FOOD AND DRUG ADMINISTRATION

 

ORALLY INHALED AND NASAL DRUG PRODUCTS SUBCOMMITTEE

OF THE

ADVISORY COMMITTEE FOR PHARMACEUTICAL SCIENCE

 

 

 

 

 

 

 

 

 

 

 

 

8:34 a.m.

Tuesday, July 17, 2001

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CDER Advisory Committee Conference Room

5630 Fishers Lane

Food and Drug Administration

Rockville, Maryland 20857

ATTENDEES

COMMITTEE MEMBERS:

VINCENT H.L. LEE, PH.D., Acting Chair

Department of Pharmaceutical Sciences

School of Pharmacy

University of Southern California

1985 Zonal Avenue

Los Angeles, California 90033

NANCY CHAMBERLIN, PHARM.D., Executive Secretary

Advisors and Consultants Staff

Center for Drug Evaluation and Research

Food and Drug Administration (HFD-21)

5600 Fishers Lane

Rockville, Maryland 20857

GLORIA L. ANDERSON, PH.D., Consumer Representative

Fuller F. Callaway Professor of Chemistry

Morris Brown College

643 Martin Luther King Jr. Drive, N.W.

Atlanta, Georgia 30314-4140

 

SGE CONSULTANTS:

RICHARD C. AHRENS, M.D.

Associate Professor, Department of Pediatrics

The University of Iowa Hospitals and Clinics

200 Hawkings Drive

Iowa City, Iowa 52242-1083

MARK S. DYKEWICZ, M.D.

Associate Professor of Internal Medicine

Division of Allergy and Immunology

Department of Internal Medicine

St. Louis University School of Medicine

1402 South Grand Boulevard-Doisy R-209

St. Louis, Missouri 63104-1028

LESLIE S. HENDELES, PHARM.D., F.C.C.P.

Professor of Pharmacy and Pediatrics

University of Florida Health Science Center

1600 S.W. Archer Road, Room MG-57

Gainesville, Florida 32610-0486

ATTENDEES (Continued)

SGE CONSULTANTS: (Continued)

DENNIS R. OWNBY, M.D.

Professor of Pediatrics and Medicine

Head, Allergy-Immunology Section

Medical College of Georgia, Building BG-1019

1446 Harpers Street

Augusta, Georgia 30912-2790

 

INVITED GUEST PARTICIPANT:

WALTER W. HAUCK, PH.D.

Professor of Medicine, Department of Medicine

Jefferson Medical College

Thomas Jefferson University #1170

125 South 9th Street, Suite 402

Philadelphia, Pennsylvania 19107-5244

 

INVITED INDUSTRY PARTICIPANTS:

IZABELLA J. ROMAN, M.D., PH.D.

Target Research Associates, Inc.

656 Ardsley Road

Winnetka, Illinois 60093

LEON SHARGEL, PH.D., R.PH.

Vice President, Biopharmaceutics

Eon Labs Manufacturing, Inc.

227-15 North Conduit Avenue

Laurelton, New York 11413

 

FDA PARTICIPANTS:

WALLACE ADAMS, PH.D.

Pharmacologist

Office of Pharmaceutical Science

BADRUL CHOWDHURY, M.D.

Medical Officer Team Leader

Division of Pulmonary and Allergy Drug Products

DALE P. CONNER, PHARM.D.

Director, Division of Bioequivalence

Office of Generic Drugs

ATTENDEES (Continued)

FDA PARTICIPANTS: (Continued)

ROBERT J. MEYER, M.D.

Division Director

Division of Pulmonary and Allergy Drug Products

GUIRAG POOCHIKIN, PH.D.

GUR JAI PAL SINGH, PH.D.

HELEN N. WINKLE

Acting Director

Office of Pharmaceutical Science

 

ALSO PRESENT:

CYNTHIA L. FLYNN, PH.D.

RWJ-PRI

1660 North Hilltop Road

Spring House, Pennsylvania 19478

PIYUSH PATEL, M.D., F.R.C.P.(C)

Medical Director & CEO

Allied Clinical Research, Inc.

JOEL SEQUEIRA, PH.D.

Senior Associate Director

Shering Plough Research Institute

2000 Galloping Hill Road

Kenilworth, New Jersey 07033

C O N T E N T S

AGENDA ITEM PAGE

CONFLICT OF INTEREST STATEMENT

by Dr. Nancy Chamberlin 6

INTRODUCTION AND OBJECTIVES

by Ms. Helen Winkle 10

BIOEQUIVALENCE CONSIDERATIONS FOR

LOCALLY ACTING NASAL DRUGS

by Dr. Dale Conner 15

THE JUNE 1999 DRAFT NASAL BA/BE GUIDANCE:

HISTORY, RECOMMENDATIONS OF THE DRAFT GUIDANCE,

LOCAL DELIVERY EQUIVALENCE,

CONTRIBUTION OF DOSE-RESPONSE

by Dr. Wallace Adams 35

DIFFICULTIES WITH SHOWING A DOSE- RESPONSE WITH

LOCALLY ACTING NASAL SPRAY AND AEROSOLS FOR

ALLERGIC RHINITIS

by Dr. Badrul Chowdhury 51

CLINICAL STUDY OPTIONS FOR LOCALLY ACTING

NASAL SUSPENSION PRODUCTS:

CLINICAL AND PHARMACODYNAMIC STUDIES

by Dr. Robert J. Meyer 69

SUBCOMMITTEE DISCUSSION 88

OPEN PUBLIC HEARING PRESENTATIONS:

by Dr. Cynthia Flynn 105

by Dr. Joel Sequeira 112

by Dr. Piyush Patel 117

SUBCOMMITTEE DISCUSSION 127

P R O C E E D I N G S

(8:34 a.m.)

DR. LEE: Good morning. I'm Vincent Lee. I'm the chair of the subcommittee, and I welcome you to this subcommittee meeting.

Before we start, I would like to turn it over to Nancy.

DR. CHAMBERLIN: Welcome.

The following announcement addresses the conflict of interest with regard to this meeting and is made a part of the record to preclude even the appearance of such at this meeting.

Since the issues to be discussed by the committee at this meeting will not have a unique impact on any particular firm or product, but rather may have widespread implications with respect to an entire class of products, in accordance with 18 U.S.C. 208(b), all required committee participants have been granted a general matters waiver which permits them to participate in today's discussions.

A copy of these waiver statements may be obtained by submitting a written request to the agency's Freedom of Information Office, room 12A-30, Parklawn Building.

With respect to the FDA's invited guests, Dr. Leon Shargel, Dr. Walter W. Hauck, and Dr. Izabela J. Roman have reported interests which we believe should be made public to allow the participants to objectively evaluate their comments.

Dr. Shargel would like to disclose that he is employed by Eon Labs Manufacturing Company.

Dr. Hauck would like to disclose ownership of stock in Bristol-Myers Squibb. Also, his employer, Thomas Jefferson University, has contracted to perform biostatistical services for various pharmaceutical companies. Additionally, Dr. Hauck has numerous consulting agreements with the pharmaceutical industry related to bioequivalence, bioassay, and content uniformity.

Finally, we would like to disclose that Dr. Roman is medical director and co-founder of Target Research Associates.

In the event that the discussions involve any other products or firms not already on the agenda for which an FDA participant has a financial interest, the participants are aware of the need to exclude themselves from such involvement and their exclusion will be noted for the record.

With respect to all other participants, we ask in the interest of fairness that they address any current or previous financial involvement with any firm whose products they may wish to comment upon.

Thank you.

DR. LEE: Thank you, Nancy.

I think it would be useful to go around the table and everybody introduce herself or himself, and who they're with and why they're here.

Dr. Roman?

DR. ROMAN: Good morning. My name is Izabela Roman. I am a medical director and principal of a CRO, Target Research Associates, and I am here as an invited industry guest.

DR. LEE: Thank you.

DR. HAUCK: I'm Walter Hauck. I'm a professor and head of biostatistics at Thomas Jefferson University.

DR. SHARGEL: Good morning. I'm Leon Shargel, vice president of biopharmaceutics for Eon Labs, a generic manufacturing company, and I'm an invited industry participant.

DR. ADAMS: Good morning. My name's Wallace Adams. I'm in FDA's Office of Pharmaceutical Science in CDER and am involved in the nasal BA/BE guidance drafting that we're discussing this morning.

DR. MEYER: Dr. Bob Meyer. I'm the director of the Division of Pulmonary and Allergy Drug Products in CDER, and I've also been involved with the drafting of the guidance.

DR. CHOWDHURY: I'm Badrul Chowdhury. I'm a medical team leader in the Division of Pulmonary and Allergy Drugs.

DR. CONNER: I'm Dale Conner. I'm director of the Division of Bioequivalence in the Office of Generic Drugs, FDA, and I'm a speaker today.

DR. CHAMBERLIN: I'm Nancy Chamberlin. I'm the exec sec.

DR. LEE: I'm Vincent Lee, the acting chair.

DR. ANDERSON: I'm Gloria Anderson, Callaway Professor of Chemistry at Morris Brown College, Atlanta.

DR. AHRENS: Richard Ahrens. I'm on the faculty of the University of Iowa in the Division of Pediatric Allergy and Pulmonary Disease.

DR. OWNBY: Dennis Ownby. I'm professor of pediatrics and medicine in the Medical College of Georgia and head of the section of allergy and immunology.

DR. DYKEWICZ: Mark Dykewicz. I'm associate professor of internal medicine and director of the training program in allergy and immunology at Saint Louis University School of Medicine.

DR. HENDELES: I'm Leslie Hendeles. I'm a professor of pharmacy and pediatrics at the University of Florida in Gainesville.

DR. LEE: Thank you very much.

Anybody else?

(No response.)

DR. LEE: Well, I guess that concludes the introduction, and I would like to invite Helen Winkle, the acting director of the Office of Pharmaceutical Science, to give us our marching orders.

MS. WINKLE: Good morning, everyone.

Unfortunately for today, my slides got misplaced, but it's understandable. I think many of you will know this is only the first of four days of advisory committee sessions that are going on this week. So, if mine are the only slides that are misplaced for four days, I think we're going to be real fortunate. It's been a lot of work.

I want to start off by welcoming everybody here. I'm really happy that the members of the subcommittee could join us to talk about this very important issue, and I especially want to thank Wally Adams, who put this subcommittee meeting together. He's worked long and hard to make sure that we're focused on the right issues that we need to address here today.

I also want to thank Dr. Meyer and Dr. Chowdhury for joining us from the Pulmonary Division. I think their input will be very helpful.

I wanted to talk a little bit about the background of the subcommittee, just so everyone here is aware of why this subcommittee exists, but before I do that, I just wanted to mention the fact that this is sort of an ad hoc subcommittee arm, basically, of the Advisory Committee for Pharmaceutical Science. In this ad hoc type of situation, we bring in experts in the field of nasal and orally inhaled drugs, depending upon what topics we're going to discuss.

So, today, basically the people that are here as members of this subcommittee or to participate with the subcommittee are people who have clinical backgrounds, since basically we're going to be talking about how to handle clinical issues, basically dose response and the merits of three-study design. So, the people that are here are basically those people who can best give us their expertise in these areas.

This subcommittee actually started as an expert panel back in 1999. It met to discuss issues relating to various guidances having to do with nasal sprays and aerosols and orally inhaled products. In 2000 we decided that probably it would be advantageous for us to start a subcommittee to address these issues, and therefore we formed the Advisory Committee for Pharmaceutical Science's Subcommittee on Orally Inhaled and Nasal Drug Products, or OINDP, which is a lot easier for us to relate to.

The purpose of this subcommittee is basically to address some of the issues that we have to do in order to have good regulatory decisions in these product areas, and these are somewhat complicated in many cases. So, although we're looking at a variety of products in CDER, we have certain products that are more complicated, have a lot more issues that have to be addressed, and this is one of the areas where we feel that we really need some expert advice from outside of the agency to help us in addressing these regulatory issues.

When the panel was formed, under the Advisory Committee Act, in order to form a subcommittee, you have to take two members from the advisory committee, and we asked Dr. Lee and Dr. Anderson if they would join us on this subcommittee, and Dr. Lee agreed to chair it. So, those are the two members from the advisory committee. They report back to the advisory committee on the recommendations of this subcommittee.

The first time this subcommittee met, which was April 26th of 2000, they addressed a variety of issues having to do with the four guidances that we are developing in this area, and these are guidances we've been working on currently. These are on metered dose inhalers, the CMC requirements, and nasal sprays and inhalation solutions, the suspension CMC, and both of those have been issued in draft and we're working on the draft, and as we're developing the draft, of course, we have questions.

We also have two other guidelines on BA and BE. They're BA and BE studies for nasal aerosols and nasal sprays for local action and BA and BE studies for orally inhaled MDIs, DPIs, and inhalation solutions for local action, and that particular guidance is in preparation form.

Today we're only going to be dealing with the draft guidance for nasal BA and BE. We're focused specifically on questions on local delivery for nasal sprays and aerosols, and basically we're looking at this issue in order that we can better determine bioequivalence of the suspension form for nasal aerosols and sprays for allergic rhinitis. One of the requirements is for in vitro studies. However, in doing these in vitro studies, it's very difficult to assure equivalence of the particle size of the suspended drug, and because particle sizes differ between the test and reference product, the potential to alter the rate and extent of delivery of the drugs to the local sites really can have some difference in clinical effectiveness.

So, basically we've put together some questions that we would like to have answered today in order to determine how best to ask for these studies to determine bioequivalence. The specific questions we have -- and Wally will again go over these questions as we move forward in his presentation today -- is, is a placebo-controlled traditional two-week rhinitis study conducted at the lowest active dose sufficient to confirm equivalent local delivery of suspension formulation nasal sprays and nasal aerosols for allergic rhinitis? And the second question we have, is a placebo-controlled park study or an EEU study conducted at the lowest active dose an acceptable option to confirm equivalent local delivery of suspended formulation nasal sprays and nasal aerosols for allergic rhinitis? We feel, with some direction from you on these particular questions, we can then move forward and make recommendations to the advisory committee.

Now, the advisory committee is meeting on July 19th, which is Thursday. It's actually meeting July 19th and 20th. So, the recommendations that come out of this group will go directly to the advisory committee on Thursday, and Wally and Dr. Meyer and Dr. Chowdhury will again present to that advisory committee and will present the recommendations from this.

So, basically, today this subcommittee is here to address the questions related to equivalent local delivery for in vitro BA and BE data and to provide recommendations to that committee.

I want to personally thank all of you for taking the time. Again, I think your expertise is very valuable to us at FDA in helping to address these very important regulatory decisions.

Thank you very much.

DR. LEE: Thank you very much, Helen. You've done extremely well without slides, and I'm beginning to worry about when will those slides turn up.

But let me reiterate two things that Helen said. Number one is that we're here to address two questions. There were specific questions. The rest of the morning will be devoted to the background information to arrive at some kind of recommendations. I'd like to remind the subcommittee that we're not here to vote. We're here to develop a consensus.

Without further ado, I'm going to call on Dr. Conner to tell us something about bioequivalence considerations for locally acting nasal drugs.

DR. CONNER: Good morning. I am leading off on this because I think it was felt that some introduction or setting the stage for a bioequivalence discussion is really kind of necessary in this topic.

Usually when I give this talk, even my simplified explanation of bioequivalence runs at least for an hour to an hour and a half, and that's really cutting it quite short. So, I'm really pressed for time here. Hopefully, I'm going to cut to the chase to a few issues that really need to be considered to set the stage for speaking about bioequivalence, because as you'll see within the next few slides, I believe that sometimes committees like this, and even we within the FDA, become quite confused between strictly bioequivalence questions and perhaps other peripheral or distantly related questions.

First off, there are many definitions of bioequivalence. I've picked only one of them, and as you'll see, I have several kind of definition or explanation slides in the next few, all trying to make the point or drive home what we're talking about here in general terms. First off, the first thing you have to understand is, and certainly when we approve generic drugs and we refer to bioequivalence, we're talking about pharmaceutical equivalents. So, that's the first point that you have to realize, and the point that often is confused by many clinicians in the outside world is when we talk about a generic drug, it's a pharmaceutical equivalent.

What do we mean by that? Pharmaceutical equivalents have the exact same drug substance. So, it isn't a question of therapeutic substitution, where you're substituting yet a totally different drug. It's in the same form. For example, a pharmaceutical equivalent of an immediate release tablet would be an immediate release tablet containing the exact same drug substance, or a solution nasal spray from a pump, the pharmaceutical equivalent would be a solution nasal spray from a pump, not a suspension nasal spray, and so forth. So, we're talking about the exact same dosage form and dosage form type containing the exact same drug substance, and furthermore, the labeling and the intended uses of that product are the same. There are some other similarities, but those are the critical parts. The first thing not to be confused about is we're not talking about different drug substances. We're talking about the exact same drug substance in the exact same form for the exact same use when we talk about bioequivalence in its application to the approval of generic drugs.

So, my definition is pharmaceutical equivalence is rate and extent of absorption are not statistically different when administered to patients or subjects at the same molar dose under similar experimental conditions. So, this would be the first definition.

But I've expanded on that to say, well, that's fine to look at the definition, but what is our endpoint? And I'm speaking specifically about generic drug approval, but this could also apply to certain aspects of NDA formulation testing, especially when a set of clinical trials that go to approve an NDA are carried on in one formulation, a clinical trials formulation, and then either the scale-up or the change in the eventual marketed formulation requires some additional connection to those clinical trials, and often it's a very similar case that that new changed or scaled-up NDA formulation has to be tied back to the formulation that was used in the clinical trial. So, that's an analogous situation to perhaps approving a generic drug, although not exactly so.

So, the main endpoint or what we're trying to achieve when we look at bioequivalence of these pharmaceutically equivalent dosage forms is at the end we're trying to achieve therapeutic equivalence, and it's important to note that when I say therapeutic equivalence, I mean equivalence of both safety and efficacy of these two products. You have to really remember because that division becomes important, as you'll see later on when we talk about nasal sprays or other topical or local delivery drug products.

This is my own definition, that bioequivalence products can be substituted for each other without any adjustment in dose or other additional therapeutic monitoring. This is an official statement that's come out for other purposes when we talked about drugs like warfarin or other controversial generic drug approvals, that when bioequivalent generic products are switched for each other, that no additional monitoring over and above what would ordinarily be done for that patient for that disease, given that drug, should be needed when a bioequivalent product is switched.

I will state that the most efficient method of assuring therapeutic equivalence is to assure that the formulations perform in an equivalent manner, and probably some explanation of what I mean by performance of the formulation -- I mean, obviously we have a drug substance that comes in a product. Sometimes that drug substance is in solution, sometimes it's in a solid form, but certainly I think we would all agree that whether it's an oral tablet or a nasal spray or any other drug product that we give, the drug substance must leave the formulation and enter the patient or interact with the patient in some way.

So, that critical step is the drug substance coming out of the formulation, and as you'll see, I have a schematic of that later, but that is the critical step we're trying to measure. How does that relate to formulations? How does the drug substance actually leave the formulation and become available to the patient? And for bioequivalent products, that process or that step should be equivalent between two comparable or equivalent dosage forms.

Now, another thing that I've seen at committees and many internal FDA meetings where we have discussions like this, one of the critical things that is often confused is the difference between bioequivalence and bioavailability. The two have some similarities, but the purpose that they're done for and what's actually what we do in the types of studies to determine bioavailability, say, of a new chemical entity for a new drug product in an NDA are quite a bit different and the endpoints of what we're trying to achieve are quite a bit different than a bioequivalence test.

With a bioavailability test, it's very descriptive. You're interested in the characteristics of both the formulation, but also the drug substance. You're trying to study how that drug substance is absorbed into the body from various forms and how it interacts with the formulations that you put it in. So, it's both involved in formulation development to find out a proper formulation for that drug substance, as well as to describe the basic characteristics of the drug substance and how it's absorbed.

So, that's very important. It's descriptive. Most of the time, it's not necessarily comparative but is simply trying to figure out the important characteristics of that drug substance and perhaps how it interacts with a particular formulation.

On the other hand, bioequivalence is entirely comparative, and not only comparative, but it is specifically referring to the formulation.

Now, if you go out of here repeating this mantra to yourself, bioequivalence is all about the formulation. Presumably, by the time you get to do a bioequivalence study, you already know, through the bioavailability and other studies, what the characteristics of the drug substance are, and the drug substance is the same in the same form in two or more products that you're doing bioequivalence testing on. So, that is more or less taken out of the equation, and the only question is, are these formulations performing or releasing their drug substance in an equivalent manner or are they not?

So, if you go out repeating to yourself bioequivalence is all about the formulation or a test of comparative formulation performance, then you'll have gained quite a lot from my talk and I'll go away happy, if no one else does.

But additional definitions or explanations of bioequivalence. The question is, are the two pharmaceutically equivalent formulations equivalent in their in vivo performance, leading to therapeutic equivalence? Again, that's just kind of a rehash of what I've said before, and again I've already defined performance as the release of the drug substance from the drug product.

First off, I'd like to set the stage for nasal sprays by discussing a somewhat simpler case. Now, first, to start off, to put things in perspective, I'm displaying a simple case, which is, fortunately for us, most of the drug products that we look at, probably 70 or 80 percent, fall into this category. They're some type of oral dosage form, usually solid oral dosage forms, and fortunately for us, we see this simple process.

Now, for my schematic, I've oversimplified the process quite a lot, but I wanted to display the critical steps in going from a dosage form or a comparison of two dosage forms to our eventual outcome, which is hopefully a therapeutic or some therapeutic effects, and to do this process in a comparative manner on two dosage forms of the same drug product. So, we see that when we start out at the left, we end up with the dosage form.

Now, in a manner of speaking, those of you from the pharmaceutical industry know that this dosage form and how it's made, and therefore how it performs in the body, is the only thing in this whole process we really have control of. The rest of it, beyond that first step where the drug leaves the dosage form, which depends on how it's made and formulated, is really pretty much up to the patient or the subject, if you're looking at an normal volunteer study, and we have very little control over all the rest of the steps.

So, this first step is the only thing we as formulators or as manufacturers of the dosage form have really any control over. The rest of it beyond here -- the gut wall, appearance in the blood, appearance at the site of activity, and eventual therapeutic effect -- are mainly characteristics of the patient or patients or normal volunteers.

So, the question is, this is the process that we are trying to make equivalent. So, when, say, a generic drug manufacturer formulates their product, they try and do the best they can to make sure that this particular step, which is the dosage form performance where the drug actually leaves the dosage form and becomes available to the patient, that that step is the same between their product and the brand name or reference product.

Again, as I said, fortunately for us, this is a relatively simple process. You could put a couple of boxes in between here if you really wanted to go into a lot of detail, but as I said, I've simplified it. The drug, usually in solid form, has to leave the dosage form and go into solution. That's one of the beliefs that we have that a drug is available for absorption through the GI tract when it's in solution, and that's true not only in the GI tract, but a lot of other areas.

