• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services


  • Print
  • Share
  • E-mail

Informal Discussion Between the Institute for Clinical PET and FDA: Development of Current Good Manufacturing Practices and Approval Procedures for PET Drugs - September 28, 1999





Tuesday, September 28, 1999

9:00 a.m.

Goshen Room
Holiday Inn Gaithersburg
2 Montgomery Village Avenue
Gaithersburg, Maryland




Jorge Barrio, Ph.D.

Ronald Callahan, Ph.D.

Peter S. Conti, M.D., Ph.D.

Robert Ferris

Jennifer Keppler

Jerry Kuhs, R.Ph.

Dennis Swanson, R.Ph.



Jane Axelrad, J.D.

Florence Houn, M.D.

Paul Hunt

Dr. Ravi Kasliwal

Dr. Eldon Leutzinger

R.K. Leedham, R.Ph.

Eldon Leutzinger, M.D.

Ravi Kasliwal, Ph.D.

Patricia Love, M.D.

Victor Raczkowski

Tracy Roberts



Opening Remarks and Welcome

Ms. Jane Axelrad 4

Opening Remarks

Dr. Jorge Barrio 5

Current Good Manufacturing Practices Update

Ms. Axelrad 9

Discussion 19

Approval Procedures Update 188

Chemistry: Dr. Ravi Kasliwal 189

Clinical Pharmacology/Biopharmaceutics:

John Hunt 220

Pediatric Rule:

Dr. Patricia Love 232

User Fees:

Ms. Jane Axelrad 240


Opening Remarks and Welcome

MS. AXELRAD: Good morning. I am Jane Axelrad, the Associate Director for Policy in the Center for Drug Evaluation and Research at FDA and the Chair of the PET steering committee.

This is the fourth in a series of public meetings that we have been holding with representatives of the Institute for Clinical PET to discuss our efforts to implement Section 121 of the Food and Drug Administration Modernization Act.

As you all know, the Act requires us to develop appropriate procedures for the approval of PET drugs and appropriate current good manufacturing practice requirements.

I suggest to start out that we all go around the table and introduce ourselves.

DR. HOUN: Florence Houn with the Office of Drug Evaluation III.

MS. ROBERTS: Tracy Roberts with the Office of Compliance in CDER.

MR. RACZKOWSKI: Victor Raczkowski, Deputy Office Director in Office of Drug Evaluation III.

MR. KUHS: Jerry Kuhs, Vice President, Regulatory, PETNet Pharmaceutical Services.

DR. CALLAHAN: Ron Callahan, Nuclear Pharmacy from Mass. General Hospital.

DR. CONTI: Peter Conti, University of Southern California. I am a member of the Society of Nuclear Medicine Board of Directors and Government Affairs Committee, President-elect of the ICP.

MR. FERRIS: I am Bob Ferris from Mallinckrodt Medical. I am the U.S. Operations Manager for PET.

MR. SWANSON: Dennis Swanson, University of Pittsburgh.

DR. BARRIO: Jorge Barrio, UCLA.

MS. KEPPLER: Jennifer Keppler, ICP.

MS. AXELRAD: What I would like to do is just welcome everybody and thank you all for coming. We are a little further out from the city this time, in Gaithersburg, but I gather you were all well taken care of last night.

I would ask Jorge if you would like to make some remarks before I explain what we have been doing over the last several months and what we hope to accomplish today.

Opening Remarks

DR. BARRIO: Thanks, Jane.

This is an important day for the PET community because we are going to discuss this morning, perhaps in the course of the day, the new CGMPs and ANDAs proposed by the FDA. As we all know, CGMPs particularly have been a point of contention for a long time. These have been related, we believe, to the history and evolution of PET.

PET was born in academia more than twenty-five years ago as a powerful tool to investigate human biochemistry, pharmacology, physiology in health and disease, a new knowledge and a new discipline that we call molecular medicine is being shaped as a result.

We have to remind ourselves that when the time came to introduce PET into the clinic about ten years ago, probably a little bit more, we were all confronted with an unprecedented situation. The lack of the radiopharmaceutical industry and also the lack of conventional clinical trials produced a regulatory paralysis and confusion.

There was no general precedent for academia developing and directly introducing drugs to the public. When compared with conventional nuclear medicine, with industry involvement, the difference was obvious and very evident.

The agency hesitated and suggested a few different alternatives initially, but when everything was said and done, they proposed very rigid regulations for PET radiopharmaceuticals. After all, that was the reasoning. Radiopharmaceuticals were drugs and they had to be regulated like drugs.

In other words, when any radiopharmaceutical produced at any academic facility was intended to be related like any drug produced by pharmaceutical giants like Dupont, Merck, and Hoffmann-LaRoche. Our analysis and thinking at the time was how is it possible that a single injection of a few nanomoles of a radiopharmaceutical into patients can be regulated similarly to drugs given in massive amounts to people for many years. That was really the center of the argument.

The PET community responded at the time by indicating that FDA shouldn't be involved with the regulation of PET radiopharmaceuticals, but this really didn't work. Reimbursement never came and the truth was that the powerful research tool couldn't be made available to the patients.

Clinical PET was paralyzed. Equally important, we were angry with each other. But FDAMA '97 came to the rescue. You recognized the special characteristics of PET radiopharmaceuticals, provided to the agency the legal tool for appropriate regulations, implicitly opened the door for reimbursement and for transferring the benefit of the technique to the public.

Our ten-member group, the committee I chair, with a combined experience in the field close to 200 years--I would like to tell you, it is not that we are old; it is that the time flies. I have been working with FDA for about two years now and we have seen a new mature approach in the agency and we would like to thank Jane Axelrad and Florence and the staff for catalyzing this transformation and making this discussion possible.

I believe the PET community has also matured and now perceive that the comprehensive regulatory solution for PET pharmaceuticals is at hand. With this background, we will review today a new CGMP draft. The PET community has said that many times, we believe that PET radiopharmaceuticals have characteristics that should be introduced in the decision piece.

We believe that these characteristics require a product-oriented decision piece with process validation that is restricted to the very essence of the preparations process under consideration. The very intense and successful efforts that we have made with FDA with the USP monographs, and they are going to continue tomorrow in Rockville, we believe are a step in the right direction.

During the course of this meeting, we will also address several points about the ANDAs, a draft affecting fluoride, ammonia and FDG, but, overall, we would like to commend the agency for an excellent job with them.

We are looking forward to maintaining this open, communicative atmosphere with the agency. I will encourage everyone in the audience to participate actively in this discussion. Let's use effectively this opportunity.

Thank you.

MS. AXELRAD: Thank you, Jorge.

Current Good Manufacturing Practices Update

Introductory Remarks

Before we get into the details of the CGMP regulations, I would like to summarize what we have accomplished at our previous meetings and what we have been working on since we last met, and also what we would like to cover today.

As you know, these meetings have been scheduled to allow for the exchange of scientific and regulatory information at the working level as we develop the new approval procedures and good manufacturing practices for PET drugs.

In August, at our very first meeting, August of '98, we discussed chemistry, manufacturing, and controls issues, and made a lot of progress in identifying some important principles that we could carry forward in our work on the approval procedures and CGMPs.

In November, we presented the preliminary results of our review of the literature on the safety and effectiveness of N 13 ammonia for myocardial perfusion and F 18 FDG for myocardial viability.

In February, we discussed the preliminary results of our review of the literature concerning the safety and effectiveness of F 18 FDG for oncology indications. We also explored in more depth the exiting new drug approval procedures to obtain a better understanding of the possible problems associated with our existing procedures and to identify some areas that we might want to change, and we began to discuss a draft outline or concept paper of possible current good manufacturing practices for PET.

In June, we presented our preliminary findings on the safety and efficacy of the three PET drugs for the named indications at a public meeting of our Medical Imaging Drugs Advisory Committee.

The committee agreed with our conclusions that the literature was sufficient to support a finding of the safety and efficacy of F 18 FDG for oncology and myocardial viability and N 13 ammonia for myocardial perfusion. They requested additional information concerning O 15 water which we intend to present to them in the spring.

As a result of that meeting, we have begun the preparation of a notice that we intend to publish in the Federal Register summarizing our conclusions regarding the safety and efficacy of F 18 FDG and N 13 ammonia for the indications that we presented to the advisory committee.

The notice will also discuss the application procedures for these drugs for these indications. We hope to publish this notice by November 21st, 1999, the two-year anniversary date of Section 121 of FDAMA or shortly thereafter depending on how long it takes us to resolve some of the procedural issues, some of which we are going to be discussing today. We also are working on a guidance document that we want to have ready.

The notice will also address how other PET centers can obtain approval for applications for FDG, for identifying the foci of epileptic seizures and also for sodium fluoride for bone imaging.

We believe that the epilepsy indication can be obtained through either an application submitted and approved under Section 505(b)(2) of the Act, which we have discussed before and we can discuss again here--it is basically something that is based on the literature--or an abbreviated new drug application depending on the timing if the submission of the application and the status of other approvals.

We believe that approval of sodium fluoride for bone imaging can be achieved through abbreviated new drug application since there is already an approved new drug application for this substance for this indication.

We also can answer questions or anything you might have about this as we talk about the procedures later.

After we present O 15 water to the Advisory Committee in the spring, and assuming that they don't raise any new issues concerning this drug, we expect to prepare another short Federal Register notice announcing our findings and inviting applications for that drug.

We are also planning to do a review of the literature on the safety and efficacy of F-dopa, and we are ready to begin doing that with the idea that when we finish it, we will present those findings to the Advisory Committee, as well. We have established a pattern, and we will follow through with O 15 water and also F-dopa.

Of course, as we have all recognized, one of the most important pieces that we are working on are the current good manufacturing practices. We discussed an outline of these in February, and we received many comments from ICP and others at the meeting.

As a result of those discussions, we reexamined our concept paper and we have significantly revised how we are approaching these regulations. At the time, many of the comments were that the draft that we discussed in February was way too detailed, so I think that you can see we have significantly reduced the detail in that and tried to be a lot less prescriptive in this newer version.

Last week, on September 22nd, we published a Notice of Availability of a preliminary draft of the codified portion of those regulations. This was a fairly unusual step. Under our regulations, the other parts of the FDA regulations, we are not permitted to provide to anyone outside the agency any draft regulations that are under development unless we get specific permission from the Commissioner or we publish a Notice of Availability and make them available to everyone.

We felt that the last time we just issued it when it was outline, it wasn't really the actual regulation, so we didn't have to go through that, but now that we had gone further and actually developed codified language, we felt that it was really important to share our actual draft working language with the PET community because the difficulties are in the details of the wording of the regulations and we really wanted to get the community's impression of where we were going on these.

So, we did take the unusual step of making them available through this mechanism, and, of course, they are also posted on our web page.

In a minute I am going to be turning this over to Tracy Roberts, who has worked really hard on these for quite some time now and will be leading the discussion of the proposed regulation. Our plan or our hope was that we would publish an official draft of a proposed rule by or shortly after November 21st, 1999, but I gather that we have still got a number of issues that we need to work out on these regulations, and so it will depend on how quickly we can work the issues out and get something that we think is going to be reasonably acceptable.

We also have an elaborate clearance process for these within the agency and the department, and so our publication date is not entirely within our control even after we finish something that we are all satisfied with.

We are also developing a guidance document that will describe in more detail our expectation concerning how to achieve compliance with these regulations. As we talked about the last time, since you indicated that you thought that the draft that we discussed in February was too detailed, and this new version is significantly less detailed, we need to address in a guidance document more specific, you know, what we mean by certain of the terms that we use in the preamble and put it is some context, so that people can understand better what is going to be looked at during an inspection, what they actually have to do.

This is similar to what we did in the rule on radiopharmaceuticals where the general outlines of how we view getting new indications for radiopharmaceuticals was laid out in the codified section of the rule that published in May, and we also did an accompanying guidance document that we issued in draft form that had more detail on what we really meant in terms of how you would apply for and study drugs for certain claims and indications.

We will follow the same pattern that we used there, and the guidance document on CGMPs for PET will be issued for comment during the comment period for the regulations, so that you will have a chance to see how the regulation and the guidance document fits together.

We don't have a draft yet of the guidance document, but we are working on it. In a way, we really have to at least get some agreement on the shape of the regulations before we can fill in the details in a guidance document.

Later on, on the agenda, we plan to give you an update on our development of approval procedures, and we have several members of our staff here to discuss the sections of the procedures for which they are responsible.

We expect to publish at about the same time that we publish in the Federal Register the notice on the safety and efficacy findings. We also plan to publish a guidance document explaining how to submit applications for the published drugs and indications.

This will go into some detail on what should be submitted in the chemistry, biopharmaceutics, and other sections of both the 505(b)(2) application and an abbreviated new drug application.

It will include the model chemistry application sections, which you have mentioned already this morning, one of which we released and discussed earlier, and two of which will we will be discussing today.

Again, we don't have a draft of that guidance yet available to share with you, but we believe that it will look something like the guidance document that we published for abbreviated new drug applications before FDAMA, which sort of walked through the different sections of the application. That was one of the only documents that we published on PET that Congress didn't tell us to revoke, so I think it is still up there on the web.

Of course, we significantly revised it to take into account our new procedures and especially our new model chemistry applications. The guidance will be published for comment in accordance with the agency's good guidance practices, so everybody will have a chance to comment on it before it actually is finalized.

We are also considering preparing a sample application based on real data that will demonstrate even more specifically how we envision an application being prepared. If we are able to do this--and it really is a question of resources and how we would get the data--the sample application would be available later this winter, would essentially be the model chemistry format filled out to reflect probably dummy data, you know, something culled from various places, but basically, what we would expect to see and how you would fill in the different blanks in the form.

We do not expect a two-year clock envisioned in the statute for the required approval of applications to begin to run until all of our procedures and regulations are in place, but we will encourage the submission of applications voluntarily as soon as the notices are published in the Federal Register and the guidance documents are completed, so that we will not receive a huge bolus of applications all at the same time.

The thought of getting something like 210 applications with due dates in that two-year period is really a nightmare, and we think that people can start submitting the applications for some of these indications early on.

We believe that the applications can be submitted and approved before the final good manufacturing practice regulations are in place. We would sort of treat that as a separate issue, and we would like to work with the community to achieve this.

That concludes my introductory remarks. Copies of all of the relevant documents that we are going to be discussing today are available on the table over there. I would like to stress that all of the documents are preliminary working drafts. They were prepared as a starting point for the discussions today and at meetings to follow.

I would also like to mention one procedural matter that doesn't seem to have been a real big problem in the past. This is technically a working meeting between the Institute for Clinical PET and FDA, but we will provide an opportunity for members of the public in the audience to comment on issues at appropriate times during the discussion.

This format seems to have worked well at our past meetings although we do have a little bit bigger audience this time. It seems to be growing. The first time we only had like one or two people, and now people are getting more interested in what we are doing.

Also, I am told that if you--I am probably not doing it myself--but if you speak back from the microphone, you won't buzz or click or whatever, get noise and static.

Before I turn this over to Tracy, Jorge, do you or any of the members of the ICP have any additional comments that you want to make?

DR. BARRIO: Any comments? Nothing.

MS. AXELRAD: I will turn it over to Tracy to talk about the current good manufacturing practice.

Also, I notice Dr. Love has joined us. Everybody knows Dr. Love is the Director of the Division of Medical Imaging.



MS. ROBERTS: I can see we are starting with the most cantankerous topic first this morning. Hopefully, we will be able to go over these regulations and come to some mutual understanding about what we think is important and what you may have problems trying to meet in this regulation.

I would like really to hear from you instead of standing up here going through the regulation part by part about what you think are the problems with this regulation that we have put before you and posted on the internet.

I would like to go section by section starting with the definitions, and I would like to hear from you all about what we think might be some of the areas that are problems. We will just jump right in there this morning.

DR. BARRIO: I would like to ask a question, Tracy. When you compare these regulations that you put forward with the ones in February, we understand these are less detailed, what do you think is, in substance, different?

MS. ROBERTS: In substance, what is different, we took a lot of your comments from the last meeting regarding personnel specifically and your resources that we have heard from you, that you may not have in order to try to meet these.

We understand that some of these firms are smaller, there is few people, and we have taken then into consideration as one of the main topics. You will not see areas where there are direct two-person checks, however, there are areas where we need a second check. It doesn't need to be done at the same time. Two people do not need to stand there and watch the operation. It is just purely a second check or a review of the documentation, such as a batch record or such as a master production record, in order for them to make sure that they are using the correct one and also to make sure that everything is in check, for example, a quality control person would review the batch records, the test results, to make sure that it's okay afterwards.

What we tried to get rid of is especially those two-person checks and requiring updates to facilities. We tried to make this regulation a minimum requirement that you would be able to meet for what we think are minimum standards for PET drug products in order to make sure that the quality and the purity are all still kept within the product.

Now, you will see that we still have kept a validation, a type or process validation for the computer systems and also for the process, and I think what might be an issue from listening to Dr. Barrio, his opening remarks, about what we mean and what we think you might need to meet in order to do that.

It is on a different scale than a huge, large drug manufacturer, however, the FDA still uses the same terminology for certain issues. We could break away from that, however, it is a little difficult to dream up new words for what essentially is the same thing. It is what we would expect for you to give us documentation to meet that.

Now, I would like to discuss some of those things and maybe have you give me an overview of what exactly you didn't like out of here, some of the biggest issues, if you would like to do it that way.

I get the feeling that you don't think much has changed from the last version. It's a lot lighter, first of all, a lot less pages, and there are different concepts here that take in a broad spectrum rather than target specific type areas.

MS. AXELRAD: I would like to just mention also that one of the things that we did was we looked at the USP chapter on PET, and we tried to make the two congruent. We tried to make sure that--there were some things in the USP chapter that we thought were really way too specific for this kind of a regulation, and we didn't pick those up, but basically, conceptually, we made an effort to make sure that our chapter and the USP chapter were congruent, that we didn't go far beyond what they did and that we picked up the important concepts that were in the USP.

MR. SWANSON: A general comment I think about the CGMPs that are presented here and then a specific question that I have.

I think part of the problem that at least I have, and I think some of the other people have with these, is now we have gone from a specific set of requirements or prescriptive to something that is, frankly, rather open-ended, and I think those of us in the PET community have a concern that we don't know how to comply with these.

That may well be addressed in your guidance document and get to specifics, but it is hard for us to review something like this without also having the guidance document available to know what you are really talking about here.

Understand that I and many other people in the room here are--I am a pharmacist and I do not come from a drug manufacturing background as I think is pretty well standard for most PET centers, so you are starting at that level. We don't understand this. We probably need more guidance than what the current set of CGMs give us.

Now, I think the guidance document will address that, but it is hard to truly comment on these without the guidance document.

A specific question. You might want to go back and tell us how you are planning on regulating PET centers, because all of a sudden in here, we start talking about compounded PET drugs, as defined as a drug that has been compounded on the order of a practitioner who is licensed by the State to compound or order compounding for a PET drug and is compounded in accordance with the State's law for a patient or research.

Now, it appears that we are mixing CGMPs with pharmacy compounding and the State regulation of pharmacy compounding, and I am utterly confused as to where you are going with this.

MS. AXELRAD: All we did was pick up the language in the statute. The statute itself, the provision deals with compounding PET drugs, and we had to come up with a definition of PET drugs and compounded PET drugs, and what we did was pick up the statutory definition of that.

We were originally looking to see, you know, the law and the conference report suggests that we should take into account the relevant differences between for-profit and not-for-profit, and one of our problems all along has been trying to distinguish what would be appropriate for the smaller operations and to distinguish them from the larger, more commercial operations, and so we were trying to figure out whether there was a basis for using the statutory language defining compounded PET drugs and distinguishing them from others for this, but it turns out that when you look at the legislative history of the statute, you can't really do that.

But these definitions and language really just pick up the definition of PET drug and compounded PET drug that is used in FDAMA itself, and if it is confusing, we could probably get rid of--

MR. SWANSON: The problem I see is, you know, you are crossing lines into pharmacy compounding by this definition. I can sit back and I can look at this and say to myself, well, gee, is this a first initiative of the FDA to develop CGMPs for pharmacy compounding in general.

Well, it can be interpreted as that, okay, because you are requiring, by the definition, people to comply with the State laws that deal with the practice of medicine and the practice of pharmacy, and where else in your CGMPs for the pharmaceutical industry do you require such a thing.

MS. AXELRAD: But the thing is we use the term "PET drugs" throughout the regulations, and when you look at the statute, the statute specifically says, Regulation of compounded--you know, it is titled, "Positron Emission Tomography"--and then it says, "Regulation of compounded positron emission tomography drugs," adding at the end, the following: "The term compounded positron emission tomography drug means a drug that exhibits spontaneous disintegration of unstable nuclei"--blah, blah, blah--and (b) "has been compounded by or on the order of a practitioner who is licensed by a State to compound or order compounding for a drug described in paragraph (a), and is compounded in accordance with that State's law for a patient or for research, teaching, or quality control."

It is just the definition in the statute. The regulations just incorporate that.

MR. SWANSON: But I don't know why you carried that definition forward here. I understand that the statute was developed as an exception to us being required to follow FDA CGMPs, and as part of that exception it defined what a compounded drug was.

If you are following State Board of Pharmacy laws, then, under that statute, you are in compliance with that statute, but that statute is also saying that the FDA needs to develop new policies and procedures to deal with how they are going to regulate PET.

So, I think it is inappropriate for you to take that definition and put it into now, what was an exception, and to put it into your CGMPs.

I just don't understand why you do that, and it creates all kinds of confusion on our part, and it makes it definitely appear that the FDA is now getting into CGMPs for pharmacy compounding. That is the way I am looking at it. Okay?

MS. AXELRAD: Well, I hear your comment and we can that into account. Again, we were trying to figure out a way of distinguishing some of the smaller operations from the larger, but we looked at the statute.

The CGMPs that we are developing apply to both. When you look at the statute and the conference report, basically, the CGMPs that we are developing apply to both, so we can probably remove this language if it is confusing regarding compounding.

One of my other jobs is to deal with compounding. Believe me, we are not interested, certainly not in this context, in suggesting that we are doing CGMPs for compounding.

DR. HOUN: Another option, too, is in the preamble, to kind of clarify when we discuss PET drugs, how it does not overlap with other fields of compounding.

MS. AXELRAD: Well, there is a specific exclusion from the compounding provisions for both PET and radiopharmaceuticals, so we have no intention of doing CGMPs in here for anything else.

MR. KUHS: It was our understanding that the compounding provision was to be an interim position until such time that the agency developed new regulations for a manufactured PET drug.

MS. AXELRAD: For PET drugs, yes, for any PET drugs.

MR. KUHS: That the compounded portion of that was an interim solution during this two- to four-year period.


DR. BARRIO: Then, the intent of agency is to--decision base are intended for both academic centers, hospitals, and industry, too?


DR. BARRIO: If you go to page 3 of 11, under Item (o), there are two concerns with that statement. "Receiving facility means any PET center," et cetera, et cetera. That seems to imply that we are really talking about only academic facilities here, but even if we do that, the term "physically located in the same building," will disqualify 95 percent of the PET centers.

Then, two clarifications here. How does this fit into the general idea that this will apply to both industry and academia? Second, in the part that applies to academia, why the same building requirement was introduced?

MS. ROBERTS: It is my understanding that these regulations will cover both compounded PET drugs and commercially manufactured PET drugs, compounded under the order of a physician with a prescription.

If you look in the back of the regulations, under Distribution, we have procedures in place to make sure that there is a valid prescription. Those things are in place.

Receiving facility, we were just trying to point out a receiving facility is any different department that receives a PET drug that has been manufactured, compounded from actual site. It could be, from my understanding, that there might be a PET center in part of a university or a hospital that will actually produce the PET drug and then send it to either another part of the hospital for imaging or across the street or maybe even across the nation.

We were trying to--what is meant by this receiving facility is all-encompassing. Whoever is actually receiving the PET drug and that is going to use it is a receiving facility.

DR. CALLAHAN: But you say that this facility must be in the same building. How does that fit? I don't understand how we can go across the country and still be in the same building.

MS. ROBERTS: Well, the actual definition is, "means any PET center, hospital, institution, imaging facility, or other entity that accepts a compounded PET drug for human use."

The second sentence within that was just a clarification that says, "A receiving facility may be in a different department but physically located in the same building as the PET center."

I was trying to make it more clear--and I see I muddied the waters a little bit--that I didn't want to make you think that a receiving facility has to be a separate physical building.

DR. CALLAHAN: To us, this implies that it may be in another department, but it must be in the same building.

MS. AXELRAD: That can be easily fixed in terms of the wording of that, because that is not the intent. The intent is that if you are keeping track of how you are distributing a PET drug, and you are distributing it to another department, you know, you are sending it from where it is made up to another department, that that receiving facility is who the receiving facility is in that particular case, but if you are sending it outside, it can be anywhere else is also the receiving facility.

DR. CALLAHAN: Since you raise the issue of distribution, and the control by prescription and practitioners and State licensed, don't you see that you have stepped over a line where a manufactured drug or even a compounded drug by your version of what that means is now getting into clearly a pharmacy practice, State license issue.

All of this has to stop at some point, and then pharmacy can take over. The manufacturing process goes to a certain point, at which point I would say that this group, you know, defers to the pharmacy boards and state control.

Here, you are in a federal document on manufacturing and producing PET drugs, we now have stopped verifying prescriptions in the licensure of the practitioners. To me, that is a major step beyond the process.

MS. AXELRAD: I think that is sort of a variation on the same comment we just had, which is that the use of the word "compounded PET drugs" is causing a problem, and we can revisit that.

DR. CALLAHAN: Not even compounded, it's in the prescription issue and the state licensure.

MS. AXELRAD: But that has to do with compounded. The only reason those things are in there is because that is part of the definition that Congress gave us on what a compounded PET drug is. That is how you distinguish a compounded drug from another drug. It's an individual prescription for an individual patient, you know, by a pharmacist licensed under state law.

That is the classic way they did in the compounding provisions, too. That is just Congress gave us here. However, distribution issues, if somebody is in a--I don't care if they are an academic PET center or they are a commercial PET center, and they are making the stuff, and they are shipping it far away, somewhere else, we have legitimate concerns, as we for any kind of a manufacturer, to make sure that the product that is shipped is done in accordance with provisions that will make sure that the patient is getting the right drug and that it is not deteriorated in quality by the time it gets to the patient, and we think that those are legitimate concerns for whatever CGMPs we write.

DR. CALLAHAN: That certainly is legitimate, and a manufactured drug being shipped to a practitioner for use, and when we ship a manufactured drug to a hospital, you certainly have something to say about that, but a manufactured drug that is being dispensed on a prescription from a pharmacy, once the tablet or the injection goes from the manufacturer to the pharmacy, then, there is some distribution issues.


DR. CALLAHAN: But once a prescription comes in and it is dispensed to the patient, that is a little bit farther.

MS. AXELRAD: For PET, it is a little difficult to draw that line since, in some cases, it goes from the cyclotron through a tube into a patient. I mean where does the manufacturing there end and the dispensing under practice of pharmacy pick up?

DR. CALLAHAN: That is a tough one, but there are easier ones we could address like the drugs that, in fact, the manufacturing process does stop, and the pharmacy practice does begin. I mean that is clear for other drugs other than ammonia or water.

MS. AXELRAD: I would like to hear your views on that. I would really want to take off the table this compounding issue. It seems like we really had a specific intent here to try and deal with a concern that we have had overall, of trying to give meaning to the language that Congress gave us about relevant differences between for-profit and not-for-profit, and also big operations and little operations. You know, somebody who is just serving patients within the PET center, you know, six, 12 patients a day versus somebody that is setting up a PET center to do a widespread distribution, we are struggling with how to deal with those, and the concern that the one size fits all CGMPs for PET makes it very difficult to distinguish between those situations.

