1

DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

CENTER FOR DRUG EVALUATION AND RESEARCH

PERIPHERAL AND CENTRAL NERVOUS SYSTEM DRUGS

ADVISORY COMMITTEE

NDA21-645, MT 100, (naproxen sodium

and metoclopramide hydrochloride)

Tablets, Pozen, Inc. for the proposed

indication of acute treatment of migraine

headache with or without aura

Thursday, August 4, 2004

8:00 a.m.

Advisors and Consultants Staff Conference Room

5630 Fishers Lane

Rockville, Maryland

PARTICIPANTS

Karl Kieburtz, M.D., MPH, Acting Chair

Anuja Patel, MPH, Executive Secretary

MEMBERS

Michael D. Hughes, Ph.D.

Carol L. Koski, M.D.

Roger J. Porter, M.D., (Industry Rep, Non-Voting)

Ralph L. Sacco, M.D., M.S.

SPECIAL GOVERNMENT EMPLOYEES (VOTING)

Larry B. Goldstein, M.D.

Lily K. Jung, M.D., MMM (Acting Consumer Rep)

Stanley Fahn, M.D.

Marc E. Lenaerts, M.D.

K. Michael Welch, M.B., Ch.B., FRCP

Sheila Weiss-Smith, Ph.D., FISPE

Mark W. Green, M.D.

FEDERAL GOVERNMENT EMPLOYEE CONSULTANT (Voting)

Dilip V. Jeste, M.D.

FDA

Robert Temple, M.D.

Russell Katz, M.D.

Eric Bastings, M.D.

Mary Ross Southworth, Pharm.D.

C O N T E N T S

PAGE

Call to Order and Opening Remarks:

Karl Kieburtz, M.D., MPH 5

Anuja Patel, MPH

Conflict of Interest Statement:

Mary Ann Killian 7

Overview of Issues:

Russell Katz, M.D. 17

Sponsor Presentation, Pozen Incorporated

Introduction and Summary:

Marshall E. Reese, Ph.D. 24

Overview of Tardive Dyskinesia:

A.H.V. (Tony) Schapira, M.D. 35

Review of MT100 Efficacy:

William James Alexander, M.D., MPH, FACP 61

Potential Role of MT100 in Migraine Therapy:

Balancing Benefits and Risks:

David B. Matchar, M.D., FACP 78

Clinical Considerations on Migraine Treatment:

Stephen D. Silberstein, M.D. 99

FDA Presentation

FDA Risk/Benefit Considerations:

Eric Bastings, M.D. 108

Overview of Tardive Dyskinesia:

Hyder A. Jinnah, M.D., Ph.D. 127

Post-marketing Review of Movement Disorders

and Neuroleptic Malignant Syndrome Associated

with Metoclopramide:

Mary Ross Southworth, Pharm.D. 138

Questions from the Committee to the Sponsor

and to FDA 160

C O N T E N T S

PAGE

Open Public Hearing

Cynthia McCormick, M.D. 217

Committee Discussion and Response to FDA Questions 225

P R O C E E D I N G S

Call to Order and Opening Remarks

DR. KIEBURTZ: Good morning. This is the

Peripheral and Central Nervous System Drugs

Advisory Committee. We here to discuss the New

Drug Application 21-645, proposed trade name of

MT100 Tablets, from Pozen, Incorporated for the

proposed indication of acute treatment of migraine

headache with or without aura.

I would just also take the opportunity to

refer people to the agenda. Incorporated in the

agenda are the questions which are posed to this

committee which will be discussing and voting on

today. We won't be discussing or voting on prior

actions of the FDA including the non-approvable

letter or any issues about a approvability of the

product. That is not our remit or discussion for

today.

So I would hope that the presentations are

focused on what the committee will be discussing

and deliberating about today.

When people speak, please speak into the

microphone and turn the microphone on. If you are

interested in speaking, raise your hand or you can

turn the microphone on. Anuja, the Executive

Secretary, will keep track and will get to people

who want to speak from the committee.

Just to the committee members, the voting

committee members, please keep in mind that, to

vote, you need to be here. So there is no leaving

of votes. Hopefully, there is no leaving until the

meeting is adjourned which is scheduled to be at 5

o'clock. Please plan your travels accordingly.

In a second, I am going to introduce Mary

Ann Killian. There is a new procedure--I think we

are the inaugural run of it--for disclosure of

conflicts of interest where Mary Ann Killian reads

a statement and then each individual member of the

committee reads their conflict statement. When

that is concluded, Mary Ann has some concluding

remarks and then we will move on with the rest of

the agenda.

The only individual who does report

conflicts of interest is Dr. Porter as he is the

industry representative.

So Mary Ann Killian, the Program Integrity

Advisor from the Ethics and Integrity Staff.

Conflict of Interest Statement

MS. KILLIAN: Thank you very much. The

FDA is convening today's meeting of the Peripheral

and Central Nervous System Drugs Advisory Committee

under the authority of the Federal Advisory

Committee Act of 1972. The advisory committee

meeting provides transparency into the agency's

deliberative processes.

With the exception of the industry

representative, all members of the committee are

special government employees or regular federal

employees from other agencies and are subject to

federal conflict-of-interest laws and regulations.

Consequently, in the interest of

transparency and the spirit of disclosure, the

following information on the status of this

advisory committee's compliance with federal ethics

and conflict-of-interest laws covered by but not

limited to those found at 18 U.S.C. 208 and 21

U.S.C. 355(n)(4) is being provided to the

participants in today's meeting and to the public.

FDA has determined that members of this

advisory committee are in compliance with the

Federal ethics and conflict-of-interest laws

including but not limited to 18 U.S.C. 208 and 21

U.S.C. 355 (n)(4). Under 18 U.S.C. Section 208,

applicable to all government agencies, and 21

U.S.C. 355(n)(4), applicable to FDA, Congress has

authorized FDA to grant waivers to special

government employees who have limited financial

conflicts when it is determined that the agency's

need for a particular individual's services

outweighs his or her potential financial conflict

of interest.

Members who are special government

employees at today's meeting, including special

government employees appointed as temporary voting

members, have been screened for potential financial

conflicts of interest of their own as well as those

imputed to them including those of their employer,

spouse, minor child related to the discussions of

today's meeting. These interested may include

investments, consulting, expert witness testimony,

contracts/grants/CRADAs, teaching/speaking/writing,

patents and royalties, and primary employment.

Today's agenda involves a review of New

Drug Application 21-645, proposed trade MT100

Tablets, proposed for acute treatment of migraine

headache with our without aura sponsored by Pozen,

Inc. MT100 is a combination of two approved drugs,

naproxen sodium, manufactured by the Albemarle

Corporation, and metoclopramide hydrochloride,

manufactured by Cosam S.p.A, a member of the CFM

group. This is a particular matters meeting during

which specific matters related to the NDA will be

discussed

Copies of each acknowledgement and

consent-to-disclosure statement signed by each

participant at today's meeting who received a

conflict-of-interest waiver along with this

statement will be available for review at the

registration table during this meeting and will be

included as part of the official meeting

transcript.

A copy of the written conflict-of-interest

waiver statements may be obtained by submitting a

written request to the agency's Freedom of

Information Office, Room 12A-30, of the Parklawn

Building.

At this time, each member will be asked to

state his or her name for the record and announce

whether his or her participation in this meeting is

based on a conflict-of-interest waiver.

Please state your name and whether you

have received a waiver from the agency to

participate in today's meeting. If you have

received a waiver, please describe the details of

the interest or interests for which the waiver has

been granted. If the agency has reviewed your

reported interest and determined that you do not

require a waiver, please indicate that for the

record.

I guess we will start with you.

DR. KIEBURTZ: I will be the exemplar. I

am Dr. Karl Kieburtz. I am a neurologist and on

the faculty of the University of Rochester in

Rochester, New York. Based on the agenda for

today's meeting and the information regarding my

financial and other interests required to be

reported to the agency prior to my participation

today as a committee member, I have not received a

conflict-of-interest waiver to participate in

today's meeting. That means I don't need a waiver.

Next is Dr. Porter.

DR. PORTER: Pass.

DR. KIEBURTZ: If you would just introduce

yourself for the record.

DR. PORTER: Sure. I am Roger Porter. I

am a neurologist twenty years at the NIH, ten years

at Wyeth. I am now a consultant.

DR. HUGHES: I am Michael Hughes. I am

Professor of Biostatistics from Harvard University.

Based on the agenda for today's meeting and the

information regarding my financial and other

interests required to be reported to the agency

prior to my participation today as a committee

member, I have not received a conflict-of-interest

waiver to participate in today's meeting.

DR. KOSKI: I am Dr. Carol L. Koski. I am

a Professor of Neurology at the University of

Maryland School of Medicine. I have received a

waiver for ownership of stock in two competing

firms. The first is valued between $5,001 and

$25,000. The second is valued between $25,001 and

$50,000.

DR. SACCO: Hi. Ralph Sacco. I am a

Professor of Neurology and Epidemiology and

Director of Stroke and Critical Care at Columbia

University. I have received a waiver for my

service as a consultant for a competing firm. I

also serve on the Data and Safety Monitoring Board

for a competing firm and I receive less than

$10,001 per year from each firm.

DR. GOLDSTEIN: I am Dr. Larry Goldstein.

I am a Professor of Medicine at Duke University and

Director of the Duke Center for Cerebrovascular

Disease. I have received a waiver for consulting

for four competing firms and I receive less than

$10,001 per firm per year from three of the firms

and between $10,001 and $50,000 per year from the

fourth firm. In addition, I serve as a member of

two advisory boards and two steering committees for

competing firms and receive less than $10,001 per

year from each firm.

DR. JUNG: Hi. My name is Lily Jung. I

am a neurologist with the Seattle Neural Science

Institute and Swedish Medical Center. I am also a

Clinical Associate Professor at the University of

Washington. I have received a waiver for ownership

of stock valued from $5,001 to $25,000 in a

competing firm.

DR. FAHN: Good morning. I am Dr. Stanley

Fahn. I am a Professor of Neurology at Columbia

University subspecializing in the field of movement

disorders. I have received a waiver for serving on

steering committees for two competing firms. In

addition, I also serve as a consultant for two

competing firms. I receive less than $10,001 per

year from each firm.

DR. LENAERTS: Good morning. Marc

Lenaerts, Assistant Professor, University of

Oklahoma, Department of Neurology, a headache

specialist. I have received a waiver for serving

on three speakers bureaus. One is between $10,001

and $50,000, and two are $10,000 or less.

DR. WELCH: Good morning. I am Dr.

Michael Welch. I am a Professor of Neurology at

Rosalind Franklin University of Medicine and

Science. I have received a waiver for serving as a

consultant for two competing firms and I am also an

advisory board member for two competing firms and

serve on the steering committee for a competing

firm. I receive less than $10,001 per year for

each firm.

DR. SMITH: Good morning. I am Professor

Sheila Weiss Smith. I am an Associate Professor at

the University of Maryland Schools of Pharmacy and

Medicine. I have not received a

conflict-of-interest waiver to participate in

today's meeting.

DR. JESTE: Good morning. I am Dr. Dilip

Jeste. I am Professor of Psychiatry and

Neurosciences at the University of California, San

Diego, and the San Diego V.A. Healthcare System. I

have received a waiver for advisory board

activities for a competing firm for which I receive

less than $10,001 per year.

DR. GREEN: I am Dr. Mark Green. I am a

Clinical Professor of Neurology at Columbia

University and Director of the Columbia University

Headache Center. I have received a waiver from my

employer's contracts and grants with three

competing firms. My employer receives less than

$100,000 from one, between $100,001 and $300,000

from a second and more than $300,000 from a third.

MS. KILLIAN: Thank you very much.

Lastly, Dr. Roger Porter is the Industry

Representative on the committee today and he will

be acting on behalf of all related industry.

In the event that the discussions involve

any other products or firms not already on the

agenda for which an FDA participant may have a

financial interest, all meeting participants are

reminded that they are required by 18 U.S.C. 208 to

exclude themselves from such deliberations and

announce their exclusion for the record.

Finally, in the interest of public

transparency, with respect to all other

participants, we ask that they publicly disclose,

prior to making any remarks, any current or

previous financial involvement with any firm whose

products they may wish to comment upon.

Thank you very much. This concludes my

statement.

DR. KIEBURTZ: Thank you everyone for

doing that. For an inaugural run, I think that

went pretty well.

I would just like to point to the agenda

before letting Dr. Katz begin which is, some people

know, the sponsor will have approximately an hour

and fifteen minutes, up until the 9:45 break, to

give their presentation. We will then break and

then there is a presentation from the FDA.

There will then be the opportunity for the

committee to ask questions about the content of

those presentations to the presenters. Then we

will break for lunch. There will be a public

hearing after that and then a discussion amongst

the committee members after that.

During those discussions, the committee

members may ask questions of the presenters

regarding details of their presentation.

Presenters may not interject or contribute to that

discussion voluntarily, just so people know the

rules of the game here.

If you have questions that arise during

presentations, the FDA's presentation slides are

numbered. You may want to note them. You may want

to note the slides of a presenter so that you can

refer back to them with reference when you pose a

question.

On the FDA side of the table, I would like

to introduce four people. It looks like it flows

from right to left from my sitting. Dr. Robert

Temple, Dr. Rusty Katz, Dr. Bastings and Dr.

Southworth.

Dr. Katz?

Overview of Issues

DR. KATZ: Thanks, Dr. Kieburtz. I want

to be very, very brief. I just have a couple of

points I want to make but, first, I want to add my

welcome to the committee. Thanks for coming.

Particularly we have several new members. I would

like to thank them for agreeing to serve on the

committee.

I would also like to thank Dr. Kieburtz

for agreeing to chair the committee. It can be a

tough job. I would also especially like to thank

our invited guests, of whom we have quite a few,

who are experts to help us deal with this

interesting issue. In particular, I would like to

thank Dr. Jinnah who has graciously agreed to

actually be part of the presentations this morning.

So thanks very much to everybody for coming.

As you know, we are here to discuss NDA

16-145 submitted by Pozen for the use of MT100

which, as you have heard and which you know, is a

combination of naproxen and metoclopramide for the

treatment of acute migraine.

Actually, we are asking you today to

address a type of question that is actually fairly

unusual for the committee to deal with and that is

because many of the questions that we are going to

be asking you to consider are hypothetical in

nature.

Those of you who have been on the

committee or have seen previous committee meetings

know that, in a typical case, when we bring you a

new drug application, we would ask you whether or

not the application contains sufficient evidence of

safety or effectiveness in order to support

marketing approval.

But, today, as Dr. Kieburtz has already

stated, we are not primarily interested in the

question of whether or not the sponsor submitted

substantial evidence of effectiveness for the

treatment of acute migraine. We have already

decided that they have not done so, in particular

because we are unsure that they have presented

sufficient evidence of effectiveness for the

combination, itself, as a treatment for acute

migraine.

But, perhaps, more importantly for today's

discussion, we have determined that they have not

demonstrated a contribution of one of the

components to the overall effect of the drug and

that component is metoclopramide. I think we will

have a lot of discussion about that particular

question today.

You will, though, of course, hear some

more or less detailed presentations of that

effectiveness data that we have already ruled on in

a sense. You will hear from the company and you

will hear, to some extent, from us as well, from

Dr. Bastings. We would hope that you would

primarily consider those data in the context of

helping to inform your answers to the series of

hypothetical questions that we are going to ask

you.

In particular, we would like you to think

about the previous effectiveness data in the

context of giving us your advice as to whether or

not, if the sponsor does perform an additional

study or additional studies in a particular

population which you will hear about, whether or

not the results--if the results of these new

studies or new study are more or less of the same

magnitude as what has been seen already, whether or

not you would think that would justify approval of

the combination given the potential risks of the

treatment.

Of course, the potential risks of the

treatment are the underpinnings for the second

serious of hypothetical questions we want to ask

you. Specifically, we are interested to know your

views about the likelihood of occurrence and,

perhaps, even estimates of the frequency of

particular adverse events that we are concerned

about which, as you know, are tardive dyskinesia,

primarily, but, in additional, other tardive

movement disorders and possibly neuroleptic

malignant syndrome associated with the chronic

intermittent use of metoclopramide as it would be

presumably used in the treatment of acute migraine.

This series of questions is hypothetical

because the current data on the risks for these

adverse events associated with metoclopramide, such

as they are, don't speak directly to the question

of what the frequency of--what they might be when

the drug is given in the regimen that the sponsor

proposes; namely, chronic intermittent use, as is

typical for an acute-migraine treatment.

As difficult as those questions might be

to answer, we would like you to go even further and

venture an opinion about what sort of possible

dosing recommendations, if any, actually could be

adopted that might reduce the risks to an

acceptable level and then ask you to discuss what

you think that possible result and level of risk

might be. So these are all, obviously, questions

for which we do not have adequate data.

That is what makes it difficult. We know

these are difficult questions, but partly because

they are so difficult, and partly because we think

these questions are very important to try to answer

from the perspective of public health, given the

large prevalence of acute migraine in the

population, that is why we have come to you today.

So, again, thank you for coming. I want

to thank you in advance for all the hard work that

you have done already in reading the documents and

in today's discussion. So thanks again and I look

forward to an interesting and productive meeting.

Thanks.

DR. KIEBURTZ: Thank you, Dr. Katz.

Actually, I realize I was a little remiss

in introducing all of you. Maybe, as we have all

had the chance to introduce ourselves around the

table, Dr. Temple, maybe you could start so that

everyone knows who you all are.

Also, to follow up on Dr. Katz' comments,

just before you do that, Dr. Temple, these are

difficult questions and they are unusual questions.

