Bentley Widened the pH Limit Instead of Finding the Cause

A pH result at the 36-month stability timepoint came back out of specification at Bentley Laboratories, an over-the-counter drug manufacturer in Edison, New Jersey. According to FDA's September 17, 2026 warning letter, the firm's corrective action was to expand the pH specification range. FDA says there was no scientific justification for it. The product did not change. The acceptance criterion did.

That one finding sits inside a letter, reference 320-26-126, that followed an inspection on March 16 to 27, 2026. FDA reviewed Bentley's April 21 response to the Form 483, called it inadequate, and cited three CGMP areas: 21 CFR 211.192 for investigations, 211.84(d) for component testing, and 211.100(a) for process and cleaning validation. It also found two products, labeled for acne treatment, to be unapproved new drugs. Read together, the three CGMP citations describe one condition. In each, a failing result or a missing control was explained or worked around instead of stopped.

Start with the investigations. FDA identified two OOS results. One was an assay failure for a lot at the 24-month stability timepoint, where the firm did not identify a scientific root cause and did not evaluate the effect on product already on the market. The other was the pH failure at 36 months, where the firm also failed to investigate adequately and then changed the limit. FDA's summary of the defect is plain: the investigations did not identify root cause, did not put an appropriate CAPA in place, and did not expand to other batches or products.

Widening a specification is a legitimate act in some circumstances. Limits get revised on data, through change control, with a justification a reviewer can read. What FDA describes is different in sequence. The failing result came first, and the wider range followed as the CAPA. An inspector reading that file sees a limit that moved after a lot failed it, and the burden is on the firm to show the movement had a scientific basis independent of the failure. FDA says Bentley did not show one.

The assay failure carries its own lesson, and it gets less attention. FDA faulted the firm for not evaluating the impact on products on the market. A stability failure at 24 months is information about lots that may have been in distribution for most of that time. An investigation that stops at the laboratory bench has answered a question about a sample. The question an inspector asks is about the shelf, and about what else in distribution shares the formulation, the process or the component lot.

Bentley's response made the gap easier to see. The firm told FDA the root cause of its inadequate investigations was deficient investigation procedures. FDA accepted none of it, because the response did not change the procedures to require root cause assessment for OOS results. A procedure that does not require root cause assessment is a finding. A response that identifies that and leaves it in place has described the problem and done nothing about it. FDA's closing language on the topic names the consequence directly: inadequate investigations lead to unidentified root causes, ineffective CAPA and recurring problems.

Here is the pattern the investigation citation shares with the next two. Under 211.84, FDA says Bentley did not perform adequate identity testing on each shipment of each lot of incoming components, and relied on supplier certificates of analysis without establishing that each supplier's test results were reliable at appropriate intervals. For glycerin and certain other high-risk components, the letter points to a limit test in the USP identification procedure that screens for hazardous impurities, and says that test was not performed on each shipment. FDA writes that the use of ingredients contaminated with those impurities has caused lethal poisoning incidents in humans worldwide. The letter redacts the impurity names. The expectation it describes is not redacted: at least one specific identity test on every incoming lot.

The firm's answer followed the same shape as the investigation response. Bentley said it opened a change control to add impurity testing and tested raw materials. FDA called that inadequate, because the firm did not document its updated raw material specifications or explain how it would perform testing on each incoming component. FDA also noted the firm did not say whether it lacks retain samples for any batches and is therefore unable to test them. That sentence deserves a second read. Without retains, a firm cannot go back and prove what was in a lot it already used. The record of that proof is the only defense available after the fact.

Then the validation findings. FDA says Bentley changed the manufacturing process for a product to rephase the formulation, without revalidating the process and outside its change management system. For another product, validation execution departed from the firm's own Validation Master Plan, which calls for three batches, and no stability samples were retained from any of the three. Cleaning validation on non-dedicated equipment produced OOS swab results for active residue in the two batches that were sampled, and those were not adequately investigated. FDA also questioned a maximum allowable carryover limit used in the cleaning calculations, saying it appeared to lack scientific justification. That is a second place where a limit does the work an investigation was supposed to do.

Taken together, the findings point where FDA's own quality systems paragraph points: FDA found the quality unit is not enabled to exercise proper authority and/or has insufficiently implemented its responsibilities. That sentence is built from specifics like these. A failing pH value that meets a revised range. A raw material that arrives with a certificate and goes straight to the floor. A process change that bypasses the system designed to catch it. The letter does not say who reviewed any of them, which is the question an inspector would ask next.

For a QA director, the practical test is narrow and checkable. Pull the last five OOS investigations that closed with a CAPA. For each, ask whether the CAPA changed a cause, a control or a limit. Where it changed a limit, find the data and the change control that justified it, and confirm they were dated independently of the failure. Then check whether the investigation reached lots in distribution, not only lots on the bench. These are the questions FDA's letter shows an inspector asking, and the file either answers them or it does not.

Bentley's letter also lists what FDA wants in return: an independent assessment of the investigation system covering scope determination, root cause evaluation, CAPA effectiveness and quality unit oversight, a separate assessment of the CAPA program, and a review of change management. The list reads as a map of where the firm's own records could not answer for themselves.

Find out whether your OOS and CAPA records would survive that read. Get your Inspection Risk Scan.