CMC GUIDANCES ON ROLLING SUBMISSIONS,

CMC GUIDANCES ON ROLLING SUBMISSIONS, pre-marketing meetings, and IND information are being considered at FDA. The three-tiered guidance approach to jump-start the NDA/BLA review process was proposed at a recent FDA/industry workshop on CMC filings. The workshop focused on the challenges involved in refining a "rolling submission" system under which firms would submit "reviewable units" of CMC data to FDA prior to completion of the full NDA/BLA. In view of the workshop recommendations, FDA is also considering ways to make pre-marketing CMC meetings with industry more fruitful and to clarify IND filing requirements at Phase II/III.

WORKSHOP BALKS AT EXPANDING IND

FDA is evaluating a "rolling submission" as a means to expedite the review process for the chemistry, manufacturing, and controls (CMC) portion of marketing applications for new drug and well-characterized biological products.

ä A guidance defining procedures for a rolling CMC submission could be one result of the FDA/industry search for more efficient NDA/ BLA reviews.

Consideration by FDA of a more formal rolling submission process was endorsed at a workshop cosponsored by the American Association of Pharmaceutical Scientists (AAPS) and FDA in Arlington, Virginia in December. FDA and industry representatives gathered at the conference to explore the feasibility of developing new procedures and guidance for expediting CMC reviews.

The workshop participants agreed that the IND may not be the appropriate vehicle to provide additional CMC information to the agency for review prior to the submission of the full NDA or BLA. That sentiment was voiced in response to a "preliminary working draft document" presented by FDA to clarify agency expectations for filing CMC information in INDs at the Phase II and Phase III stages of clinical development.

Participants felt that the FDA proposal potentially blurred the distinction between the safety concerns inherent to the IND under section 312 of the federal regulations and the CMC component of the marketing clearance process which is regulated under CFR section 314. The rolling submission concept emerged as a potential vehicle for keeping the distinction in tact.

Under a rolling submission system, firms would submit "reviewable units" of CMC data to the agency prior to the completion of the marketing application.

ä FDA New Drug Chemistry Director Eric Sheinin discussed how the original purpose of the meeting - to develop IND Phase II and III guidance - evolved into the idea of a rolling NDA/BLA submission.

"We were talking about developing a guidance on information that should come in during various phases of the IND," Sheinin said. "Now we are talking about perhaps moving that into the NDA - pre-submission of the NDA."

Instead of focusing only on an expanded IND guidance, the workshop has led FDA to consider developing three separate guidances on the handling of CMC information during the drug development process. The three documents would potentially cover (1) the submission of Phase II/III information directly related to product safety in INDs, (2) CMC-related meetings during the various stages of application development, and (3) rolling CMC submissions for NDAs and BLAs.

"So I guess the bottom line is what we really need to address along with safety is this rolling submission concept," Sheinen said. "Maybe what will come out of this is not one guidance, but two or three guidances - one on the safety aspects, one on the rolling submission…and perhaps some guidance on the CMC-type meetings that were discussed over the last few days."

ä [EDITOR's NOTE: The three topic areas were discussed in break-out sessions during the three-day AAPS/FDA workshop. The box beginning on p. 13 contains summary points from the sessions on CMC-related meetings (pre-IND, end-of-Phase II, and pre-NDA). Workshop wrap-up remarks by meeting coordinators from FDA (Sheinin) and industry (Lilly Director of Regulatory Affairs Tobias Massa) begin on p. 4]

Sheinin explained that the new FDA guidance potentially would concentrate on the criteria and procedures for pre-submission of the CMC information rather than on the content of the submission. Content requirements are spelled out by the existing NDA CMC guidances such as those on the drug substance, drug product and packaging. "A number of these are outdated," he acknowledged, "but they are being updated."

Many Issues Confront FDA About Rolling CMC

While the rolling submission concept was recognized as a potential vehicle for expediting the CMC review process, both industry and FDA representatives at the workshop pointed out a variety of issues that would need to be worked out before a system acceptable to both sides could be developed.

ä Center for Biologics Evaluation and Research (CBER) Associate Director for Policy Rebecca Devine discussed FDA's concerns regarding the proposed system.

Devine commented that it is imperative for industry and FDA to not lose sight of the overall purpose of the rolling submission. "The whole point of the pre-submission," she said, "isn't the continuation of clinical trials [and] isn't for the review of safety. It is to try to do something ahead of time to get the application approved at a later date when it comes in in its complete form."

ä The agency is not sure that every product type would qualify for the rolling submission. "I think from the regulatory standpoint," Devine maintained, "it is probably fair to say that [rolling submissions] are not going to be used for everything."

To truly "capitalize" on the system, Devine suggested, FDA will want to "use it in those situations were it is going to make the biggest impact for both the public health and getting things out there quickly." Potentially, the rolling submission may be limited to fast-track, or priority drug products.

There also could be a benefit in extending rolling submissions to complex products. "For example," Devine explained, it could be used "where there might be particularly complicated microbiology issues [and] biopharmaceutics issues [or] where there might be a combination product which involves intercenter collaboration."

In such cases, the rolling submission would give the agency "earlier availability" of the CMC information with the intent of preventing a delay later on in the application clearance process. "In other words," Devine clarified, "many of the complicated issues would be worked out" prior to the submission of the completed application.

ä Certain tests were also identified by Devine as possible candidates for a rolling submission; for example, an assay "that might be very important for evaluating product, and you want to make sure that all of those issues are worked out before the NDA or BLA comes into the agency."

Before FDA can determine which product categories will qualify for rolling submissions, Devine maintained, it "will have to come together and decide what the definition for all of these types of products (fast-track, priority) is and consolidate our regulations." The regulations currently provide for expedited development and review of products to treat

serious to life threatening conditions and those that address an unmet medical need.

Current regulations (CFR 314.50 & 601.2) do allow pre-submissions to NDAs in some circumstances. Under this framework, both centers have accepted parts of an application prior to the completed submission in the past. Prior experience will influence FDA in refining the rolling submission concept. "Some of our biases," Devine said, "are going to be based on that previous experience."

PDUFA Pressure Could Stymie Rolling Submissions

The limited resources and time available to Center reviewers also will be a major agency consideration in framing the guidance for rolling submissions.

FDA will be cautious in placing additional burdens on reviewers who already are saddled with compressed performance goals, instituted under the first and second installments of the Prescription Drug User Fee Act (PDUFA I & II).

