By Brent Conatser, Senior Consultant - Regulatory Surveillance and External Engagement, Elanco
At the FDA/Xavier PharmaLink conference at Xavier University in Cincinnati, Ohio, in March 2017, FDA National Expert Investigator Robert Tollefsen presented a case study on a pharma company with compliance problems, and led a group discussion with the audience on what the appropriate actions would be for the company to undertake, given the circumstances.
The presentation took place as part of a panel discussion that also included FDA Cincinnati District Supervisory Investigator Dell Moller, and Elanco QA Compliance and Auditing Regulatory Surveillance and External Engagement’s N. Brent Conatser. At the same session, Moller gave a presentation on the top ten FDA inspection findings in FY 2016.
The Case Study
The case study used a fictitious company “Always Correct Pharmaceuticals,” a manufacturer of non-sterile prescription oral and topical solutions and suspensions. Tollefson said that the issues were based on real-life inspection findings, but that Always Correct Pharmaceuticals is not an actual drug manufacturer and is only used for demonstration purposes for this case study.
A previous inspection at the firm revealed, among other things, the firm’s regular use of trial injections of samples, which was allowed by the firm’s SOP. The trial injections are HPLC injections of samples performed prior to official analysis. This allows test failures to go unreported and not investigated. The inspection was classified as Voluntary Action Indicated (VAI).
Another recent inspection—also resulting in a 483—cited the following broad categories, with the 483 language describing the detail of the citation in quote marks:
- Failure to adequately investigate the impact of the performance of trial injections of samples, noted during the previous inspection, on product quality (21CFR 211.194(a))
“You only reviewed data for 12 batches representing seven finished drug products produced at your facility during 2013 and 2014. Over the same period, you manufactured more than 1,100 batches representing 34 drug products, but you did not include the majority of these batches or products in the assessment.”
- Failure to provide reliable data to the agency, including the firm’s use of a secondary Internal Audit CAPA system (21CFR 211.22(d)).
“The practice of conducting some CAPA under the Internal Audit program effectively concealed these activities from oversight by both Quality and the FDA. All corrections to cGMP related issues that are identified during an Internal Audit must be managed under the existing CAPA system.”
- Failure to submit a Field Alert Report within three working days when a USP Drug Product batch was found to be OOS for assay at the 3-month stability test interval (21CFR 314.81(b)(1) (NDA), 21CFR 314.98(b) (ANDA))
Management explained this investigation unintentionally “fell through the cracks.”
- Failure to keep up with workloads as evidenced by the firm’s failure to analyze in-process samples prior to release and the common practice of using the deviation procedure to manufacture products at risk (21CFR 211.110(c), 21CFR 211.110(a))
- Failure of the firm’s management to account for material, including a rejected batch of raw material (21CFR 211.22(d))
- Deficient Out-of-Specification (OOS) investigations (21CFR 211.192)
“Investigations were not thorough, timely, or based upon scientific rationales and did not adequately determine root cause. Typically only re-test results were reported as part of batch release decisions and the original results were not reported or considered in evaluating the quality of drugs for release.”
- Failure to adequately validate the manufacturing process for one Drug Product (21CFR 211.110(a))
“Raw material specifications for one raw material were inadequate to assure the drug product would meet final release specifications. The firm performed small scale production of finished products using various batches of this raw material in order to release raw material batch for use. This practice was not a defined specification and FDA had not been made aware of this unique practice.”
Regarding the firm’s 483 response, FDA said that the responses “were incomplete, lacked supporting documentation, and did not fully address all issues.”
The participants at the conference were divided into small groups, and allowed time to discuss the 483 findings and formulate answers to four questions.
Discussion question #1: What is the potential significance of performing trial injections with samples?
One participant commented that performing trial injections with samples generates OOS (out-of-specification) results for the finished product, which are not investigated. Also that the results could show that there is a manufacturing problem. “The data may no longer be available if it is not on the audit trail, so there is no way to have lab management or quality review it.”
Others said that the practice allows the analyst to preview system suitability without documenting it, and that it helps to reinforce bad practices in the lab and circumvents procedures for quality assurance oversight.
“What are they thinking?” asked another attendee. “What is prompting them to do this? Are they questioning the validity of their methods? The qualifications of the lab? The product validation? If this is happening in the lab, it is happening in operations, engineering, EHS, everywhere. It is part of the culture. What is the message that is being communicated?”