Some oral drugs are already in solution, so in a way we skipped this step, and the regulations that we operate under allow us to assume that for a drug in solution, the bioequivalence, when you've left that part out, is self-evident. So, the regulations allow us to waive in vivo bioequivalence testing for many oral drugs that are in solution.

So, once you get to that point, you go through a variety of steps. The drug in solution passes across the gut wall into the blood, and eventually the blood carries it to the site of activity, and when the drug appears at its site or sites of activity, we get a pharmacodynamic or clinical effect, be it a desirable one or an undesirable one.

This is a relatively simple process. We've chosen, I think -- and I could go into a long explanation of why, but to me it seemed perfectly logical -- to measure this process and to assess this step back here through the blood. Because of a variety of characteristics of blood, it's easily measured. For most drugs, not all, it's a fairly linear process. If I give a greater dose or I see a greater amount delivered from one formulation versus another, it's reflected on a linear scale here, so it's easy to relate differences in dosage form performance or relative bioavailability or bioequivalence in the blood. It's very easy to do, and the variability is quite low compared to going through a few other steps.

It's important to point out that every one of these steps adds variability as I go along. So, by the time I get to site of activity and therapeutic effect, my variability of these effects is quite high. So, again, that increases the difficulty of doing a study.

The other is sometimes, as you'll see for the nasal sprays, we really can't do this effectively. Either it's not valid to assess therapeutic effect or perhaps there might be other technical problems where we aren't able to do it in blood, and therefore we have no choice but to do either a pharmacodynamic or a clinical test of the eventual therapeutic effect to make sure that those products are therapeutically equivalent.

As you'll see, part of the whole thrust of this discussion about dose response is that the therapeutic effects or pharmacodynamic effects don't fall on a nice straight line. As most of us remember from our pharmacology textbooks, if you display it on the right type of graph, they usually have a sigmoidal or dose-response type of curve, and where you are on the dose-response curve really determines how good your measurement is. I have some blowups of these here to more characterize the difficulties of doing it on this particular graph, which is one of the questions that we have to deal with today.

So, now, I showed you a very simple case, one that we're all familiar with and fortunately we deal with most of the time. However, now we see a simplified schematic of what I've done to describe what happens with a nasal spray. There are somewhat similar things that you could draw very similar schematics for what happens with, say, an inhaler or perhaps other locally acting drug products.

This one is specifically for nasal sprays, and again it's oversimplified. Those of you who really know this process and deal with it with patients or in developing products know that there are a lot more steps here, but I did this schematic to show that, number one, we again have this critical step here, which is again the only one we really have any control of. We as U.S. manufacturers or formulators and we as regulators can control how this dosage form performs, how it releases its drug substance to the patient.

However, it then goes through a variety of processes, and as you'll see, there are branch points here that are not necessarily there in an oral product. For example, what we're really trying to achieve with most nasal sprays is a local administration to or very close to the site of activity.

So, on the top, it shows that particular process. The drug -- say it's from a suspension nasal spray -- first has to go into solution, then cross the nasal membrane to its site of activity, and you'll note here that there's no blood in between there. It simply passes across a membrane and achieves in very close proximity the site of activity, and eventually achieves a therapeutic effect. That's essentially what we're trying to achieve.

Unfortunately, knowing the characteristics of this dosage form and the way that the nose and the body handles this, this is not the only process that we have to deal with. When you give a drug from a nasal spray, or an inhaler, for that matter, but certainly from a nasal spray, a certain percentage of the dose usually is swallowed. So, perhaps not even the majority ends up in this pathway, but down in this pathway. It's swallowed. It goes into the GI tract. It can also, as another pathway, be inhaled into the lung.

Again, these could be considered undesirable, but effects that we simply probably can't get around. I mean, they do happen, and sometimes they happen when a major part of each nasal spray may go into the GI tract or the lung.

Now, the importance of this is those pathways also end up in absorption into the bloodstream. As I said, the nasal spray itself goes through the site of activity. Also, a portion or perhaps even all of the drug that's in the nasal membranes goes into the blood as well, and that blood concentration then carries it to distant sites of activity, which can result in usually toxic effects, but it can result in some therapeutic efficacy as well from systemic effects. So, we have here a much more complicated picture than what I originally outlined for an oral drug.

The advantage of an oral drug for doing blood concentrations determined by equivalence is we have one test, and through that one test we can determine both safety and efficacy equivalence. So, the entire package of therapeutic equivalence is all wrapped together very nicely and conveniently in one single test.

If we were to take the exact same rationale here and simply try and measure blood, we wouldn't really be covering all the bases. We would not be adequately assessing the therapeutic effect part, the therapy that we're trying to achieve. We probably would be correctly assessing the systemic toxicity. So, this is definitely not to say that a blood concentration test of equivalence is not useful. It simply is not a single test that encompasses or answers all of our questions.

The other question, and perhaps equally if not more important, is how much drug comes out of this formulation to reach its site of activity in the nose, and therefore create the desirable therapeutic effects.

So, we're left with a point of having no single test that can answer all our questions about therapeutic equivalence. This blood certainly doesn't answer the question about equivalence of this, but it does perhaps give us the other half about equivalence of toxic systemic effects. So, what we're left with is probably doing at least two separate bioequivalence tests to be able to answer our question, which is overall therapeutic equivalence of two pharmaceutically equivalent nasal spray dosage forms.

Now, just to finish up and discuss some of the problems, as I said, blood concentration is usually very nicely behaved, and even when we have a nonlinear drug, we can deal with that effectively. But you'll see that this dose-response PD or clinical dose-response curve does give a few problems that we have to deal with.

Now, I've drawn it on a log scale, so it has a nice sigmoidal appearance. If you don't do it on a log scale, it more or less rises and comes to a plateau, but this is the way that most pharmacology books or pharmacologists like to display it. It adequately represents the fact that this test and where I conduct the test on this curve really turns out to be quite critical. If I wanted to use this as a single test of bioequivalence and make a strong bioequivalence statement, the portion of this dose that I study it at is quite critical, and I'll explain why.

If I study it at a fairly high dose -- and you must remember that for nasal sprays, often the minimum dose that you can possibly give is probably somewhere around up here. Simply, the characteristics of many drugs available in nasal sprays, and that's part of what you'll be discussing today.

The problem that must be overcome or dealt with is the fact that if I'm up on top of this plateau, I can have two products that are quite different in their delivery of the drug substance. Well, to state an extreme, I could have several hundred percent difference, as long as I was up in this region, and the clinical effect I'd get from those two products would be virtually identical, simply because I'm up on top of a plateau and no matter how much I work my way out on this and how much difference I have between these two products, both of which are at the dose up at that plateau, the effect that I'm trying to measure, which in this case is either a clinical effect or a pharmacodynamic one, is not going to show any difference at all. So, studied in this range, the comparative test considerably lacks sensitivity.

However, if I had enough knowledge of these products and I had enough flexibility to give the products at whatever dose I wanted, which obviously is not true for a nasal spray, I would pick a dose to do my study down in here, and we have a relatively steep part of this dose-response curve, and even a tiny difference shows quite a significant difference in effect.

Granted, I've drawn this dose-response curve in a very extreme manner to be very steep. We might see a variety of different forms of this, from a very shallow to a very steep dose-response curve, but the best chance that we have of doing a good, sensitive test is to somehow establish our testing range or our dose of testing down in here.

As we all know, with a nasal spray, you can give one spray, two sprays, three sprays, and if that one spray is already up here, you have a problem. You really don't have the flexibility you might have with some other drugs of working yourself down and picking an exact dose that's ideal, even if you were able to determine what that dose is.

So, this is the problem with many nasal sprays, that even at the minimum dose, we're already up here. That's part of what we come to you today about, is dealing with this in an effective manner and putting together a package of studies that will still be convincing as far as bioequivalence of these products.

This is simply an illustration of what I said about the blood. If we have a drug with linear pharmacokinetics, it really doesn't matter what the dose is. As long as I have a good assay and can measure that dose, the same difference -- say these are two products, the blue product and the black. If I have a nice linear response and the products are this much different as far as their delivered dose, if I study it at this dose versus this dose, I'm still going to get the same response, which in this case the response is the plasma concentration difference. So, that's one of the many advantages of blood concentration monitoring for equivalence purposes, and that's why for oral products, when we can do it, we virtually always go with blood concentration monitoring to show equivalence.

In this case, as I said before, to reiterate, the blood only shows us a portion of what we want to know. It's not a single test that can answer all of our questions about nasal sprays.

That's it.

DR. LEE: Thank you very much.

We have time for maybe one question if there is any. Yes, Les?

DR. HENDELES: What is the evidence that a drug inhaled into the nose gets into the airways?

DR. CONNER: Well, I listed that as a possibility. I think Wally has some papers. Mainly, that particular route depends on particle size. It's not one of the things we're discussing today, but if you're aware of the in vitro tests we do, there's quite an effort to determine particle size and to try and make sure that the amount or the percent of fines, which are the things that presumably will get down into the lung, are the same between two products.

If you were using those same techniques to develop a new product, you'd probably want to reduce or eliminate that particular range of particle size, but part of the in vitro testing we do is try and make sure that if a reference product has that particular characteristic, that the test product also has it.

But I think, Wally, if you want to answer that?

DR. ADAMS: Some of the evidence for that, Les, is sinographic evidence. The University of Maryland has done some studies in which they've looked at nebulizers, a nasal nebulizer with a very fine, slow spray and effect. In that case, where the mass median aerodynamic diameter is down in the few micron range, they did find a substantial amount of drug getting into the lungs.

But for properly formulated nasal sprays and MDIs and DPIs, we would not expect for there to be much pulmonary deposition at all and, as Dale has indicated, we do have in vitro testing that helps protect against that.

DR. LEE: I think that we have to move on. I think Dale has done a very good job in setting the stage, and I would like to remind everybody to pay attention to the top slide on page 4, which will be reiterated by Drs. Chowdhury and Meyer later on.

The next item on the agenda is the presentation by the architect of the draft nasal BA/BE guidance, Wally Adams, and he's going to give us some history and not take too much time, but he's going to tell exactly why they posed those two questions before the subcommittee.

While we have some dead time, let me remind the subcommittee that I do have an electronic gavel. I have a button that can cut you off.

(Laughter.)

DR. LEE: And I hope I won't have to use it.

DR. ADAMS: Well, good morning, everyone. Thank you for coming. I'd like to thank the subcommittee for participating in today's activity.

Also, I'd like to thank Helen Winkle for her opening presentation with regard to the objectives of today's meeting, and for Dr. Conner for laying a very strong background for the bioequivalence issues that we're talking about today.

The title of this talk refers to the nasal bioavailability/bioequivalence guidance, and I'd like to talk about the history, recommendations, and local delivery issues, and this will lead up to presentations by Dr. Chowdhury and Dr. Meyer.

I would like to emphasize what Ms. Winkle had indicated, that today's discussion will be centering solely on nasal aerosols and nasal sprays and does not involve the orally inhaled products whatsoever.

The guidance covers four groups of drugs for local action -- corticosteroids, anticholinergics, antihistamines, and cromones -- and in three of those groups all of the presently marketed drugs are solution formulations. Only the corticosteroids exist either as solution or suspension formulations. We have a different path for solution formulation bioequivalence than we do for the suspension formulations, and I'll mention that later on. The topic today is restricted to the suspension formulations.

The outline and history of guidance recommendations for bioequivalence and local delivery bioequivalence issues. I have three slides on the history of this.

I wanted to emphasize that the issue of establishing bioequivalence for nasal sprays goes back for many years. In fact, Beconase AQ, a Glaxo SmithKline product, went off exclusivity back in July of 1990, and at the present time, 11 years later, there is still no generic product for this innovator product. So, we are still struggling with issues with regard to establishing bioequivalence for such products.

In September of 1993, the Generic Drugs Advisory Committee, with Pulmonary and Allergy Drugs Advisory Committee representation, meeting was held, and at that meeting it was determined that bioequivalence for nasal solution formulations may be established with in vitro testing only. That recommendation is reflected in our June '99 draft nasal BA/BE guidance.

In April of 1995, there was a CDER internal memo which made the recommendation for bioequivalence of generic formulation aqueous suspension nasal sprays, providing that the generic version would be qualitatively and quantitatively the same, that there were comparative in vitro data which were acceptable to the Office of Generic Drugs, and a multiple-dose PK study, and that that information establish bioequivalence. You'll notice what's not in that list is a clinical study for local delivery or rhinitis.

The second slide on the history. In December 1996, there was a letter which was received by FDA from the innovator industry. It said that OGD requirements for bioequivalence of aqueous suspension nasal sprays do not require data on drug particle size distribution. We did ask for information on droplet size distribution, but not for the particle size or particle size distribution of the active pharmaceutical ingredient, and therefore this letter contended that the requirements from OGD were not adequate to assure bioequivalence.

They made an argument, which in fact is plausible, that drug particle size distribution can affect both the rate and extent of dissolution and absorption from the aqueous suspension nasal sprays to sites of action. It is plausible. It's an argument which we took very seriously. However, that letter had no accompanying data to support the claim. It was rather a theoretical or pharmaceutically-based scientific argument in the absence of data. But we did take the issue seriously.

In May of 1997, the Orally Inhaled and Nasal Drug Products Technical Committee was organized, and then in June of 1999, the draft nasal BA/BE guidance, which came out of that organization of the technical committee, was issued. It's been somewhat over two years now, and it immediately preceded an AAPS workshop, which many of you may have attended, which talked about in vitro and in vivo issues, CMC issues, compendial issues, and in vitro and in vivo BA and BE. It was a long two-day meeting which had much valuable information in it.

The last history slide. In November of 1999, the OINDP Expert Panel was organized, as Ms. Winkle has indicated, and that was subsequently changed into a public subcommittee format, of which this is the second meeting. April 26 of 2000 was the first meeting. That was a very ambitious meeting, and we took in vitro and in vivo CMC and BA/BE issues and questions to that subcommittee meeting. A report of that subcommittee was then made to the full Advisory Committee for Pharmaceutical Science in November of 2000. That takes us up to date, then, as to where we are today.

What I'd like to do now is to talk about some regulatory issues with regard to establishing bioequivalence, and to indicate that, according to the CFR, there are four basic methods for establishing bioequivalence, and they are: pharmacokinetic studies, which is the first bullet; pharmacodynamic studies, the second bullet; comparative clinical trials, the third bullet; and comparative in vitro studies, the fourth bullet.

That list in the CFR is in descending order of accuracy, sensitivity, and reproducibility, so therefore that says, according to our regulations, that we'd prefer to establish bioequivalence based upon pharmacokinetic data when and if that is appropriate. As Dr. Conner has indicated, for locally acting drugs delivered to the nose, there are issues with regard to efficacy and issues with regard to safety, and those issues for the suspension formulations cannot be answered with a single study. So, in fact, the approach that's in our nasal BA/BE guidance uses several of these bullets in order to completely establish bioequivalence.

Bioequivalence can play a role for NDAs. It could ask questions about a to-be-marketed product. Is it comparable to the clinical trial product? For ANDAs, is a generic product bioequivalent to the innovator product? For NDAs and ANDAs, it could be used for certain post-approval changes where appropriate.

Now, the bioequivalence recommendations in the guidance are as follows, and as Dale has indicated, we are asking for in this guidance a package of information, and I'm going to go through that now.

Qualitative sameness, Q1, identical active and inactive ingredients as in the reference listed drug, and that is a key aspect of the recommendations which we make for bioequivalence because we're aware that different inactive ingredients can alter the absorption and the efficacy of a particular product. So, if an excipient or inactive ingredient is present in the reference listed drug, these recommendations say it must be present in the test product. If an ingredient is not present in the reference listed drug, it may not be present in the test product.

That's the recommendation. There could be some exceptions with regard to establishing safety and efficacy, but these are our recommendations.

A quantitative sameness, what we call Q2. Our recommendation is that each of these inactive ingredients be present within plus or minus 5 percent of the concentration in the reference listed drug. We recognize that while the labeling for the product indicates the inactive ingredient composition, it does not provide the concentrations of those ingredients, and that does require some analytical work on the part of the generic applicant in order to determine what is the concentration of each of the inactive ingredients. And that is a doable recommendation.

The guidance makes recommendations with regard to device. "Assurance of equivalence," and I'm quoting, "is greatest when the test product uses the same brand and model (particularly the metering valve or pump and actuator) as used in the reference listed drug."

If that's not feasible, we recommend that the metering valve or pump, the pump spray device, and the actuator designs should be as close as possible in all critical dimensions. Those would include such things as metering chamber volume, actuator orifice diameter, and nominal spray angle of the actuator insert.

Now, comparable in vitro performance. We have six tests that are asked for to assure equivalence in terms of in vitro performance.

The first, dose content uniformity through container life, our working group has recommended that we change that to unit spray content because in fact that is not a content uniformity test. It's not meeting a nonparametric test as recommended either by FDA's CMC guidance or the USP, but rather it's an equivalence issue. Is the test product delivering out of the actuator the same amount of drug as the reference listed drug? We have statistical criteria for that, and so we would have confidence, then, that through this first bullet, the unit spray content, that test and reference products are delivering the same amount of active drug from the actuator.

Droplet and particle size distribution is another attribute which we ask for. I'm going to come back to the particle size distribution in a moment, but first I'll flip to the next slide.

We ask for spray pattern, plume geometry, priming and repriming, and tailoff characteristics as well, all of these being comparative. The droplet size distribution, the spray pattern, and the plume geometry all impact where the drug is going to be deposited in the nose. So, if the droplet size distribution, the spray pattern, and the plume geometry are the same, then we believe that the test and reference products will deposit in the same regions of the nose.

However, particle size distribution, as Ms. Winkle indicated, is the problem that we're dealing with which brings us to this meeting today, and that is that the center is unaware of any validated method for determining particle size distribution of the drug in the nasal spray or, for that matter, nasal aerosol products. In the case of the nasal spray products, in addition to the drug, the product contains inactive ingredients, and there's an issue about determining the drug from the inactive ingredient.

I should say, too, with regard to particle size distribution that a generic firm, in order to match the particle size distribution, which can affect the rate and extent of absorption to sites of action, if the generic firm had access to the bulk drug of the innovator, then it would be an easier issue of being able to match that product before it went into the formulation. That, of course, is not the case. The generic firm has only access to the marketed product, and therefore it's a question of determining the particle size distribution of the drug in the marketed product, and that's the challenge, because this validated methodology is not available.

As Dale has indicated, differences in particle size distribution can affect rate and extent of dissolution and rate and extent of reaching the sites of action, whether it be the local sites for efficacy or systemic sites leading to toxicity, and consequently when particle size and particle size distribution cannot be determined, there are issues with regard to efficacy and safety. That's why, therefore, for these suspension products we ask for additional in vivo studies.

The guidance asks for dose response to document sensitivity, and it talks about for the local delivery three types of study designs, which Dr. Chowdhury will describe and discuss in some detail.

It also asks for a systemic exposure study, a pharmacokinetic study, and I won't go through all the bullet points here, but that study is a single- or multiple-dose study. Multiple actuations per dose to achieve measurable plasma concentrations, if necessary. These products, of course, are not intended to deliver drug systemically. The levels are very low, and therefore the PK study would be a high-dose study. We would look at AUC and Cmax measures and apply a statistical criteria to it for equivalence.

In the event that pharmacokinetics is not possible -- and we know that it is for a number of these nasal sprays, but for some it may not be possible -- when that's the case, since we're talking about corticosteroids, we recommend an adrenal axis suppression study, and the endpoint there would either be 24-hour urinary free cortisol or 24-hour serum cortisol.

So, the package of information, then, that we're talking about for establishing bioequivalence of these products is Q1 and Q2 sameness, device recommendations for comparability or equivalence, comparable in vitro performance, and for solution formulations those consist of our complete recommendations for bioequivalence.

For the suspension formulations which are on the table today, as I've mentioned, we go on to ask for in vivo studies for local delivery and systemic exposure, so there's a comparable in vivo performance for local delivery for suspension formulation products and comparable in vivo performance for systemic exposure or absorption for the suspension formulations.

Now, the guidance indicates, for the topic of local delivery issues, that a clinical study may be crucial to establish bioequivalence for local delivery. What it's getting at there is the issue of unknown or unvalidated particle size distribution between test and reference products. There are issues with regard to equivalence of local delivery.

Regarding dose-response relationship, it says that it may not be possible to show this or that it may not be consistently reproducible. So, those are some substantial challenges with regard to dose-response relationship.

It says that the clinical study should document sensitivity between different doses, that doses may differ by two- to four-fold, and it also says that that two- to four-fold range of the two active doses, that the low dose in fact may be below the minimum labeled dose, and that's consistent, then, with a recognition, however, that the minimum dose could be not less than one spray per nostril daily.

Part of that's what Dale was getting at with regard to these products. You can't arbitrarily lower the dose lower than what the product can deliver. It delivers a particular amount of dose in a spray. There's one spray per nostril for these products, or more, and you can't get lower on daily dosing than one spray per nostril. So, it limits how far down you could get on a dose-response curve, if in fact there were a substantial dose-response curve.

Well, I put this slide in. I didn't want to be outdone by Dale's slide.

(Laughter.)

DR. ADAMS: I liked his slide so much, but let me try and comment additionally on this.

First off, let me describe this. This is a dose-response slide with a dose on the x axis and percent maximum response on the y axis, and I've shown dose-response curves both for efficacy -- that is, the local activity -- and also for a safety curve where, as you go up the safety curve, you're getting increased concerns about the safety. That could be, for instance, for the corticosteroids, adrenal axis suppression.

Further, if we were on the rapidly rising portion of the efficacy curve, then clearly we would have a more sensitive means of distinguishing a test and a reference product. We would know that they're delivering different doses. But as indicated for the nasal sprays, it may very well be that we're up on the plateau of response and we cannot conduct a study in the more sensitive region where we would desire.

Now, recall from the in vitro studies, which all of these products must provide, we know that the test and reference products are delivering the same amount of drug X actuator, they're delivering the drug to the same regions of the nose, based upon the spray pattern, plume geometry, and droplet size distribution.

What we don't know, however, is that because the particle size distribution may differ, the amount of drug reaching the active sites in fact may be different between these products, even though the drug X actuator is the same and the distribution to the various regions of the nose is the same. The particle size is an issue, and consequently the position of test and reference products up on the plateau could differ, and we do not have a sensitive methodology for distinguishing where we are on that plateau.

But we would propose putting a bioequivalence criterion on the test-to-reference ratio, so that there is an equivalence criterion, and we know therefore that the test and reference products are both equally efficacious. They could be tested only at one dose, but they're both up on the plateau and they both would demonstrate equal efficacy, even though the amount of drug reaching the sites of action may be somewhat different as a result of these potential particle size differences.

Well, just as a different amount of drug could be reaching the active sites because of particle size differences, similarly systemic absorption could be different between the test and the reference products, and that puts us down onto the safety curve. We could have a situation where if these differences caused a difference in systemic delivery, that it could put us in a different place on the safety curve.