Again, we are dealing with PET as we know it now. There is a concern that with all of the work that is being done on PET, that huge commercial manufacturers are going to spring up and take advantage of this, and we are having a hard time figuring out why those people ought to be following special CGMPs instead of, you know, 210 and 211, and we have also had questions raised by traditional radiopharmaceutical manufacturers that why should they have to follow 210 and 211, why don't they get the special GMPs, and what we are trying to do is design a current good manufacturing practice regulation for PET that we think provides adequate protection for the industry for which they are meant and we are trying to define that.

DR. CALLAHAN: And I think the key is the language because certain terms mean different things to different people, and certain groups have ownership of certain terms and have a long history of defending those positions, and I think when we see prescription and licensed practitioner and state regulation, and all these other things in a document like this, it just--

MS. AXELRAD: Let's get that off. We will try and stay out of compounding to the extent that we can.


MR. KUHS: Could I ask just one question of clarification before we do that? Does the agency view, then, that at the culmination of all of these regulations, that there will be actually two classes of PET drugs, there will be a manufactured PET drug and a compounded PET drug? Is this term "compounded PET drug" going to survive the new regulations, and if so, I think that is where the concern is, is there actually going to be two classes of a PET drug?

MS. AXELRAD: We don't think so. The statute is clear that CGMP requirements that we are developing will apply to all PET drugs. So, whatever this regulation looks like at the end, it will apply to all PET drugs, period.

MR. SWANSON: So, I will go back to the basic question then. What are the differences in the way that you are going to regulate not-for-profit entities versus for-profit entities?

MS. AXELRAD: Well, the statute tells us that we should take into account any relevant differences, and we would be interested in hearing from you as to what relevant differences you think there are, because from my perspective, if I am the patient lying on the table getting a PET drug, I don't really expect to get a lower quality PET drug because it was produced by a not-for-profit PET center than if it was produced by a for-profit PET center.

So, we are having a little trouble identifying relevant differences from a scientific standpoint, and if you have suggestions--we have talked about this I think at previous meetings--if you have suggestions for what those relevant differences might be, we would certainly like to hear them.

MR. SWANSON: Personally, I agree with you. I think it is going to be hard to sell to the public two sets of standards for the production of PET radiopharmaceuticals, and you have heard me before. I think where your differences are going to come in is how you require these people to register with the FDA and what kind of fees are going to be associated with whatever submissions you require is probably where you are going to have to cut, make the differences.

MS. AXELRAD: Well, we think we have addressed the fee issue. We will talk about that later today. The registration issue, all registration means is that they have to tell us that they are in the business and what drugs they are making, and we will be inspecting them.

But even if they didn't register, if we knew they were in existence, we could inspect them. Clearly, we are not going to be going to all this trouble to put our regulations that we are not going to be inspecting against.

So, I think that the issue of registration and inspection is one that we don't really have it on the agenda for today. We can talk about it if you want to, but basically, I don't think that there is going to be a lot of room there.

We are looking at the registration form. We don't think it is a big burden. We understand there may be some problems with that form for certain PET facilities in terms of some of the boxes or whatever don't fit, but the reality is all the registration form is, is that you tell you that you are in the business of making drugs and what drugs you are making, and then the inspection requirements come along anyway.

MR. SWANSON: Well, I think it goes beyond that. Again, if you require my institution--I said this at the last meeting--to register as a drug manufacturer, and it is going to be subject to all FDA's other requirements placed on drug manufacturers, and I will give you the best example of the problem that I see, is the physicians at the University of Pittsburgh Medical Center, who work for the medical center and, hence, are now agents of that manufacturer, and you are not going to allow our own physicians, then, to use approved PET radiopharmaceuticals for off-label uses because as agents of the manufacturer, they can't promote that.

So, fundamentally, a medical center is not a drug manufacturer. This is the argument that we made 10 years ago. That is not what we are, and to require an academic medical center to register as a drug manufacturer, you know, I don't understand. Have us register as something else. Have us register as a PET drug preparation center, okay, that's fine, but as soon as you start putting us in the role of a drug manufacturer, then, you make us subject to all other types of requirements related to drug manufacturing.

Again, let me emphasize, we are an academic medical center, we are not a drug manufacturer, period. We are not a pharmaceutical company. That is not our role.

MR. HOROWITZ: I am David Horowitz with FDA's Chief Counsel's Office.

There has been a lot of confusion over the years about what registration means, and I think it comes in part from confusion about the similarities or differences between state licensure and FDA registration.

FDA registration really is a way just for FDA to get the addresses of the facilities that, in this case, are making PET drugs and find out what drugs they are making. There is no connection between registration and other FDA requirements. That is a myth. I don't know why it is so pervasive.

You would be subject to the FDA requirements without regard to registration, and whether you register or not will not change your status. Now, you don't register as a manufacturer. Registering doesn't mean you have conceded you are a manufacturer and given up any kind of rights. You register with FDA, you don't register as anything.

MR. SWANSON: I just want to make sure those comments are in writing in a public record, because the first time an FDA inspector comes to my medical facility and attempts to regulate me as a quote, unquote "manufacturer" on issues such as promotion of alternative uses, you have got a big problem on your hands. Okay?

MS. KEPPLER: I guess one of the questions that I didn't hear from David, though, answered, was registration doesn't start the clock rolling or the rules rolling as far as promotional things, but would academic medical centers that are producing these drugs, would their physicians have the limitations that Dennis was describing?

MR. HOROWITZ: The limitations will not be determined or affected by registration. The limitations will be determined by the nature of the drug products and whether those drug products qualify under the statute for the various requirements or whether they are exempt from them.

So, if there are advertising restrictions that apply to drugs, to the extent that PET drugs fall within the scope of those requirements, they will be subject to those requirements, and they always were. Whether they register or not is simply not relevant to those other requirements.

MR. SWANSON: So, now we are arguing semantics. So, if I register, that is not an issue, but if I file an NDA or an ANDA for the drug, then, I am subject to issues such as promotion.

MR. HOROWITZ: I don't think it works that way either, but we can talk next time. I will be better prepared, and we can talk about specific promotion issues, if you have specific questions about what promotion requirements will apply and what triggers them, I would be happy to answer those questions.

But one thing I can tell you today is that promotional requirements are not linked to registration. If you are going to be subject to any kind of promotional requirements--and I am not sure whether you will or not--but it will not be affected by registration.

MS. AXELRAD: And this meeting is being transcribed, so we will have a written record of these statements.

DR. CONTI: I want to get back to this compounding issue one more time and the definition of a receiving facility.

Can you describe to me what a facility is called if they accept a manufactured PET drug?

MS. ROBERTS: They are also a receiving facility.

DR. CONTI: So, the issue really is to replace the word "compounding" with a "manufactured drug" because there is a very big difference in--

MS. AXELRAD: No, the idea would be to just drop the word "compounded" and we will just use PET drug, which is what we were doing before, before we introduced this confusion into this. You know, you just delete the word "compounded" and we will deal with the definition, and the deal is we are talking about PET drug, period, any PET drug.

DR. CONTI: The other thing I wanted to just point out, I picked up on something that you mentioned earlier, and I think that we have to make sure we all understand that the statute does not imply that PET drugs, because they have been or are compounded in the past or currently, is the rationale for new CGMPs or for unique CGMPs. It is the unique nature of the PET drugs as the rationale, so that distinction needs to be--everyone needs to be cognizant of that.

Is that clear? I mean am I confusing you? You made a statement earlier that manufacturers are complaining why are they not allowed to have shortened or simplified CGMPs for their manufacturing process, and that was in the context of discussion of the compounding issue.

I want to make sure that we don't confuse the two issues and make sure that we understand that the reason that PET drugs have been carved out in the statute is because of their unique nature, not the fact that they have been compounded.

DR. BARRIO: We, as a group, over a period of probably two years, tried to define what the same facility means. We can use the example the University of California. You can be very lax and say, okay, I could send radiopharmaceuticals to the same facility within the same institution, therefore, you will be without commercializing, sending from San Francisco to San Diego, and you are still preparing the compound for your own subjects or patients.

Then, that definition is very crucial because I am not so sure we really have a clear idea what that definition should be. Certainly, we know it is not the same building, clearly, it is not the same building, but it could be defined as the same physician group, for example, nuclear medicine center, that could be in the main hospital or in any satellite hospital. That may be one mile, two miles, 10 miles, 20 miles from where the main facility is.

That has caused significant amount of confusion among ourselves because many people believe that the term nonprofit refers to the entity. I have received many phone calls saying, you know, I work in that private hospital, and this hospital is for profit. Then, I am considered to be a commercial manufacturer when I use the pharmaceuticals for my patients.

The way I understood this to be is that what this not for profit is the production of radiopharmaceutical to that particular group or entity, in other words, the cyclotron facility is not selling to the hospital the radiopharmaceuticals to make a profit, but just to cover the cost. That is what our understanding of not for profit is, and will differentiate this from the commercial entity, of course.

But the definition of the location of the same entity, hospital or whatever it is, still is an issue that needs to be discussed because we might restrict PET centers that operate in different sites without realizing that this may be happening. That is something that we need to think of.

MS. AXELRAD: I think when you focus on some term here like receiving facility, the only relevance or meaning that it has is how that term is used here in the regulations. I wish I had a computer, so I can see the places, you know, easily pull up the context in which it is used, but the only main one that I saw is the one that we talked about earlier, which is you have to have the name and address of the receiving facility, you have to keep a record of where the drug is going when it leaves the production area and goes somewhere else. That is all it is.

So, we are not trying to deal with this whole issue of for profit or not for profit, or any of these global issues by using that term receiving facility, so whenever you focus in on these definitions, the only meaning they have is whatever context they are used in the regulations, and we are just trying to define it, so that is clear when we say you want the name and address of the receiving facility, what it is we mean by that.

DR. CALLAHAN: So, the issue of stratifying a facility relative to commercial or not-for-profit is not address here at all then.

MS. AXELRAD: That's right.

DR. HOUN: Just inherently, it may be more complex for people who are doing commercial distribution and getting these names and addresses of receiving facilities and tracking. That might be more of a problem if it's in another state than it's just upstairs or downstairs or across the street.

MR. KUHS: There is a couple of issues that we have here. The first one is that FDAMA stated that you had to take into account the relative difference between the not-for-profit and the for-profit institutions, when, in face, everybody around this table recognized that it wasn't the for-profit or not-for-profit status that was in question, that it was in the relative difference between an academic institution that prepared drugs solely for the use within their own institution, however that is defined, and the institution or facility that prepared drugs for commercial distribution.

In fact, we had just exactly that scenario play out in Peoria where we administer the NDA for Peoria, and the inspector wanted to know the names and addresses when we had taken a prepared vial for of radiopharmaceuticals and sent it to a radiopharmacy for redistribution to authorized recipients. The inspector asked for the names and addresses of the people who got the prepared drug in the unit dose form, when, in fact, using traditional practice of pharmacy and medicine in manufacturing segregation, the manufacturing process ended at the preparation and quality control of the bulk vial, and the redistribution of that product was controlled under practice of pharmacy, and subsequent to that, we were not required to give the names and addresses of those patients because it wasn't within our purview to do that. We didn't know who it went to for final patients.

I think this is the issue that we are struggling with - how do we take FDAMA, which says for-profit and not-for-profit, and apply that to academic versus commercial preparation of the drug.

MS. ROBERTS: I am a little confused. Maybe you can help sort some of this out for me then.

From my standpoint, writing a regulation, I am writing a regulation for a drug product. I expect that drug product to have all of its quality attributes that it should purport to have.

Part of what we want to regulate is how it is manufactured, where it is manufactured, who is manufacturing it, making sure that you have appropriate resources, facilities, to produce this drug to have the desired end product. In order to do that, there are certain laboratory controls, production and process controls, and other things that I would expect that any drug product that a person, a human is going to receive should meet.

I don't think for the purpose of making a regulation for the drug product itself, that there is a relevant difference between an academic institution, a for-profit, a not-for-profit. I looked at it as the drug product, and what I think is a reasonable minimum standard that we should have to make sure that that drug product at the end comes out the way it should be meeting the specifications that you have set up for it to meet.

Now, is there any reason why an academic institution or a for-profit or not-for-profit would have problems meeting these regulations?

MR. KUHS: I think there are relevant differences between the scope of the practice, and the scope of the practice makes a big difference in how many doses you are preparing for how many patients on a particular day.

If you are making a batch of FDG every two hours, as large as you can make it, and redistributing that product, the scope of that practice is a lot different than making a single batch of FDG that may be administered to one or two patients, and the scope of that practice requires a different regulatory strategy. It doesn't require different drug standards, it requires different regulatory standards.

MS. AXELRAD: Could you elaborate on that? What do you mean by different regulatory standards?

MR. KUHS: This is something that I have struggled with a for a long time about how you could make that relevant distinguishing distinction. The thing that I see that is most easy to accomplish is some type of regulatory discretion within the FDA as to how it examines the CGMPs process in an institution that has a relatively small and limited resource base versus a commercial manufacturer, which has a larger resource base and has a much larger scope of practice.

I don't know where that dividing line lies, and I think there have been a lot of attempts to do that by saying, well, it's prepared for patients within their own institution, and then you have the problem, well, what's your own institution, if you are a member of Kaiser, does that mean that your institutions stretch all the way across the United States, and that is your own institution, or does it mean the same building, which clearly that doesn't work. Does it mean two miles? I am not sure that you even want to put something like that in regulation.

But there are relevant differences between commercial preparation for distribution and preparation for within a single institution or the patients of a single institution, and I think that is where we are all struggling here.

MS. AXELRAD: I would like to get to the words of the CGMPs themselves and the requirements that we have in here. So far, I have heard one thing that you said, which is that once you finish producing the PET drug product and you ship it out somewhere, you don't know, you are essentially sending it to a distributor, and you don't know where that distributor further sends it, you know, like what patients are going to get it, but you presumably know who you ship it to, you know, when you put it on a plane, you know who you are shipping it to.

So, the issue is where we draw the line as to holding the PET center responsible, you know, how far down the distribution chain they are held accountable in terms of knowing to whom it is going.

That is a very specific issue, but I would like to focus on the other words in here to see what other things in here, you should have adequate laboratory controls, you need to make sure that you use the right equipment, that your equipment is calibrated. I really don't see having different requirements of that nature depending on whether you are doing one vial or 10 vials.

That is where I want to focus the discussion.

DR. CONTI: I think part of the issue is that I could draw an analogy of making 10 vials of FDG or doing 10 activation studies in a patient or 12 activation studies running the cyclotron 12 times in an academic institution.

So, the actual numbers produced at a single site per se is not necessarily a discriminating factor. A possible discriminating factor to consider is the infrastructure required to operate multiple centers, whereupon, you now have to impose a regulatory structure to make sure that all those multiple sites are in compliance. It is a different level of regulation above and beyond what would occur at a single site, a single cyclotron, or production center, because I think you are going to have a very difficult time discriminating between turning on a cyclotron for one issue that is regulated versus another issue regardless of how many times one does it or how much you actually make.

MS. AXELRAD: The whole idea here, if it's a commercial manufacturer of a radiopharmaceutical, they are under 210 and 211. Congress has said do something special for PET, and we are doing that, we are not using 210. It is significantly different than 210 and 211, and we think that when you see the guidance document, you know, we are making a really good-faith effort to really differentiate PET as we now know it from commercial radiopharmaceutical manufacturing, and we believe that is sort of the real distinction of why there is a separate section carved out here. The radiopharmaceuticals were dealt with in a separate section of FDAMA and PET has its own section of FDAMA, and so we have special CGMPs for them because we were told to do that.

So, we are already making that distinction just by developing a whole new set of regulations that are going to apply to PET. But I really would like us to focus in on the words here and see where, you know, other than the one that you just mentioned, what are the specific problems here.

That is what we did the last time, and you pointed out a lot of issues, so I would like to get into the specific language here.

DR. HOUN: One option is to start with Subpart J on Distribution, because that is the issue about once it leaves production and going to more complicated distribution systems versus upstairs.

If we look at 212.90, the control of distribution of PET drugs, we would love to hear your comments on this section and how it might be different for larger manufacturers versus academic institutions.

MS. KEPPLER: One thing, too, that we should keep in mind as we are going forward, too, I think where the concern is, is that over the time, without any differentiation between a commercial production that might be just a powerhouse, you know, sending out thousands of doses a day versus an academic center that is doing their four or five patients, is that somehow the recordkeeping standards over time for that academic institution that is doing four or five, is going to raise to the level of whatever recordkeeping standards are appropriate for something that is, you know, piling out thousands of doses.

So, it might be something that isn't different in the regulations, but is differentiated in the guidance because I think--and you guys correct me if I am wrong--I think that is the concern that people have is that somehow the academic sites are going to be, you know, down the road, be at the same level of a commercial site as far as whatever, you know, recordkeeping requirement.

MS. AXELRAD: When we are making comments on the regulations, let's assume that these are the CGMPs that would be required for the smallest academic PET center, and what problems do they cause if they are applied in that particular situation.

Then, take it to the next level. Let's say it is an academic PET center that is doing some distribution of the drug small scale, like leftover stuff might be being shipped out, you know, then, what problems would this regulation have if it was applied to them, and then get to the bigger size company.

I mean really we are not trying to write this regulation for the larger commercial people, and we don't exactly know where to draw the line. To be honest, we are going to struggle with that for a longer time, but we are writing this for the small academic PET center, but we believe that it may be acceptable to cover most situations, so let's focus when we do this on that.

DR. CALLAHAN: To go where you suggested, at 212.90, I mean I see a simple change that might relieve some stress. In (a), "You must establish, maintain, and follow procedures for the control of distribution of PET drug products shipped from the PET center to ensure that ... approved for use," if you strike the next phrase, "that prescriptions are reviewed to assure that they have been properly filled," that doesn't belong in there for any reason.

Then, you could include the last part, "the process of shipping will not adversely affect the quality," we could see that you are not going to ship this in an improper container, but that phrase in the middle is irrelevant.

MS. AXELRAD: That relates to the compounding issue. We hear you on that.

There is somebody in the audience who has really dying to say something.

DR. HUNG: This is Joseph Hung of the Mayo Clinic.

I think, the FDA, that you just mentioned that you have difficult time trying to differentiate the difference between a small PET institution versus commercial manufacture, but I think if you read the Section 121, that Congress already gave you a very clear indication there, the difference between those two facilities.

One is the not-for-profit and compound the drug for their own patient versus commercial manufacture, which is for large-scale distribution and also patient use. I think that is already very clear direction from the Congress.

The other comment, if I may, is the definition of the compounded PET drug under the Section 121, talking about practitioner, and I want to ask you whether the practitioner in your definition include pharmacist or not.

MS. AXELRAD: Thank you for your comment. I hear you. I don't think it is anywhere near as clear. We have done some legal research on the terms used "for-profit" and ot-for-profit," just for an example, what is PETnet, you know, what is PETnet, when PET net is operating a facility for an academic PET center. It is really not very clear. The terms for-profit and not-for-profit are used other places, too.

Anyway, it is not clear, and I really don't want to get into the issue of the practitioner. You have sent us a letter with some questions, and we will be trying to answer those questions and get back to you on that, but I really want to keep the focus of this on the language of the GMPs, so if we could proceed.

MR. WATKINS: I am Len Watkins from the University of Iowa.

I have a problem with the definition of a PET drug that you have here. This is what we are discussing, and this definition here doesn't make any sense to me at all.

MS. AXELRAD: I think it's a statutory definition.

MR. WATKINS: I mean we are talking about apparatus, computer programs. That is not a PET drug. If you take the definition and you stop at the diagnostic imaging part, I think that is an acceptable definition, but to include all these other things in there, it seems to me not a definition of a PET drug.

MS. AXELRAD: Like I said, I can't help that. Congress gave it to us. That is a verbatim definition from the statute. If you look at the statute, Section 121(a)(ii)(2), that is what Congress told us, so we really can't do anything about that. All we can try and do is, in our implementation, make it make sense when we apply it.

DR. HOUN: Can we go back to comments on distribution?

MR. SWANSON: I think one of the issues that came up last night that we probably need to get clarified under Distribution, is it may be especially for certain PET drug products that you are going to want to start the distribution process prior to actually releasing them in view of their short half-lives, it may be that you are completing some testing procedure and you want to get them in the distribution cycle prior to completing that.

Now, the way I read this, it would actually permit that because it says, "You must establish, maintain, and follow procedures for the control of distribution to ensure that only those products approved for release are used."

So, I suppose under this you could have procedures that would allow you to do that, but I think we just need to get that on the table, so that people here understand that that would be possible, and that is your thinking also, okay?

MS. ROBERTS: That is exactly what is intended by culminating out release and distribution. You may begin to distribute the product prior to release of the product, and by "release," we mean somebody with authority to do has reviewed the batch record, reviewed all of the testing, and has decided that it's okay to be given to a human patient.

Now, we understand and have taken into account that you need to ship the product that has short half-lives. That is why there is this provision for a procedure to control it, that you have whatever procedure works for your particular institution that you can track and make sure that you have some way of retrieving product if it's not releasable or communicating to the receiving facility not to use the product, it's okay to use the product.

MR. FERRIS: So, you really mean prior to human administration.

MS. AXELRAD: If you look at the way release is defined, that is exactly what it says, release, "to permit the use of a batch of finished PET drug for patient administration."

MR. SWANSON: The release definition probably needs to be changed a little bit. It says, "Release means," and if you go down to the last, "finished PET drug product for human use," because understand that your regulations apply to research, so it may not be a patient.

I have a lot of specific comments. I don't know if you want to go down these one by one. Would that help you out, to see if anybody has got specific comments, we can bring them up. Some of them are probably not worth lengthy discussion on. So, if you want to do that, maybe we will just start with Subpart A.

I have one general comment. I think you need to be careful. You define a lot of terms, but then you don't use those defined terms throughout your document, so I think you need to go back and clean up the document a little bit to make sure that you are not introducing terms into your subsequent statements that aren't previously defined or conversely, that you are consistently using previously defined terms throughout your document, and I realize that comes with cleaning it up.

MS. AXELRAD: Well, it would be useful for you to point out to us anyplace that you think we haven't defined a term that you think needs to be defined. We put up-front that we need to add definitions of active ingredient, PET center, and maybe not prescription, if we take out the word prescription, but any terms that are not self-explanatory here, please tell us and we will provide definitions.

MR. SWANSON: Subpart A, 212.1(a), acceptance criteria, "decision to accept or reject a unit, lot, or batch," you don't define a unit, for example. I don't know what a unit is.

You have got to be a little bit careful like in your NDA submission, you talk about sub-portions, okay. In other words, your document is just going to have to be consistent.

MS. AXELRAD: I wish we could all agree on this. This is another issue that I swear we have talked about at every single meeting we have had, you know, about batch and sub-batch, and whenever we were talking about the USP monographs, this issue.

If we could just agree on some--we ought to be able to just agree since this is not a controversial issue. If you would just suggest what the right terminology is, we would be happy to try and put that in.

MR. SWANSON: It's a quality control lot. Your definition of lot includes a portion of a batch, that's okay. So, basically, your QC sub-portion, sub-batch is a QC lot of your drug. It is a lot of your drug that you use for QC purposes. Well, actually, it's a little amount of a drug, but you would call it a lot of a drug, okay?

MS. ROBERTS: So, are you saying that each, what we have called a sub-portion of batch per se, technically, you just call a different lot, you are going to give a different lot number to each?

From what I understand, like just when we are talking about the water or even a gas that is administered when there is no particular vial to call a batch or a lot, and we were talking about the testing where you need to take samples of it like beginning, middle, and end, what do you consider a lot, a sub-lot, a batch? That is what I don't understand.

MR. SWANSON: I am trying to put it back into your defined terms. You defined a, "Lot means a batch, or a specifically identified portion of a batch."

MS. ROBERTS: Okay. That is enough for you. You do use specifically identified portions of a batch.

MR. SWANSON: Well, we refer to them as QC sub-batches and by various names, and we will just have to use your terminology, how you defined it, okay.


MR. SWANSON: But you need to yourself use your terminology how you defined it. Okay?


MS. AXELRAD: We found in our experience in going over this, that we could spend most of the meeting going over the definitions. We might want to start with Subpart A, and if we get through one or two subparts, we can all take a break.


MR. SWANSON: Well, I will keep speaking up unless somebody jumps in. If you go down to (m), under Definition, "Quality control means a system for maintaining the quality of drug substances, drug products, intermediates, raw materials, analytical supplies, and other components"--can't you just say drug product, doesn't drug product mean all other ingredients anyway?


MR. SWANSON: Can you just say components and ingredients, because components you define as everything else. It is confusing because you have, "Quality control means a system for maintaining the quality of"--blah, blah, blah--"container-closure system and in-process controls, through procedures, tests, analytical"--I guess what I am saying shouldn't it read, "container-closure systems, through procedures, in-process controls, tests, analytical methods," aren't your in-process controls part of your quality control?

MS. ROBERTS: That is where we define that indeed in-process controls are part of the quality control.

MR. SWANSON: Okay, but you are not maintaining--it's a semantics thing--you are not maintaining the quality of the in-process controls. In-process controls are part of your quality controls.

MS. ROBERTS: That is exactly what it says quality control means. We hear you.

MR. SWANSON: Okay, I am with you.

MS. AXELRAD: We can consider whether we could just say components and ingredients. Drug product usually means a finished dosage form, period. That is sort of a term of art we use throughout the regulation, but we can look at the other list of things in there and see if they are necessary.

MR. SWANSON: Okay. Does anybody else wish to comment on the definition of theoretical yield and whether or not you can actually do that?

DR. HUNG: Could I ask Tracy to, under Section 212.1, to define manufacturing versus compounding, and also practitioner, please?

MS. AXELRAD: I think, as I pointed out at the beginning, this is really a meeting between the participants at the table, and we will provide an opportunity as we sort of finish sections of this to give people in the audience a chance to make comments, but I really want to keep it focused on the wording of the GMPs, and not get off on some of these other issues.

Would you address the theoretical yield question? Is that a problem?

MR. SWANSON: Yes, I am not sure that we can do that. I mean taking into account the nature of PET production and the fact that our--I mean if you were to say a range of theoretical yield, then, we could probably could that, but that doesn't say that, at least that's not the way I interpret this.

I mean to define a specific amount as a theoretical yield is going to be difficult to do.

MS. ROBERTS: How do you determine, then, what your specific target range is if you don't have for your process a theoretical yield, one number that you think, normally, it's 100 percent, now, what theoretical yield would you, from your known process, you have a process, you know how it works, I would assume that you put certain portions of (a), (b), and (c) in, you are going to get a specific portion of (d) out, and then based--is that impossible for you to do?

DR. BARRIO: This is, of course, okay in a chemistry process where you start with a known amount of a precursor, let's say a known amount of ion or something like that, but if you include the target process here, there is no way to define the theoretical yield, because the target is not the sign, different targets will produce different yields, and the age of the target will also produce different yields or the precursor.

Then, if you define the theoretical yield beginning, not with the target production or the precursor, but rather with the amount of precursor you have in the vessel for that reaction to occur, I think the second one is possible.

MS. ROBERTS: So, then, we just need to distinguish between where we are starting your theoretical yield from.


MS. ROBERTS: Either it's from the beginning of the cyclotron process, what you are saying depending on the target, but wouldn't that give you invaluable information that if your target is starting to deteriorate, you are not getting up near your theoretical yield?

MR. KUHS: But that is perfectly normal to happen, and it's an expected occurrence.

MS. ROBERTS: I am not saying anything about it not being normal or expected.

MR. KUHS: The problem is with a theoretical yield, you could have a theoretical yield anywhere between 25 percent and 90 percent of your starting precursor, and all of those are acceptable, and you may get 25 percent one day and 90 percent the next day.

MS. ROBERTS: Isn't there a predescribed theoretical yield when you know your target is 100 percent and everything is set up the way it should be?


DR. CONTI: There may be an expected range of yields that either through experience or through the manufacturer, you can obtain basic data on what you would anticipate if you operated under these configurations for that particular device. That might be something that we could use. The range has to be broad enough, and it is actually really critical because that is going to be part of your criteria for accepting or rejecting a batch. If it falls below some certain level of yield, you probably will reject it, and that will tie into your process.