James O’Reilly, GMP Law Professor at the University of Cincinnati commented, “But there are generations that are changing. And as the old fart’s generation retires, and the young Millennials are taught on the way out the door to do things this way, the new generation will be cultured to doing things this way, and it is going to have serious long-term consequences.”
US WorldMeds Sr. Quality Director Bill Webb commented, “I am not disagreeing with you at all. You are thinking of normal people working at 11:00 on a Wednesday doing this test. They are going to follow it. But if you have an analyst working at 1:00 in the morning performing this test, other people are out, he is trying to set it up, doesn’t know exactly how to set it up. So, not thinking, he says, ‘OK, I am going to try this and see if it works.’ That is not an excuse, but a scenario where it could happen.”
Tollefsen noted that the firm had been previously warned not to do trial injections. “If somebody is doing something like you described, that is a legitimate concern—that is a training issue. They should know not to do it. And when the supervisor comes in and sees what they have done” he should point it out and have retraining done. “A quality system should have checks and balances. The supervisor needs to know that it occurred, and then can implement appropriate corrective measures.”
Webb further commented, “I understand the point, but the reality is that situation is more than just training. For example, if the person has made the error, or made errors before, he may be under some kind of HR plan. ‘This one is the one that is going to take me out of the building.’ Or, ‘this is the only workaround I can come up with. I have had this conversation with lab management. This is the only way I can determine whether or not this batch will potentially fail.’”
“The critical point for the analyst at that point is, if that fails, everything now stops. ‘Everything is focused on me. I have to explain why it failed with all of the pressures that are associated with that. From my history, it is more than the training.’”
“The lab management has to give that person support,” Webb continued, “and say, ‘it does not matter what the outcome is. We are going to support you. But you have to do it right.’ Without that, the analyst is left on their own to determine what they need to do to survive. If he is not getting that support and thinks that he is going to get thrown under the bus, then if he has to work around, he will work around.”
A participant commented, “Validated methods are the key. If the methods are validated, there should never be a reason to inject a sample. The very fact that they do inject a sample, for me, being a chemist, and auditing chemistry laboratories for years, the only time a chemist is going to do it is because they think that it might not work. If that is the case, then that is a big issue.”
Another person commented that some analysts think they can do it because they have always done it this way.
FDA’s Moller commented, “This is a pattern. You were warned last time on a 483, and you are still doing it. All it takes is one system to warrant an OAI classification. This, for me, would be recommended as an Official Action Indicated for the lab system only, even if everything else is OK.”
Moderator King & Spalding Lead Quality System & Compliance Consultant Steve Niedelman asked Elanco’s Conatser for an industry perspective.
Conatser commented, “I would venture to guess that most people in here at one time may have received a 483 from the FDA. You send your response. And I am pretty sure that most people track those commitments that you have made to the agency. And you put those very visibly so that you do not let those go undone.”
“The worst thing you can do is make a commitment to the agency, and the agency comes back in. They trusted you that you could fix this the first time. They are going to check these things first. If you did not do what you said you were going to do, that does not start your inspection off on a good foot. That does not give the agency a great feeling that you are capable. Do that. Make it visible. Keep track of it. Always meet your commitments.”
Complete Mastery CEO and President Jim Wulfeck pointed out that “We have been listening to this for 40+ years – the ‘why’ question, in particular. What I have learned from other people is, number one, if you have the system in place, but not in use, and you really think that it is a performance problem, not a system problem, then why isn’t somebody doing what they are supposed to be doing? Ask that question.”
“There are three primary reasons: First, look at the reward system. Are they being rewarded for doing it? Or are they being rewarded for not doing it? A lot of the root cause of the problem will actually be in the reward system itself, when you really break it down and do the cause analysis.”
“Number two: Do they know why, what, when, or how? If they don’t, then we need to train them. That is a proper corrective action. If you are training when the person knows the why, what, when and how, you are wasting your time and money.”
“But they still might not be doing what they are supposed to be doing because they have what we call a ‘thinking problem.’ They think that they are already doing it. Or they think that their way is better and this way won’t work. Or they think that something else is more important. Or they think that they cannot do it. That is not a training problem. That is a coaching, counseling problem.”
Discussion question #2: What is the impact for all of these cGMP issues?
One participant commented, “Of the four systems inspected, it appears that all four failed. There will most likely be a warning letter in their future, since this is the second inspection. We wondered if the company would actually take the initiative to recall the product.”
The spokesman for another group said, “We believe this is a hot mess. There is a high probability that there will be a warning letter. Either the quality unit does not know what they are doing, or they are actively trying to cover something up—hiding the information with regard to the validation, and what appears to be hiding OOS and CAPAs in the internal audit program. It could be fraud.”