That is why we have to ask for a systemic exposure or systemic absorption study, preferably a systemic exposure study. By that, I mean a comparative pharmacokinetic study, because any of the adverse effects which are systemically mediated would be covered by pharmacokinetic equivalence, and so we would prefer to use a PK study, rather than an adrenal axis study, in order to assure equivalent systemic exposure.

But as I've mentioned, there are some drugs, some of the more recently developed corticosteroids, where the systemic absorption may be so low that the levels just can't be realistically measured in the plasma. In that case, we would ask for an adrenal axis study and assure ourselves through those studies that the test and reference products are equivalent.

So, the proposal for a bioequivalence study, then, for the nasal suspension aerosols and sprays is this. Formulation recommendations. That's the Q1 and Q2. Device recommendations. That's saying we want for the devices to be the same or as close as an applicant can get it. In vitro studies which assure us delivery is equivalent and that distribution to the various regions of the nose is equivalent.

Then for the suspension products, in vivo studies. One, to assure equivalent local delivery. That's the rhinitis study, and two, a pharmacokinetic study to assure equivalent systemic exposure. The rhinitis study, we're recommending a low dose in order to try and put us down onto the more sensitive region of the curve to the extent that that can be done, and the pharmacokinetic study, in order to get measurable levels, would be a high-dose study. The alternate to that would be a pharmacodynamic study. So, that is the proposal that we are making today.

Now, I'd like to acknowledge the participation in this development of the nasal BA/BE draft guidance of a large number of individuals within the Food and Drug Administration, within the center and OPS, and that would be our OINDP Technical Committee, and I have a subsequent slide to show that; Helen Winkle and Ajaz Hussain, both of whom strongly supported the development and the continued development of this guidance as it's dealt with a number of issues, each one of which has had to be discussed by our working groups; and Roger Williams, who initially made the suggestions for and saw the need for a guidance to provide industry with bioequivalence approaches for these locally acting products.

This last slide shows our working groups. There are six working groups which look at various aspects of the guidance, both in vitro and in vivo, CMC, and so on. Each of the individuals on here has participated in the development of this guidance and the further refinement of this guidance.

So, I'd like to stop there. Thank you.

DR. LEE: Is it really the end?

DR. ADAMS: That's the end.

DR. LEE: Thank you, Wally. You left a couple of minutes for some burning questions.

(No response.)

DR. LEE: Well, if not, I think that Wally did impress upon us what needs to be done in his slide 19, and also he stressed the difficulties that we should anticipate encountering with a PD study.

So, Dr. Chowdhury is going to tell us whether it is really difficult, or the difficulties with showing a dose response with locally acting nasal sprays.

DR. CHOWDHURY: Good morning. I'm going to talk about the dose response that we have been talking about for so long, and for the next half an hour or so I'll show some data about which I'll try to explain why showing a dose response with these locally acting drugs such as nasal sprays and aerosols are so difficult in allergic rhinitis.

Before I go into the specific data where I will talk about dose response and the difficulty in showing those, I would like to talk with you briefly about nasal sprays and aerosols in general using one slide, and then I'll talk about allergic rhinitis, specifically the study design aspects, so that when I go into the data, those will be more clear.

Now, as we heard, the nasal sprays and aerosols are in different chemical forms, and the ones we're talking about today are the suspensions. However, the nasal sprays are solutions also, which is not really the big issue, and the nasal aerosols are also suspensions. Just to make the point, the nasal aerosols are the ones which have some propellant in it like CFC, and the nasal spray solutions and suspensions are usually aqueous formulations. Throughout my talk, I'll be referring to them as solution nasal sprays, suspension nasal sprays, and nasal aerosols.

Some examples of these. The nasal sprays which are solutions, examples are the antihistamine azelastine, the anticholinergics, cromolyn sodium, and some steroids. The suspension nasal sprays are all steroids and the aerosols are, again, all steroids. Again, the focus is on the suspensions for my talk and today's discussion.

Now, on to allergic rhinitis. Allergic rhinitis studies can be done in different ways, and the way that we look at it, mainly the sponsors do it to gain approval of drugs. Typically, there are three kinds of rhinitis studies and I'll walk you through those. We just call them a natural exposure study, a day-in-the-park study, or an environmental exposure unit study, or an EEU study.

The natural exposure studies are typically done as an outpatient setting in the natural environment of the patients where they're exposed to the allergens in the natural setting. Typically, these are parallel-group studies. The patients are recruited and they're treated for an initial time period with no drug or perhaps a placebo, and we have a placebo run-in period where the patients score their symptoms or some of the efficacy measures, and we get a baseline.

Then the patients are put on drug or controls, and they're again treated for a duration, and again the same measures are recorded again, and we get a treatment effect. The difference of the baseline and treatment is the drug effect or the drug efficacy.

The duration of studies typically for a seasonal allergic rhinitis study are 2 weeks. The perennial allergic rhinitis studies are 4 weeks. However, there are some exceptions to that.

The next study design that we usually see is the day-in-the-park study, which is again a natural exposure study. The patients who are allergic to some environmental agents are taken to a park and at the time they're taken, presumably the allergen exposure is very high, and they're given the drug, and again they score symptoms. These studies last for a day, 2 days, or 3 days, and the symptom scores here are done very, very frequently. Again, these are usually parallel-group studies.

The third kind, the EEU kind, is an artificial setting. The patients are usually studied out of season, which means they're not exposed to the usual allergens. They're brought into the allergy unit, the EEU unit. They're exposed to the allergen to sensitize them, and once they're sensitized, they're brought back and given the drug or the placebo, either in a parallel fashion or in a crossover fashion, and then the effect of the drug is studied.

Now, when we go from the natural exposure to the EEU, we're actually migrating from a natural study more to a pharmacodynamic study, the EEU being a classically pharmacodynamic study. The EEU studies are typically used to look for pharmacodynamic questions, such as onset of action.

The common studies that we see are typically natural exposure studies. During drug development, sponsors and companies usually do a dose-ranging study to pick an effective or therapeutic dose, and they are typically done in the natural exposure setting.

Now, the second point I want to cover are the endpoints, and as I said, there can be varieties of endpoints one can look at. The ones that we typically use and the ones that are validated are based on patient scoring of symptoms. Depending on the drug, they can be nasal symptoms or non-nasal symptoms, and a scoring scale can be done in a variety of ways. Typically now, more and more people are using the 0-3 scale. When I show you examples, I'll show these scales. Maybe in some studies different scales were used.

The symptoms typically for a nasal drug would include nasal symptoms, such as itching, sneezing, rhinorrhea, and congestion.

Now, there are potentially other objective pharmacodynamic measures. For example, measuring nasal passage patency or inflammatory markers, such as cytokines, chemokines, cells, and nitric oxide. At this current time, they are a pretty useful experimental scientific tool. However, they are not validated to the extent that it can be used in clinical trials. Therefore, they are not typically used for drug approval.

Now, on to my examples, and I will use three drugs or drug substances to show why and how it is significant to show dose response. Some of the data I will show are proprietary. Therefore, I'll be using made-up names, A, B, and C. However, some of these, actually the results are published or available in the public domain.

Just to make a point that these problems with dose response are not something unique to the suspension. I will show an example with a solution also.

The studies that I will show are pretty large studies. I'll have five clinical trials that I will run through. For the solution, it will be a day-in-the-park study. If you remember, this is more of a pharmacodynamic kind, short duration of exposure. For the drug B and drug C, I will show one natural exposure dose-ranging study for B and for C, and then I'll also show a study where actually a suspension spray formulation was compared with an aerosol formulation. So, a lot of data. I'll go through the five clinical trials that are described one by one.

The first one, this is a solution nasal spray. I'll call it drug A and this is a day-in-the-park dose-ranging study.

Now, this study was conducted in two U.S. centers about 11 years ago and they were conducted on seasonal allergic rhinitis patients, ages 12 and older. The patients were in the park for 2 days and the drug here was used on a b.i.d schedule, which means on day 1, they got a dose in the morning and in the evening. On day 2, they got a dose in the morning. The three dose levels will be clear when I show the results.

The efficacy measured here was instantaneous scoring. By instantaneous, it is meant how the patient is feeling at the time when he is scoring, like how do I feel now?

In this particular study, they used six symptoms, which are listed here, mostly the nasal symptoms. There were also eye symptoms. The scale here was not 0 to 3, but is 0 to 5, which was used at that time, and the symptoms were scored hourly when the patients were in the park in the morning of day 1 and day 2, and when they went home they were scored less frequently. To show the results, they were all summed up, and this sum of scores that I'm going to use is called major symptoms complex.

This is the baseline that I'll show first. For this slide, and throughout the presentation, the legends here from top to bottom will follow the bars which are from left to right. So, left-most will be the top, and again throughout my presentation, the placebo will be the top or the left, and the color will be blue.

In this particular study, there was an active control, which was chlorpheniramine, and as I was talking about earlier, there were three dose levels and these are the three dose levels. This is one spray b.i.d., two sprays q.d., and two sprays b.i.d. The number of patients were approximately 50 in each group, a pretty large study.

This is the mean score, so they were not really very comparable. However, they were pretty close, between 9.5 to about 10.5.

Now, this is the result on treatment. Since the baselines were not really very much the same, I'm expressing the results here as mean percentage change from baseline. If you see the result here, between placebo and active drug, there's approximately a 30 percent difference. We are working in a 30 percent range to show an efficacy and we are actually trying to show a dose response here.

These are the three active drugs, and if you look through, it is very difficult to pick up a dose response. If you see, for example, the yellow bar, which is one spray b.i.d., and then the dark red, two sprays b.i.d., you perhaps see a dose response. However, if you look at this, which is two sprays q.d. once a day and two sprays b.i.d., the two sprays once a day appears to be more efficacious than two sprays b.i.d., which goes against it.

So, really, this is almost like a random phenomenon, and I'll show more examples, because here one can argue the separation of the doses were not that large. One spray, two sprays, just two-fold difference.

Let me show another example. This I'm calling drug B, and again there will be two trials I'll show. The first one will be a natural exposure dose-ranging trial and the second one is going to be a comparative study.

Now, this is the dose-ranging study, and as I said before, this was a natural exposure study, which means long duration. It was done in 14 centers in the U.S. about seven years ago. The patients were ragweed-sensitive seasonal allergic rhinitis patients ages six and above. There was a 1-week baseline period for the baseline scoring and there was a 4-week treatment for the drug effect.

The treatment here was q.d. There were four doses used and, as I will show you in the results, the separation of the doses here were actually over an eight-fold range, pretty large.

Efficacy here was 12-hour reflective, and the reflective means how the patients are feeling or had been feeling for the last 12 hours, almost like an area under the curve for efficacy.

They measured here three nasal symptoms -- runny nose, nasal congestion, and sneezing -- on the typical 0-3 scale. They were all summed and the total sum is called the nasal index score.

This is the total result and this is the nasal index score. This is the baseline treatment and the change, which is the difference between treatment and baseline. The baselines are very comparable. I'm showing raw data here. The point to look at really is the change, because this is the difference between baseline and treatment, and the doses here are between 32 and 256, eight-fold, and the placebo response size here is about less than 1. If you look at all four, it's basically a flat curve. The lowest and the highest dose, virtually there is no difference.

I tried to look to see if looking at individual symptoms that this composite is made up of would show anything, and it actually did not, and here it is. Rhinorrhea score, sneezing score, congestion scores. If you look at any of these, if there were any hints, perhaps it was for congestion, but again the lowest dose was more effective than the higher dose. Really, it's almost like random for almost all the doses.

Well, this is another study with the same active drug substance, drug B, and there are two formulations in the same study. One is a suspension, aqueous, and the second one is an aerosol.

The design, this was a natural exposure, seven-center Canadian study. Patients were ragweed-sensitive seasonal allergic rhinitis, ages 12 and above. It had a 1-week baseline period, followed by 3 weeks double-blind treatment.

The three dose levels, again the same drug, q.d. dosing. The efficacy here was the same: 12-hour reflective, three nasal symptoms, and the scale is 0 to 3.

If you look at the results, it's really the same story. These are three nasal symptoms: rhinorrhea, sneezing, congestion. This is the sum of these three, and this is eye symptoms.

Let's just pick up this one and look at it, because they all are the same. This particular second bar here, the dark red, is 256 q.d. This is spraying. The second one is 400 q.d., again the spraying. So, this is the lower dose and this is the higher dose. It goes in the opposite direction.

This one is 400 q.d. spraying. This one, 200 b.i.d., which is equal to 400 q.d. total dosing. This is an aerosol. If you look at it, the same dose virtually. You see some separation. So, bottom line, this is all almost waving around the baseline.

Let me go to my last example, which is drug C. I picked up these examples because I looked through almost all the drugs which are approved in the country and picked up three just to make the point. The first one I picked up was a solution. The second one I picked up was a classic case representative of almost most of the drugs. The third one I picked up because this particular drug moiety perhaps has a hint as to what is a dose response, so this may be the best case scenario. And let me show you two studies with this drug.

This was a natural exposure study done in 15 centers in 1992, pretty large. SAR patients, 18 and above. There was a 1-week baseline period, followed by 4 weeks of treatment. Q.d. dosing and the four dose levels. And the range here is higher, 16-fold, so there's a wide spectrum of dose ranges covered here.

The efficacy was almost like before, 12-hour reflective. Here they looked at eight symptoms, runny nose, congestion, sneezing, itching, and a couple of eye symptoms, and they were scored on a 0-6 scale every morning.

This is the result. The primary efficacy measurement here in this particular study was done by physicians. Typically, we would like patients to rate because patients know the symptoms better, and the others which I showed earlier were all patient recording. This one was physician recording, although the study also had patient recording and I'll show that later on. The primary was physician, so let me show that first.

The results are mean change from baseline, and I'm showing it sliced on different days, day 3 through day 28. The points are almost all the same. I'll just pick up on day 21 and show the results.

This is the change with placebo, and the effect size for placebo is about 30 percent, which we have seen typically with the nasal placebo spray because they are really also active.

Here, as I said before, this is a drug that has got some hints of a dose response. However, they are within approximately 7, 8, or 9 percentage points, so it's a very tight range. And if you look at other days -- for example, day 14 and day 28 -- it is not that consistent anymore. So, again, it is almost at the top of the dose-response curve, and what one is seeing is almost a fluctuation around that, and even if this really had a dose response, the margin between the lowest dose and the highest dose, and the difference being 16-fold, are so close that one would not be able to pick a difference between these two extremes of doses.

As I said, I'll also show the patient scoring, because they were done, and the point is here again the same. If you look at day 21, for example, there is a hint, but again it is within a very tight range, within perhaps about 10 percent or so.

The last study that I will show is another dose-ranging study where a suspension and spray were used. It's a pretty recent study done about two years ago, a very large study done in 32 centers. It was again a natural exposure study. Patients were seasonal allergic rhinitis, ages 12 and above. The study had a 1-week baseline, followed by 2 weeks double-blind treatment.

It was q.d. dosing of three dose levels from two devices. Again, a pretty wide spectrum, eight-fold range.

Efficacy was the same as before, 12-hour reflective. However, they only measured four nasal symptoms -- rhinorrhea, congestion, sneezing, and itching -- on a 0-3 scale.

This is the result. I'm showing this result as mean percentage change from baseline, and the primary was day 1 to 15. The first week and the second week is also shown here. Let me just go through this, which is day 1 through 15.

The first bar here is placebo, and there are two formulations here, so three of these would make a pair. Second, third, fourth, and fifth, sixth, and seventh. These are two. These three are with the aerosol and these three are with the spray.

If you look at it, the placebo response was again in this study about 15 to 20 percent, and if you look at the day 1 to 15, there is again a hint towards a dose response. However, we are again working within perhaps even 4 percentage points here. So, virtually, if you do statistics on that, I don't think one would ever be able to pick a difference between the lowest and the highest dose, and chances are that if the study was done again, it may come another way. And if you look at the second week, day 9 to 15, it does not really hold up. So, for an eight-fold difference, there is no dose response.

So, the bottom line here is for these nasal sprays, either solutions or suspensions or aerosols, there is perhaps no dose-response relationship, and if it exists it is very difficult to show, irrespective of what study design you pick up. I showed a natural exposure study because that's what we have data on. I showed one day-in-the-park study where one could not see a dose response.

The question comes up really why we don't or we fail to show a dose response. The reasons may be perhaps that the symptom score that we have at this time for assessment of efficacy for these drugs is not a sensitive enough measure to show a dose response. For now, that's all we have, because the other measures which I talked about earlier -- for example, the mediators or other PD measures -- are not validated to be used to assess dose response.

The second, which was touched upon earlier, is for many of these drugs, perhaps we are working at a high flat portion of the dose-response curve, and Dr. Conner showed towards the end of his talk the flat portion, and perhaps we're working there. So, it doesn't matter if you make a difference between the drugs 10-fold which I showed here, they are the same. I believe even if it was not the case, the assay that we have is perhaps not sensitive enough.

That's all I had, and thanks for your attention.

DR. LEE: Well, thank you very much.

Les?

DR. HENDELES: I have two quick questions. What is the impact of the pollen count? Does that change the dose-response relationship? And then also, is there a frequency-response relationship? I recall early data with beclomethasone showing that three times a day was better than once a day.

DR. CHOWDHURY: First of all, the pollen count on these studies were looked at, and of course the pollen count results come after the fact, and the pollen counts indeed were high in most of the studies. The day-in-the-park studies, typically pollen counts were high that day. So, pollen counts typically do not have any impact on this. However, if you move to an EEU setting, that may be a different question.

The frequency I do not think has an effect either, because we are on the flat portion, so it doesn't matter, and the formulation itself limits how much really you can go down.

DR. LEE: Other questions? Yes, Dr. Roman?

DR. ROMAN: Actually, Leslie, in one of the studies on drug B, there was q.d. versus b.i.d. and there was no difference. So, even the presented studies showed that frequency really doesn't mean anything.

I just want to make one point, and I know we will be discussing it later, so not to belabor it. One of the conclusions you made is that potentially we're on the flat curve part of the curve for dose response and we do not study the lowest effective dose. One of the limitations of what we define as the lowest effective dose is a dose which separates statistically from placebo, and throughout all your presentation we see that effectiveness is here, at the best, 10-20 percent or twice the placebo or less, and the lowest dose, if we go lower, will just simply not separate from placebo. So, I think it's a method problem, rather than the dose selection problem.

DR. CHOWDHURY: That is correct. That is correct, and perhaps with the drug C which I showed, I mean, there is a hint towards a numeric dose response, but again, if you go down in the dose, you'll be hitting placebo. Your point is actually well-taken, but for perhaps some of the older drugs which have been in the market perhaps for a longer time and the safety was not that much appreciated, it is an open question where they are on the dose-response curve.

DR. ROMAN: Exactly, and also I assume that whenever you said no difference, you mean of course no statistical difference, and whenever you said difference means statistically significantly different.

DR. CHOWDHURY: That's absolutely correct.

DR. ROMAN: Not that I think that "statistical" is the only way of believing that that works, but it's our way.

The last quick statement is also that the pollen counts you said several times were measured and they were not really different.

DR. LEE: I just noticed that Walt was motioning to say something.

DR. HAUCK: Just a quick question, actually. This might be more for Wally. Since you did have placebo in all these studies, I couldn't tell, looking through your notes, Wally, as to whether placebo is going to be recommended in your clinical studies in fact.

DR. CHOWDHURY: Yes.

DR. LEE: Very well. I think we do have time set aside for discussion at 11 o'clock, and I think now I was told that we have to take a break. Is that right? So, would you please come back at 10:30 and we will hear a presentation with Dr. Meyer on the study design.

(Recess.)

DR. LEE: Will the subcommittee members please take a seat?

I think that Dr. Meyer is very interested to get going. He is ready. He is going to tell us about study design and the topic he will address is clinical study options for locally acting nasal suspension products, clinical studies and pharmacodynamic studies.

DR. MEYER: While that's being done, while they get my slides actually to show, I just wanted to basically recap and set the stage for what I'd like to talk about.

If you recall from Wally Adams' slide about the history of how we came to be here today, in about 1995 my division, which was then the Division of Pulmonary Drug Products, sent a memo of advice on what we would consider sufficient to establish bioequivalence for the nasal sprays, given the relative short distance the drug product had to travel and the less complex anatomy than, say, the upper airways and in fact the lower airways, and the other characteristics of the nasal sprays. We had felt at that point that even for the suspension nasal spray products, that a package consisting of in vitro sameness, pharmacokinetic sameness, and Q and Q sameness, and so on was sufficient to establish bioequivalence and did not feel a clinical study was necessary.

However, with the concern raised about the possibility of particle size distribution in the formulation itself making a difference in terms of efficacy, we had shifted to working with the OGD folks to argue that perhaps if that is the case, that if one wanted to convincingly establish bioequivalence, one must do an in vivo study. And if you're doing that to establish the bioequivalence, you have to do it in a manner that demonstrates the study itself was sensitive to dose response, and in fact the response for the test and the reference were the same, given the sensitive study design.

However, at the same time that we said that and the same time we put that in our guidance, we were also cognizant of the fact that showing dose response with clinical studies, as Dr. Chowdhury has reminded us all in his talk, is extremely challenging.

Now, one could argue that if you looked at the data that Dr. Chowdhury showed, that if in fact an eight-fold dose really doesn't make much difference, why are we all that concerned about the possible impact of particle size distribution in a nasal spray suspension? And the fact of the matter is that I think everybody in my division and elsewhere is concerned that although on the mean dose doesn't seem to make much difference, individual patients are likely sensitive to dose. So, even if you can't easily show a dose response in a large study on mean, that does not negate the fact that dose may matter to individual patients.

So, that's just kind of setting the stage for how we came to be here today, and following on to the talks given this morning, I'd like to touch on four basic areas in my talk today.

I wanted to talk about what the options are for a "clinical" study. And you'll notice that I have that almost in parentheses there, or certainly in quotations, because largely what I'm talking about will be study designs with clinical endpoints, but two of the study designs that we in the division consider more, in some ways, pharmacodynamic in characteristic because they're more controlled, less natural, less generalizable. Then I will talk a little bit about the details of the clinical studies we normally see.

After I've done that, I'd like to really address this bullet, which is what is the question that you're expecting these studies to answer. Because clearly the study design that you choose will be critically dependent on the question that you're putting to that study. What answer are you expecting to get out of the study? Once you clarify the question, then what is the best study design to accomplish the end result that you're looking for? And then I'll close with some observations and come back to the recommendation that Wally Adams had shown earlier.

Well, in terms of what are the options for study design, the disease in question here is allergic rhinitis, as you've already heard, and that disease is primarily experienced and assessed subjectively, although there are ways of assessing airway patency or perhaps disruption of the nasal mucosa in terms of scoring allergic rhinitis. Primarily, again, as patients experience it, it is a symptomatic disease and that is how we have historically assessed it and how we are most comfortable. In essence, the most clinically validated pathway for assessing allergic rhinitis is through subjective measures.

