What regulatory submission management covers
Regulatory submission management is everything between “we need to file” and “the agency has it”: planning the content, collecting documents from authors, reviewing and approving them, formatting and publishing the dossier, validating it against technical rules, sending it through the agency gateway and tracking what happens next.
In drug development the standard format is the electronic Common Technical Document (eCTD). For many medical device submissions to FDA, it is the eSTAR template. Both put more structure into submissions and more weight on getting the structure right.
What is changing
eCTD 4.0
RAPS reported on February 3, 2026, that FDA has accepted eCTD v4.0 submissions since fall 2024, plans a compatibility function to convert existing applications in fall 2026, and intends to make v4.0 mandatory by 2029. ICH’s regional implementation table shows the EU strongly recommending v4.0 for new centralized marketing authorization applications from Q1 2027 and requiring it from Q1 2028. Japan has required v4.0 for new applications since April 2026, and Health Canada plans voluntary use in 2028 and mandatory use in 2030. An FDA official told the forum: “we are accepting 4.0 submissions, we want to allow people to start testing this out and to send it in.”
eSTAR for devices
According to Emergo by UL, FDA has required eSTAR for 510(k) submissions since October 1, 2023, and for De Novo requests since October 1, 2025. eSTAR is a dynamic PDF with built-in logic that prompts for required content.
AI-assisted review at FDA
FDA launched its internal generative AI tool, Elsa, on June 2, 2025. Morgan Lewis advised sponsors that because the tool parses electronic documents, “well-structured, complete files with accurate formatting and metadata matter more,” and gaps or inconsistencies are more likely to be noticed. Faster review cycles also mean less time to answer questions.
The submission lifecycle
| Stage | Activities | Common problems |
|---|---|---|
| Plan | Define submission type, content plan, timeline, authors | Unclear ownership; late scope changes |
| Author | Write and compile documents to templates | Inconsistent templates; missing sections |
| Review and approve | Scientific, quality and regulatory review; sign-off | Review cycles that miss cross-document inconsistencies |
| Publish | Convert to submission-ready format; build eCTD structure; hyperlinks and bookmarks | Broken links; wrong granularity; lifecycle errors |
| Validate | Run technical validation against agency criteria | Late discovery of technical errors |
| Submit | Send through the agency gateway (for FDA, the Electronic Submissions Gateway) | Gateway or certificate issues |
| Track | Acknowledgments, questions, commitments, approval | Questions answered from outdated documents |
What submission software does
- Content planning: a submission plan mapped to the CTD structure, with owners and status
- Document management: versioned, approved source documents
- Publishing: conversion, bookmarks, hyperlinks, eCTD backbone and lifecycle operations
- Technical validation: checks against FDA, EMA and other agency validation criteria
- Gateway submission: transmission and acknowledgment tracking
- Archive and viewer: the full submission history as the agency sees it
What to automate
Automation in submissions has three layers.
Technical automation (mature). Publishing, bookmarking, hyperlink generation, technical validation and gateway transmission are well automated in established tools.
Content assembly (maturing). Templates that pull structured data (for example specifications, batch data or product details from a RIM or LIMS system) into documents, so the same value is not typed in several places.
Content checking (emerging). Checking that the content is complete and consistent before publishing: the same strength, specification limits and site names across modules; every claim supported; every referenced document included. This is where AI tools are being applied, and where they can save the most rework.
What should not be automated is the regulatory judgment: the content of the argument, the classification of a change, commitments to the agency and final sign-off. The April 2026 FDA warning letter to Purolea Cosmetics Lab, though about GMP documents, stated that AI-generated documents must be reviewed for accuracy and compliance. The same standard applies to submission content.
Submission quality checklist
- Content plan mapped to CTD or eSTAR structure, with owners
- Source documents approved and version-controlled
- Cross-module consistency checked (product names, strengths, specifications, sites)
- All referenced documents included and correctly linked
- Technical validation run early, not only at the end
- eCTD 4.0 readiness plan in place (Japan since April 2026; EU new centralized MAAs Q1 2028; FDA 2029; Health Canada 2030)
- Device submissions built in eSTAR where required
- AI-assisted content reviewed and signed by responsible staff
- Question response process ready for faster review cycles
Check submission content before you publish with AskDona
Publishing tools check structure. They do not check whether Module 2 summarizes what Module 3 actually says. AskDona reads submission documents together and flags inconsistencies, missing content and unsupported statements, each linked to the page it came from, before the dossier goes to publishing. Run a sample submission readiness assessment on a draft module. AskDona shows the evidence; your regulatory team decides.
FAQ
What is regulatory submission management? The planning, authoring, review, publishing, validation, transmission and tracking of submissions to health authorities.
When does eCTD 4.0 become mandatory? Japan has required it for new applications since April 2026. ICH’s regional table lists Q1 2028 for new EU centralized marketing authorization applications and 2030 for Health Canada, and FDA officials said v4.0 will be mandatory by 2029.
Is eSTAR mandatory? For FDA 510(k) submissions since October 1, 2023, and for De Novo requests since October 1, 2025.
What can be automated in submissions? Publishing, validation and transmission are highly automated. Content assembly and consistency checking are increasingly automated. Regulatory judgment and sign-off are not.
Does FDA use AI to review submissions? FDA uses an internal generative AI tool, Elsa, to support staff tasks such as summarizing adverse events and comparing labels. Reviewers remain responsible for decisions.
References
All pages opened and checked by October 14, 2026.
| Source | Type | Used for |
|---|---|---|
| RAPS, FDA and EMA officials encourage companies to pilot eCTD v4.0 (Feb 2026) | Professional society news | eCTD 4.0 timelines and quote |
| ICH, eCTD v4.0 regional implementation information | International harmonization body | eCTD 4.0 regional dates (checked October 14, 2026) |
| Emergo by UL, De Novo final guidance and eSTAR implementation date | Consultancy news | eSTAR dates |
| Morgan Lewis, FDA’s AI-assisted review | Law firm analysis | Elsa and sponsor implications |
| FDA warning letter, Purolea Cosmetics Lab (April 2, 2026) | FDA warning letter | Review of AI-generated documents |