Incorrect Specifications, Process Validation Issues at CMO Lead to Adverse Events, Recall

At the International GMP Conference held virtually in early March 2021 co-sponsored by the University of Georgia at Athens and FDA, FDA Office of Regulatory Affairs National Drug Expert Captain Ileana Barreto-Pettit presented FDA drug GMP Warning Letter trends and provided in-depth case studies from recent drug GMP inspections illustrating agency concerns and findings.

Barreto-Pettit has been with FDA for 22 years.  She has been a drug investigator since 1999 and a Drug National Expert Investigator since 2017, and is a Captain in the U.S. Public Health Service Commissioned Corps.

The GMP inspection case studies Barreto-Pettit provided include an in-depth analysis of the findings, lessons learned, and how companies can avoid similar shortcomings.  Areas examined in the case studies are:

• The first countrywide import alert issued by FDA
• Inadequate deviation investigation
• Inadequate product specifications and a product recall
• Process validation
• Different perspective on process validation and the culpability of the quality unit

The first two case studies were discussed in the first two parts of this story, available here and here. The third and fourth case studies are discussed and explained in this part of the story. The final case study is detailed in Part IV.

[Related: Check out our latest report, a comparison of two of the top CDMOs, here.]

Case Study 3: Inadequate Specifications and a Contract Manufacturer

The third case study Barreto-Pettit presented at the UGA conference involved a scenario that resulted in a mission critical inspection conducted during the COVID-9 pandemic due to adverse events that were reported to the agency after patients took a drug product that was intended for the treatment of thyroid disease.

In this case, the firm, which was both the product owner and the distributor of the product, had established the wrong assay specifications for thyroid tablets in that the upper limit was too high.  21 CFR 211.160(b) requires the establishment of product specifications but the company failed to do so.

The product in this case study is an unapproved prescription drug product for which there is no approved application.  Therefore, the manufacturer must follow the USP monograph for the specifications.  The product is contract manufactured. “Both the product owner and the contract manufacturer share the responsibilities for adequate specifications,” Barreto-Pettit stressed.

The USP specification for assay for this thyroid product is 90.0% to 110.0%.  However, the firm erroneously established its specifications with a higher and wider range.  “This was probably one of the reasons patients were experiencing adverse events when they were taking this medication,” she postulated.

In the Warning Letter, which was issued to the product owner, FDA acknowledged that the product owner changed its specifications to the correct ones and recalled some lots that had exceeded the newly established USP specification upon release.  However, since this was a product owner and not the manufacturer, the agency conducted a subsequent inspection of the contract manufacturer and found 13 additional lots manufactured with an API from a previous supplier that also exceeded the USP specifications but were not initially recalled (Author’s Note: See the next case study below for more detail).

In addition, FDA investigators collected samples of the product for testing and found several batches that had low out-of-specification results that also had to be recalled.  “This was interesting because their specification had a higher upper limit, but when we tested the products, they would come in low. We knew that there was an issue going on with content uniformity,” she commented.

Case Study 4: Process Validation Issues Investigated at CMO

The fourth case study is the follow-up inspection of the contract manufacturer of the thyroid tablets that was the subject of the third case study, where the agency identified an additional 13 lots that exceeded the USP assay specification of 90.0% to 110.0%.  Since the agency had seen variable assay values ranging from the upper 80s to values over 110%, investigators reviewed the process validation documents for that product and found other deficiencies.

For example, it found that the firm lacked evidence to demonstrate the homogeneity of the powder blend it was formulating into tablets after it was blended in the V blender.  And even though the company collected samples from multiple locations within the blender, personnel combined the samples in the laboratory and tested them as a composite sample, only getting one assay result.

Both the product owner and the contract manufacturer share the responsibilities for adequate specifications

Agency investigators found upon further review of the validation documents that some of the validation batches also exceeded the correct assay specifications of 90.0% to 110.0%.  But these were not identified during the company’s investigation, which did not extend to the process validation or stability lots or any other batches that were distributed and were still within expiration dating.

FDA found that the most recent validation batches after using a new API supplier were not placed on long-term stability until a year later, which was about a month before the inspection discussed here. Therefore, there was no stability data.

The agency also cited the quality unit under 21 CFR 211.22(b) for not ensuring that the product was manufactured in accordance with CGMPs, that investigations were complete, and that their product met specifications.

Warning Letter Focuses on Process Validation, Blend Uniformity, and Control Strategy

A Warning Letter was issued to the contract manufacturer in January of this year, 2021.  The first citation was 21 CFR 211.108, which is inadequate process validation.  The Warning Letter states that the company lacked data to show the homogeneity of the powder blend, as the samples that were collected at different locations within the blender were tested as a composite sample and therefore the company only had one result and was not able to detect variability within the blend that could have led to content uniformity issues of the tablets.

During the inspection, FDA collected samples of three different batches, two of which were from the process validation batches.  All three samples had results that were low and the samples were subpotent for the active ingredient.  One of them also failed content uniformity.

“It is important to understand that blending is a very critical step in the manufacturer of oral solid dosage forms, particularly for narrow therapeutic drug products like these thyroid tablets,” Barreto-Pettit pointed out.  “And when the blends are not handled adequately, it can promote segregation, increase the moisture levels, cause particles to aggregate, and can lead to inconsistent flow characteristics when they are being compressed.”

Agency investigators found upon further review of the validation documents that some of the validation batches also exceeded the correct assay specifications

Additionally, because of the narrow therapeutic range of this product, content uniformity is critical.  It is especially important to prevent patients with hypothyroidism from receiving insufficient or excessive doses.  As a result of these content uniformity issues, the company had to recall all the lots on the market.

The Warning Letter also stated that the firm’s response to the FDA 483 was not adequate in that it did not provide sufficient data to show where the variability was occurring in its process and what its control strategy was before the company revalidated the process.

Concurrent Process Validation Not the Best Choice

The firm stated it would perform a concurrent process validation instead of a prospective process validation.  In concurrent validation, the firm manufactures validation batches but each batch is evaluated concurrently after meeting specifications but before completing the process validation.  “This approach should be rarely used, especially where process knowledge is limited and a control strategy has not been identified fully,” the FDA drug expert stressed.

The Warning Letter also cited the firm for not having continued process verification, which is stage three of process validation, to ensure its manufacturing process remains in control and produces a drug product that consistently meets specifications.

The Warning Letter required the contract manufacturer to provide the following:

• A data-driven and scientifically sound analysis identifying all sources of variability, including but not limited to the raw materials and any type of manual steps such as the hand scooping personnel used to transfer the blend to the hopper, etc.
• A determination of the capability of this manufacturing process step and a CAPA to reduce the process variation
• A detailed summary of the validation program to ensure the process remains in control throughout the product life cycle along with any associated procedures
• More specifics about the process performance qualification and how the company is going to monitor both intra- and inter-batch variation to ensure they continue in a state of control

Evaluating State of Control in Manufacturing

The following is a set of questions Barreto-Pettit prepared for firms to use in evaluating manufacturing practices:

• Do your quality unit and the responsible departments have sufficient knowledge of the manufacturing process and the sources of variability?
• How do you monitor the sources of variability and control them?
• What systems do you have in place?  Do you have systems that are adequate as far as procedures to detect a shift in the process and identify new risks?
• And if those risks are identified or shifts are detected, how is that being communicated to the appropriate parties?  And how do they correct it?  How are they documented and monitored?
• How often do you evaluate your process to determine whether process improvements, new equipment, and/or new technologies are needed?  Do you have a system in place in writing for systematically doing this type of evaluation?

[Editor’s Note: Part IV will provide a detailed analysis of a separate case study involving process validation.]

[Related: Check out our latest report, a comparison of two of the top CDMOs, here.]