"We have many, many performance goals and we are trying to do our part to succeed - and we have succeeded so far in meeting and exceeding all of the PDUFA goals," Devine explained. The concern is that, by accepting pre-submissions, the agency could be "biting off more than we can chew, and then not be able to meet performance goals."

FDA is concerned that it not loose industry confidence by failing to keep within strict PDUFA timelines. "PDUFA is the top priority," Devine asserted, "so if a rolling submission were to come in, then one of the issues is, where does it fall in the queue, when will we look at it? And it will probably be as resources permit. It would not be a top priority."

ä In light of the PDUFA deadlines, FDA would want to ensure that valuable review time is not wasted on pre-submissions for products that fail to lead to the submission of an NDA or BLA.

To help prevent such situations, Devine stressed that FDA is willing to work with industry to identify reasonable milestones that help a firm determine the appropriate point in the development stage to submit a reviewable unit.

From CBER's point of view, the use of milestones is especially important because biotech firms experience more clinical delays and failures than their chemical drug counterparts.

Devine described the type of milestones that the agency will be considering. For example, she suggested that firms might be required to wait until interim results are available from pivotal trials before submitting a reviewable CMC unit. Another milestone could be the point at which firms have "no more anticipated changes to [the] CMC" pre-submission data.

While willing to entertain the concept of a reviewable unit, FDA will need industry's input in creating a functional definition. "I have to tell you," Devine said to the audience, "FDA has not decided what a reviewable unit is. We don't have any preconceived notion necessarily of what it should or would be."

Reviewable Units Should Be Complete & Independent

No matter what type of CMC data are eventually targeted for pre-submission, the reviewable unit would have to be complete, independent and not likely to change, Devine stressed.

She defined "complete" as meaning that reviewers will "have to be able to make some determinations or decisions about this package of information that you got in order for it to be useful."

The reviewable unit should be independent of the other CMC and marketing application data. "One thing I do know," Devine asserted, "if I get a reviewable unit and I have to look at it, I have to be able to look at it out of context without the rest of the application."

ä Devine's concern was echoed in the workshop breakout sessions where it was noted that it is difficult to evaluate information such as that on impurities without considering the toxicology or medical section of the larger application.

The CBER official added that the pre-submission should not need to be re-reviewed later in the NDA/BLA clearance process. "I think from our standpoint," Devine maintained, "if you have to re-review it later, you have not gained much. That is a very important determinant of whether or not [the reviewable unit] is useful."

A key factor in limiting re-reviews is that firms understand that they should not submit a reviewable unit if they intend to make changes to it prior to the completion of the NDA/BLA review.

"There needs to be some certification that changes will not be made" to the [continued on p. 9]

