Common issues found during agency inspection of cleaning programs and cleaning validation at medical device manufacturers include choice of the wrong worst case device for validation studies, introduction of contaminants during the cleaning process, inappropriately selecting “visually clean” as a cleaning endpoint, and a lack of understanding of cleaning load size and tank cleanout frequency.
At the FDA/Xavier MedCon conference held virtually in April 2021, FDA Office of Regulatory Affairs (ORA) Office of Medical Device and Radiological Health Outcomes (OMDRHO) Division I Investigator and Medical Device Specialist LCDR Tom Peter from the agency’s Detroit duty station shared his insights on CDRH field investigator expectations for cleaning programs and cleaning validation.
Definitions and Standards
Peter began by sharing pertinent definitions from ASTM medical device cleaning standard F3127. He explained it is an FDA consensus standard that the agency recognizes “in whole.” It defines “clean” as, “having a level of residues and environmental contaminants which do not exceed a maximum permissible level for the intended application.”
He pointed out that the definition “does not necessarily mean 100% free of all contaminants. It is up to your organization to define what the maximum allowable level of contamination is based on the risk involved in the intended use of your device.”
The standard defines a “contaminant” as, “any material that potentially adversely impacts the assembly, the functioning of the device, and/or shows undesirable interaction with the host,” meaning that it is not biocompatible.
Peter provided an overview of some of the more common contaminants investigators see during medical device inspections (Figure 1). He pointed out that they will vary depending on the nature of the device and the manufacturing methods involved in making those devices.
Machine lubricants and mold release agents are some of the more common contaminants. “Notice that detergents and cleaning solvents appear on this list,” Peter pointed out, “which is notable because they are contaminants that are introduced by the cleaning operation itself. It highlights the importance of having a rinse step after your cleaning operations have been completed.”
Most of the common contaminants are chemical or physical in nature, but some like microorganisms and endotoxins are biological.
Device Functionality and Biocompatibility
“Why is cleaning important?” Peter asked. “There are two things we really need to focus on here. One is biocompatibility, the other is device functionality.”
From a biocompatibility standpoint, failure to remove contaminants to an acceptable level can lead to an adverse biological response. Depending on the contaminants pertinent to a device, those should be assessed in the biological evaluation for all relevant endpoints—typically, cytotoxicity, irritation, and sensitization, at a minimum.
Contamination could also adversely affect the functionality of a device. For example, if two components need to be bonded together during a manufacturing operation and there is some level of residue on the components prior to bonding, it might be critical to remove the residue to make sure it does not compromise the bond strength. If the bond fails during the use of the device, it can affect its functionality.
When it comes to biocompatibility, it is common to hear people only think about the materials that go into making the device—the materials that comprise the device itself. But the ISO 10993-1 standard and the associated FDA guidance are clear that there should also be a focus on the manufacturing methods used to make a device as well as any residuals that could be introduced during manufacturing when performing a biological risk evaluation.
Cleaning and Sterilization Are Not the Same
While cleaning and sterilization may complement one another, it is important to keep in mind that they are two distinct processes, Peter stressed. “That is something that we commonly see misconstrued during inspection.”
Thinking back to the overview of some of the more common contaminants that investigators see, one of those was microorganisms. Sterilization is killing contaminant microorganisms to achieve the target sterility assurance level. The risk of not doing so is infection.
When it comes to cleaning, however, the agency is not just concerned about removing biological contaminants below a defined level, but also the physical and chemical contaminants. The risk of not doing that is not just infection, but an adverse biological response or toxicity—also known as biocompatibility.
One area where cleaning and sterilization complement one another is cleaning operations where the goal is to reduce bioburden. That is especially important for sterilization methods such as gamma irradiation, which is highly dependent on bioburden on the device. If bioburden is not reduced to the expected level, the sterilization dose may not be sufficient to achieve the target Sterility Assurance Level (SAL).
Machine lubricants and mold release agents are some of the more common contaminants
Failure to adequately clean a device could lead to nonviable contaminants entrapping microorganisms on the device and making them more difficult to kill during sterilization. In the case of ethylene oxide (EO) sterilization, contamination could impede the flow of EO gas and not allow it to contact the entire device surface, which would compromise the sterilization process.
“One of the most important points I want to make,” Peter stressed, “is that it is possible for a device to be sterile, but not clean.”
For example, consider an implantable medical device that has some level of particulate contamination on it, and it is shipped out for sterilization. It becomes sterilized. The device and the particles are both sterile, so infection is not really a concern. But when the device is implanted, it is possible that depending on the nature of the contamination, it could elicit a foreign body reaction or some other adverse immunological response. That would mean that the device is not clean, but it is sterile.
Peter reported that the agency has had firms try to diminish the importance of some of the cleanliness concerns that it has uncovered during inspection by saying that the device is sterile. “But it is important to keep in mind that sterility and cleanliness are two separate issues and they need to be thought of that way.”
A Risk-Based Approach for Determining Cleanliness Requirements
The OMDRHO investigator provided an overview in the form of a flow chart on how to take a risk-based approach to determine cleanliness requirements (Figure 2).
He recommended starting off by evaluating the manufacturing process flow, i.e., the steps involved in making the device. For each of those steps, identify the potential contaminants that could be introduced. They could be manufacturing residues, environmental contamination, microorganisms, etc.