I hope the committee members feel comfortable

voicing if they are uncertain about that and I will

be happy, as chair, to direct back to the FDA

questions about clarifying as to whether we are

answering the questions they had in mind and

getting clarity that we are providing them the

advice that they are seeking from us because it is

a little bit unusual.

So, if people are a little uncomfortable

about that, that is how we can do that. We can ask

questions of them to be certain we are addressing

the issues at hand.

So, Dr. Temple, please.

DR. TEMPLE: Good morning. I am Bob

Temple. I am the Director of ODE I. That is

office in which the Division of Neurology Products

lives. I have not received a waiver.

DR. KATZ: I am Russ Katz. I am the

Director of the Division of Neurology Products. I,

too, am not allowed to have a conflict of interest.

DR. BASTINGS: I am Eric Bastings. I am a

clinical team leader in the Division of Neurology.

DR. SOUTHWORTH: I am Mary Ross

Southworth, a safety evaluator in the Office of

Drug Safety.

DR. KIEBURTZ: Next on the agenda is

presentations from the sponsor.

Sponsor Presentation, Pozen, Incorporated

Introduction and Summary

DR. REESE: Good morning and thank you.

(Slide CC-1-2)

Pozen wants to thank the FDA for

assembling the Peripheral and Central Nervous

System Drugs Advisory Committee today to review our

naproxen-metoclopramide combination product called

MT100 for the acute treatment of migraine with and

without aura.

(Slide CC-3)

Let me briefly review an outline of

Pozen's presentation for this morning. Following

my introductory comments, Dr. Schapira, Professor

and Chair of Neurology at the Royal Free and

University College Medical School in London and

Professor of Neurology at Queens Square, will

present an overview of tardive dyskinesia with

metoclopramide use.

Dr. Alexander, Senior Vice President and

Chief Medical Officer at Pozen, will briefly review

the efficacy data for MT100 as contained in our

NDA. Dr. Matchar, Professor of Medicine and

Director of the Center for Clinical Health Policy

Research at Duke University, will discuss the

potential role of MT100 in migraine therapy and the

benefit-to-risk ratio of MT100.

Dr. Silberstein, Director of Jefferson

Headache Center in Philadelphia and the current

President of the American Headache Society, will

review clinical considerations in migraine

treatments. Then I will close our presentation of

this morning.

(Slide CC-4)

A bit of history. Pozen filed the IND for

MT100 in 1997 and undertook a preclinical, clinical

and pharmaceutical development program. There were

several discussions in meetings with the FDA over

the next six years which culminated in the

submission of the NDA in July, 2003. Pozen

believed that the totality of the data in the NDA

supported approval of the fixed-combination

product. However, the FDA did not agree with Pozen

and issued a not-approvable letter in May, 2004.

A critical-path meeting was held in late

October, 2004 with the Division Director, Dr. Katz,

and the Office Director, Dr. Temple. As a result

of that meeting, the FDA suggested an

advisory-committee meeting be convened to address

the potential risk of tardive dyskinesia with MT100

before we undertook any additional work.

(Slide CC-5)

That brings us to today's meeting which

really revolves around one central question; does

the potential risk of tardive dyskinesia preclude

the ultimate approval of MT100, whether for all

patients or for a readily identifiable group of

patients who receive the maximum benefit.

(Slide CC-6)

MT100 is a patented pharmaceutical tablet

formulation which is basically a pill inside a

pill. The core consists of the 500 milligrams of

naproxen sodium that is sprayed with an insulating

coat followed by a spray coating of 16 milligrams

of metoclopramide hydrochloride, which is

equivalent to 13-and-a-half milligrams of

metoclopramide base, then followed by a color coat.

The tablet is designed to release

metoclopramide immediately into the stomach to

alleviate the gastroparesis often associated with

migraine following the release of the long-acting

drug, naproxen, after it leaves the stomach.

Please note that the doses of both components are

well below the maximum daily doses approved for

these two products for other indications.

(Slide CC-7)

In May, 2004, the FDA issued a

not-approvable letter for MT100 citing both

efficacy and safety concerns. The FDA concluded

that the efficacy data for MT100 provided only

modest benefit over naproxen at 24 hours and that

this benefit, coupled with the possible risk of

metoclopramide-induced tardive dyskinesia, did not

warrant approval of MT100.

The not-approvable letter also stated that

the data submitted in the NDA did not provide a

significant benefit for all of the

migraine-associated symptoms at two hours versus

placebo in two well-controlled studies. The FDA

did agree that one study was considered to have met

all the endpoints necessary for approval.

Therefore, the FDA felt that the potential

risk of developing tardive dyskinesia was not

outweighed by the 4 to 6 percent benefit of the

MT100 over the active control, naproxen, at 24

hours.

(Slide CC-8)

Now, regarding tardive dyskinesia, the

not-approvable letter states, "The absence of any

detected cases among 300 patients is consistent

with the true rate of TD of about 1 percent, an

unacceptably high risk in the absence of any

advantage of the product."

The FDA's mathematical calculation of 1

percent is derived from the upper limit of the 95

percent confidence interval around zero which we

believe is based on the 300 subjects in the

long-term safety study. Any implication that the

true rate approaches 1 percent is unfounded based

on the available scientific data in the literature,

the spontaneous case reports from the U.S. and the

U.K., national safety databases and our own

clinical-trial experience in treating over 3700

patients with MT100.

We feel that the risk of tardive

dyskinesia is very low and, certainly, much less

than 1 percent. While approximately 2700 of these

patients treated only single attacks, our 12-month

safety data that we conducted was actually three

times larger than the FDA had requested.

This study exposed over 1000 subjects to

MT100 for three months, over 600 subjects for 6

months and over 300 subjects for 12 months treating

over 23,000 individual migraine attacks and there

were no reports of tardive dyskinesia in these

studies.

(Slide CC-9)

Now, metoclopramide had been on the market

for over 20 years when Pozen submitted the NDA and

there were never any concerns raised by the FDA as

far as I am aware regarding tardive dyskinesia

during the development of MT100. Even though we

saw no cases of tardive dyskinesia during the

development program, to be conservative, Pozen

mimicked the current metoclopramide labeling found

in Reglan, from the Warnings Section of the

approved label, regarding any possible risk of

tardive dyskinesia.

The Reglan label states, regarding tardive

dyskinesia, that both the risk of developing the

syndrome and the likelihood that it will become

irreversible are believed to increase with the

duration of treatment and the total cumulative

dose. Less commonly, the syndrome can develop

after a brief treatment period at low doses. In

these cases, the symptoms appear more likely to be

reversible.

I would like to stress, again, that the

use of MT100 in the migraine population exposes

patients to both a low dose, 16 milligrams, of

metoclopramide hydrochloride and to an episodic use

of about three to six times per month.

(Slide CC-10)

Based on the available scientific

evidence, Pozen submits that the risk of tardive

dyskinesia associated with metoclopramide use is

very low and should be even lower with the episodic

use of MT100. The therapeutic dose of

metoclopramide hydrochloride, as I said, in MT100

is only 16 milligrams. The data from the long-term

safety study indicates that the expected use of

MT100 is only about four doses per month.

Dr. Schapira will review the spontaneous

national safety databases from both the U.S. and

U.K. and the scientific literature. There have

been very few cases of tardive dyskinesia reported

from the chronic use of metoclopramide as a single

ingredient over the past 40 years and, to our

knowledge, no cases of tardive dyskinesia have been

reported with the episodic use of metoclopramide.

As I said, there were no cases of tardive

dyskinesia seen in our clinical-trial program

either. Therefore, to the best of our knowledge

from the literature, the national safety databases

and experts in the field, the risk of developing

tardive dyskinesia from the episodic use of MT100

should be lower than currently approved

metoclopramide-containing products. Therefore,

Pozen feels its potential risk of tardive

dyskinesia should not preclude the ultimate

approval of MT100.

(Slide CC-11)

Since MT100 is a fixed-combination

product, it must also satisfy the FDA combination

policy as shown on this slide which simply states

that, "Two or more drugs may be combined in a

single form when each component makes a

contribution to the claimed effects and the dosage

of each component is such that the combination is

safe and effective for a significant patient

population requiring such concurrent therapy as

defined in the labeling."

We believe MT100 satisfies this policy.

(Slide CC-12)

Dr. Alexander will review the efficacy

data for MT100 in a few moments, but I would like

to share a few highlights of what he will show you.

There was a significant improvement in the

primary endpoint of sustained pain response over 24

hours in five of six studies versus placebo or the

pseudoplacebo metoclopramide. One study did not

achieve significance and the p-value was 0.054.

The data from the two component studies

both demonstrate that each component of

metoclopramide makes a significant contribution to

the claimed effects for all patients but an even

greater effect in a significant patient population

experiencing migraine attacks without nausea.

In addition to the primary 24-hour

sustained pain endpoint, the FDA requested that we

evaluate migraine efficacy endpoints at two hours

versus placebo. In all six efficacy studies, MT100

was always significantly better than placebo for

pain at two hours. We also showed improvement over

the associated symptoms of nausea, photophobia and

phonophobia at two hours.

Although these studies were not powered to

show a difference in these secondary symptoms, in

most cases, MT100 was numerically, if not

statistically, superior to placebo.

(Slide CC-13)

In conclusion, I believe that the

potential risk of tardive dyskinesia should not

preclude the ultimate approval of MT100.

(Slide CC-14)

Next I would like to introduce Dr.

Schapira, Professor and Chair of Neurology at the

Royal Free and University College Medical School in

London and Professor of Neurology at Queens Square,

who will summarize the available information on

tardive dyskinesia associated with metoclopramide

use.

Thank you.

DR. KIEBURTZ: Dr. Reese, before you--does

anybody have just a quick clarification or--okay.

Thank you.

Overview of Tardive Dyskinesia

DR. SCHAPIRA: Thank you, Dr. Reese, and

thank you, Dr. Kieburtz, and thank you to the

committee for the opportunity to come and speak to

you this morning.

I guess I am coming here wearing two hats.

The first is of a neurologist, a general

neurologist, in the U.K. who, in outpatient clinic,

sees a spectrum of neurological disorders, a

significant proportion of which, of course,

includes headache and a significant proportion of

that, in turn, includes migraine.

The second hat is that of a neurologist

with a specific interest in movement disorders. So

it is with those two hats, if you wish, that I am

going to cover some specific areas this morning.

(Slide CC-15)

The first is to address the issue of why

use metoclopramide in migraine and the second is

specifically to address the risk of tardive

dyskinesias, or TD, with metoclopramide use. I

would like to divide my comments on this into three

areas; the chronic, intermittent and episodic use.

I will come back each of those in turn.

(Slide CC-16)

Just to begin with why use metoclopramide

in migraine.

(Slide CC-17)

I will cover this only briefly because

others will also comment on this, but we know that

it enhances absorption of orally administered

analgesics. It is an anti-nauseant and

anti-emetic. A meta-analysis indicates that

parenteral metoclopramide seems to have a specific

anti-migraine activity on its own.

(Slide CC-18)

The advantages, if you wish, of

metoclopramide in migraine have actually been used

in the U.K. because we have, for 25 years, actually

had access to three drugs, all of which are

metoclopramide analgesic combinations. The first

is MigraVess. The second is Paramax, and MigraMax.

MigraVess was available between 1980 and 1999 and

was then withdrawn in favor of Migramax because of

the higher dose of aspirin compound in the latter.

All of these three compounds, as I say,

contain 10 milligrams of metoclopramide per dose

and the maximum recommended dose in the U.K. is

three dose per 24 hours, so a 30-milligram-per-day

dose of metoclopramide.

There is no restriction in the U.K. on the

number of times a patient may take this compound

per week, per month, et cetera, so long as they do

not exceed the three-times-per-day, 24-hour, dose.

I should also point out that, for general use,

metoclopramide has been available in the U.K. since

1964.

(Slide CC-19)

The use of these metoclopramide-analgesic

combinations in the U.K. have been found useful.

In fact, they have now been incorporated into the

U.K. Guidelines for the management of acute

migraine. The first step is a simple analgesic.

The second step is, then, the

metoclopramide-analgesic combinations given orally

or, if necessary, given by suppository. The third

step is the use of a triptan.

We have found, in clinical practice in the

U.K., that that middle step, that Step 2, is a very

useful practical intermediate step between the use

of simple analgesics and the use of a triptan.

(Slide CC-20)

Now, I would like to come on specifically

to address the issues of tardive dyskinesia. In

terms of the use, I will focus first on chronic

use. This I am going to define, really, as the

most frequent, most common, use in the U.S.,

particularly, of Reglan, or metoclopramide, for its

gastrointestinal uses, and also in the U.K. we have

an equivalent drug which we call Maxolon, again

with the same range of uses for gastrointestinal

disturbances.

(Slide CC-21)

Let me, first of all, though, before

moving on to the surveillance data, begin with a

view of tardive dyskinesia. There are several

different definitions of tardive dyskinesia, so

what I have tried to draw out is some of the

commonalities between them.

I think we could say that it is a syndrome

consisting of potentially irreversible involuntary

dyskinetic movements which can affect any part of

the body but which predominantly affect the

orolingual-buccal region. It has traditionally

been associated with chronic, and that is 30 days

or more, use of a dopamine antagonist, generally

speaking, at the higher dose ranges of the those

antagonists.

But some definitions of TD also include

daily use for three months, or daily use for one

month if the patient is 60 years or more, onset

during use or, alternatively, onset with four to

eight weeks of cessation.

The pathogenesis of tardive dyskinesia is

not fully understood but it is thought to include

the development of supersensitivity of the

dopaminergic system. The prognosis of TD, once it

develops, is variable and, again, the precise

handles on this can vary. Two studies, for

instance, quoted in the helpful FDA submission,

suggest that 33 percent of patients may resolve

spontaneously in two years and another 29 percent

over six months.

But, certainly, TD can be irreversible and

can be extremely distressing.

(Slide CC-22)

I would like to just now move quickly to

some of the surveillance data that is available on

TD, the first of which, looking at the association

between TD and metoclopramide came from

Scandinavia. Between 1977 and 1981, there were

established 11 million doses and they identified 11

cases of TD.

Then the first of two U.K. studies. The

first was a retrospective analysis of the Committee

of Safety of Medicines. This is a yellow-card

system whereby medical practitioners will send in a

yellow card to the CSM when they identify an

adverse drug reaction.

Looking at the years between 1967 and

1982, so about 15 years, of Maxolon only--so this

is looking at the use of, if you will, the Reglan

equivalent in the U.K.--it established 15.9 million

prescriptions over this 15-year period, so just

over 1 million per year. They identified four

cases of TD.

Then there was a prospective study by the

same author looking at a time point in 1986 over a

six-month period where they prospectively looked at

prescriptions, again for Maxolon, the Reglan

equivalent, not for the metoclopramide-analgesic

combination. So, for this Reglan equivalent, they

identified just over 2-and-a-half thousand

prescriptions or patients who were given

prescriptions and found 25 extrapyramidal events

with 12 dystonias, eight akathisias, five

drug-induced Parkinsonism but no case of tardive

dyskinesia.

It might be helpful just for me to set in

context the dosage issues of metoclopramide in the

form of Reglan or Maxolon and that suggested for

MT100.

(Slide CC-23)

Reglan, here, I understand, is used at a

recommended dose of 10 to 15 milligrams per day, in

some cases up to 20 milligrams, but the general

recommendation is for 10 to 15, up to four times a

day. So the maximum dose would be 60 milligrams a

day. Then the course of the medication varies

according to the indication it is used for, up to

eight weeks or up to 12 weeks.

If you look at the maximum calculated

recommended exposure for one course, you come to

just over 5 grams of metoclopramide. But, if you

take a conservative estimate of usage--let's say if

you half that recommended maximum--you would come

out with, let's say, 10 milligrams four times a day

and for eight weeks rather than 12 weeks.

You come out to about 2.24 grams, so that is 45

percent of the maximum recommended dose on that

schedule.

Just to put this in context, that

half-exposure, if you wish, half of the maximum

recommended exposure, is the equivalent to treat

166 doses of MT100, 166 migraine attacks, or, if a

patient were to take MT100 at its maximum

recommended dose of 6 tablets per month every

month, they could take 2.3 years of MT100.

In fact, the median number of doses per

year of MT100 in the 302 study was 22, so if you

translated this into practical MT100 usage, this

would be the equivalent to seven-and-a-half years

of Reglan at half its maximum recommended dose or

15 years of practical use of MT100 at the maximum

exposure of Reglan in one specific course.

So that just sets the sort of dosage

issues in context.

(Slide CC-24)

I would like to now come to this very

helpful review by Shaffer, Dr. Shaffer, who was--he

and two other colleagues from the FDA and another

from Duke published a paper looking at the U.S.

reporting system for the period 1968 to 2003, so

over 35 years.

Now, just for the 10-year period between

1994 and 2003, they estimated that there were about

42 million scripts for metoclopramide. They

identified in their database 87 cases, 40 of which

made that predetermined definition of TD. But I

will talk about this in a little bit more detail.

(Slide CC-25)

This is a 35-year review. Interestingly,

just when they looked at all the scripts for

patients who were given metoclopramide, 62 percent

of those were intended for women and 24 percent,

almost a quarter, were intended for patients who

were age 70 or over. The authors actually didn't

include the use of migraine in their estimations

but I understand from the FDA submission that they

have now calculated that 2 percent of this use was

for migraine and, no doubt, they will address that

issue specifically themselves.

Now, the predetermined definition of TD

that these authors used to identify their cases was

metoclopramide exposure for 30 days or more and

documented involuntary movements or symptoms. As I

say, they identified 87 separate reports but 60 of

these had involuntary movements and 53 had duration

of use of 30 days or more.