FDA & INDUSTRY ASSESSMENTS OF AAPS/FDA CMC WORKSHOP
	The following is a transcript by "The Gold Sheet" of concluding remarks delivered by Office of New Drug Chemistry Director Eric Sheinin and Lilly Director of Regulatory Affairs Tobias Massa (then with Schering-Plough) at the AAPS/FDA workshop on the CMC submission process held in December 1997. Massa, who was co-chair of the program planning committee for the AAPS/FDA CMC workshop, has the responsibility for reviewing the technical committee reports from the various meeting breakout sessions and consolidating them into an overall report. Massa's consolidated draft will, in turn, be submitted to the FDA co-chairs of the meeting, CDER Division of Chemistry I Director Charles Hoiberg and CBER Office of Therapeutics Associate Director of Science Gene Murano, for further review. The final consensus report will help provide a basis for future FDA deliberations and guidance on the CMC review process. In his concluding remarks at the workshop, Massa referred to an outline prepared by FDA suggesting CMC information that should be submitted in INDs for Phase I, Phase II and Phase III/pivotal studies. FDA presented a "preliminary working draft document" to the workshop for consideration as a basis for a formal guidance on the subject. The third column of the FDA chart, which addresses CMC information at the Phase III/pivotal study stage, states that "a full description of the physical, chemical, and biological characteristics of the drug substance should be provided." Examples cited include neutralization equivalents, solubility, hygroscopicity, crystal properties/morphology, particle size/surface area, stereochemical considerations, and biological activities. Sheinin, as co-chairman of the CDER CMC Coordinating Committee, provided an initial assessment of how the conclusions and recommendations that emerged from the workshop might impact the refinement of FDA policy and guidance related to the CMC component of NDA/ANDA submissions. Sheinen commented specifically on the success of the workshop in meeting the "goals and objectives" established by the workshop steering committee. These were to: "n n improve the quality of the original NDA/BLA submission which can facilitate CMC evaluation; n n present a proposal and guidance on utilization of measures to facilitate drug development, end of Phase II meetings, pre-NDA meetings and other strategies in an effort to improve CMC NDA/BLA filings; n n Present for open discussion current FDA and industry thinking on the concepts for CMC IND information as it relates to Phase II and III studies for drugs and biotechnology-derived products; n n provide an opportunity for interactive discussions and feedback with industry participation; and n n prepare a summary report of the discussion which can aid in finalizing the guidance."
CONCLUDING REMARKS BY INDUSTRY'S MASSA
I am going to try to give an overlay to what we have heard today and where we think we are. This started out as an exercise in trying to determine what parts of the CMC should be reported into what phase of the IND. When it started out we had the chart [prepared by FDA on IND CMC submissions] that was given to everybody as part of the package. When the industry group that was involved in the early discussions saw that and we looked at column three, we said why are we doing this? We are not really sure what you [FDA] are trying to accomplish. Is this a safety related issue or is there some other reason why this information is being requested? As [FDA Division of New Drug Chemistry I Director Charles Hoiberg] pointed out on day one, there is a lot of latitude in [CFR section 312] in terms of what should be submitted and when. Things like "essential new data" and the "amount of data should be commensurate with the phase of development" leave an awful lot of room for interpretation. And by the way, there is nothing that says that as a result of this conference, we can't go back and rewrite 312 at some point, although that takes a lot longer.... I think what we started out with was that FDA's priority concerns, especially during Phases I and II were safety, and how do we define what safety is? What parameters in CMC contribute to safety concerns? And that is what day one was really all about. So the feedback we got back was that this was kind of soft and fuzzy. We don't really have a good handle on it. And maybe some of that was our fault, because we didn't really make it clear what we were trying to do and what information we were trying to get back. But one thing that was clear was that, almost to a person, everybody felt that, regardless of the phase of study, if something was a safety issue in Phase I, it was a safety issue in Phase II, it was a safety issue in Phase III. And if you submit information in Phase l and say, "this is the critical information we think you need to have," if you make changes to any of those parameters during the course of development, that requires an information amendment to FDA so they know what we are doing. I think FDA's concern is maybe that those updates would not be coming, and that, if you look at the information that is in that Phase I CMC document, that that is all the information that will come forward. Certainly that is not what industry had in mind. We had intended that information amendments would come forward. I think we have fairly good agreement on what needs to be there from the 312 standpoint. Keep in mind what [Guilford Senior VP for drug development] David Savello said yesterday - that we need to look at what do we need to satisfy the requirements of 312 vs. what do we need to satisfy the NDA requirements of 314. So I think we have a basis of saying that at least that part of the guidance document probably can go forward with some of the inputs that we have provided. Once you get to things like the end-of-phase II meeting and what do you do with all of that information that is in column three, I think that is still where we have a little bit of debate. FDA's concern certainly is that they want to know that the material that is tested in our pivotal clinical trials, be that Phase II, Phase III - Phase I/II/III knowing how some of these clinical programs get compressed - that that material is representative of what we test commercially. And if it is not, how are we going to get from whatever we tested in the pivotal trial to commercialization? They want to know, have we got the product adequately characterized when we do our pivotal clinical so that we know what the differences are between the pivotal material and the commercial material. And how do we get that information to them? What level of detail do they need to have in order to make that assurance? Overlaid on that are all of the PDUFA regulations and timelines that FDA has to deal with. We have heard both from the speakers as well as people on the breakout groups: there is a ten month review clock for standard review, there is a six month review clock for priority review. And although it hasn't been mentioned, there is a four month review clock for CMC supplements that require prior approval. Add to that meeting timelines, response timelines, and all of the chemists that have been here and all of the FDA people that have been here have said, PDUFA guidelines are going to come first. FDA has a certain number of finite resources right now. Those resources certainly are not going to expand in the future and may shrink in the future. Even if they don't shrink, there will be additional INDs being filed. We know that the rate of IND filings is increasing. So the workload at FDA is going to be ever increasing. In the face of that, how do we work out a system with FDA that allows them to have the comfort level that the material we are testing is well characterized, particularly in Phase III, and balance that against their limited resources of being able to review information. Some of the comments we have heard are that IND amendments don't necessarily get reviewed to a high level of detail unless there is something that really jumps out at the reviewer. All of us have filed annual reports to NDAs and maybe two years later you get a comment back. Part of that is because of workload. And it is not meant to be a criticism of FDA. It is the reality that they just can't deal with everything that they are being sent. So the question becomes, from a pragmatic standpoint, what do we submit to them, when do we submit it, and how do we submit it? We had some rather interesting debate over dinner last night. I think it helped us understand a little better what FDA is trying to obtain when they ask for all of that information in column three. Part of that is a perception on their part that is it safety related. Because you gain more critical information about the drug substance and the drug product during Phases l and lI, and some of that information may be critical for the drug to have certain safety and efficacy characteristics. We understand that and we agree with that. The question becomes, what do you do with those data? If FDA comes to us as a result of an end of phase II meeting and says we would like additional characterization done, or we would like X, Y and Z done, that may be done on that drug substance prior to the NDA filing, but the reality is you are going forward with your pivotal clinical trial with whatever you have at that point in time. You may have to do some "bridging study," and I shudder to use that term based on our conversation yesterday - you may establish I think a preferable term - you will establish some sort of comparability protocol to show that that material and your commercial material are comparable to one another. But the reality is that you are going forward with whatever you have available at the end of Phase II. So how is that information going to be used? I think that is something that still needs to be worked out. I don't think we are that far away, but I don't think we have adequate resolution on how we need to work this out. One of the things that I think we are going to propose as a result of this meeting is that this guidance document not be published in the format that it is currently in, or at least in the scope that it is currently in. I think we feel that there is probably enough good information that the 312 part of it - what do you submit when based on the feedback that you are going to hear about today - can go forward and can go to completion. But there are two things I think that need to be said. First of all, I don't think that I feel that we are where we want to be with regard to end-of-Phase II and all of that information that is in column three - I don't want to call it Phase III information, but the information that is in column three. I don't think we have quite resolved how that information ought to be submitted and how it should be used. I think we can probably get two additional guidance documents out of that. One is end-of-Phase II and NDA. The other: FDA expressed concern, that for the larger companies, people who have had an awful lot of experience with the NDA process, we know pretty much how things have to proceed as we go through an IND and get into an NDA. We know - well we think we know, sometimes FDA thinks we don't know - we have a lot of experience in bringing drugs from IND through NDA. FDA's concern is that some of the newer, smaller companies may not have that same level of expertise just because they haven't gone through the experience. There is concern that the guidance would have to cover not only the experienced folks but also the newer folks, so that they have an understanding of what it is that needs to be done. I think our feeling is, that is kind of a separate guidance. That is a "how-to" guidance document, a drug development guidance document. In 1987 there were a number of guidances that came out that are in the process of being revised that talk about what do you submit in an NDA and how do you submit it, what is the format, level of detail and all of that stuff. Maybe that guidance document needs to be rewritten to accommodate some of the things that have been addressed here. So we are going to go forward and make those recommendations and certainly we are going to incorporate everything that we have heard here. I think even though we may not have gotten to resolution on where we started out with in this guidance document, we have collected an awful lot of good information. And if nothing else, there has been a lot of good interchange between FDA and industry, and I think we have a much better understanding of where each of us all on all of this. So from that standpoint this conference was a big success. So I think we should take this information and go forward from here.
