How Early Communication with FDA Sped Up One RMM Implementation
Rapid Microbial Methods (RMMs) offer another tool that can facilitate Lean Manufacturing in biologics and pharmaceutical manufacturing, and taking the industry a few steps closer to parametric product release. Proactive communication with FDA reviewers was instrumental in getting approval for one RMM implementation project at a major pharma company. Rich Boehler discusses how his team at a previous (unnamed) employer went about the implementation.
By Agnes Shanley, Editor in Chief
Microbial contamination is a real and present danger for manufacturing many biologics and pharmaceuticals. Rapid Microbial Methods (RMM) systems and methodologies can help manufacturers release product faster and manage the risks of contamination while ensuring product quality. While they're not necessarily a “slam dunk” to implement, RMM’s can be far less difficult to work with than current guidance documents, such as PDA Technical Document 33, might suggest. The key is focusing on the elements of PDA Technical Document 33 that are most important, and communicating proactively with FDA.
At the Lean Pharmaceutical Manufacturing conference in Princeton, N.J. on Nov. 9, Rich Boehler, Microbiology Group Leader with Luitpold Pharmeceuticals Inc., discussed his experiences as part of an RMM implementation for a major pharmaceutical company earlier this year. Boehler could not disclose the name of his former employer, but the project focused on applying ATP bioluminescence (for details on ATP bioluminescence, see www.celsis.com/rapid_pharma.htm), which he had studied in graduate school. In this brief interview, he recaps best practices and lessons learned.
Ph.M – What attracted you to RMM, and where did you want to apply it?
R.B. – Naturally, one of the major draws was faster product release, but RMM’s sensitivity and ability to zero in on specific organisms were also key factors in the decision. We wanted to use RMM instead of traditional microbial limits test for end release products, in our processes and in raw materials.
Ph.M – What was your overall strategy to implement RMM?
R.B. – We knew that we needed to get approval from the FDA to utilize RMM. Since we were already releasing product using traditional methods, we decided to file a supplemental NDA to use the Celsis Rapid Detection system. Since the FDA had recently introduced the PAT initiative to drive product quality, we decided to work with this dedicated FDA team on the project. We also decided to take a product-by-product approach. We’d start with our highest volume, highest revenue product. Later products would follow the first prior approval supplement. When you file an NDA or SNDA you must mention the specific technology you are validating. This is the reason it is appropriate to mention Celsis.
Ph.M – What overarching themes guided this implementation project?
R.B. – We knew that we’d have to address both validation and regulatory issues and that we’d need a strong scientific rationale for making a change. We also knew that we’d have to be committed to working through any and all potential issues during validation. We found a great deal of resistance to change within our QA and QC departments. To overcome this, we observed the need to provide a pragmatic overview of the rapid technology and ensure timely and thorough training of the QA staff.
Ph.M – How helpful was PDA Technical Document 33 in helping you establish goals for this project?
R.B. – PDA Technical Document 33 is important, but I don’t think it’s possible to do everything that it lists. This is a reason why, I believe, many RMM projects fail to move forward in pharma companies. People will meet to discuss RMMs, look at 33, see that there’s just too much to cover, and give up.
Ph.M – Why is that? Can you give an example?
R.B. – Well, for instance, to satisfy 33 requirements for one criterion, you’d need different RMM machines to run studies under different conditions. How many companies launching a new technology are going to invest in more than one piece of the equipment at first? That was one reason why we decided to contact FDA’s PAT Team, to open up discussions and get feedback on our plans and how reasonable they were. Every company worries about whether any process change will meet regulatory expectations, and direct communication with FDA can sort out the issues.
Ph.M – How much data should you provide to FDA?
R.B. – You can’t submit everything. By communicating closely and carefully, our team leaders and the company’s regulatory department determined what information would be best to show them for the initial discussion. We decided to provide an extensive literature review, and data gathered after one product was run through the entire validation process.
Ph.M – How did you communicate with FDA?
R.B. – We started with a teleconference in which we presented our validation plan. Dr. Bryan Reilly, a reviewer at the FDA, really liked what we had submitted. He had some suggestions and initial concerns, which we welcomed and addressed. At times, our project’s voice became that of our internal regulatory department, which also met separately with FDA. But we took pains to ensure that our team took a collaborative approach, kept an open mind, and maintained open lines of communications between technical and regulatory staff and management.
Ph.M – Were there any side benefits to this approach?
R.B. – The FDA’s positive view of our initial validation plan was enough to convince some skeptics that RMM deserved attention. Once FDA accepts your proposal, it becomes official that, “Yes, you can do that.”
Ph.M – How do you validate the method?
R.B. – Validation requires evaluating sample effects, to establish that the product doesn’t inhibit or enhance the testing reagent’s activity. This establishes whether the drug is a good candidate for the methodology. The other key steps to validation are performing inoculation studies and doing parallel testing of non-sterile materials. However, parallel testing occasionally caused problems. For example, due to the biological nature of non-sterile material, there is a potential for uneven distribution of low-level indigenous flora.
We did some research and found that Dr. Reilly had already published a document that addressed those issues. If we hadn’t done a thorough review of the literature, we might have duplicated work that had already been done, unnecessarily. Research is essential to supporting a strong scientific case to present to senior management.
Ph.M – What is the most important aspect of the validation studies?
R.B. – Setting acceptance criteria is extremely important. With ATP Bioluminescence, a baseline must first be established, based on non-microbial ATP. A reading two times that amount or greater should be considered a positive bioluminescence result; other levels (i.e. less than two times the baseline luminometer measurement) would mean that the material has passed testing and can be released.
Extensive studies performed by Celsis support the calculation (two times the baseline luminometer measurement) of a positive bioluminescence result. However, in-house validation work needed to be performed to analyze, interpret and confirm this acceptance criteria.
All that information must be spelled out clearly in your documentation. Each product should undergo this validation process, and it should be written up in all summary reports submitted to management for review, including reports for QA and in-house regulatory specialists.
The question of how to set these criteria is an obstacle for some companies. People usually wind up saying, “The old way works.” Some also worry that with new, more sensitive methods, they’ll get false positives.
Ph.M – But what do you do when you get a positive result?
R.B. – We established a two-tier approach. If you get a positive RMM reading on a sample, you revert back to traditional testing methods for that specific sample.
Ph.M – But doesn’t reverting imply that RMMs aren’t that effective?
R.B. – Actually, the two-tiered approach with RMMs is extremely effective. 95% of the time our products tested negative for contamination and we could release those products within 18-48 hours. For the 5% that we found a positive reading, we’d need to revert back to traditional testing methods for end product release. This way, we only needed to delay release on a very small percentage of our products.
Ph.M – What additional in-house tests were required to support validation documents?
R.B. – We generated test data after we had a solid validation plan in place, and we also ran additional support studies, and in this case, it was important to keep an open mind. For instance, why use TAT and not Latheen? One could reference literature, but we also performed tests to demonstrate whether the two broths exhibited different levels of organism recovery.
In addition, mold studies were extremely important. Although they weren’t required in addressing questions of potential risk, by doing these studies we were able to demonstrate recovery of mold within 48 hours.
Other support studies looked at incubation time and temperature, and controls. Additional studies provided relevant support for our RMM initiative.
Ph.M – What would be a realistic time frame to expect for a project like this?
R.B. – You meet with FDA to discuss an initial proposal and you have a strong validation plan in place, approval is possible in a little over a year. But success requires ensuring that technical and regulatory professionals on the team are communicating well and often, and that even minor issues are addressed proactively before they can snowball. The keys to success are ongoing communication and establishing a dialogue with FDA.