Another commented, “The management of the company does not appear to understand GMPs – their purpose, their intent, or even their value. If the management is clueless, there may be hope. If they are not clueless, there is probably not.”
FDA’s Tollefsen interjected, “I don’t care whether they intended to do it or not, the facts are the facts. The facts do clearly suggest that there was intent – but that is a lot harder threshold than I have time to try to establish. It is certainly something that you could ask for.”
“Keep in mind that FDA does not need to show intent – it is strict liability,” King and Spaulding’s Niedelman pointed out.
Moller recounted that he was in industry for ten years, and his firm received 483s. He emphasized that FDA “understands the impact of what we are giving you. In our training program we put a lot of emphasis on making sure that you have the right observations. The last thing I want to do is call something wrong if it is right. I also do not want to call something right if it is wrong.”
Niedelman again requested an industry perspective from Conatser. “Brent, from an industry perspective, how would you instruct them to address these GMP observations?”
“We have been talking this week a lot about culture,” Conatser reminded the participants. “In my opinion, this is a cultural issue, where they are just not doing what they are supposed to be doing. And it has been perpetuated throughout the company. It is going to be much more than rewriting and retraining. The firm has to be taken apart and put back together the right way.”
“One of the things that I have been seeing in warning letters more recently is ‘you have a poor culture of quality at your firm.’” They can retrain, rewrite SOPs, build a new plant, but they will not be different with this culture. “Until you change the mindsets and people want to do and are expected to do the right things, this is going to keep happening. I think this is a big job. It is not an easy fix. It is a long-term fix. It is a long-term commitment. And firms that want to stay in business are going to have to be willing to commit the proper resources to doing it and doing it right.”
FDA Ombudsman Jessica Zeller noted, “Ultimately it can impact the patient and your brand, and could end up in litigation.”
Discussion question #3: What would you need to do in order to properly address the cGMP findings?
One participant suggested that the firm open a CAPA in the CAPA system. “We need to understand the impact to the product out in the field. The firm needs to conduct a risk assessment to determine if the product is at risk. Is there an adequate number of Quality personnel? Are they trained appropriately?” The firm needs to look at the test injection process, the OOS process, and the validation process, just to begin with, he emphasized. “They need to get a baseline assessment and bring in an outside consultant for a few weeks to do an evaluation.”
Suggested by another participant was to have the consultant look at all the lots and batches, the procedures, and the validation processes. The consultant should also look through the internal audits and pull out CAPA items to review, and help develop a plan for adequate staffing and training of the staff.
Another said, “I think there are systemic issues and I do not understand why the internal audits are not picking up the issues.”
King and Spaulding’s Niedelman pointed out that internal audits are the opportunity to find problems and fix them yourself. “It is a critical process that is often undervalued. Sometimes it pays to bring in outside people for the internal audits.”
FDA’s Tollefsen emphasized, “I consider the QA unit at your facility to be my eyes and ears. I expect you to be doing the job so that when I come along I do not have to write a 483. You have to have the systems in place and they have to be robust. What I have seen is that the internal audit programs at a lot of companies tend to be very weak...I should not be the first one to find anything. You should know everything that is going on at your facility, and all of the problems.”
FDA’s Dell agreed. “When I look at annual reviews, I look at what you are covering. Are you trending? Are you trending from line to line? Are you trending from facility to facility? Are you trending from lot to lot? Give your internal auditors enough time and resources to do their job the way that they want to do it. They want to find the issues so that we don’t find them.”
Niedelman stressed the importance of making sure that the people brought in from the outside are qualified to perform these tasks.
An audience member pointed out, “Sometimes the company is not a bad company. It is a good company that has made a big mistake. In your organization there may be someone making a big mistake while you are sitting here. You do not know what is happening all the time. But you do not want to be blind when it happens. You want to go back and ask yourself if it could happen to you, and how you find out, and what do you do about it. So don't think that we are talking about bad companies and yours is not.”
Discussion question #4: Do these findings affect your trust in this firm?
The final question for discussion in the case study exercise was, “Do these findings affect your trust in this firm?” There was general agreement among the participants that they had no trust in the firm as portrayed in the case study.
The issue of trust—and the depth and breadth of the problems that can arise from a lack of trust by FDA—was addressed by former FDA official John Taylor.