CONCLUDING REMARKS BY FDA'S SHEININ
The literature that was distributed about this meeting [recognized] five goals, five agendas. How did we do in meeting those? The first one, to improve the quality of the original NDA/BLA submission: In my mind, when we developed this workshop we were looking at this rolling submission concept as information that would come in during the Phase III IND which then could lead, because of interaction between the agency and the sponsors, to improved quality of the NDA/BLA when it comes in. Now what we are hearing is, at least from the industry perspective, this rolling submission really applies to the NDA, to a pre-submission to the NDA. So have we done anything here that is going to improve the quality of that rolling submission? Maybe to some extent, but not to the extent that I was hoping for. If we are talking about a rolling submission being a pre-submission to the NDA, that is in fact part of the original NDA then. Because what we are looking at is reviewing it in that pre-submission portion and finishing with it. So we kind of addressed this, but maybe not totally. To present a proposal and guidance on utilization of measures to facilitate drug development and meetings and so on to improve NDA/BLA filings: I think we have done that. We have talked about all these types of meetings and how this can help to provide guidance to the industry that will ultimately improve the quality of those submissions, which in the long run is going to lead to quicker approvals. So we have accomplished that. We certainly have provided a forum for open discussion of FDA viewpoints, industry viewpoints on CMC IND information which relates to Phase II and Phase III studies. That is a win. We accomplished that. We had a lot of information that we are going to go back and sift through and evaluate. Providing an opportunity for interactive discussions and feedback with industry participation: You have had every opportunity possible to present your viewpoints over the last two and a half days. So we have accomplished that. To prepare a summary report: That we haven't accomplished yet. We have individual summaries of breakout sessions. Our goal is to have an overall report of this workshop which I believe will be made available to FDA and probably to all the attendees. So that will be achieved at some point down the road.... It is our preconceived concept as we came into this workshop that as you go through Phase I to Phase II to Phase III, that the amount of safety information that you should be providing to us would increase, because you would be learning more about the various components of the drug substance/drug product portion of the application. What we are hearing, for the most part, is that it is your perspective that what is a safety concern [at Phase l], is a safety concern [at Phase II], is a safety concern [at Phase III]. It is not really growing except to the extent that, perhaps as changes are made, there should be updates as you get more information that is related to the safety aspects.... We also looked at information that you as the industry have in hand but did not necessarily have to submit to us. This is the gist of the Phase I guidance and was continued into our draft guidance for Phase II and Phase III - that there would be some information that you would possess, yet is not necessary for you to provide to us until the NDA or BLA. This is even in the face of the regulations in [CFR section] 312 that talk about submitting really essential information. It was our hope to use this guidance to give our interpretation of that portion of 312 which provides some relief to the industry - that you would not have to submit every little piece of information as you learned it if it was not related to safety or...was not information that we felt could come in early - sort of a rolling type of submission to aid us to finalize more of the CMC portion of the application before the NDA comes in or the BLA comes in. But what we are hearing as a result of this meeting is, this...information that didn't have to come in immediately...really should not be part of the IND, it should be part of this rolling submission, or early submission, or pre-submission of the NDA/BLA document. So, as I indicated earlier, we are going to have to go back and reexamine the principles that are shown on this portion of this overhead. Perhaps, yes, this does have to move further out in the process. Maybe there is going to have to be an overlap - some information coming in during Phase III, some of it coming in as a pre-submission. Part of it is going to be timing. If we are going to be getting information in very early, you may not have an NDA number yet. I don't know what the regulations are or policy when you can get an early number assigned to your NDA that is going to come in...There is a process. When you do a pre-submission under the regulations now - they say if you come in 90 to 120 days before you expect to file your NDA, you get an NDA number. So it is actually coming in to the NDA. But if there is some of this information that we have discussed as part of the rolling submission that will come in earlier than that...and there is no NDA number, where is it going to be submitted? I think the only place that I can think of where we can accept it and review it is going to be to the IND that is going to form the basis for that NDA. So these are things that we are going to go back and look at.... The IND portion of drug development...before we get into the NDA...is going to continue to be compressed. What we are trying to accomplish is to find ways to handle this compression - still give you the guidance that in some cases that you need, and, at the same time, be able to meet our PDUFA goals and have quality NDA and BLA submissions come in. One of the things that you know...from listening to the summary reports: in our template in our draft guidance, we talked about what we would expect for the drug substance, what we would expect to see for the drug product. It is clear, I think, from the discussions at the first session this morning that we are going to have to look at various types of drug products on their own basis. Just even looking at the safety information - what is a safety concern for a solid oral dosage form is considerably different than what might be a safety concern for a metered dose inhaler. We are going to have address that in this guidance in whatever form we end up with. Moving more of this guidance, the pre-submission and rolling submission into the NDA phase - what came to my mind this morning [is that] we were talking about developing a guidance on information that should come in during various phases of the IND. Now we talking about perhaps moving that into the NDA, pre-submission of the NDA. Do we really need a guidance that tells you what should be in the pre-submission of an NDA? We have a drug substance guidance. (A number of these are outdated, but they are being updated.) We have a drug product guidance. We have a packaging guidance. The updated packaging guidance is out for comment and our packaging technical committee is now evaluating the comments. We have a draft stability guidance that is going to be out very soon for comment. We have a microbiology guidance for what is in the NDA. Do we really need to spend the time to develop another guidance that essentially sums up all of what is in the other guidances to tell you what should be in the pre-NDA submission? I don't think so. So if we are not going to address what should be in the Phase II and Phase III IND beyond safety, then our topic is going to be a lot easier. That is something that I think we are going to have to consider when we go back to the office.... So I guess the bottom line is, what we really need to address along with safety is this rolling submission concept. As we talked last night, maybe what will come out of this is not one guidance but two or three guidances - one on the safety aspects, one on the rolling submission, if there is going to be some of that rolling submission that is going to have to come into the IND, not wait for the NDA. And perhaps some guidance on the meetings, CMC-type meetings, that was discussed over the last few days. We heard some talk about resources and how can we accomplish what we discussed given our current resources, your resources and the PDUFA goals that we are going to be under. If we could get something started soon, even maybe without a guidance, getting some trial applications going where we could work together to do this - this is going to put some burden on you to get these into us in relatively short order if there is anything in the works. Our PDUFA goals this year are exactly the same as they were the last year. PDUFA 2 does not change anything in the first year of its life. So what was a 12 month standard review in fiscal 97 is a 12 month standard review in fiscal 98. The same thing with the supplements. The prior approval supplements are still six months for this year. Starting in fiscal 99 there will be a gradual phase in of those new goals. Just like PDUFA 1 had a gradual increase in the percentage that we were expected to complete within a certain timeframe, the four month reviews and the 10 month reviews will be phased in - I believe it is something like 30, 50, 70, 90 over the last four years of PDUFA 2. So it is not going to hit us all at once. If we can work with you and you can work with us to try to start moving some of this information earlier in the process, whether it is in Phase III or it is earlier than a normal pre-submission, we will strive to find some way to coordinate our resources to do this review. As was indicated, one way of possibly improving our resource situation - we are no longer doing a huge review in the Phase I stage...We had some Phase II INDs that were [very big] in the past. We are not reviewing those anymore and you are not submitting those. That is a freeing up of some resources. I indicated at least in one of the breakout sessions I was in yesterday, it is going to be a gradual process. It can't happen overnight that all of a sudden we are going to have these resources to look at early submissions. But if we can do it gradually, case-by-case basis, have one here, one here, one here - over a period of time, that is going to make the NDA portion go faster which will free up some resources which in turn will allow us to do more of this experiment, more of these trial drugs to see if we can do more earlier in the process. So it won't happen overnight, but it can be this gradual process [to] achieve this goal. The bottom line is that along with these guidances we need a rolling submission guidance if it is going to be a different concept than the pre-submission, which is just as I said - we already have guidances for various portions of the NDA.