In practice, 40 of the 87 met the

predetermined criteria of TD. I note that, in the

FDA submission, their number is 68 and, again, no

doubt, they will address that separately.

Of those that did develop TD, the mean age

was 60 with a range of 11 weeks to 95 years, and 65

percent of the TD patients were women which

corresponds, actually, quite well with the 62

percent women that were given the scripts in the

first place.

The mean dose was 33 milligrams per day,

the duration 753 days although, again, the FDA

submission, I note, identifies the median as 180

days. Six of the patients were on anti-psychotics

as well as metoclopramide and 22 of them were

considered to have permanent disability, eight of

whom needed a visit to the emergency department or

hospitalization because of their TD.

(Slide CC-26)

I would like to now move from what I have

considered in terms of the Reglan or Maxolon type

usage in the U.K. and the U.S. to the intermittent

or episodic. Here I would like to draw my own

distinction between these.

In my understanding, intermittent

pharmacotherapy is a course of treatment separated

by a period of treatment followed by another course

of that same treatment so, over a prolonged period,

intermittent doses with periods in between without

the medication. I contrast that with episodic PRN

or as in "when required" use such as, for instance,

as used in acute migraine attacks. That is what I

am going to refer to as episodic use.

(Slide CC-27)

Let me just remind you that, in the U.K.,

we have, for the last 25 years, had access to these

metoclopramide-analgesic combinations for the

treatment, the episodic treatment, of acute

migraine the dosage of which, in any 24 hours, is

30 milligrams. Looking at the equivalent,

incidently, in MT100, the maximum daily dose is

13.5 milligrams in terms of the base of

metoclopramide which is the equivalent in these

combinations.

In the U.K., it is estimated almost

100,000 patients receive a total of about 8 million

doses of these combinations per year. In the

five-year period 1999 to 2003, there were estimated

to be a total of 40 million doses. So these are

drugs which are used relatively commonly for the

treatment of acute migraine in the U.K.

(Slide CC-28)

Now, the ADROIT database is a physician

database. It records physician-identified and

reported adverse events to a central, now

computerized, database and it records prescriptions

as well as adverse events, so it is particularly

helpful.

In the period 1964, when metoclopramide

first became available, to 2005, so about a 40-year

period, they were able to collect data on

metoclopramide. But what is, I think, of

particular interest this morning is that this

database is able to discriminate between the

Maxolon-Reglan type use in the U.K. and the use of

metoclopramide-analgesic combinations for acute

migraine. So the database discriminates between

those two uses.

They found almost 3000 adverse-event

reports by any route of which 156 were related to

the acute-migraine metoclopramide-analgesic

combinations of which 69 were neurological over a

period from 1980 to 2005 which is when these

combinations have been available to us.

(Slide CC-29)

Just to look at little bit more closely at

these 69 neurological events over that 25-year

period reported to this database, there were 26

dystonias or oculogyric crises, eight

extrapyramidal disorders not specified, three

dyskinesias which were not classified as TD--they

were reversible after the patient stopped their

medication--one of Parkinsonism, one of akathisia

but no reports of choreiform movements and no

reports over this 25-year period of tardive

dyskinesias.

(Slide CC-30)

There were a collection of other

neurological events; acute extrapyramidal disorders

were numbered 14 and this may well include things

like oculogyric crises, and then a variety of other

neurological features. So that totals a number of

69 none of which were TD.

(Slide CC-31)

Just to make a comparison between acute

episodic use of metoclopramide-analgesic

combinations for acute migraine and the other

general use of metoclopramide, I have listed there

the adverse events. You will see that there have

been reports, of course, of a variety of

neurological events including the

dystonia/oculogyric crises with the more chronic

type of metoclopramide use, the sort of

Maxolon-Reglan type use, and 24 cases of tardive

dyskinesia with the non-migraine use of

metoclopramide compared to the zero for the

migraine use.

(Slide CC-32)

I would just like to very briefly cover

the MT100 experience; nine phase 3 studies, 3,700

subjects, over 25,000 doses and a study which took

just over a 1,000 patients to follow them up over a

period of up to 12 months.

In the MT100 studies, there were two

patients that experienced acute dystonic reactions

but no patients that experienced tardive

dyskinesia.

(Slide CC-33)

Just looking at the longer-term study,

1,000 patients recruited, 621 were followed over

six months, 329 over 12 months, treating 23,000

migraine attacks. As I mentioned before, the

median number of doses per patient over the 12

months was 22 and the mean number of days between

each dose was almost 10.

(Slide CC-34)

So just looking at the--one has to accept

somewhat limited MT100 data. We haven't seen any

cases of TD. But just looking at the U.K. data

where we have got data now for over 25 years, and

there is that period 1999 to 2003 where

specifically, just for that period, they have

estimated 40 million doses, we haven't had any

reports to the ADROIT database of any cases of

tardive dyskinesia over that period.

(Slide CC-35)

I would like to summarize. I think we

have to accept that the MT100 experience is

insufficient to exclude a small risk of TD with its

usage. But, moving to the larger U.K. experience,

I think we have had no reports of

analgesic-metoclopramide combinations causing TD

and that is use for migraine over 25 years and at a

very conservative estimate, over 100,000 million

doses.

This, remember, is using a dosage and a

frequency for these analgesic-metoclopramide

combinations in the U.K. which is greater than that

which is proposed for MT100.

(Slide CC-36)

The FDA briefing documents raised some

important topics and I would just like to address

three of those specifically. The first is the

question that they asked, is there sufficient

evidence that the chronic intermittent

administration of metoclopramide does not carry the

same risk of TD as the chronic administration.

I can say that the experience from the

U.K. over the 25 years that we have had them of

these metoclopramide-analgesic preparations, the

answer is yes. Yes; we do have sufficient evidence

that the chronic intermittent administration of

metoclopramide does not carry the same risk of TD

as the chronic administration.

(Slide CC-37)

So, if the answer is yes, what is the

maximum number of recommended monthly doses to

avoid the risk of TD? Well, the answer to that is

not known. But I have to come back to the U.K.

experience just to mention that, over the 25 years,

there have been no cases of TD using the

metoclopramide-analgesic combinations at their

recommended dose and schedule which exceeds that

for MT100.

(Slide CC-38)

Finally, this is an issue which will be

addressed by other experts specifically and that is

on medication-overuse headache, but the question is

posed, do you believe that, based on the existing

data on medication-overuse headache, there is

evidence that the proportion of patients prescribed

MT100 will likely take a number of monthly doses

higher than that recommended.

Well, I can't answer this question

specifically, of course, but I can only say that if

this does happen, even if it does happen with this

type of combination, the U.K. data don't indicate

that it should lead to TD.

Thank you very much.

(Slide CC-39)

I would like now to pass to Dr. Jim

Alexander who is Pozen's Chief Medical Officer. He

will review the data on the efficacy of MT100 in

migraine.

DR. KIEBURTZ: Same thing. Anyone have a

point of clarification?

DR. SMITH: Could you go over with me, on

Slide CC-21, the definition of tardive dyskinesia,

please--the definition. My question is, you say

some definitions include the duration of exposure.

When do the definitions include that? In other

words, is that a common use definition?

DR. SCHAPIRA: The definitions vary. As I

say, some definitions, looking at the case studies

that have been published on TD and metoclopramide

have required that the patient has been taking

metoclopramide continuously for two months, others

for three months. Some of the other studies like

the Shaffer review have said that the patient

should be taking it for 30 days or more.

So there is some variation in how people

define the requirement of metoclopramide exposure

before they will associate it with TD.

DR. SMITH: I see. So, if it doesn't meet

the duration of use, it would be dyskinesia, not

TD? Is that correct?

DR. SCHAPIRA: I think that would depend

on the individual study and the interpretation of

the authors. For instance, in the Shaffer paper,

they identified that they would use the definition

of 30 days or more. But they also recognized--for

instance, I think they reported on three juvenile

cases, two infantile and one adolescent case, that

developed tardive dyskinesia, I think the two

infants following an overdose of metoclopramide and

the adolescent also had some other features.

So I think it depends clearly how strictly

you want to define and whether you will comment on

other cases that fall outside your definition.

DR. SMITH: Okay. Thank you.

DR. KIEBURTZ: Let's hold on that because

we will hear more about definitions. If the

question is about TD definitions--no? Go ahead.

DR. LENAERTS: Dr. Schapira, in sharing

your U.K. experience, what is your estimate of the

prevalence of specifically migrainers either

overusing metoclopramide-analgesic combinations or

staying frequently or constantly at the maximum

recommended dose, because you mention--

DR. KIEBURTZ: Excuse me. I am just going

to stop. If you have a clarification on what he

presented, that is one thing. Additional questions

about something he didn't talk about, not yet.

DR. LENAERTS: Thank you. I will hold.

DR. KIEBURTZ: Just clarifications of the

presented material. Dr. Katz?

DR. KATZ: A couple of questions. On

Slide 22, the second Bateman study, just for

clarification, what the design was. That was a

prospective study?

DR. SCHAPIRA: No; that is a retrospective

study. The CSM, yellow-card system.

DR. KATZ: The second one is the

yellow-card system. I thought you said it was a

prospective study.

DR. SCHAPIRA: No. I'm sorry. The second

Bateman study that you see on the slide there, the

one published in 1989, that was a prospective

study.

DR. KATZ: Right; I am talking about the

second study.

DR. SCHAPIRA: I'm sorry. I thought you

said second on the list.

DR. KATZ: Oh; I'm sorry. The second

Bateman study, the 1989. So that is prospective,

so those patients were followed and their adverse

events were recorded contemporaneous with their

occurrence.

DR. SCHAPIRA: Yes.

DR. KATZ: It was a true prospective--

DR. SCHAPIRA: Yes.

DR. KATZ: Okay. Thanks. One other

question. Slide 28, which looks at the reports of

these events over a 40-year period, do you know

anything about the temporal pattern of those

reports? In other words, were there more reports

earlier on and then reports started to wane over

time which sort of happens all the time, we think,

with spontaneous reports? Do you know anything

about that?

DR. SCHAPIRA: No; I can't comment on

those. I can only say that the system has been in

place, of course, for all of that time. More

recently, over the past years, it has been

computerized. So the ADROIT system is a fairly

responsive system which is linked to primary-care

computers throughout the U.K. But I can't tell you

about the pattern of those over the years.

DR. KATZ: Just one other, if I can,

question. The previous slide, 27, which looks at

the combination, the actual acute-migraine

treatments, do we know the actual doses that people

took? As you say, the maximum dose, I guess, is 30

a day. Do we know? I don't know. Maybe we figure

it out from the numbers, but do we know what people

actually took?

DR. SCHAPIRA: No; we don't know precisely

how many they took, only how many were prescribed.

As I say, it is estimated as an average of 85 per

person, but that doesn't tell you how many they

took in an individual dose. So I don't have the

data on that.

DR. KATZ: Dr. Fahn?

DR. FAHN: If we can go back to slide 22,

again, for a clarification, the second Bateman

study, the 1989 study, zero cases of TD, do you

know what definition of TD they used to come to

that number?

DR. SCHAPIRA: No. They did not specify

their definition of TD.

DR. KIEBURTZ: Dr. Goldstein?

DR. GOLDSTEIN: I am not all that familiar

with your drug-reporting system. Two questions

about it. One is how compulsive is the use of

this? In other words, how often do you think you

are actually getting reports about things that are

actually happening.

The second question related to it is does

the system allow for validation somehow of these

reports because, especially with primary-care

physicians, it is not clear to me how accurate

these reports may be about particular types of

problems.

DR. KIEBURTZ: It is a little evaluative.

It is a good question. Can we hold on it because I

am conscious that the sponsor only has a certain

amount of time to present. I don't want to

infringe on that.

One last question about the second Bateman

study that you have already had questions about.

Was that only new prescriptions?

DR. SCHAPIRA: Yes.

DR. KIEBURTZ: Thank you.

DR. SCHAPIRA: I'm sorry; can I just

clarify. That was the number of prescriptions that

were given during that six-month period. So it

didn't specify whether that was a renewed

prescription for that individual or not.

DR. KIEBURTZ: Oh; I see. Okay. Thank

you.

DR. SCHAPIRA: Thank you. I will now hand

over to Dr. Alexander.

Review of MT100 Efficacy

DR. ALEXANDER: Thank you, Dr. Schapira.

Although the potential risk of tardive

dyskinesia is the primary focus of this meeting,

when Pozen and the FDA discussed the meeting, we

agreed that the committee should have the

opportunity to review data described in the

efficacy of MT100.

(Slide CC-40)

My presentation will, therefore, summarize

the results of studies designed to evaluate the

efficacy of MT100 using two different trial designs

which evaluated the acute treatment of single

migraine attacks.

First, I will show the results from the

phase 3 studies which evaluated MT100 versus

placebo or metoclopramide as a pseudoplacebo.

These studies examined the efficacy of MT100 as a

migraine drug using those endpoints that are

usually required for the approval of new migraine

therapies.

Secondly, I will review the data from the

two component controlled trials which I will call

the factorial studies. These are the trials that

compared MT100 to its two individual components.

Now, as you have heard, the efficacy of naproxen

sodium as a component of MT100 is really not in

question. So my focus in discussing these data

will be on comparisons between MT100 and naproxen

sodium which directly address the contribution of

metoclopramide as a component of MT100.

(Slide CC-41)

The MT100 phase 3 program was quite

extensive and almost 6,000 subjects were enrolled

in six controlled trials treating single migraine

attacks. Four studies directly compared MT100 with

placebo while, in the two factorial studies shown

below, 301 and 304, metoclopramide was considered a

pseudoplacebo.

2,355 subjects received single doses of

MT100. Did these studies provide evidence that

MT100 was an effective migraine drug? Well, Pozen

believes that the data clearly showed this.

(Slide CC-42)

This table lists the six studies in the

left-hand column. It is arranged to show the 30

individual different primary and secondary

endpoints in the five columns to the right. Study

306, which is at the top, is the study that was

accepted by the FDA as demonstrating the efficacy

of MT100. The two columns on the far left show the

key pain endpoints--that is, sustained pain

response at 24 hours, which was the primary

endpoint in four studies, and the two-hour pain

response in the second column was a key secondary

endpoint in five studies.

As shown now on the slide, in 11 of 12

comparisons, MT100 was significantly superior to

the comparator for each of these pain endpoints.

In Study 303, which had an unbalanced randomization

with a smaller number of placebo recipients, the

p-value for sustained pain response was 0.054.

But in all six studies, the efficacy of

MT100 over the comparator for the two-hour pain

response, shown in the second column, was

significantly superior. These results provide

clear and compelling evidence that MT100 provides

effective two-hour pain relief, the usual

regulatory endpoint in migraine trials, as well as

providing sustaining pain responses at 24 hours.

I will provide a better definition of

sustained response in a few minutes. I want to

mention the efficacy on the associated symptoms.

Efficacy for the associated symptoms of nausea,

photophobia and phonophobia, are also for a

migraine drug. But, in contrast to pain, these

symptoms are not always present in migraine attacks

and, in fact, none of the Pozen studies were

powered to detect differences for these endpoints

but all were specified as secondary endpoints in

our studies.

Nevertheless, significant differences in

the incidences of these symptoms were seen among a

number of these studies at two hours after dosing,

as shown now. In additional comparisons, shown in

yellow, the p-values were between 0.05 and 0.1.

The p-values, finally, in orange, are above 0.1.

As is reviewed in your briefing document,

by three or four hours after dosing, significant

benefits on all of these associated symptoms were

usually present with MT100 treatment.

So, to summarize, the totality of the

evidence from these six studies clearly shows that

MT100 is an effective acute treatment for migraine.

(Slide CC-43)

I will now show the comparisons of MT100

against naproxen sodium. These comparisons, again,

reflect the direct assessments of the contribution

of metoclopramide within MT100 in order to satisfy

the combination drug rule.

The two phase 3 factorial studies were

each performed at sites in the U.S. Subjects were

randomized to treatment with either MT100, naproxen

sodium 500 milligrams, or metoclopramide 16

milligrams, the identical doses of these component

drugs that are contained within MT100.

Subjects treated a single migraine attack

of moderate of severe pain intensity and symptom

assessments were performed at baseline and hourly

for 24 hours. Rescue medication was permitted

after at least two hours had elapsed after dosing.

(Slide CC-44)

I would like to take a second and explain

the pain assessments in these trials, the primary

endpoint as well as the secondary endpoints.

Sustained pain response at 24 hours was agreed by

Pozen and the FDA as the appropriate measure to use

to assess the efficacy of MT100 versus each of its

two components.

Sustained pain response is a composite

measure of efficacy and is defined as pain relief

at two hours--that is, no pain or only mild

pain--and no relapse or moderate or severe pain and

no need for the use of rescue medication over the

next 22 hours after the two-hour assessment. The

efficacy of this endpoint is judged by how many

subjects meet this definition at 24 hours.

I would like to stop at this point and

explain why Pozen and the FDA agreed that the use

of the two-hour pain response endpoint would not be

acceptable for the comparison of MT100 with

naproxen sodium. This was because both treatment

are active due to the presence of naproxen in each

drug and should, in fact, produce very similar pain

responses at the time point of two hours after

dosing.

Two-hour pain response was a secondary

endpoint in these studies and was used to evaluate

MT100 versus metoclopramide as a pseudoplacebo, as

I have previously shown.

In contrast to the sustained pain response

and two-hour response rates which measure the

number of subjects responding, at the bottom of the

slide, you will see three secondary endpoints that

were also evaluated in these trials. These are the

Pain Intensity Difference score, PID, the Sum of

Pain Intensity Difference scores, SPID, and the

TOTPAR scores, or Total Pain Relief scores, over

time.