The basis of approval has been, then, with clinical studies with subjective symptom scoring, and Dr. Chowdhury took us through that earlier, but typically these would include a total nasal symptom score where you have component scores for runny nose, for congestion or nasal blockage, sneezing, and nasal pruritus.

Pharmacodynamic questions, such as things like onset of action or the appropriate dose or the dosing interval, frequently, although they may still be addressed through a clinical endpoint, are addressed with differing designs in our applications. But even if we do have some day-in-the-park studies or some EEU studies, we really expect the basis of approval to come through the natural clinical study because this really integrates safety, efficacy, and a more generalized setting.

As Dr. Chowdhury had said earlier, these are generally short-term studies. Two weeks is typical for seasonal allergic rhinitis; four to six weeks for perennial allergic rhinitis, often parallel group and looking at comparative changes in total nasal symptom score over the treated period. In other words, change from baseline for active versus the placebo.

Patients are enrolled prior to or at the start of their season, they go through a baseline period to establish their symptoms, and then are randomized.

As I said, this really allows for the assessment of efficacy, but really fully, for a season anyway, it would allow a reasonable assessment of safety and tolerability. Perhaps not fully elaborating on systemic safety, but certainly on local safety and tolerability.

The EEU study uses a clinical endpoint, but in some ways we regard this more as a pharmacodynamic tool and not as a firm basis of approval, because it does not fully integrate the clinical response and is not perhaps as generalizable, but a useful tool nonetheless. As Dr. Chowdhury had previously conveyed, this is typically done out of season and exposes patients with known sensitivity to an allergen to high levels of a specific pollen, where everyone is exposed in a chamber at the same time to the same high level of pollen, and then symptoms are assessed over a very short-term period, typically over a period of hours following their exposure within the chamber.

These are often used for assessing onset of effect in dose finding. However, I would say that I'm sure that we have very firm data to establish that this study design is much more sensitive to dose response than is the typical natural study.

Then somewhere in between the two studies, in terms of generalizability and intent, is the day-in-the-park study, where a cohort of patients with known allergen sensitivity, but typically a fairly low level of symptoms, are taken to an outdoor setting where they're all exposed, because they're all in the same place, to the same pollen counts. But this is in essence a natural short-term exposure and it does allow for some short-term efficacy and safety. Again, typically in our packages, although it sort of crosses between a true clinical study and a more pharmacodynamic study, it is used more for pharmacodynamic-type questions, such as onset of effect and assessing dose response and duration of effects.

So, from our divisional standpoint, from the Division of Pulmonary and Allergy Drug Products -- and actually, in the intervening time since the guidance that we're discussing today was published, we have published a draft guidance on the evaluation of products for allergic rhinitis -- we regard the natural clinical study as most informative for approval purposes. Again, we regard these, even though they use clinical endpoints, as more pharmacodynamic in nature, and to date we've not approved new drugs solely on the basis of things like EEU, although those kind of studies are often in the NDA packages.

Other objective endpoints in any of the study design, be it the natural study or one of the shorter-term studies -- and those would be assessments of nasal patency, such as acoustic rhinometry or markers of inflammation -- we would regard as interesting, but not clinically validated. And I would say that I'm not aware of data that would establish these as being particularly more sensitive to dose response than our clinical study either.

So, let's turn to the question that we're bringing to the clinical study for bioequivalence. I think the way to really look at this, although I don't want to perhaps hang my hat too firmly on these terms, is are we looking at the study to really confirm a data package that has otherwise established pharmaceutical equivalence and bioequivalence, or are we looking at this as the primary means for establishing bioequivalence, and everything else is sort of background, but viewed as less important. So, really that depends on one's interpretation of the in vitro and BE comparisons.

But if one takes the pathway that the clinical study is confirmatory, then your question that you take into the clinical study is you want it to confirm a lack of important clinical difference from any of the unknowns that might remain after you've fully assessed through in vitro and PK comparisons.

If you look at it more as pivotal in establishing bioequivalence, then you really have to have a study that can do that, and I'll get to what I mean by that in a second.

So, for the confirmatory role, we could really look to a design that broadly assures no important clinical differences, given an already established background of pharmaceutical equivalence and systemic pharmacokinetic sameness.

If you were looking at a confirmatory setting, then, you really would not have to show a rigorous dose response or show sensitivity to dose of the test so much, because you're not going to be looking at the test and reference in such a discriminating manner. So, therefore, the comparisons could be done at a low-dose level in looking for comparable efficacy, safety, and tolerability in that study, taken on the background of everything else that has been done to date to establish equivalence.

If the intent taken to the clinical study, however, is to establish bioequivalence, then you really must show that the study design could discriminate between rather small differences in doses. In some of the studies that Dr. Chowdhury showed, we were looking at eight-fold differences in dose with very little change in clinical response. Clearly, a drug product that was being developed would have to be very, very different from the reference product to show up in these clinical study designs as being different. So, when you go to wanting the clinical study to establish bioequivalence, really where we're hanging up is coming up with a design that has sufficient sensitivity to assess differences in dose.

Now, of course, even in this scenario, a clinical study would still compare some relative safety and tolerability, but in essence the difference in the question you're bringing to the clinical study critically changes the design that would be expected or the outcome that would be expected.

Again, as Dr. Chowdhury I think showed from his survey of studies, and we have many, many others, bioequivalence would be very difficult to establish with standard designs and endpoints. So, the task set out in the draft guidance of two years ago of using the clinical study really to establish bioequivalence, because of this concern about the particle size distribution within the drug formulation, is very daunting indeed, and in fact in our experience may prove to be impossible.

Now, we did allow in the draft guidance for the possibility of using either EEU or day-in-the-park-type studies to look at this, but I must emphasize that we don't really have data to say that they are much more sensitive to dose effect. So, it's not entirely clear. They may be a better approach to establishing bioequivalence, but it's really somewhat of an unknown. But clearly, they could have a role in a confirmatory setting, just as a natural exposure clinical study would.

Now, we have had some comments to the draft guidance about the possibility of using a true pharmacodynamic-type endpoint, such as markers of nasal inflammation or measures of nasal patency. But as I've said before, these are really unproven in sensitivity to dose response and/or they're not clinically validated, so any differences detected in those may or may not be informative about the clinical differences we're concerned about.

Other potential endpoints in standard trials, such as well-validated and constructed health-related quality of life instruments, could be potentially useful, and this was again a comment raised to the docket, but I think that they're unproven as superior in sensitivity to dose response.

So, really the question that we're bringing to the subcommittee today is that we have a concern raised by comment from industry that the uncertainty about particle size within the suspension formulation for nasal sprays could impact on the local bioavailability of the formulation, but as Wally Adams pointed out, that contention was made without supportive data to show that particle size within such a suspension nasal spray would, in fact, lead to clinical differences. But it is at least a scientific concern and one that we've taken very seriously in our draft guidance.

However, given the difficulties of establishing dose response within clinical studies and using the clinical study in this package to actually establish bioequivalence, and given the fact that we're not entirely certain that this concern is in fact clinically important, we're now contemplating shifting the question of the clinical study in the BE package.

I must say, however, that no matter how one regards the clinical study, it would not trump the lack of equivalence from in vitro or systemic bioavailability. In other words, if the sponsor of the test product couldn't show that they had similar or reasonably comparable in vitro characteristics to define the drug as being the same, and they were not able to show comparable systemic activity, then they wouldn't even get to the point of doing a clinical study in either of the paradigms that we've discussed, either in the draft guidance, where we're asking for dose response, or in the paradigm that we're presenting today.

I did also want to stop and make one other point that follows on to Dr. Conner's presentation earlier, or perhaps this is actually from Wally, but what we're asking for in this draft guidance is Q and Q sameness. Dr. Conner was very careful in his talk to talk about having the same drug substance, same route of administration, same dosage form. We're actually going a step beyond what is required for oral products, for instance, where the excipients can differ. We're saying the excipients have to be the same, and in fact they have to be very close in their proportion in the drug formulation. So, we're already going a step beyond that in terms of Q and Q sameness, and then asking for characterization of the in vitro characteristics of the device itself, since that's critically important in how the drug is delivered to the patient and released, and then looking at systemic exposure.

If all of that gets you to comparability or sameness, then what we're saying is perhaps we could look at the clinical study as being confirmatory in nature, and if one wanted to do that, given the dose-response curves that Wally had shown us earlier, one could examine the lowest dose of the test versus the reference and do a statistical comparison between this lowest dose, as they relate to placebo, ensure therefore that the test or the new product is not meaningfully different in this clinical study from the reference product, and given that, with all this background, one could say that one has equivalence.

Now, if we went forward with this paradigm, one of the questions, or actually the heart of the questions presented, is what then would be the best way to do that? Would it be through a clinical study, an EEU study, a day-in-the-park study, or perhaps some other study design?

So, that is a summation of how we got here and the background to the questions that we're presenting today, and I'm happy to entertain a few minutes of questions, I guess, if anybody has any.

DR. LEE: Well, thank you very much, Dr. Meyer.

I think that the conclusion I can draw from this morning is that the PowerPoint presentation really keeps everybody on time, ahead of schedule.

I see that Les is ready for some questions.

DR. HENDELES: Is there any evidence that nasal challenge studies produce a dose response? I personally haven't seen that data. I was wondering if you have. Any form of nasal challenge.

DR. MEYER: I'm not really aware of data that has really established that, number one, and number two, I guess, depending on the particular challenge that you're talking about -- frequently we're talking about sort of Q-tips dipped in allergen and applied topically -- I think we'd still have the question of clinical validity of that, too. So, to me, it sort of has two strikes against it right now, not to say that it couldn't be useful in the future, but I'm not sure how well it's been related to clinical differences and I'm not aware of data that firmly establish that could easily show dose response either.

DR. OWNBY: I've got a question. If you've got Q and Q identity between the reference and the test product and you've got a device that's substantially the same, if not identical, is there any evidence to say you could then generate an aerosol that would be substantially different?

DR. MEYER: The question that really brings us here is not so much about the aerosol itself, but about the particle size within the droplets, and if those were to be substantially different, you wouldn't detect that through any of the testing we do because in the suspension products, for instance, the excipients, such as methyl cellulose, really interfere with your ability to assess in a validated manner the particle sizing of the drug substance.

The concern, then, isn't that the aerosol is different or that the deposition is different. It's that when the particles get to where they're going, because they might be different sizes, they might have different rates of an extent of absorption.

But our concern isn't, given Q and Q sameness, and in an ideal situation even the same metering pump and so on, that we would suspect the aerosol itself will be different.

DR. LEE: Dr. Roman?

DR. ROMAN: Yes. I would like to sort of organize my understanding and thoughts after your presentation, and if you can help me with this.

So, I understand that you believe that there are two reasons why the company would do clinical studies, one which you say is confirmatory and one when you do not have any previous bioequivalence based on the blood sort of measurements. And if there isn't any previous bioequivalence established by blood measurements, then you are suggesting that clinical studies of nasal allergy will include dose response.

DR. MEYER: Not exactly. What I'm saying is that we in the draft guidance have said that everything else has to be the same, the in vitro package and the systemic bioavailability, either directly measured by the blood, preferably, when that's possible, or established through HPA axis or some other measures of systemic effect in a pharmacodynamic study. So, that is given in anything that I said. So, if one were to fail to show sameness to any of those, to the in vitro package or to the systemic bioavailability, you're not equivalent and you couldn't go the generic route under this paradigm.

So, the main thing is the question that you're putting to the clinical study, and that really gets to the uncertainty that's remaining after you do everything else. If you feel like there's a lot of uncertainty, then you really need the clinical study to establish bioequivalence, and therefore you need a very sensitive study design that can clearly separate small differences in dose and then show sameness.

If one takes the remaining questions after you've looked at all the in vitro package and the pharmacokinetics are very few, then you're just looking to any clinical study to just make sure there aren't any important signals that we've missed in everything else we've assessed, because clearly, although what Wally Adams has presented is a very comprehensive package that we're expecting from folks, including Q and Q the sameness all the way down, there are always going to be some remaining uncertainties.

So, the question really is what are you expecting the clinical study to do? If you're only expecting it to really kind of put the products into comparison where you're looking for just any important signals of difference, you're really expecting much less of the data than where you're having to establish sensitivity to dose response and then critically look at that dose relationship.

Now, I do want to emphasize, we're still talking about statistical comparisons of this lowest dose in what we're asking about today, but it's really kind of a shift in what you're asking the clinical study to demonstrate.

DR. ROMAN: Okay, because one of the obviously reasons of the design -- and if you don't mind, I will concentrate on the natural exposure, because I agree with you that potentially the EEU and the park study, it's more of a pharmacodynamic in terms of questions you ask. However, since we are talking about products which are mostly taking some time to act, such as steroids, a single dose or very few doses which you can deliver during the park study really excludes the study designs for any meaningful clinical study with steroids, because even in the data presented previously, you can see that there is an increase in effect with time. So, indeed, the natural exposure of 2 weeks for steroids is probably the minimum which we can use for establishing efficacy, not to mention differences in doses.

But one of the major reasons we have a problem with dose response is that this nice sort of S-shaped dose response which we would all like to see does not work with intranasal delivery systems, and one of the reasons actually is that placebo is an effective treatment. A nonclassical way of treating nasal allergy is by washing the nose with water or with saline, and that is an effective, nonmedical, if you wish, treatment. So, delivering vehicle really acts almost as an active treatment. Therefore, on an S-shaped dose response, the placebo will be around 20 or 30 percent of the maximum response with an active drug. So, we can never get into the lower levels of concentration of the active formulation because we have an active treatment, which is placebo.

DR. MEYER: By and large, I agree with that. I think the interesting thing to me in that is that if you think back to the first study that Dr. Chowdhury showed, the placebo in that was vehicle. Correct?

DR. CHOWDHURY: Yes.

DR. MEYER: And actually, there was a pretty good separation between placebo and chlorpheniramine in that study. So, I think certainly the installation of the vehicle can wash away some allergen and be sort of an emollient in effect. I'm not sure that that's the whole answer to why we're not able to show a difference. That study would suggest that the placebo response, or actually the effective placebo, is not as huge a factor as one might suspect.

DR. ROMAN: Actually, if I recall, this was the product A, which Dr. Chowdhury presented, and since they used chlorpheniramine as an active comparison, I would suspect that it was antihistamines which they were studying and it was in a park study.

DR. MEYER: It was.

DR. ROMAN: So, indeed, in a park study with a single dose, you have sort of an enlarged response with an active.

DR. MEYER: Right. It gets you into a paradox with the nasal steroids because the longer you go, perhaps the more the placebo actually acts as a true therapy.

DR. ROMAN: Exactly.

Now, what I am discussing actually here is how to do the sort of well-designed clinical studies for bioequivalence with steroids, specifically. Again, I agree with you that we have to characterize them as well as we can, but there are limitations of methodology which we sort of cannot overcome with the subjective nature of an endpoint.

DR. LEE: All right. I think that we should give Dr. Meyer a break. Thank you very much.

Let me summarize what we heard this morning. We began this morning with a charge from Helen Winkle about why we are here, and then we went into the background to address two questions in relation to the development of the draft guidance on nasal BA/BE. Wally Adams gave a very comprehensive history about that, and also we had presentations by FDA scientists and clinicians about the background information.

So, now we're moving into the session on the agenda which is committee discussion, and the purpose of this is primarily to clarify any questions from any of the presentations made this morning. Then we're going to open up the floor to the public hearing, and then we adjourn for lunch. So, basically, this is an opportunity for the subcommittee members around the table to pose questions to the presenters.

And may I suggest that you identify yourself when you speak for the minutes takers, and also, just because you do not have a microphone in front of you, doesn't mean that you cannot speak. All right?

DR. SHARGEL: Hi. I'm Leon Shargel, Eon Labs. In general, bioequivalence studies, the in vivo bioequivalence studies, have not been clinical studies. As Dr. Conner mentioned, they were very often for oral dosage forms, blood level kinds of studies. And in terms of those studies, generic manufacturers do not do confirmatory clinical studies or look at clinical differences between the test and reference product, even though my brand-name counterparts feel that it should be done.

In this particular case, we have been talking about using confirmatory clinical studies. However, in the case of pharmacodynamic endpoints, they're often unsuitable, and again, if we did pharmacodynamic endpoints, we wouldn't see the need in previous type products to do confirmatory studies. It's usually assumed, such as in a blood level time curve.

My question really is, in looking at a slide of Dr. Meyer, I noticed that the environmental exposure under one of the slides says, "Often used for assessing onset of effect and dose finding." Is that a better endpoint and more objective than a clinical study?

DR. MEYER: I think at this point we really don't have enough data to look at whether differences in formulation -- I would say from the innovator perspective, we don't have enough data to say that that study design would be discriminating in terms of differences in formulation or other aspects of the drug product.

I think it is conceivable. Certainly as we've wrestled with this, it is conceivable that such attributes as onset of effect or offset of effect, in fact, might be more informative or more sensitive to dose than some of the other things we've traditionally looked at, but we really don't have a lot of data to strongly state that that is the case at this point.

DR. LEE: Yes, Dr. Ahrens?

DR. AHRENS: Dick Ahrens. I've got a couple of questions for Dr. Meyer, again to follow up on the talk. One is you noted that there would be statistical comparisons between the lowest dose of test and reference versus placebo to assure that there is not a meaningful difference of test from reference. Is it possible to give any idea of how "meaningful" is going to be defined and what kind of comparison is going to be made there? Is it simply failure to show a statistically significant difference or will there be criteria in terms of showing sameness? You get what I'm after here, I think.

DR. ADAMS: Dr. Ahrens, Wallace Adams speaking. I think we'd have to look at that issue in more detail, but at the present time, we have been discussing in our working group about a statistically significant difference between the low dose and the placebo as showing that there is sensitivity to the study.

The difference between the test and the reference?

DR. AHRENS: Yes, that's what I was interested in.

DR. ADAMS: How would we assure equivalence for that?

DR. AHRENS: Yes.

DR. ADAMS: I presume that would be a standard two one-sided test procedure equivalence study looking at the test-over-reference ratio.

DR. AHRENS: With the equivalence criteria yet to be determined.

DR. ADAMS: Yes, that's correct. As Dale has indicated, in the past the default criteria for that is 80 to 125, and so we most likely in our guidance would not include an equivalence criterion or limits for that 90 percent confidence interval to meet, but we would have to look clearly at that issue. You raise an important point.

DR. AHRENS: And one more question, if I might.

DR. LEE: Sure.

DR. AHRENS: And that relates to, if you look at inhaled steroids as used in the treatment of asthma, it seems to be clear now that one of the reasons it's so difficult to show dose-response relationships there is because most of the studies that have been done as parallel studies in the past, and the between-subject noise, just difference in asthma from one subject to the next, appears to bury the very real dose-response relationships that are seen, at least in some subjects, that can be identified when crossover studies are done.

So, my question here is are you aware of any data looking at nasal steroids where that issue has been looked at where there are crossover, as opposed to parallel, studies? I realize it would probably have to be something like an exposure unit study.

DR. MEYER: I'm not aware of any data. I don't know whether any of my colleagues from the FDA are, but I'm not aware of any such data. Even given an EEU study design, I think it would be still quite complex because of the issues of priming and so on. There would be, I would suspect, a significant period effect in such a trial that would have to be dealt with.

DR. DYKEWICZ: Mark Dykewicz. A question about I guess the efficacy versus safety assessments in terms of comparing the test versus the reference drugs. The proposal is that for assessing efficacy, you would look at a single dose of the test versus reference drug, and am I understanding correctly that in terms of assessing safety, you'd look at a single dose of the test versus reference drug, but using a dose that was the highest labeled dose or even beyond that? What I'm thinking of is that in actual use, there may be some patients who are using more than the labeled use of the drug, and so there might some additional safety issues about the labeling.

DR. MEYER: According to the draft guidance currently, and we're not, I don't think, contemplating any particular change to this component, the systemic safety would be approached through pharmacokinetics measures at a high dose, and whether that would have to be within the label range or whether there might in fact be mechanisms where you might have to go higher than label, I think we're cognizant of, but I'm not sure that the draft guidance really is very clear on that, the reason being that as one gets beyond the labeled dose, you get into issues about if you were to do, say, eight sprays of a product in very close succession, one would be concerned that you might be swallowing a larger proportion of the dose than you would be otherwise. Now, there may be ways to get around that where you wait 20 minutes in between sprays and so on.

But the upshot is that we're aware that you're probably going to have to look at a higher dose than you would be for efficacy, for instance. So, it really gets to wanting a dose that could be properly characterized through pharmacokinetics, ideally.

I would also add, though, that if one does a 2-week clinical study, you also get some important local safety information from there, although presumably, with a product that is Q and Q the same to a reference product, that shouldn't be much of an issue, but it's there nonetheless.

DR. LEE: Dr. Ownby?

DR. OWNBY: Dennis Ownby. Following up on that, if you were talking about the typical 2-week study and you talk about local safety, what kind of parameters would you look at in terms of local safety that you think would be meaningful?

DR. MEYER: What actually typically is done is the patients have daily diaries that they record adverse events in, including things like epistaxis and so on, and then it's not unusual for the clinical studies to include local observation by the clinician, just basically looking at the mucosa. We're not talking about biopsies or anything more elaborate, but basically a clinician assessment of the nasal mucosa, along with patient adverse event reporting.

DR. LEE: Of course, a question from Les.

DR. HENDELES: I'd like to pose this to Dale. If you had a drug that had 99 percent first-pass metabolism, why wouldn't you be able to use a pharmacokinetic approach to document bioequivalence, since the drug in the blood has to come from the nose?

DR. CONNER: Well, unfortunately, we don't have many that are -- I can think of one that I personally worked on when I was in Bob's division, which is fluticasone. If you've looked at the clinical information and the labeling of that drug, if we can use that as an example, the oral availability I think, if I'm remembering the labeling correctly, is less than 1 percent. So, if you simply give orally the whole dose, you do not get very much of the parent compound absorbed. So, that would be the type of drug that you're talking about.

In theory, you might say, well, I could then assume that the blood concentration that I see is primarily from the nose. I'm not really sure you would be entirely firm on that because, as I said, there are other possible -- although in your other question, we said that it's unlikely, but still possible -- that there are other routes other swallowing.

That's the first point. So, I can't be 100 percent certain that all of the drug you see in the blood simply went into the nasal mucosa and entered the blood, and all that I see is through that route.

The second is really a much more fundamental question about locally acting products, and I was talking to someone during the break about if you look at perhaps a more simple case, such as absorption through the skin, where there are no really peripheral routes of entry into the blood, the blood still ends up being a point or event, if you will, that's after the site of activity, and in many cases appearance in the blood is actually an undesirable phenomenon. You'd like to supply just enough drug to the local area to create the effect you want and minimize that absorption. So, it actually is, in most cases, undesirable and it's actually an event that happens after the event that you're really interested in, which is appearance at the local site of activity.