MS. AXELRAD: Do you do that now? I mean that is the question. I mean this concept of theoretical yield is something you use traditionally in production of regular pharmaceuticals, and it is a way of showing whether your chemistry process has worked the way you thought. You know if you are way off, you have got a problem.

DR. CONTI: These decisions are not made based on theory. It is based on what is the anticipated expectations of the device that you are working with, the target systems, the precursors, as Dr. Barrio mentioned.

We have a track record established for whether we are going to make or break a particular batch as to whether to accept it or not accept it, so that type of data would be I would think more relevant than theoretical yield.

MS. AXELRAD: I going to let Ravi, our chemist, say something about this.

DR. KASLIWAL: I think we have discussed with you a couple of times, there are two issues. One is the percent yield that is calculated on the basis of the corrected known or expected amount or radionuclide, starting radionuclide that is synthetically incorporated. I think especially for master record localization, some of those things, it is important.

The second is the target yield. Based on your historical data, you are right, you probably have to have a range of yield there--okay, we will take that into account, some action level.

MR. SWANSON: I will keep going. You define verification, and I am not sure what the difference is between validation, and then you go on, and throughout your document you never use verification anyway, so I am not quite sure why it is there.

If you get down to Subpart C--do you want to keep going or do you want to stop at Definitions? Has anybody got comments on definitions? Is that okay?

MR. BRESLOW: Ken Breslow, PETnet Pharmaceutical Services.

In terms of the words units, lots, and batches, I don't think we need to dwell on this. I think we could probably all agree that a batch is the vial of product you get out of the synthesis regardless of what the product is.

A lot can be a subdivision of the batch. Now, let's say a commercial vendor is not operating as a pharmacy, but as strictly a manufacturer, and he gets a vial of 2 curies of FDG, and he wants to sell that to three dispensing pharmacies. He will subdivide that vial into three additional vials. Each one of those vials is a lot.

A uniquely identified fraction of a batch or a lot that is taken for QC is just that, a QC sample. The word unit usually refers to a product where you are making 10,000 vials of a batch, and a vial is a unit.

So, I think if we stick with those simple definitions. When we are talking about the ultra short half-lives, like water, when you are doing a consecutive series of O 15 water, you have got your test batch and then subsequent batches or you could term them sub-batch if you wish to make the entire number of units for a given patient or number of patients. That definition can be open for discussion, whether you want to call it a sub-lot or a sub-batch, but I think that is a unique circumstance with the ultra-short life.

When we talk about yields, I think we have talked about yields several times, and I think Ravi has a good understanding of the yields. In the recent past, the FDA has relieved certain control requirement limits on cyclotron yields and radiochemical yields in regard to the Peoria NDA where there used to be action and control limits for cyclotron and radiochemical yields, and the FDA has relieved the requirement of control limits for cyclotron yields and radiochemical yields. We are specifically talking about FDG at this moment.

I think yield again is impossible, theoretical yield is impossible to calculate. You are looking for a radiochemical yield that is historically within a range that you are known to perform at, and if the radiochemical yield falls below that range, it doesn't necessarily require you to reject the drug product. It requires you to take note and perhaps investigate if it's below your action limit.

Let's keep the term yield in perspective of what we are talking about, and as defined in your proposal, it is very misleading and not appropriately defined.

I also want to make one other comment about different types of commercial ventures as far as the distribution and where the manufacturing ends and where pharmacy practice picks up, for example.

If there is a PETnet, for example, a facility that is operating as a pharmacy in addition to being a manufacturer, and the product has been manufactured and the quality control testing has been done, and the sampling has been completed, now, this material is going to be dispensed on the other side of the room, so to speak, by the pharmacist, and that person, the pharmacist might be the same person that has produced the drug, but it's a pharmacist.

I would suggest that the control of the manufacturing processing ends at the time that the QC sample is removed and the QC testing is done. Subsequent handling of that material would be under the practice of pharmacy.

However, if a distribution center was not operating as a pharmacy, but rather strictly as a manufacturer, where they would sell a vial, a batch, or a lot of a product to either a pharmacy or directly to the physician's office, then, the distribution would be encompassing in the definition of manufacturing, and the control of the distribution would be as ordinarily is controlled by the agency.

Did I make myself clear on that?

MS. AXELRAD: Could I ask you a question? What difference does it make? You know, I mean what difference does it make to the patient in terms of what they are actually getting whether you draw this line? Everybody is sort of forcing us or trying to force us to draw this clear line. What difference does it make?

MR. FERRIS: I would like to address that. The difference is that if you stop manufacturing at a bulk finished drug vial, and then turn it over to the practice of pharmacy, then the practice of pharmacy, under the State law, is able to distribute that drug in patient unit doses accordingly.

We ought not to have the regulations rescind the option of being able to prepare under manufacturing, patient unit doses, but under manufacturing, where FDA GMP would extend through, then, you folks would be coming in and looking at how we did our aseptic transfer, breaking down the bulk finished drug into patient unit doses under GMP, whereas, if a center chose to stop it at the bulk finished drug and turned it over to the practice of pharmacy, then, FDA stops there.

MS. AXELRAD: Okay. I hear you and I think I have a better understanding, but in terms of the patient getting a quality product, it doesn't seem to make a lot of sense to me.

I mean if there is a concern that in breaking down the product into, you know, taking a vial and breaking it down into unit doses, there is some concern that you could alter the sterility of the product or do something that could affect the quality of the thing, then, why shouldn't GMPs, why should it allow--let's take first a small center.

Okay. Here I am, I have my cyclotron, I have my operations. I get the stuff, you know, the sterile filter, you know, into the vial, and then I take off my manufacturing hat and put on my pharmacist hat and start sticking syringes into the vial to draw out individual things, and in doing that I mess up the whole thing.

Now, why should it be permitted that we are going to stop and we are going to say, okay, when you fill the vial it stops there, but as soon as you take off your hat and put on your other hat, we don't have any concern about that?

MR. FERRIS: I am not saying we shouldn't have concern about that. What I am saying is, is that when the pharmacist hat goes on, the regulatory authority moves to the States just the way it does now. If I receive, if a nuclear pharmacy receives a bulk vial of thallium and dispenses that thallium on prescription to specific patients, that is under the practice of nuclear pharmacy.

The same way is if a person received a bulk vial of FDG and they dispense it on prescription, that is one option of a way to distribute.

The other option of a way to distribute is in any drug manufacturing situation, there is the opportunity to be able to prepare and market unit doses for a patient. This is a patient dose. In that instance, the manufacturing process carries right through to the distribution.

MS. ROBERTS: I have a question regarding in that instance, when you are sending the bulk vial to the pharmacy, at which point does release occur? Does release of the product to be used in the human happen after the pharmacist has done what they needed to do and put it into unit doses, or does it happen before it reaches the pharmacy?

DR. CONTI: As soon as regulatory jurisdiction changes.

MS. ROBERTS: So, then you are saying that you have released the bulk product to the pharmacy, and at that point, then, you have actually released, it meets all of its criteria, and that it goes there.

DR. CONTI: That's correct.

MS. ROBERTS: It doesn't happen once they have already done that.

DR. CALLAHAN: If somebody is in there dispensing unit doses improperly and contaminating and reducing the quality of the product somehow, the State Board of Pharmacy has to deal with that issue, and the profession has to deal with that issue.

MS. AXELRAD: Why do you rather have them than us?


MR. KUHS: Don't ask that question.

MS. AXELRAD: I mean just as a possible patient who someday may be taking a PET drug, I mean the only thing I can think of is that you would really--you know, that they are not going to be in there looking at you as often or they maybe are not going to be as rigorous. I mean I really have a hard time understanding why that is.

MR. KUHS: This is a long, long, long fought, hard battle, and it goes back to the 1984 Nuclear Pharmacy guidelines that attempted to distinguish where the FDA's regulatory authority stopped and the State Boards of Pharmacy--and I think that those, by and large, have served us pretty well as definition guidelines, and I think we ought not try to tamper with that because that brings all of the boards of pharmacy and all of those other folks, and rightly so, into this discussion.

The manufacturing process is separate and distinct from the distribution administration process that is covered under practice of pharmacy, but it does not preclude making unit doses, and usually, the designation is that distinguishes practice of pharmacy and medicine from traditional manufacturing is that there is a specific patient in mind, and when there is no specific patient in mind, it becomes manufacturing.

If there is a specific patient in mind, it becomes practice of pharmacy and medicine when done through the proper channels and the licensed individuals within the State, and those have been fought through numerous court battles, so I don't think we need to try to redefine that.

DR. CONTI: I have a comment on the practice of medicine that I wanted to get an FDA opinion on because just picking up on what was said a few moments ago about a manufacturer preparing a unit dose for a patient, is this considered under the practice of medicine if it is done under the jurisdiction of a physician and does the FDA have purview in that transaction just as we are talking about the practice of pharmacy.

MR. FERRIS: I want to clarify that one thing, is that a manufacturer won't prepare a patient-specific dose. They will prepare a patient dose, but there is no name. In other words, it is a dose. This is a single administration, and a manufacturer can do that, but they don't prepare patient-specific doses.

DR. BARRIO: Then, what we mean is practice of pharmacy and practice of medicine can be depending upon we have a pharmacy or a chemist under the physician prescription dose, then we could use. For the record, then, we should understand the definition of practice of medicine and practice of pharmacy.

DR. CONTI: The same issue as is with the practice of pharmacy, you can receive, say, a lot, the physician can receive a lot and then draw from that lot a series of patient-specific unit doses, but again the manufacturing process stops at the distribution of that lot as opposed to monitoring what happens when the physician takes that lot and prepares it for individual patient doses, so it is exactly parallel to the practice of pharmacy.

MS. AXELRAD: I think I hear you. I think we need to figure out how, you know, exactly to draw this line. I think what I see what you are saying is the issue is where do GMPs stop and something else, the practice of--I will set my skepticism aside--and we will try and address that.

MR. FERRIS: We should be able to preserve the option of turning over, because there may be a PET center that does not want to register as a pharmacy.

MS. AXELRAD: Well, like I said, if you have some definitions, if you would like to try and--what we need is clear dividing lines, and it is going to be very difficult, you know, here, if you want to do both at the same time, you are going to cross the line, and then there are things like this is O 15 water that goes direct, or is it ammonia, one of them goes--O 15 gas that goes direct into the patient, you know, tell me where the line is.

Anything you can do to help us clarify how you would draw the line in words, I mean because that is the challenge. You know, we may all sort of say we know it when we see it, but we have to write it in words, and then we have to train our inspectors, so that they understand where they are supposed to be looking and what they are supposed to be looking at and where it stops.

So, we need to try and figure out how to focus on the words in a way that would articulate that. Believe me, it is complicated.

Are there any more questions on the definitions because I would like to take a break and then move on to the next section.

MR. KUHS: I would just refer you again to the 1984 Nuclear Pharmacy guidelines and to look at those definitions in there, because I think those can be helpful because this was really hashed out a lot.

MS. AXELRAD: We are going to be looking at that guideline because parts of it have been rendered obsolete by more recent legislation, and it needs to be revised, so we need to look at that. Again, unfortunately, it is where all these compounding radiopharmaceuticals and PET all sort of get together. You have to look at all the issues and how they all fit together.

DR. HOUN: But we would welcome, if there is a draft on this division, on how we would define this division between manufacturing and practice of pharmacy, if ICP wants to draft that, we would welcome that.

MS. AXELRAD: Where does manufacturing end and practice of pharmacy or medicine pick up, specifics.

DR. CALLAHAN: It is really not that difficult, I mean from our point of view. It is where a process-oriented activity and a product-oriented activity diverge, and we don't want the process controls concept, if I even understand that, to apply to the dispensing of unit dose radiopharmaceuticals on prescription. That is the fear. So, once that goes from a process-controlled manufacturing to a professionally regulated and State regulated dispensing process, that is equally focused on the quality of the product, but not equally focused on the process to get to that. I mean I don't think it is that difficult to define myself, and we will take a whack at it.

MS. KEPPLER: We will come up with some words.

DR. CONTI: I want to look at the manufacturing from the other end, too, because where does manufacturing begin is also an issue particularly as we get into the NDA discussions later on, because there are certain things in that process that you might want to exclude from manufacturing per se, because there may be production of raw materials, and it may not be necessary to have that within the confines of manufacture of a PET drug.

I was curious to see what the FDA thought about where this should begin and where it should end. If you read the NDAs, it is deciding how many micro amps do you use on the cyclotron. I am not sure if that is the right starting point.

MR. SWANSON: I just might say about pharmacy, also remember that the definition of pharmacy is based upon a physician-patient-pharmacist triad. That defines pharmacy practice, anytime that you have an external party that enters that, so if you have got a pharmacy dispensing a product to another pharmacy, that is outside that triad, that is manufacturing. Okay. The pharmacy is defined by that triad.

MS. AXELRAD: We went there, we tried that. That is where you got the prescription, the individual patient, and whatever, you know. Congress's way of describing that relationship was to call it compounding, and we spent the first half an hour of the meeting with you all telling me you didn't think we should go there.

So, your challenge is to write the distinction between manufacturing and the practice of pharmacy without using a reference to the prescription for the individual patient relationship.

MR. MOCK: Bruce Mock, Indiana University.

To pick up on some comments that Jane made earlier and her concern, a valid one, being should there be two sets of standards for the profit of not-for-profit, these kinds of issues. If you are the patient on the table, you want to make sure that you have got the proper juice being injected into you. We all agree on that.

We have traditionally, under pharmacy and medical practice, have tried to assure that by doing the end product testing. Now, the FDA's position has always been, well, we want to ensure quality, we want to build in up-front, so we are going to kick the regulatory powers in, in the process of manufacturing this material.

So, our concern is, we share the same concern about the quality being no different, but where the FDA kicks in is in the regulatory power of the process controls, and so whether I am a single individual making FDG for my own patient or whether I am manufacturing it for distribution to half a dozen hospitals throughout the city or throughout the state, the end product needs to be of the same purity.

Our concern is the way the FDA has traditionally said the way we are going to ensure that it is the same purity is require everybody to do the same type of production controls, process controls, process validations, therefore, I am no different than Eli Lilly in terms of what is expected of me to do to make sure that that--oh, by the way, do the end product testing, as well, we are going to require you to do that.

We have always wanted to do that in the first place, and the agency's position has been, well, no, we want to put the regulatory control in up-front. So, that is our concern, and then your definitions are so broad that I don't see how we can--the small unit, the small entity can be treated any differently than the big entity.

You see where I am coming from? I agree that the product should be the same purity regardless of how it is used, but where you guys kick in, your authority is in the process to get to that point.

MS. AXELRAD: But the reason we do that, the reason why we focus on process validation is that we have discovered through long experience in regulating all kinds of pharmaceuticals from small companies, big companies, generic drug companies, radiopharmaceuticals, non- radiopharmaceuticals, that end product testing does not always identify problems that got built in and introduced during the process. That is why we focus on process controls.

I would like to say that for a drug product that is supposed to be sterile and is injected into patients where you don't get the results of the sterility tests until 14 days after you have injected it into the patient, that process controls, particularly with regard to sterile processes, take on an even greater importance because there your one barrier of testing the drug to see if it is sterile, you are not going to have that before you inject a patient.

So, that is some of the reasons why I think we focus on process controls and don't rely solely on end product testing, but I think that we are trying to be reasonable and recognize that the process controls, when you are making two vials in a small room, is different than when you are making a million dosage units with a two-year half-life and shipping them all over the country, and that is what we are trying to do in this document is to try and make that kind of a distinction between the traditional 210 and 211 regs and something special for PET.

Dr. Hung, one more question, and then I think we are going to take a break.

DR. HUNG: Again, this is Joseph Hung from Mayo Clinic.

As you know, I am from Mayo Clinic, and patient care is our number one mission. I agree with you totally that patient care is the number one issue that we are talking about here today. We like to provide the best quality of product to the outpatient, but I think there is a reality issue, practical issue here.

You have to design a regulation to be practical and also sensible, and the other issue is risk versus benefit. You don't want to set up a very restrictive regulation and force those small PET centers to be shot down.

Let me give you an analogy. It may not be a good example, but let me try to do this. I mean you eat food at home prepared by your wife, by yourself. If someday the Department of Health implement the sanitary standard, the same standard that applies to the commercial restaurant, we would be forced to eat at the restaurant. You know what I am talking about.

So, I think there is a practical issue here that I wish the FDA should consider that, the difference between those two.

MS. AXELRAD: Why don't we take a 15-minute break and be back between five after and ten after 11:00.


MS. AXELRAD: We are going to get started again. My plan is to go through until probably about 12:15 and then we will break for an hour for lunch.

Go ahead, Tracy.

MS. ROBERTS: I would like to begin by starting with 212.2 and going through each section and seeing if you have any comments or particular problems with each section.

DR. CALLAHAN: What happened to 212.1?

MS. ROBERTS: That was the definitions.

DR. CALLAHAN: Subpart B, 212.1, what personnel and resources must I have?

MS. AXELRAD: That comes next. First is 212.2, which is right above that.

DR. CALLAHAN: I am sorry.

MS. ROBERTS: In 212.2, it is the basic question. You might have noticed that the format is different than other formats that you have ever seen for regulations before.

MS. AXELRAD: They didn't seem to notice that. They didn't give us any credit for that. It is supposed to be plain English.


MS. AXELRAD: We obviously didn't go low enough.

MS. ROBERTS: I would like to pick up with 212.2. What are the current good manufacturing practices for PET drugs?

This just goes over the scope of what we will be covering, manufactured, as well as compounded PET drugs, and the different sections that we expect for them to meet, personnel and resources, and we will be going over each of those sections one by one.

Are there any comments on this particular section?

Okay. 212.2. What personnel and resources must I have? Basically, this is a minimum requirement. Does anybody have any problems or questions with this one?

DR. CALLAHAN: Is n equals 1 adequate?

MS. ROBERTS: If that is a sufficient number of personnel, and they have sufficient background in order to enable them to perform their assigned function, if one person can do the job.

DR. CALLAHAN: So, there is nothing in these proposed GMPs that would preclude a single person operation for the manufacture or production of the PET radiopharmaceuticals?

MS. ROBERTS: No. The only thing that comes in, there could be one person--what I had envisioned for this is there may be one person if that is adequate for your facility, however, there are other provisions later on that require the review of the records, testing, review of the testing and release. If one person is adequate to do that, then, that's okay.

MR. FERRIS: So, the maker can be the tester.

MS. ROBERTS: That's possible, yes.

MS. AXELRAD: Let me turn that question around and ask you, do you think that it is adequate to have one person? I mean it is the collective view of the people at the table from the ICP, that it is adequate in some cases to have just one person operating a PET facility, period?

DR. CALLAHAN: I think in some cases, that is possible. It depends again where we define the manufacture process, but if it goes back to maintaining and documenting and everything in the cyclotron, maybe not, but in terms of the production of the drug product on a day-to-day basis, then, it is theoretically possible, I think.

DR. CONTI: Let me answer that in a little different way. Currently, it is adequate, under the current configuration with minimal recordkeeping and no additional amount of work. If you add this additional amount of work, I don't know the answer to that.

MS. AXELRAD: Well, also, keep in mind the title to these is current good--good manufacturing practices, and as I was just explaining to someone in the audience, and we have talked about this before at other meetings, there is a wide spectrum I think of operating practices in the industry, and I am not sure that we are trying to design regulations for the poorest operating facility.

You know, we are not trying to go to the absolute lowest common denominator here. We are trying to set what we think are current good manufacturing practices that are representative of what is used across the industry. Maybe not the cadillac, you know, 20-people operation, but maybe not the lowest with only one if it is not believed that you can adequately run a PET manufacturing facility with a single person.

DR. CONTI: Again, I think it depends on what level of recordkeeping and all the other constraints you place on that single person. An academic institution, the one person is completely adequate, not to do a lot of these as we currently practice.

But I just don't know based on this document how much more work is involved at this point, and I think it is going to be difficult for us to answer that, and I think you have to recognize that, as you said earlier, you yourself stated that these documents are really trying to get to the bare bones minimal configuration. That was how we were supposed to take this.

So, if that is the case, I can't answer your question yet.

MS. AXELRAD: I was just trying to get at whether he was actually suggesting that one should be adequate and that we should adopt that, because we have heard previously, I think, that--well, we have never really addressed the question of exactly how many people, but we have talked about that certain operating--what people consider to be well-operated facilities have several people, and they keep procedures, and they do a lot of what we have in here in terms of recordkeeping, so what I am trying to get at is if we are trying to find something that sort of is the way it is done, the average, what people would consider to be a well run facility, it may be that some of the facilities that aren't to that standard might need to do something to comply with it.

MR. KUHS: I think you need to look at again the scope of practice and about how many doses or batches are being produced in a single day. There are many academic institutions that prepare a single small batch for a relatively small number of patients, and one person is completely adequate to do that, but as you increase the scope of your practice, it becomes physically impossible to do all of the testing, all of the requirements, and things by a single individual, but by and large, it can be done with a single individual.

DR. CALLAHAN: And I think one of the restrictions, and where I was going with that, was that at any given time, you know, while a process is being performed, there may indeed be one person. That is not to say there aren't people in the facility, on the staff, that can provide other functions, but is it possible for a single person to be there operating this without running into kind of regulatory problems with checks and rechecks, and things like that.

MS. AXELRAD: I believe the way this is drafted, we have answered that question to say yes, we have heard a lot of the comments about the 2:00 in the morning runs and the difficulty of having a second person, and I think we have tried to make sure that where there is a double check, it doesn't necessarily mean a second person.

DR. HUNG: If I may, I would like to go back to Section 12.2. I know the CGMP applies to manufacturing and the compounding of PET drugs. I really don't like the word distribution because it is simply applied to manufacturing practice.

If we talk about compounding a PET drug, I think another word would be a better one, is dispensing. So, I think that dispensing should be included in that 212.2(a).

MS. ROBERTS: Dispensing is usually used with the practice of pharmacy, isn't it, and then we are crossing that line again? That's why we use distribution. Distribution, how we see it here is the physical movement of the product to the receiving facility.

DR. KASLIWAL: I think that dispensing here also could mean if you are a manufacturer, and dispensing into unit dose vial for distribution, that could be dispensing into unit dose vials. I don't know, you don't like dispensing term. That's packaging, okay.

MS. ROBERTS: I just wanted to make one blanket statement, though, that I am sure most of you are not used to being regulated by FDA whatsoever. Is that true except for maybe the gentleman from Mallinckrodt?

I work in the Office of Compliance, and I just wanted to give you a little background, that when I see firms that come in to me, they have already been inspected, they go through a district compliance function. A 483 might be given, which is a notice of observations to a firm based on the regulations. That goes through their compliance function in the district. They decide if any regulatory action needs to be taken.

Then, it comes into the center where I am. We do another review of everything in front of us, and we decide. There is regulatory discretion that we use for under 210 and 211. There is regulatory discretion in everything.

I want you not to think about these are regulations where we will be going in nitpicking you and getting you per se on certain little issues, because that is not the case.

There is certainly many levels of review that reports and actions go through before that ever happens. Then, we also have to deal with our general counsel, like David Horowitz, who I am sure wouldn't allow some things. So, there is a bunch of different levels. It is not just a 483 is the end of your world. There will be a lot of training just so that you have a basic understanding of kind of the different levels that we go through.

MR. FERRIS: So, this is a kinder, gentler FDA?

MS. ROBERTS: It depends on.

DR. CONTI: Actually, I have a question. Jorge, I am sorry, we may want to bring this up now since you have talked about the 483 and the regulatory discretion, whether or not we want to talk about what activities the PET community can play in terms of an appeals process.

MS. ROBERTS: We are not going to talk about that and discuss that today. What I want to get through is the actual GMPs. That will be discussed later once we can agree on the regulations that you should be subject to.

So, moving forward then, if we can start with--

MR. SWANSON: Can I ask you one question while you are on the topic, though? Again, it comes back. I mean the CGMPs are difficult for us to have a full understanding without the guidance document, and that is a point I am trying to make, but to take it one step further, do your inspectors inspect according to the guidance document or is there a separate inspector guidance document?

MS. AXELRAD: Well, the guidance document will be one of the pieces that they will use. They may also develop a compliance program that would tell the inspectors what to look for.

MR. SWANSON: Can we assume that this community would be involved in working with that also?


MR. SWANSON: Okay. That would help a lot, I think.

MS. AXELRAD: We intend to do a major effort to get a cadre of inspectors who are specially trained in dealing with the new good manufacturing regulations, in dealing with PET facilities, in conducting inspections. I mean the last thing we need to do is, you know, go to all this effort to put this stuff in place and then have an inspector go out and apply 211, you know, kinds of concepts. I mean we obviously are going to a lot of effort here, and we are going to make sure that there is adequate training of inspectors, and we will be working with you on all the implementing documents.

DR. HUNG: Jane, could I ask you another question on that, please? Okay.

MS. AXELRAD: Could you hold your questions, and we will leave five minutes right before we break for lunch to get questions from the audience? We really would like to move on.

DR. HUNG: That's fine.

MS. ROBERTS: So, moving forward with the Subpart C, Quality Control, 212.20. I wanted to know if there is any comments or questions, clarifications needed to for that section.

MR. SWANSON: If I could just give you a quick example there. I said earlier you have defined terms, and I think you need to go back and incorporate your defined terms. A good example is under 212.20(b), to be very simplified and say, "The quality control unit must have the authority to examine and approve components," okay, which include all the other stuff by your previous definition, and drug product to ensure compliance.

MS. ROBERTS: I guess we have a problem with definitions here because the definition of components is particular to FDA and drug products in particular, and I think then we might need to move the definition of components into here.

DR. CALLAHAN: And just to clarify again, there is nothing that says that the quality control unit can't be the same person who does the production.

MS. ROBERTS: No. If you notice, we specifically didn't name who would have to do it. We just say that you must have a quality control unit. It could be the same person wearing seven different hats as long as, upon inspection, we find that it is adequate for your facility.

Any other comments or questions under Quality Control?

Good. Moving right along, Subpart D, Equipment and Facilities, 212.30.

DR. CALLAHAN: There is reference to environmental conditions. What doors does that open in terms of environmental controls?

MS. ROBERTS: We have listened to your concerns over the past, and in working with David Hussong and working on that section of the USP, we took a lot of that into consideration. Something like that section will go into the guidance document also that adequately explains what we expect, and I think after the last public meeting, I kind of told you what we expected. We just want to make sure that you are assuring that there will be no microbial contamination into a product that purports to be sterile.

Now, however you can do that, that's fine, and from what I have understood from past discussions, looking at PET centers, a lot of what I have seen might just be adequate. You just need to ensure either with procedures or with different equipment, depending on what is particular for your facility, that you are just able to meet the requirement that you are going to control contamination however you do that.

DR. CONTI: Specifically, I want to ask about dual use of high-cost equipment for research purposes, as well as for production of these radiopharmaceuticals, to clarify this. It will not be the intent of academic institutions to duplicate many of these pieces of equipment, so you can take that into consideration that these can be used for dual use purposes.

For example, an HPLC unit could be used for research purposes or for production and quality control of the manufactured pharmaceuticals.

MS. ROBERTS: That is fine as long as it meets the requirements under laboratory controls.

DR. CONTI: I just want to clarify that because the academic institutions will not be able to duplicate these types of facilities for production purposes.

MS. AXELRAD: But presumably they will make sure the HPLC is working properly and does what it is supposed to do even for the research.

DR. CONTI: Yes, the quality control issue is a separate issue. I am just talking about dual use and this type of equipment.

MS. AXELRAD: As long as it meets the requirements that we have here, it is not going to be a problem.

MS. ROBERTS: Any other comments under Equipment and Facilities?

I just wanted to point out, though, that if you are using technically a contract laboratory almost to test for your sterility per se, even if it is a different part of the institution that you are using, they still fall under all of the regulations here for PET drug.

If you are using a separate laboratory anywhere, within your facility, outside your facility, another part of the academic institution, these regulations also would apply to what they are doing there, if they are doing the quality control or release testing for your drug product.