These are the measurements of how much

pain relief is obtained, not of how many subjects

have a specific pain response at a given time.

These are the accepted general endpoints for

analgesics within the FDA. They are recognized as

very sensitive for detecting differences between

individual active analgesic drugs.

But let's first look at the agreed primary

endpoint and that was sustained pain response from

these two studies.

(Slide CC-45)

Shown here are the data from these studies

with a percent of responders plotted. First, note

that the metoclopramide-alone treatment produced

sustained pain response rates of 19 and 20 percent

which are similar to what might be expected of a

placebo.

The responses to naproxen sodium alone

were 9 to 10 percent higher than metoclopramide and

the rates were actually 28 percent and 30 percent

in the two trials. These were significant

differences over metoclopramide. The sustained

response rates for MT100 were 6 percent and 4

percent higher than those for naproxen sodium in

these two studies.

I am sure you have noted that Pozen and

the FDA arrived at different p-values for these

comparisons. But both parties agree that the

absolute differences are 4 and 6 percent for this

endpoint. Are these differences confirmed by other

analyses? The secondary endpoints provide support

for these findings.

(Slide CC-46)

The mean SPID scores at 24 hours show

significant differences for MT100 versus naproxen

sodium in both studies. So these analyses of a

secondary endpoint, a valid measure of pain relief,

support the findings of the sustained endpoint. I

would also note, and not shown, but the fact that

the differences were significant in the SPID scores

at two hours after dosing in both studies.

(Slide CC-47)

A third dataset, the 24-hour TOTPAR

scores, is also supportive with mean TOTPAR scores

at 24 hours for these two studies showing

significant differences between MT100 and naproxen

sodium. So these additional analyses, which were

secondary, support and confirm the results seen

with the sustained pain-response endpoint and

substantiate the contribution of metoclopramide to

the effect of MT100.

But, even if this were not the case, there

is a subgroup pseudoplacebo that seems to respond

much better to MT100 than the naproxen sodium.

Now, the reason that we can discuss this subgroup

is the following: at the outset of the phase 3

program, Pozen theorized that metoclopramide might

contribute not only to better pain relief but might

also contribute to the relief of nausea that may

accompany migraine attacks.

(Slide CC-48)

For this reason, one of the three

pre-planned analyses that were used in all of the

phase 3 studies include analyses of pain endpoints

within two subgroups of migraine attacks--that is,

those with nausea and those without nausea at the

time of treatment.

(Slide CC-49)

These are the results for subjects whose

migraine attacks were not accompanied by nausea.

This type of migraine attack made up one-third of

the attacks treated in Study 304 and one-half of

the attacks treated in Study 301. The number of

subjects in each study is shown with the figure on

the left being 301, the figure on the right, 304.

In these subgroups of attacks, the

differences between MT100 and naproxen sodium for

sustained pain responses were essentially doubled

to 9 and 10 percent. In this instance, the

differences were highly significant, with p-values

less than 0.01 in both studies. This was seen in

both studies and, therefore, is not likely to be a

chance occurrence.

Pozen took a further step of providing its

phase 3 data to Drs. Richard Lipton and Ken

Kolodner who conducted independent analyses of

these data and confirmed these findings. The odds

ratios and the significant p-values are provided in

your briefing document in Table 11.

(Slide CC-50)

As additional confirmation, the mean SPID

scores in these subjects with attacks without

nausea also showed significant differences in these

sensitive measures of pain relief for MT100 versus

naproxen sodium at 24 hours. When Pozen met with

the FDA in late 2004 and the data for this subgroup

of attacks were presented to the agency, Pozen was

asked if the same effect was seen for MT100 across

the phase 3 studies.

The answer is definitely yes. Pozen

performed a pooled analysis of phase 3 trial data

and these results were obtained.

(Slide CC-51)

These studies were conducted in the same

time period. They all treated subjects with

migraine attacks of moderate to severe intensity

and there were similar entry criteria and

evaluation criteria. The comparators included

placebo, metoclopramide and naproxen sodium.

As you can see, there was a significant

difference only in the treatment with MT100 for the

comparison of the treatment of attacks with and

without nausea, again, highly significant.

So why would these effects be present?

The only plausible explanation is the 16 milligrams

of metoclopramide contained within MT100.

(Slide CC-52)

Therefore, the unique contribution of

metoclopramide may be described as counteracting

the gastric stasis associated with migraine,

enhancing the rate of absorption of naproxen,

providing better pain relief in the overall

treatment population and, finally, enabling maximum

benefit to be obtained in migraine attacks without

nausea.

(Slide CC-53)

So where does this leave the efficacy of

MT100? Pozen believes that the data show that

MT100 is an effective migraine treatment, that

MT100 provides an absolute 4 to 6 percent

improvement in sustained pain response over that

for naproxen sodium, that MT100 provides absolute 9

to 10 percent improvements in sustained pain

response over naproxen sodium in migraine attacks

without nausea.

Secondary endpoints, SPID and TOTPAR,

confirm the superiority of MT100 over naproxen

sodium. Finally, the contribution of

metoclopramide to the primary endpoint of sustained

pain response is demonstrated in two studies.

Thank you for your attention.

(Slide CC-54)

I would like now to introduce--it is my

privilege now to introduce Dr. David Matchar,

Professor of Medicine at Duke University School of

Medicine. Dr. Matchar is Director of the Duke

Center for Clinical Health Policy Research and,

over the past two decades, he has made significant

contributions in the field of evidence-based

decision making in medical care. In the migraine

area, he has been a member of the U.S. Headache

Consortium and was lead author of the group's

evidence-based guidelines for the treatment of

migraine, a collaboration among eight professional

societies.

Dr. Matchar was invited by Pozen to

provide his perspective on the potential role of

MT100 in the treatment of migraine and his view on

the balance of benefits and risks of this

treatment.

DR. KIEBURTZ: Just real quickly, any last

clarifying questions? Dr. Welch?

DR. WELCH: The nausea versus the

non-nausea studies. Was that a prospective nausea

versus non-nausea?

DR. ALEXANDER: The studies were both

designed to have a preplanned analysis of the

subgroups of attacks with nausea and without

nausea.

DR. WELCH: So you didn't look for

separate populations.

DR. ALEXANDER: I'm sorry; I didn't

understand.

DR. WELCH: You didn't look for separate

populations. It was all in the same study.

DR. ALEXANDER: Oh; I'm sorry. It was the

same study. It was certainly not randomized

between nausea and no nausea.

DR. WELCH: Did you look at the time from

the start of the pain to the onset of the nausea in

the nausea group?

DR. ALEXANDER: No; we didn't.

DR. KIEBURTZ: Dr. Temple.

DR. TEMPLE: Maybe you will think this is

too much discussion, but when you separated out the

nausea people, my assumption always was you thought

the drug would work better in people that had

nausea, not less.

DR. ALEXANDER: That is exactly right. I

mentioned that--I may not have emphasized it enough

because initially Pozen believed that

metoclopramide would have an anti-nausea effect in

migraine. The thought was, we will look at those

with nausea and those without nausea.

We did that. As it turns out, if there is

an anti-nausea effect, it occurs after two hours--

DR. TEMPLE: No; I don't even mean that.

You have said that the effect on pain is better in

people with nausea.

DR. ALEXANDER: That's correct.

DR. TEMPLE: And you did, as you showed,

have groups with and without nausea separated for

analysis. But what happened was the opposite of

what you expected. Maybe that is not a major

point.

DR. KIEBURTZ: Dr. Koski?

DR. KOSKI: I assume that your patients

within this study had more than one attack of

migraine.

DR. ALEXANDER: That's not correct. This

was a single-attack study.

DR. KOSKI: It was single attack. Thank

you.

DR. KIEBURTZ: Dr. Goldstein.

DR. GOLDSTEIN: You may also want to defer

this question for later, but the preparations that

you used in these comparator studies, you went

through, or somebody went through, in the beginning

talking about how the MT100 is put together. You

have a core. Then it is sprayed and sprayed again,

and then there is another spraying on top of that.

In these studies, how is the

metoclopramide put together? Was this done with a

blind core that was then sprayed in the same way so

that the pharmacokinetics would be the same?

DR. ALEXANDER: Yes; they were identical

in visual appearance and the placebo--excuse me;

the comparators were identical and the

metoclopramide was around a core, a blank core.

DR. GOLDSTEIN: Thank you.

DR. KIEBURTZ: Dr. Katz, did you have

something?

DR. KATZ: No.

DR. KIEBURTZ: Just to remind the sponsor,

we will stop in half an hour. Just if you want to

think about your presentations, we will be stopping

at ten of the hour.

Potential Role of the MT100 in Migraine Therapy

Balancing Benefits and Risks

DR. MATCHAR: Good morning. I think, in

addition to the introduction that Dr. Alexander

gave me, I would just like to comment that I am

also a principal investigator of the three-city

study of headache management that is funded by the

Agency for Healthcare Research and Quality and that

is in an effort to link the evidence-based

guidelines that have been developed to actual

clinical practice. So it is in that context that I

will make my remarks this morning.

I guess, also parenthetically, I should

mention that I am the husband of a migrainer and

the father of a migrainer so I guess I have both a

clinical, a research and also a personal interest

in this topic.

(Slide CC-55)

My task that I have been asked to fulfil

today is to talk about the clinical trials and the

safety studies in a clinical-practice context. In

thinking about this, three questions really arose

in my mind that I felt were particularly salient.

The first is is there really a role for a

new migraine therapy above and beyond what we have

available. We have seven triptans that are out

there, for example. Do we really need something

else?

The second question is, when we look at

clinical differences in clinical-trial results of 4

to 6 percent, what, really, does that mean to

patients. Is that something really worth pursuing?

Then the third question is how should we be

thinking about benefit to risk in the particular

scenario of an acute migraine treatment.

So, in talking about these three

questions, or in addressing these three questions,

I am going to follow the following outline which is

first describing just some context of the clinical

burden of migraine, efficacy in clinical trials

focusing on the relationship between the measures

and the meaning those measures might have in a

clinical setting, and a little bit about available

oral treatments including something about adverse

effects of available treatments, and then, finally,

talk a little more about this issue of balancing

benefits and risks and a clinically useful

conceptual framework that I have, I use, and I find

useful in thinking about benefit and risk.

(Slide CC-56)

Not to really belabor the obvious to a

group of neurologists about headache, headache is

about pain. The definition from the International

Headache Society places pain as key. It is an

episodic disorder lasting 4 to 72 hours with two of

any of the following pain characteristics;

unilateral location, pulsating quality, moderate or

severe intensity and worsened by movement.

In addition, there are the associated

symptoms which were described earlier, specifically

photophobia and phonophobia together or nausea

and/or vomiting. So that constitutes a definition.

But, again, the key element from a clinical

perspective, and from the diagnostic perspective,

is pain.

(Slide CC-57)

It might go without saying that migraine

is not a homogenous disease. While pain is nearly

always present, what is less consistent is the

presence of associated symptoms. Here the

phonophobia or photophobia, the punch line,

basically, is that most people typically do have

these symptoms whereas, in the case of nausea, most

people typically don't have nausea. So the data

here is only 38 percent reported nausea or vomiting

in more than half of attacks and only 32 percent

reported nausea in all attacks. So that is just,

again, the point. The nausea is not uniformly

present and that migraine really is a syndrome with

a variable syndrome cluster presentation.

(Slide CC-58)

The question I am moving on to now is the

issue of the unmet need. I don't know if anyone

cited the statistic of 25 million people in the

United States having migraine. That is based on a

very high-quality epidemiologic study done by

Richard Lipton and colleagues.

Of those 25 million, 53 percent of these

individuals describe a disability, significant

disability, or the need for bed rest. Now, I think

this is going to be described a bit later, but

there needs to be some understanding of the true

magnitude of a migraine for most migrainers. These

are very severe headaches. They are very

disabling. In fact, a day is sliced out of that

person's life.

In addition to there being a lot of

migrainers and the disability being quite severe,

patients tend not to be satisfied with their

treatment. We will go into that a little bit

later. I will mention--I will expand a bit on the

issue of adverse effects, in particular, but there

is good evidence that patients are not getting

effective care in their early visits, that

physicians are finding it difficult to take the

medications that are available to them and create a

mix that is useful to a large majority of patients.

One of the issues at the bottom here that

is cited, and I realize it is not a FDA concern,

per se, but it certainly is a concern for our

patients, is that the medications available are

very expensive and often interfere with patients'

willingness and ability to take them regularly for

their severe attacks.

(Slide CC-59)

What do patients need? What patients

need, effectively, is what they want. What do they

want? They want pain relief. Again, this is from

a survey done by Dr. Lipton and colleagues.

Patients surveyed with migraine, they say the most

desirable outcomes in an acute migraine therapy are

rapid onset of pain relief, their freedom from pain

and there is no recurrence of pain. So it is the

notion of a sustained response to pain and

sustained response that goes into the definition of

what patients are asking for from a migraine

therapy.

(Slide CC-60)

Do clinical measures, or do measures used

in clinical trials, address what patients want?

Now, the standard measure that is used in clinical

trials is the ordinal rating system in which pain

is rated 3, 2, 1, 0 from severe to none. It is

important to point out that 3 to 2 is not

especially valuable for patients but going from 2

to 1 is something that patients would clearly

desire and, therefore, the criteria for entry into

clinical studies would be having severe or moderate

pain and the criteria for response is going from

severe or moderate to mild or none.

So the measure that is typically used, the

standard measure that has historically been used,

is pain response rate. It is the proportion of

subjects who achieve mild or pain-free status two

hours after dose when pain was either moderate or

severe at baseline and no rescue medications were

allowed in that period. But it is a two-hour

measure.

(Slide CC-61)

Let's turn to that other issue about

sustainability of the response. Let's start with

the measure I just mentioned which is a good

measure. It is a two-hour pain relief. It is a

good start. Historically, it is what has been used

as the regulatory endpoint. Triptans, for example,

were approved on the basis of the two-hour

response.

But a better response takes into account

this time-course issue that patients care about.

Sustained pain response at 24 hours includes mild

or no pain at two hours, so it is what the

preceding measure includes, but also includes to

relapse to moderate or severe pain and no use of

rescue medications. This means you get relief.

You continue to have relief.

Again, from a clinical perspective, the

notion that you are not going to have a recurrence

is extremely important because the possibility of

having a recurrence is a very ominous concern for

patients. If you know that there is a good

likelihood that this is going to come back again,

you are not going to be able to experience your day

in a normal way.

This also raises this concern about, well,

is 5 percent more people having this response

really worthwhile. I would suggest, well, if we

were only talking about 5 percent of people, or 5

percent of pain, being better, going from 100 to a

95, or going from 95 to a 90, that would not be

particularly worthwhile.

But what we are talking about is 5 percent

more people, so we are talking about people, in

this case, they get relief and they continue to

have relief. Again, this is a point of

differentiation that distinguishes MT100,

potentially, here.

At the bottom, I have here what would be

considered the best outcome which would be

sustained pain-free at 24 hours. I think this

constitutes our vision for what we would like to

see in migraine therapy and I think we are moving

towards that as a more standard measure in future

clinical trials.

(Slide CC-62)

Briefly, on the issue of associated

symptoms, we talked about the three photophobia,

phonophobia and nausea. In clinical trials, these

symptoms tend to be more commonly reported than

they are in community samples of migrainers. But,

again, even in trials, these symptoms are

associated only with a fraction of the patients.

They are recorded as present or absent so

the all-or-none measure is a relatively crude

measure of response to treatment. Again, efficacy

is assessed at two hours which has a concern from a

clinical perspective that some of these patients

who won't have nausea at the outset will start to

have nausea after and will have nausea two hours,

but then might have it relieved at three hours

after their pain is relieved.

So I think the point here really is that

the measures of associated symptoms--it is not that

associated symptoms aren't important. They are

important. But the measures that tend to be used

and are standard in clinical trials are relatively

crude and more so than the measures used for pain.

(Slide CC-63)

So what do we have currently for migraine

therapy that is oral and FDA-approved for migraine

indication? What is currently available includes,

on the left side, the over-the-counters, which are

ibuprofen, which are two products, acetaminophen,

aspirin-caffeine combination. That is one side.

On the other side, and I would say,

actually, far on the right side, are, then the

prescription medications. There are seven triptans

currently FDA-approved for migraine and the point

here is there is a paucity of approved oral drugs.

I don't know any clinicians who would say they are

particularly happy with the variety of medications

that are available.

In light of the fact that most patients

presenting to a doctor have failed over-the-counter

medications for at least their worst headaches,

then there really, truly, is a big gap in what is

available when a patient presents to you. In

effect, the only thing you have available, as a

migraine-specific therapy in this case, is going to

be the triptans.

I will mention in a moment that that is

not always a satisfactory solution for patients.

Unfortunately, what happens clinically, when this

gap is not filled with another more useful

medication, physicians are tending to use--continue

to use--narcotics and barbiturates which are

undesirable for lots of reasons, three of which are

that they have not been studied in clinical trials.

They are not FDA-approved, so that is a concern.

And they, obviously, have undesirable adverse

effects.

(Slide CC-64)

This clinical impression that there is a

therapeutic gap is supported by empirical evidence.

This is a couple of studies in which, they point

out, in the real world, half of patients will often

delay treatment with prescribed medications. They

will have a prescription in hand and 69 percent of

them will wait and see if the headache is really a

migraine. About half of them will want to take

their medication only if the attack is severe.

This is not the sign of a very healthy

environment, that people have prescriptions and

they are not wanting to take them even though they

are having, in this case, at least moderate to

severe pain.