Some would argue and have argued that that still may contain enough information to make inference backward, but it isn't as clear cut as in oral product, where that's the intervening compartment, and you assume that there is an equilibrium between that compartment and what you're interested in, which are sites of activity.

So, there are some more technical problems before we ever accepted that for local products in general. It is an interesting question, but it's not something we're ready to necessarily assume at the current time.

DR. HENDELES: It seems that even for those drugs that have larger oral bioavailability, you could artifactually change that situation by giving a dose of activated charcoal along with a test dose and have essentially produced the same thing you have with fluticasone.

DR. LEE: Dale, do you have a response?

DR. CONNER: The activated charcoal blockade is brought up, and I think you and others have done work on that. I'm not totally convinced that that's a pure and effective way of totally blocking drug input and it doesn't otherwise affect the pharmacokinetics or disposition of the drug over and above what's blocked from getting in. Perhaps I'm not up to date on this and you know more about it. I think your early work with theophylline and some other drugs showed that even if you put the charcoal down, it tended to enhance the clearance over and above the prevention of absorption. So, I would worry about at least the theoretical aspect of that happening.

DR. HENDELES: That was a multiple-dose phenomenon. A single dose of charcoal blocks the absorption without altering the metabolism of that particular drug, but it would be really easy to show. You could give a drug orally with charcoal and without and see what you get.

DR. LEE: Wally, you want to comment?

DR. ADAMS: Yes. I wanted to respond to that also. In addition to what Dale said, Les, is that the PK data would not tell you necessarily anything about the site-specific delivery of the product to the particular regions of the nose. I believe you could get comparable PK plasma concentration versus time curves, and yet the two products may have delivered the drug to different regions of the nose, and so it wouldn't tell you anything about that, even if you were to use charcoal block and prevent the drug coming in from the gut.

We certainly know that various products can deliver differently to different regions of the nose, depending upon whether it's a nasal spray or an MDI, for instance, or depending upon a spray angle of the actuator. So, we would have that comfort or that confidence. Of course, our in vitro data helps us with that issue.

DR. CONNER: Another theoretical example that might be slightly closer to this is, if you look at systemic absorption from the lung and say we'll have a hypothetical case where drug only gets into the lung and it's only absorbed from the lung, still the lung is a very large area with different segments, and your aim may be to deposit drug or a sufficient amount of drug in certain areas of the lung and you do not want to do it in others. So, two products could perform very differently in that case. You might in theory see the same systemic absorption, and the systemic absorption doesn't necessarily feed back and say I've gotten the exact amount of drug from these two products into the exact same part of the lung that I'm trying to do. It doesn't necessarily prove that.

The same thing, in certain respects, with the nose. I could have absorption from the nose as a whole, but it might not necessarily be from the areas that I want it to be absorbed from or to reach.

DR. HENDELES: Is there any evidence that you have to apply the reference product to a certain part of the nose?

DR. CONNER: To gain efficacy? Well, that's what we're assuming here.

DR. HENDELES: We're dealing with a disease that is relatively benign, although uncomfortable. But it certainly doesn't result in emergency room visits, hospitalizations, or deaths, so it's very different from asthma.

On top of that, you have patients who are very poorly adherent on average. The adherence to this formulation, to an intranasal steroid, for example, is dramatically less than it is to an oral inhaled steroid. So, you're talking about people not taking it consistently. You're talking about a relatively benign disease, and you're talking about, at least from what I've seen, you don't have the ability to distinguish between a huge difference in dose.

So, yes, it's confirmatory, but why do you need to confirm it? If you have a good bioavailability study, where you've taken into all the accounts and you can reproduce something like the fluticasone situation, I just think that that's all you really should need and everything else is overkill, in my opinion.

DR. LEE: Are you moving for adjournment?

(Laughter.)

DR. ADAMS: Les, just for clarity, you said that you would need what in your opinion? You wouldn't need the clinical study.

DR. HENDELES: I think if you were able to do a well-controlled bioavailability study -- pharmacokinetic, blood level -- if you could do that and address all of the concerns, and there was a real tight relationship between the two products in terms of AUC and Cmax, et cetera, I think that anything more than that would be overkill.

DR. ADAMS: You're talking about with the charcoal block?

DR. HENDELES: It depends on the drug, but if you were talking about triamcinolone, you would need to do that, and maybe one step before that would be to prove in a first study that the charcoal block really blocks, that you couldn't get any drug into the blood if you took it with charcoal.

DR. LEE: Bob?

DR. MEYER: I just wanted to maybe press a couple of questions or points with you and have you respond. Would you want to look at the concentration-time curve? You mentioned the metrics of Cmax and AUC, but one of the issues that was raised by the letter about the particle size distribution within the formulation was rate of absorption, not just extent. The metrics you mentioned may or may not get to rate of absorption. They really focus more on extent of absorption. So, would you want similar curves as well as metrics for Cmax and AUC?

DR. HENDELES: Yes. I think you would have to include Tmax as well. If you had differences in Tmax, then there may be a reason to then require a clinical study, because that would probably reflect differences in release of the drug or particle distribution.

But I think if you had really solid Tmax, Cmax, AUC, and of course the appropriate design -- if it was a drug that had an accumulation factor, you may need to do a steady state instead of a single dose. I mean, you'd take that into account, but if you did a quality study and you could not show a difference, including Tmax, then I think that that proves bioequivalence to me.

DR. MEYER: Then let me press a second point. If one were to take a drug that is reasonably orally bioavailable -- triamcinolone, or beclomethasone -- has some oral bioavailability and you do a charcoal block, are you then properly assuring the bioequivalence for the systemic safety if the information you're focusing on is the pharmacokinetic characteristics of what's getting in through the nose?

DR. HENDELES: That's an interesting question, and maybe you'd have to do it with and without the charcoal block to answer that question or you'd have to do a second study at the higher dose range. Maybe you could combine them both in one study, but you'd have to have a third arm or an arm where you did it with and without the charcoal block.

DR. LEE: Leon, you have a follow-up question?

DR. SHARGEL: I just had a follow-up, more or less, comment. I agree with Dr. Hendeles about the idea of doing systemic blood levels. However, I disagree about measuring Tmax. If we do that kind of approach, we'd do the same as we do for oral drug products. Rate is generally done by Cmax, even though it may not be the best metric for rate, but it has generally been the acceptable metric, as well as AUC for extent. There's no particular reason to be different for these products if we're just talking about a blood level time curve comparison. In oral drug products, we do see occasionally differences in particle size, but we can see superimposable blood levels within statistical values. So, I would leave the criteria the same. I wouldn't try to be anything novel, since this is rather tried and true.

I also agree we do not need necessarily confirmatory clinical studies if we have an objective bioequivalence study.

DR. LEE: I would like to change -- okay.

DR. DYKEWICZ: Just one last comment.

DR. LEE: Identify yourself.

DR. DYKEWICZ: Mark Dykewicz again. One of the concerns that I have is that we're looking at the prospect of having drugs which may not have a great deal of systemic absorption, and so if we're talking about maybe some of the traditional nasal steroids, we can do all these nice pharmacokinetic studies, but that may not be the case as we get further into some new drugs, and in there I really do think it is going to be more important to have the clinical studies available for confirmation of the relative efficacy between the test and the reference drug.

DR. HENDELES: What kind of drugs did you have in mind?

DR. DYKEWICZ: The question from my colleague was what type of drug class did I have in mind? Well, for instance, in terms of the systemic absorption from nasal cromolyn, that's relatively minor. I think some of the newer nasal steroids also are having very little evidence that you can pick up blood levels with that. If I'm wrong on this, please let me know, but I think we are looking at the prospect of this type of scenario where we're not going to be able to have the information coming from systemic bioavailability studies to enter into the fray, so to speak.

DR. LEE: In the last five minutes, I would like to make sure that every speaker will have a chance to be questioned. Wally Adams has not been asked any questions, although he responded to several.

(Laughter.)

DR. LEE: Are there any questions for Wally?

(No response.)

DR. LEE: No questions for Wally. So, maybe save that for the afternoon.

Any questions for Dr. Chowdhury?

(No response.)

DR. LEE: No questions.

Yes?

DR. ROMAN: Izabela Roman. Actually, I think that I would like an official statement. Dr. Chowdhury, do you believe that with existing methodology we can measure dose response in intranasally delivered drugs?

DR. CHOWDHURY: With the existing methodologies, I do not believe that one can, and by that I mean the traditional outpatient 2-week study using the symptom scores, and looking through the completed studies that have been done for the NDAs, we don't see that. With that experience, I don't believe with some new drugs you would be able to see that.

DR. ROMAN: Thank you very much.

DR. LEE: Any other questions for anyone?

(No response.)

DR. LEE: Hearing none, we're going to move into the next item on the agenda, and that is the open public hearing. As you know, or you might not know, but there are three individuals who have expressed interest to speak, and the first two represent the Inhalation Technology Focus Group of the AAPS and the IPAC-RS. Cynthia Flynn is going to be talking about review of the CMC OINDP issues addressed by her group. She has six minutes and the timer has started.

(Laughter.)

DR. FLYNN: Good morning. My name is Cindy Flynn and I will be speaking on behalf of the ITFG/IPAC-RS Collaboration.

ITFG is an organization which is a subset of the American Association of Pharmaceutical Scientists. IPAC-RS is an industry association. These two groups have formed a collaboration in 2000 to address the various CMC and BA/BE issues which are contained in the FDA draft guidances. The technical teams had previously presented their concerns to this subcommittee in April of 2000, concerning the issues which are contained in the draft guidance.

My objective today is to provide the subcommittee with an update on the work and proposals that have been completed to date by the CMC technical teams. In addition, my colleague, Dr. Joel Sequeira, will be presenting the views of the BA/BE technical teams on dose-response studies.

As has been mentioned by Dr. Lee, I am limiting my time, so I would like to mention that additional information is contained in our written statements, which have been submitted to the FDA and the committee, and those statements are on the table in the back. Actually, outside the back door.

There are four critical CMC issues which I'd like to discuss with you today. The first issue is that of dose content uniformity. The collaboration has collected and analyzed a dose content uniformity database and we have found that 68 percent of the products analyzed do not comply with the FDA test requirements.

Subsequent to this finding, we have then met with the FDA twice to discuss the findings and to plan the work for the future. The outcome of these meetings has been that we have decided to develop an improved dose content uniformity test.

This improved test is based on a parametric tolerance interval approach, which is very similar to the approach presented by Dr. Walter Hauck at the subcommittee meeting last April of 2000. Our approach also uses test design concepts which are very similar to those proposed by the ICH. Our improved test uses quality standards which are superior to the current test that is contained in the FDA draft guidances, and we have developed our test keeping in mind the consideration and capabilities of modern inhalation technology.

The parametric tolerance interval test that we have designed allows for increased efficiency in the use of sample information. In addition, it provides improved consumer protection as compared to the current test listed in the draft guidances. It also provides improved producer protection.

In our test, we have defined quality in terms of the proportion of doses within a batch that will fall within a given target interval. We ensure this quality by having instituted three acceptance criteria. Those acceptance criteria have been established for the sample mean, the sample standard deviation, as well as a term called the "acceptance value." These three criteria ensure that the dose that will be delivered by the product will be very close to the label claim, that the variability of the dose within a batch will be very minimal, and that the frequency of outliers will be limited.

Our test provides for a consistent quality standard, regardless of the type of product tested. So, it doesn't matter if it's an MDI or a DPI, single dose, or multiple dose.

The test does have flexibility, though, with regards to the testing schedule that can be used by a producer. Our test, as designed, requires only a single test to look at both the within-unit and between-unit variability of a product. The current draft guidance actually requires two separate, independent tests.

Of course, I've just mentioned to you that we feel that there's quite a lot of advantages with our test, but there's always a tradeoff, and that tradeoff with regards to our test is that the sample size, on average, is increased as compared to the current guidance.

We anticipate that we will be providing a report that fully explains this test to the FDA in the fall, and we anticipate meeting with them to discuss this new test. We would like to very strongly recommend that this new test replace the one that is currently listed in the draft guidance.

The next issue which I'd like to discuss is that of particle size distribution. The current guidance has a requirement that the mass balance must be within 85 to 115 percent of the label claim. We feel that this is not appropriate as a drug product specification. Rather, we feel that the label claim of a product should be controlled by the emitted dose test. It might be appropriate to use the mass balance criteria as a system suitability test, but then, through validation studies, the exact limits on the mass balance must be established.

We have come to these conclusions following analysis of a database that we have collected, which showed that in general compliance with this requirement was not feasible. In fact, only 11 percent of the products within our database would meet this requirement.

A second particle size issue that I'd like to discuss very briefly is that of in vitro bioequivalency tests. The current guidance requires that a test and reference product be compared using the chi-square test, and this recommendation has been made, to the best of our understanding, based on analysis of a single product -- that is, albuterol -- using a single test method. We feel that to generalize this conclusion to all product types and using all different types of testing equipment is not appropriate at this time. Rather, we are recommending that additional investigations into alternative tests, in addition to the chi-square, be carried out to determine which is the most appropriate test for comparing in vitro bioequivalence.

The third CMC issue that I'd like to discuss, then, is that of tests and methods contained in the draft guidances. The current guidances require that the exact same battery of QC tests be performed for all products. We are recommending, by contrast, that only appropriate QC tests be selected, based on the development database.

We had provided a report to the FDA in May and this report contains recommendations concerning the eight tests that are listed on this slide.

Lastly, before my six minutes is over, I'd like to just discuss the last point, which is leachables and extractables. The key concern with regard to leachables and extractables is that the current guidance does not contain a reporting, ID, and qualification threshold for leachables and extractables. In addition, there is not a very clear and precise definition of two very important terms. That is, a "correlation" and a "critical component." We have submitted a report to the agency just this past March in which we highlight various points to consider, and I'd like to just review some of those with you.

First of all, in our paper, we are recommending that toxicological qualification be performed only on leachables. We also in our paper recommend specific reporting and qualification thresholds for leachables, and we have provided justification for our selection of those values. We have developed a process for the qualification of leachables. Our strong recommendation with regards to this point is that the guidances need to be updated to incorporate a leachables qualification program, and that reporting and toxicological qualification thresholds for leachables need to be included in the guidances.

I'd like to thank you all very much for your attention, for listening to these very critical issues for the CMC team, and we are confident that if we work together with PQRI, the subcommittee, as well as the agency, that we will be able to resolve these.

Thank you.

DR. LEE: Thank you, Cindy.

Any questions? Just one or two?

(No response.)

DR. LEE: No questions. Thank you very much.

We move on to Joel Sequeira, who's going to be talking about BA/BE team work and their comments on the issue of dose response.

DR. SEQUEIRA: Good afternoon. As mentioned by Dr. Flynn and by Dr. Lee, I'm speaking here as a representative of the BA/BE technical team of ITFG and IPAC-RS.

In our one-and-a-half-year history, the BA/BE team has been very productive and has worked constructively on this very difficult issue of bioequivalence of locally acting nasal drug products.

As you can see listed on this slide, there were three face-to-face meetings, one with the OINDP subcommittee, the Advisory Committee for Pharmaceutical Science, and the agency. The BA/BE team has also prepared three reports which were submitted to the FDA on this topic.

After review of the current literature, scientific literature and medical literature, in this area, a task which has been taken over the year and a half, we do not have any substantive new approaches on dose response for efficacy, but feel that risk assessment and risk management must be done first to put this whole issue of nasal drugs into proper perspective, as discussed later in my presentation.

In vitro study designs in draft BA/BE guidances are useful for determining comparability of products, but unproven in value for establishing clinical equivalence and substitutability.

We support inclusion of at least two doses of the reference and test product in the clinical dose-ranging study, and at least one of these doses should be representative of the currently approved dosage regimen for the reference product.

At this point in time, we agree that the traditional treatment study offers the most appropriate study design for assessing nasal drug products intended for local delivery. We agree that the 2-week duration for the study is appropriate.

However, there is a need for the draft BA/BE guidance to further develop the statistical requirements for this study if it is to be used for equivalence testing, so as to appropriately link to the guidance on allergic rhinitis without confusing the issues of equivalency and comparability. As most of you know, weaknesses of this design include dependence on seasons and a measurable placebo effect.

Since the last advisory committee meeting, the BA/BE team has sought additional information to answer the questions posed in connection with dose-response studies, in vivo study waivers for locally acting nasal products, and test metrics for in vitro as well as in vivo comparisons. This effort continues to reinforce the earlier findings that the development of robust clinical protocols, the availability of reliable metrics, and the establishment of relevant in vitro test platforms are in fact lagging behind present regulatory needs.

In my next slide, we put forth an example which highlights the need for additional work in this area. This is a study published in the Annals of Allergy, Asthma, and Immunology in 1999, and it's by Casale, Azzam, and coworkers on the demonstration of therapeutic equivalence of generic and innovator beclomethasone in SAR.

On reviewing this paper, we see three issues with this kind of a study. The first is that, as stated by the authors, the primary objective of the study was to compare two doses of the test product -- in this case, the generic -- versus the placebo. It was a secondary objective of this study to compare the reference product -- that is, the innovator product -- against the test product. We think that in this case a reversed hierarchy is more appropriate, in that it should have been the primary objective to compare the reference versus the test product.

The second issue is one of sample size. The study was designed as a study to study differences and not equivalence. The sample size was adequate to distinguish between active and placebo, but inadequate to distinguish between either type of BDP preparation or between the two doses of BDP, had there actually been a difference.

Whereas the study detected differences between active and placebo, it failed to statistically differentiate between the different actives. Failure to differentiate in this case does not mean that a difference does not exist, had the design been more appropriate in order to detect one.

The third issue with this was the dose of administration. The administration of active was followed by a placebo, and the treatments were not randomized. This brings up the issue of bias, in that the placebo could have a washout effect on the drug treatment.

I mention this paper not to reiterate or critique this particular paper, but only to use it as an example of the need for further work in this area.

This leads me to the key issues to confirming a correct study design, which are summarized on this slide. Firstly, the draft guidance must address the issue of substitutability and not confuse this with comparability, and secondly, we need to develop statistical requirements for this study design for use in equivalence testing.

Now, one way to deal with open questions in bioequivalence study design is to use risk management to focus scientific investigation on those critical elements whose uncertainties should be given priority as the development of guidances progresses.

Three risk areas that are present with locally acting nasal sprays in the context of dose response and clinical equivalence include the primary local effect, the local side effects, and systemic side effects resulting from absorption of a fraction of the applied dose.

While the first two risk areas can possibly be grouped together and dealt with in a single trial, the third must be treated independently. In fact, the types of clinical trials needed to address each risk area may be very different in nature and construction. It cannot, therefore, be presumed that an in vitro test that correctly correlates with the local actions will also be predictive of the systemic exposure.

In conclusion, the BA/BE team agrees that development and validation of an appropriate model for assessing dose response as a model for in vivo equivalence is an important element in the development of equivalence standards for this group of products.

The BA/BE team also believes that a high-risk area in the establishment of product equivalence is the systemic absorption component. We suggest the design of studies to assess systemic availability and equivalence between nasal solutions for local actions deserves appropriate attention.

Thank you for allowing us the time to present the views of the committee to this distinguished group and the FDA experts who are leading this guidance.

Thank you.

DR. LEE: Thank you, Joel.

Any questions for Joel?

(No response.)

DR. LEE: If not, thank you, Joel.

The last one for this morning is Dr. Patel, and he's going to talk about an environmental exposure chamber and the design of a pilot study to determine dose response and response variability with topical nasal steroids.

DR. PATEL: Thank you very much.

My name is Dr. Piyush Patel. I'm the medical director at Allied Clinical Research. We're a CRO in Toronto. We have a lot of experience in doing allergy and asthma studies. In fact, that's what we specialize in. We've done over 350 studies and about 15 or so rhinitis studies in the last four or five years.

What I wanted to do today was to discuss the functionality of the exposure chamber that we've just developed over the last couple of years, and share with you the thoughts of a pilot study that we've designed for this chamber that we're going to be doing this coming fall.

This has already been discussed today, but obviously there are limitations of doing studies in the traditional way or in the park setting related to the unpredictable nature of pollen exposure. Typically, pollen counts are very, very variable depending on the weather and pollen exposure is submaximal. Symptom scores are not the maximum that you would get. As the symptom scores are smaller, the degree of sensitivity of the assay isn't very good.

There are also issues of patient compliance, the issue of not being able to do the study in a timely manner in terms of seasonal dependence, and also the need for multiple sites.

To give you an idea of the variability of the pollen counts, this is the pollen counts in Southern Ontario in '99, and as you can see, there are very, very large differences in pollen counts, and you can see it reaches up to 3,000 particles per cubic meter on one day, but if you look at it two or three days later, it's down to about 50 or so particles per cubic meter. So, if you happen to schedule your day-in-the-park study on a day when pollen counts are very low, you're going to get very poor sensitivity of your study.

We've been designing the exposure chamber over the last couple of years and we have gone through several versions of a pilot chamber in an academic setting, and we're now in the process of validating our 3,000-square- foot chamber. It's been designed with a unique air flow system to exclude external allergens, mold or diesel particulates or pollution, and we've really designed it specifically for ragweed.

We have been able to show that it can consistently deliver a pollen count of anywhere from almost zero to four and a half thousand grains, but the working range is about two and half to four and a half thousand grains, which if you look back at the slide or if you think back, it's about the highest pollen count you're going to see on a very heavy pollen day.

Capacity for our unit is 110 persons.

We feel that there are a number of major advantages with using this model in doing studies with rhinitis, specifically that we can have the maximum possible symptoms. There are differences in terms of patient response to a given pollen level, and I think that one of the drawbacks of traditional studies is the assumption when you analyze them that every patient with rhinitis responds in the same way, but there is a huge variability in response for a given pollen count. Different patients will have different degree of symptoms, and so we can push the pollen count up to a level where everybody will have as much symptoms as they're going to have.

Symptoms are very typical of rhinitis symptoms, and in fact it is more sensitive, I feel, because we are excluding external other allergens, such as pollution and diesel particulates, and so it's more specific in that way.

Compliance obviously is 100 percent, and we can have accurate readings of diary and peak flow.

We've been thinking about bioequivalence, looking at the draft guidance, and grappling with issues of dose response. We feel that one of the drawbacks, as has been mentioned, is the assumption that everybody with rhinitis responds in the same way. In fact, I'm an allergist, so I work in a clinic, and we know that there is a huge difference in the degree of symptoms people get with rhinitis, and also the response to the drug is different. So, there's a variability in response to the drug and a variability in response to the allergen. Put these two together and the difference is large.

What we can do is look at onset of action to look at dose response and look at clinical efficacy, measuring symptom scores of peak nasal inspiratory flow rates.

I wanted to just briefly present this article done by Jim Day's group, a colleague of mine, in Kingston, and what they've done is in an exposure chamber setting, taken a group of allergic rhinitis ragweed-sensitive patients and exposed them sequentially every day to ragweed pollen and treated them in the morning with triamcinolone, which is a nasal steroid.