For each of the contaminants identified, assess the risks, both from a biocompatibility and a device functionality standpoint as if that contaminant were to remain on the device. If the risk is negligible or there is no risk, if it could be perfectly acceptable that the spec is just that it needs to be visually clean. Maybe it is just a cosmetic speck, something that can be visually inspected and fully verified. In that case, the cleaning process might not need to be validated.
If, on the other hand, there is a higher risk of either biocompatibility or device functionality, quantitative limits may need to be defined for how much of the contaminants can remain on the device to mitigate the risks.
“In my experience, typically these limits are in the milligram or microgram level—not something that can be visually verified.” Peter commented. “In those cases, there is a need to validate the cleaning operation to assure FDA that your cleaning operation is consistently resulting in devices that meet your cleanliness requirements.”
Common Cleanliness Issues Seen on Inspection
The following are some of the common issues Peter provided that the agency sees during inspections from a device cleanliness standpoint.
The agency investigator previously emphasized the importance of identifying the contaminants that could be introduced during the manufacturing process. “It is quite common those are missed,” he said.
If potential contaminants are not being identified, then it is not possible to assess the risk of those contaminants remaining on the device. “So, everything kind of breaks down from there.”
Sometimes investigators will see that firms establish cleanliness requirements as visually clean. However, based on the nature of the device—for example, if it is implantable—there should be quantitative limits based on the risk involved.
It is important to keep in mind that sterility and cleanliness are two separate issues and they need to be thought of that way
The concept of “visually clean” takes on special significance when operators report gross contamination on devices or customer complaints alleging visual contamination. Peter characterized this as a “big red flag,” because that is just the contamination that someone can see.
“It begs the question of, what are we not seeing?” he asked. What is being shipped out the door is conforming based on just being visually clean, whereas there could be levels of contamination that present a patient risk. “So that is a concern.”
Issues with cleanliness requirements will cascade into cleaning validation, “because if your requirements are not squared away, of course, your acceptance criteria and validation will have nothing to trace back to,” Peter emphasized.
Sometimes investigators will see firms run many different types of devices through a single cleaning process depending on the production schedule. And they will establish what is the worst-case device and challenge that during their cleaning validation.
That can be an appropriate approach. “But I would like to remind you to document justification for the worst-case device. Consider things like the size of the device—the surface area—and difficult to clean geometries and surfaces and parts such as threads or lumens,” Peter advised.
If two different devices go through the same cleaning operation but one of them goes through some additional steps upstream whereby additional contaminants could be introduced, that could be a worst case. Many things need to be considered when determining what is the worst-case device. And the justification must be documented.
After validation, it is common to see products adopted into cleaning families. Ensure each is justified and documented and that a new device does not create a new worst case. Agency investigators sometimes see that that is missed, “or worse yet, something is adopted in and it creates a new worst case and the process was not revalidated.”
The concept of ‘visually clean’ takes on special significance when operators report gross contamination on devices
Inadequate process characterization is a common issue. Before validating a process, critical process parameters must be identified and then challenged during the validation.
“The biggest blind spot we see during coverage of a cleaning operation is typically the load size and tank cleanout frequency—how many devices are allowed to go through a cleaning tank before it needs to be cleaned out.” As the volume of product going through the cleaning tank increases, the conditions of the tank will start to deteriorate. The worst-case tank conditions should be challenged during the validation.
There are also potential issues with test method validation. For example, the test method used to verify cleanliness must be validated prior to cleaning validation and shown to be suitable for its intended use, including evaluation of the limit of detection and recovery efficiency. These parameters are needed to calculate an appropriate correction factor to get the finished result.
Purchasing Controls, Process Inputs and Monitoring
Regarding purchasing controls there are a couple of considerations. “If you are an Original Equipment Manufacturer (OEM) using a contract manufacturer and their manufacturing process involves one or more cleaning steps, make sure that you are communicating cleanliness requirements to them. They must have that information to validate their cleaning operations appropriately.”
On the other hand, “if you are a contract manufacturer and you have not received that information from your OEM, it is fair to ask why not. And please do. Because sometimes it is something that is just not considered by the OEM and eliciting that type of requirement could trigger them to think twice and give them some pause.” The OEM and contract manufacturer have mutual responsibilities for product quality.
When using a contract lab for testing, make sure that the methods they use are suitable for your intended use. Sometimes investigators will see that firms confirm that the contract lab has a method in place, but there is no assessment of whether it can detect and measure specific residues that are relevant to a specific device.
Regarding monitoring the cleaning process, when periodically testing for cleanliness as a means of monitoring the cleaning operation, “please make sure that whatever devices you are testing are going through the normal manufacturing process flow so that they are representative of production devices.”
The agency has seen this as an issue, especially when firms use simulated products. Sometimes the devices may take shortcuts and some steps might be omitted. When that happens, it is not representative of a production device and monitoring such a device does not provide assurance that the process is in a state of control.
“Finally, please make sure that the water system you are using for cleaning and rinsing and the compressed air systems used for drying are validated and monitored,” Peter stressed. “It is important to keep in mind that it is not possible to demonstrate a state of control of your cleaning operation if you do not know for certain that inputs to that operation are in a state of control. And in the case of cleaning, that includes your water systems and your compressed air systems.”
[Related: Medical device manufacturers can click here to get a FREE 483 Observation Report focusing on specific observations, such as cleaning validation.]
Subscribe to Redica Insights
Get quality and compliance insights from our experts in your inbox