As a consequence, I would presume, that

four out of five migrainers have expressed an

interest, a specific interest, in trying a novel

product with similar efficacy to what they have in

hand, the prescription they have in hand, but has

fewer adverse effects.

(Slide CC-65)

This then turns us to the issue of

bothersome adverse effects. Why don't migrainers

like what we have available?

On the right side, you see the non-triptan

products which include the over-the-counters I

mentioned, nonsteroidals, but also include opioids

and barbiturates. As one would expect, the side

effects are sleepiness, nausea, difficulty

thinking, inability to function, and so on.

Not too dissimilar, even, are the triptans

on the left side. But one syndrome which is

particularly bothersome to many of my patients--I

know it is extremely bothersome to my daughter--is

this chest-pressure phenomenon.

Yes; there are coronary effects of the

triptan. Some patients--and, indeed, it is

contraindicated with patients with coronary-artery

disease--but, for the vast majority of people who

are having these chest-pressure syndromes, they

have no coronary disease. These are not coronary

symptoms. What they are, again, is a bit of

conjecture, but they are extremely frightening and

most people who experience them find them

sufficiently disturbing that, even if you try to

convince them endlessly that they are not having

cardiac ischemia, they are frightened and they

won't want to take the medication.

So that is a concern and, as I say, other

symptoms are sufficiently aversive for patients

that they will delay their therapy or not take the

medications prescribed at all.

(Slide CC-66)

Now let's turn to the issue of balancing

benefits and risks of acute therapy. To think

about this, I would like you to imagine, first of

all, another scenario entirely. This other

scenario entirely is a stroke-preventive

medication.

A stroke-preventive medication might work

and it might not work. How do you know that it

doesn't work? For the most part, you know it

doesn't work because the patient has a stroke.

Okay; you lose. And that is how you know that your

drug is a failure.

Well, we have a very lucky circumstance

with migraine in that migraine lends itself to

tailoring. There are multiple episodic attacks

over many years. You get immediate feedback on the

efficacy of the acute treatment. Tailoring here,

then, is specifically aimed at maximizing the

chance that the therapy will work for a given

attack.

The idea, basically, is patients don't

like to take medications that don't work,

especially if they don't have any other effect that

you kind of like. So an opioid you might take even

if it doesn't really--well, not me, personally, or

you, personally, but, certainly, some people will

take them just because they have another effect

that they like.

Consequently, with this tailoring

occurring, the benefit-to-risk margin actually

improves over time for each of our individual

patients.

(Slide CC-67)

Recognizing that some people don't like

words as much as they like pictures, I have here a

picture that basically raises this concept as the

filter of clinical experience. We start out

basically saying, look, from population studies,

from clinical trials, we realize that not all

patients are going to respond. But we are going to

try it. We are going to treat all these patients

within some set of characteristics.

We have some set of characteristics and,

of course, it wouldn't have been approved if we

hadn't considered the benefit-to-risk to be

acceptable. Now, after some period of time,

patients decide this works under this condition,

this doesn't work under this condition, and they

pick and choose, and what we end up with is

patients taking medications for which they tend to

respond.

Consequently, the clinical benefit-to-risk

ratio improves over time and is ultimately

maximized. Again, the point I want to make is that

patients don't take drugs that don't work for the

most park.

(Slide CC-68)

As suggested earlier, from the experience

in the U.K., as Dr. Schapira mentioned as well as

using the components in the United States, the

notion is that MT100 would fill in this gap that is

currently basically being filled with opioids and

barbiturates which is a bad scenario. The notion,

again, is that, amongst the various options, what

we allow by making this new drug available is to

fill in the gap and to offer an opportunity for

patients to create a mix for themselves that makes

the most sense for them.

Not all patients, certainly, will respond

to this. Those who will respond to it will take

it. The benefit, again, as I mentioned earlier, or

the risks, will only accrue to those people who

achieve benefits.

(Slide CC-69)

So, in summary, I am going to just cover

those three questions real quickly. Is there a

role for a new migraine drug? I believe the answer

is unequivocally yes. Migraine is a common

disorder. Patients have significant unmet needs.

The available oral medications are very limited

and, unfortunately, the gap that exists is now

being filled by undesirable drugs.

The second question is what is the meaning

of the clinical-trial difference, this 4 to 6

percent everyone is talking about. Well, not

quibbling over whether you buy the 4 to 6 percent

statistical significance or not, what does

5 percent mean. Let's just say 5 percent. 5

percent is not 5 percent of pain. It is 5 percent

of people. That is an important point from a

clinical perspective. That is meaningful.

Now, the last point, or the last question,

is what is the meaning of a benefit-to-risk ratio

in clinical practice. I just want to mention again

this concept of tailoring. Migraine treatment

lends itself to tailoring. Patients don't take

drugs that don't work and thus, in clinical

practice, we have the lucky circumstance that

benefit-to-risk ratios can be optimized.

Thank you very much.

DR. KIEBURTZ: Any--Dr. Sacco?.

DR. SACCO: Dr. Matchar, just a

clarification, maybe, on Slide 63 for part of your

talk. I assume most of your talk has been

indicated for acute migraine attacks. You haven't

dealt with any of the FDA-approved medications for

migraine prevention, of which there are some.

DR. MATCHAR: Oh, sure; yes.

DR. SACCO: That would just be a

clarification.

DR. MATCHAR: Right. These are oral

products for acute indication, acute migraine.

DR. KIEBURTZ: Dr. Lenaerts?

DR. LENAERTS: Thank you. I have a

question regarding Slide 57. Could you confirm the

38 percent of patients reporting nausea and then 32

percent, actually, reporting in all attacks. I

have some other information that says up to 90

percent of people have nausea occurring. So

migrainers have up to 90 percent.

DR. MATCHAR: Right. The point that I am

making here has to do with the patterns, the

typical patterns, for patients, not the average for

all migraines. So having nausea is a typical

pattern in a minority of patients. Actually, Dr.

Silberstein did one of these studies and he might

be able to clarify that later.

DR. KIEBURTZ: Dr. Jeste.

DR. JESTE: I have a similar question. If

you look at your Slide 62, you said nausea

incidence is 40 to 70 percent.

DR. MATCHAR: Right; and this is in

clinical trials. So the population you are going

to see in clinical trials is going to be different.

So this says, basically, as a patient enters into

these trials, the presence of nausea is going to be

more likely than it was going to be when you are

asking the question, what is the typical pattern or

cluster of symptoms among migrainers. So, yes;

patients who are in trials will typically have the

symptoms more commonly.

DR. KIEBURTZ: Thank you.

DR. MATCHAR: I am going to turn to Dr.

Silberstein.

(Slide CC-70)

Dr. Silberstein is actually a colleague

working on one of the clinical trials that I

mentioned earlier and he is the Director of the

Jefferson Headache Center and the Department of

Neurology and is the President of the American

Headache Society.

DR. KIEBURTZ: We see your number of

slides in the book, but just so you are--

DR. SILBERSTEIN: I have cut them.

DR. KIEBURTZ: Perfect. Thank you.

Clinical Considerations on Migraine Therapy

DR. SILBERSTEIN: I want to thank

everybody for having us here today. Looking at the

time, I have tried to cut and I will try to talk

reasonably quickly.

(Slide CC-71)

I am going to talk a little bit about the

rationale for the use of metoclopramide. I am

going to briefly talk about attacks without nausea.

I am going to spend most of my time talking about

medication-overuse headache of which I have a

100

particular interest and then summarize a possible

benefit of MT100.

(Slide CC-72)

We learned about metoclopramide and

migraine many years ago from, actually, our

colleagues in London. Marshall Wilkenson and Nat

Blau who run the City of London Migraine Clinic

made it part of their everyday treatment and it got

introduced, like many things do, on the basis of

anecdote.

Many of us continue to use it in the

absence of trials until you saw the evidence today.

It is used to prevent nausea. It enhances the

absorption of nonsteroidals. Many headache experts

continue to use metoclopramide for those reasons.

(Slide CC-74)

I think you have seen the evidence to show

that MT100 is more effective than placebo. One can

argue about the statistics, but you see in the

evidence that MT100 is more effective than naproxen

sodium and clearly more effective than

metoclopramide.

101

(Slide CC-76)

One of issues is its 4 to 6 percent

response, clinically significant. I think, in

part, it depends on how seriously you view migraine

as a disorder. If you had a patient who has had

cancer of the brain and you had a survival rate of

10 percent and you went to 14 percent, nobody would

argue that that is clinically significant. So take

into context what migraine is to the sufferer and

take into context that migraine is often considered

not a serious disorder.

One of the ways of looking at it is to

look at all attacks and look at the absolute and

relative differences. If the 4 to 6 percent really

means in patients getting 14 to 20 percent relative

increase, and if you look at the subset of attacks

without nausea, you are assuming that the data is

correct because the subset analysis was not the

primary endpoint, you are talking about a third

improvement.

This, to me, is clinically significant.

(Slide CC-78)

102

I would like to spend a little bit of time

talking about the concept of medication-overuse

headache, for that was one of the questions. What

is it? First, many patients have chronic daily

headache which, by definition, means nothing more

than headaches occurring more than 15 days a month.

In the clinic, it is the most common cause

of chronic daily headache. I was fortunate enough

to be the head of the International Headache

Society Classification Committee on Chronic Daily

Headache. The criteria we came up with were the

following: headache has to be there more often

than not greater than or equal to 15 days per

month; regular overuse for more than three months

of acute medication; the headache is actually

developed or worsened coexistent with overuse;

lastly, you stop the overuse medication and the

headache reverts to its previous form.

(Slide CC-79)

The next issue is how much medicine.

First, triptans, ergots, opioids or

butalbital-containing analgesics taken on a regular

103

basis ten or more days per month. What we don't

mean is ten days in a row. We mean ten days

divided up. Two, other analgesics 15 or more days

a month for a total exposure of 15 or more days a

month. That is the definition of

medication-overuse headache.

(Slide CC-80)

The next issue is which are the drugs that

are most likely to produce medication-overuse

headache. The first caveat is there are absolutely

no placebo-controlled, well-designed clinical

trials of medication-overuse headache in the world,

yet. High probability based on a series of

anecdotes, opioids or narcotics, ergotamine and

butalbital-containing compounds.

Chris Diener from Germany said the best

thing he ever did was get butalbital-containing

compounds removed from the market in Germany. That

is his legacy.

Caffeine is associated with

medication-overuse headache. Lower probability;

aspirin, acetaminophen, and triptans. Unlikely and

104

controversial, other non-steroidals, DHE or

neuroleptics are even associated with

medication-overuse headache.

(Slide CC-81)

In summary, MT100 in migraine therapy. I

think it could be a primary therapy when simple

analgesics fail. By the time patients come to the

physician, they have failed simple analgesics and,

as Dr. Matchar showed, there is an area in between.

Triptans can't be used, don't work or are overused.

The reason for this is, we believe, that

nonsteroidals and neuroleptics, metoclopramide, in

particular, are unlikely to produce

medication-overuse headache. It is common among

clinicians who are interested in headache--we use

this class of drugs to prevent medication-overuse

headache or to treat medication-overuse headache.

Lastly, we believe it can fill the gap between

simple analgesics and triptans that is now being

filled by opioids and by butalbital-containing

compounds.

(Slide CC-82)

105

I think it is important to realize the

World Health Organization has said that migraine is

one of the four most disabling disorders know to

mankind and that a patient with a severe migraine

attack has the same degree of disability as

somebody who has quadriparesis, dementia or acute

psychosis.

Thank you.

DR. KIEBURTZ: Thank you.

Any clarification questions? Dr. Temple?

DR. TEMPLE: One of your slides, and a

number of people have shown the same one, was the

attractiveness of oral metoclopramide in migraine.

DR. SILBERSTEIN: Correct.

DR. TEMPLE: Counteracting gastric stasis.

DR. SILBERSTEIN: Correct.

DR. TEMPLE: Treating or preventing

nausea, enhancing absorption of NSAIDs and a lot of

people use it.

DR. SILBERSTEIN: Right.

DR. TEMPLE: I guess what are you saying

about those things? Are you saying that is part of

106

the evidence? Or what?

DR. SILBERSTEIN: What I am saying is the

following. Until these trials were done, we were

doing this on anecdote. Physicians continue to do

a number of things in the absence of evidence-based

medicine. I think what you have seen today is

evidence-based medicine. I think the questions are

going to be, there are a lot of patterns of

behavior. The pattern of behavior in the United

States today for taking care of most migraine

attacks is to either give a narcotic or opioid or

butalbutal-containing in the absence of scientific

evidence.

What I am suggesting is this is an

alternative and I think it is the job of this panel

to see whether it is a good or a bad alternative.

DR. TEMPLE: Okay. But you are not

suggesting any of those reasons are the reasons or

true or--

DR. SILBERSTEIN: I am suggesting that

this is the anecdotal lure and the basis of why

this compound has been commonly used in the past in

107

the absence of good scientific evidence.

DR. TEMPLE: Okay.

DR. KIEBURTZ: Thank you.

We will break now for fifteen minutes. I

will just remind the committee members that our

discussions only happen in public. During the

break, you are not to discuss with other committee

members or, in fact, anybody, the presentations or

your views on things. The point of having a public

meeting is our discussions are public. So, just

avoid that in the interim and we will start at

10:05.

Thank you.

(Break.)

DR. KIEBURTZ: Why don't we get started.

Dr. Bastings will be our first presenter, the

clinical team leader. We will have, just to

clarify the agenda, about an hour-and-15-minute

presentation from FDA including an invited speaker.

Then we will have time to question, for the

committee to question, both the sponsor and the

FDA.

108

Some of the questions I kind of suppressed

earlier about interpretation, context, and so

forth, that is our opportunity to do that.

So, Dr. Bastings, please.

FDA Presentations

FDA Risk/Benefit Considerations

DR. BASTINGS: Thank you. Good morning.

(Slide 1)

I will now present you some FDA

risk/benefit considerations for MT100.

(Slide 2)

As you know, MT100 is a combination of

naproxen sodium 500 milligrams and metoclopramide

hydrochloride 16 milligrams. The proposed

indication is the acute treatment of a migraine

headache with or without aura.

The division issued a not-approvable

action in May, 2004 mostly because the review team

determined that the contribution of both active

drug components to the claimed effects of the

product had not been established.

(Slide 3)

109

According to the FDA Combination Policy,

two or more drugs may be combined in a single

dosage form when each component makes a

contribution to claimed effects and the dosage of

each component is such that the combination is safe

and effective for a significant patient population

regarding such concurrent therapy.

(Slide 4)

To address the Combination Policy

requirements, Pozen conducted two factorial studies

of similar design. These were Study 301 and 304.

In both studies, patients were randomized to MT100,

naproxen or metoclopramide. The primary endpoint

was sustained pain response.

(Slide 5)

Sustained pain response is defined as a

moderate or severe headache at baseline with mild

or no headache at two hours and no relapse and no

use of rescue medication between two and 24 hours.

(Slide 6)

This slide shows you the key result of the

two factorial studies. For Study 301, the

110

sustained response rate for MT100 was 35.6 percent

as compared to 29.8 percent for naproxen. So the

contribution of metoclopramide was 5.8 percent and

this was not a statistically significant difference

according to the prespecified analysis plan.

For Study 304, which was a much larger

study, the sustained response rate for MT100 was

31.8 percent as compared to 27.9 percent for

naproxen. So the contribution of metoclopramide

was 3.9 percent and this was not a statistically

significant difference according to the

prespecified analysis plan.

(Slide 7)

This slide shows you the two-hour

endpoints in the factorial studies. I must stress

that Pozen was not required to show a contribution

of metoclopramide on these endpoints. However,

these are highly relevant endpoints in migraine

studies. These are the ones typically used to

approve migraine drugs. Since the primary endpoint

did not show a significant contribution of

metoclopramide, it is useful to examine these

111

typical endpoints.

What you can see on this slide is that, in

both studies, there was no significant difference

between MT100 and naproxen for the two-hour pain

response, the incidence of nausea at two hours, the

incidence of photophobia at two hours and the

incidence of phonophobia at two hours.

(Slide 8)

As you know, sustained pain response is a

composite endpoint. To better understand the

changes seen with that endpoint, it is useful to

look at the individual components which are the

two-hour pain response and the use of relapse or

rescue medication.

So, in Study 301, you can see that the

two-hour response for MT100 was 48.1 percent as

compared to 46.6 percent with naproxen. So the

contribution of metoclopramide at two hours at 1.5

percent. This was not statistically significant.

The use of rescue medication or the

relapse of the headache after a response at two

hours was seen in 12.6 percent of MT100 patients

112

versus 16.8 percent of naproxen patients. So the

contribution of metoclopramide there was 4.2

percent and this adds up to 5.8 percent of

difference in the sustained response rate.

(Slide 9)

In Study 304, you can see a contribution

of metoclopramide for the two-hour pain response of

3.1 percent and you see that the difference in the

relapse or rescue-medication use is less than 1

percent. This adds up to 3.9 percent of difference

in the sustained response rate.

(Slide 10)

Finally, this slide shows the sustained

responses for the associated symptoms. Sustained

responses here are defined in a similar manner as

for sustained pain response. For example,

sustained nausea-free means no nausea at two hours

with no relapse of nausea between two and 24 hours

and not use of rescue medication.

What you can see is that, in both studies,

there was no significant difference between MT100

and naproxen for sustained nausea-free, sustained

113

photophobia-free, and sustained phonophobia-free.