They plotted here the cumulative percentage of patients who reach 25 percent improvement in nasal symptom scores. 25 percent was felt to be a significant improvement, so that was chosen. As you can see, 40 percent or so roughly get a significant improvement on day 1, but there are a number of people who take 6, 7 days to reach 25 percent significant improvement levels. So, there is a variability of individual response to the given constant pollen count in the exposure chamber. We feel that there may be a way of using this type of a model to show a dose response with controlling other variabilities, such as pollen, et cetera.

Incidentally, there was quite a large placebo effect as well here, as we've discussed before.

This briefly is just the raw numbers of other symptoms. Congestion, rhinorrhea, itching, et cetera, and all of these also show an improvement.

So, briefly, just to present our pilot study, we're in the draft stages of designing a dose-response and a response variability study of a nasal steroid in the treatment of rhinitis in an exposure chamber.

The objectives of the study are to determine if there's a dose response for nasal steroids using a placebo, 100 milligrams every other day, 100 micrograms a day, 200 micrograms a day, 200 micrograms a day, and 400 micrograms a day, and also we would like to look at the response variability or the CV of response.

Briefly, this is the design in the exposure chamber. Each one of these arrows represents an 8-hour session in the exposure chamber where we will be measuring rhinitis nasal symptom scores, peak nasal inspiratory flow rates, patient global assessment, and physician global assessment. There is a priming session here of between one and five sessions. We feel an average of about three priming sessions of about 3 to 4 hours each, followed by weekly changes in the amount of doses the patients will get. Each one of these we feel is a similar design to Jim Day's study that we presented, so that we can assess the response of daily treatment with nasal steroid over a week period.

Now, as this protocol is drafted, we actually are looking at putting in a washout here just to reduce the carryover from one group to the next, but we would like to look at from one group to the next the degree of response and see if there's an earlier onset of response in the subjects who were responding at 6 and 7 days if we can get them to respond earlier with higher doses. We've actually looked at every other day, but looking at Wally Adams' presentation, that's something that we need to sort of discuss, but certainly in the clinic we know that patients tend to use inhaled steroids intermittently, either consciously or unconsciously, because they're noncompliant, and certainly get a good clinical response from it.

So, finally, what we were planning on analyzing is the percentage change from baseline for each day at each treatment level, calculate the AUC of the rhinitis index score, and peak nasal inspiratory flow rates. Incidentally, that's been shown in a couple of studies by Jim Day to be a significantly more reliable measure than actual nasal congestion. We'll be measuring them for each 8-hour session in the EEC, and compare that for each day and each treatment level and plot mean symptom scores for the study for each symptom.

That's it. We're hoping to do this study this fall, and I just wanted to share the design and see if we can get some feedback from the committee on this. We're hoping to do this this fall and hopefully we'll get the data by the winter.

Thank you.

DR. LEE: Thank you very much. I think you need to show us the data first.

DR. PATEL: We will.

DR. LEE: All right. Thank you very much.

Any questions?

DR. AHRENS: Can I make a comment?

DR. LEE: Yes.

DR. AHRENS: In the pilot design that you showed there, one of the concerns that I would have is that you have progressively increasing doses as time goes on, and so that you have the time effect that you showed in your previous slides to be thoroughly confounded with the dose effect. So, a randomized allocation of the treatments, as opposed to sequentially increasing, so you could put things like first-order carryover in the model and look at that, would be I think extremely important. It's a valuable study, but I think the tweaking of the study design is really important.

DR. PATEL: Thank you. I appreciate that. I think we need an adequate washout between the periods also to reduce the carryover effect.

DR. AHRENS: That would certainly help, but not be a substitute for randomization of the doses.

DR. LEE: One last question. Dr. Roman?

DR. ROMAN: It requires some tweaking indeed, because if you go do the washout, then priming will have to be repeated. Secondly, if you do it in the ragweed season, which you are saying you intend to do, then obviously they will be primed in various ways, different patients. So, just for your thinking, not so much discussion.

DR. PATEL: Well, the priming seems to last, from Jim Day's work, about 40 days, so that is probably not an issue. If we start during the ragweed season, that means that we don't need to prime as much as we would do if we do it off-season, because the subjects are naturally primed.

DR. ROMAN: However, the washout of steroids is at least 2 weeks.

DR. PATEL: Yes.

DR. LEE: Thank you very much.

Before we adjourn for the morning, I would like to remind the subcommittee that we have work to do in the afternoon, and Wally has proposed two specific questions for us to address and they are addressed in the context of allergic rhinitis. It was prompted by this issue of particle size distribution and I think that you can read over those two questions, but we do have to come back and provide him with some guidance. The purpose is to develop a consensus. We are not going to take a vote.

So, we're going to adjourn for lunch. We are way ahead of schedule, so to anticipate what will be a long afternoon, I would just say that we are going to come back at 1 o'clock instead of 1:30.

Thank you very much.

(Whereupon, at 11:58 a.m., the committee was recessed, to reconvene at 1:00 p.m., this same day.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AFTERNOON SESSION

(1:10 p.m.)

DR. LEE: Will you please take your seats so we can continue with the deliberation? As I mentioned before lunch, the subcommittee has work to do and we hope that we can quickly come to some consensus, so that we can all move on to something else.

But on the screen is the background to the two questions that the subcommittee is asked to address, and Wally, correct me if I misspoke, when you were in the process of evolving the guidance, you encountered an issue of particle size distribution which was difficult to handle experimentally, and therefore the thought was should a clinical study be used to determine whether the particle size distribution difference, if any, might affect performance. Is that right?

DR. ADAMS: That's correct.

DR. LEE: That's correct, okay. So, that being the case, can we have the next slide?

DR. ADAMS: Before we get to those two questions, I'll read that preliminary information.

DR. LEE: Okay.

DR. ADAMS: Because that's critical to the responses to these two questions, to realize what we're proposing here as a package of information necessary to establish equivalence. So, this information that's projected before you now is the lead-in for those two questions, and I'd just like to read through that, just so that we all understand what that's saying.

"To establish bioequivalence of suspension formulation nasal aerosols and nasal sprays for allergic rhinitis, the June 1999 draft guidance, Bioavailability and Bioequivalence Studies for Nasal Aerosols and Nasal Sprays for Local Action, recommends the following: equivalence of formulation, both qualitatively and quantitatively; equivalence of device; equivalence of in vitro studies; and equivalence of systemic exposure or systemic absorption."

So, going in, then, to these two questions is the idea that, as we indicated, prior to a clinical study being done, as Dr. Meyer pointed out this morning, that there would be Q1 and Q2 formulation equivalence, device comparability or equivalence. There would be all of the in vitro studies that we request, which do not include a validated particle size distribution, however, and also a systemic exposure or systemic absorption study, preferably a PK systemic exposure study.

"The in vitro studies, however, do not assure equivalence of particle size of the suspended drug. Because particle size differences between test and reference products have the potential to alter the rate and extent of delivery of drug to local sites of action in the nose, differences in clinical effectiveness could result. For this reason, the draft guidance also recommends the conduct of a clinical study for allergic rhinitis to confirm equivalent local delivery."

And recall what we said this morning was that the Q1/Q2 in the case of the formulation is the same, the in vitro assure that the amount of drug X actuator is the same between test and reference products, and assures that the distribution of the drug to the various sites of action in the nose is the same, and that a PK study or, if necessary, systemic absorption study, is done in order to assure equivalent exposure systemically. Given all of that information as a package, then the clinical rhinitis study would be conducted.

"For this reason, the draft guidance also recommends the conduct of a clinical study for allergic rhinitis to confirm equivalent local delivery. Providing equivalence of each of the items" above in the rhinitis study -- and then we turn to the two questions which we have in the other format, so if we could do that.

Then the two questions to the committee are, first, "does the committee believe that a placebo-controlled traditional two-week rhinitis study conducted at the lowest active dose is sufficient to confirm equivalent local delivery of suspension formulation nasal sprays and nasal aerosols for allergic rhinitis?" And second, "does the committee believe that a placebo-controlled park study or environmental exposure unit study conducted at the lowest active dose is an acceptable option to confirm equivalent local delivery of suspension formulation nasal sprays and nasal aerosols for allergic rhinitis?"

So, I felt it was important to read that introductory, lengthy paragraph to set the stage for the questions that we're asking in 1 and 2, which is given that prior package of information that we have on the formulation, the in vitro and the PK study, given all of that, is the rhinitis study conducted at the lowest active dose appropriate under those circumstances?

DR. LEE: Everybody understands that? Leslie, do you have a question for Wally?

DR. HENDELES: It just occurred to me that the whole morning we talked about seasonal allergic rhinitis, and I guess my question is whether everything you said about seasonal in terms of flat-dose response curve is also true for something like nasal stuffiness and perennial allergic rhinitis.

DR. CHOWDHURY: The answer is yes, that whatever is true for SAR usually translates, and it does, to PAR.

DR. LEE: Dr. Ownby, you have a question?

DR. OWNBY: The question I really had, are we assuming that given all of the in vitro data and the dosing study, a PK study, that we will then automatically move on to a clinical study, that the clinical study is actually necessary, or do you need consensus on that question first?

DR. ADAMS: I would say that the package of information which we have in the June 1999 guidance and which we're presenting at this subcommittee meeting is the result of many deliberations from the technical committee or the working groups within the technical committee, and it's our feeling that because of this particle size distribution issue, that a rhinitis study at the present time should be conducted.

Does that answer the question?

DR. OWNBY: My follow-up on that, though, is that in all the information presented this morning, that there was nothing about the degree of particle size disparity within essentially identical systems, and can you give us any idea of how much variation you really think exists in those, once you've got Q1 and Q2?

DR. ADAMS: Yes. I would say that they're two separate issues, because Q1 and Q2 doesn't speak at all to particle size and particle size distribution of the drug within the formulation. We do know that the drug, the active pharmaceutical ingredient which is formulated into these suspension formulations, is micronized drug, down in the low numbers of microns in the median diameter range, even though the droplets from these drugs are much larger, up in the 30 to 40 or so micron range. The micronized drug that's formulated into the products is much smaller than that. But there's clearly a potential for test and reference products to differ substantially in that degree of median particle size.

DR. HENDELES: I just want to comment. You say there's the potential for them to differ, but how would you possibly measure it if you can't distinguish between 256 and 32 micrograms of budesonide?

DR. ADAMS: Well, in fact, we're not asking that the particle size distribution be determined by the firms in a validated method. The guidance does ask for studies to be done by the firms to examine particle size distribution of the active pharmaceutical ingredient, but that is for their own benefit and information, because there are concerns about the potential for differences in -- possibly PK may be more sensitive in the clinical rhinitis study. So, there are potential for differences between the products. So, that is a recommendation to assist firms, but we will not be asking for that to be compared in a statistical sense, I believe.

DR. LEE: Let me set a stage because I think it turns into a grilling session for Wally.

DR. ADAMS: That's fine.

DR. LEE: Let me say this. We have the question in front of us. We're looking at a condition of allergic rhinitis that's very specific. We're looking at local delivery. We are not looking at systemic delivery per se.

We have an issue about particle size distribution. I think the in vitro data suggested by the guidance already have shown there might be equivalence. However, there's no assurance that this dosage form or this formulation, upon administration in the nose, might behave differently.

I'm going to go around the table and I think what I'd like to do is to have the committee address the question. What is your personal feeling? Are there sufficient scientific reasons to believe that the first question posed is appropriate? And then also express, if any, some concerns.

Is that all right?

DR. ADAMS: Vince, if I could just interject one comment before you start that. The paradigm that we're talking about does not necessarily require that the particle size distribution be the same between test and reference products. It requires total nasal symptom scores to be equivalent and it requires the PK to be equivalent, but if that can happen with different particle size, then that would be acceptable.

DR. LEE: Shall we proceed? Leslie, you were about to say something, so let me start with you.

DR. HENDELES: In considering this, is there the possibility that a firm could choose for one of the treatment arms a below-labeling dose of the medication? In other words, you could say that 16 is no different from placebo, but 32 is. Some statement like that. Let's use budesonide as the example.

DR. ADAMS: I would say, to start on that question, Les, one of the requirements is that the lowest active dose that's on the slide be statistically greater than the placebo dose, so it has to be an active dose.

DR. HENDELES: But how do you know that that dose is not on the top of the dose-response curve? I mean, the lowest dose in the labeling may be at the top of the dose-response curve.

DR. ADAMS: Well, it may be. It may be at the top of the dose-response curve.

DR. HENDELES: So, all you want to do is show that it works.

DR. ADAMS: All we want to do, in this paradigm that we're talking about, is to show that that low dose has equivalent efficacy in terms of total nasal symptom scores. Just show that the two of them, the test and reference products, are equivalent. Why the lowest dose? Because it puts us as far down on the dose-response curve as we can get, but all we're asking in that paradigm is that the test and reference products be equivalent in their total nasal symptom score. That assures equivalence of efficacy and then we move on to the other studies for equivalence of safety.

DR. HENDELES: In my mind, it doesn't show equivalence of efficacy if they're at the top of the dose-response curve, because that means that the test product could deliver a fraction of the drug that the reference product delivers to the active site and still be equivalent, and that's okay.

DR. ADAMS: That's correct. You know, we have to live with the products that are marketed in the lowest possible dose that we can give, and if that lowest daily dose is at the top of the dose-response curve and the study is done and shows equivalent total nasal symptom score for both test and reference, we know then that it meets the criteria for equivalence, even though the fraction of the drug which reaches the local sites of activity in the nose may be different, but they're still equally efficacious.

It also raises the question, of course, that that difference in particle size distribution which would cause that effect of differences in the amount of drug reaching the local sites of action may also have different systemic absorption. That's why we need to couple this with some measure of systemic absorption to assure that indeed they're just as -- have equivalent --

DR. HENDELES: And I don't have any trouble with the systemic absorption part of it. That seems justifiable to me, but I still have a problem with the topical efficacy part.

DR. LEE: So, you're not comfortable with that?

DR. HENDELES: I don't think it's necessary. I think that it's just way overkill for all of the reasons I listed in the previous hour. I don't think this question is relevant because I don't think that, by and large, unless there were some exceptional circumstances where you couldn't use a more sensitive assay and do pharmacokinetic studies, et cetera, that maybe it would be reasonable, but we're talking about a disease that is mild.

Let me give you an example. Hydroxyzine is used for treating acute urticaria. You have approved generic hydroxyzine based upon in vitro dissolution tests. You didn't even require those products to be tested for bioavailability.

So, I don't understand why you're applying such a stricter criteria in this situation when the clinical circumstances don't warrant it.

DR. ADAMS: I'd certainly be interested in other of our FDA colleagues responding to that question, but let me just indicate that we're trying to put in place a practical approach at the present time.

Now, Les, you talked about PK studies as one possibility, and indeed we had taken that question about comparable in vitro performance and in vitro studies only to our April 26 of 2000 subcommittee meeting, and certainly there is some possibility of doing that, but there are some practical issues involved here.

Of course, we're dealing with drugs which are intended for local action which have very low levels in the plasma, and it's a challenge to develop sensitive analytical methodology to adequately characterize that PK profile.

So, if you were to do something that you are talking about, where we're dealing with levels already that are very low and in some cases may be approaching the limit of quantitation, if you then use an approach of saying, well, I will use charcoal block as part of my approach such that the plasma levels are due only to the nasal or nasopharyngeal or whatever absorption and not gut absorption, then to the extent that the drug being seen systemically is due to gut absorption, you've cut that amount of it out of there and you've made the analytical challenge all that much greater.

Dale talked about fluticasone propionate, where the oral bioavailability is less than 1 percent and the nasal bioavailability, according to the labeling, is less than 2 percent, and so if just roughly 50 percent of that area under the curve is getting into the gut, you've cut that portion out.

DR. HENDELES: There's no area under the curve with fluticasone getting into the gut from oral inhalation. Are you saying that when you administer it nasally, you're getting drug into the --

DR. ADAMS: I don't have the data at hand, but when I look at the approved labeling, it indicates that the nasal bioavailability is less than 2 percent after nasal dosing, and after oral dosing it's less than 1 percent.

One more point. Just to further elaborate upon that, there was a paper published on beclomethasone dipropionate very recently by Glaxo, and in that they did PK studies with and without charcoal block, and what they found when they used charcoal block was that the nasal bioavailability was less than 1 percent. Most of the drug is coming in through the gut.

So, while you could have measurable levels after nasal dosing without charcoal block, with charcoal block, I think it's going to cut the levels way, way down. So, there's certainly a practical issue involved in that.

DR. LEE: We will come back to Les.

DR. HENDELES: Well, just to respond to the practical issue, I think it's not practical to do these clinical rhinitis studies, given all the information that I've seen here today. We'll have to agree to disagree, Wally.

DR. LEE: All right. That's what we're here for.

Mark, are you ready to offer your comments?

DR. DYKEWICZ: I can make some comments, yes. What I'd like to do is just to step back and take a bigger picture view of this dilemma, and that is what we're really trying to do is to establish whether these nasal drugs are safe and effective, and by extension, whether a new formulation or -- I shouldn't say that, but a new test product would be relatively equivalent in effectiveness and equivalent in safety.

To me as a clinician, the most straightforward thing is to do a study in the human being in terms of looking at effectiveness and safety, and as elegant as many of these pharmacokinetic studies may be, we've discussed some of the misgivings that we have in terms of particle distribution and size and so forth. I don't think for a sponsor it would be that odious to require that there be one clinical study demonstrating reasonable equivalence in safety and efficacy.

So, my feeling is that it is an appropriate question to ask. I think ultimately clinicians, and probably patients, would have greater confidence in a new test product if it were demonstrated, if you will, in real life, with some misgivings, in a clinical trial with kind of a standard 2-week assessment. So, that's the way I would do it.

DR. LEE: Mark, this is a very specific suggestion, which is one single dose.

DR. DYKEWICZ: Yes. I'm comfortable with what we've talked about. If we're looking at some sort of equivalent efficacy, going to the lowest dose may not be beyond the top part of the dose-response curve, but in real life that's the lowest dose people would be taking -- one puff or one spray or whatever -- and although you could say, well, why don't we go down to one-tenth the dose and see whether there's an equivalent sort of response, that in my mind is probably not clinically necessary either. We're trying to look at what people are actually going to be using.

In a similar way, looking at toxicity, safety, the second big question, I think that would be reasonably met by looking at the high end of either the labeled dosing or, as they mentioned, maybe even just a little bit beyond to see if, because in real life people might be taking more extra puffs, that that was equivalent in safety with the reference product.

DR. LEE: Thank you.

DR. OWNBY: Well, I'm still having misgivings. I'm not entirely sure that a clinical study is necessary from what I've seen, but I would be willing to accept a clinical comparison that showed no difference, provided we have a high-dose study showing that there is equal safety from these.

It just seems that the biggest problem is that, as we've seen, these clinical studies are very, very blunt instruments, and we've talked about a couple of the reasons. One, that there's probably an active placebo effect, which makes it harder to distinguish active drug. The second thing that was mentioned is the imprecise measurement, that we're using questionnaires and summing the scores, which have questionable validity.

But the other problem is that there's a small response, and that could be either that these drugs are just not very effective, which I think a lot of my patients would tell you is true, or that we have a poor choice of subjects. No one has mentioned the fact that as you enroll patients in these clinical studies, I don't think a lot of these patients have true seasonal allergic rhinitis or it's very mild, and that's why you either have to superchallenge them to see an effect or you have to be more selective. So, I think there are still a lot of issues in how we set up clinical trials that should have been answered.

DR. LEE: Thank you.

Dr. Ahrens?

DR. AHRENS: For the reasons that have already been mentioned, I would come down on the side of I think you do need a clinical trial, and because it hasn't been brought up yet, I would like to briefly examine maybe the other side of that, of would you need to go so far as to do a dose-response relationship, which is I think one of the questions that's been put to us today.

It seems to me that, for all of the reasons that have been said, it would be, given the current state of the art, very difficult to accomplish that, and while it is entirely possible that there might be a few people who have real dose-response relationships that are getting lost in between-patient variability and a bunch of patients who are at the top of their dose-response curve, we just don't have the technology to look at that at this point.

I guess, taking that a little further, I'd like to ask the clinicians in the group here how many of you think that you see dose-response relationships with intranasal steroids? I think when you ask that question of most clinicians about inhaled steroid use in the treatment of asthma, most will respond indeed dose does make a difference there. Maybe not all, but most.

But I'd like to now ask the question about nasal steroids. Do any of you think you see dose-response relationships within the nasal steroids?

DR. DYKEWICZ: I think it may occur in individual patients. I mean, there are situations where I'll start somebody at a lower dose and I'll bump them up and they seem to do better, but then the question, of course, is are they going to be doing better anyway because that's kind of the progress of the allergy season.

I guess it also brings up the other point. Whenever you're looking at these mean results, you may be missing differences that might exist for individual patients that are just kind of getting averaged out.

So, there's no clear way of demonstrating this objectively, I suppose, other than you can probably in individual patients have different washout periods and different doses, maybe for perennial allergic rhinitis, but it would be very difficult to truly assess this analytically for seasonal allergic rhinitis.

DR. HENDELES: In Florida, we don't have seasonal allergic rhinitis, or at least not very much. We have a big problem with dust mite, and in that situation, because it's ongoing, the biggest problem is patients not taking their medicine. So, when somebody's not responsive, we call their pharmacy and find out how often they're getting their prescription refilled, and it's always a month's supply is lasting six months.

So, I don't know of any patients in our clinic where we've ever tried raising the dose. Usually we try and get them to take the medicine.

DR. AHRENS: I have fiddled around with that some. I guess I have not ever felt that I have seen a dose-response relationship either. So, it sounds like as close as I got is maybe a qualified "it is possible."

With that as a given, it seems to me that this is one more reason why this is a very different situation than topical steroids for the treatment of asthma. I don't think we have this clinical impression even that is very strong that there is a dose-response relationship here. So, to go through all the difficulty that would be created to try to come up with some suitable model that would just about have to be some kind of complicated crossover design, probably be an environmental challenge kind of system, you would have to deal with the issues that Dr. Roman brought up about how do you deal with priming, how do you deal with period effect. It may be possible to do that, but it would be a major challenge.

Given the fact that we don't even seem to have much of a clinical impression that this is important at all, again unlike asthma, as somebody who has spent a major part of his professional career thinking about how you make dose-response relationship comparisons between topical drugs mainly in asthma, I don't think you need that kind of crossover study here.

So, I think you do need some kind of study. Again, for reasons that have already been mentioned, I am not comfortable with just going with the pharmacokinetics, but I think the kind of simple single dose versus single dose -- or one dose level of each would be sufficient.

DR. LEE: Thank you very much.

Gloria?

DR. ANDERSON: For the two questions, my answer would be yes, I would like to qualify the "believe." I have a concern about that, and that is more as a chemist than anything else. I had a professor once who said that scientists don't believe things.

