Your Complaint Investigation Is Closed. FDA Says the Risk Is Not.

FDA's January 2026 warning letter to Cohance Lifesciences Limited rejected the firm's complaint investigation twice. The second review had the same gap as the first.

Warning letter 320-26-40, dated January 30, 2026, cites two violations at Cohance's finished-dose facility in Hyderabad: deficient investigations under 21 CFR 211.192 and inadequate equipment cleaning under 21 CFR 211.67(a). Each of those is a common GMP deficiency. Together, in a single enforcement letter, they document a more specific failure: a quality unit that treated complaint handling and cleaning validation as separate systems when the product complaint required them to work as one.

The complaint investigation was opened, closed, then re-opened after FDA reviewed it. FDA rejected the re-opened version too. The reason was precise: the investigation still did not include reserve-sample testing, did not address remediation for the raw-material issue identified in the complaint, and did not assess impact on other products manufactured in the same environment. These are not advanced requirements. They are baseline expectations for any investigation where contamination is a credible hypothesis. The quality unit answered the narrow question, what happened to this batch, and FDA said that was the wrong question.

The cleaning finding sharpens the picture. Equipment documented as clean still carried multiple active pharmaceutical ingredients above allowable residue limits. That is not a cleaning SOP that failed to execute on a given day. That is a verification loop that produced a false result on its own check. The equipment was marked clean because the cleaning records said it was clean. The residue data said otherwise. FDA's response to that kind of finding is not limited to "fix your cleaning procedure." It extends to why the verification system told the quality unit the equipment was clean when it was not.

FDA treats these as connected findings, and the logic is straightforward. A complaint involving a product manufactured on shared equipment requires the investigation to account for what else ran on that equipment. If cleaning validation data is not reviewed as part of the investigation, the investigation is incomplete by definition, not because of a technicality, but because the cleaning record is the primary evidence for or against a cross-contamination hypothesis. Cohance's complaint investigation closed without examining the data that would have answered the most important question the complaint raised.

That gap is the structural failure the warning letter documents. Two quality systems, complaint investigation and cleaning control, existed and operated in parallel without communicating during a contamination-adjacent event. Both generated documentation. Both followed their respective SOPs. Neither SOP required them to talk to each other.

The word that matters in 21 CFR 211.192 is "thorough." The regulation requires written records documenting a thorough investigation into any unexplained discrepancy or failure affecting the identity, strength, quality, or purity of a drug product. What "thorough" means in the context of a shared-equipment facility is not left ambiguous in the Cohance letter. It means reserve-sample testing, raw-material source tracing, and an explicit impact assessment on every other product that could have been exposed to the same risk profile. Following the complaint pathway is not enough. A thorough investigation follows the contamination pathway.

The cleaning violation under 21 CFR 211.67(a) reinforces this from the opposite direction. The standard requires equipment to be cleaned at appropriate intervals to prevent contamination that would alter the safety, identity, strength, quality, or purity of a drug product. Cohance had cleaning procedures. The equipment had cleaning records. The residue data contradicted both. When a cleaning control system produces documentation that conflicts with residue test results, the deeper regulatory concern is not just why the equipment was dirty. It is why the verification step signed off on equipment that was not clean.

For a quality director at a multi-product or shared-equipment site, the Cohance enforcement record is worth reading carefully. The central problem is not a single deficient investigation or a single cleaning SOP that needed revision. It is two quality mechanisms operating without an integration point between them. A complaint that should have triggered a cleaning data review did not. Equipment that should have prompted a deeper scan of complaint trends was documented as clean and closed. The quality unit had both systems. Neither system required the other to be consulted.

Quality units that manage complaint investigations and cleaning validation as separate domains, separate SOPs, separate owners, separate review cycles, build exactly this kind of structural gap into their operations. When a contamination-adjacent complaint arrives at a facility running multiple APIs on shared equipment, there needs to be a defined handoff point between the complaint investigation SOP and the cleaning control system. If that handoff depends on individual investigator judgment rather than an explicit procedural requirement, the risk of an incomplete investigation is not an exception case. It is the expected outcome.

The question a quality director should be able to answer after reading this letter: when your site receives a complaint involving a product manufactured on shared equipment, does your investigation SOP specifically require a cleaning data pull for that equipment and a review of all other products that ran on it? If the answer is "it depends on the investigator," that is the integration gap FDA documented for Cohance.