(Slide 11)

Pozen met with the division in October,

2004, and, at that time, they presented these

subgroup analyses which suggested a contribution of

metoclopramide in patients with no nausea at

baseline. At that time, the division considered to

accept the prospective replication of these

findings to fulfill the Combination Policy

requirements but we assured Pozen that we would

need to bring this to an advisory meeting because

this is an unusual patient population and we need

to make sure the benefits in that population

outweigh the risk related to metoclopramide.

(Slide 12)

I will briefly show you these subgroup

analyses that Pozen made. You already know that,

for the combined patient population, there was no

significant difference between MT100 and naproxen

for sustained pain response and for the two-hour

pain response.

If you look at the patients who did not

114

have nausea at baseline, you see about 10 percent

difference between MT100 and naproxen with a low

p-value. But, if you look at the two-hour pain

response, there was no significant difference

between MT100 and naproxen even in that subgroup.

For patients who had nausea at baseline,

you can see that there was less than 1 percent

difference between MT100 and naproxen for sustained

pain response. For the two-hour pain response, the

rate was actually numerically higher for naproxen

but the difference was not statistically

significant with MT100.

(Slide 13)

Similar findings for Study 304. You know

that, for the combined patient populations, there

was no significant difference for sustained pain

response and two-hour pain response. For patients

with no nausea at baseline, again, there is about a

10 percent difference between MT100 and naproxen.

For the two-hour pain response, there is no

significant difference between MT100 and naproxen

for that subgroup.

115

For patients with nausea at baseline,

there was about a 1 percent difference between

MT100 and naproxen for sustained pain response and

the two-hour pain response. These differences were

not statistically significant.

(Slide 14)

I would like to give you some thoughts on

an indication limited to patients with no nausea at

baseline. In a survey of 500 self-reported

migrainers, nausea occurred in more than 90 percent

of these patients and nearly one-third of these

experienced nausea during every attack.

Less than 10 percent of patients

consistently had migraine with no nausea at

baseline which is the indication for which MT100

which is being considered today. In line with that

survey, there was a 45 to 69 percent incidence of

nausea at baseline in the MT100 phase e studies.

(Slide 15)

Migraine patients, in the majority of

them, may have some attacks with nausea and other

attacks without or nausea may develop during the

116

attack. Patients would, therefore, need two

different treatments based on the presence or

absence of nausea or they would treat their attacks

with nausea with a combination product containing

metoclopramide which has no established

contribution for efficacy for the type of attack.

Yet, they would be exposed to the risk of

metoclopramide.

(Slide 16)

As you know, our main safety concern is

tardive dyskinesia. Tardive was originally

intended to emphasize a late appearance during

neuroleptic treatment. However, there have been a

number of reports that TD may appear early during

the neuroleptic treatment and there seems to be no

fundamental distinction between cases appearing

early and those appearing late.

(Slide 17)

In addition, there have been a number of

TD variants described and these include tardive

dystonia, tardive akathisia, tardive myoclonus,

tardive tics, tardive tremor, and it is very much

117

unclear how well these different variants have been

captured in the post-marketing reporting systems.

(Slide 18)

TD is a well-known side effect of

metoclopramide. Its exact incidence remains

unclear. There was no case reported in the MT100

database but the database was too small to detect

rare events such as TD.

(Slide 19)

The current metoclopramide labeling

includes a warning which I am going to read to you.

"Tardive dyskinesia may develop in patients treated

with metoclopramide. Although the prevalence of

the syndrome appears to be highest among the

elderly, especially elderly women, it is impossible

to predict which patients are likely to develop the

syndrome. Both the risk of developing the syndrome

and the likelihood that it will become reversible

are believed to increase with the duration of

treatment and the total cumulative dose. Less

commonly, the syndrome can develop after relatively

brief treatment periods at low doses. In these

118

cases, symptoms appear more likely to be

reversible."

(Slide 20)

Because of these safety concerns, the FDA

limited the indication of oral metoclopramide for

short-term therapy for gastroesophageal reflux for

up to 12 weeks and only when conservative treatment

fails for the treatment of diabetic gastroparesis

for up to eight weeks. The recommended dose is 5

to 15 milligrams up to four times a day.

(Slide 21)

As I mentioned earlier, there have been a

number of cases reported in the literature of the

relatively short durations of treatment, sometimes

as short as one or two weeks. We also asked our

colleagues from the Office of Drug Safety to look

for cases of movement disorders associated to

metoclopramide in the AERS database and you will

hear a presentation with much more detail on that

topic later in the morning.

In that analysis, the first quartile of

duration of treatment for the cases of TD was 19.5

119

days.

(Slide 22)

This slide show you a breakdown of the

treatment duration. You can see that there are

quite a few cases with duration of treatment less

than the 90-day definition which has been used in

some of the earlier studies.

(Slide 23)

We also asked our colleagues from ODS to

look at the patterns of use of metoclopramide.

Metoclopramide is mostly used for GI indications.

Migraine use, up to now, is quite limited. It is

less than 2 percent. 13 percent of patients

appeared to have received prescriptions for more

than 90 days and 7 percent of patients for more

than 180 days, so exceeding the labeling

recommendations.

Over a three-year period, cumulative

therapy was longer than 90 days for almost 20

percent of patients and greater than 180 days for

over 10 percent of patients, again exceeding the

recommendations.

120

(Slide 24)

What about the risk of TD associated with

chronic intermittent use of metoclopramide as has

been proposed in this NDA. This is very difficult

to evaluate for a variety of reasons which include

that there is no current indication for chronic

intermittent use in the United States and that

there is no specific capture of chronic

intermittent use in the AERS database.

Some animal data suggests that the

intermittent use of neuroleptics may be no safer,

or even riskier, than continuous use in an animal

model of TD. In the psychiatric population, the

number of interruptions in chronic treatment--so

this is slightly different from chronic

intermittent but this may be suggestive--the number

of interruptions was the second factor after age in

predicting the occurrence of TD.

(Slide 25)

Another concern that we have is the

overuse of acute migraine drugs.

Medication-overuse headache was recently introduced

121

in the IHS classification. There is a subcategory

of analgesic-overuse headache. According to

experts, there is substantial evidence that all

drugs used for the treatment of migraine may cause

medication-overuse headache and the prevalence of

medication-overuse headache in the general

population is around 1 percent.

The IHS also said that the overuse of

symptomatic migraine drugs is the most common cause

of chronic daily headache. We are not that much

worried that MT100 could cause chronic daily

headache. We are just worried that there could be

a similar abuse of the drug as has been seen with

the other approved or non-approved migraine drugs.

(Slide 26)

So we have the following questions for the

advisory committee. The first one; in a recent

submission to the NDA, Pozen estimated an annual

incidence of TD of up to 0.038 percent for

metoclopramide at a daily dose of 30 to 40

milligrams per day for 72 days per year which

corresponds to up to 380 cases of TD per million

122

patients per year.

Do you think that this is a reasonable

estimate? If MT100 were to carry the same risk,

would such a risk level be acceptable if the only

contribution of metoclopramide is a 5 to 10 percent

improvement on sustained headache relief with no

effect onto our endpoints? Is any risk of TD

acceptable for a migraine population?

(Slide 27)

Question 1; is there sufficient evidence

that the chronic intermittent administration of

metoclopramide does not carry a risk of TD? Is it

possible to define a maximum recommended number of

monthly doses of MT100 to avoid the risk of tardive

dyskinesia?

(Slide 28)

Question 3; do you believe that, based on

the existing data on medication-overuse headache,

there is evidence that a proportion of patients

prescribed MT100 will likely take a number of

monthly doses higher than recommended?

(Slide 29)

123

Question 4; all currently approved acute

treatments of migraine are indicated without

restriction regarding the presence of absence of

nausea at baseline. Given that patients may have

nausea at some attacks and no nausea at others,

does an indication limited to the subpopulation of

migraine patients with no nausea at baseline

represent a clinically meaningful and acceptable

indication?

(Slide 30)

The last question; if Pozen shows

prospectively in a new clinical study in migraine

patients with no nausea at baseline a significant

contribution of metoclopramide on sustained

headache pain relief of 5 to 10 percent with no

contribution at two hours and no contribution on

relapse rates or rescue-medication use in the two

to 24 hour period, would the demonstrated benefit

outweigh the risk related to TD? If not, what

additional data or desired primary outcome, or

desired effect on sustained relief, could provide

evidence of safety and efficacy?

124

(Slide 31)

Finally, I would like to thank the

following FDA colleagues who have contributed to

this presentation . Thank you for your attention.

DR. KIEBURTZ: Thank you. Same deal. If

there are some clarifying questions. Let me just

point out some things about the questions which Dr.

Bastings has presented. After the public hearing

this afternoon, that will be the time we have to

spend a great deal of time discussing these.

We can clarify points of these questions

at that time rather than at this time because it

will be immediate to our discussion.

I would just add at this point my approach

to this, or our approach to this, should be that

there are some assumptions made in here. We are

not debating whether those assumptions are good

ones or bad ones. The question to us is, if that

was assumed, how would you think. This is what Dr.

Katz referred to earlier in terms of this being

somewhat hypothetical.

I think we could spend a lot of time

125

arguing about whether the assumptions are good ones

or not. I don't think that is the meat of the

matter here. If one assumed those things, then how

would you make decisions about that. I just want

to put that out there.

There are some questions that are asked of

us about estimates and whether those are

reasonable. But, again, much of the clarification

on the questions, I think it would be better to do

at the time we discuss the questions individually

unless you have a burning question about those.

Certainly, they are open to questions about the

rest of Dr. Bastings' presentation.

Dr. Green.

DR. GREEN: I have a regulatory question

about the Combination Policy. It has to do with

the contribution of each drug and what is

acceptable. Is one drug increasing the

bioavailability of another? How would that be

interpreted?

DR. KIEBURTZ: Dr. Katz or Dr. Bastings?

DR. KATZ: I think the contribution, as

126

defined in the reg--it is ill-defined in the reg.

It just says "some contribution." So it could be

in any one of a number of clinical areas, safety,

efficacy. But, in and of itself, increasing the

bioavailability probably wouldn't be particularly

helpful unless that resulted in some sort of

clinical advantage; faster onset, or more sustained

onset, or fewer side effects.

So, in and of itself, increasing the

bioavailability in a typical case, at least off the

top of my head, wouldn't be, necessarily,

considered useful.

DR. TEMPLE: But you can imagine cases,

and there have been cases, where inhibiting

metabolism might lead to a more sustained and less

variable blood level. In some sense, what does

carba dopa do? So there could be cases. But, as

Russ says, you would have to weigh the disadvantage

of adding another therapy and, if the alternative

is just tasking 20 percent more, you would probably

find that not worth it.

DR. KIEBURTZ: Okay. Why don't we go on.

127

Our next speaker is Dr. Jinnah from Johns Hopkins

Hospital.

Overview of Tardive Dyskinesia

DR. JINNAH: Good morning.

(Slide)

Thanks for the invitation to come and

speak here. I have been asked to give a brief

summary of the clinical condition of tardive

dyskinesia.

Normally, my presentation on this topic

would be about an hour, but I am going to limit my

comments to ten minutes and, in so doing, I am only

going to be able to touch on the highlights. I am

going to skip a lot of details but I can certainly

answer questions later if necessary.

(Slide)

So, with that, let me just proceed to

review this topic. The term "dyskinesia" refers to

any abnormal movement and the term "tardive" refers

to late or delayed. What I would like to do is

first address the nature of the abnormal movements

and then go on to describe when it occurs.

128

(Slide)

The movements vary quite a bit depending

on different patients. By far the most common

abnormal tardive syndrome is the so-called

buccolingomasticatory syndrome, which is a bit of a

mouthful, but it basically refers to abnormal

movements of the face and tongue. I will show you

an example of this on videotape in just a second.

Less common tardive movement disorder

syndromes include the ones listed there including

tardive dystonia, which refers to mainly twisting

and bending movements, tardive chorea, which

resembles dancing, tardive tourettism which

resembles Tourette syndrome, tardive tremor or

myoclonus which, simply put, are shakes and jerks.

In addition to this group of broader

movement disorders, there are some tardive

syndromes that are not necessarily classified as

abnormal movements but, rather, psychological or

psychiatric manifestations. These include

akathisia, which is a sense of severe restlessness

that prevents people from sitting still. It

129

includes unusual tardive pain syndromes which have

an unusual anatomical distribution that include the

oral or perineal regions, and, finally, respiratory

irregularities referred to as tardive respiratory

dyskinesias.

Now, most clinicians will recognize the

most common form here and that is the top one, the

face-and-tongue syndrome. But the less common

forms are far less well appreciated.

(Video)

Let me show you an example of two of

these, first the most common one. This is a

videotape of a man who has two problems. I hope it

is not too small to see. If you can see his mouth,

it is in constant motion, jaw, lips and tongue. He

came complaining that he had trouble talking, he

had trouble eating and he was biting his tongue.

You can also see his hand. One of his

hands is shaking. He has a tremor that resembles

Parkinson's disease. This man got his condition

after two years of metoclopramide use. He was,

unfortunately, unaware that he was only supposed to

130

be taking the medication for three months and his

doctor was unaware that metoclopramide was on the

list of medications that can cause a tardive

syndrome like this.

(Video)

Let me show you a second example of the

tardive dystonia. You can see this man's problem

is much more severe and, perhaps, more disabling.

This is an example of dystonia or the

bending-and-twisting syndrome. You can see that he

has great difficulty standing up straight because

of extreme arching of his back and backward bending

of his head and neck.

Here you will see a closer view of the

nature of this problem. He is like this more than

80 percent of his waking hours, standing or seated.

He gets relief only if he lies down. He can only

temporarily bring his head to the midline position

voluntarily and then it just goes back and it is

too much effort to keep straight.

This man got his condition from the use of

a classic neuroleptic agent which is a more common

131

source of this problem.

(Slide)

So what exactly causes these syndromes?

It is widely recognized that they are most commonly

caused by the neuroleptics. These are the

dopamine-receptor-blocking agents that Dr. Schapira

alluded to in the initial presentation.

Essentially all classes of neuroleptics will cause

this syndrome although some are less likely to

cause it than others.

These are widely recognized but what is

less well recognized is a much longer list of

agents that also have the potential to cause these

syndromes. On this list include anti-emetics such

as metoclopramide and prochlorperazine. Other

medications used in psychiatry such as

anti-depressants and several others of which I have

provided a partial list here.

(Slide)

So when do these problems actually arise?

They are referred to as tardive syndromes because

they usually require chronic administration of the

132

offending drugs. They occur with a wide range of

prevalence according to different studies from less

than 5 percent to more than 50 percent with an

overall average being somewhere in the 20 percent

range of chronically treated patients.

These numbers here were not derived

specifically from metoclopramide use but rather

from all neuroleptics and other drugs capable of

causing tardive syndromes. The incidence is

estimated to be about 5 percent per year during

chronic daily treatment. The treatment duration is

somewhat arguable and varies from report to report

and definition to definition. The most

conservative one is that treatment requires at

least three months of constant therapy.

But, as noted before, there are lots of

cases out there who have developed tardive

syndromes with much shorter durations of action,

sometimes just a few days. We can then ask whether

or not these disorders should be classified as

tardive syndromes or not, but that is generally not

done in most of the epidemiological reports or

133

collecting databases.

There are some known risk factors of

developing this condition. They include older age,

female gender and several other less

well-understood phenomena such as duration of

treatment, the total cumulative dose of treatment,

prior brain injury or other organic problems,

diabetes, mood disorders and others.

(Slide)

How are these treated once they arise? In

general, these recommendations come from psychiatry

where this problem is most prevalent. They refer

to all neuroleptics, not specifically to

metoclopramide. Generally speaking, the

recommendation is to attempt to discontinue the

offending agent if possible.

If it is not possible, or if

discontinuation does not cause resolution of the

syndrome, as it often does not, there are several

other things that are often tried, but there is

very little evidence supporting a beneficial effect

of any of these. Individual patients may respond

134

somewhat to one or another of the items on this

list, but, by no means, can these be considered

reliable treatments.

Instead, at least in the psychiatry

literature, the belief for the use of neuroleptics

and related offending medications that can cause

tardive syndromes is that prevention should be one

of the key aspects of treatment.

To prevent the disorder, the general

guidelines are that the neuroleptics should not be

used unless there are no other alternatives. When

they are used, they should be used at the lowest

dose possible. Some even recommend intermittent

withdrawal of the neuroleptic to make sure that

ongoing therapy is still needed. Since this

disorder is quite difficult to treat, prevention is

really quite an important element.

I believe that is all I have here and if

there are any questions, I could take them.

DR. KIEBURTZ: Dr. Jeste.

DR. JESTE: One question. Did you imply

that antidepressants and antibiotics cause tardive

135

dyskinesia?

DR. JINNAH: Causation is difficult to

establish. If you look in the literature, you will

find many cases of tardive dyskinesia reported that

are due to a whole variety of different

medications. The frequency of some of these other

medications--for example, antidepressants or

antibiotics, or I should say the frequency of the

reports, is quite low. We could argue whether or

not the patients who were presented in those

reports really were tardive dyskinesia or not.

I am not passing judgement on the

diagnosis of those. But I think it is generally

well-accepted that tardive dyskinesia does not just

come from neuroleptic medications and that was my

point.

DR. JUNG: Can you clarify what mood

disorders are associated with the development of

tardive dyskinesia?

DR. JINNAH: There appears to be a slight

epidemiological risk associated with affective and

schizo-affective disorders as opposed to, for

136

example, schizophrenia.

DR. JUNG: So that includes just general

depression which is very prevalent in the

population.

DR. JINNAH: It does.

DR. JUNG: Thank you.

DR. SMITH: When a case comes up where

isn't chronic exposure to a drug, is that typically

categorized, then, as a dyskinesia as opposed to a

tardive dyskinesia?

