How far is 2041? Update on data submission to health authorities

Back in the summer 2023, I was invited to present “Standards and Open-Source Hand-in-Hand: Leveraging Automation to Expedite Drug Market Request Review Process” at PharmaSUG-China. I was trying to imagine the future of data submission, travelling to 2041 and envisioning how AI can support and expedite the regulatory drug submission process, and how AI could enhance the preparation and review of data submission packages. I then brought the discussion back to the present, sharing some reflections on the journey ahead — a journey that will inevitably require better use of standards, open-source adoption and solutions, and collaborative industry initiatives.

About 18 months later, the topic of AI became predominant in our industry. This is clearly reflected by the growing number of AI-related presentations at conferences, including the recent PHUSE US Connect Conference held last March in Orlando, and the upcoming CDISC-EU Interchange this May, just a few steps from our offices here in Geneva.

Here, I would like to provide a brief overview of the latest updates on data submission requirements, as well as industry initiatives aimed at improving how we create clinical data packages for submission to health authorities in support of market drug approval.

FDA data submission requirements update

Regulatory data submission requirements, more specifically those from the US FDA, have been refined through various updates of their guidance. Since my January 2024 summary of the latest changes, the following additional requirements have been added:

• Submit a dataset, LC, copy of LB with US conventional unit as standard unit (March 2024)
• Viral load results should be placed in the MB domain, confirming there is still misuse of specific laboratory related data domain e.g., LB, IS and MB (October 2024)
• The requirement for US conventional unit was recently extended to ADaM, with ADLC dataset (March 2025)

See here the latest March 2025 version of the FDA Study Data Technical Conformance Guide.

It’s also worthwhile to mention the FDA’s “Protocol Deviations for Clinical Investigations of Drugs, Biological Products, and Devices,” which provides various recommendations around the management of protocol deviations. This includes some specific recommendations for SDTM mapping, such as including a variable in the DV domain that provides the sponsor’s determination of whether the protocol deviation was important.

EMA data submission requirements

While the EMA has not made data submission mandatory — nor specified a required data format — the European Medicines Agency launched the “Raw Data” pilot proof-of-concept project about two years ago. In this initiative, selected applicants were invited to submit structured clinical trial data as part of their initial applications and post-authorization procedures. Clinical trial data in this context refers to individual patient-level data, including:

• Clinical laboratory results
• Images
• Medical records

The aim of the pilot is to assess whether the use of structured clinical trial data can help speed up and improve the drug assessment process.

An initial outcome of the project was published in a report released last October. It summarizes lessons learned from five data submissions received between September 2022 and December 2023, out of the ten originally planned. Among the key learnings and outcomes, CDISC standards, namely SDTM and ADaM with define.xml and a data reviewer’s guide, were confirmed as suitable formats for data review. The software tools being explored included SAS and R for statistical analysis, and SAS JMP Clinical for visualization. While SAS XPT files were required, other transport formats such as XML or JSON were also accepted, upon mutual agreement between EMA and the applicant.

Although these standards and formats are not yet mandatory, additional guidance has been provided in a Q&A document (e.g., regarding maximum data package size). Since then, the EMA has decided to extend the project’s duration. Final recommendations are expected in 2025 — potentially with some early updates to be shared at the upcoming CDISC EU Interchange in May.

Industry initiatives update

Since my speech at PharmaSUG-China, the industry initiatives I discussed there have progressed quite rapidly:

•   The R Pilot Submission Experience: All four planned pilots have been completed, and in February a fifth pilot was announced. This time, the goal is to establish the new dataset-JSON format as a CDISC standard for clinical data submissions (see here a report from successful pilot submitting data with the new format to the FDA.

•   R Packages for SDTM and ADaM: Both the SDTM (oak) and ADaM (admiral) R packages are now widely used in our industry for submission projects.

•   The Analysis Results Standard (ARS): The first version of the ARS was released in April 2024, along with a new initiative, the eTFL Portal, which shares examples and templates for the most common TFLs.

•   The CORE Project: The project continues its mission to develop Open Conformance Rules, alongside a growing number of Open Source initiatives.

Interested in learning more?

Get your copy of Angelo Tinazzi’s latest ebook, “The Good Data Submission Doctor on Data Submission and Data Integration to the FDA”:

Share Via