A December 2024 warning letter to Brands International Corporation cites the firm under 21 CFR 211.67(b) for failing to perform cleaning validation studies on non-dedicated equipment — and asks for a cleaning programme built around worst-case products, worst-case swab sites and maximum hold times before cleaning.
What happened
The Warning Letter identifies failures at Brands International Corporation, FEI 3007389902, at 594 Newpark Blvd, Newmarket, Ontario, Canada. FDA inspected the site from 17 to 21 June 2024; the letter itself (Warning Letter 320-25-25, reference 689983) is dated 17 December 2024 and was posted to the FDA website on 24 December 2024.
Under Violation 5 — addressed to the quality control unit (21 CFR 211.22) — the letter states that the firm “failed to perform cleaning validation studies for non-dedicated equipment used to manufacture your drug products (21 CFR 211.67(b)).” In the same section FDA records two further gaps: no adequate procedures to validate processes and qualify equipment (21 CFR 211.100(a)), and no ongoing testing programme to support the labelled expiry of distributed products (21 CFR 211.137).
The letter also documents that the firm limited and/or delayed the inspection, with records withheld and hostile conduct toward investigators; FDA therefore declared the drugs adulterated under section 501(j) of the FD&C Act. Products from the site were placed on Import Alert 66-40 on 28 October 2024. Two voluntary recalls are referenced: a 10 June 2024 recall of a body lotion due to mould contamination, and an expanded recall on 13 November 2024.
Why it matters
FDA did not ask simply for a cleaning protocol. It asked for a lifecycle programme built on explicit worst-case justification. The letter requires improvements to the cleaning validation programme “with special emphasis on incorporating conditions identified as worst case”, naming: drugs with higher toxicities; higher potencies; lower solubility in their cleaning solvents; drugs difficult to clean; swabbing locations for areas most difficult to clean; and maximum hold times before cleaning.
That is the standard non-dedicated equipment invites. Once a train, tank or filling line carries multiple products, a single cleaning procedure is only defensible if the product it was validated on is genuinely the hardest to remove.
How such a gap can realistically develop
Rarely as a decision — usually as drift. Shared equipment is used product by product while nobody owns the matrix or bracketing exercise. Worst-case selection is skipped because potency, toxicity and solubility data sit in separate files and are never consolidated. Hold times before cleaning are never established, so the procedure is validated against an interval no real campaign observes. Swab sites are chosen where they are easy to reach rather than where residue accumulates. Cleaning verification is substituted for validation, and the quality unit lacks the authority to stop release.
Here the gap did not stand alone: with process validation and equipment qualification also absent, there was no baseline state of control for cleaning to defend. FDA's requested response reflects that — the cleaning programme is to be presented inside a lifecycle strategy, including process performance qualification and ongoing monitoring of intra-batch and inter-batch variation.
Questions for our programme
- Which product is our documented worst case, and what data — toxicity, potency, solubility, cleanability — supports that selection?
- Are maximum hold times before cleaning established, and were they challenged in the validation runs, or inherited from a procedure?
- Are swab locations chosen from cleanability data and equipment geometry, or from convenience?
- If equipment is shared, does our matrix or bracketing cover every product that touches it, including low-volume campaigns?
- Is the cleaning validation anchored to a validated, controlled manufacturing process — or is it defending an unqualified state?
Wider context
This is a recurring theme rather than an isolated finding. Cleaning validation deficiencies appear repeatedly in FDA Form 483 observations and warning letters, and the Agency's framing has shifted from a discrete study to a lifecycle control element, consistent with the risk-based expectations of ICH Q9 and the lifecycle model of ICH Q10. The same site was cited in 2023 for laboratory controls, including failure to test for Burkholderia cepacia — suggesting systemic weakness rather than a single procedural lapse.
Sources
- FDA Warning Letter, Brands International Corporation, 689983, 17 December 2024 (primary; direct fetch blocked, verified via mirror).
- einpresswire.com — full-text mirror of Warning Letter 320-25-25.
- pharmabiz.com — USFDA warning letter posting list, December 2024.
- keypedia.com — FDA inspection and 483 records for the site.
Validation takeaway
Cleaning validation on non-dedicated equipment is only defensible when the product and locations used in the study are demonstrably the worst case for that train. FDA now asks for that rationale explicitly — higher toxicity, higher potency, lower solubility, difficult-to-clean products, worst-case swab sites and maximum hold times before cleaning — and expects the cleaning programme to sit inside a lifecycle control strategy, not beside it. A cleaning gap found without process validation and equipment qualification in place reads as an absent state of control, and the resulting remediation burden is far larger than a single protocol rewrite.
Questions worth asking in our own programme
- Which product is our documented worst case, and what toxicity, potency, solubility and cleanability data supports that selection?
- Are maximum hold times before cleaning established and challenged experimentally, rather than inherited from an existing procedure?
- Are swab locations selected on the basis of cleanability data and equipment geometry, or on the basis of accessibility?
- If equipment is shared, does our matrix or bracketing exercise cover every product that contacts it, including low-volume campaigns?
- Is our cleaning validation anchored to a validated and controlled manufacturing process, or is it defending an unqualified state?