DR. JINNAH: It is a very good question.

I think different experts would answer you

differently here. Some people use a very strict

criteria that it has to be at least 30 days or you

call it something else, dyskinesia and acute

abnormal syndrome. Some people are a little bit

less strict in their criteria and say, well, if it

looks and behaves like tardive dyskinesia, then,

perhaps, one week exposure is sufficient. But these

are not generally agreed-upon timetables.

DR. JUNG: You didn't talk about this on

your slides, but could you discuss a little bit

137

about the kindling phenomenon with intermittent

use, or is this an appropriate time?

DR. KIEBURTZ: Can we hold that one for

the general question session.

Dr. Sacco?

DR. SACCO: On Slide 3, I think it is

Slide 3 of yours, can you clarify what you mean

by--it is actually blurred on my

page--restlessness.

DR. JINNAH: Restlessness is exactly what

that sounds like. These patients report that they

can't sit still. When you watch them, they often

rock in their chair. They stand up. They pace

around. They just can't sit still and they

describe a severe inner sense of restlessness.

DR. SACCO: Thank you.

DR. KIEBURTZ: Just a point of

clarification, a follow up on Dr. Jeste's comment.

Linking exposure to a phenomenon or establishing

causation is a long road. There are various

degrees of that road being established for various

agents in tardive dyskinesia.

138

But, if I understood you properly, would

you say that it is generally accepted that

metoclopramide, as an agent, can cause tardive

dyskinesia?

DR. JINNAH: I believe most would agree

with that.

DR. KIEBURTZ: Thanks.

Our next speaker is Mary Ross Southworth.

Post-Marketing Review of Movement

Disorders and Neuroleptic Syndrome

Associated with Metoclopramide

DR. SOUTHWORTH: Good morning.

(Slides 1 and 2)

As we have heard this morning, MT100 is a

combination of metoclopramide and naproxen that is

being evaluated for the treatment of acute

migraine. The proposed dosing is an a chronic but

intermittent matter based on episodes of migraine.

The proposed dosing of the drug recommends no more

than six tablets be used per month and more than

one tablet used per single migraine episode.

We were interested in looking at this

139

chronic intermittent dosing, whether we could look

in our adverse-event database to see whether we

could elucidate the risks associated with this kind

of dosing.

(Slide 3)

I think we are pretty clear on the fact

that metoclopramide is well known to cause movement

disorders. In fact, the program labeling is

specific on several movement disorders including

extrapyramidal symptoms, Parkinsonian symptoms,

tardive dyskinesia and neuroleptic malignant

syndrome. The labeling for metoclopramide

recommends a daily dose of 5 to 20 milligrams QID

with a duration of therapy not exceeded 12 weeks.

(Slide 4)

This slide shows number of prescriptions

dispensed for metoclopramide over about the last

ten years. You can see that it exceeds 7 million

in the year 2004. This jump in number of

prescriptions dispensed in 2000 coincides with the

withdrawal of cisipride from the market.

You have to keep in mind that the numbers

140

represented here are prescriptions dispensed, not

patients. Also keep in mind that, although this

usage data slide only extends for ten years, my

adverse-event review will extend farther than that.

(Slide 5)

When developing our case series for

metoclopramide-associated movement disorders and

neuroleptic malignant syndrome, we wanted to focus

on several points. First, could we ascertain what

the reversibility of the reaction was, whether it

be treatment with another pharmacologic agent or

withdrawal of the offending drug.

We also were very interested in

associating the dose and duration reported in the

adverse-event reports, themselves, to the proposed

dosing for MT100 which, as you have heard, is in a

chronic intermittent manner. We also wanted to

focus on any associated risk factors that were

apparent in the cases such as concomitant drugs or

concomitant disease states that the patients might

experience.

(Slide 6)

141

So the purpose of our review is to

characterize the cases of some specific adverse

events that were reported in the adverse-event

reporting system, or the AERS database, that were

associated with metoclopramide.

(Slide 7)

The AERS database is a computerized

database which contains reports of adverse events

for all U.S. marketed drugs. It contains over 3

million adverse-event reports. The reporting in it

is largely spontaneous meaning that healthcare

providers are not compelled to report adverse

events. However, sponsors, when they become aware

of adverse events through a variety of sources, are

required by regulations to report those to the

agency.

Consequently, the source of the reports,

for the most part, come from sponsors. However,

there are a good number that come from healthcare

providers or lay people like consumers, patients,

patient's families or lawyers.

(Slide 8)

142

The left side of the screen shows the

adverse events that were specifically searched for

in our database. They included neuroleptic

malignant syndrome, acute dystonia, akathisia,

Parkinsonism and tardive dyskinesia. For each of

these adverse events, we looked at the total number

of case reports with specific focus given to daily

dose of metoclopramide, the duration of treatment,

any risk factors that might be present and the

reversibility of the reaction.

(Slide 9)

In order to do this, we ran a search of

the database using each of the movement disorders

and NMS as a search term plus metoclopramide. We

classified the cases into movement-disorder

categories based on the diagnosis in the case,

itself. I think it is pretty clear that there is

substantial overlap between some of the reporting

of the different movement disorders and, in order

to keep clarity, we just used the diagnosis that

was used of the case thinking we could capture the

most number of cases that way.

143

Some points to remember when reviewing our

case series were there could be what could be

considered case misclassification because some

cases might have reported a movement disorder after

several doses of drug but may have caused the

tardive dyskinesia where it could have possibly

been called an acute dystonia. But we used the

diagnosis made in the case.

Another thing to remember is that the way

cases are reported in AERS, we frequently know

dose, duration or frequency of the dose given but

we very rarely know whether the dose was given

continuously or intermittently.

Obviously, because these are labeled

adverse reactions, there is going to be

under-reporting of adverse events and also because

the drug has been on the market for a long time, we

are not going to get a maximum number of reports of

adverse events.

The quality of the reports varied, as you

might expect. There are several data points that

seem to be more inconsistent and some of those

144

included status of recovery. Some cases may not

have reported whether the patient recovered or not,

and also time to that recovery.

(Slide 10)

This slide shows the number of reports we

retrieved for each of the adverse events we

searched for. There were 37 cases of NMS, 203

cases of acute dystonia, 57 cases of akathisia, 35

cases of Parkinson's and 68 of tardive dyskinesia.

(Slide 11)

Using those reports, we developed our case

series which I will present to you. The case

series is going to include demographics of the

patients and clinical characteristics including any

information we have on recovery. I am also going

to spend some time talking about cases that

reported continuing symptoms at the time of

reporting. I will present some representative

cases and then focus a little bit on cases

associated with short-term metoclopramide therapy.

(Slide 12)

The first adverse event that will be

145

presented is neuroleptic malignant syndrome. We

had 37 unique cases. The age represented was a

mean of 49. The range of daily dose ranged from

7.5 to 80 milligrams with a mean of 33, mostly I.V.

dosing represented here and mostly G.I.-related

indications which will be very common in the next

few slides. The range of duration of therapy was

from 1 to 196 days with a median of three days.

(Slide 13)

In these 37 cases, concomitant medications

that were associated with the development of NMS or

NMS-like symptoms was reported in 20 and they

included anti-depressants, anti-emetics and

anti-psychotics.

One thing to remember is that not all

cases reported whether there were concomitant

medications or not, so I have just provided

information on the cases that have.

Drug therapy was used to treat the adverse

event in 18 cases and it largely consisted of what

would be considered standard of care for

neuroleptic malignant syndrome. The symptoms were

146

reported or resolved in 11 of the cases. The

symptom was reported as continuing in one of the

NMS cases but the symptom that was reported as

continuing was more of a dystonic jaw clenching.

In this series, eight patients died.

(Slide 14)

To look a little bit closer to the

patients that died, in those eight patients, the

daily dose of metoclopramide ranged from 10 to 40

milligrams with a mean of 32, kind of a mix of oral

and I.V. dosing used, and the duration of therapy

was short and ranged from two days to 15 days.

(Slide 15)

The first movement disorder, in our view,

is acute dystonia. There were 203 unique cases.

Acute dystonia was reported in a younger population

with a mean age of 32. The range of daily dose was

0.6 to 800 milligrams with a mean of 71 milligrams.

Largely oral dosing reported here. The range of

therapy from one dose to over 2000 days but a short

median duration of therapy of two days.

Again, you see mostly G.I. symptoms being

147

treated here although there were a sizeable number

who were getting pre-treatment for chemotherapy

with metoclopramide.

(Slide 16)

For these 203 acute dystonia cases, there

were 64 cases which reported concomitant

medications that were associated with movement

disorders, mostly anti-depressants and

anti-emetics. Drug therapy was used to treat the

adverse event in 115 cases. For these acute

dystonia cases, 115 cases reported the symptoms as

improved or resolved. But symptoms were reported

as continuing in 12 cases.

(Slide 17)

In those 12 cases which reported

continuing symptoms, the daily dose ranged from 10

to 40 milligrams with a mean of 25, mostly oral

dosing, and duration of therapy ranged from one day

to over 2000 days with a median of 2.5 days.

(Slide 18)

We have 57 unique cases of akathisia. The

mean age seen in this case series was 45. The

148

daily dose ranged from 5 to 200 milligrams with a

mean of 42, mostly oral dosing again. Duration of

therapy ranged from one over 2500 days with a

median duration of 17 days. Again, mostly G.I.

indications.

(Slide 19)

For these 57 cases akathisia, concomitant

medications associated with movement disorders was

reported in 23. Drug therapy was used to treat

akathisia in 29 cases. Symptoms were reported as

improved or resolved in 31 cases but symptoms were

reported as continuing in nine cases.

(Slide 20)

In the nine cases that reported continuing

symptoms, the daily dose ranged from 8.6 to 40

milligrams with a mean of 25 milligrams, mostly

oral dosing. Duration of therapy ranged from 17 to

over 2500 days with a median duration of 119 days.

(Slide 21)

We reviewed 35 unique cases of

Parkinson's. This was in an older population with

a mean age of 60. Daily dose ranged from 10 to 80

149

milligrams with a mean dose of 36 milligrams per

day, mostly oral dosing. The duration of therapy

ranged from one to over 1400 days with a median of

60 days.

(Slide 22)

In these 35 cases of Parkinson's, there

were 13 cases which reported concomitant

medications that were associated with movement

disorders. Drug therapy was used to treat the

adverse event in 18 cases.

Symptoms were reported as improved or

resolved in 15 of the cases but symptoms were

reported as continuing in eight cases.

(Slide 23)

In those eight cases that reported

continuing symptoms of Parkinson's, the daily dose

ranged from 20 to 40 milligrams with a mean of 32,

mostly oral dosing, and the duration of therapy

ranged from one day to 203 days with a median

duration of 81.

(Slide 24)

There were 67 cases of tardive dyskinesia.

150

The mean age was 57. The daily dose ranged from 5

to 80 milligrams with a mean of 35, mostly oral

dosing again. Duration of therapy ranged from one

to over 4700 days with a median of 180 days, again

G.I. indications for the metoclopramide.

(Slide 25)

25 cases reported that the patient was

taking concomitant medications associated with

movement disorders and drug therapy was used to

treat the adverse event in 19 cases. In 12 of

these cases, symptoms were reported as improved or

resolved. In 20 cases, symptoms were reported as

continuing.

(Slide 26)

In those 20 cases with continuing

symptoms, the daily dose ranged from 5 to 80

milligrams with a mean of 53, mostly oral dosing

and a duration of therapy from one to over 4700

days with a median of 165 days.

(Slide 27)

After we developed our case series, we

wanted to look at a more focused group of these

151

cases to see if we could approximate the dosing

seen in the MT100. So we looked at two further

subgroups of our case series. One, we looked at

characteristics of cases that specifically reported

symptoms as continuing at the time of reporting.

Then we also looked at cases with the diagnosis of

tardive dyskinesia that were related to short-term

metoclopramide therapy.

(Slide 28)

There were 50 cases out of 400, over 400,

that reported continuing symptoms. They were

represented by eight Parkinson's, 20 tardives, nine

akathisias, 12 acute dystonias and one NMS which

was actually likely a dystonic reaction. A little

over half of the cases with continuing symptoms

reported a duration of therapy of greater than 30

days.

(Slide 29)

So 15 cases in our series reported

continuing symptoms with a duration of therapy of

less than 31 days. Eight of those cases reported

continuing symptoms with a duration of therapy of

152

less than three days, what we would call very short

durations of therapy. That included one

Parkinsonism case, two tardive cases, four acute

dystonias and one NMS. Most of those eight cases

occurred after at least three doses of

metoclopramide.

(Slide 30)

I have a few representative cases to

describe to you. The first one is a 49-year-old

female who received two doses of metoclopramide, 20

milligrams orally, over two days for treatment of

gastric reflux. Concomitant therapy included

cimetidine. On Day 2 of therapy, she developed

dystonic reactions consisting of torticollis and

trismus. Her dystonic reaction was reversed by

diphenhydramine. However, she subsequently

complained of left-sided weakness and temporary

loosing of the teeth.

(Slide 31)

The second case is a 34-year-old female

with nausea who received metoclopramide, 10

milligrams, orally three times a day for three

153

doses and experienced difficulty breathing,

extremity shaking, head and neck jerking back. She

went to the E.D. where she was treated with

benztropine, after which she started to relax.

However, symptoms still occurred. She was

subsequently treated with lorazepam and paroxetine

which did not completely relieve the symptoms. She

was seen in the E.C. and by neurologists several

times for reactions milder than the first reaction.

Approximately three months later, she still suffers

from head pain, dizziness, tingling, pressure,

fatigue, agitation, involuntary shaking, muscle

spasm and neck pain among other symptoms.

(Slide 32)

The third case, a 27-year-old male

received three doses of metoclopramide, 10

milligrams orally, over two days for diabetic

gastroparesis. He experienced a dystonic reaction

with psychotic tendencies, agitation and agitation

with suicidal tendencies on the second day of

therapy.

He was treated in the E.D. with

154

diphenhydramine and lorazepam. Once discharged, he

continued to have symptoms of inability to

concentrate, slowed mental processing, difficulty

focusing, eye strain, vertigo, loss of equilibrium,

fatigue, dizziness and hallucinations.

(Slide 33)

The second subgroup analysis of our case

series that we did looked at specific cases with

the diagnosis of tardive dyskinesia that were

associated with short-term therapy of

metoclopramide of less than 30 days. What we were

trying to look at here was trying to approximate

what kind of dosing regimen would be seen with

chronic overusers of migraine therapy because there

is a certain population of migrainers who might use

this drug prophylactically in a manner similar to

other migraine therapies.

We chose tardive dyskinesia as our adverse

event because the diagnosis of tardive dyskinesia

infers a long-term or permanent adverse event. We

also noted that about 25 percent of our cases had a

duration of therapy of less than 30 days.

155

(Slide 35)

You have seen this slide before, but you

can see that this is the distribution of our cases

based on the duration of therapy of metoclopramide.

The large number are reported with durations of

therapy of less than 90 days, but there is a

significant number with durations of therapy of

less than 30. Actually, there are 15 such cases of

tardive dyskinesia with a duration of therapy of

less than 31 days.

Of these 15 cases, the status of recovery

was not reported in nine of them. Symptoms were

reported as resolved in one case but continuing

symptoms were reported in five of these cases.

Some of the characteristics of these cases include

two out of the five cases reported symptoms as

continuing but improved. Two out of the five cases

reported I.V. dosing and four out of five cases

reported daily doses of 40 milligrams.

The important thing to remember is, again,

we are not really able to discern, because of the

AERS data, whether this was chronic intermittent

156

use of metoclopramide or chronic continuous use of

metoclopramide in these cases.

(Slide 36)

In fact, we found no cases in AERS that

specifically linked intermittent use of

metoclopramide with any movement disorders. There

are maybe several reasons for this. First, and

probably most likely, is that AERS--the way data is

reported in AERS does not make the distinction

about intermittent dosing so it just wasn't clearly

described in the report.

It could be that intermittent dosing is

not commonly used or the adverse events seen with

intermittent dose are not commonly reported or that

there may be few movement-disorder-related adverse

events with intermittent dosing.

(Slide 37)

So, in conclusion, most of the reports

that we saw with continuing symptoms of the adverse

event involved long-term therapy of greater than 30

days with the caveat, again, that we didn't know

whether it was intermittent or continuous therapy.

157

There were eight cases which reported

continuing symptoms with very short-term therapy.

There were five cases of tardive dyskinesia that

were associated with therapy of less than 30 days.

Concomitant medications associated with movement

disorders were frequently present in the cases and

there were two out of eight deaths from neuroleptic

malignancy syndrome that occurred after less than

three days of therapy.

(Slide 38)

That's it.

DR. KIEBURTZ: Thank you.

Questions or clarifications for our last

speaker?

DR. KIEBURTZ: Dr. Jeste.

DR. JESTE: In these cases of acute

dystonia and neuroleptic malignant syndrome, some

of the patients have some concomitant therapy as

shown, in these cases, the side effects occur after

metoclopramide therapy.

DR. MATCHAR: Right; they were temporarily

associated with metoclopramide. Yes.

158

DR. KIEBURTZ: Dr. Porter.

DR. PORTER: Metoclopramide is not widely

used in the U.S. You didn't spend a lot of time on

the primary diagnosis. I presume that you found no

migraine-related movement disorders in this search.

DR. MATCHAR: There were very, very few

cases that reported adverse events related to

migraine. As you saw from the indications, they

were mostly G.I.-related indications.

DR. KIEBURTZ: Dr. Welch.