Let me just say that I think that you ought to proceed.

I have a few concerns. One, of course, is that there are so many variables that apparently cannot be controlled the way that some of the things that I do can be controlled. I have a concern with the "sufficient to confirm equivalent local delivery."

But more importantly, the concern I have is related to your last statement before the questions, and that is that the assumption is that the items in sentence 1 are equivalent or have been determined I guess I should say.

The one that I have a concern about is systemic absorption. You go on to say that particle size has effect on the rate and the extent of delivery of the drug to the desired site. I am sure that is correct because that is a fundamental theory. The problem is, it seems to me like from what I know, that absorption also is affected by particle size. This is particularly a problem in suspensions. I haven't seen anything down here about homogeneity in terms of suspensions as well, which may be a problem.

I am not suggesting that you look into them. I would ask you to comment on systemic absorption. My concern is that the answer to questions 1 and 2 depend on an assumption that these things are equivalent, and it is the systemic absorption that I would like for you to comment on.

DR. ADAMS: I am not sure what you meant by homogeneity of absorption. Could you explain that please? You mean the physical product? The particle size distribution?

DR. ANDERSON: I did not mean for you to comment on that. I guess I threw that in because in a suspension you have got a lot of things happening, but my question is related to absorption because the size of the particle should affect the rate of absorption. The questions that you are asking require an assumption that all these things are equivalent. And I am just asking you in your expert opinion to comment on systemic absorption.

DR. ADAMS: Well, of course, that is one of the elements in this package, the pharmacokinetic study, and if there are differences in the rate of absorption which would be likely if particle size distribution were substantially different, that would be captured in a PK study. So, we do look at that.

DR. ANDERSON: In fact, then the answer to 1 and 2 may not have the assumption of all of these things in sentence number 1. That is my concern.

DR. ADAMS: I think there is a distinction here that perhaps you are going down the same path that Dr. Hendeles was going down with the use of the PK study. If you don't use charcoal block, then the rate and extent of absorption which you see in the PK study would be either due to the nasal absorption, which may be very low, and/or gut absorption. It would be a combination of both of those, or possibly pulmonary absorption, if that happened.

DR. ANDERSON: I am not an M.D.

My concern is this statement, providing equivalence. I don't mean to belabor the issue here, but the answer to questions 1 and 2 depends on the statement above, which includes systemic absorption.

DR. ADAMS: Yes, it does.

DR. ANDERSON: And all I'm saying is that there should be a problem with this. If I answered this yes, I'm assuming that there is no problem. Now, I'm answering this yes, but I want to point that out. That is all I'm saying.

DR. ADAMS: Well, again, the PK study with comparable AUC and Cmax and Tmax, if that were to be looked at, does not necessarily assure that the drug is getting to the sites of local delivery at the same rate and extent because the plasma levels which you see are a combination of drug which comes in via more than one route potentially, as Dr. Conner had in his slides. So, that systemic PK study does not separate out nasal absorption from gut absorption. So, there is still an issue about rate and extent to local sites of action.

Dr. Meyer, did you have an additional thought on that?

DR. ANDERSON: Excuse me. I'm not going into that much detail. I think the problem here is a different one for me, and that is what you're seeing down here is that there must be equivalence in terms of items in that sentence to the extent possible. That's the problem I have.

DR. ADAMS: Well, to the extent possible, for the PK study, it refers statistical equivalence.

DR. MEYER: I was just going to reflect that I think what we are saying there -- if I hear you correctly, you are saying particle size does matter. It is going to impact on the absorption. What we are saying under our supposition or the way we framed these questions is that although we can't assess the comparability of the particle size in the test and reference product in a suspension nasal spray because of the difficulties of the interference from the excipients and so on, it has gotten to the point where that systemic bioavailability has not shown a difference.

So, at the point where we are asking you to be, in terms of answering these questions, we are saying that we have not assessed in vitro in a validated manner the comparative particle size, but we have gotten to the point where whatever pharmacokinetic data have been generated don't show a difference. So, yes, particle size could impact on that, but we're down a pathway where, for whatever reason -- maybe it is similarity of the particle size that we haven't been able to assess in vitro -- we're not seeing a difference in systemic bioavailability.

DR. ANDERSON: Maybe my question is then, what do you mean by "providing equivalence of each of the terms in the first sentence exists"? Maybe that's the question I need to raise. Maybe I'm not understanding it.

DR. ADAMS: Maybe it could be written more carefully, although there was an attempt at that.

DR. ANDERSON: That doesn't mean the test and the reference -- all these things are equivalent in the test and the reference. Is that correct?

DR. ADAMS: Yes, it does.

DR. ANDERSON: Okay. That's the way I interpreted it. The device, the Q1, the Q2, the in vitro studies, all of that.

Then you come to systemic exposure and systemic absorption, which to me, according to this sentence, says that they're equivalent, but in the next sentence you talk about particle size.

I'm not going to belabor it because you guys are the experts.

DR. MEYER: I think the point is maybe the part about the systemic exposure and absorption should have followed the sentence about the in vitro studies, but the point is you do the in vitro studies, and if everything you can assess looks equivalent, you then do the systemic bioavailability studies. If they then look equivalent, then you get to the clinical study. The clinical study doesn't trump these and the systemic bioavailability study does not trump the in vitro. You are building to this.

DR. LEE: Is it clear?

DR. ANDERSON: Well, I'm finished. I still say particle size will have an effect on that, maybe that's not a part of that sentence.

DR. LEE: Let me summarize the prevailing opinion on this side of the table, which is that there is a need for a clinical study.

So, let me now turn to my left. Leon?

DR. SHARGEL: I have some concerns whether the clinical study, as designed here, is really appropriate for determining bioequivalence. In terms of bioequivalence, as Dr. Conner mentioned, we're looking at the performance. We are comparing the bioavailability of a test drug product against a reference drug product. Generally when we do that kind of test, whether we do a blood level time curve for all drug pharmaceuticals or pharmacodynamic kind of study, we make the assumption that if it is shown to be bioequivalent under whatever standards we have, which is usually Cmax/AUC, that the performance of these products will give a similar clinical efficacy and safety as well.

Now, if we look at this study -- and the question seems to be this morning a discussion whether it really can be sensitive to look at different doses, and even if we use only one dose and we see no difference in the clinical effect, can we then assume that the products are bioequivalent? And that is what we seem to be doing. We can say, well, both products give the same clinical endpoint. But you really don't know where you are on the log dose response for sure.

So, unless you're able to really have some more objective measurements, in my mind, to distinguish whether we're at the appropriate dose and at small differences in dose or particle size, which in a sense -- and other kinds which lead to changes, say, in the rate of absorption, will make a difference in terms of bioequivalence because one of the parameters or metrics is rate of absorption. Many times we see small changes in particle size makes no difference. In other cases, it may.

So, are we really measuring particle size here? I'm not sure in my mind that we are. Or rate of absorption? We are saying, okay, we've got a clinical endpoint. Both give the same clinical efficacy, and therefore we're bioequivalent. I'm not sure I'm happy with that.

I am happy, though, if you say bioequivalent on some objective means such as, say, the blood level time curve. I think throughout the years in the last 20 years, we've seen that that is a pretty good guess, particularly if the Q1/Q2, quantitative/qualitative excipients are the same, there is less likelihood of any untoward effects on that case and the dose is about the same and the pharmaceutical equivalence and follow all the in vitro things.

So, I agree with Les. I think this is somewhat of an overkill, that the clinical endpoint really achieves what we're really looking for.

DR. ADAMS: Walt?

DR. HAUCK: Well, I will start with what I'm going to call question 0, the one the committee has put on the table as to whether the clinical studies should be done at all. In my mind I've not heard enough to reach a judgment on that.

The problem is that what we have heard that is that particle size distribution matters, but what we haven't heard or seen data on is for two products that are Q1 equivalent, Q2 equivalent, equivalent devices, equivalent on this battery of in vitro tests, and equivalent on the systemic exposure, whether they could still differ enough in particle size distribution to matter. I don't know. Data probably doesn't exist on that, but that seems to be the relevant question, and I don't know that we've heard that. So, I'm going to restate the question and not try to answer it because you're also moving outside my expertise.

In regard to questions 1 and 2, what is the alternative? The alternative is to study more than one dose. It would seem sensible to say that you would want equivalence of the dose-response curves. It would not seem sensible, though -- let me add this -- the statistical significance of the dose-response curves is irrelevant for a bioequivalence study. The test and the reference could have equally flat dose-response curves, and that's equivalent and that should be fine.

I think that what we're hearing is that even the lowest recommended dose is still on the flat part or not far off of it, so that studying more than one dose then becomes a waste of time because you know you're on the flat part. And if you've shown it for the one dose, you've shown it for the rest of the doses.

So, from that perspective, my answer to questions 1 and 2 would be yes. I just don't know which of 1 of 2 I would answer yes to because there are really two parts to the question. Do we study one dose and then which design? And on the design issue, I haven't heard enough to make a decision on that.

My only concern would be that if a guidance is being developed based on the current state of products, that they need to have some sort of mechanism in place to worry about when it's no longer the case that the lowest recommended dose is right near the flat part of the dose-response curve. I mean, 5 years from now or 10 years from now, if a new body of products is coming up, that maybe that doesn't hold, you may want to rethink your choice of using a single dose.

DR. LEE: So, Walt, am I hearing that you're in favor of the --

DR. HAUCK: I'm ambivalent. Or no. I can't decide on the need for a clinical study. If there is a clinical study, then I say yes to one dose, and I've not heard enough to say for me whether or not it needs to be the 2-week study or the natural study or the EEU study. I don't think we heard enough. So, that's what I'm saying. I'm saying yes to 1 or 2. I just don't know which of the two I'm saying yes to.

DR. LEE: Okay. Maybe Dr. Roman is going to help us in that regard.

DR. ROMAN: Yes, and I hope I help Walter and I help the rest of you.

I know through doing the study and reviewing literature that the second question at the moment cannot be answered successfully. We do not know if the unit exposure or park study can give us any better answer than natural exposure 2-week studies, particularly for steroids. So, in terms of which study to choose, I will say, yes, natural exposure design, which is point 1.

Since we are not able to determine dose response with existing methodology, if you choose one, three, or five doses, 8 times difference, 16 times difference, and twofold difference, you will still, more or less, see a flat dose response as we saw so far.

I agree again with Dr. Hauck that maybe in the future when the methodology will be more sensitive to determine differences between the doses, we could go back to the dose-response question.

However, in my mind exactly as I asked Walter, what is the difference between particle sizes in low dose and high dose of the same product, when you study dose response for an innovator product, do you study particle sizes when you do such a study? Do you know that the low dose new steroids have similar particle size distribution or just a correlative number, or whatever you measure, as high dose? Do you correlate the particle sizes in different doses to the efficacy and safety? And this is obviously a question to pharmacologists, pharmacists.

DR. MEYER: I have to apologize because we were having a side bar in the earlier part of your discussion. But if you would repeat the question, I would appreciate it.

DR. ROMAN: Do we know what is the particle size distribution and contribution in the efficacy of low dose and high dose of the same innovator drug?

DR. MEYER: For the nasal products. I don't believe we do know, no.

DR. ROMAN: So, we're asking a very important question to be compared for generic and marketed product, but we don't know what is the contribution of particle sizes in even an innovator product to the dose response or efficacy.

DR. MEYER: Right. I think the place where we end up with the innovator product is we know what the particle size distribution looked like in the clinical trial batches and we know what they look like in the marketed batches. And we put some criteria on those to keep them within the same range, as much as feasible, so that hopefully we don't see a shift. But we don't know where in that distribution of particles the efficacy is coming from. If it is restricted, if it's broad, we don't know that, so we don't know what would be critically different if you were to go to a test product or a generic product that would be substitutable.

DR. ROMAN: So, as I say, medically it makes sense to do the study, and number 1 is the one I will choose. That's the opinion I have. Thank you.

DR. LEE: So, does the subcommittee know where we stand? We have two members who feel that this is overkill. We have one member who says it all depends, and then we have the others who feel that it's appropriate.

DR. HENDELES: If you are going to overkill, then I think number 1 is appropriate.

(Laughter.)

DR. LEE: Yes.

DR. ROMAN: Dr. Lee is looking at me so I will have first chance.

We know where we stand in terms of development of models for sensitivity for studying nasal allergy in terms of an endpoint, which is very subjective, symptomatic and all that. Is there at all the possibility to develop the comparison of the particle sizes? Isn't it more you attach something to it that you can compare A to B, or is the science so difficult that it is impossible to compare particle sizes or look for the methodology to do so? Or is it too naive a question?

DR. ADAMS: No, I don't think it's an unreasonable question. In fact, we are very interested in that issue, Dr. Roman, and we are giving it consideration in our own laboratories. But at the present time, we don't have a methodology which is validated.

DR. OWNBY: I've got a follow-up on a comment Dr. Meyer made and Dr. Adams just added to. It's my understanding you said that for innovator products, you looked at the particle size distribution in the test lots that were used for clinical studies, and then when they went into manufacturing, you also looked at the particle size distribution and said they were substantially the same. Therefore, they could go ahead with their distribution. Is that correct?

DR. MEYER: Let me be clear about that because I may have said things in an unclear manner. There I was talking primarily about the micronization of the drug substance before it's put into the formulation because the issue about a validated manner to assure the sameness of the particle size in the formulation itself holds no matter what product you are talking about.

But where an innovator firm can and does monitor the micronization process in their drug substance, a generic firm has no way of acquiring that data from the innovator product or from the reference product to assure that whatever micronization they achieve is the same as the micronization of their reference product. So, we're actually talking a step back. I'm talking about drug substance before it's formulated.

DR. OWNBY: Would you clarify? In my mind, it seems like we're talking about two different things. One is the particle of the steroid itself, the micronized steroid that's going to be put in the suspension, and the second thing is going to be the droplet size as that particle is delivered with the excipients and the vehicle around it from the nozzle of the device. Is that correct?

DR. MEYER: What we're focusing on today, though, is the particle size of the drug particles in the suspension. We feel like there are ways to characterize the droplet distribution. There are techniques for doing that, and I would leave it to my colleagues who know more about this than I to talk about that. But the unknown that we end up with at the end of the in vitro assessment is the particle size of the drug within the formulation itself, not so much the droplet size, plume geometry, those other things that we feel we can characterize.

DR. OWNBY: As a follow-up to that, when you talk about the particle size, you say that a secondary firm doesn't have access to the micronization statistics. Do you also think that once it's put into the formulation, that the micronized particles again begin to clump together so they change their size once they're in the formulation?

DR. MEYER: There probably is some agglomeration and aggregation of the particles, but because of the difficulties of assessing that in a suspension formulation with other particulate excipients, I think the degree of that may be difficult to know with surety.

DR. LEE: It seems to me that we should have scheduled someone to come and talk about suspensions.

DR. ADAMS: We may be able to have some additional thought on particle size distribution if Dr. Poochikin would like to make a comment. Let me ask. Guirag, would you be interested in making a comment with regard to this topic?

DR. POOCHIKIN: There were too many questions asked. I don't know where to start from.

As a matter of fact, with regard to particle size for these type of drug products, we're dealing with many aspects.

First, we're dealing with the particle sizing of the incoming drug substance, as Dr. Meyer and you were talking about.

Second, we're talking about we have to consider also particle size of the excipients. In this case, methyl celluloses, for example. Those are critical. They need to be controlled adequately.

Third, we have the particle size of the active in the formulation, as you were referring.

And fourth, we have the droplet size of the emitted spray.

On top of it, of course, we have a fifth complication which is the manufacturing aspects. Once we have the active and the excipient with appropriate particle size, will it create agglomeration? And that will depend on the environmental condition after formulation, but on top of it, it will also depend on the manufacturing procedures. So, that's a fifth complication that you will have to achieve similar spray from these type of drug suspension products.

If there are other questions on that issue, I can elaborate.

DR. OWNBY: To follow up on that, is there information about how much this affects the systemic availability of the drug or the local delivery of the drug?

DR. POOCHIKIN: As was discussed all day long, it depends on the particle size of the emitted or sprayed dose. That was discussed quite frequently this morning. That's the extent I can comment on that.

DR. ROMAN: Could you tell me then, since there are so many various things, variables, which you deal with when you produce intranasal steroids, since we are discussing them, that you could assure the characteristic of this delivery system from batch to batch for the innovator product from one formulation of a dose to another formulation of a different dose by the same company? So, are there enough guidelines, requirements, that they characterize the delivery of intranasal steroids?

DR. POOCHIKIN: Of the same product.

DR. ROMAN: Of the same product.

DR. POOCHIKIN: Yes, because as I indicated earlier, all those aspects that I discussed are monitored and controlled at the manufacturing level, including manufacturing itself. So, if you control everything that goes into the product, hopefully what comes out will be of the same quality.

But on top of it, of course, there are certain additional tasks which can be measured with regard to the pump or spray, as well as certain attributes of the drug product after it's manufactured, and of course, we have stabilities studies also. So, there are additional aspects that are being monitored.

Of course, you have to assume that it is always being done in the same environment by the same personnel, by the same experienced people. So, that's the assumption there.

DR. ROMAN: If you don't mind, this will be my last question. I won't say anything else. This was my question actually to you, Robert, and now I will ask it again.

So, what you are controlled is a new product when it's manufactured. During the clinical studies, when you do dose response for a new product, do you know if the particle sizes of a low dose formulation versus high dose formulation, when you do dose response, because they formulate different concentrations, are of the same size? Is this at all looked upon?

DR. MEYER: I think that we have the same technical difficulties assessing that for a nasal suspension in terms of comparing two different formulations from an innovator company that we would a test and reference in a generic sense. So, I guess in essence we really don't fully know that, but I think we are cognizant of the fact that alterations in the concentration and other aspects of the formulation can lead to different performances of those formulations.

One of the discussions earlier got to could we use even lower doses of the test or the reference. Well, in the generic world, you're not allowed, as a generic company, to manipulate your comparator product, number one. But number two, doing so may change the characteristics of that formulation anyway. So, we are cognizant of that being an issue, but we don't have strict assurance, no.

DR. ROMAN: Could it be, in part, responsible for this flatness of dose response? Theoretically I can imagine that if you load more product, you change particle sizes, don't you or not?

DR. MEYER: You might. I think it would confound the interpretation if you saw a dose response. I don't think it likely would lead, in my mind, to a flattening of the dose response. I suppose it could.

DR. CHOWDHURY: I just want to comment on that because for the dose-response studies many times is the same drug product which is used, but differing the number of sprays.

DR. ROMAN: But then you can only modulate two times, twofold, fourfold. When you got to 16-fold, it's a different concentration.

DR. CHOWDHURY: That's correct, but again, what you see for a flat dose response actually even goes for the lower like one-fold, two-fold changes.

DR. POOCHIKIN: With regard to different strengths of the same product, if I understood correctly, what Badrul was talking about, what Dr. Meyer was talking about was the same product given one spay versus two sprays versus four sprays and eight sprays, et cetera, as opposed to the same product manufactured at different strengths. As you know, there aren't that many nasal suspensions with different strengths. Those that are available, the concentrations are so low, I suspect if that will make any particle size distribution differences under the same conditions.

Having said that, of course, we have to consider the limitations of the test that we have. Based on the available data that we have on those very limited number of products, there isn't that much difference for different strengths of the same product because they are manufactured under the same conditions using the same spray, the same pump, the same excipient, the same drug substance.

DR. LEE: Dale, you wanted to make a comment?

DR. CONNER: One thing that impresses me about the last 10 or 15 minutes' discussion is we seem to be concentrating on strictly particle size like it is an end in and of itself. The only reason that particle size is important is because what it implies to what we're really interested in, which is are the two products, the resulting products, of which particle size is one component, therapeutically equivalent. Will they give the same therapeutic responses? So, particle size is simply something that contributes to that endpoint, which is what we're really interested in. We're saying, obviously, if we had our choice, the easiest thing is just to measure particle size directly, and if it's exactly the same, it essentially answers our worries about what could happen down the road about what we're really interested in.

In that we've said we can't do it directly, then we have no choice to jump over that and measure what we're really interested in anyway, which is therapeutic equivalence. That's what we're proposing to do, to say now we're kind of going a little bit further down the chain actually measuring what's happening presumably in the patient and assuring that it's the same. When we do that effectively, we say it really doesn't matter if there's a difference in particle size. I'm measuring what I'm actually interested in, the endpoint.

It's not the other way around, that the particle size is the end that I'm interested in. We were only interested in particle size because it has some implication on what we really want to know. So, we've just simply jumped over it and said, well, we're going to measure what we want to know more directly. Maybe there's no particle size difference at all. Maybe it's huge, but I've shown that it doesn't matter.

DR. ROMAN: Yes, but this is the only thing which we cannot characterize properly. I understand that everything else can be done in vitro with the exception of this one. So, if you have the same active moiety and the only difference is particle size, so we are doing this clinical study just because we cannot characterize this particular aspect --

DR. CONNER: If some brilliant person -- and many have tried over the past few years unsuccessfully -- if we could get a very convincing, validated measure of particle size and be able to compare it in the finished product, we may not need to do this. We were hoping that that effort would be successful and that we could not have to do this clinical trial, that we would have the last piece of the puzzle. We don't have that right now. So, we're forced to confirm it through other means by looking at the endpoint of what we're really interested in rather than measuring that particular factor directly.

DR. LEE: Okay. I think that we are going around and around expressing our discomfort about our respective positions.

Let me state, in all fairness -- yes, Wally.

DR. ADAMS: Yes. I think listening to each of the subcommittee members, I think I hear what the recommendations are or the feeling of the various individuals. I'm wondering if we could, however, put this on the record that in Dr. Meyer's presentation, he talked about the rhinitis study as being either confirmatory or pivotal in terms of the bioequivalence assessment. I would be very interested in hearing specifically that issue addressed. I think it has been said in so many words, but if we could have that on the record as to how the people feel about that, it would be helpful. I think that also gets to the issue of one dose versus two doses.

DR. LEE: This is exactly what I'm about to say. Thank you very much for framing the statement.

Wally stated a point about your feeling about confirmatory versus pivotal. Should I start with Dr. Roman?

DR. ROMAN: I always have an opinion to share, so let me share my opinion with you.

The way I understood Dr. Meyer's presentation of this pivotal versus confirmatory is that their situation, even at present, is that you cannot do pharmacokinetic comparison. That is, the drug levels in blood are so low or the test method to measure drug level is so insensitive -- whatever, we don't know any better -- that therefore the complete profile of blood level is not possible to determine.

In this case, you don't have any in vivo data for bioequivalence, and the only one you have is clinical study. In this situation, I will call it pivotal.

If on the other hand, you have a sensitive method to determine the pharmacokinetic profile of the product, that for me would be only confirmatory because of this particle size situation.

So, it can be defined different ways or with a different definition for a different situation.