reviewable unit after it is sent to the agency, Devine maintained. "Now, that is not to say that you can never make a change once we review something, but you might want to handle it more like you handle the supplements now. As opposed to resubmitting the whole thing, you submit more like a change supplement - so that you only give information about the change and what you have done to show that the change does not have an adverse effect."

ä Determining at which phase of development to submit reviewable units is another issue that needs to be worked out. Devine noted that CMC data most likely would have to be submitted following Phase II, since most of it would not be available before then.

Devine also expressed concern that industry might have unrealistic expectations of FDA's ability to turn around reviewable units in a certain period of time. With the resources and staff of the reviewing offices tied to user-fees, FDA would have to set practical targets.

To emphasize this concern, Devine elaborated on the workload faced by reviewers. They may be "asked to review an IND in 30 days, review a response to a clinical hold in 30 days, do a 10 month review on an application, and they have another application that is a six month priority, and they are getting ready for an advisory committee." The problem is sorting out "what to do first."

What it comes down to is "management," the CBER official maintained. She added: "How do we distribute the workload, and how do we forecast the types of reviewers that we need in the situations…that we are getting?"

"Reviewable Units" In Search Of Regulatory Home

Another challenge in pursuing the rolling submission concept is determining where the reviewable unit would fit into the existing IND/NDA regulatory structures.

"That may not be as important to [industry] as it is to FDA," Devine explained, "because we have an IND application and we have an NDA application and a BLA application. Those are separate applications that have their own lives, are filed in separate areas, and the IND might go away later after the NDA has been approved and has its own storage time."

In his wrap-up statement, CDER official Sheinin commented that the timing of the pre-submission could

be a problem. "If we are going to be getting information in very early, you may not have an NDA number yet. I don't know what the regulations are or policy is when you can get an early number assigned to your NDA that is going to come in."

ä Sheinin noted that under the current process for pre-submissions, firms are assigned NDA numbers at the earliest "90 to 120 days before you expect to file" the application.

However, he asked, "if there is some of this information that we have discussed as part of the rolling submission that will come in earlier than that...and there is no NDA number, where is it going to be submitted? The only place that I can think of where we can accept it and review it is going to be the IND that is going to form the basis for that NDA. So these are things that we are going to go back and look at."

Reviews Cannot Be Piecemeal

Another FDA concern is that the rolling submission process does not lead to "piecemeal reviews." Devine explained that she experienced such reviews earlier in her FDA career, which proved "very inefficient."

Devine raised other questions about the viability of rolling submissions: "What about reviewer turnover?....Also, if there is really a long lag time, has the expectation for the technology changed in that time so that now when you review it, you might review it differently?…If we have this pre-review that now you go back and re-review, what does it mean if we missed anything? Or have we made an error and what do we do when we get the whole application?"

FDA will also have to guard against a "bombardment" of submissions "which cannot be reasonably reviewed out of context," Devine argued. "That's why I'm very concerned about limiting this to those situations where it would really be useful."

ä A final concern raised by the CBER official is the possibility that firms could "game" the pre-submission system.

"We know from past experience," she said, "that companies have their own internal milestones. They have their own internal financial things going on.…These milestones might be things like filing an application or filing a submission, or going to an advisory committee meeting, that are very important to the company." FDA does not "want pre-submissions to be used to kind of game the system, if you will, and

waste our resources just because you've made a commitment to file something.…And so we have to look at what the deficiencies are with gaming."

ä In a presentation at the December workshop, Guilford Senior VP for Drug Development David Savello addressed some of Devine's concerns about the rolling submission process.

Savello noted that there currently are pre-submission models that give clues about how FDA and industry can "engineer a better way to do CMC from IND to product approval." For instance, he noted, FDA has accepted pre-submissions of CMC information for "fast-track" drug products identified as filling a pressing health need.

Fast-Track Review: A Model For Rolling Submission

Prior to working at Maryland-based Guilford, Savello participated in a fast-track review as regulatory affairs VP at Glaxo.

He recalled that while preparing the application for a priority drug product, Glaxo met with an FDA reviewer to discuss the drug substance section of the CMC. The agency and the firm agreed that the drug substance portion, which was ready about six months before the full NDA, could be pre-submitted to help speed up the clearance process.

During the meeting, Glaxo agreed not to make changes to the drug substance component "until after the NDA was filed," Savello explained. Following the pre-NDA meeting, the firm received a letter from FDA asking for a written confirmation of the agreement "So we sent a letter," Savello said, "saying that we will be good boys and girls and that we will do what you request of us in terms of no more changes, and timing, and all kinds of things."

Savello recognized that this type of interaction might be viewed by some firms as FDA "bullying." However, he noted, the result of the process was a much more efficient review for the priority drug application.

ä Although the rolling submission improved the review time for Glaxo's priority product, Savello noted that putting together the reviewable unit was challenging.

"CMC are not milestone events, it is all evolutionary," he commented. "There is no stand-alone activity except the continual monitoring of safety issues around CMC during the IND process. And then finally the separation when IND stops and NDA starts."

Devine's and Savello's presentations set the framework for discussions of the rolling submission during breakout sessions at the AAPS/ FDA meeting. The sessions focused on pre-NDA/BLA filings in the context of biotech and conventional drug substances and finished drug products.

