26th Annual ISPE Aseptic Conference

FDA Panel Q&A

Disclaimer: This is an abridged, unofficial summary of FDA regulator’s responses during a panel dialogue at a conference that has not been vetted by the agency. The responses below are an informal and brief synopsis of the panel’s views, and do not represent official guidance or policy of the FDA.

In an effort to disseminate the information shared during the regulatory Q&A session, ISPE is making strides to publish the discussions. These comments are considered opinions only, and cannot be viewed as statements by the FDA; they do, however, provide insight into current directions and risk points to be considered when operating aseptic facilities.

An annual highlight of the ISPE Aseptic Conference is the regulatory discussion panel. Attendees can submit questions to the regulators in advance via the ISPE website; there is also an open microphone option for those brave enough to ask their questions directly.

This year’s FDA representatives were:

• Thomas Arista, Consumer Safety Officer, FDA/ORA/ORO/DMPTO
• Rebecca Dombrowski, Facility Reviewer, FDA/CDER/OPQ
• Lynne Ensor, Division Director (Acting), FDA/ CDER/OPQ/OPF/DMA
• Richard Friedman, Deputy Director, Science and Regulatory Policy, FDA/CDER/OC/OMQ
• Robert Sausville, Director, Div. Case Management, FDA/CBER/OMPT/OCBQ
• Jeremy Wally, CDR, PhD, US Public Health Service, Director, Regulatory Operations Officer, FDA/CBER/DMPQ

How do regulators review process design and what are the critical stage parameters that should be considered in designing an advanced aseptic process?

• On the review side, two internal groups (OPF’s Division of Process Assessment and Division of Microbiology Assessment) determine whether a good manufacturing design concept will ensure that the products meet specifications.
• If applications are vague, then the agency asks the owner for clarification, but many times the owner cannot clarify.
• For advanced aseptic processing, a meeting with the agency beforehand is highly recommended—especially if the approach is novel—to verify that the firm is on the right track.
• Our inspections and reviews are also interested in things like pressure differentials, transfer points, and other elements that can increase or decrease sterility assurance. On the inspection side, we evaluate procedures and controls as well as air patterns.
• During PAI [preapproval inspection], we will also refer back to the application to verify that you are doing what you said you would.

What are the agency’s expectations on the management of glove holes? Does the agency think a robust inspection process is better than using a glove tester? Are the expectations for interventions performed during media fills different for isolator vs. a cleanroom line (i.e., do we need to simulate every intervention in an isolator media fill)?

• The two potential weak points of an isolator have always been 1) glove holes, and 2) transfer ports.
• Both glove testing and visual inspection are important. As you heard in the conference, automatic testers can identify circa 30–70 μm breaches, while the eye can detect breaches around 300 μm.
• Expectations are clear in the FDA’s 2004 aseptic guidance: (https://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM070342.pdf). We expect you to look at gloves each day of use (preferably observing the gloves anytime you go into them), as well as via a routine automated/mechanical test.
• The whole point of preventive maintenance is to avoid making the difficult decision of what to do after finding holes in your gloves. So these expectations are basic to quality assurance. Prevention is better than detection. GMPs, quality engineering, and QA all say same thing: Build in quality by prevention, not by reaction.

Is there an expectation that interventions and operator proficiency in performing interventions are challenged in a media fill in the same frequency and manner as in a conventional filling line?

• Along with same aseptic technique, whether you call it a media fill or a process simulation, the point is it’s a simulation. The same aseptic technique is expected in an isolator as in a conventional cleanroom.
• Yes, you should simulate planned and unplanned interventions that can occur.
• Plan your media fill to represent what you will see and do in production. FDA always compares the simulation with production.
• If done right, you might be able to compress things (shifts) on isolators, so two shifts might be simulated in one semiannual media fill (instead of two). This is because isolators are not as vulnerable to shift changes as are less protected processes. This guidance, however, does not apply to RABS [restricted barrier access systems]. A RABS process should still be simulated using two media fills per shift for each line each year. See the FDA website for Q&A guidance.