Now, I would like to hear, because I think that I also understood from Dr. Meyer that if it is pivotal, he would like to see dose response, and he would be willing to agree with one dose for a confirmatory study. This is what I was trying to ask you actually immediately after your presentation.

DR. MEYER: I think from my own viewpoint that if one is asking the study to establish bioequivalence and that's going to be your primary basis for that, then I think you need to show sensitivity to the effects of dose so that you can assess what the meaning is of any different scene between the test and the reference in the clinical study.

DR. ROMAN: And here, of course, we share the understanding that with the available methodology and sensitivity of the methodology, it's almost impossible, and with the limitation of having an already formulated product so the fold difference could be no more than four in terms of different dose levels.

DR. LEE: Walt?

DR. HAUCK: Well, we spent the afternoon pretty much trashing the clinical study. So, given that we've got a clinical study now that at least, I would say, about half the committee views is of uncertain value, confirmatory would be the best we could call it, certainly not pivotal.

On a different level, if we were to set a principle of calling the clinical study pivotal, which would almost be reopening all the bioequivalence guidances that we've got, so as far as I'm concerned, the entire tenor of all the bioequivalence approaches is the clinical approaches are not pivotal. Everything is trying to find an alternative to doing a clinical trial.

DR. SHARGEL: I would go along with confirmatory, looking at what Dr. Meyer has here. I think the whole idea of bioequivalence studies in any regard, the final analysis or assumption of a bioequivalent test, if you do any other kind of thing, is that predictably it's going to have the same clinical effect. Therefore, if you put it in a clinical situation, it has the same effect, that's the endpoint that you're looking for in terms of substitution of a generic brand or different lots of the brand or different lots of the generic. So, I would tend to think that this would confirm the fact that both products behaved in a clinical situation in a similar fashion within perhaps statistical parameters. Whether it has slight differences in bioavailability I'm not sure you'd get from that kind of study. Therefore, I don't look at it as pivotal.

DR. LEE: Thank you.

Gloria? Richard?

DR. AHRENS: I'd agree that it's confirmatory without additional comment. I think it has all been said.

DR. OWNBY: I would agree that confirmatory is the best I think we could reasonably expect in this circumstance.

DR. DYKEWICZ: Confirmatory.

DR. HENDELES: Ditto.

DR. LEE: There's a consensus on confirmatory I suppose.

I think what we would like to do is to propose a break so that we can clear our heads. Then we'll come back and once again we address this issue. But I'd like to come a little bit to some kind of a consensus about how we feel about the first two questions posed.

Yes, Wally.

DR. ADAMS: Just one point for clarity to Dr. Roman. You indicated that provided the plasma levels could not be measured, the rhinitis study would be pivotal. But could that be broadened to say that whether a PK study or an adrenal axis suspension testing were not possible, that you would view it as pivotal?

DR. ROMAN: You don't have anything else then. If you cannot do plasma levels, so you're really comparing based on all in vitro work, and therefore I dare to call it pivotal, but I don't know if those responses are doable as a pivotal. Maybe it will be defined pivotal for efficacy, this one dose level study, and then we should discuss the safety study, which is a different issue. What we were touching on this morning is that the low effective dose, which I assume will be close to the labeled dose or will be from an efficacy standpoint, and then some kind of a high dose for systemic safety.

DR. ADAMS: So, if either the high dose PK or the high dose adrenal axis suspension testing could be done, then the rhinitis study is confirmatory. Is that what you're saying or not?

DR. ROMAN: Let me repeat it. If PK cannot be done and HPA axis cannot be done --

DR. ADAMS: Can be done, then would you view the rhinitis study as confirmatory?

DR. ROMAN: Right.

DR. ADAMS: Okay, good. Thank you.

DR. LEE: Wally, let me ask you a question before we break so that we can really think about it. Is it fair to say that these two questions would not have been posed if we had access to information about the particle size distribution?

DR. ADAMS: That's a very interesting question, Vince.

(Laughter.)

DR. ADAMS: I would deflect that question to say that should at some time a validated particle size distribution method become available, then we will take that to our working group and discuss it. It's not an issue that we've had the luxury of addressing at this time.

DR. LEE: All right. I think it is time for a time-out. Let's say that we come back at about -- would 2:45 be too generous?

DR. ROMAN: It would be right.

(Laughter.)

DR. LEE: Okay. 2:45, and we come back and address those two questions one more time. Thank you.

(Recess.)

DR. LEE: So, we have to reconvene. I hope that this time-out was helpful to everybody.

Let me try to bring some focus to this concluding session, the final session. I promise that we will be done before 4:00 p.m. today.

We have two questions posed to us, and let me start out by saying that if we were to address one of these two scenarios, which one would you prefer? Would you have much confidence in the EEU and the park study? Dr. Roman?

DR. ROMAN: Yes, again, the question is number one point or number two, a clinical study of 2 weeks' duration at least, et cetera versus EEU or park study. My answer is number one, a classical or traditional exposure study.

DR. HAUCK: I'm going to go along with whatever the rest of the committee decides on that.

(Laughter.)

DR. HAUCK: I'm getting my arm twisted. It's number one.

(Laughter.)

DR. SHARGEL: I'm twisting his arm also. I go along with number one. I think it confirms that both products have similar clinical endpoints.

DR. LEE: Thank you.

Gloria?

DR. ANDERSON: Number one.

DR. AHRENS: Number one.

DR. OWNBY: I'll go with number one also. I hope it's the right door.

(Laughter.)

DR. DYKEWICZ: Number one.

DR. HENDELES: I would go with number one also unless somebody came up with an innovation on number two that allowed you to look at like an inhaled steroid with its onset of action.

DR. LEE: Okay. I'm a bit worried because you're the one who has to leave at 4:00.

So, there seems to be some consensus developing for number one. Is that right?

DR. ROMAN: Number one at the moment.

DR. LEE: And the follow-up question is are you comfortable with the statement made.

DR. HENDELES: What do you mean?

DR. LEE: The lowest active dose is sufficient to confirm equivalent local delivery of a suspension formulation intended for allergic rhinitis.

DR. SHARGEL: May I make a comment on that? In terms of lowest effective dose, if we're dealing with a bioequivalent product of another manufacturer that's coming out, it would be the lowest effective dose or the lowest dose that's on the label. Indication. Is that what we're talking about, not necessarily the lowest effective dose, but the dose that the manufacturer of the brand has already established? Are we distinguishing between lowest effective dose, which means I have to find the lowest effective dose, or the dose that the manufacturer has indicated on the label?

DR. LEE: What I'd like to do is to take the question as stated and then you can agree, disagree. And if you disagree, please propose an alternative.

DR. SHARGEL: Well, I would like to propose the alternative being the lowest dose that the innovator has proposed.

DR. ADAMS: Lowest labeled dose.

DR. SHARGEL: The regulators will give me the proper term, but the lowest labeled dose, whatever appears on the insert.

DR. LEE: Walt?

DR. HAUCK: I'm generally okay. The statement "confirm" really seems overly strong for me given the nature of the study. We were given a choice of confirmatory versus pivotal, and there probably should have been a third choice there. So, with the caveat that all we're really doing is saying there's been an opportunity to find a large difference in particle size and whatever else and we didn't find it, and if that's what's meant by confirmatory, then it's okay as worded. Otherwise, I would work on that wording.

DR. LEE: Izabela?

DR. ROMAN: I would agree that if the label states the dose range, it should be the lowest approved dose. The problem I have is what if there is no dose range in the labeling but a dose or the dose. Will we then accept that this is the optimal dose?

DR. LEE: Thank you.

Les?

DR. HENDELES: I have a small problem with the wording. It says "confirm equivalent local delivery," and I don't think you can do that. I guess maybe if you replace the word "equivalent" with "to confirm therapeutic comparability" or something like that. What you're doing is confirming therapeutic comparability, but not local delivery because it's possible that you may deliver half as much drug and not detect that difference.

DR. ADAMS: Les, we should have had you help write the question. That thought hadn't occurred to us. You're absolutely right.

DR. HAUCK: Let me add that was the better stating of what I was trying to get to.

DR. DYKEWICZ: And that was stating what I was about to state. I don't know if it's an incorrect semantic approach of saying "to confirm bioequivalent local delivery" or that's really a misappropriation of the term "bioequivalence," but that's exactly the concern, that you may not be getting similar delivery but it's becoming equivalent in terms of clinical outcome.

DR. LEE: Dennis?

DR. OWNBY: Yes. I agree that we're really talking about therapeutic endpoints. That's the only thing you can say is equivalent.

DR. AHRENS: I wasn't about to say Les' comment in a different way, but I agree with it. And with that said, the lowest labeled dose.

DR. LEE: Gloria?

DR. ANDERSON: I agree as well. I still have problems with "believe."

DR. LEE: You don't believe?

DR. ANDERSON: I don't think I have enough information to say I believe that. I just wish you could rewrite that word.

DR. HENDELES: And this committee member does not believe.

(Laughter.)

DR. ROMAN: Dr. Conner came with this nice definition of therapeutic equivalence, which is what we are studying in a clinical study, therapeutic equivalence.

DR. MEYER: If it makes people feel more comfortable, I think what I had tried to say in my talk -- and perhaps we should have chosen wording to better reflect this -- is that where we are with a confirmatory trial is really saying that given whatever unknowns may still exist from all the other data that get us to that point, we're confirming that those differences don't matter clinically with this confirmatory trial; that therapeutically whatever unknowns remain from all the in vitro parameters, from the PK being the same, everything else being the same, the unknowns, such as particle size distribution, in the formulation don't matter clinically. So, it's not the establishment in a bioequivalence. Perhaps this is overworded. It's really just to say that whatever differences might remain or whatever we don't know about, we've taken to a clinical trial and we've not seen an important difference.

DR. LEE: Very well. So, we throw out question number two, and we all feel comfortable about the 2-week study. There's some discomfort about the wording, and we all propose the writer of that question ought to be sent to English school.

(Laughter.)

DR. LEE: I think that he got the gist of how we felt about the wording, and he will come up with different wording for the public record.

DR. ADAMS: Dr. Lee, I would point out with regard to the wording of that question --

(Laughter.)

DR. ADAMS: -- somehow our sign-off initials were down in the lower left of that page, and so you can see that that was not a single person's effort.

(Laughter.)

DR. ADAMS: It managed to get past a number of individuals. But I certainly take the comments which have been made around the table and completely agree with them with regard to the wording of the question.

In view of the fact that two days from now on Thursday there's to be a report of the OINDP Subcommittee to the full ACPS, and it is a short time period and we will not have the transcript in order to make sure that this report is accurate. Would we be able to present our summary bullets for you and the committee to consider and make sure that we have these right and make any changes to them? If we could just spend a few minutes doing that.

DR. LEE: Sure. Who will be doing that?

DR. ADAMS: That would be Dr. Singh.

DR. LEE: Dr. Singh, please be to the point.

DR. SINGH: I was given the task of keeping quiet till the end and then give my perception in the form of a couple of conclusions. I'll put three or four statements, as you said, to the point, and these statements are statements only relevant to the main issues discussed this morning and partly this afternoon.

This has been a key issue although it's not directly stated in the two questions that Wally put up and the two questions that we just finished discussing.

The key issue in my opinion has been demonstration of the dose response with the difficulties associated with it. What I've heard is that based on the current technology and methods, the demonstration of dose response may not be possible at this moment. That's the number one conclusion I made, and it's open for the committee's comments, Dr. Lee.

DR. LEE: Is that what the subcommittee says? They all agree. Good job.

DR. HAUCK: Well, with a caveat. At least part of my comment was it's also irrelevant for the bioequivalence context whether you can demonstrate a dose response or not.

DR. SINGH: Yes. I think what we heard from Dr. Chowdhury and Dr. Meyer and others is that whether it's with regard to the bioequivalence determination or bioavailability, thus far we cannot determine dose response for at least these steroid formulations. Am I right in that, Dr. Chowdhury?

DR. CHOWDHURY: Yes, you are right on that.

DR. SINGH: Thank you.

DR. MEYER: I just wanted to comment on what Dr. Hauck just said that it is not material to the actual determination of bioequivalence, but the use of it is actually to establish the study could have detected a difference if a difference existed. It's to show assay sensitivity in the study, and then you can do your bioequivalence based on a determination of how the two single doses relate.

DR. HAUCK: That's where your placebo comes in and why I asked about the placebo earlier this morning. It's the fact that you have the placebo control which is giving your assay sensitivity.

DR. MEYER: I actually don't agree with that because if you have a binary answer where 32 micrograms looks no different from 256, and you are studying, say, 128 micrograms, the failure to show a difference in that doesn't suggest that no difference exists or doesn't establish that no difference exists.

You can show a difference from placebo with a corticosteroid with fair regularity. There are very few well-done trials that fail with a steroid, but they act very binary. There's either an effect or there's not. So, you need to show sensitivity to dose. It's not just whether there's sensitivity to active versus placebo; you have to show sensitivity to dose if you wanted to establish bioequivalence in a pivotal setting.

DR. SINGH: Then the committee went into discussions like there were some opinions expressed, do we really need a clinical study or in vitro and PK can be enough. Dr. Hendeles took the lead on that. I think what I heard the committee say, based on that initiative by Dr. Ahrens, is yes, under circumstances a clinical study is needed to establish equivalence of suspension nasal products. Any comments on that?

DR. SHARGEL: I have one comment on that because I think we wound up saying that it was confirmatory to the bioequivalence rather than establishing equivalence, which is a little different.

DR. SINGH: I'm coming to that point. This is in a slide later on.

DR. HAUCK: Wait a minute. I'm not sure I agree with this statement. At best, I thought this was an open question. I can understand that there's enough information to make this a question as to whether there's a need for the study, but for me there wasn't enough information to -- I was going to say confirm --

(Laughter.)

DR. HAUCK: -- to conclude that the clinical study was going to add anything in the context that it's being used, that is, on top of all the other equivalence studies. So, I don't buy this for me. Maybe it's a consensus, if not unanimous, for the committee.

DR. LEE: I think the majority of the subcommittee felt that a clinical study would be necessary.

DR. SINGH: That's what my perception was. I think the difficulty may be with regard to the word equivalence here, which I'm coming to later on whether it's confirmatory.

DR. MEYER: Right, but if we're trying to capture the sense of the committee, I think we would have to reflect perhaps that the majority felt a clinical study would be useful in the examination of --

DR. HENDELES: But not establishing equivalence.

DR. MEYER: Not establishing equivalence, but useful as a part of the comparison between a generic and a reference product for nasal suspension sprays.

DR. SINGH: That's right.

DR. ADAMS: Gur Jai Pal, could you just modify that sentence, because as we move through this, if we could get the wording correct, as we go through it, it will be very helpful.

DR. ROMAN: If you would add to establish, as it was stated, therapeutic equivalence, would it be sufficient?

DR. SHARGEL: I think that the clinical study doesn't actually establish bioequivalence as a pivotal equivalence. We're taking the total data and submission, which includes a lot of in vitro measurements, and this clinical study only confirms that there's no difference between the products as opposed to looking at rate, extent, and those kinds of parameters. So, as it was written, I don't agree with that. I'm not sure I see the need for a clinical study, but if we were doing a clinical study, it's basically for a confirmation that there is no difference between the products, that whatever was found in vitro in terms of particle size and characteristics of plume geometry and all the good stuff that's in there, we are confirming it with the clinical study.

DR. SINGH: I think all my comments reflect is what Dr. Lee said, the majority of the committee said that some kind of clinical study is necessary.

Now I deleted two words "establish bioequivalence," and I just put "to compare suspension nasal products," whatever manner we want to compare it. Wally, do you want to modify it further?

DR. MEYER: I thought perhaps the committee might be more comfortable with the word "useful" rather than "needed," a clinical study is "useful" in the comparison of nasal suspension products. I don't want to put words in anybody's mouth, though, because I know some people were more definitive about this than others and that's softening the language.

DR. ADAMS: Does the subcommittee feel comfortable with a clinical study is "useful"?

DR. SHARGEL: I disagree with any term that just says "useful." There's a lot of stuff that's useful. I know that when we submit applications, I get feedback from the regulatory people that things are recommended as useful. I think we really need to know whether it's needed scientifically.

DR. HENDELES: As I understand it, this clinical study is going to confirm that there are no problems with this product. So, something along that line I think was the wording you used. So, instead of useful, it's going to confirm that there are no apparent bioequivalence problems. How about that?

DR. ADAMS: So, Les, are you saying that a clinical study is needed to confirm?

DR. HENDELES: I'm not saying that.

(Laughter.)

DR. HENDELES: It's still hard for me to understand how you can put a bolus of this stuff in somebody's nose at the site of action and question whether it's going to work or not.

DR. MEYER: Just to help, because I think we're going to get to some of what you were just saying, Les, in the later summary of slides. I think the crux of this is does the committee feel like there is a role for a clinical study in what we're proposing or not. My sense is that there was a majority of opinion that there was but not a consensus, if you use consensus to be synonymous with a unanimous agreement. I think we can reflect that to the full committee.

DR. SINGH: Okay. The next bullet is with regard to the two types of studies that were put up there in question two, what I heard was for study number one, which was a placebo-controlled traditional 2-week rhinitis study may be appropriate for whatever we want to achieve here, and a lot has been said about that. And the dose that should be studied where the generic and the reference product should be compared is the lowest active dose and that is the lowest dose in the label. If there's only one dose, then that's the dose, as it was said.

DR. ADAMS: Gur Jai Pal, that should read -- I think what I heard was lowest labeled dose is the feeling of the committee.

DR. SINGH: I have that in the parentheses, "label."

DR. ADAMS: Strike the word "active."

DR. SINGH: Okay.

DR. ADAMS: We're trying to get this wording so that it correctly reflects at the present time before we wrap up today. So, the "lowest labeled dose."

DR. SINGH: And the last one is what is the significance of the rhinitis study, the clinical study, and I think we heard that it's not a pivotal study. It's a confirmatory study.

And those were my bullets.

DR. LEE: Dr. Ownby, you have a question about the previous slides?

DR. OWNBY: Yes, could we have the previous one back up? I thought this was a comparative study between two active drugs and not a placebo-controlled trial, or did we decide that a placebo arm was needed with the two active?

DR. ADAMS: Yes, it is a placebo-controlled study, as reflected by our guidance, and the intention here is that this be a placebo-controlled study.

DR. DYKEWICZ: My only other comment is about "may be appropriate." I think we were coming to some sort of a consensus or opinion that that was the most appropriate of the various types of studies that were being proposed.

DR. MEYER: Maybe the correct way for us to word this later is if a clinical study is done, a placebo- controlled, traditional 2-week rhinitis study is appropriate or is the most appropriate.

DR. DYKEWICZ: I like that.

DR. LEE: Yes, Walt.

DR. HAUCK: On the last transparency you had, Gur.

DR. SINGH: Okay, this was the last one. It's not a pivotal study; it's confirmatory.

DR. HAUCK: The committee did take a vote on that, but then there was a later phrasing by Dr. Meyer that actually expressed both the purpose and what was being accomplished with this clinical study. I'd rather see you capture that than to have confirmatory, which has all sorts of common language meanings in addition to whatever specific meaning we may or may not have agreed on today. So, I would find the single word potentially very misleading, and Dr. Meyer's summary would be something I would be very happy with as a statement of the purpose and significance of the rhinitis study.

Should I summarize so everybody is clear about what I'm referring to? It was the notion that there were potentially many unknowns, of which particle size distribution was one, and this is just ensuring that whatever the impact of those unknowns were, that they were not clinically important.

DR. ADAMS: Gur Jai Pal, I took a note when Bob said that. That statement could read that the rhinitis study is useful to confirm whatever differences exist don't matter clinically. Whatever unknowns remain don't matter clinically.

DR. OWNBY: Is it better to say don't matter clinically or don't materially affect the clinical outcome?

DR. MEYER: Whatever wording you're comfortable with. I actually like the way Dr. Hauck said it and the way Wally captured what I said probably better than what I said.

(Laughter.)

DR. HAUCK: I think I said "was not important clinically," something like that, which is closer to your phrasing rather than it does not matter. It should be an equivalence phrasing rather than a no-effect statement. You could say "was not clinically important" or "had no material clinical effect," either of those.

DR. SINGH: I think, Walter, that when we put it in writing --

DR. ADAMS: Well, if we could just make that statement. To capture it now would be helpful. "Are not important clinically." Walter, was that what you said?

DR. HAUCK: That's all right with me.

DR. LEE: Wally, may I make a proposal?

DR. ADAMS: Please.

DR. ADAMS: Everyone is going home except Gloria and myself. Would it be possible to have these written and e-mailed to the subcommittee and give everybody a day to respond so that when we do have to report to the full committee on Thursday morning, that we have at least some sense for the degree of enthusiasm for the wording?

DR. ADAMS: I think that's an excellent idea. So, you're proposing that sometime by maybe tomorrow morning --

DR. LEE: Yes. Somebody get busy tonight and start writing.

DR. SINGH: Yes, I think we can handle that.

DR. MEYER: I would just make the point that as somebody who's contributed in virtual working groups by e-mail, the one problem with doing that is that if people do minor wordsmithing, you end up in a position where you don't have time to then go through with reiterations and so on. So, people would just need to confine their comments to important points and maybe really focus on the substance and not have specific wording recommendations, but the substance of what we would need to change back so we can integrate and come up with something that's satisfactory to all.

DR. LEE: The other alternative is to empower me to speak on their behalf.

DR. HAUCK: I'm willing to empower you.

(Laughter.)

DR. HAUCK: Not that we could stop you anyway.

(Laughter.)

DR. LEE: Don't forget, English is not my native language.

(Laughter.)

DR. LEE: Any other comments, questions?

I understand that some of you might be on vacation. For those of you on vacation, please do not feel compelled to respond, but in case you do, we obviously would welcome your input.

Anything else? Wally, do you have enough to move on to the next phase?

DR. ADAMS: What is that phase?

(Laughter.)

DR. LEE: You should know. To finish the guidance.

DR. ADAMS: I'd like to ask Ms. Winkle if she has any comments to add.

MS. WINKLE: No. I just want to thank everybody today for their participation in the discussion. I think it will be extremely useful in helping us in CDER to finalize this guidance and to be able to capture exactly what we need to ensure equivalence. So, I appreciate everyone's attention to these issues and look forward to coming out with a guidance soon that is able to directly talk to these issues. Thank you.

DR. ADAMS: Yes, I also would like to reflect Helen's comments, that we're very appreciative to the subcommittee for their willingness to come on short notice I guess and participate in these deliberations. It moves the draft guidance one step further along towards the process. So, we're very appreciative of that and thank you very much to each of you for your expertise in this matter.

DR. LEE: Okay. Thank you very much. Is there any further business?

(No response.)

DR. LEE: A move for adjournment?

DR. DYKEWICZ: So moved.

DR. LEE: So moved.

Second?

DR. HENDELES: Second.

DR. LEE: Let's go. Thank you very much.

(Whereupon, at 3:20 p.m., the subcommittee was adjourned.)