MR. FERRIS: I have a question on (c) where you talk about reactive, additive, and absorptive with respect to the equipment, the materials that come in contact with the drug product.

That can be a phenomenal job with respect to automated synthesizers with tubing and the pathway of the drug synthesis. If that has to be evaluated on the context of a piecemeal basis, that would be a tremendous burden.

If you envision simply evaluating the synthesized product and its purity on that basis, then, it would probably be doable.

MS. ROBERTS: What I thought of and I envisioned when this was written, I know the topic of validation is going to come up. You have been producing these products for years and years and years.

What I am going to address will be a type of retrospective validation, so that you have been doing this all along, the same with this, to show that it is not additive, absorptive, or anything of the like. It might just be easy for you to show that what you are getting out at the end, instead of looking at each little piece, should be okay, but you have to break down the processes what I envision, like the black box is one component and that what goes in and what goes out is what you expect in that piece, so therefore, there is nothing that is reactive, absorptive in there.

Do you understand what I am saying?

MR. FERRIS: [Comment off microphone.]

MS. ROBERTS: It could be based on how a traditional drug manufacturer would have done it or would be expected to do it, but since we have all of the years of experience with the PET drugs, it may be okay to do what we would explain as a retrospective in that what you are getting out of that particular piece under a protocol or whatever you would write up is okay.

DR. CONTI: Let me just ask another question on this issue. In terms of the isolation of the compound, let's say, and you had some of the compound retained on the tubing, theoretically, if you include quality in terms of the actual amount produced, if that is within the spectrum of your quality control, that could theoretically alter the quality, if you will, by leaving some retained on the tubing or something like this during isolation, and this is a routine occurrence.

So, if quality includes something like that, then, we can't do it.

MS. ROBERTS: What is exactly meant by this is that it is not going to affect the quality of the product, and that is why we talk about theoretical yields and that you know what your yield is going to be, and that it won't affect the end product per se.

DR. CONTI: It could affect the amount of the end product is my point.

MS. ROBERTS: But as long as you have that specification in your range taking into account what may be left behind, that is fine.

Any other comments on that section?

MR. FERRIS: The devil will be in the details.

MS. ROBERTS: But we will be all here to work them out together, I am sure.

Moving along to Subpart E, Control of Components, Containers, and Closures, 212.40.

DR. CONTI: I had some concerns, I guess on (1) and (2), Section (c).

I am just concerned about these identity tests on some of these species. There may be some extreme expenses of testing certain types of these, inorganic reagents, and things of this nature that would not be appropriate for a PET facility in an academic institution to embark upon, and I need to figure out some way to relieve us of that burden and maybe use the manufacturer's Certificate of Analysis in place of the tests performed.

MS. ROBERTS: Is there some other test that you could perform on that specific product? One test within the COA, that you could say okay, we have looked at that one, maybe an identity test might be too much, is there something else?

DR. CONTI: If the identity test means looking at the vial, checking the label to see if it matches the invoice, yes. If it is a matter of doing an elemental analysis on something that may not be feasible, that is not going to be possible, and in particular, if you are telling us the lab is performing some of these tests have to be in compliance, I mean we have no way of following through on that.

MS. ROBERTS: Is there some way that you could list out for me some of those areas or the compounds that you would have problems with, so that we could take it up with our chemists here and try to work through that with you? I would like to know what specifically might be the problem.

DR. CONTI: I would actually like to throw it back in your court and tell me why a Certificate of Analysis is not adequate, and if there is a rationale for doing the test, then, we would like to hear it as opposed to us telling you which ones are not going to be feasible or not.

If you identify something within FDA that says you must do something beyond a Certificate of Analysis for X, Y, and Z reasons, then, I think we can accept that and work with you on it, but otherwise, I think it's in your court.

MS. ROBERTS: Okay. We will work on that for you.

Other specific comments on that section?

DR. KASLIWAL: We can just comment on identity tests. You really need to do identity tests on the components that you are using for formulation of drug product. If it is a stabilizer or preservative, or if you are making up your saline solution, you are not using a commercial product, I mean you are really taking a risk in directly adding something just on the basis on the visual observation there. So, that is there now.

It is usually a good practice to know what you are using in advance, all the reactants or whatever else that you need to, and you can have identity tests that works for you, and the other thing is the universities, most universities that I know, usually, you will be able to find equipment to be able to perform identity tests.

DR. CONTI: The issue is give us a specific example in the proposed drugs that we are dealing with here where you feel it is essential that we have something beyond the manufacturer's Certificate of Analysis, and given the fact that we test each batch, those two stipulations justify this additional testing.

DR. KASLIWAL: For example, my understanding is, for example, FDG. If you are using a saline, if you are using, let's say, a commercially available approved drug product, you know, 0.9 percent sodium chloride or something, I would agree you don't have to do an identify test there, but if you are making up a saline solution, you buy sodium chloride, you buy water for injection, and you mix it up and prepare a solution, yes, you do have to do an identity test on sodium chloride.

MR. FERRIS: There would be no manufacturer's certificate, you are the manufacturer, so you would have to do it, of course.

DR. KASLIWAL: Right, but where do you take into account the bottle, the white powder, crystalline powder is sodium chloride? Do you go and qualify your manufacturer, your suppliers? No, you are just buying off randomly.

DR. CONTI: But they do provide an analysis.

DR. KASLIWAL: Yes, but that is taking into assumption that the bottle actually contains that, the analysis that you have got the material for.

DR. CONTI: I assure you if it doesn't, we are not going to get the product.

DR. KASLIWAL: If it's a reactant, but if it's a component that you are adding directly to the finished product for formulation.

DR. CALLAHAN: Anyway, you were specifically talking about adding materials to the--

DR. KASLIWAL: To the components, drug product components.

DR. CALLAHAN: So, what you just said, that whole discussion was for products using the formulation of the final drug product, not in the production of the drug substance, because, in a way, not everyone is adding anything after the fact to formulate the final drug product. We are asking about identity testing on reactants and components, as well.

DR. KASLIWAL: I think that we will probably need to discuss. Currently, most everybody does that, and that may be in part because there is a lot of money invested in production of a batch for a regular pharmaceutical, so they want to be sure of that, whether, you know, for the sake of one identity test, whether they want to sacrifice the whole batch or not. Now, here, you may have a smaller batch and there may be some different issues here. That is something we need to discuss probably.

MR. FERRIS: In some instances here, too--what you say is absolutely true in the cold pharmaceutical production business, but if we talk about some of the unique characteristics of radiopharmaceuticals in general, and PET in particular, the synthesis itself can be destructive to the final drug product if you use the wrong component. That is inherent in these procedures, and measurable, as well.

DR. KASLIWAL: Right, and it is all integrated. If you put robust, for example, radiochemical yield limits, you can pick up that, but if you are going to propose at the same time a limit of 10 to 100 percent, it all goes together.

The other thing is that is the risk you take. If you don't want to perform it, so you assume that risk.

MS. AXELRAD: They are going to change the sound system over lunch, so if you could just bear with us until the lunch break, then, they are going to hopefully fix it, but you have to use the mike even if it has feedback, because otherwise the transcriptionist can't pick it up.

MR. SWANSON: Another question under Subheading (2). "Instead of such testing a report of analysis may be accepted from the supplier provided the PET center establishes the reliability of the supplier's test results," could you expand on that a little bit as to how you see us establishing the reliability of the manufacturer's test results?

MS. ROBERTS: By periodically testing certain--retesting for the results of that COA by doing or picking up and checking on the COA, and finding out if it is indeed reliable. Another way that you could it also, which would not be acceptable to you, I am sure, is to go out and audit what they are doing there to make sure that--and I know that that is not acceptable, so the least that we could do is to try to make sure that their COA for you is reliable by doing one of the tests on the COA.

DR. CONTI: I think I have just told you that we can't do all of these tests, so you need to tell us which ones you are worried about, so we can figure out whether we can actually do the tests or not because, frankly, I am not in the position to develop a test for metal hydrides or something like this at an academic institution necessarily, it is just not possible to do some of these things.

MS. AXELRAD: You could contract it out, for example. We are not just dealing--I mean if we wanted to limit this regulation to the drugs we know, FDG, ammonia, water, maybe we could do that, but we are not doing this. We are writing GMPs for all PET drugs, both those we know now and those that may come in the future, and so it puts us in a very difficult position because we don't know whether you might start manufacturing a PET drug which has an ingredient or a compound in it that might not be picked up by just--you know, the process just doesn't work, and that could cause a real problem.

DR. CONTI: That is not what we are talking about here, Jane. We are talking about issues related to the materials used in the preparations of a drug and making sure that they are what they, in fact, are.

You are putting an extra burden on the academic centers to come up with some way of doing a quality control check on manufacturers that are producing these materials on the outside, and I am telling you that that is not going to work, so, you need to work with that.

DR. KASLIWAL: I think one way maybe ICP can help out there is in qualifying the vendor for everybody, so individual person doesn't have to do that, establish the reliability on a central basis. I mean what is reliable? Just looking at somebody's face, is that reliable? Sometimes it is.

MS. ROBERTS: We understand what you have said so far about the testing and the COAs. We will have to take that back for internal discussion, however, we think it is important for you to test, to do one specific test off the COA for the major components that are going into the drug products.

MR. SWANSON: One area of relief there, a specific comment that you definitely need to look at is if we are using any component that is already FDA approved, is either a drug or a device, okay, it seems like you ought, I mean as part of your FDA--normal saline, 0.9 percent sodium chloride for injection, sterile vials, okay, that are commercially available. I mean you are requiring all these people to go through all the same process controls to ensure that what they are releasing meet acceptance criteria. Now you want us to go back and retest these things?

So, I think that is one area where you could probably provide some relief.

DR. KASLIWAL: Dennis, if you look at the model applications, I think there is a relief for presealed, if you buy pre-sterilized, sealed vials, as well as saline, if it is indeed an approved product, and by that means, 0.9 percent sodium chloride, let's say, from a manufacturer that has an NDA for that or an ANDA for that, if you want to prepare the 0.9 percent sodium chloride in-house, that is not an approved product.

MR. SWANSON: I don't have a problem with what you just said, but I understand your model ANDA submissions are not the regulations, so perhaps your regulations could help address that issue.

MS. ROBERTS: That point is well taken. We will look at that.

MR. KUHS: On the final section, Section (e), the last sentence that we need to keep records of the disposition of rejected material and the expiration date, we ran into a situation in Peoria where we had some glass vials that had a Certificate of Analysis that were shorted in another facility, and they sent 25 vials to another facility to use, and the issue comes with reconciliation of the inventory, of control of all of the components, containers, closures, material, and the inventory that was at the end did not reconcile with what was received, and so there was a question on why didn't we have a reconciliation of inventory.

I think that we have resolved this. There was a 483 issued on that. I would have to ask Ken for sure, but I think we resolved it in that we weren't required to keep an inventory or reconciliation of inventory at the end of disposition, in other words, if we had a bad vial of Kryptofix, people just toss it, they don't mark it and say there isn't a reconciliation of inventory, and I think that that could be a problem.

MS. ROBERTS: What this specifically is talking about here is if you test a component and it doesn't meet, it is rejected, it fails, then it should be marked as rejected, you should write in under wherever you tested it that it wasn't any good, that this failed and it was disposed of.

MR. KUHS: Let me give you an example of that. Oftentimes you will get a shipment of a case of vials that are used for collection, a final container, and it was dropped during shipment and two or three vials are broken. You throw you the vials you didn't use, but we don't make a note anywhere that two vials were broken.

It is an issue that while it seems insignificant now, when we talk about reconciling inventory at the end, it does become a problem, and I don't think anyone keeps track of vials that for some reason when you pull the pop top off, you didn't do it right or the container crimped or something else, you just toss it, you don't make a record in your inventory anywhere that something was wrong with it or that you destroyed it.

MS. ROBERTS: That wasn't the intended purpose of this specific part, and that is something that I can definitely address in the guidance, which I agree with you that if you throw out a vial if it's broken, you are not going to be able to use it anyway, so that is something that I will take under advisement and make sure that it's delineated in the guidance document.

MR. KUHS: Okay. I think it's just the final reconciliation doesn't really lend itself well to small amounts of inventory that are generally kept on hand at a PET center.

DR. CONTI: If I could go back for one more minute to (2) again, on this issue of testing. One issue comes to mind where you have a commercial entity that may buy these components in bulk, and then is able to, because they have purchased a certain amount or they have a certain level of operation, have the infrastructure in place to do such identity testing, and that may amount of the same bulk shipment that is used for 20 cyclotron operations.

In that configuration, perhaps that is one way of looking at the difference between a commercial entity and an academic institution, which would have to actually duplicate at each site that same testing or contract with somebody to do that type of testing. So, I think that needs to be taken into consideration, as well.

MS. ROBERTS: Any other comments on this particular section?

Then, we will move on to Subpart F, Production and Process Controls, and as quickly as we can get through this, then, we can go to lunch.

MR. SWANSON: Item (c), you talk about information that needs to appear on the master production and control record. In Sub-item (1), you have the name and strength of the PET drug. The strength is going to be batch-specific, so we can't define a strength in a master production record per se.

MS. KEPPLER: For those of us that are a little less familiar with this, could you explain the difference between the written production and process control, the master production and process control, and the batch production and process control? It wasn't clear to me.

MS. ROBERTS: The master production and process control generally is your master record of how, based on your knowledge of the product, you are going to produce that product. It includes all of the steps that you are going to follow, everything that is going to be done, spaces on the formulation page for how much product, how much component you will be putting in, spots for weighing out the component, and a detailed production listing of the steps that you are going to follow with signoffs, initials, parts for, you know, that we have completed that step.

Usually, what it also consists of is whether it is the most recent production record. You have a master. You keep the one master, and how we usually see it done is off the master you make your copies for your batch production records, and that batch production record, you actually fill in when you are producing your batch from a master.

A master lays out the guideline that you will follow for making that particular batch or that particular size of the batch.

That is a master that is kept, and each time you may change your formulation or you change a different processing step, you need to make a new master production record, keeping in history the old ones also, and they are usually dated as to which is the most recent one to follow.

MS. KEPPLER: So, the master is a compilation of all your batch production records with the most recent one on top as being the one done?

MS. AXELRAD: No, I think the master is like the recipe, and then the batch record is a notation that you followed the recipe and what you did for the specific batch.

MS. KEPPLER: Are they on the same sheet of paper or you would have your recipe--but she was talking about filling in the blanks and places to initial on the master.


MS. KEPPLER: That you would then xerox and use for your batch.

MS. ROBERTS: That is exactly what it is. The master production record is a template of everything, of the whole recipe that you are going to follow with all the blanks. You keep that as your master, and every time you go to produce a batch, you make the copy, you give it out, and that is the one that they will fill in the blanks for when they produce the batch.

MR. SWANSON: You might also state that when you make that copy, you are required to certify that that copy of your batch record is an accurate reproduction of the current master formula card, so somebody has to sign off on that, okay. Another step.

DR. KASLIWAL: Just a comment on that inclusion of the strength on the master production record. You can probably include the range of the strength that you would put in your model application that you have validated, and your batch's specific strength would be on the batch record.

MR. SWANSON: All I am saying is it can't be a specific value.

MR. FERRIS: With respect to the issue of strength, the name and the strength of the drug, as you define strength here, you are talking about concentration. If you carry that over to this particular section, that would basically indicate to me anyway, in my read of this, that you wouldn't have a batch size, you would have a concentration that you would specify.

In other words, I could make 25 millicuries to 24 curies. As long as I diluted that in my final finished dosage form to the appropriate concentration, what you call strength, then, that would be okay, but that doesn't make sense.

DR. KASLIWAL: You would define a batch size in the model application.

MR. FERRIS: That should be millicuries. That is millicuries.

DR. KASLIWAL: Right, but the strength for purpose is the concentration, but the batch size will be the total, yes, millicuries.

DR. CONTI: I have another comment on the (5), the theoretical yield again, as we talked about earlier, that needs to be modified, so just to note that.

Down on Section (7)(h), the two tests, are these duplicate or different tests, and where do we get that information from?

MS. ROBERTS: This is just explaining the reserve sample portion. It is a reserve sample that you would keep in case you need to retest for any particular reason, and we say that you should keep enough for two tests in case the one is a failure or a problem, you will have enough there to do a retest to confirm whatever result you got. That is what is meant by that.

MR. KUHS: I have two comments on that. Number one, it is very often that the entire batch may be administered to a single patient, and that came from a different definition a long time ago.

The second one is if we are making entire vials to be redistributed by a radiopharmacy, oftentimes the entire manufactured batch goes to that radiopharmacy. Are you saying that we need to take a sample out of that batch and put it into a separate vial for a reserve sample?

This is an issue that did come up again in Peoria, that they were making a separate entire vial, manufactured vial, which was delivered to a radiopharmacy for redistribution without keeping a reserve sample out of that, and if you keep a reserve sample out of it, and transfer it to an additional vial, doesn't that destroy the integrity of the original sample?

DR. KASLIWAL: I think that probably we will need to discuss whether a sample ought to be withdrawn or the vial ought to come back to the manufacturer. We don't know what is going to be the proper way.

MR. KUHS: There are a number or regulations that actually prevent return of vials to the manufacturer, so that is probably not an appropriate solution, but it was an issue.

DR. CONTI: The other thing about that question is whether or not you do these pre-tests, pre-batches for certain isotopes also for the O 15 in particular. You are not going to be taking a sample necessarily out of the one you are actually delivering, but you have done your analysis on the prior one.

If it is a gas or something like that, how do you save that sample for a retest later?

MS. ROBERTS: I actually had the same question that you all are bringing up, and I wanted your input on this issue, on how feasible it is, how you would do it, if it is absolutely necessary.

I would think that it might be in certain instances, for certain products, where there is different ligands or there is other problems where a product may--there could be something in the product that you don't find about until later and you want to retest the product to find out what it exactly was, so you can fix it later.

Is there any instances where that would happen, and it would be beneficial? I thought it was in the Peoria instance.

MR. KUHS: I think you can probably address that by doing a periodic reserve sample, and not necessarily one out of every batch, and that may be something that is defined in the way the application is submitted, that you agree on a monthly, bimonthly, six-month basis to keep a reserve sample for periodic testing.

I am not sure how to resolve that, but that is just a suggestion.

MR. FERRIS: It is intriguing to me, with 110-minute, half-life material, doing an investigation 30 days later, and typically, the tests exclude sterility and pyrogenicity, that is not what you usually mean for the testing, so if we exclude those two, what tests might be useful, that would give us insight as to what happened?

MS. ROBERTS: Is there any instance where it would be important that where an organic solvent would have gotten into the product, that you would want to retest to find out if that indeed was the instance or the problem?

That is the specific thing that I was thinking of in this case.

MR. FERRIS: It may have dissipated 30 days later.

DR. KASLIWAL: I think the chemical testing obviously, because radioisotope is gone. You could repeat chemical testing, and if the safety problems arise, more than likely they probably arise due to chemical.

DR. BARRIO: We seem to be focusing clearly on radiopharmaceuticals that are being used in the clinic, but, of course, this is intended to be used in research, too. If a PET center produces 10 or 15 compounds a day either being used for the clinic or be used for research or be used for animals or be used for studies, all of these have to apply.

Therefore, you know, I think it is difficult to envision how you could do many of those things even for research preparation, that have less control, of course, because they are research preparations. Sometimes we are beginning to produce a product, and they are more difficult to get this kind of control we should have daily for human injections.

MS. ROBERTS: We will have to revisit this, and I will take a lot of your comments. There might be a way where we can deal with it on an application by application basis, but my fear with that is for products that don't have an application or may not in the interim, so it is something that we will need to talk about.

I would appreciate also--I guess you are all staring at me like why do you think we need to keep this--and the particular instance was from the research that I have done with Peoria and I thought that there might be other organic solvents or other problems even if you get a pyrogenic reaction in a patient, you might want to go back and retest your sample also.

I can think of a couple of reasons why. If you could tell me why it is not feasible to do that?

DR. CALLAHAN: For one thing, organic solvent contamination is a release criteria, so you have done that before you released it. They are not going to grow into it, if anything, they are going to go away over time, so that example in the application, residual organic solvents is a release criteria, so you know that before you even release it.

MR. SWANSON: Okay. So, I have a reserve, I mean what is the purpose of the reserve sample testing, am I going to recall this lot? I can't. It has already been used. It is a batch size of one, it has already been injected in everybody, so it seems to me like if I have suspicioned a problem with organic solvents in my product, I am going to run my next batch, and I am going to take--I am taking a look at it as an end release criteria anyway, okay, I am going to go back and look at my process.

I am not sure what the purpose--yes, reserve samples have a purpose, I think, in traditional drug manufacturing, but I am not sure if that purpose applies here, I am not sure what I can do about it.

MS. ROBERTS: Does any PET center at this time keep any reserve samples for any purpose?

MR. SWANSON: We keep vials to go back for sterility testing, but I am not sure that is an appropriate use of reserve samples.

MS. ROBERTS: Why does anybody else keep them? Is it only for sterility? Is it in case there is a problem for corrective and preventive action? I would like to hear why anybody else is keeping reserve samples.

MR. WATKINS: We keep them because we thought it was a requirement to do so, but for sterility testing, number one, your sample would have to be kept under sterile conditions if it was going to be of any use to test afterwards.

You require us to test as quickly as possible after we have made the dose. There are things in th literature which would suggest that the various bugs will disappear over time, so it may not have any significance at all doing your tests later on.

I think most of us keep them purely because we thought it was a requirement to do so.

MS. ROBERTS: Okay. I thought I might have gotten like a revelation, but I guess I am not, so we will take that under consideration.

DR. CALLAHAN: Do you require that other manufacturers retain samples beyond expiration, they have to keep them way beyond expiration or not, after the product has expired, they have to retain them?

MS. ROBERTS: Yes, they are kept beyond expiry.

DR. CALLAHAN: (7)(e), production and dispensing, I would recommend maybe changing that to production and packaging area or something like that rather than the word dispensing again.

MS. ROBERTS: In here, the particular word dispensing was used for those instances where it is directly dispensed from the manufacturing area into the patient, whether it is a gas or a liquid. That is what was meant by that.

DR. CALLAHAN: Your dispensing area is going to be inside a very, very small tube that is going to go into the patient's nose or something like that. I mean this is really impractical. Dispensing in my mind means when you are preparing unit dose syringes and under pharmacy or medicine. I think we need to reserve that term for that specific activity, and not confuse the issue. That is why I suggested packaging.

In your example, that is really patient administration, not dispensing, right, because you are administering it, it is at the interface between the drug delivery system and the patient if they are breathing a gas or a continuous--

MS. ROBERTS: That is what we had meant by that, so maybe dispensing area is the wrong word to use.

DR. BARRIO: I would like to go to (7)(f) and (g), I guess, both of them. For chemists in the field, we do validation all the time because we want to understand how our process works, of course, but what done means to us is--if I could use a very simple example--use the FDG synthesis that I am sure everybody uses.

We have a system that we understand, has components that accomplish several steps in the process, synthesis, hydrolysis, purification, utilization, but what we would like to consider validation, I guess, that is easy to understand, is we have a system. We know what we put in, and we validate that system that produces always a certain amount of a pharmaceutical that is always sterile and pyrogen-free. That could be considered an appropriate validation procedure. This is something that every one of us will deal with every compound.

I think the issue of interpretation of what validation means for the agency is a thing that I try to address, because validation could have a different meaning if we have to address issues of minimum or large amount of activities, I mean low and large, to produce a certain amount here and there.

The other one could be validate every single step in the process or validate every instrument we use or whatever. It may be computers, as you indicate here. What I think is not only very impractical, I think it is from my personal perspective, I guess it is completely unnecessary.

I think it would very important if you could define what validation means for you and then for every one of us to really understand how to approach that topic.

MS. ROBERTS: During the last public meeting, we had talked about validation and the definition of validation, and what I said that I might expect from the standpoint of a process validation and from sterile product validation, from your environmental monitoring per se.

I had asked for examples of things that you have done as far as validation, so I could look through them and tell you, give you a read of whether it is more than I expected, what I expected, less than I expected, and I had never received anything like that.

So, now, and at this time, it is difficult for me to sit here and tell you this what I would think how you should validate your process, because each process in each center is extremely different, and your validation is based on what you are doing and what you want to prove, your process.

It could be important to validate each particular little step if it's a critical processing step. If it is not a critical processing step for your operation, you may not need to include it in a validation.

When we talk about validation, it is the definition that is written in here, yes, it's pretty vague, however, what we want to make sure that you are doing is consistently producing a product to what your specifications are.

DR. BARRIO: In the preparation of pharmaceuticals that include a synthesis process, other elements in the synthesis, not only their reaction, but sometimes modification of intermediates, all the steps. Then, based on your definition, then, we have to monitor every single step, and I feel that that is probably unnecessary because if we achieve the result that is expected, and, of course, one has to expect that the chemists or pharmacists doing that will really understand how the system works.

I mean this is a black box, but we know it is not a black box, it is a series of chemical components there, but there is a difference between going through every single component in regards to regulation versus if you go to any PET center you want, you are right, every PET center does it in a different way, but if you get from whatever system everybody is using, you go there and say, okay, you give me 10 studies in which you obtain always the same part, then, why do we need to go to every single step in the process?

I think that is crucial because this will add not only a tremendous amount of work, but I feel it is probably certainly unnecessary in many circumstances, and I agree with you, maybe in some circumstance.

I could see, for example, that when you are investigating a particular new synthesis, this is something we do all the time. I would like to do how every component works. I would like to understand whether that column does the separation I want or not. I mean this is something we do all the time, but that becomes a routine procedure, something we have done a zillion times. To require this kind of documentation is going to be so burdensome that this is difficult to believe that anybody would like to do that.

MS. ROBERTS: I understand what you are saying. You have brought up a very important point about new entities that you are manufacturing, new synthesis that you are going to validate. You are validating while or before you are starting your production.

For products that you have already been producing for so long, it wasn't my intention to require prospective validation, what we normally call prospective validation where you would go and start validating every single little piece.

I would expect that for new drug products that you will begin to bring up on line when these regulations are in effect.

What I have envisioned for products that you are already making and have a lot of history on, what you were talking about would be like a retrospective validation where you wouldn't go back to each particular little step because you have all these years or experience making the product, and you basically know what your results will be.

A retrospective and a prospective are different in those respects, and I would think that it would be fine for you for products that you have already been making for this long, to gather a lot of the data that already exists.

What I envision is that you would only have to write a protocol about what you are going to do to show that this process is indeed validated, is reproducible, which may mean that you don't need to do any additional tests. You may just need to write your protocol about what you expect it to meet, how you are going to do this, and then gather all the data to show me that indeed this process is validated.

We have done that with other older products in the drug arena, as well as a lot of firms also--it hasn't been a problem in the past, and I really don't envision you having to do prospective validation for every product that you are currently manufacturing.

MR. FERRIS: If you are going to accept retrospective validation, then, typically, that will not include extreme limit testing.

MS. ROBERTS: I am sorry?

MR. FERRIS: If you are going to accept retrospective validation, it is not going to include extreme limit testing with respect to components, and which is typical in regular prospective validation.

MS. ROBERTS: Right, and I understand that.

MR. FERRIS: And that is acceptable.

MS. ROBERTS: We are going to have to take that under advisement, and it may be acceptable, and I envision doing a retrospective for all the products that you already have up and running, and have all of these on, unless you do encounter a problem with your process or along the lines, and you think that is part of the process, and you would need to do the extreme testing.

But if you are not encountering any problems, I wouldn't expect that you would have to go back and do that as part of retrospective.

DR. BARRIO: Then, in this context, if you read (7)(f), how do you interpret that section?

MS. ROBERTS: In that section, we are talking about in-process controls, if your PET center thinks that you need the in-process control. For some processes you may need them, and for other processes you may not need them.

This is one is basically, whether for the critical components that you think may be a problem, that you definitely have to watch to make sure that you are going to get the product that you expect to get out at the end. That is what I would think. That is the intent of this.