Taylor is President and Principal, Compliance and Regulatory Affairs at Greenleaf Health. He was in leadership roles at FDA for 20 years. From 2009–2014, he held three high-profile positions at the agency: Counselor to the Commissioner, Acting Deputy Principal Commissioner, and Acting Deputy Commissioner for Global Regulatory Operations and Policy.
“Both Bob and Dell emphasized that their job is to go in and make observations. And certainly they will draw their own conclusions regarding the trust issue based on their findings and what is on the 483,” Taylor commented.
“But remember that there are others in FDA and throughout government who are reviewing these findings and making their own trust judgments, like the compliance offices and the attorneys at the Office of the Chief Counsel (OCC). To Brent's point, you can regain some of that trust by doing a good job with your corrective action plan and remediation plan. That is the way to claw some of that trust back.”
The effect of not doing that is that the trust issue starts to be more severe, Taylor asserted. “And it is not just trust by FDA. There are others who are starting to be more interested in this information. The Department of Justice (DOJ) is starting to focus more on quality issues now. The uptick in False Claim Act cases has to do with quality. Payers are paying more attention to quality... These warning letters can kill an acquisition deal. The analysts are paying more attention.”
My point is that it is not just trust of any one individual that you are trying to address and restore—it is institutional trust. And there are a lot more players than just the FDA people you are seeing in this room. There have been dire situations in which DOJ has been actually looking at whether or not the trust issue is suggestive of bad intent, which is the difference between merely a civil problem or a criminal problem.”
Taylor concluded, “Being able to show that you have done a good job in your remediation, that you have made commitments and actually followed through on them, might be the very thing that keeps that case from being a criminal case.”
Others experienced in the industry also weighed in on the topics of trust and company culture.
James O'Reilly, University of Cincinnati Professor of Law and Public Health – who has been involved with FDA issues since 1974 – pointed out that in recent years, “Firms that have been bought by venture capital firms or hedge funds get the message that they need to get product out the door, keep the numbers up, and keep the share price up. They are facing culture wars between new and old owners.”
Speaking from an FDA investigator’s perspective, Tollefsen said, “When I walk into a facility, I am going to know in the first day whether we have a [trust] problem or not. I would cite, for example, the crowd of people who are always following the FDA investigator around. I am paying attention to how people are interacting with the senior managers. The quality assurance manager is always there. The regulatory affairs manager is always there. And I have been at firms where the quality assurance manager is literally disrespected in front of FDA. Or they have never seen this person and do not know who it is. That is a scary thought, because they get responsibility and authority from people recognizing what your authority is and what you can control.”
King & Spalding Lead Quality System & Compliance Consultant Steve Niedelman commented regarding occasions when he has been called to testify in court, and the difficulty of defending repeat compliance findings.
“Regarding John's point about expanding beyond FDA, I have reluctantly testified as an expert witness on a few occasions. It is always to defend the industry. But it is really tough to defend the industry when you have had repeat 483s. Try to explain to an attorney why that is not significant. Or how a warning letter that says you are operating in violation of the Act is not significant. These create patterns beyond the FDA realm in trying to defend any lawsuits that you may have out there as well. So it does expand. And to John’s point, the False Claims Act has really been growing. It is big money.”
Complete Mastery CEO and President Jim Wulfek—a student of the late W. Edwards Deming—suggested reading the 2008 book The Speed of Trust, by Stephen M. R. Covey, son of noted author Stephen R. Covey.
“Covey talks about 13 behaviors. In the real world, if I have taken the time to build up ‘deposits of trust’ to these different behaviors over time, if something happens—an isolated incident or whatever—that trust will not necessarily be destroyed by that one incident. It will take some other things.”
“The first program I wrote in 1976,” Wulfeck recalled, “was ‘Integrity and Trust – Our First Concern.’ That is how important these elements are. We have to work hard at trust as an industry and as individual companies... I would urge you to take a look at what Covey has written and apply it to your particular organization.”
FDA’s Moller tied company culture to the character of the individuals in the organization.
“Thoughts lead to actions. Actions lead to habits. Habits define your character. If you want a quality culture, have a company full of character.”
There was general agreement among the participants that they had no trust in this firm as represented in the case study. Comments included:
- “We have no trust in this firm. They pick and choose data.”
- “Arrogance and not understanding by senior management is the hardest to deal with, and the culture that flows from that.”
- “They are concealing investigations. There are systemic issues and cultural issues. It gets to the essence of GMP: Are they operating in a state of control? We do not trust this firm. There is no integrity, and not just of the data.”