What is considered an acceptable practice for the handling holes in RABS gloves? Is there a recommended frequency for challenging the gloves? Is a standalone qualitative visual inspection an acceptable practice? What frequency is expected for resterilizing gloves?

• Sterilize and install gloves for every batch or campaign. If you campaign (anything more than one day) any aspect of your RABS operation (e.g., hoppers), it should be only a few days, never as long as for an isolator.
• If you encounter problems, response is similar to the situation with isolator (see Q2).
• Like an isolator, gloves should be visually inspected and subjected to automated testing. If you are going to campaign, you should use an automated test during the campaign.
• For RABS, it is critical to disinfect installed gloves during a campaign (after open-door interventions) as well.
• We recommend automated testers. The same concepts apply for either RABS or isolators. The principle is that the doors/walls provide a barrier for sterility assurance; to the extent the barrier is compromised, then nonintegral gloves are a tough situation.
• Need an understanding how often gloves could be resterilized, how much wear and tear they could take. You must sterilize RABS gloves before a campaign. You must know the maximum n times that gloves may be sterilized.
• The other way to compromise a RABS is by opening a door. See discussion of open and closed RABS in FDA’s Compliance Program Guidance Manual 7356.002a: “Sterile Drug Process Inspections.”
• If you have holes, then you must investigate the deviation and address the problem. If you open the doors even to the Grade A/HEPA filtered air outside the RABS, then you must disinfect gloves frequently, same as a conventional cleanroom. You must use testers for RABS, same as for isolators.

How far are we from the world that requires isolators only for aseptic filling? For how long is there room for closed-door RABS with glove ports and rapid transfer ports? What are the expectations?

• We are all proponents of 1) isolators and 2) RABS. There are circumstances in which isolators may not be possible, so well-designed RABS are fine. The main point is GMP—we will hold you to that!
• There are opportunities where even those companies can move to next level. We are encountering the conventional equipment less and less. So such firms can move to glove boxes and RABS, rather than using the old paradigm of a hood or a processing line with limited barriers.
• More advancement in protective technologies should help shorten inspections! RABS/isolators are incentivized because you are likely to have more reliable performance and lower scrutiny. There is no set date to require isolators.
• For years many conventional cleanrooms have operated on the edge of failure, so overall, we do expect some sort of barrier technology for aseptic processing (e.g., glove boxes, RABS, isolators) in nearly all cases. See the cGPM documents on the FDA website.
• The bottom line is that advanced aseptic technology means at least RABS, and particularly RABS with no open-door interventions.
• It really depends upon circumstances of the product and production.

Do you support the use of RMM [rapid microbial monitoring] for EM inside an isolator? If used appropriately, do you feel it could have the potential to reduce the frequency and number of points during isolator EM [environmental monitoring]?

• The methods we know about for EM are continuous monitoring devices. We were a little puzzled in the question about less frequency, etc. You are going to get a lot of data points. It’s difficult to answer except to say there are going to be fewer classic surface samples in an isolator just by design due to sampling at the end of a campaign, so you have some savings there. RMM includes lots of data and continuous monitoring that provides valuable insight into state of control of the process environment—that is great! Two different concepts are being conflated: RMM vs EM.
• Continuous monitoring with RMM can be a valuable method in conjunction with appropriate routine EM methods to detect any microbes that might be present.
• Re: the number of points—if they are the validated locations that are important to monitor, how do you conclude that you no longer need to sample those locations? Why? If you have justified valid locations for EM, then how could RMM eliminate the need for monitoring?

Should the room for an aseptic isolator for highly potent product (sterile powder handling, liquid, or freeze-dry processing) be pressurized positive or negative?

• This is tough, right? You don’t have to use a negative pressure isolator—we have seen companies have difficulties with them. You could use a closed positive-pressure isolator, or you could use open positive pressure and control egress points. The operator could use PPE [personal protective equipment] (done classically before there was a negative-pressure isolator).
• Those options and others are all there before you go to negative-pressure isolator.
• If you decide a negative-pressure isolator then you would have and antechamber, surrounding room, and an airlock into the room.
• The bottom line: negative-pressure isolators should likely be your last choice, but they could work if other preferred options are not suitable.