How TCG v6.2 Shifts SEND from Compliance Deliverable to Review-Ready Package
- eniivancesolutions
- May 5
- 3 min read
Introduction
FDA’s March 2026 release of the Study Data Technical Conformance Guide (TCG v6.2) has been circulating for a few weeks now. For most teams, it reads like a routine update some language refinements, a few new sections, nothing that demands immediate attention.
With the addition of Appendix I, that reading is worth revisiting. FDA’s expectations around SEND are shifting from technical compliance toward review readiness. Those are not the same thing, and the gap between them is where most submission problems live.
Here is what Appendix I introduces, what it actually means, and what teams should be doing about it.
The change that matters most: Appendix I
This is the most significant addition in TCG v6.2, and the one with the clearest implications for how SEND submissions are prepared.
Appendix I is a new section titled Best Practice — SEND Data Review Prior to Submission to CDER and CBER. It formalizes something that many teams have historically treated as optional: reviewing your SEND package the way an FDA reviewer would, before it leaves your hands.
The motivating statement in the guidance is worth noting when SEND packages arrive with certain errors, the document states plainly that SEND may be unusable. Not delayed. Not flagged for correction. Unusable. That is the stakes FDA is articulating.
What makes a SEND package unusable in FDA’s view? The guidance groups the most common issues into three categories:
Submission issues: Wrong file locations, incorrect file naming, missing files, or multiple studies in the same submission folder.
Basic SENDIG conformance issues: Record uniqueness violations (two findings' records sharing the same USUBJID and sequence value), empty required variables like DSDECOD, values exceeding the eight-character limit such as LBTESTCD, incorrect data types where a numeric variable has been defined as character (for example, PPSTRESN submitted as a character field). Invalid references between files such as POOLID values in a PP domain that don’t exist in POOLDEF domain.
Study design representation issues: Terminal and recovery animals placed in the same set, mismatched code/label values where a single SET value maps to more than one SETCD, and multiple records for the same animal, endpoint, day, and timepoint with conflicting results.
These are not edge cases. They are the categories of errors FDA is actually encountering at submission receipt. The FDA expects that the SEND study data should be fully traceable to raw data and the study report, compliant with the applicable SENDIG version, and accurately reflect the study design to support reproducible analyses. The nSDRG should clearly document data creation, verification processes, and any deviations or anomalies with proper justification. Overall, the dataset must be complete, consistent, and transparent to enable FDA review and validation.
What this means in practice: Pre-submission review can no longer be an informal final check. It needs to be a structured, documented step, and it needs to mirror the review FDA itself will conduct.
Appendix I outlines two areas of verification every team should build into their pre-submission process:
Verification of the Submission:
Ensure all SEND datasets and supporting files (nSDRG.pdf, define.xml) are correctly placed within the m4\datasets\[study]\tabulations\send directory, with proper naming conventions and no misplaced or empty (0 kb) files.
Verify consistency between dataset files and define.xml, including matching file listings and uniform STUDYID values across all datasets.
Perform direct folder-level checks (not just via eCTD tools) to confirm accurate file structure, completeness, and compliance with submission requirements.
Verification of the Contents of SEND Dataset Files
Ensure SEND datasets comply with FDA Business Rules and accurately represent study design, including correct trial design datasets and inclusion of all reportable endpoints (with any exclusions documented in the nSDRG).
Verify that SEND data supports proper analysis by checking groupings, counts (N), and summaries (means/incidence) against the study report, ensuring consistency and explainable differences (e.g., NOT DONE, exclusions, early deaths).
Confirm both summarized and individual animal-level data in SEND align with the study report for representative endpoints, days, and timepoints, with any discrepancies or special grouping logic clearly described in the nSDRG.
A final thought
TCG v6.2 makes one thing clear: SEND is no longer just about passing compliance checks; it must be review-ready. FDA now expects datasets to be immediately usable for analysis, fully traceable, and accurately reflective of study design, with no ambiguity left for reviewers to resolve.
The introduction of Appendix I reinforces that pre-submission review is not optional or superficial anymore. Sponsors must proactively validate both the submission structure and data content through a structured, FDA-like review process ensuring consistency with the study report, adherence to business rules, and clear documentation in the nSDRG.
In practical terms, success with SEND now depends on a shift in mindset: from “Is this compliant?” to “Can an FDA reviewer confidently use this data without rework?”
Comments