of the collaboration between the R Consortium and the FDA, the latest FDA Study Data Technical Conformance Guide (August 2025) now explicitly allows R-related file extensions for electronic submissions. This conversion step is therefore no longer necessary for future eCTD packages.
Key Takeaways for the Community
The presenters offered several recommendations for future R submissions:
- Leverage Open Source Repositories: Use CRAN or curated repositories to simplify the setup process for reviewers.
- Documentation and Communication are the Key: Update the ADRG with clear instructions on package installation and environment reproduction. J&J used the R Consortium Pilot 3 ADRG as a template. It is always a good idea to communicate with FDA staff early on your submission plan.
- Minimize Dependencies: Use standard packages where possible and minimize reliance on complex proprietary packages to reduce friction.
Watch the Presentation and Learn More
To see the full technical breakdown, including directory structures and code examples, watch the full recording of the presentation:
For those interested in the foundational work that made this possible, J&J has also shared resources on their validation and adoption processes: