GMP Case Studies Presented by FDA Investigator Ileana Barreto-Pettit: Part II

Execution of proper cleaning regimens and cleaning validation continue to be problematic in the pharma industry and the subject of FDA Warning Letters. Why do these practices continue to be deficient after being cited in Warning Letters over many years? And how can companies evaluate their cleaning practices to ensure they are compliant and are not causing cross-contamination that can endanger product quality and patient health?

[Related: Download a FREE compilation that includes all the case studies in this series by Jerry Chapman here.]

At the International GMP Conference held virtually in March 2022, co-sponsored by the University of Georgia at Athens and FDA, FDA Office of Regulatory Affairs National Drug Expert Captain Ileana Barreto-Pettit presented four in-depth case studies from recent drug GMP inspections that illustrate agency findings and concerns.

Barreto-Pettit has been with FDA for 23 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:

Inadequate change management systems
Cross-contamination in solid oral dosage facilities
Inadequacies and potential contamination hazards in aseptic manufacturing
Perforated tables and work surfaces in cleanrooms

Part I change management findings in Warning Letters is available here. Part II of this four-part series covers Barreto-Pettit’s discussion and analysis of cross-contamination in pharma manufacturing facilities. The remaining case studies will be covered in the subsequent two parts.

A Persistent Problem

Barreto-Pettit characterized the issue of cross-contamination in solid dose operations as “a persistent problem” that has been “happening for years.” She began her presentation by referencing the talk she gave at the Georgia Conference in 2020 using fluid bed dryers and ductwork as examples.

[Related: According to our data, 87 inspections at Human Drug GMP sites resulted in 483s with the primary citation of 21 CFR 211.42(b)(c). Want to dive in and see the details? Sign up for a personalized walkthrough of the report used to find this data in Enforcement Analytics.]

“We keep finding this,” she pointed out. “I am going to keep reminding companies that you need to look at your cleaning programs. This issue has occurred for decades, including multiple times in the last few years.”

For example, process equipment ductwork has been found contaminated with residues from multiple drugs. Agency investigators have observed visible residues in fluid bed dryers as well as tablet coaters and have confirmed the presence of material from more than one drug with rinse and swab samples.

I am going to keep reminding companies that you need to look at your cleaning programs

“Residues that have been swabbed from ductwork have shown up as having contaminants. Samples have also confirmed cross-contamination between finished drug products. Many firms have had to recall numerous products and batches,” Barreto-Pettit said.

She cautioned that cross-contamination cannot be assumed to be uniformly distributed, characterizing that mindset as “an oversimplification of risk assessment.”

Poor Cleaning of Equipment Draws Warning Letters

Warning Letters have been issued to multiple firms for equipment cleaning issues, she said–for example, to companies that produce drugs with significant potential toxicity or pharmacological activity on multiuse equipment, including products such as hormones and other potent compounds.

Barreto-Pettit provided some general inspection findings related to cleaning. She explained that agency investigators have found:

Fluid bed dry cartridge filters without unique IDs that are labeled as clean and ready to use but have foreign materials such as black substances inside the filter groove as well as powder residue in the bottom of the storage container. When the filters are washed and then dried, they still have some powder residue within the filter that then falls into the storage container.

If the fluid bed cartridge was dedicated and the powder residue is from the same product all the time, “it may not be such a big deal,” she pointed out. But the case in point was a filter that was unidentified and therefore there was no assurance that it had been used for only one specific product and not caused some cross-contamination as could happen if it had been used in a different product and not properly cleaned.

Powder residue in multiple types of manufacturing equipment that has been labeled as clean.
Clean equipment that dripped what looked like accumulated water from the interior product contact surfaces when it was dismantled. When the agency investigator asked the firm to dismantle a particular piece of equipment, water came out from inside it.

“Think about it,” she said. “Water has been sitting there for days or even weeks. So, most likely, it contained some sort of microbial contamination that would then cross-contaminate the next batch if the water was not properly flushed out of the equipment before storage.”

Various pieces of equipment that show what looks like rust on product contact surfaces.

“When things this obvious happen, I wonder if anybody is really taking a close look at the equipment and inspecting it properly before using it. This must be done every time,” she stated.

…cross-contamination cannot be assumed to be uniformly distributed…

Here is one of the citations in the Warning Letter issued to this firm: “Your firm failed to establish and follow adequate written procedures for cleaning and maintenance of the equipment (21 CFR 211.67(b)).”

“This is a citation that has been in the top 10 since I have been at the FDA, and that has been over 20 years,” Barreto-Pettit lamented. “We continue to find issues with cleaning programs.”

[Related: Below are the Top 5 FDA Enforcement Analytics issues found in the Top Issues Report (Filtered by Date, Human Drugs, GMP, Form 483). SEE THE FULL TOP ISSUES REPORT.]

The citation further reads, “Your cleaning procedures for non-dedicated equipment, including your tablet presses, are inadequate. Our investigators observed drug residue from previously manufactured drug products inside one of your tablet presses, which was documented and released as clean by your quality unit. You use this tablet press to manufacture several potent and non-potent drug products.” Barreto-Petit characterized this as a “significant” risk.

The Warning Letter continued, “Investigators also searched residue on the equipment as well as on the duct of the equipment. Cleaning and preventive maintenance procedures for this equipment which was installed 14 years ago in 2008 did not include cleaning instructions for these areas or regular inspection of the air ducts.”

Common Deficiencies Found by Barreto-Pettit

The FDA National Expert Investigator shared some of the most common deficiencies in this area that she has found during inspections, along with her comments:

Equipment erroneously labeled as clean: “We see visible residue, which shows me that the visual inspection practices at the company are not adequate. If we go to where the equipment is stored and is cleaned, and we just select a random piece of equipment and we look at it and we see residue, that shows to us that the visual inspection practices are not adequate. So, why is that happening?”
Equipment use and cleaning logbook deficiencies: Logbooks are often judged to be deficiently written, incomplete, illegible, and not reviewed by anyone. Sometimes they are uncontrolled, such as loose pages in a binder. “These equipment logbooks provide a lot of useful information that companies need to capture–for example, was it a major or minor cleaning? I have found many equipment logbooks that do not include that information. It should include the lot numbers and products that were manufactured at the time of cleaning. All that information needs to be there as well as who inspected it.”
Documentation of failed visual inspection: “If equipment fails visual inspection and needs to be recleaned, why is that not documented in the logbook? I would say probably 90% of the time recleaning is not documented in logbooks. I just find out because of interviews or maybe because of some other reasons, but not because it is documented in the logbook. And it should be.”
Failure to follow cleaning procedures: Barreto-Pettit has found that personnel do not always follow cleaning procedures. “This is so easy for you to find out the same way we do,” she pointed out. “Interview your cleaning personnel, observe what they are doing, ask them what they are doing, and compare that to the cleaning procedures. Many times, they do not have the cleaning procedures with them when they are cleaning. I review them and I find out that whatever they were doing at the time is not necessarily the same as the cleaning procedures. And why is that? You do not want your personnel deviating from your validated cleaning process because you may not have the same outcome.”
Availability of cleaning SOPs: “I have found that sometimes SOPs for cleaning are not easily accessible by personnel during cleaning activities. This is a very common observation that I find. Sometimes I find that the SOPs are stored in the supervisor’s office in some binder where employees cannot easily access them unless they ask a supervisor. Sometimes the SOPs are electronic, and they are accessible only through a terminal that is in a hallway, not always near the washrooms.

“There have been many times I have asked operators to access cleaning SOPs for me, and it is unbelievable to me to find out that sometimes it seems like that was the first time they even accessed the system–they do not remember passwords or know where to go. Even when they get the list of SOPs, they do not know how to find them or which one applies. That gives me the impression that they are not accessing the SOPs to perform these activities.”

Incomplete equipment disassembly: “I have found equipment that is not sufficiently disassembled for cleaning. This can cause problems like the example that I provided with the pump, where it was not sufficiently disassembled and there was water that remained within the interior of the pump. You need to know this, and you need to identify which pieces of equipment require more thorough disassembly.”
Inadequate cleaning validation: Some companies even today have not performed cleaning validation or whatever they have is inadequate–for example, their limits are not appropriate or not appropriately justified. “Maybe the cleaning procedures do not ensure that the cleaning processes are consistent because they are too vague,” Barreto-Pettit commented.
Incorrect preparation of detergents: “Preparation of detergent is very important. The company went through the trouble of validating a cleaning process with a certain concentration of detergent. If the personnel are not following the same instructions, the detergent will be either too diluted or too concentrated and might not be easy to remove.

“I find that sometimes there are no instructions within the procedures for the preparation of detergent, or sometimes the instructions are confusing. They may require the operator to do a calculation in their head. For example, the SOP may instruct the operator to prepare a detergent in a three to one concentration. The operator may not understand what that means.

“Why do you not specify what it is that you want to prepare? ‘Prepare five gallons by using five gallons of water and adding a half a cup of this detergent.’ That would be a lot easier to follow than specifying a percentage of a concentration that would require the operator to do math in his or her head.”

Improper equipment storage: “Sometimes I find that clean equipment has been stored improperly in that it is wet and covered, so it cannot dry properly. Or sometimes it is not covered when dry and properly stored and can get dusty.”
Incorrect performance of surface swabbing: “Swabbing of surfaces for cleaning validation or verification sometimes is improperly performed; either the swabbing technique is incorrect, or the wrong type of swab is used, or the swabbing locations are not representative of the worst cleaning locations. Sometimes the swab areas are too small. And I sometimes find that companies do not have recovery studies for the swabbing, which are necessary to ensure that the testing that can capture the residues.”
Monitoring of cleaning effectiveness: “Another thing I find is no monitoring of cleaning effectiveness. I find that companies sometimes validated their cleaning process maybe 10 years ago and have never looked at it again. There is no monitoring and no swabbing to check that the cleaning process is still effective.”

Equipment Cleaning Self-Evaluation

In addition to providing inspection findings on equipment cleaning and cross-contamination, Barreto-Pettit presented a list of questions she suggested companies ask to conduct a self-evaluation of their cleaning practices.

“Have you identified all cross-contamination hazards, whether chemical, micro, or physical, associated with each piece of manufacturing equipment, including those areas that are not routinely disassembled? Look at this–especially procedures that have been in place for a long time–to see if they are adequate to ensure that all cross-contamination hazards are addressed.
“Are your cleaning procedures detailed enough? Are they easy to follow? Do they contain any pictures that are helpful for operators to follow for disassembly of the equipment or swabbing? Consider including some checklists. I have seen some companies that have implemented cleaning batch records, where the operators are required to check the things that they do to assure consistency. I think that is a great improvement. And are the cleaning procedures accessible, as I mentioned before?
“How do you ensure personnel are properly trained, and always, not only when you are watching, follow the validated cleaning procedure? How do you know that? How do you know that they are rinsing the equipment for the correct number of minutes? How do you control that human variability?
“Is there an opportunity to automate some of these cleaning processes? And if so, why are you not doing that?
“How do you monitor equipment cleaning effectiveness, and how often do you have justification for the frequency of your cleaning effectiveness or cleaning monitoring program? How often do you reevaluate your cleaning process based on the data that you are gathering? You need to have a program that evaluates your cleaning process every so often to see if it can be improved. Technology changes and you need to keep up with those changes.”

In Part III, find out what happened when a company’s response to a 483 observation involving discolored HEPA filters referred to them as “recolored.”