ä During the conference wrap-up, Janssen Tech-nical Regulatory Affairs Group Director Jeffrey Blumenstein, formerly with CDER, presented a summary of these three breakout discussions.

Regarding the issue of what parts of the CMC could stand alone as a "complete" reviewable unit, there was a general consensus that the drug substance portion is a good candidate. Blumenstein explained that the groups agreed that most of the drug substance section "is done reasonably early enough…that you can say that the vast majority of the...section is complete and available for a pre-submission."

The only data not available for drug substances prior to the submission of the marketing application is final stability and final specifications. "Even then," Blumenstein noted, "you may get early reports of provisional specifications that can be submitted for further review in context."

In the biotech breakout session, the sentiment was that pre-submission of the drug substance CMC section would be useful only if it was "fully complete." Blumenstein explained that "while it wasn't a full, absolute consensus, the strong feelings were probably a reasonably complete drug substance section is what we should focus on."

Priority Drugs Are Good Candidates

Regarding the types of applications that would fit under the rolling submission umbrella, "there was a general feeling that priority applications would certainly" fit, Blumenstein said. "Dave Savello's presentation the other day raised that point that you could really put things together when you are trying to look at a very aggressive development cycle and work with the agency to break those down. And those would be certainly the target applications."

ä The participants, however, were not as confident that rolling submissions would be feasible for "standard reviews."

"I think there were concerns about resources both on the FDA and industry side," Blumenstein said.

"To say that you are really going to fix things, that everything is going to be in place, the documentation is all done ahead of time,…is the industry always going to be in position to do that for standard reviews and is the FDA going to be able to really shift the paradigm of their user-fee goals and timelines to be able to review that in time?"

ä On the other hand, the discussions left the door open for the extension of the rolling submission process to standard reviews. "In many cases during the development cycle," Blumenstein maintained, "the drug substance section is reasonably completed at an early enough stage that it may be set and ready for submission at some time before the NDA."

The groups concluded that after more experience is gained with the system, it may be possible to extend it to all application types. "At this point, there may be opportunities," Blumenstein explained, "but I think we probably have to go through some living experience and additional learning perhaps on some priority applications before we delve into the standard reviews."

Drug Substance Section Could Be Reviewable

Regarding the timing of the reviewable unit, Blumenstein stated that the working groups generally felt "a reasonably complete drug substance section" could be submitted a year to eighteen months "before you had a final NDA submission."

However, the groups saw the potential for "diminishing returns" in too early a submission. Most participants agreed that there would be no value in submitting a reviewable unit earlier than six to 12 months before the application was complete depending on the project.

ä It was generally agreed at the breakout sessions that firms should not make a pre-submission if there are "planned changes" to the reviewable unit.

"We are always going to have outliers," Blumenstein explained. "Things will creep up for those few rare cases, and we cannot focus on those when we develop a guidance for this situation. But we should not anticipate that there will be changes when we have a completed section, and that is necessary to get a valuable FDA review."

Session participants acknowledged that the rolling submission would have to alleviate pressure on the agency and not further diminish resources nor affect increasingly compressed review periods.

ä Blumenstein pointed out that industry participants showed awareness of the risk the agency would take by accepting pre-submissions. Should a marketing application fail to evolve, he explained, FDA would not receive compensation as required under PDUFA for the time the reviewer spent evaluating the reviewable unit.

There was general agreement, Blumenstein maintained, that "in most cases, we probably can set up some milestones when you have a useful expectation that you are going to be submitting a successful NDA or BLA, especially in the priority applications [with] accelerated development plans. That is pretty much easy to identify." For standard applications, however, it "might be a little more challenging."

The groups acknowledged that both FDA and the industry would have to be committed to the efficient handling of pre-submissions.

"It is very easy to say that FDA has to commit once you make a submission to turning around and providing feedback on that submission within three months, four months, six months, or whatever the time is." However, Blumenstein noted, industry also "has to be very committed to a timely response back to the FDA."

System Needs Monitoring, Formal Communication

The discussions also focused on the characteristics of a potential rolling submission scheme deemed important.

ä For example, participants discussed the need for a tracking system that would monitor the success of rolling submissions in reducing review times. There would have to be some "new systems" to "measure the [initial] payback from the early review," Blumenstein said.

The criteria for measuring the success rate of the rolling submission would have to account for a transition period.

"We cannot turn off what is in [FDA] now and turn on a new system immediately," Blumenstein said. "So if you start pre-submitting on top of [applications] that are already out there, there is going to be a certain transition period were you will have...a big tsunami" of submissions hit the agency.

The need for formalizing the communications between reviewers and industry regarding pre-submissions was also considered important.

"Probably, informal communications with this are not going to be sufficient," Blumenstein said. "That probably came out as much on the FDA side as on the industry side: [the need for] form letters - not telephone calls, not faxes - official letters, official correspondence to make sure that we all have a common understanding of what has been done and what the outcome is."

The workshops stressed the importance for FDA to be clear and firm when making a decision about a reviewable unit. Reviewers will need to "make very clear that agreements are final on both sides," Blumenstein said. "If the FDA asks the industry to do something, [it should] make very clear that it is not a request." Similarly, he added, "when the FDA commits to something," there needs to be "an understanding that that doesn't change as it goes on."

ä Industry will be looking for "clear guidance" from FDA to make the rolling submission work. The guidance should delineate the ground rules for the system.

"This is not something that can be left ad hoc," Blumenstein summarized. Even though the procedure has been used on a case-by-case basis for accelerated applications, "if we are going to expand the scope, we are going to need formal guidance to make sure that we are all on the same page with regard to this."

AAPS/FDA Workshop Report To Guide FDA

Technical committee reports from the various meeting breakout sessions are being drafted into a final consensus report. The workshop report will, in turn, help provide a basis for future FDA deliberations and guidances on the CMC review process.

ä CBER official Devine, in summarizing her presentation, discussed the issues that FDA needs to consider in framing future guidances.

For safety data, she said, FDA has "already made much progress in terms of defining information [required] at Phase I" ("The Gold Sheet" December 1995). "We now need to move forward into Phase II and III."

"We have the CMC guidances at the end of the process for the NDA/BLA," she said, and "we have the Phase I guidance document for your IND" at the beginning, "but you got this black hole in the middle. So we do need to make some progress in this area."

