TL;DR
- Company: Genzyme Ireland Limited, a Sanofi subsidiary, Unit 701 IDA Industrial Park, Old Kilmeaden Road, Co. Waterford, Ireland.
- Date: Warning Letter CBER 26-728681 issued June 22, 2026 (MARCS-CMS 728681).
- Classification: CGMP / Deviations / Biologics License Application (BLA).
- Key citations: Quality control unit failure to enforce CGMP (211.22), incomplete laboratory records covering undocumented particulate excursions and repeated test failures (211.194(a)), and unexplained discrepancies closed without root-cause investigation (211.192).
- Consequence: FDA rejected Genzyme's written response as addressing individual findings rather than the underlying quality-unit failure, and demanded a facility-wide data-integrity assessment before the file closes.
| Company | Location | Letter # | Issued | Issuing Office | Inspection Dates | Classification |
|---|---|---|---|---|---|---|
| Genzyme Ireland Limited | Unit 701 IDA Industrial Park, Old Kilmeaden Road, Co. Waterford, Ireland | CBER 26-728681 (MARCS-CMS 728681) | June 22, 2026 | CBER | January 12–20, 2026 | CGMP / Deviations / BLA |
A batch record shouldn't need eleven attempts to pass. At Genzyme Ireland's Waterford facility, FDA investigators found (b)(4) testing repeated up to eleven times on the same material, with no documentation explaining why any of the first ten runs failed or what changed by the eleventh. That finding sits inside the FDA Warning Letter CBER 26-728681, June 22, 2026, addressed to site head Cian O'Brien and issued not by CDER but by the Center for Biologics Evaluation and Research, the branch that oversees licensed biologics rather than conventional drugs.
That distinction matters here. Genzyme Ireland manufactures two BLA-licensed products at the Waterford site: Thymoglobulin (Anti-thymocyte Globulin [Rabbit]) and Altuviiio (antihemophilic factor [recombinant]). FDA inspected the facility January 12–20, 2026. The agency's three citations, taken together, describe a quality system that could observe its own failures but consistently declined to explain them.
The three citations, in order
21 CFR 211.22 — the quality control unit failed to ensure CGMP compliance. FDA frames this as the umbrella finding: the site's quality unit had the authority on paper to enforce manufacturing controls but did not exercise it in practice. The two citations below are the evidence FDA cites to support that conclusion, not separate, unrelated problems.
21 CFR 211.194(a) — laboratory records did not include complete data from all required tests. The letter is specific about what "incomplete" meant in practice. Multiple non-viable particulate excursions and repeated (b)(4) test failures, visible in the instrument histories themselves, were never written up. Some of that (b)(4) testing was repeated as many as eleven times without any record explaining the repetition. Separately, the number of active air-monitoring samples the firm documented did not match what the instrument histories showed had actually been run, a gap FDA can only read one way: the paper record and the equipment record disagree, and nobody reconciled them. Cleanroom logbooks were not always consistent with recorded entry and exit times either. Perhaps the most telling detail: quality control staff kept informal "Review Checklists" to track their own record reviews, then threw them away once finished, so the checklists never became part of the permanent record and no single reviewer could be tied to any given check.
21 CFR 211.192 — failure to thoroughly investigate unexplained discrepancies. FDA states plainly that the firm cancelled numerous deviations without investigating root cause or assessing product impact. A cancelled deviation is not a resolved deviation. It is, per this letter, a discrepancy the firm chose not to explain.
Why FDA rejected the response
Genzyme had 15 working days to respond, the standard window FDA sets in nearly all warning letters. The company did respond, and FDA reviewed that response between February and April 2026. The agency's verdict, in its own words as summarized in the letter, was that the response addressed individual observations rather than the systemic quality-unit failure FDA had actually cited.
That's a distinction with real consequences for how a firm should read any 211.22 citation. Fixing the eleven-times-retested (b)(4) test is not the same as explaining why the quality unit allowed eleven undocumented retests to happen in the first place, or why nobody flagged the pattern earlier. FDA is not asking Genzyme to patch the specific instances documented in the letter. It's asking for a comprehensive data-integrity assessment across all operations at the site, an expanded retrospective review of deviations reaching further back than the window Genzyme initially selected, and revised documentation practices for how investigations get recorded going forward. Point fixes don't satisfy a systemic finding. That gap, response-to-symptom versus response-to-system, is the single most useful thing a compliance team can take from this letter.
The letter closes with FDA's standard warning that failure to promptly and adequately address the violations may result in legal action, including seizure and injunction, without further notice.
What this means for other BLA manufacturers
Three things in this letter apply well beyond Genzyme's Waterford site. First, CBER's citation logic mirrors CDER's: a quality-unit-authority finding (211.22) sits at the top, and specific documentation or investigation failures underneath it function as supporting evidence, not standalone problems. A firm that treats each supporting citation as its own fix-it ticket, rather than asking what allowed the quality unit's authority to erode in the first place, is likely to get the same "addressed the individual observations, not the systemic failure" response Genzyme did.
Second, the discarded "Review Checklist" detail is worth flagging internally at any biologics site. Any informal document your quality unit uses to track its own work, and then throws away, is a data-integrity gap waiting to be found. If it exists to prove a review happened, FDA expects it to survive as long as the batch record it supports.
Third, cancelling a deviation is not the same as investigating one. FDA's language here, "cancelled numerous deviations without investigating the root cause or assessing product impact," describes a pattern regulators have flagged at other biologics and drug facilities in 2026. A deviation-cancellation rate that climbs without a matching rise in documented root-cause analysis is a metric internal audit teams can and should track before an inspector does.
FAQ
What products does the Genzyme Ireland warning letter cover?
Two BLA-licensed biologics manufactured at the Waterford, Ireland facility: Thymoglobulin (Anti-thymocyte Globulin [Rabbit]) and Altuviiio (antihemophilic factor [recombinant]).
Why did CBER issue this letter instead of CDER?
Thymoglobulin and Altuviiio are both licensed biologics under a Biologics License Application, which places them under the Center for Biologics Evaluation and Research rather than the Center for Drug Evaluation and Research, CBER's counterpart for conventional drugs.
Did FDA accept Genzyme's response to the inspection findings?
No. FDA reviewed the firm's response between February and April 2026 and found it addressed individual observations rather than the systemic quality control unit failure the letter cites under 21 CFR 211.22.
Source: FDA Warning Letter CBER 26-728681, June 22, 2026. Byline: Andy Gaber. Published 2026-07-06.