DR. KASLIWAL: I think it is related to testing of in-process materials, if they are isolated. In the radiochemical operation, you probably don't have any isolatable in-process materials, but I understand the regulations are--to be used and the facilities and controls to be used.

That whole aspect starts from a starting material, was defined as a starting material, so your individual process may have or may not have in-process materials. In-process materials, we usually can designate them as components. They are in-process materials, anything beyond a starting material if it is isolatable and kept, it is in-process material and needs to be tested, according to what you designate them as is called the parameters.

DR. CALLAHAN: Could you give us an example in, say, FDG terms, is there anything that relates to that based on your definition of in-process materials that are isolatable?

DR. KASLIWAL: If it was defined as a starting material, anything beyond the starting material is an in-process material. For example, the mannose triflate is an in-process material, it is not a starting material according to the definition, if you go back and look at the definition in Drug Substances guidelines, and we will take it as a key intermediate. It's an in-process material, and you can accept it on the basis of COA.

DR. BARRIO: You are saying that mannose triflate is not a component of the final preparation. It is a material that is used to produce a radiopharmaceutical.


DR. BARRIO: And therefore, should be controlled with a Certificate of Analysis or whatever.

DR. KASLIWAL: You could do that, right, and in the application are the criteria for it. If you read the definition, component means any ingredient intended for use in the production of a PET drug, including any ingredients that may not appear in the final PET drug product as well as any packaging materials and container-closure system.

DR. BARRIO: Well, this is similar to the discussion we had a few minutes ago. I mean we are talking about the same thing, right?

DR. KASLIWAL: How you define that.

DR. BARRIO: I mean how to really define the quality of the product.

DR. KASLIWAL: For mannose triflate, if you look at model application, I think you have to define how you accept it, and, you know.

DR. CONTI: To be honest with you, I would rather be considering testing the mannose triflate than sodium chloride.

DR. KASLIWAL: And I agree with that.

DR. CONTI: I think there is a level of comfort.

DR. KASLIWAL: I agree with that except that if the sodium chloride is used in the formulation.

DR. CONTI: I am talking about the components, the manufacturing process. I mean there are certain things that just seemed a little bit onerous.

MR. FERRIS: In the discussion about validation, are you including computer system validation, as well, with respect to retrospective?

MS. ROBERTS: Yes. If you have already been using that same program to produce that same drug product for all those years, that system would need retrospective validation also.

DR. CONTI: What if you upgrade your software?

MS. ROBERTS: If you have already prospectively validated the whole system, and you upgrade to a different version, that would still be a validation or more of a verification that this upgrade is still working, you are still achieving the same thing.

DR. CONTI: Wait a minute now, because now if I have a 10-year track record of making FDG with a certain piece of software, you are telling me I now have to do a prospective validation when I upgrade my software? That changes the whole configuration.

MS. ROBERTS: If you have already retrospectively validated that program, and it's working fine for what you needed, when you upgrade to a new version, you have to do a smaller validation, it is still a validation, to make sure that everything is still working the way it should be working for that change in software.

DR. KASLIWAL: I think if you read the USP, even USP says you do have to verify that upgrade if there is a change in computer software program. If you want, I can read the USP language.

DR. CONTI: But the validation could be just again the reproducibility of the--

DR. KASLIWAL: Verification of the batches.

DR. CONTI: Right.

DR. KASLIWAL: I think that is what Tracy was saying.

DR. BARRIO: Then, you upgrade your system, and if it produces FDG according to the specification, you are done. You don't need to go and really reanalyze.

MS. ROBERTS: In essence, yes, if everything is continuing to work and you expect--but you need the paperwork behind it that says I put in a new version of software, I am going to make sure that my next three batches still work the same way, all my testing is still right, there is no glitches.

You produce your three batches. Your validation is done. It is signed off, everything looks okay. There is no problems with the new upgrade.

MR. SWANSON: Along the same lines, the USP statement that you quote was actually criticized in several comments that came back in that there probably needs to be some cutoffs for types of changes that require revalidation versus types of changes in the computer software that wouldn't require revalidation.

You may change the software for insignificant reasons, so there needs to be some clarifying information in your guidance document again I think along that line, because I think that is a valid criticism.

MR. KUHS: I have a question on (c)(2). We don't have a good definition in our definitions in front of what constitutes a dosage unit. That is a little vague to me. I am not quite sure what that means.

We are dealing here with something that changes over time, we are dealing with amount of radioactivity for a certain weight or volume, but that changes over time, and I am not quite sure how this addresses this.

We can give you the total weight or the total volume of what the dosage unit is depending on what that is, but is the dosage unit an entire vial, is the dosage unit a single injection, is the dosage unit a millicurie or a mL? That whole section doesn't mean much to me.

MR. SWANSON: I really think it needs to be changed to specify per batch or lot, and not dosage unit.

DR. KASLIWAL: I think this is intended to be--and we will look at that--it is intended to be the batch formula that you use.

MR. SWANSON: Along the same lines, what is the difference between (2) and (4)? You don't differentiate between an active ingredient and a component in your definitions, so I am assuming a component is an active ingredient, and so I don't see any difference between the two.

MS. ROBERTS: I think there is a differentiation in there for that, and if there is not, we will make one.

MS. AXELRAD: We do say any ingredient, but traditionally, in our regulations, there is different kinds of ingredients. There is active ingredients and other components is usually I think the way it is done.

So, we had indicated that we needed to add a definition of inactive ingredient, which would differentiate it from a compound.

I don't know about the rest of you, but I am badly in need of breaking.

Are there any very quick, maybe one or two questions from the people in the audience?

MR. CHALY: I am Thomas Chaly from North Shore University Hospital.

I don't think it is fair to ask us to do the chemical testing of all the reagents that we are using for PET production. For example, we are using anhydrous ether, acetonitrile, Kryptofix, and the precursor for that.

If we have to do all the testing for all these, we need a lot more staff, I don't think that a small hospital like us can afford that, and I don't think it is necessary. We have been doing this for the last 15, 20 years, using the same kind of ether, same kind of acetonitrile, and it doesn't make any sense to me.

They are manufactured by good manufacturers, Aldrich, Sigma, and all these companies.

Another thing is the validation for each indices, we do that in the beginning of a new synthesis. When we develop a synthesis, particular synthesis, we do like a four of five synthesis in the same fashion. We take the sample out. We do a sterility and a quality control testing.

For example, in the case of when we develop F-dopa, we had to analyze the mercury situation there. We did four or five synthesis like that. We send out the sample outside the company, and did the checking for the mercury, the amount of mercury that can be found in the sample. This, we validate all the time. We have done that.

If we change from one to the other, suppose we buy a new synthesizer, we validate that machine before we start using that for patients. We do three or four synthesis. We do the sterility testing on that one. So, that all is validated.

MS. ROBERTS: Then, I guess you would meet the requirements under retrospective validation. That is exactly what we mean is that if you have done all this validation testing, you just need to put it together and be able to put your hands on it when somebody comes in and asks for it.

MR. WATKINS: I am a little bit confused at the moment as to what is a starting material and what is an in-process component. I guess the only starting material for fluoride, for example, is O-18 water, but I think chemists would normally think of triflate as being a starting material, and not an in-process component.

MS. AXELRAD: I think we need to discuss that among ourselves. I have some questions, too.

MR. WATKINS: The other thing was on identification. The indication was we could some tests ourselves to identify a material, but each one of those instruments, if I take it to the chemistry department, I run an NMR or something, which would be a good way to determine the purity of triflate, that instrument would have to be validated, as well, so it is not quite as simple as it is made to appear here.

MS. ROBERTS: I just have a question, that if that instrument is being used for any other testing within a facility, I would think any academic facility or anybody that is using it for research would have that machine qualified and calibrated to meet most of the standards to make sure it is working properly, and if that is the case, indeed, then you should have no problem then in meeting the requirements under the laboratory control for the calibration and making sure that the equipment is okay for its intended use.

DR. CONTI: But by the very same argument, though, I could say exactly the same thing for the people that we bought the supplies from. It is a circuitous argument you are making here, because I could say that Aldrich also does quality control on their instrumentation when they produce these materials and test them.

So, if we get a Certificate of Analysis, it is sufficient given the fact that we also test the final product, in my opinion, and I think the opinion of both the public here, as well as this table.

DR. KASLIWAL: I guess the issue is most drug product manufacturers, they do identity tests, as a precaution, I would say, so that their final batch, because there is a lot money invested, doesn't go bad.

So, you build that quality in. So, this is the risk you are taking whether, you know, and you need to evaluate that in light of your batch sizes.

The other issue is the reason why we require that to do is I think in my mind at least, is that sometimes if people don't do it, and if there is a lot of money at stake, people--if the batch is borderline or failing, and that is the reason we require them to do that.

DR. CONTI: If the batch has failed, the system, because you test every product, it is rejected by your own criteria that you have established. Again, I mean this is the difference between testing every batch and testing only samples of batches produced in the pharmaceutical business.

MS. AXELRAD: I think we should break for lunch.

You will have a chance. Is it on this particular section?

DR. HUNG: Yes. I just want to ask a question about when you submit an ANDA or NDA, do you have to submit a very specific GMP plan for your facility, and if so, can the inspector use that plan that you submit to the FDA to inspect the facility?

MS. AXELRAD: No, and especially in this case, since we won't have written the GMPs yet, you won't be expected to submit anything with regard to compliance with GMPs because we will have to figure out what they are going to be.

Let's meet back here at 1:30.

[Whereupon, at 12:39 p.m., the proceedings were recessed, to be resumed at 1:30 p.m., this same day.]



MS. ROBERTS: We are going to start with Subpart G, Laboratory Controls, 212.60.

If there is any comments on this section, I would like to go over them at this time.

DR. BARRIO: Do you think we could revisit briefly a couple points that we have on the previous sections?


DR. BARRIO: The comments will come from Dennis.

MR. SWANSON: I think a couple things that we summarized from this morning's conversations, we just want to have a record of summary comments. We definitely have concern regarding reserve samples and really can't see the purpose for them, so we suggest that any statements related to reserve samples be removed.

We have a major concern regarding the testing required for the acceptance of components and in-process material. We must definitely make efforts to minimize the testing required when you get a Certificate of Analysis associated with the product. There should be no requirements for additional testing with the emphasis again on end product validation, the quality of your end product.

We would definitely--I think Tracy mentioned this morning about the 483 process especially as this evolves over time, there is definitely a need to keep I think this committee or some kind of an advisory committee actively working with you in compliance at the FDA to take a look at 483s, so that there is some public input on your evaluation of 483s instead of just your input would be an important point.

I think we have a concern, and this may sort out in later discussions, but how these CGMPs relate to PET drugs as the subject of INDs or RDRCs. Again, we have a lot of questions about as we develop new agents for research, what kinds of product validation, procedure validations are going to be required, and what I think we heard this morning is that there may be quite extensive validation associated with new drugs under development versus drugs we have been involved in, and again, you know, you have developed research agents which may have a very limited application as far as human subjects are concerned.

You may do one research study with a certain agent that involves a maximum of 30 subjects, and to require extensive validation of the process, et cetera, for a research study that involves a limited number of people just doesn't seem to make a lot of sense to us.

Again, I will reemphasize, and I think the committee will reemphasize, there needs to be, and the FDA really needs to think in terms of end product validation, the emphasis has to be on testing the final product to ensure that it is, in fact, the product that we say it is and has appropriate strength, quality, and purity.

That concept needs to definitely be applied to research agents also.

MS. ROBERTS: I am sure you are not familiar, then, with our INDs and how GMPs apply to them in normal drug realms, and things like that, so I won't go into that now or how that is done, but for your own peace of mind, you might want to research that so far and see how we deal with that, but we do have policies and procedures of how we normally do that.

I think a lot of what you are fearing, we don't normally look at for an IND anyway during that time, but we can explain that and talk about that at other times, but we do understand what your fear is, but it is not totally founded. We will discuss that and take that into consideration.

Now, we will begin with the Laboratory Control section.

DR. CONTI: 212.60(d). The identity, purity and quality of reagents, et cetera, must be adequately controlled. Maybe it's just me, but what do you mean by "controlled" in this context, since we have little to say about many of these issues?

DR. KASLIWAL: In the solutions that you prepare, you will label them with the correct label, you know, identifying what the composition is, the reagents that you use, you will specify their quality, and that is what you will stick with

DR. CONTI: Am I safe to say cataloged is a better word as opposed to controlled? Maybe I am just misunderstanding the language.

DR. KASLIWAL: For example, if you specify for a reagent ACS grade, that means that is what you will use, and, you know, actively, that is what you will use, you control that at that level.

If you specify certain grades, that is what you will use to control it.

DR. CONTI: So, it is really the specifications.

DR. KASLIWAL: You can take it out of COA, yes.

MS. ROBERTS: In addition, what is also covered under this Subpart (d), when we talk about reagents, solutions and supplies used in testing must be adequately controlled.

If for any reason, for example, the media used in the sterility tests, that is one example that we could use, that you need to do growth promotion on the media to make sure that it is going to work. You have an adequate control over the way you have stored it, so it doesn't deteriorate.

Those are the fundamental purposes behind this, as well as the solutions that you make, that the reagents or the supplies that you use in some of the testing are actually part of the validation of that test, and that you are controlling those things.

DR. CALLAHAN: Regarding the growth promotion, once again, that would be something that could be done by the provider of the medium, and not in your own laboratory, correct?

MS. ROBERTS: It depends on how you store it and what you do with it. Sometimes inherently in the sterility tests, there is a growth promotion, and there is a built-in validation that you should be doing, and that is what is meant by that also.

DR. CALLAHAN: Right. Regarding the sterility issue, David Hussong and I put together that piece of the USP chapter, and there was a mention earlier that something very much like that would be included in the guidance for the sterility testing.

I don't think there would be individual growth promotion tests performed on site. I think that was something that was referred to through a certificate from the manufacturer.

MS. ROBERTS: That is just one example that I could think of off the top of my head, but in the guidance document there will be more examples of what we mean. However, that is like the blanket statement that would cover reagents or supplies that could be expected to have some kind of deleterious effect or if you don't control them per se, either temperature control, that is what is meant by this point.

MS. AXELRAD: I should mention that we have put out over there a piece entitled, "Microbiological Validation of Sterilization, Sterility Assurance," attachment to an application for FDG F 18 that describes how this issue would be addressed. It is sort of an addendum to the model chemistry thing that deals with sterility.

Unfortunately, the person who was here, there was somebody here this morning, Paul Stinavage was going to address it, but he had to leave, so we didn't get to that. But anyway, you should know it is over there and get a copy of it.

DR. CONTI: Section 212.60(b). Two comments. One is I assume this will be in the guidance, but "scientifically sound sampling" needs to be defined someplace along the line.

Then, there is a qualifier at the end of that, "when such standards exist." Is there a guidance or some sort of comment on when they don't exist as to what to do?

MS. AXELRAD: That phrase just relates back to when there are standards for identity, strength, quality, and purity. That is what you are testing against if there are such standards, and if there are not, then, the correct wording is conform to appropriate standards, whatever they may be, whatever it is that you think is appropriate for the particular drug product or thing that you are testing.

DR. CONTI: So appropriate standards could be defined by the applicant then.

MS. AXELRAD: Yes, although if there are existing standards of identify, strength, quality, and purity, they have to meet those.

DR. KASLIWAL: A lot of these standards are probably going to be defined in your application, so for components and container-closures, it was finished product.

DR. CONTI: How long do we keep these records for?

MS. ROBERTS: That is in the record section, three years.

DR. CONTI: Oh, three years, I am sorry.

MS. ROBERTS: Are there any other comments specifically in this section?

MR. SWANSON: Under (g) and (g)(1), "Each laboratory performing tests related to the production of a PET drug product must keep complete records of all tests necessary to ensure compliance with established specifications and standards, including examinations and assays, as follows:

"(1) A description of the sample received for testing including its source, batch or lot number, date and time the sample was taken, date and time the sample was received for testing, and its quantity."

Most of us do produce like FDG and do QC testing as a contiguous process. I mean it is overkill in documentation to now require us to document the description of the sample received for testing. I understand we produce it and then we test it. We are not sending it somewhere else for testing?

MS. ROBERTS: That could be easily then captured in the procedure that you would have for this, is that will test as per the batch record. We are testing the whole sample that we produce. But I think the important part here also is the date and the time the sample was taken, and I would assume that when your QC is testing the sample, don't they write on the results the batch or lot number, the date and the time that it was tested? That is what is meant by that.

MR. SWANSON: It's all part of our batch record. It is one contiguous batch record.

MS. ROBERTS: That would meet the requirement then. We are not asking for separate pieces of paper. If there is other means to meet the requirement, if you are saying it is contiguous and it is in the batch record, that would meet this requirement.

MS. AXELRAD: If the batch record has a step in it that says, okay, and after you have filled the vial or whatever, you pull out some and you do the following tests on it, and then your batch record reflects it for batch number, you know, 222, you pull the sample and you did the tests on it, that is all you have to do.

MR. SWANSON: As part of the contiguous process, I don't record the date and time I took the sample, I don't record the date and time I received the sample. We are just doing it, okay.

MS. AXELRAD: But there just has to be some record for somebody who comes back to see if you have actually been doing it, to verify that, in fact, for each batch it was done. We can work on the wording, but that is the idea. I mean you have to have a record. You can't just say, oh, well, we are supposed to do it, and therefore we must have done it.

MR. SWANSON: I don't have a problem with documenting. What I have a problem is over-documentation of things as now required by your regulatory wording here.

DR. KASLIWAL: I think the intent is that, for example, at the end of synthesis, you are going to do an assay, so you have to have a calibration time there, so you will have that in there someplace, things like that, and then when you testing is finished, what was the end of synthesis time, and things like that. We will look at the wording there.

DR. CALLAHAN: You have the data for the test result, so that suggests it was done, so just saying that you do it doesn't add much to it since the blank for the test results, there is data there.

MS. ROBERTS: It is important in the batch record, then, you are going to include this, it is part of the batch record and that on your results that you get, that you are printing out, there is identification of what test result that is for, which particular batch it is for, what size sample you use. That is what we are asking for, is that on the actual laboratory results, there is identification, so that we know what the sample is that you were testing.

DR. BARRIO: In the same section, (g)(1), what do you mean by "quantity"? Many times we have no idea how much. Do you mean volume, activity?

DR. KASLIWAL: Volume, I would think because especially if you are making your batch in a vial, in the model application, I would think you would specify, for example, you take a mL out, whatever volume that you take out for QC testing, so that volume you took out, you would specify you took out that volume.

MR. SWANSON: But you are not talking about getting down to the level of the volume that I spot on the TLC strip.

DR. BARRIO: No, no.

MR. SWANSON: That is what it says.

MS. ROBERTS: What is the quantity of the sample you took to do that specific test, was it a mL, was it one vial, was it--

DR. BARRIO: I think it refers to the quantity of the sample received for testing in general, right? This is my understanding. Is it right?


MR. FERRIS: In other words, the sample used that is extracted from the batch that is used for quality control testing may very well be used--you would run several tests on portions of that sample. You want how much was taken for quality control testing, period.


DR. KASLIWAL: And the model application, you would describe how it is distributed among individual tests.

MR. SWANSON: I am back to the same issue. You want me to record a volume for a drop I put on pH paper, you want me to record a volume for a drop I put on the TLC strip? That is the way this reads, and that is what we are hearing from you, and that is kind of absurd.

MS. ROBERTS: What this really refers to is if you have a separate quality control unit, you are taking a sample, you are sending it to them a different unit where you are kind of losing your control, they are receiving the sample in.

This is when this documentation comes into effect, how much of the sample they received from you, the date and the time that it was received, the batch, the lot number. That is what this number 1 refers to.

If you are taking that sample out, you are sending it to your QC lab, they have to document when they received the sample, the date, the time, the quantity that they received, so that you have a control over where those portions of the sample went. That is what is specifically meant by this.

MR. SWANSON: And as I said, that is not applicable to the way most places do it, which is as a contiguous process. I understand where you are coming from, but the way this currently reads is everybody would have to do this, okay. So, you need some additional wording to somehow indicate that if you are transferring a sample to a separate testing facility or unit.

MS. ROBERTS: That will be made clear.

DR. CONTI: No. (2) specifically says used for each test, so that is the issue, I think.

MS. ROBERTS: In (2), if you need one mL to do a certain test, we would expect that you are always going to use the 1 mL. It's our way of, you know, why did you take 2 mL at this time, why did you only use a drop this time, if your test method calls for a specific quantity. That is what this is in reference to.

DR. BARRIO: The next item is (g)(2), mentioned or determined by your chemical purity, why is it important to weigh the sample under (2), "statement of the weight or measure of the sample used for each test," because it's a relative term, of course, relative by your chemical purity, chemical purity, or solvent percentage, whatever it is.

MS. ROBERTS: This particular section of the regulation for laboratory control is the blanket for all tests that you could possibly do under this, not just specifically for radiochemical purity, a sterility test. This is just the basic documentation that we expect to see when we come in and that we think it is necessary for you to have.

Like I explained, No. 1 is when somebody receives the sample, they have to document what they have received, a weight, a measure, the size, physically identifying it, so they can tell you what they have received.

No. 2 is when you do, when you are actually doing the test on the actual raw data that you are getting. We expect that there is a description of the method that you are using whichever method. It could be a number, it can refer back to the USP, a record of all the calculations performed, whatever you did in hand, we want to see that, and then a statement of whatever size of the sample that you tested, whether it's a weight, a measure, a milliliter, a drop, whatever that method that calls for, we want to know how big was the sample size, if it is relevant of course.

DR. KASLIWAL: Another thing, it is a vague "or measure." A measure could be CPMs.

DR. BARRIO: It doesn't make any difference provided that you are within the limit of sensitivity of your procedure. You are looking at relative terms.

DR. KASLIWAL: That is the idea, that you putting a certain limit of detection, that you are able to pick that up.

DR. BARRIO: You have to remember though, Ravi, that every time we take extra time to measure, weigh, or whatever the sample, this sample is frequently radioactive, then, we don't want to necessarily expose people to radioactivity when these procedures are not really fundamental for the tests we are going to perform. That is the point we are making.

MS. ROBERTS: Then, how do you know how much of the sample you need in order to test it?

DR. BARRIO: It's in your procedure. You have a procedure that says in order to test this solvent, whatever the solvent may be, then I inject in my HPLC so many microliters, but you need to know the volume you left behind.

MS. ROBERTS: Oh, no, that is not what we are asking for. We are asking for the amount that you are required to inject.

DR. BARRIO: It's in the procedure. It will tell you how much you inject.

DR. HOUN: This is the description of each method.

DR. BARRIO: Right.

DR. HOUN: A statement of weight or measure would be in that description.

DR. BARRIO: Barrio.

DR. HOUN: It is not saying at each time you must measure and write it down.

DR. BARRIO: Thank you, Florence.

DR. CONTI: Can I go on to 212.61? Under the Section (a), I have a question about whether or not we have to account for what a distributor may do with the manufactured vial.

We may be able to do a stability study at room temperatures in our hood over a period of time, but obviously, a shipped vial undergoes different environmental challenges between the time it has left the manufacturing facility to the time it is injected into the patient, and that theoretically would fall under the jurisdiction of the state boards of pharmacy.

So, how do we reconcile this?

MS. ROBERTS: The way that I envision this to be done is you make the product and you need to ship it somewhere to the receiving facility. Whoever happens to be the receiving facility, before you release that product, it is still technically in your possession, and you must control what happens to that product.

You are responsible for the fluctuations in temperature or making sure that you know what that is or putting on the carton that it must be shipped within a certain temperature, because you only have stability for it within that temperature range.

DR. KASLIWAL: I think what you would do is when you ship it out, you put a time or, for example, let's say eight hours from time of calibration, your product expires, and as a manufacturer, that is what you need to validate that for eight hours to support that label that your product is good.

DR. CONTI: What I am saying is to do the testing, I can keep that for eight hours in my hood, as a vial, and do the stability testing, during that eight-hour period it's fine, and so that's the standard. But what happens when I ship it to a distributor in the first hours, there are seven more hours to go, and that vial undergoes all kinds of transportation. It is not in my hands, it is not in my control any longer, yet it's in that expiration time?

DR. KASLIWAL: The way manufacturers cover themselves with that is they do testing under accelerated conditions, elevated temperatures.

DR. CONTI: So, it's not really suitable, it's extremes we have to actually do.

DR. KASLIWAL: You can do that if you want to label, let's say, 15 to 30, then do extremes. In the USP currently, what we recommend if 25 degrees plus or minus 2, or if you want a 30 degree, it has to do with the label, and then accelerated 15 degrees higher above that at 40 degrees, but we are not requiring accelerated conditions for these drugs.

MS. ROBERTS: Did we adequately answer your question?

DR. CONTI: I am just concerned that if a vial of FDG is sitting out in the sun somewhere and gets hot, and suppose that other temperature range of what our expiration criteria are going to be tested within, and then all of a sudden, something goes wrong with that product, we are responsible for it, yet, it was the distributor who mishandled the product.

MS. AXELRAD: But you are responsible to the extent that you need to put on the label what the conditions are of storage. If you put on it certain conditions, and you have justified those conditions by doing stability testing, you don't have any control over what they do with it, and nobody is going to hold you accountable for that.

The idea, though, is to establish storage conditions and give directions to the people who are going to be handling it to make sure that that doesn't go out of spec.

DR. CONTI: I don't have a problem with that.

MS. AXELRAD: And that is no different than a regular pharmaceutical that is given a two-year or whatever expiration date, and then shipped through multiple hands on trucks and whatever and ends up on the pharmacist's shelf.

We are not going to hold the person accountable for what happens all in between, but we are going to make sure that when they establish the two-year expiry date, that they have a basis for testing it under reasonable conditions to make sure that it isn't going to just easily lose potency or whatever.

MR. FERRIS: So, you anticipate basically that most stability studies will be done at ambient room.

DR. KASLIWAL: The other precautions that you can take is there are a lot of indicators you can put on the box that will tell you if you exceeded temperatures significantly.

MR. KUHS: Is this something that you have to do periodically, in other words, how often does this need to be done if we are talking about different conditions that may exist for shipping, in other words, if you are shipping in Arizona where your storage conditions might well exceed 40 degrees and where it is likely that the container would be tipped upside down at one point or you are expecting that stability studies would be done in an upside-down vial as well as a right side up vial, and at extreme storage conditions?

DR. KASLIWAL: The model application, yes, we require that it be done upside-down bottle at least the stability batches, and what we require is if your proposed drug for these three drugs is within the strength of reference listed drug one batch at the time of submission, and a minimum of one batch per year after that.

But if it is a higher strength rather than the reference listed drug at the time of submission, three batches at the highest rate of concentration, and again a minimum of one batch per year.

MS. ROBERTS: I see that we have moved into the next section on stability. Are there any other questions or comments under stability?

DR. HUNG: Section 212.60, Subsection (d). I believe the proper labeling for the solutions you have expiration date included.

MR. CHALY: Thomas Chaly from North Shore University Hospital.

Regarding the testing, for seven pharmaceuticals we take less than a drop of sample to do TLC testing, and there is no way that we can record how many mL we took for those testing. We have already written in our procedure that at the end of the synthesis, for example, in the case of FDG, we take 10 microcurie of the sample or 10 mL of the sample, and do the testing like that.

MS. AXELRAD: I think that is what we just said, that that's fine, if your procedure says that is what you do, that is all this means is that you are supposed to describe the method. By saying that the method says take 10 whatever and inject it, that's it. That meets this requirement.

MR. CHALY: But we are afraid that if this kind of wording is there, the inspector comes, and we will be in trouble at that point.

MS. AXELRAD: We are going to train the inspectors, so that they know what to look for.

MS. ROBERTS: We will move on to Subpart H, Finished Drug Product Controls and Acceptance Criteria, 212.70.