ä Guilford's Savello suggested that industry and FDA may need to consider a more comprehensive overhaul of the CMC submission process.

Savello asserted that the ever-increasing size of CMC sections has become a problem for industry and FDA. He characterized the process as a "vicious circle" that leads to larger CMC sections year after year.

When at Glaxo, he explained, the number of pages in the CMC section of NDAs that were submitted increased yearly. "You can trust me," he said, "from 1991-1996, they went up even higher. CMC amendments by year were increasing."

In turn, CMC deficiency letters during this timeframe "showed a dramatic increase," Savello added. "And of course that begets the cycle all over again, because in trying to avoid deficiency letters, we made even bigger CMC sections. The whole thing is a vicious circle and we cannot seem to get out of it."

"Re-Engineering" Of CMC Needed?

Savello commented that the CMC process was never really "engineered." Instead, the CMC review process has evolved, and the "driving force" has been "FDA's reaction to disasters, catastrophes, and product failures in the marketplace or during inspections."

ä In this context, a true "engineering" of the content and format of the CMC section of NDAs/BLAs is something Savello thinks will be necessary in the future. He provided some examples of areas that could be targeted in this process.

To help FDA reviewers, Savello suggested that firms could link parts of the CMC with the safety and efficacy portions of the marketing application. "Just talking to a couple of reviewers at FDA," he said, "they have to go hunting all through the application to do the linkages themselves."

The process could be made more efficient if firms included in the marketing application, for example, "a section called 'impurity linkage' or 'impurity justification' linking all the safety data, all the toxicology data, with the impurity levels, and put that all in a nice package in an easily reviewable form." He noted that the European Expert Reports could serve as models for this approach.

Regarding potential areas of revision to the content of the CMC, Savello pointed to the amount of impurity, bioavailability/dissolution, and stability data required in applications. "This continues to be, I think, an issue that is evolving, and it is not having a good, healthy, scientific debate."

Savello expressed a desire to participate in a meeting to address the content and format of the CMC section. "Let us look for the scientific basis for all of the things we are presently putting into these submissions and start to examine if there are good scientific reasons for them, or maybe some of these things can be eliminated or backed-off from."