DR. CALLAHAN: I just have a comment under (a). It says, "For each batch of drug product, you must establish criteria." You is defined actually under the definitions as us, I guess. That seems like we have the ability to set the standards for strength, quality, purity, et cetera, as opposed to someone like the USP.

So, could we set our own standards for the quality, strength, purity, or do we say that we are going to comply with USP or do we have an option to say something else?

MS. ROBERTS: With any product that you put an application in for, you have the option of whether you are going to follow the USP monograph in which it is usually labeled as USP, which means you will comply with all the specifications in the USP monograph.

You also have the option to determine your own specifications for your own product that you develop, and that goes through the review process.

DR. CALLAHAN: That means for a product that is not in the USP or can that be for a product that is listed? You are saying we can set a different set of standards for a product that is listed in the USP?

DR. KASLIWAL: Right, yes, you can submit in your application a different method of testing or if that is what you are implying. Our view is that USP is the minimum standard for us, and whatever else you want to go from there. That will be on an individual basis in your application.

MS. ROBERTS: And if your standards are different than the USP, you just can't label it as a USP product or claim it's USP.

MR. SWANSON: Along the same line, under (c) it says, "You must conduct laboratory testing to demonstrate that each PET drug product meets acceptance criteria before release. You must establish and document the accuracy, sensitivity, specificity, and reproducibility of the test methods."

If I am using USP test methods, am I required to do all of this documenting the accuracy, sensitivity, specificity, and reproducibility?

MS. ROBERTS: Only to the extent that the USP methods are usually you would have to show that they work in your facility under your establishment with the equipment that you are using. There wouldn't be a full method validation per se if it's in the USP, however, you would need to do like a qualification to show that you are able to test this product in your facility using the USP method.

MR. SWANSON: So, tell me for a TLC test, give me the specifics of what you would be looking for.

MS. ROBERTS: I can't give you the specifics right now, at this time. I would have to look to the USP and sit down and look at that. However, in the USP, there are sections that require you to do the accuracy, sensitivity, specificity, and reproducibility of at least HPLC method and the chromatographic. It is built into the USP.

DR. BARRIO: Then, you mean if we use a USP procedure, we have to do all this, right? What would be sufficient information to satisfy that requirement?

MS. ROBERTS: I can't answer that exactly without having the USP in front of me, but within the USP, under HPLC methods, there is a section, I believe, that speaks to making sure that it's accurate reproducibility, there are system suitability tests that need to be done if you are going to use those USP methods.

DR. BARRIO: Then, this is going to be the standard.

MS. ROBERTS: That would be adequate if you are doing the USP method as long as you are following all of the USP validations that are required.

DR. KASLIWAL: I think you probably need to do some validation of the USP method, but it depends what it is that you are doing. A lot of people don't always use the method that was used in the USP, so the impurities and other materials may be different depending. So, whether your method still holds or not, you need to validate that.

Sometimes the USP methods, you know, and that is assuming that the USP method was validated to begin with, hopefully, USP will have that validation information, but assuming it is validated, you may have to do some minimum validation depending on your conditions of use, for example, let's say TLC method, you spotted so many counts, and you have certain sensitivity that you want to get to be able to pick up some other materials, so you have to validate that you can pick up that amount of material.

For example, GC, you may have a specification set over here, but your working levels are way down here, and that is where you see materials, but since your specifications are set way up here, you are going to have to demonstrate that the method is good from here to there, is it linear or not. Otherwise, it is very difficult to accept whether those specifications, you are going to still pick up those amounts accurately by those methods.

DR. BARRIO: But the USP established that. We have done these with all the procedures I guess for everyone I believe. If it is explicitly indicated, should we redo it again. You are saying yes.

DR. KASLIWAL: You may or may not depending on what validation USP has, and when the question comes, we always go up to USP and we do ask them that we want to take a look at their validation file, and they do comply with that.

DR. CALLAHAN: Can we go back up to (a) for a moment, 212.70(a)? Just a point of clarification, "to ensure that each batch of PET drug...before it is released." In the case of ammonia, it is a sub-batch or some test batch that we run first, whatever we call that.

Again, we are getting back to the terminology, but the point is that there will be a sample of drug that is tested, then, a series of samples will be used, and so it is not actually each batch.

MS. ROBERTS: I think what we discussed at the last meeting was that it was described to us that technically, it is a whole batch, and that is where there is this sub-batch came in.

DR. CALLAHAN: Okay. So, by testing the first one, we have met this criteria.

MS. ROBERTS: By testing the first one and then some middle or the end, or whichever you had described, and we had agreed that that would constitute the testing of the batch.

DR. CALLAHAN: And pyrogens are not included here because they are not a release test, is that true? It said pyrogens do not have to be determined prior to release, is that true?

MS. ROBERTS: Yes, that would be true. The sterility test is the only one that doesn't need to be completed prior to release.

DR. CALLAHAN: I think in the application, pyrogens, it is stated don't have to be done prior to release.

MS. ROBERTS: Is that for every PET drug product? I am sure the LAL test takes 60 minutes, and in other circumstances, it can even be validated for a shorter time period.

DR. CALLAHAN: Pyrogens are not included in 212.70(a).

MS. AXELRAD: We could correct that.

DR. CALLAHAN: In the application, it is stated that it is not a release criteria for FDG, for example.

DR. KASLIWAL: I think basically, on that, the working philosophy, having talked to David Hussong, who has worked on that, we will probably more than likely follow what agreements we reach with USP on that and the component chapter.

DR. BARRIO: On the same page, (d), when we say, "You must establish and follow procedures to ensure," et cetera, (d)(2), for short-life nuclei, for ammonia, that may be a deadly requirement.

MS. ROBERTS: The intent of this purpose was to make sure that even though the laboratory testing was done, that the calculations are checked, it is reviewed, and indeed it is correct testing, everything was running the way it should be, now, you are saying for which product would it be a problem?

DR. BARRIO: O 15, N 13.

MS. ROBERTS: I think there might be in some instances.

DR. KASLIWAL: Let me ask you this. When you are making this quality control, initial quality control lot batch, a sub-portion or whatever, so you are not completing the testing before you start making your regular batches.

DR. BARRIO: Well, we have requirements or we indicate which are the appropriate laboratory testing, that is under (1), for the different nuclei.

I am referring to (2) that goes beyond that, associated laboratory data, and this may be truly impractical. I am not sure what you are referring to beyond the laboratory testing, but, you know, it may be difficult to do that.

MS. ROBERTS: How long would it take, what do you do now with respect to reviewing laboratory data, does anybody review it, or does just the person that does the test, and then it is never looked at again?

DR. BARRIO: In general, the laboratory testing the quality control is the trigger for release if the product is appropriate and meets the standards, we just go ahead and send it out.

MS. ROBERTS: It never gets a second review to make sure the calculations are correct or the right method was used or anything to that effect?

DR. BARRIO: The HPLC or GLC will tell you right away whether you have impurities or not. I mean you don't have to wait too much.

DR. CONTI: In addition, you can take this to the extreme and say associated laboratory data could be the documentation of the HPLC fidelity and quality control, and the starting materials--I mean you can go on and on. I mean how many things do you want to review before you release the product.

I think Jorge's point is that if the laboratory testing is appropriate, that is the release criteria, period, and you can retrospectively review things, but I think particularly with short-lived isotopes, it doesn't make any sense to do this.

MS. ROBERTS: And then what happens for these products that you review them at a later date and find out that there was a problem? They are already gone.

DR. CONTI: Your laboratory testing has already been completed, and it met the release criteria.

MS. KEPPLER: I think we are talking about two different things. I think that Jorge is looking at the laboratory testing is completed, means that he looks at the results of the HPLC printout and evaluates it and says it is good. I think that is what you mean by saying that you are reviewing the laboratory data.

It is not like he is injecting his sample into the HPLC, doesn't look at the results, but releases the product.

I mean do you mean by the fact that your data is reviewed, just looking at the results of that test and seeing that indeed it is water?

MS. ROBERTS: No. What is meant by this is that you are going to review the laboratory data, the HPLC chromatograms, they are signed off on, everything is okay, it wasn't done off-line, all your calculations are correct, and that is indeed the answer that you should get, and I think I am hearing that that never gets done here.

DR. CONTI: I guess the issue is in the language as usual. The appropriate laboratory testing is completed and reviewed might be a way of putting it, such that you check the results to make sure they make sense before you release the product, and then any associated laboratory data that may be related to quality control procedures or other things could be reviewed retrospectively or just this thing just deleted.

MS. ROBERTS: So, you suggest that as a quality control function, they would look at this data after the product is already gone other than like looking at all the HPLC chromatograms or calculations that would need to be done?

DR. CONTI: Again, I think the testing that we are doing is such that you can look at it and see if it makes sense in a very short time frame without sitting there and checking the calibration of the HPLC and doing the other quality control tests associated with this.

MS. ROBERTS: Maybe we are misunderstanding each other. I don't mean all of the quality control tests. I mean you test the product, you get an answer. You have that answer, and then there was documentation to get there, not related to the calibration of the machine, just related to the actual tests, maybe looking at a peak height or making sure there is not other impurity.

I think that is what you mean by makes sense?

DR. CONTI: Right, that the RF on the TLC is where it should be, that is the end of it. You don't need to do anything more with it. That is an instantaneous determination when you do the test.

I guess I don't know what review means here.

DR. BARRIO: I think that is a problem as suggested by Peter, would be testing is completed and reviewed or analyzed, I don't know, something simple.

DR. CONTI: And noted.

DR. BARRIO: Or noted.

MS. ROBERTS: I think we are on the same page. It is just kind of the words, and I will try to work on that.

Is there any issue with the authorization by dated signature?

DR. CONTI: That could be a scribble on the HPLC.

MS. ROBERTS: Acknowledging that the test was completed and acceptable.

DR. CALLAHAN: That would be for each test or the signing off for the release on the batch record.

MS. ROBERTS: It says for the release.

MR. CHALY: For the quality control thing, for example, when you take FDG after the preparation, we take a sample, we inject it into the HPLC, we see the peaks coming out, and depending on the ratio of the peak, whether it is 95 percent pure or 99 percent pure, that is recorded in the batch sheet, and based on that one, will release the sample and we say that this compound is good for patient use. That's the way we do. We don't call anybody else to look at it because we know if it is not good, we are not going to release it.

These are the standard values, 95 percent or 98 percent or 99 percent.

MS. ROBERTS: Thanks. I think that is what I really meant here. It's just a bunch of extra words.

MR. SWANSON: A couple quick comments under (b), "Sterility testing need not be completed before release but must be started as soon as possible," I think USP says within 24 hours because of radiation safety considerations associated with sterility testing, so there is inconsistency between here and what we are seeing in the USP guidelines.

If you go down to (d)(3), "Release is authorized by the dated signature of a designated, qualified individual." Your previous requirements say it has to be by somebody from the Quality Control Unit or person, so you need to say release is authorized by the dated signature of the designated Quality Control Unit or person.

MS. AXELRAD: What are you referring to as the previous? You want it to say that?

MR. SWANSON: Say what?

MS. AXELRAD: You want it to say release is authorized by the dated signature of a designated qualified individual from the Quality Control?

MR. SWANSON: I think your previous statement said that only the Quality Control Unit can authorize release.

MS. AXELRAD: What previous statement?

MR. SWANSON: In this document.

MS. AXELRAD: I don't see that in here.

DR. HOUN: We can take a look at that again.

MS. AXELRAD: We can look at it. I think we were again dealing with the issue, the suggestion that we were requiring two people, and we were trying to eliminate that wherever possible, so we just said release is authorized by the signature of a designated qualified individual. We weren't saying that you had to have a separate person do it.

DR. HOUN: I just want to get the ICP's comment on (b) in terms of if there is a sterility problem, a notification of the doctor who wrote the prescription, is this acceptable?

DR. CALLAHAN: In the case when that product is distributed to a nuclear pharmacy for subsequent dispensing on prescription, it might be adequate for the PET drug producer to notify their pharmacy, because the manufacturer/producer wouldn't necessarily know the prescribing physicians nor the patient.

DR. HOUN: So, in cases where the product is released to a pharmacy, the notification would be to the pharmacy--to the receiving unit.

DR. CALLAHAN: Right, that language would cover it all, if it was the clinic or if it was the pharmacy, even though we don't describe a receiving unit as a pharmacy, I don't believe, maybe we should.

DR. HOUN: And in the case where you have a smaller operation where the receiving unit is actually--

DR. CALLAHAN: Upstairs.

DR. HOUN: --or you are directly doing this into a patient.

DR. CALLAHAN: Well, then, we would have that data, and we could contact the referring physician.

DR. HOUN: Maybe that could be put in guidance in terms of who the appropriate receiving--

DR. CALLAHAN: But just denote that there are some cases where the manufacturer would not have that information available to him.

MR. FERRIS: That notification, as it is written here, happens if there is a sterility positive, and you are saying even without investigation. My point is, is suppose it is anasterile that is causing contamination, are we notifying the physician that there is a sterility failure or that there is a potential sterility failure?

MS. ROBERTS: What is intended by this is a true sterility test failure. We wanted to waste time, however, if it is the laboratory's fault, we wouldn't want to unduly alarm the receiving facility either, but I can see a problem in the case where if it is taking the laboratory a week to do the investigation, I would find that a problem, if it takes that long to notify them of a sterility test failure after you have figured it out.

This is what is envisioned, is that if there is an initial sterility test failure, it triggers some kind of investigation, and then it is determined whether--if you can rule out the laboratory or if it is clearly the laboratory's fault, that is easy enough, it is not a true product failure, however, if the investigation isn't sure right away, I would assume that you would notify them of the potential then or the actual sterility test failure, because I would hate for you to wait two more weeks for them to retest and especially since you are not keeping or you don't want to keep retains.

So, that would be troublesome to try to do a retest if you cannot duly rule out a product sterility failure by a true laboratory error, you should notify them as soon as possible.

DR. HOUN: And I think probably in guidance, we can describe what does immediately mean or with some assurance. I don't think we should put it in the regulation 24 hours or 12 hours or whatever.

DR. HUNG: What happens if the failed product, the failed sterility test has been injected into a patient, should the patient be informed by the prescribing physician, and if so, what sort of a time period?

MS. AXELRAD: Well, we were just suggesting that the physician ought to be notified, but the comment we just got back was that we shouldn't notify the physician, we should only notify the pharmacy who received it.

I guess I would wonder since sterility test failures are supposed to be incredibly rare and very unusual events, and since that is probably the single most significant failure you could have in one of these products that would actually affect a patient, who would have already gotten it two weeks ago, that perhaps notifying the physician in these very unusual circumstances, and perhaps even the patient, might be appropriate.

DR. CALLAHAN: What I suggested was that the manufacturer may not have access to that information. That is why we notify the pharmacy. If the physician is available to us, we would notify the physician. However, I don't think we would directly notify a patient. That is the physician's prerogative based on his clinical judgment of the patient's condition, what the risks are, and so it is not appropriate for us to contact patients.

MS. AXELRAD: But you could probably get the name of the physician from the pharmacist and make sure that that notification is made to the physician.

DR. CALLAHAN: Again, we are crossing this line that we want to draw somewhere.

MR. SWANSON: When you have a product failure with a traditional drug, do you require the traditional drug manufacturer to go out and identify each physician that wrote a prescription for that drug? Of course, you don't.

MS. ROBERTS: The requirement as it reads here in the regulation is to notify the receiving facility as soon as possible, immediately, and the physician if you know who the physician is.

DR. CALLAHAN: That's fine.

MS. ROBERTS: Are there any additional comments on this section?

MR. SWANSON: But that is not what that says, though, so be aware of that.

MS. AXELRAD: Right, but on the one hand, you are arguing that you are not like manufacturers, that you really more like small operations in the practice of pharmacy, but then you are now arguing that, you know, well, manufacturers aren't required to notify the physician or the patient, why should we.

MR. KUHS: I think there needs to be a clarification of the receiving unit, and the receiving unit could be a pharmacy, in which case you would not know some of the end users. The receiving unit could be a physician in which case you would document that, and you would notify him. I don't think they are mutually exclusive of one another. We are not saying we will do one and not the other. It's that you use the distinction based on where you distributed it to.

DR. CALLAHAN: You can only act on the information that you have in hand, and it may be impossible to gather beyond a certain level. What's all we are trying to say.

MR. SWANSON: Fundamentally, in response to what you just said, we are conceding the fact you are going to regulate us as a manufacturer. That doesn't entitle for you to regulate us as both. Okay?

MS. ROBERTS: Are there any more comments under that section, under Controls and Acceptance Criteria?

Then, we will move on to 212.71, What other actions must I take if a batch of PET drug products does not meet the acceptance criteria?

DR. CONTI: In regard to complaints, if it hasn't met acceptance criteria, would it have been released in the first place unless I guess there is potentially a test that would be done retrospectively, but I am reading this as acceptance criteria for release, and in which case, why would--I mean I guess our distributor might complain that they didn't get the vial, but what are you referring to here as far as complaints are concerned?

MS. ROBERTS: Where are you reading?

DR. CONTI: 212.71(a).

MS. ROBERTS: Oh, what this means is that when you are doing an investigation of a failed drug product, you are required under these regulations to keep complaint files for the specific drug product. We would expect that an investigation would include a review of your complaint file for that drug product, not for that particular batch, but for the drug product as a whole.

MS. KEPPLER: And this would be I think, Peter, not if they released it, this would be for probably unreleased product, you know, you had a run failure, you should keep track of run failures, investigate the causes.

Is that the purpose of this?

MS. ROBERTS: That is the purpose of that. Also, we ask that if you do have a run failure, that you look back at product complaints to see if you started to have a problem maybe and that is evident by complaints, you might never have a complaint, and then that will be easy, but if you do have complaints, you should see maybe if it is attributable to this same problem that you had that the batch failed.

Are there any other comments or can we move on to Labeling and Packaging?

MR. CLANTON: Jeff Clanton, Vanderbilt University.

I don't see where this particular section allows for distribution pre-release, which is a common practice.

MS. ROBERTS: I don't understand what you mean by pre-release.

MR. CLANTON: In other words, the material has gone through the production cycle, it is packaged for shipment in interstate commerce, it is shipped, and the QC is ongoing during the process of while it is in transit.

MS. ROBERTS: We discussed that earlier, and we have made a distinction between distribution and release. You are allowed to put the product into distribution prior to release. You just cannot administer it to a human prior to release.

MR. CLANTON: I just didn't see in the section where it allows that, though.

MS. AXELRAD: I think it is under Distribution, 212.90. "You must establish, maintain, and follow procedures for the control of distribution to ensure that only those products approved for release are used, and that the process of shipping will not"--I mean that is just one of the places, but that is where it really appears I mean other earlier section.

MR. SWANSON: Under (b), the date and time it was prepared, do you really mean date and time of calibration? Prepared is when do I start that.

MR. FERRIS: And why use that on the label? It should be calibration time.

MR. SWANSON: It should be calibration time, I think. Probably in your guidance document, you are going to have to address, do you mean in the label affixed to the actual immediate container of the drug or the shielding of the drug? For example, you require strength, okay, and strength is something is we determine at assay, and if we have to put that label on that vial, that is a fairly significant radiation dose. So, that is one where you may want the strength to appear on the lead shield associated with the product.

DR. KASLIWAL: Some people use the strength test.

MS. ROBERTS: Any other comments on labeling?

MR. FERRIS: On (d), when you say "labeling and packaging operations must be controlled," I would like to get a sense of--to me, that means reconciliation, label reconciliation. Ultimately, a label is issued, counted.

MS. ROBERTS: That is not the intent here because we understand that a lot of times you just handwrite out a label and slap it on the bottle.

This is meant--and it will be covered also in the guidance--that if your operation is large enough where you are using preprinted labels for what you are doing, then, it is more of a labeling reconciliation control, but for the purposes of what I think the PET center operations are here, this would also cover, if they are handwriting them all beforehand, they are not going to be mixed up. There is some way to prevent that.

Now, if you are just writing one label and sticking it on the bottle, there is really no possibility for a mixup, so that would cover it.

MR. FERRIS: So, that is going to be applied in a variable way depending upon the scope.

MS. ROBERTS: Yes, depending on the scope of the operation.

If there is no other comments on labeling, I think we covered distribution earlier. Are there any other comments on 212.90, Distribution?

MR. SWANSON: The only comment that I would have would be under (b)(2). Again the patient's prescription, if applicable, or any control numbers is not within your purview.


We will move on to Subpart K, Complaint Handling, 212.100.

MR. FERRIS: Could we go back to 212.90(a). You have that "prescriptions are reviewed to assure that they have been properly filled."

MS. AXELRAD: We already took that out per this morning's comment.

MS. ROBERTS: Complaints? I meant Complaint section. I think we are all starting to fall asleep here and we still have a lot more to cover.

MS. AXELRAD: One page.

MS. ROBERTS: No, not this. If you don't have any comments on complaints, we will move on to Records, Subpart L, 212.110.

DR. HUNG: Do you accept computer records?

MS. ROBERTS: We have a whole regulation that covers computer records, and that is across every center. That is our Part 11 on Electronic Signatures and also on all Electronic Records. If you wanted to keep electronic records, you would have to comply with that Part 11.

MS. AXELRAD: Are you planning on giving us written comments? We have indicated that we will accept written comments on this preliminary, sort of unofficial draft on or before October 13th is what we said in the notice. So, we will take into account everything that was said at the meeting on the draft and anything that is submitted to us in writing before that date, and we will go back and work on this some more.

MR. KUHS: I think there is one specific charge of drafting a statement on the end of the end of manufacturing process and dispensing under practice of pharmacy and medicine. I think there is a lot of room for clarification on that issue.

DR. HOUN: And any other specific revisions, the more specific you are in terms of wording, the closer we will get.

MS. AXELRAD: Why don't we open it up to the audience.

MR. CHALY: How long do we have to keep cards?

MS. AXELRAD: October 13th.

MR. CHALY: No, how long all the records are maintained in the PET centers.

MS. AXELRAD: Three years.

MR. MOCK: One very specific question. With all the different QC testing that is required, what happens if my GC column, capillary column used for doing residual solvents or whatever happens to break or the computer fails and the piece of equipment doesn't work that particular day, yet, my track records shows that I haven't had any problem at all with this particular test, can I still release the sample for use because I am not going to get that GC fixed until tomorrow maybe or next week or when the new part comes in?

That might be the place where we can keep a residual sample to do some of these tests after the fact, but I am just concerned with all the different documentation, you know, tests that need to be done, if for some reason a device fails, am I shut down for however long it takes to get that GC fixed or the computer that the software is operating on, the hard drive crashes, and I can't get it fixed immediately, what are the limitations on product release based upon past history?

MS. ROBERTS: None. Skip block testing is not permitted for these drug products. All release criteria must be met prior to release of the drug products.

MR. HAMMES: Dick Hammes, University of Wisconsin.

Just a general comment about the process. I listened to what you had to say about process validation versus end product quality testing, and I think you need to listen to the reality of what kind of resources we have available out there in the smaller institutions just doing our own in-house use.

I can definitively tell you that the University of Wisconsin does not have the resources to do the documentation and validations and everything that you are talking about today. Granted, we may be at the low end of the curve here and you might want to bring us up.

I have been trying to bring us up to that level for 18 years now. We have been making FDG for 18 years. We have never had an adverse event. We have had successful scans clinically, and if this goes through as it is presented today, I fear that our PET center is not going to continue.

MR. BRESLOW: Ken Breslow, PETnet Pharmaceutical Services.

I see in the proposed GMP two terms, validation and verification, and I am used to thinking in terms of validation and qualification. I am a little puzzled why a new term of verification was introduced when the term that is most widely used, qualification, would be acceptable.

I think the committee and other people in this room have had a hard time understanding what the difference is between validation and qualification, and I will assume that your definition of verification is equivalent to the definition of qualification.

There was discuss earlier about what do I have to validate, what do I have to verify or qualify, and I think an accurate differentiation could be made in considering the USP test methods as valid methods. USP has published them, and therefore they should have on record the validation of those procedures.

If, as a manufacturer, we determine that the USP methods will work for us in evaluating the drug, then, no further validation of the USP method is required, however, we must demonstrate as a manufacturer that the equipment and personnel that are being employed in the testing of the final product according to the USP validated procedure is qualified to perform that validated test.

Here is where we have to evaluate the sensitivity and specificity and linearity and reproducibility, and all the same things that we would ordinarily also have to do if we were going to validate a new test method.

MS. ROBERTS: I think I was trying to say that when I was explaining what needed to be done for USP. We are going to look at the terms and reclarify. Verification is listed in the GMP I think once, and so therefore it was defined.

We are going to revisit that to decide what we are talking about on verification and validation, but I don't disagree with what you have been saying about the USP method. I just think there is some different wording in what we have been talking about as what is required.

We are taking that under advisement, and I had addressed that earlier.

MR. BRESLOW: Okay. One last comment on this point.

The valid points that several persons in this room raise as far as the economics and the staffing and equipment availability and the expertise to do some of the equipment qualifications is at the same level as if we were going to validate a test method initially.

I mean the effort that needs to be expended in qualifying the test method, equipment, and personnel is at an equally high level, and the economics behind that is significant, and the expertise available at many traditional PET sites is lacking because in many instances, you don't have an analytical chemist available at these sites, and are we supposed to go hire analytical chemists who are expert chromatographers that have experience in the GMP regulated industry?

So, I want to reinforce the point that other people made, is that it is a significant and costly exercise.

DR. BARRIO: I would like to address the comment made a minute ago about the possibility of equipment failure. I just say that this is probably a very rare occurrence, but I think the question was very good because that possibility always exists.

One alternative in an emergency situation could be to allow the PET center for this period where this equipment may need to get repaired, is to allow an alternative procedure that may replace, for example, a GC that is not working, another procedure that could allow to have an idea or a good idea of how that radiopharmaceutical is in terms of quality control.

I am not proposing this as a loophole, but rather as an emergency consideration for this kind of situation.

MS. ROBERTS: I don't think that would be a problem as long as long as you are doing a release test for that specific product, as long as you have thought about from since you have so much history with these drug products and with your equipment, you should know which ones will usually break down, and I would expect that you would have had then, if you are accustomed to this, an alternate method already prepared that you would use in case of an emergency.

DR. CALLAHAN: I don't know of an alternate method for GC for organic solvents, and that is a good example, because that is the only piece of equipment, and in our institution, we do not have a redundancy. I mean it is one unit. I have got six patients upstairs who haven't eaten since 10 o'clock this morning, and they are waiting for their FDG, and I can't do an ethanol concentration because that GC died, and I haven't had an ethanol even approaching the limit for the last 10 years, then, I have real hard time saying that those patients have to go home and not get their diagnostic study and come back some other time.

That is a specific example, and it could be another piece of equipment, but that is a good one because most people don't have a bunch of GCs laying around.

MR. FERRIS: If a trend analysis is done on a periodic basis, one can consider an emergency parametric release.

MS. ROBERTS: Not at this time under these regulations, but we will take it under consideration, however, there is no provisions for it in this current regulation.

DR. CONTI: But you have an opportunity to write the regulation. Just go ahead and edit.

MS. AXELRAD: We will think about it.

MS. ROBERTS: I just have a question about what happens if there is too much organic solvents in a product, what is the health and safety risk?

DR. CONTI: If your trend analysis says it is unlikely to be occurring is one issue, the likelihood of it being there is very small. These particular solvents, even at these concentrations or even above this, pose very little risk to the patient.

So, even if you put in--I don't remember the exact concentrations, what we actually use in the process--

DR. CALLAHAN: It's 0.4 percent and 0.5 percent for acetonitrile and ethanol respectively, for example, and those are huge. I mean we never approach those kinds of limits. For the ethanol certainly, that is not an issue.

Acetonitrile, again, if you can go up to the level of 0.4 percent acetonitrile in a product and accept it, I mean that suggests that it is not very risky, and the fact that we are down way below that constantly, consistently, to reject an entire batch of material and deprive the patients and inconvenience--it's a case if we were not under the GMP veil here, we would exercise professional judgment.

I know this is not the arena maybe to discuss professional judgment, but this is the place where I would say that I am releasing this product, and I don't think that I am increasing the risk to my patients whatsoever, and I would go ahead and do that.

MR. FERRIS: Send a sample for analysis later.

DR. CALLAHAN: And whatever else we do, but I mean just at that point, where it is a decision process, you have got to go and decide whether you are going to take care of the patients or not, that is where I would invoke my own professional judgment.

MS. ROBERTS: We take your comments and we will think about it, and we will take what you said, but I just wanted to clarify that parametric release requires a laboratory determination in order to release a product. It's not a skip lot that is the absence of testing.

MR. FERRIS: The laboratory determination I was making there was trend analysis, that's all.

DR. CONTI: The other thing is if you would consider doing this, you may want to include as sort of a mechanism to notify the pharmacy that is receiving the material that this is the case or the physician that is receiving the material that this is the case. Then, it becomes their discretion whether to administer it to the patient or not.

MR. CHALY: I have a general question. We haven't heard anything, how much we would have to modify our laboratory to come into compliance, because I am very concerned about that one because recently we had to spend about $300,000 to satisfy the New York Environmental Agency for the emission of the radioactive materials, and now we have to spend another $500,000 for these things, most of the hospitals haven't approved any of these things, so we would like to get an idea of how much laboratory modification is required to come into compliance, so that they can close the facilities as early as possible.

MR. SWANSON: I guess I still have a concern about whether you call it qualifying or validating all of your test procedures. Here is the scenario I see. Take something like FDG. We now maybe have 100 different manufacturers, 100 different PET facilities, to get each of the 100 PET facilities validating or qualifying what may be essentially the same test procedures, I mean to me that--I mean it is not like traditional manufacturing where you have one drug product made by one manufacturer.

Can't we think outside of the box here and try to come up with a way to say okay, if you are doing this test procedure this way, under these conditions, there is not a requirement for you to independently, at each site, validate or qualify the test procedure at least for our traditional products, because this scares me a lot.

DR. KASLIWAL: I think some of the methods, literature has some very good validation. You can probably use that as a reference point.

MR. SWANSON: Who? What?

DR. KASLIWAL: For some of the methods, the literature may have very good validation data. You could use that.

MR. SWANSON: But you are still requiring each of the 100 facilities to do essentially the same thing.

DR. KASLIWAL: You can obtain that centrally if you want at some point, it can be validated centrally and given to you for those conditions, the specific conditions that you use.

DR. BARRIO: That means that a procedure that has been validated by USP, centrally, as you say, will not need localization.

DR. KASLIWAL: For example, the USP method to me seems like it is valid for a no carrier added method of synthesis using acid hydrolysis. FDG, right, that is what it is in the literature and that is where it is coming from.

Now, if you use a very end of that method, under the conditions that you use, for example, in TLC, if you go to basic hydrolysis, whether you can pick up mannose triflate, which you can form by isomerization under basic conditions, there are literature references for that, but whether your method can pick that up or not, that will be an issue, and at that point we are going to have to see under your conditions of use whether your method is good or not.

DR. BARRIO: We discussed this issue, as you remember, in the context of the USP. If you have a fluoro-mannose, that would be only if you have an electrophilic procedure going on, and the procedures we put in the USP monograph are not the same to detect that isomer.

One aspect of our discussion was if any center decided to use the electrophilic procedure, then, they should provide the procedure to verify the presence of whatever isomer or impurity may exist on that particular procedure.

DR. KASLIWAL: Right, and basically, that is the philosophy we follow, under the method that you use to manufacture and the conditions that you use in the procedure, whether those things are still valid.

MS. AXELRAD: I think that ends our discussion of this draft of the GMPs. I am going to suggest we take a five-minute break, very short, come right back, and spend a little time on the procedures.


Approval Procedures Update

MS. AXELRAD: For this next part of the agenda we are going to try and cover some issues essentially with the approval procedures. We are just going to sort of update you on chemistry, clinical pharmacology, and biopharmaceutics, pediatrics, and user fees, and answer any questions you may have.

Keep in mind that, as I said this morning, what we are doing is developing a guidance document--I don't know if it will be one or two--but basically, that will lay out the procedures for submitting an application under 505(b)(2), which is something that is based on the literature, and also an application under 505(j), and the differences between those are that for (j), you have to demonstrate sameness to a reference listed drug.

So, the first application for ammonia, the first application for O 15 water, the first application that comes in based on a literature review will be a 505(b)(2) application, and after that the next applications that come in, if they can demonstrate sameness to the reference listed drug, the first one that has been approved, or in the case of FDG, sameness to FDG in the Peoria application, then, they could come in as abbreviated new drug applications under 505(j).

Anyway, we are going to lay this all out. There is not a huge amount of difference procedurally between those two types of applications. They will both have basically the same chemistry submission and they will basically follow that form. We will go through all the different provisions of the regulations - you will have a debarment certification.

We went through these at the February meeting, but the guidance will lay out specifically exactly what you have to do - the patent certifications, the debarment certifications, the financial qualifications of investigators, lots of very, sort of relatively small procedural things in our regulations, and the guidance will sort of step through that for each section of the application.

One of the biggest parts, of course, is the chemistry section, and Ravi is going to tell you what he has been doing in terms of the model chemistry application.


DR. KASLIWAL: I think everybody has a copy of the three applications that we would like to have your comments on - the F 18 FDG, N 13 ammonia and F 18 sodium chloride.

F 18 FDG, we discussed that in our last meeting. Since then, we received a number of comments for two ICP, as well as two other people, and we have taken those into consideration and incorporated the relevant comments into this draft.

Basically, each application covers what your drug product is, your components, and what the drug product's quantitative composition is, provides for control of components and raw materials that you use, for reference standards that you may use.

For example, in FDG, if you use a process that is different than some of the compounds that are listed in the reference standard, you don't have to use those. Provides for manufacturing testing facilities. You have to tell us where it is manufactured and where it is tested.

If there is more than one facility within your application, you can include that. Provides of the manufacture of drug substance, what happens in your CPCU, the batch formula that you use, all the controls, and then once the product comes out, how is it formulated, in what vehicle, or whether it is not formulated, it remains as it is you could describe here.

Also, the container/closure information. Accordingly, if you use a pre-sterilized container/closure from a manufacturer in good standing, you could provide accordingly limited information versus if you want to make your own container by container/closure separately, you want a seal and you sterilize it accordingly, the information will need to be much more in nature.

Also, provide for control of finished dosage form and the description of analytical test procedures, and each procedure, what you need to have. Also, the microbiological validation and the validation data, that needs to be included here.

Basically, in the last discussion what Jane mentioned, a two-page document that is a draft document that is also available on the table, provides a little help, a guidance, what should be included in that section.

There is a table at the end of the batch data, and as I said previously, it provides for the conditions you should store under, your vial. Really, your conditions depend on what you want to label, and how is the product that is going to be shipped then stored, and to support an expiration dating period.

Basically, one batch, if the batch is going to be manufactured within the strength limits of the reference listed drug. That is applicable to the NDA, but if it is a 505(b)(2) application, and you are bringing it especially in the higher strengths than what is listed, then, you are going to have to provide three batch data at the highest rate of concentration.

The current reference listed drug for FDG has a range of 4 to 40 millicuries per mL at the end of synthesis time, so if it is higher than 40 millicuries, you are going to have to bring it up to three batches per mL.

MR. SWANSON: Say that again.

DR. KASLIWAL: The current range of strength is 4 to 40 millicuries per mL at the end of synthesis time.

MR. SWANSON: That is the current NDA?

DR. KASLIWAL: That is the reference listed drug, yes. That's in the package.

MR. SWANSON: How are PET centers going to be made aware of what the characteristics of the current NDA-approved drug are?

DR. KASLIWAL: I think whatever is available in the package insert, that is disclosable, and we can disclose that from the agency's point of view.

MS. AXELRAD: I think that people here in this room are affiliated or associated with this application. One question would be whether the parameters in there, that it eligible as a reference listed drug could be made publicly available, so that people could reference it, and we could give a list of what kinds of information people would need to know.

It is a very different situation than the standard generic, where they way they figure it out, is they go buy the stuff off the shelf and they analyze it. That is the way a generic usually demonstrates that it is the same as the reference listed drug, but this, it is a little difficult to do. So, we ought to find some other way for the parameters of that reference listed drug to be made known, so people can see whether they are the same or not.

MS. KEPPLER: Is it in the chemistry DMF, and if so, I mean ICP owns it, so we might be able to make it available through the ICP, the characteristics of it. Is everything in the DMF?

MR. KUHS: The original DMF--the NDA has been supplemented once since then--the original DMF contains information at a lower batch strength in a reference to specific concentration at end of bombardment rather than end of synthesis, and the supplement that we filed was specifically to change the range of concentration and the reference to end of synthesis rather than end of bombardment.

MS. KEPPLER: Maybe the two of us could get together and put together a--

MR. KUHS: I don't see a problem with that.

MS. KEPPLER: --a descriptor of the reference listed drug.

DR. KASLIWAL: The criteria for generic, maybe somebody can explain--

MS. AXELRAD: We are going to provide this. We are also probably going to be presenting at the ICP meeting in Vancouver some more details about how you demonstrate sameness, but we can definitely get you a list of the parameters that need to be known to somebody in order to demonstrate that they are the same, strength, and some of the other characteristics.

If we made those available, then, perhaps you all would be willing to make available to people what they are.

DR. KASLIWAL: But I think whatever that is needed to show that, it is available on the package insert. The other impurities, if present, they can be controlled in guidance document what the limits are allowed.

Basically, after that, a draft copy of the vial and outer packaging labels, and basically a claim for categorical exclusion from performance on EA, and we have provided the statement here, which you can simply fill out.

With FDG, I think the thing that I want to mention is that the model application allows no carrier added method of synthesis, and the specifications are drafted from a point of view that it involves acid hydrolysis, and it is clearly stated if you use any other alternate method of synthesis, then, your specifications and method need to be appropriately evaluated in light of that to show that it's okay.

MS. AXELRAD: Is there any way for us to find out how many people are using the electrophilic process, if anyone? We are hearing nobody, then, I have heard four. Maybe we could ask the question at the ICP meeting.

DR. CALLAHAN: Using electrophilic or a base hydrolysis? Is that two different things? Of course, it is. So, I think we have got a problem here.

MR. JACKSON: Mark Jackson. The only electrophilic that I am still aware of is at Vancouver. They still make FDG in that method. I know of no one in the U.S.

DR. KASLIWAL: They are in Canada, so we don't have to worry about it.

DR. CALLAHAN: Hearing from people that I have discussed it, the base hydrolysis issue is going to become more of an issue.

DR. KASLIWAL: No, the only issue that are published, under basic conditions, depending how you employ, you can have inversion of configuration, and then you are going to have to show that. You could still use it, I am not saying you can't use it, but if you use it, then, you have to show that actually you don't do that.

DR. CONTI: We actually went through these documents in detail, so I might suggest in order to move things along, we actually just go right to some of the points that we had, if that would be reasonable.


DR. CONTI: Jorge, do you want to lead that, do you want to go through that?

DR. BARRIO: We had a few comments. We were wondering--I was not following this carefully--where the 2 percent fluoride ion impurity came from. Do you remember, Ravi?

DR. KASLIWAL: No. We haven't discussed that in the past, and that is one of the points that I want to discuss. Where it came from was the recommended dose of FDG for scans is 5 to 10 millicurie in the package insert, so if you have 10 millicuries without the limit, and the radiochemical impurity allowed is 90 percent, that means without a limit to that, you can have up to, let's say, a millicurie of free fluoride, you can still pass the product.

At the same time, if you go and look at the package insert of sodium fluoride, the recommended dose for imaging with sodium fluoride is half a millicurie to 2 millicuries. So, we have to the free fluoride amount below what you could get a useful image.

DR. CALLAHAN: I would like to comment on that. That package insert is based to 1974 or something when people were using rectilinear scanners to do bone scans in a planar mode. It has nothing to do with doing PET scans with F 18 fluoride. So, that had more to do with the instrumentation and how much you could get, and it was distributed around the nation, and there were a lot of issues of why that was for I think it was 4 millicuries, as I recall, because actually, I have been around long enough to actually have dispensed a lot of that material, and that has nothing to do with this and is irrelevant.

If you were doing bone scans with F 18 fluoride, I will defer to--

DR. CONTI: I would say at least 10 millicuries to do a good scan, to do a fluoride bone scan. You can get away with a little less, but--

DR. CALLAHAN: One is an impurity and one is a desired product, so I don't see how they relate. I mean if it is a radiation dosimetry issue or what.

DR. HOUN: Would it interfere in terms of 10 percent sodium fluoride with an FDG product the way it would appear on the scan?

DR. CALLAHAN: Probably not although I can't validate that

DR. CONTI: It can be visualized. The question is does it interfere with the diagnostic quality of the scan, and that is subject to question. I don't know if that is true or not, because you are visualizing bone, which normally you are not going to visualize with an FDG scan for the most part. So, it could potentially interfere.

DR. CALLAHAN: But when the radiochemical purity limit was set in the USP for FDG, I assume that there could be up to 10 percent of something else, and that could be fluoride or it could be partially hydrolyzed FDG.

So, now, under this guideline, under this application, you could have a 97 percent radiochemical purity and fail, because that would mean you had 3 percent fluoride.

DR. KASLIWAL: Well, the fluoride amount is basically--I explained where it's coming from--the recommended doses, but you are right, you can have a partially hydrolyzed product, as well.

DR. CALLAHAN: I would also challenge that logic to get to that point using the original package insert from sodium fluoride from the mid-seventies.

DR. KASLIWAL: I think that problem that we have is that's the only document, the evidence we have or information we have.

DR. HOUN: We can ask this committee in terms of if sodium fluoride would interfere with the FDG imaging and if that was a possibility, at what limit would people comfortable, or the other issue is that we have to think of the pediatric group, too, and if they were being imaged with FDG and had sodium fluoride, is there a particular concern about their exposure to sodium fluoride you would want to limit.

DR. CALLAHAN: It becomes a radiation dose issue then and the dosimetry from fluoride is different, but still the critical organ is still the bladder, so if the total administered dose is 10 millicuries of the substance, I bet if you did the numbers, the bladder dose isn't going to be all that different from the contribution from 10 percent fluoride compared to the dose to the bladder already from 10 millicuries of FDG is probably going to be a small factor.

DR. CONTI: The bottom line is no one knows clinically because this has never been studied before to my knowledge. I know people do occasionally combine the tracers because you get anatomical delineation of bone, which is sometimes helpful in the diagnostic test with FDG when you combine the two.

So, that is sort of a trick of the trade, so to speak, but the fact is that there is no clinical data to show at what threshold free fluoride interferes with diagnostic interpretation of an FDG scan that I am aware of.

DR. CALLAHAN: And you do have another document, you have the USP specifications for FDG, which says it allows 10 percent radiochemical impurities.

DR. BARRIO: Sorry for asking the question.

DR. KASLIWAL: Do people see a lot of fluoride in their product?

DR. CALLAHAN: A couple of percent, yes, absolutely.

DR. KASLIWAL: What is the normal level people see generally?

DR. CALLAHAN: Anywhere from zero to 2 to 2 1/2 percent.

MR. BRESLOW: Good point. We do see counts of zero, which could be fluoride, may not be fluoride, but most likely it is, and it is not unusual to see zero counts and it is not unusual to see 2 percent, 2 1/5 percent on occasion, rare, but it does happen.

MR. KISELEV: May I make another comment? Maxim Kiselev, Eastern Isotopes, Sterling, Virginia.

I think I have a unique experience in this area because we are manufacturing at somewhat higher levels than most other facilities. As far as the release testing is concerned, there is no problem. You can have 99, 98 percent. However, in our experience, the stability of the product is not good enough to maintain over 99 percent over the long period of time.

We have done some extensive stability testing. It appears that the reasonable level which could be achieved without considerable amounts of stabilizer is probably about 95 percent, but if you specify anything more than that, then, they end up either adding stabilizers, which not everybody likes because they are not in use traditionally, and therefore the clinical data may be not relevant to compare with old results, or again having to dilute the product and inject large volumes, which also not everybody likes. So, I think that 95 percent is probably the reasonable amount, and as a number of people pointed out, there is no clinical data to point out that there is a problem at that level.

Well, 90 may be a bit too low, but it is traditionally level, and there is a big body of data that we had analyzed or the agency had analyzed and concluded that the drug is safe and effective, whereas, it was produced under the 90 percent limits.

DR. KASLIWAL: Do you know what concentration you think the stability problems have?

MR. KISELEV: The stability problem starts at the concentration exceeding about 100 millicurie per mL. In order to make the FDG a truly clinically useful drug and make it available to the wide patient population, I think they are shooting at concentrations in excess of 200 for logical in distribution, logical and cost effective distribution.

MR. SWANSON: Also, correct me if I am wrong, doesn't the drug have to meet its acceptance criteria throughout its expiry period?

DR. KASLIWAL: That's right.

MR. SWANSON: So, then you are dealing with a radiochemical impurity of 0.25 percent at time of calibration in order to maintain it below 2 percent throughout an eight-hour expiration period? Pretty difficult to achieve.

DR. KASLIWAL: Maybe the solution is when you get to very high radioconcentrations, you have to use stabilizer, who knows.

DR. CONTI: Again, the experience is such that at the concentrations that we have been traditionally producing these isotopes at, which is below this level, at the 90 percent radiochemical purity level, it has not interfered with the clinical utility of the test, so I would encourage you to focus on that piece of information as the baseline.

If someone goes to an extreme, then, they would be then required to document issues, such as stability or the impracticality of doing a scan with that level of contamination.

MR. WATKINS: I have a comment. My name is Len Watkins from the University of Iowa.

We have done quite extensive studies in this area, as well. We don't make by far the amounts that Maxim uses, but with a 500, 600 millicurie batch in 18 mL, we get most of the time zero percent fluoride in our product when we start.

I have taken samples throughout the day. Whenever we inject a patient, I have taken a sample at the same time and measured it. As the day goes along, we see increasing amounts of fluoride, usually not exceeding 2 percent, but the higher amount of activity you start off with, the more radiolytic the composition we have.

I have asked the physicians who read the scans to tell me if they see any difference during the, and we have done I would guess probably 100 or more, and the physicians have never reported back that they have seen any difference in the scans between the early part of the day and the late part of the day.

MS. AXELRAD: We hear your comment. I think we will have to look into this some more, and we may have further discussions on this.

Do you have other comments?

DR. BARRIO: Yes, on the same page, under pH, when you refer to pH paper and pH reference standards, you are referring to the color scale when you use the pH paper or you are talking about something else?

MS. AXELRAD: Is it page 11?

DR. BARRIO: Page 11 under pH.

MS. KEPPLER: You are talking about the color scale on the box?

DR. BARRIO: pH paper and pH reference standards.

DR. KASLIWAL: No, I was talking about the standards, the pH standards, the drops.

MR. SWANSON: Do you have to do that with each pH test or is that part of your validation of your pH paper?

DR. KASLIWAL: You could probably do it as part of your validation.

DR. LEUTZINGER: I think so. I think you can just validate, doing the validation.

DR. KASLIWAL: The specific paper you are using.

DR. BARRIO: Also, at the bottom, you mean osmolarity, I guess.

Finally, the measure of glucose concentrations, why would anybody like to or would have to calculate the amount of glucose present, who cares really?

DR. KASLIWAL: No, the calculated amount is based on your batch formula. In the batch formula, you are specifying the amount of substrate on that, you can calculate and specify here the maximum amount of glucose present assuming everything hydrolyzes.

DR. CONTI: What is the purpose of knowing how much glucose?

DR. KASLIWAL: That is the description of your product. You have to know what is in your--you know, your definition of the product, what's in the product, and we are not requiring that you need to test it. You need to calculate and just specify there that will be the maximum amount present. It's a calculated amount from the amount that you use.

DR. CONTI: Dr. Barrio mentioned osmolality, should the correct thing be osmolarity?

DR. KASLIWAL: I thought in the package insert it's osmolality--it's molarity or molality?

DR. CONTI: It's with an "r," it should be with an "r."

DR. KASLIWAL: Okay, whatever is in the package insert. Usually, we have osmolality.

DR. CONTI: It is not really practical to measure osmolality in certain circumstances, so I would suggest you stick with osmolarity, and change the package insert.

MR. MOCK: Do we need to test for the amount of water in the dose? That's an ingredient. If we have to test for glucose, why not water?

DR. KASLIWAL: No, you test for active ingredients or any functional inactive ingredients.

DR. CONTI: Back on page 10, the Appearance, the Procedure, validation, I am wondering what the validation is. It says, "Visual observation under adequate light."

DR. KASLIWAL: What is your question, are you asking what would be the validation for that?


DR. KASLIWAL: If you look at what we have asked that you submit data, validation data, show suitability--we haven't asked for that validation data, if you look underneath the analytical procedures.

MR. SWANSON: Under Residual Solvents, the not more than limits are percents. I think percents are independent of total volume.

MS. AXELRAD: Could you give us a page? It is really hard to follow.

MR. SWANSON: I am sorry, page 11, the bottom of the page, Residual Solvents, we have not more than 0.04 percent, 0.5 percent, but we have per volume, and a percent is independent of a total volume measurement, so it doesn't make sense to have per volume there.

DR. KASLIWAL: We will correct it to reflect that.

DR. CALLAHAN: I have a point on the radionuclidic purity on page 11. You talk about gamma spectroscopy of a decayed sample. If it were a completely decayed, would you really have a 511 photon? There is a specification in the USP about radionuclidic purity, which suggests that you do a sample and look at the spectrum, but it doesn't say decayed sample. I know why you do a decayed sample, to look for the long-lived, very low level trace materials, but if you need to look for those, you have no positron emitters left there, so I don't understand this.

I can see doing a radionuclidic purity on an active sample and making sure there is not significant amounts of something else there, but if you are going to do it on decay, then, the acceptance criteria should be that there is nothing there.

DR. KASLIWAL: I think you are right. The real intent is to decay the sample and see what you have got.

MR. CLANTON: Jeff Clanton, Vanderbilt.

Would it be fair to say that the test for 2-chloro-2-deoxy-glucose could be replaced if you are doing base hydrolysis with a test for the mannose derivative?

DR. KASLIWAL: Right. If you look at the top, if there is no possibility that your method is going to provide an impurity, you don't have to test for it. If you are using a procedure where you cannot form an impurity, for example, if you don't use solvent, you don't have to have a specification for that.

DR. CONTI: Back to radionuclide identify, bottom of page 10, we discussed this last night. We need to make sure that this is alignment with USP because according to my calculations, it is not possible to measure in 10 minutes with 3 percent accuracy this half-life. So, I think we had some other numbers to take a look at.

DR. KASLIWAL: We will make that consistent with USP, no problems.

MR. CHALY: Thomas Shaly from North Shore University Hospital.

I would like to know whether these tests have to be done on each sample or these are validation testings.

DR. KASLIWAL: I think if you look at the testing schedule, it says that.

MR. CHALY: You will appreciate if you don't include that osmolarity testing and the radionuclide testing on a routine basis.

MR. WATKINS: I would like to just return to this osmolality. As far as I know, in the most recent USP there is no mention of measuring osmolality. Is this going to be a separate issue?

There is no requirement as far as I know in the current USP to measure osmolality, and why are we asking to have it here.

DR. BARRIO: It's calculated, I guess. I would have the same question, yes.

DR. CALLAHAN: We just removed the requirement, the word isotonic solution in the description of the drug product in the monograph, so it no longer is defined as an isotonic solution.

MR. SWANSON: A point of clarification. The USP PET compounding guidelines do require you to calculate an osmolarity as part of your initial validation procedures under product, but there is no requirement for you to routinely test for that, and there is no requirement for the product to be isotonic.

MR. WATKINS: Thank you.

MR. SWANSON: Ravi, a little semantics. On page 3, for example, you have under Name of Target Material, 18 Water, you have test and acceptance criteria. To me, that implies that you have acceptance criteria, and then what tests are you going to perform to ensure the acceptance criteria is met for that component.

That is not really what we are saying we are going to do because basically, you may not have to do all those tests, okay, so I think test is probably an inappropriate word and it probably ought to be something like characteristic and acceptance criteria is what you are saying, because appearance is not a test.

DR. KASLIWAL: Usually, it's test procedure and acceptance criteria. It's not a procedure test identity, and this is the criteria for identity. You have to establish that, and the information for that, you can get it out of COA and make sure that COA data is consistent with your established criteria.

Underneath that section, there is then a section that says, okay, that's fine, but then exactly what would you do to release the product for use. So, in that, whatever it is that you are doing, you need to describe, identity test performed to release each lot for production use. This is for 18 water.

MR. SWANSON: I think you still miss my point. I don't want somebody that fills this out for a component to think that for each of the stated acceptance criteria, they have to do a test. You know, your test means what is the test you are going to do to evaluate that specific acceptance criteria. Okay.

So, I think there is just probably a better word to use than test there. Okay.


MS. AXELRAD: Is that it for this one? We have an option. We could either go through specific comments on the other two, or you could just submit them in writing. If you feel there are things that need to be discussed, we can talk about them, but if you--

DR. BARRIO: We probably could go quickly through them, if you don't mind.


MR. KUHS: Before we leave that, on page 8 of the draft procedure, there is operating parameters under high-pressure targets, it is under Operating Parameters, and you have a number of different parameters of the targets that you are using, and those certainly need to be ranges, and the operating pressure often changes during the irradiation cycle, so I don't think that you can say that there is one specific pressure.

I am not sure why you even distinguish between a high-pressure and a low-pressure target, and without any definitions of what high pressure or low pressure are outside of they are called that, it is really meaningless information.

DR. CONTI: I would also like to consider maybe going more to manufacturer specifications and operating of the devices as opposed to this type of setup.

DR. KASLIWAL: That's fine. What you can then do is just state the manufacturer's specifications in here, what those are.

MS. AXELRAD: Which one are you going to do next?

DR. BARRIO: Sodium fluoride.

DR. CONTI: I actually had a question on this in terms of the issue regarding the source of the F 18 fluoride. There is an inconsistency between the FDG documentation and the sodium fluoride documentation.

In the fluoride documentation, it requires us to list a drug master file for receiving it from another entity, but it doesn't give us the options of if we receive it from a facility that does not have a DMF to go ahead and do the testing for acceptance of the material, just like you would with the FDG.

There is a page here on the FDG, for example, on page 4, that says, "If yes, provide the following information for the supplier," but that is not an option under the sodium fluoride package.

DR. KASLIWAL: I guess the difference is in FDG, fluoride is a reagent. Here, it is the drug substance itself, and that is a very, very critical difference, and you need to have that additional information in this case. I mean you are buying a drug substance from somebody, and the level of information in that case is more than if you are buying a reagent that you are going to use in a synthesis to make something else.

DR. CONTI: I understand that, so you are eliminating the possibility of us doing an acceptance criteria for using this in patients by requiring us to only get it from a drug master file provided facility.

DR. KASLIWAL: I suppose you can provide the information that is listed there right in this application, but you have to be aware that--the reason we have drug master file is because the drug master file holder has the agreement.

They sign a certification that if they change anything, they will notify you that they changed, which then, in turn, you can notify us. They would also notify the drug master file, so if there is any change, it is to protect you. They don't do it without telling you as the user of the product.

If you provide the information in your NDA, I suppose you can have an agreement with them.

MS. AXELRAD: You don't have to get it from a drug master file holder, you can supply it yourself, but then you are held accountable if there is any changes that the supplier makes and they don't tell you about it, you are going to be held responsible for knowing about those things, I mean if you provide it directly yourself.

You always have that option. A drug master file is simply a mechanism that the agency has, so that the supplier can keep the information confidential. You can always just get it, they can always give it to you, and then you can incorporate it in your application, and then you can just have an agreement with them that they will tell you whenever they change anything, and then you can submit a supplement to your application to take that into account.