POINTS TO CONSIDER ON PRE-MARKETING CMC MEETINGS
	The following is an outline of points to consider in approaching pre-marketing meetings with FDA on CMC issues. The outline is divided into four sections: (1) general points for all meetings, (2) pre-IND meetings, (3) end-of-phase II meetings, and (4) pre-NDA/BLA meetings. The outline includes summary points provided at the December AAPS/FDA CMC workshop by FDA Division of Drug Chemistry II Chemistry Team Leader Stephen Moore followed by workshop consensus points from breakout groups which addressed the three types of meetings from the point of view of both conventional drugs and biotechnology-derived products.
GENERAL POINTS FOR ALL MEETINGS FDA FDA regulatory resources: l 21 CFR 312.47 Investigational New Drugs, Meetings l CDER Manual of Policies and Procedures (MAPP) 4512.1, Training and Communications, Formal Meetings between CDER and External Constituents (March 7, 1996) l Division of Pulmonary Products, Points to Consider, End of Phase 2 Meeting (August 30, 1996) Purposes/goals/attributes: l Facilitate development of new drugs through proactive interaction on scientific and regulatory issues l Evaluate sponsor's plans relative to agency policy l Provide regulatory input and guidance resulting in improved submissions l Be accessible to the extent that resources permit l Conduct meetings so that there is free, full and open communication l Generally instituted by sponsor l Consultants may be present Protocol: l Initial written request and scheduling l Meeting package: - Statement of purpose - Listing of specific objectives/outcomes - Agenda - Attendees - Background material l FDA topics communicated to sponsor (if applicable) l The meeting:		- Objectives - Discussion points - Decisions (agreements) reached - Unresolved issues - Action items l Minutes Summary of pre-marketing meetings: l Pre-IND meeting: Focus on safety issues, avoidance of clinical hold l End-of Phase II meeting(s): - EOP II multidiscipline meeting is critical evaluation and planning meeting for Phase III/pivotal studies - EOP II CMC meeting is a valuable evaluation and planning meeting to discuss detailed CMC issues - All sponsors are highly encouraged to participate in EOP II multidiscipline and CMC-specific meetings l Pre-NDA/BLA meeting: Focus on filing issues, improve NDA/BLA submissions Workshop l Provide clear written objectives and meeting agenda l Briefing package: - Keep package brief and concise - Provide references to IND and other complete literature - Provide specific questions/proposals - Send to FDA in advance - Closure - Recap key points/agreements at the end of meeting - Create action item/responsibility list - Written minutes: l Minutes MaPP defines FDA minutes as official record l Company can submit minutes (perhaps more detailed) to IND file with desk copy to participants l FDA minutes will be sent to company
PRE-IND MEETING FDA l Focus on safety issues l CMC portion generally held in conjunction with other disciplines l CMC portion generally brief for conventional synthetic drugs l CMC portion discussed in more detail for: - drugs/biologics from human/animal sources - biotechnology drugs - reagents from animal/cell line sources - novel drug-device delivery systems - novel dosage forms Workshop (Conventional Drugs) l Not needed for innovator companies, except for some specific safety issues (e.g., impurities) l May be needed for companies on the learning curve and for special/unusual drugs l No need for formal guidance l If a multidisciplinary meeting is held, it is recommended that a CMC representative attend Workshop (Biotech) l The applicant should supply the agency with a summary of manufacturing information, characterization data, impurities identification and also any information on unique characteristics which differentiate the product from other similar entities l The more invested in identifying the issues and approach prior to the meeting, the more effective the meeting will be at resolving the issues l An effective mechanism is to capture on a flip-chart the issues and their resolution during the discussion l Could "good meeting practices" be developed? l Feedback for pre-IND meetings - Briefing document l Background is critical l Should include plan, but be issue driven - If FDA has experience base, it will be shared at the meeting to guide sponsor - For new technology, industry needs to provide critical background information - Meeting minutes l Sponsors want copy of the minutes (implementation of meeting requirements as dictated by PDUFA 2 - Get right people at the meeting l Sponsor should request specific expertise be in attendance l Agency should ensure that appropriate disciplines are present		END OF PHASE II (EOP II) MEETING FDA General EOP II meeting: l Purpose/Goals/Attributes - A working meeting - Critically review the results of the drug development program to date - Prospectively review proposals for Phase III/pivotal studies - Evaluate sponsor's plans relative to agency policy - Identify scientific issues and potential problems - Obtain upper management involvement, as necessary - Create a time line of future needs for meetings and discussions - All sponsors are highly encouraged to participate CMC EOP II meeting: l Purpose/Goals/Attributes - Discuss approach to test methods and acceptance criteria - Discuss approach to stability protocols - Discuss correlation/link between formulation(s) - Emphasize need to coordinate all activities with contractors - Identify any other issues - Important for all drugs - Reduce number of review cycles - Multi-discipline and CMC-specific meetings highly encouraged l Meeting package - Background l Overview - Summary of the product - Drug development to date - Time line for future development l Drug substance l Drug product l CMC issues and questions to be addressed l Meeting package: Drug substance - Conventional - Description and characterization - Manufacturer - Syntheses/method of manufacture - Process controls - Reference standard - Regulatory specifications/analytical methods - Container/closure system - Stability l Meeting package: Drug product - Components/composition - Specifications and methods for inactive components - Manufacturer - Methods of manufacturing and packaging
- Regulatory specifications and methods - Container/closure system - Stability - Microbiology l Typical discussion points: Drug substance - conventional - Facilities - Drug master file cross-references - Designation of starting material(s) - Synthetic scheme - Effect of scale-up - Unique physical/chemical issues (e.g., polymorphs, chiral purity) - In-process controls - Validation studies - Methods and specifications (e.g., reference standard., identity, assay, impurities) - Stability protocol (e.g., stress testing, degradants, site-specific batches) - Anticipated CMC changes - Bridging studies needed l Typical discussion points: Drug product - Facilities - Formulation(s), linkage - Complex dosage forms - Novel excipients - Sterilization/aseptic filing - Effect of scale-up - Methods and specifications (e.g., reference standard., assay, impurities, dissolution) - Novel container/closure system - Devices (if applicable) - Stability protocol (e.g., matrixing, bracketing, degradants, site-specific batches) - Anticipated CMC changes - Bridging studies needed l Typical discussion points: Drug substance - Biotech - Characterization (e.g., 1° , 2° , 3° structure, glycosylation) - Bioassay - Host cell/vector expression system - Master/working cell banks (e.g., viral and pathogenic agents, genetic stability) - Process validation of viral removal/inactivation - Fermentation/cell growth - Harvesting/extraction - Specialized purification steps - Biosource reagents - Product-related substances and degradants (e.g., activity) - Host cell DNA removal - Host cell proteins contamination l Typical discussion points: Bridging studies needed - strategy (biotech) - Types and level of comparative testing needed (in vitro < animals < humans) l Analytical testing l Chemical/physical testing		- Characterization (1° , 2° , 3° structure, glycosylation) - In-process controls, specifications - Impurities profile - Stability l Biological assays - In vitro (biochemical, cell lines) - In vivo (animals) l Pre-clinical testing (PK, PD, TOX) l Clinical studies (PK, PD, immunogenicity, safety, efficacy) Workshop (Conventional Drugs) l CMC should be represented at multidisciplinary EOPII meeting (i.e., clinical meeting) l Different names - Pre-pivotal - CMC strategy - Pre-registration - Pre-EOPII l Separate CMC meeting recommended, but optional l Much more important that pre-IND and pre-NDA meetings l Draft CDER MaPP is under preparation l Purpose/Scope - Clarify critical issues, such as: l Definition of starting materials l Sites of manufacture l Dissolution l Stability protocols (e.g., matrixing, bracketing) - Review protocols (e.g., stability) - Identify any "show stopper" issues l Inherently unstable drugs l Drugs difficult to formulate l Novel delivery systems and/or excipients l Closure - Recap key points/agreements at the end of meeting l Create action item/responsibility list l Written minutes l CDER MaPP on minutes defines FDA minutes as official record l Company can submit minutes (perhaps more detailed) to IND file with desk copy to participants l FDA minutes will be sent to company Workshop (Biotech) l The EOPII meeting may facilitate and speed new drug development and evaluation. It provides an opportunity to identify and resolve CMC issues prior to Phase III studies: Specs, dissolution, manufacturing site change, test methods, synthesis change, impurities l Feedback for EOPII meetings - Typically, the company sets the agenda, and FDA reacts to it. Sponsors would like FDA also to communicate important issues to sponsor before the meeting
- Sponsor doesn't want FDA to remain silent in the meeting about issues that they know could be a problem - However, sponsor doesn't want to make the meeting a "fishing expedition" - Apparently, some sponsors avoid EOPII meeting to avoid agency requests for information that is perceived as overly burdensome PRE-NDA MEETING FDA l General: purpose/goals/attributes - Focus on filing issues - Exchange information about proposed marketing application - Avoid delays associated with initial review of marketing application - Avoid "refuse-to-file" - Facilitate review of marketing application - Uncover any major unresolved problems - Improve marketing application filings - Reduce number of review cycles l Typical discussion points: CMC - Brief update of manufacturing sites - Brief update of manufacturing changes - Brief update of bridging studies - Stability data available at time of submission - Need for environmental assessment - Organization/format of CMC sections - Time lines (e.g., NDA/BLA submission, plant readiness) Workshop (Conventional Drugs) l Purpose - Review of EOPII plan		- Discuss changes/deviations from EOPII plan - Update linkages: stability, impurities, dissolution/bioavailability - Input on fileability/approvability l E.g., stability complications - Clarify timelines for NDA filing - Discuss planned NDA amendments l E.g., update of stability data - Discuss manufacturing facility readiness Workshop (Biotech) l The purpose of the pre-NDA/BLA meeting is to address major unresolved CMC issues, to familiarize the reviewers with the general CMC information to be submitted, and to discuss the formatting/data presentation in the marketing application. l Typical issues - Data package: stability, batch data - Validation strategies - Manufacturing site/scale changes - ICH compliance l Normally, this meeting is to discuss - Filing issues - Follow-up on issues discussed earlier in EOPII meeting, such as cell bank passage limits - However, in many cases, issues that generally would have been handled earlier actually need to be addressed at this time l Sponsors want a streamlined process for managing the review: - Small and separate questions/issues to be handled by FAX and telephone conferences - Formal meeting for list of issues