MR. KUHS: Couldn't that information be just delivered with the--we are thinking of, in this case, an occasional use of fluoride from someone else, where they just gave you the parameters that they operated under that particular day, and oftentimes that is available on a printout. They can also give you the parameters under which it was made.

There is also one other assumption that the F 18 is going to be delivered in a solution, and probably that is not true. Most of the time it would be delivered as an ion and an ion column, an ion exchange column.

DR. KASLIWAL: Thank you.

MS. AXELRAD: Ravi, I will have to think about those things and figure out how to incorporate it.

Someone from the audience, go ahead.

MR. MATTMULLER: Hi. I am Steve Mattmuller from Kettering Medical Center in Kettering, Ohio, not to be confused with our small cousins in New York, Sloan Kettering.

In FDA, Section 6, FDG, Manufacture of Drug Product, B. Reprocessing of PET Drug Product, I was curious if we could get some additional information on this and also make sure I am on the right page with you all on this.

I am thinking if, for example, the bubble test fails, that we could reprocess the solutions for a new filter, the new bubble test passes, then, we are okay. Is that what you had in mind for something like that as far as reprocessing?

DR. KASLIWAL: Which drug are you looking at?


DR. KASLIWAL: And you said what page?

MR. MATTMULLER: I downloaded it from the web. It's page 9 on mine. It might be page 10 of yours, I believe.

Section 6. Manufacture of Drug Product. Part B, Reprocessing of PET Drug Product.

DR. KASLIWAL: Right. That will be one scenario where you can filter it and be able to use it, but you need to state that under this condition you will do that.

MR. MATTMULLER: Thank you. I am curious. Do you have any other potential reprocessing steps that might be acceptable?

MS. AXELRAD: We welcome you to suggest some to us that we can look at.

MR. MOCK: There are a number of examples of additional reprocessing other than sterility. If the fluoride level is too high, run it through another silica cartridge or a luminar cartridge.

If the intermediate--you question whether that worked--another C 18. The pH wasn't quite right, you know, there is the number of things that can be done, so I don't think you are restricting it to any one particular type of reprocessing. I hope that was not the intent.

MS. AXELRAD: No, it isn't. You can write in here whatever things you--circumstances under which you might want to reprocess, and we will look at it when we look at the application.

DR. KASLIWAL: I think the only issue we may have with the reprocessing, what you just mentioned, yes, that will get rid of fluoride, but so would probably chloride and other ions, and you would probably change the whole osmolality of the solution.

DR. BARRIO: Not very much.


DR. BARRIO: Going back to F 18 fluoride, page 7, something very trivial, I guess, under 6A. You mean for each batch of fluoride F 18 injection, right, not FDG?

DR. KASLIWAL: You are right.

DR. BARRIO: Do you guys have any other comment?

MS. KEPPLER: I think it was just the same comments about osmolarity, as well as the radionuclide identity test being in conformance with the USP that we also picked up in this.

MR. SWANSON: In other words, some of the comments we made under FDG would generalize to all of these, and you just need to take a look at those.

DR. BARRIO: Can we go to ammonia then? On page 11, under Radionuclidic Identity, we say yes/yes. It should be yes/no, I believe, because in F 18 fluoride we have yes/no. It should be the same, I believe.

MR. SWANSON: The same thing on that page for osmolality, it would be a calculate. You basically need to go back and make the tables standard.

DR. BARRIO: We discussed yesterday the issue of--let's go to page 8. A specific activity, I think it should not be determined if we are using this procedure--because it is similar to the others.

DR. KASLIWAL: In reading the literature, my understanding you can form actually some ammonia during the radiation, don't you? I mean that is some of the literature, some of them do indicate you actually can.

DR. BARRIO: Are you saying that we are forming--

DR. KASLIWAL: I don't know, that is what the literature says.

DR. BARRIO: Ammonia, mass amount of ammonia?

DR. KASLIWAL: Mass amounts of ammonia. I think there is no way you could do that. There is a possibility of--how can you form ammonia--I don't think there is any way during bombardment you can form ammonia unless you have nitrites and nitrates already in your water, and then during the bombardment conditions and the alcohol present, you can form massive amount of ammonia, but I don't see this as being--you have to remember that we have large amounts of ammonia now in circulation. This is like the glucose issue, it really doesn't matter.

DR. KASLIWAL: I remember reading a procedure that they seemed to state that you could actually form ammonia.

I think it was a no carrier added where you can't form, we will probably accept your no carrier added statement there, but if there is a possibility of forming, then, we are going to have to stick with something like that.

DR. BARRIO: Certainly, this is a problem with carbon 11, let's say, CO2 in which you contaminate or you could contaminate your sample with CO2 from the atmosphere or whatever, but I don't think that is the case here.

Now, the other issue we discussed yesterday, of course, is in the US monograph, is the limitation for nitrites and nitrates to be 2 percent each. The bottom line is that this probably is relevant where this 2 percent each or 4 percent, 1 and 0 the other, and things like that, but we became a little concerned because the array of chemical impurities stated are 94 percent, and we can reject a batch simply because he has more than 2 percent nitrites or nitrates. This is clearly an inconsequential issue.

One thing we could do is to discuss tomorrow under the USP, I mean during that meeting, but I don't know if you guys have some comments on that, but I think this is not a very important issue.

MR. CHALY: I am Thomas Chaly from North Shore University Hospital.

We have been using N 13 ammonia using the [Dewaters elismotad] for the last 10, 15 years. We haven't seen any great amount of nitrite or nitrate in our product.

DR. BARRIO: But in the alcohol procedure--

MR. CHALY: We haven't done the alcohol procedure.

DR. BARRIO: Right. But in the alcohol procedure, it depends upon how you do it. You can see a little amount of nitrite and nitrates. Then, some centers will pass ammonia through a column to remove the anions and leave the cations like ammonia go through. That procedure will be mostly affected by this.

MR. CHALY: In the Dewaters process, we are distilling it out completely.

DR. BARRIO: That's right.

MR. CHALY: So, distilling it out, so we are not contaminating--

DR. BARRIO: But with the alcohol procedure, that is a problem.

If you guys don't have any comments, we are very much done here with this except what Dennis has said just to make sure that it is consistent with the others, and thanks very much.

MR. FERRIS: Is the comment period for this document October 13th, as well?

MS. AXELRAD: Yes, I think that we will say the comment period for all of these are October 13th. Furthermore, on that one, which I didn't really get a chance to look at, it needs to be somehow merged perhaps with the chemistry section. I mean it asks, for example, for the name of the manufacturer, and so does the chemistry section. So, we will square those and try and make sure that all the pieces of this sort of fit together, and you don't have to have redundant information in different sections of the application.

MR. FERRIS: Thank you.

MS. AXELRAD: Let's move on to Clinical Pharmacology/Biopharmaceutics. We just want to briefly alert you to the fact that there is this requirement. We think that it will be fairly easy to deal with in the applications for FDG, ammonia, sodium fluoride. We want to tell you what the requirement is and how we are going to be approaching it.

Clinical Pharmacology/Biopharmaceutics

MR. HUNT: I am John Hunt. I am from the Office of Clinical Pharmacology and Biopharmaceutics. As Jane has indicated, I am going to talk on the area of clinical pharmacology and biopharmaceutics particularly related to the regulatory umbrella we are working under and as related to what kinds of information needs to be provided in an NDA or an ANDA.

I have a four-page handout that I will talk through. On the second page, I have highlighted the section of the Code of Federal Regulations, particularly Part 320, that addresses the bioavailability and bioequivalence requirements.

Under that part there is a Section 320.21 that states that when a sponsor submits an NDA or an ANDA, they either need to provide in vivo bioavailability data, that relates to the NDA, or bioequivalence data, which relates to an ANDA. Alternatively, you can submit information to allow a waiver for not meeting in vivo bioavailability or bioequivalence information.

This morning we had a lot of discussion on definitions, so I included one here to focus on the term bioavailability. As stated in the regulations, it states that bioavailability means the rate and extent to which the active ingredient or active moiety is absorbed from a drug product and becomes available at the site of action.

Now, for I.V. products, historically, the agency has assumed that an I.V. product is 100 percent absorbed, so although the definition addresses the percent absorbed concept, we assume that an I.V. is 100 percent absorbed.

Also, in this section of 320, there is a section related to waivers, and there is two scenarios. Particularly for these kinds of products, which are parenteral products, there is one section that says for the drug product, (1) is a parenteral solution intended solely for administration by injection.

The second component of that requirement for a waiver is that it contains the same active and inactive ingredients in the same concentration as a drug product that is the subject of an approved full new drug application. That is the listed reference drug.

There is also another component of the waiver criteria, which is general, for good cause, for the public health. Historically, the agency hasn't really used that when there is the ability to measure something.

On a few occasions, it has been used when there is a critical need to get a product on the market and there is not technological methods available to quantitate a drug in terms of in vivo performance.

In the area of sodium fluoride F 18, there is an approved product, and as Jane indicated, I am guessing that there is going to be procedures worked out where the information about the approved product would be made available.

Since there is only one synthesis procedure method for this agent, and lastly, if there are CMC limits that are set, if a sponsor can meet those three criteria, that is, provide information on the ingredients of their product related to the reference listed drug, that it is identical, again, it's the same synthesis process, and it meets the CMC specs, then, all that would be needed is citing this section of the regulations and getting a waiver for it. You would not need to do in vivo kinds of studies.

DR. CONTI: The synthesis process for sodium fluoride, the reactor produced versus the cyclotron produced, you are calling equivalent?

DR. KASLIWAL: Where is the reactor produced product?

DR. CONTI: Reactor produced sodium fluoride is what was used years ago, as well as cyclotron produced materials.

DR. KASLIWAL: I think the package inserts say that it is a cyclotron produced product.

DR. HOUN: It does say this solution contains no carrier added fluorine 18 as the fluoride ion in isotonic sodium chloride solution. The 18F is produced by bombardment of neon gas accelerated in a cyclotron.

DR. CONTI: On your documents here, as I recall, it said something about either FDG or fluoride, it says something about reactor produced material. Yes, on page 4 of the FDG, Chemistry, Manufacturing, and Controls document, Item 3 in the lower large box, it talks about reactor produced fluoride.

DR. KASLIWAL: You are looking at FDG.


DR. KASLIWAL: That is if somebody wants to buy fluoride from a reactor produced product, then, we are going to have additional issues with it, specifically, some radionuclidic impurities.

MS. AXELRAD: So, we are just asking them to let us know if it is reactor produced.

DR. CONTI: But in this document here, I want to make sure I understand, the only way that you can get equivalence is that the material is being compared to the original cyclotron produced material. Okay.

MR. FERRIS: You are saying the same thing is true with spectral lineac, as well?

DR. LEUTZINGER: I don't know about the lineac. We would have to look into that, but I presume that there wouldn't be any--

MR. FERRIS: They are combined in this statement of reactor or lineac.

DR. LEUTZINGER: We would have to look into the lineac, but I would presume--what I know about the lineac, there wouldn't be any reason to suspect anything different than you would in the cyclotron at this point.

DR. CALLAHAN: When you read the package insert there you stated that it was the deuteron on neon reaction, right? That is not the reaction that anyone would use today to make fluoride, I don't believe. There goes the waiver process.

MS. AXELRAD: Basically, I think that what we are trying to say is that it is very easy if you could show sameness. If you can't show sameness, we are going to have additional issues with it.

Can you explain that, John?

MR. HUNT: We have had a lot of discussion internally that if it isn't identical and what would we consider, and going again back to the regulations, if it doesn't fall under a waiver, then, we need some kind of a study, and one thought is a dosimetry type study, but we are certainly open to any thoughts where you think the appropriateness might be related to that if it can't fall in the window of being a same product as a reference listed drug.

MR. BRESLOW: Ken Breslow, PETNet.

Regarding the criteria under (b)(2), that it contains the same active ingredients in the same concentration as the reference drug, with FDG that is going to be an issue, because the reference product currently is a relatively low specific concentration range, which is probably lower than most distributors would require to produce.

Here again, we must keep in mind that we are talking about a radioactivity concentration range where the actual physical amount of the drug, which is carrying three levels and very small, minute physical amounts of FDG, isn't really different. It is the amount of radioactivity as the strength, as it is defined, and so it is inappropriate to set that standard in dealing with the definition of strength with a PET radiopharmaceutical especially FDG.

It might be appropriate when there is a PET pharmaceutical that has pharmacological impact or potential to elicit pharmacological response, but not with FDG.

MR. KUHS: It is only the radioactivity that changes in the strength. The molecular composition stays the same. The concentration, by definition, is the amount of radioactivity per unit volume, and when typically, you are talking about concentration, you are talking about different molecular concentrations. That stays the same. It's the radioactivity that changes.

So, a specific concentration shouldn't be an issue in determining bioequivalence or bioavailability.

MS. AXELRAD: I think what we are trying to do here is explain that we have run across, you know, we have been looking at all the different requirements that are in our existing regulations and trying to figure out how they would apply. We are sort of trying to give you an overview of what we see and really to identify the problems we recognize that there are issues associated with this.

So, basically, let's go through what the requirements are and then we can get some comments on what problems those might pose that we can look at.

MR. HUNT: Continuing on to the ammonia N 13, there is no approved product at this time, however, the agency has, as you are aware, gone through and looked at the literature. There has been a review that was prepared that addresses this, and the thinking is that the information that is available via this review process that has gone on internally would be the basis a firm or sponsor could cite from the Federal Register notice where this will, in the future, be referenced.

So, that would meet your in vivo requirement. It would be based upon information that has been already reviewed in the agency and found to be acceptable.

Once the first NDA is approved, then, that puts us into the mode where you can get approval of the ANDA, again showing you have got an approved product, all you have to show is that your product is similar or identical to the reference listed drug.

Again we are dealing with a one-synthesis process and again you have to meet the CMC limits that would be set based on the reference listed drug. So, again, you can get away with a waiver and not needing to do an in vivo type study, and even in the first case, the information that is in the literature has been found to be acceptable to satisfy that need.

Lastly, is the FDG. I have two scenarios here, but it sounds like the last one is not really relevant because it doesn't appear that the electrophilic procedure is being used here in the U.S., so again, that falls into the former scenario of a waiver, again, that, in fact, a reference listed drug can be established and which is available, and that can information can be disseminated, and then you just have to show that you meet the CMC specs and you are using the same synthesis procedure.

DR. BARRIO: Let me ask you a question about sodium fluoride. Let's go back to this. The original synthesis or rather nuclear reagent being used when large cyclotrons were available, there was, of course, the deuteron neon reaction.

The one that we normally use right now with the smaller cyclotrons being used is the O 18, maybe O 18, mostly O 18 water. What you are trying to say is that you like to demonstrate that you have the two-synthesis procedures, the one that was done before, the deuteron, and the one that is done right now, that the product has the same biological properties.

MR. HUNT: I hope I am correct on this. If you can show that what is made, and if they fall within the CMC limits that are set by the agency, you can meet those, then, there is probably not a problem unless there is another impurity or something that is formed that we would not expect to see.

DR. KASLIWAL: I think if both the methods are no carrier methods, then, they are pretty much deemed to be the same.

DR. HOUN: I think why we brought this up is just to say that in terms of biopharmaceutic requirements for FDG, ammonia, and sodium fluoride, we are going to be handling it this way, but certainly if new PET drugs are developed, then, the bioavailability, bioequivalence issues come into play, and we would just remind the community that these are other requirements.

MR. SWANSON: I am sorry, I didn't understand that. Doesn't bioequivalency come into play when you submit an ANDA? So, don't these waivers have to apply to equivalency to whatever we grant 505(b) status to or whatever currently has an NDA, is that not correct?

MS. AXELRAD: Bioavailability requirements apply to NDAs; bioequivalence requirements apply to ANDAs, Abbreviated New Drug Applications. Basically, if somebody comes in with a new NDA, maybe not based on the literature, but based on regular clinical studies, there would be certain bioavailability requirements, and then is subsequently, a generic comes in to the reference listed drug, then, it would come under--usually, it would get a waiver if it's a parenteral kind of a product, but we want you to be aware of what the regulations are.

MR. SWANSON: But then it seems like we have got certain problems now because, as was pointed out, our current NDA FDG probably doesn't represent what a lot of the syntheses are going to be, the same concentration, et cetera.

Our current sodium fluoride approval doesn't represent what is being done out there, and we don't have one, so it seems to me like we are going to end up having to submit--are we going to have to submit multiple NDAs at different concentrations in order to make this work?

What if my PET center is making it at 40 millicuries per milliliter, and they are making it at 100 millicuries per milliliter, and if I want to go the ANDA route, then, I am going to have to tie together with an NDA that is doing that 40 millicuries per milliliter, right?

DR. KASLIWAL: One is the strength, has to be within the strength range, so that is one aspect for ANDA. The other is the composition, if your composition changes, then under certain circumstances, some things are allowed in ANDA, other things are not allowed in ANDA.

MS. AXELRAD: But that is why the (b)(2) route, I think is available. If you can't show that you are the same as a reference listed drug, then, you can't come in as a generic, but you can come in as a (b)(2).

In this case where all the clinical safety and efficacy data is based on the literature anyway for the three drugs that we are talking about here, there is not a huge difference between a (b)(2) and a (j). It just doesn't really matter that much.

You just have to be aware that if you are going for a (j) and you are going for the sameness, you know, trying to show sameness, then, you have to be aware of the criteria for sameness and see if you are the same. Otherwise, you come in as a (b)(2), and would address the differences.

DR. KASLIWAL: I also just want to clarify when I say composition, we are not talking about impurities that are present. Impurities, you can control. Composition is the active and the inactive ingredients.

MS. AXELRAD: You can sort of mull over, and I would like to move on and mention another little issue, pediatrics. Dr. Love is going to tell you about the requirements. You are probably aware of the Pediatric Rule, or may or may not be aware. We published a final rule on pediatrics. It deals with pediatric studies for drugs, and requires new applicants to address that in a certain way. Dr. Love is going to summarize that and tell you again how we are going to try and deal with in the PET context.

Pediatric Rule

DR. LOVE: The agency has been concerned about the need for pediatric labeling, as Jane was just saying, and in December 1998, there were regulations published under 314.55 that talk about required pediatric studies for new drugs, new indications, new dosage forms, and the like, and it specifically related to the drug and the indication particularly.

That regulation identifies the fact that there are methods for dealing with this, there can be deferrals, again waivers, full or partial waivers. Those can be initiated either at the request of the sponsor or on the agency's own initiative, and we would look at such things as whether or not the indication is relevant to pediatrics, whether or not the number of pediatric patients that might receive a particular drug for a particular indication is appropriate, the safety of the product, and whether or not it is practical or impractical to do pediatric studies.

Also, in that FR notice, there is a list of indications or diseases for which the agency is expected to or apt to provide a waiver.

So, what we have done is look at these particular drugs that we are considering for this FR notice - FDG, sodium fluoride, as have been mentioned, and ammonia, and looked at the indications that were discussed at the MIDAG meeting and considered where these drugs and indications would fall in this format.

One of the listed indications in the preamble to the FR notice is atherosclerosis. So, we have looked at the fact that the FDG indication for hibernating myocardium and the indication for ammonia for myocardial perfusion are certainly associated with that, and we feel that we would be able to waive any requirement for those two drugs.

We are taking that information to our Pediatric Committee in the agency that looks at all of this, and will be discussing it with them in two weeks, but that is our expectation at this point in time.

As far as FDG epilepsy, that is already labeled for pediatric use, so that is not a concern.

The other drugs and indications for which we are seeking some information at this point happen to be the FDG for oncology, we certainly expect that it would be used in pediatrics, and the FDG for bone imaging, again, that would be used in pediatrics--I am sorry, sodium fluoride for bone imaging.

What we are doing at this point is seeking information from Oak Ridge and also contacting NIH looking for dosimetry information on these uses, and considering just putting that information in the labeling and trying to address it from that standpoint.

Where we are running into a little bit of a challenge is finding information on sodium fluoride since that is an old product, and we are looking for information on sodium fluoride in the pediatric population. We are still seeking it from the two sources that I have identified, but also if you have some other information that you could provide to us, that would be helpful. I would also like to get your comments on whether or not you think dosimetry information would even be the appropriate way to go in trying to finalize the labeling for the pediatric population for these two drugs and indications.

DR. BARRIO: But when you are looking for dosimetry information, you mean in children or in adults?

DR. LOVE: Pediatric population specifically, which in the regulations is defined as under 16. Normally, what we do when we are looking at a pediatric population is think more specifically about which pediatric age group is apt to function like the adult population and where you might see differences, so for bone imaging, it would be issues where there is epiphyseal closure has not occurred or where there is a rapid growth spurt or something, and what is happening in that population.

If it's FDG, it may be an issue of whether or not certain tumors may metabolize the product in a different way in a pediatric population, something that is more specific to pediatrics, or perhaps where the metabolic process on the basis of age may be different.

If those things are not issues, then, we wouldn't worry about them, but that is the general approach we take to thinking through the issue, but specifically, we are looking at dosimetry, probably in a younger age population or smaller body surface size population, not so much the 16-year-olds that are comparable to adults.

MR. SWANSON: Could you summarize for us what the pediatric regulations say, do they actually mandate that industry must do pediatric studies? I thought that there was a series of incentives associated with it.

DR. LOVE: What you are talking about there is the exclusivity process. The Pediatric Rule itself does require pediatric studies for new drugs, new indications, new dosage forms, and the like, and it says the information is required, but the manufacturers, the sponsors can identify situations in which it may not be relevant, and that is when the waivers come into play.

So, as I was mentioning earlier, if an indication is not relevant, if it is a very small population, an orphan indication, that sort of thing, where it is either not wise, unsafe, or impractical to study, but we would need information that showed that it is impractical.

What we are trying to do is do this across the board and address this up-front. This is not going to be an issue for each individual site to address. This is something that we would like to take care of in the FR notice on safety and effectiveness of these particular drugs and indications.

Our goal here is to also along with all the other notices that would be coming out is to actually publish the labeling, and the labeling would already contain the statement for pediatrics, so this would be done ahead of time.

DR. HUNG: Dr. Love, did I hear you say that you will use body surface to adjust a dosage for pediatric patients?

DR. LOVE: No. I am saying that those would be the kinds of things we would think of in general for pediatrics when we are looking at it, not specifically for--

DR. HUNG: So, you don't have any specific method for adjusting the dosage?

DR. LOVE: There are a number of different approaches and algorithms that can be used to adjust dosing. I think again we look at the specific drug and its mechanism of action and what is taking place. So, that was more of a general comment.

DR. CONTI: I think I said this last time, that dosimetry I think is the key issue. There are pages of ways to calculate this for pediatrics, and it is done routinely for radiopharmaceuticals, such as a technetium bone scan. It is very traditional to adjust the pediatric doses. There are standard ways of doing that.

I would also note to you that since you have already put FDG, according to your label, in children with epilepsy, that since the injected dose is identical whether you are doing an oncology study, epilepsy, or a heart study, that you can just dismiss that now as equivalent if you are using dosimetry per se as the criteria. So, it is done.

DR. HOUN: The label doesn't have, I guess, the information on how much radiation is received to the critical organ for kids. Is that something the label should have or not?

DR. CONTI: It is going to be the same thing as in adults. The biodistribution is essentially identical in an adult. The only difference is going to be the total amount of activity that you are injecting and what that activity is going to expose the critical organs to, which in most cases is the bladder for most of these drugs.

DR. LOVE: And certainly we thought about that particularly for FDG and wondered what is the relationship. One of the key issues in the Pediatric Rule is the indication, so the issue there is not just the safety in terms of the elimination through the bladder, but also are we going to get a different biodistribution pattern on the basis of disease.

If what you are talking about is true, then, what we simply need are some data to try to support it from administrative record perspective, but I understand what you are saying.

DR. CONTI: I think what I am telling you is I don't think you are going to get it because the only alteration in distribution is going to be in a child with a cancer, and you are going to see uptake in that cancer as opposed to it not being in the cancer.

It is essentially the uptake in the heart, the uptake in the brain, and the normal organs are going to be identical across the board, and I would venture to say that there is probably an insignificant change in dosimetry irrespective of whether they have a cancer or not to the normal organs.

DR. LOVE: What about sodium fluoride in pediatrics--

DR. CONTI: Again, I mean you can use the technetium bone scanning as a means of calculating the same biodistribution and adjusting it exactly the same way as an MDP dose is adjusted, because again it is just a matter of the total activity that is going to change, not the biodistribution in an adult versus a child.

I mean you are going to see more uptake in the epiphyses just as you would see in a technetium bone scan. So, again, it is the same issue, it is just a matter of adjusting the dose according to the standard calculations.

The only thing I would add to this is that you are not looking at ammonia for the pediatrics, is that--

DR. LOVE: We were considering waiving it because of the indication, atherosclerosis. We actually had data presented at the MIDAG on dosimetry that went down to the age of the ones we actually have some information on.

DR. CONTI: There is little reason to do it for atherosclerosis in children.

DR. LOVE: That is the basis of the waiver.

DR. CONTI: But there are indications, though, in children that would require that you use this drug.

DR. HOUN: Not the one that was reviewed in the literature. If you want to come with a supplement to that--

MS. AXELRAD: You might have to do a study.

DR. CONTI: For coronary artery disease?

DR. HOUN: Everything was looked in, in people with angiography as a gold standard with coronary artery disease.

DR. CONTI: What about Kawasaki's disease, for example?

DR. HOUN: Not one article had that, not the ones that were reviewed, that met the standard for prospective.

DR. CONTI: Okay.

MS. AXELRAD: If you want to come in and add that indication to the label, you might have to do a study in kids. That is basically what is going to happen.

DR. CONTI: Okay.

User Fees

MS. AXELRAD: The last issue on the agenda for today is user fees. Again, what we want to do is try and address this in an overall fashion in a Federal Register notice. We are examining the possibility of giving a waiver of the application user fees, which I think are going to be the biggest issue. It is under the barrier to innovation provisions under the user fee law that provides that we can give a waiver for anything that is a barrier to innovation based on insufficient resources or other circumstances.

What we are looking at is whether the provisions of FDAMA that tell us to specially regulate PET facilities and to deal with them in a special way, and the equities of the situation in that whoever happens to come in first would happen to pay an application fee, but once the first (b)(2) application is approved, nobody else would have to pay.

It seems sort of unfair that it is just an accident of whoever steps forward and wants to be first, so it is sort of based on that sort of combination of circumstances, as well as the fact that there isn't going to be any clinical safety and efficacy data in the applications, it is all based on the literature, literature that we ourselves have already reviewed.

We are going to try and see if it would justify a blanket waiver of application fees, and we would try and address that in the Federal Register notice. Once two products are approved, no one pays any--well, ANDAs don't pay any fees anyway, they don't application fees or any other kind of user fee, and really application fees is what the issue is here.

So, I think we can solve that problem if we can determine how to justify that. Again, we think this is a very unique situation, and we are basing it on the sort of uniqueness of this.

That is hopefully where we are going to go on that. Again, it will be addressing it in there. We will let you know if there is some change in that.

DR. HUNG: Could I make a comment? This is Joe Hung from Mayo Clinic.

I know that in the past, collecting the user fee has been very successful in cutting down the review process for the new drug application. By not collecting this user fee for the PET new drug application, can we anticipate that this will not affect the review process, and how are you going to cope with the inspection process without the additional fund from this user fee, collecting from the users?

MS. AXELRAD: Well, we would love to have somebody voluntarily pay it. We would really like to have the first application and each of these areas pay the fee. You know, we spend quite an amount of our agency resources on this whole project, but we can say that if you do get a waiver, you get the same review, and you are subject to the same standards in terms of timeliness of review whether the fee is waived or the fee isn't waived basically.

I think that covers all the issues that we have. I think we have covered an awful lot of ground today, sort of several steps forward and much more to go, it seems like, but I guess we will go back and try and absorb everything that we got today in terms of feedback.

We will look forward to getting your written comments on all these documents by the 13th, and we will try and revise the documents and see where we go next, after we have had a chance to go over the record.

Thank you very much.

[Whereupon, at 5:50 p.m., the meeting was adjourned.]