DR. WELCH: Were there any different

characteristics of the patients who had the T.D.

after the short-term as opposed to the long-term?

It is almost like a biphasic population response

there.

DR. MATCHAR: No; it was a very

heterogeneous population. There were different

presentations, different durations of therapy, so I

am not really sure whether you could say those that

experienced it earlier had similar characteristics

than those that presented after 90 days of therapy.

DR. KIEBURTZ: Dr. Katz.

159

DR. KATZ: Just one point of clarification

which I think is true. Maybe you said it and I

missed it, but, obviously, we focused on the

patients with continuing symptoms, at least in

part. As a general matter, I think, for these

reports, and correct me if I am wrong, what we

don't have which would be nice to have in terms of

information about those cases is sort of the

latency between the onset of the event, the

movement disorder, and the time of the report.

So it could be that an event happened on

Day 1 and the report is made on Day 2, in which

case, it might be continuing. But if the report

had been made on Day 47, after the drug had been

discontinued, for example, it might be that they

were discontinued. So I think, as a general

matter, we don't know this duration of continuation

of symptoms because we don't know the link between

when the event happened or stopped and the time of

the reporting.

There were a few cases, I think, where we

do have that but I think, in many cases, we don't.

160

DR. MATCHAR: Right.

DR. KIEBURTZ: Dr. Hughes.

DR. HUGHES: You mentioned under-reporting

in the AERS database, a well-known problem. But

when you have an adverse event which is labeled,

what sort of characterizes the types of adverse

events that might then be reported? We are looking

at a very peculiar set here.

DR. MATCHAR: Probably. There were a lot

of reports from lawyers. I mean, that is one.

Looking at other case series that I have done, it

seemed like there were a lot of reports from

lawyers. But we looked at, actually, quarters of

years, the reports, and there really didn't seem to

be any change, like an increase in reporting after

cisipride came off. It seemed to be fairly steady

so I am not sure that I could say that there was

one specific thing.

DR. KIEBURTZ: Good. Thank you.

Questions from the Committee to the Sponsor and FDA

Now, we have time to ask questions of the

presenters without me suppressing you about them to

161

be about clarifications. We have about an hour to

do that. I suggest we start principally with

questions to the sponsor and immediately with those

individuals who I suppressed.

I recall stopping Dr. Goldstein when you

were asking Dr. Schapira a question about the U.K.

reporting database, I believe, and Dr. Lenaerts, I

stopped you in the middle of a question, too, but I

can't remember the context.

DR. LENAERTS: It was in the same context.

DR. KIEBURTZ: Dr. Goldstein, would

you--and you can draw anyone from the sponsor or

from the FDA if you would like to them to repeat or

present material again.

DR. GOLDSTEIN: I guess what I was

actually asking for was clarification about the

validity and accuracy and reporting rates in the

U.K. system, especially now as contrasted to the

last presentation we had from data here in the

United States which is a similar sort of reporting

system, but the numbers seem to be quite different.

DR. SCHAPIRA: Thank you. The U.K.

162

reporting system to the Committee of the Safety of

Medicines, the so-called yellow-card system, is a

system by which physicians send in to the CSM

documentation of an adverse event. So it is

physician-led. It is not spontaneous in terms of

including patients. But it is spontaneous in the

sense that physicians have to send in the yellow

card.

Those that do send in yellow cards often

get a response back from the CSM asking for further

clarification if all the relevant information is

not included in the original yellow form that they

have submitted.

As for a proportion of reporting to all

potential cases, of course, I can't comment on

that. I don't know the data. Obviously, that is

not possible to obtain.

DR. GOLDSTEIN: The second part of that

question--it was sort of a two-parter--was, in

terms of validating the conditions that are being

reported, we have this clear problem with

categorizing a lot of these movement disorders.

163

There is this point of definition about how much

exposure is enough exposure to be categorized as a

given type of condition. Are these validated in

any way, or is it just based on individual

reporting.

DR. SCHAPIRA: It is predominantly based

on physician reporting.

DR. KIEBURTZ: It sounds analogous to what

we heard from Dr. Southworth. Dr. Bastings.

DR. BASTINGS: I have a comment regarding

the U.K. reporting system. We have a fair idea of

the incidence of acute dystonic reactions with

metoclopramide. In the Pozen study, it was 0.05

percent. I find it surprising to see that only 26

cases were reported with migraine product, if you

consider the exposure, to have so few cases and it

suggests that there was a vast under-reporting of

these adverse reactions.

Do you have any comment on that?

DR. SCHAPIRA: This, remember, was the

number of reported cases with migraine preparations

as opposed to those with all metoclopramide

164

preparations. Do you have the slide number from

which that is taken?

DR. BASTINGS: Yes; it is Slide 31, CC-31.

DR. SCHAPIRA: Just looking at all

metoclopramide preparations, that is to say the

non-migraine ones, there were 478 of those. I

don't know how that compares to the U.S. data.

DR. SOUTHWORTH: There were 203 unique

dystonia cases.

DR. SCHAPIRA: In the U.S. data.

DR. SOUTHWORTH: Over about the early

'80's to present.

DR. SCHAPIRA: So this is actually twice

the number here in the U.K.

DR. KIEBURTZ: Over a longer period.

DR. SCHAPIRA: Over a period of 64 to 45.

So this is 40 years and the U.S. data was over 35

years, I think. Is that right?

DR. SOUTHWORTH: The earliest reports we

have are from the early '80's.

DR. SCHAPIRA: So about 25 years.

DR. KATZ: I think the point is that we

165

have a good--an estimate, anyway, from the Pozen

control trials of a particular adverse event

associated with the acute use of the product. It

is 0.05 percent or maybe it is 0.1 percent,

depending--we have seen those numbers vary.

DR. SCHAPIRA: Yes.

DR. KATZ: But, when you look at, for

example, over the 40-year period in the U.K., and

if you look at dystonia with episodic use of the

combination products, which are the migraine

products, there are 26 reports. So I don't know

what the percentage of use that is, but it is

probably less than 0.1 percent.

I think that is the point from the

controlled trials, which are the best way to get an

estimate of these events, you see some finite risk.

It is relatively low but finite that the reporting

rate seems to be, perhaps, orders of magnitudes

less than that, the point being that

under-reporting may be a sizeable--there may be

sizeable under-reporting for these events which are

known to be associated with these treatments.

166

DR. SCHAPIRA: Yes; I agree entirely. It

was just the comparison between the U.S. and the

U.K. reporting system. It seems that there isn't a

difference between them here.

DR. KIEBURTZ: Dr. Hughes, did you have a

comment?

DR. HUGHES: A related question. I think

you mentioned there were no cases of tardive

dyskinesia amongst the combination users. I think

you mentioned something like 100 million doses had

been prescribed. That was an estimate.

DR. SCHAPIRA: That was an estimate.

DR. HUGHES: But 24 cases amongst those on

chronic use.

DR. SCHAPIRA: Yes.

DR. HUGHES: Do you have an idea of how

many doses have been prescribed for chronic use? I

am trying to get some sense of--

DR. SCHAPIRA: Right. Of course, I am

going to have to speculate on this, but the

proportion of use of metoclopramide in the U.K. for

migraine as a precaution for all of its other uses

167

is about 20 to 25 percent of metoclopramide

prescriptions are for migraine, approximately. So

I suppose that one could extrapolate from that.

DR. KIEBURTZ: Dr. Temple.

DR. TEMPLE: Just a comment on reporting

rates and spontaneous reporting systems. We spend

endless hours and months agonizing about this. It

is very clear the rates are different for different

kinds of reactions. There is something called the

Weber curve that was derived mostly from British

data that says reports of any given reaction

decline after the first three years because people

all know about it. It is in the label.

You can easily imagine that people

wouldn't report very much of something like a

dystonic reaction which everybody knows about. You

would hope that they would be more likely to report

TD because it is not as clear that everybody knows

about it. But there is just no way to know these

things and it is a constant source of difficulty.

The other thing that is going on, in the

United States, in the '80s, there were maybe 20,000

168

reports to us a year. You probably have the

number. We are now up over 400,000 a year. So we

have a belief that reporting of all things is going

up. It is very impossible to reach conclusions

that are valid, however.

DR. KIEBURTZ: Dr. Jung.

DR. JUNG: I will pass.

DR. KIEBURTZ: Dr. Fahn.

DR. FAHN: Regarding Slide 31 and then 22,

where there is no tardive dyskinesia being

reported, Slide 22, the Bateman second article

there, the Bateman 1989, the second of the Bateman

articles, fortunately, and I must congratulate the

sponsor because we do have the reprints of these

articles so I was looking at that because that came

as a big surprise where there were zero TD cases.

But, actually, what this survey was is

that they wrote to the physicians who prescribed

the drug Maxolon for migraine--I assume it is all

for migraine--and got the responses back. But they

grouped the responses. They didn't list tardive

dyskinesia as one of the responses. They just said

169

dystonia-dyskinesias and they grouped all that

together. So they are listed here in your table as

12 dystonias. But a lot of them--and some of these

are older people. It is very likely that a lot of

these were persistent dyskinesias and not just

acute dystonic reactions or something that this

table may imply.

So I just wanted to say that, looking at

this, you can't really clearly say how many--there

was no zero TD cases. So that was one comment.

DR. SCHAPIRA: Yes; they classified

patients as dystonia-dyskinesia which, if you look

back at the first Bateman paper here, they also

used that classification and separated out from

tardive dyskinesia. So, in the first paper they

did, they did clearly separate dystonia-dyskinesia

and tardive dyskinesia. In the dystonia-dyskinesia

they noted that--I think the dyskinesia resolved

and made the comment--and this is with respect to

their paper in 1985--that this was more likely an

acute type reaction.

DR. FAHN: But the other thing on this

170

table, too, in the second Bateman article, they

have 46 other events. So, perhaps, within that,

there might be TD. That is what I am saying. From

this paper, you can't really say there was no TD

because they specifically failed to talk about TD.

That is the problem.

If I can also make a comment at this time.

The definitions of TD, that is another thing that

can confuse people. If you are going to use the

definition of TD that you have to be on the drug

for 30 days or more or three months, as some

definitions have used, I am not sure any

neurologist used that kind of definition. We

tended to look at tardive dyskinesia as persistent

dyskinesia.

The name tardive was given because it

wasn't seen until later on after the neuroleptic

drugs had been on the market for a while. That is

when tardive came on. But, with more experience

with this, recognizing what the syndromes are, we

sort of consider, now, tardive dyskinesia really

should have been properly named persistent

171

dyskinesia, meaning that it lasts longer than 30

days.

It doesn't matter how long they have been

on it for. They can be on it for one day and still

have persistent dyskinesia. To us, that is what we

now refer to as tardive dyskinesia. So we don't

give up the name tardive dyskinesia. We use the

definition differently than what was stated here.

I just wanted to make sure the concern I would

have, as a neurologist, who treats tardive

dyskinesia as meaning persistent dyskinesia, that

these can be irreversible.

When I heard the comment about how

devastating migraine is, and I agree it can

be--severe pains can be very bad for people,

devastating, but so can severe akathisia including

tardive akathisia, I consider that equally as bad.

If you ever induce this for a drug you might not

have needed, then that is a concern for us that we

have to face.

DR. KIEBURTZ: I am going to follow your

comment and pursue questions with a couple of

172

people. Neurologists fall into camps of movement

disorders and stroke and headache. I think, just

to set tone, I think everyone would agree that

migraine is a serious and disabling disorder that

needs good treatments. I don't think anyone is

debating that. And there are unmet therapeutic

needs.

I think everyone also admits, but might

not be so familiar with, that tardive syndromes,

tardive dyskinesia and other tardive syndromes, are

also very disabling, much less common than

migraine. The issue before us is how to balance

those issues. Neither is more important or less

important than the other. They are both very

important clinical issues. I don't think anyone is

doubting that. So I don't want to have that sort

of be on the table, that we are undervaluing one or

the other.

Dr. Lenaerts.

DR. LENAERTS: Thank you. In the U.K.

experience, could Dr. Schapira possibly--I don't

know if it is hypothetical or known

173

information--the comment on the use of

metoclopramide specifically in migrainers and,

namely, the frequency of use as that may factor in

the relatively low reports of adverse events in

migrainers as compared to the U.S. experience.

DR. SCHAPIRA: Thank you. Well, as I

mentioned, the combination of metoclopramide and

analgesics is the second step in the

U.K.-recommended guidelines for the management of

acute migraine. The estimate of prescriptions of

combinations, of the two combinations currently

marketed, is approximately 200,000 per annum. It

is estimated that about 95,000 patients are taking

these products each year.

DR. LENAERTS: How does that do--in terms

of frequency of use per patient because that would

be really the relevant issue there.

DR. SCHAPIRA: Of course, that is

undoubtedly going to vary from one patient to

another depending upon the severity of response,

frequency of attacks, et cetera. The sort of

bottom-line figure, if you like, that we have is 85

174

doses per annum. But, between that, there is going

to be a substantial range.

For instance, one has patients who will

take two doses at the onset of an attack and have a

good effect from that. Another proportion of

patients will need to take two doses separated by

four hours.

But what I don't have is the data to tell

you what proportion take just one dose at the

beginning, two doses, et cetera. I don't have the

data and don't think it is available to tell me how

many migraine attacks they are treating per year.

I can only give you that bottom line of about 85

per person per annum.

DR. KIEBURTZ: Dr. Welch?

DR. WELCH: Do you have any comment on any

difference in prevalence of medication overuse

between Britain and the United States?

DR. SCHAPIRA: I think I should pass that

to one of my U.S. migraine experts because I think

they will probably know more about that in terms of

the prevalence of medication overuse.

175

DR. KIEBURTZ: If you haven't spoken

before, just please introduce yourself for the

record.

DR. SAPER: My name is Joel Saper. I am a

neurologist specializing in headache and Director

of the Michigan Head Pain and Neurological

Institute in Ann Arbor. Dr. Welch, could you--I

got part of that question. I didn't hear the whole

question.

DR. WELCH: What is the prevalence of

medication overuse in the U.K. and what is the

prevalence of medication overuse in the United

States?

DR. SAPER: I am not sure that I can

answer the question with specific statistics. I

think Dr. Silberstein showed a slide, was it 1 to 2

percent, Steve, on the medication overuse in the

United States?

DR. SILBERSTEIN: The data is

extraordinarily limited. The two studies I know

of, population-based studies, in medication overuse

headache, Ann Scher's study done in this region.

176

What Ann and Richard Lipton showed is about 4

percent of the population have chronic daily

headache. In the clinic, it is estimated that most

of the patients we see with daily headache overuse

medication, what Ann and Richard estimated in the

population is about 30 percent.

I don't know of any good estimates in

Great Britain. I do know that, in other parts of

the world where the problem has been studied, all

worldwide estimates of chronic daily headache are

about 4 percent and half are migraine, and of those

patients, about 30 to 40 percent in

population-based studies.

Anecdotal experience from Peter Goadsby in

National Hospital in Queens Square suggests that is

clinic-based population is about the same as here.

Mike, do you know of any population-based

estimates? Anyone?

DR. SCHAPIRA: No. So, in other words, we

don't know whether the medication overuse is less

in the United Kingdom than here and we are being

given data on the adverse events in Britain.

177

DR. SAPER: If I may just finish the point

I wished to make when I asked Dr. Silberstein for a

clarification, I do know this. In our system, in

the United States, medication overuse is a

significant problem of the referral base that is

referred to us, perhaps 60 to 70 percent of

referrals to a tertiary quadrinary will have that

problem.

In the U.K., I do not know for other than

opioids. In opioids, which is the dominant overuse

in this country, in my referral base, which is very

large. In the U.K., opioid use is much more

limited. So, to the extent that we can compare

opioid overuse in the United States to that in the

U.K., there would be much more overuse in the

United States, but specifically targeting the

opioids.

DR. WELCH: Do you think that speaks to

the difference in physician practice in the United

Kingdom--in other words, major primary-care control

of the prescriptions in the United Kingdom as

opposed to the United States?

178

DR. SAPER: There is, I think, a

difference in prescribing habits, Dr. Welch. I

think, in the United States, opioids, particularly

in the last ten years, are being much more

aggressively prescribed for non-cancer pain than

they were prescribed ten or fifteen years ago.

In the U.K., from my understanding and

talking to colleagues, the opioids are restricted

much more so for non-cancer pain.

DR. KIEBURTZ: Dr. Hughes, you had a

question.

DR. HUGHES: Sort of two related

questions. Going back to the Reglan warning label

and the comment in that label about risk of TD

being increased after relatively brief treatment

periods at low doses, is there data available that

is underpinning that particular comment?

DR. BASTINGS: We don't know what that was

based on. We did not write that label. It is a

different division and we don't have that

information.

DR. KIEBURTZ: I am not sure it is

179

data-driven but, even in the comments you saw in

the Pozen briefing booklet, I believe, amongst

movement-disorder neurologists, there is the

impression that even one dose of exposure can

possibly cause tardive syndromes.

To answer your question, I am not familiar

with data to address the question but the label may

have incorporated clinical acumen or anecdote as

part of the label.

DR. HUGHES: I guess the related question

was the open-label safety study that you have

conducted, I was wondering if there were

eligibility restrictions or the

treatment-management approach within that study was

conservative to try and avoid risk of movement

disorders.

DR. ALEXANDER: Thank you. This study we

call 302 enrolled 1,006 subjects. They were

enrolled on the basis of having a history of

migraine. But there were no restrictions on

previous presence of movement disorders or any

persistent neurologist deficits. We had not, at

180

the time of that study being initiated, anticipated

that this would be an issue in the development and

so there were no particular requirements.

DR. HUGHES: And there weren't any

particular requirements in terms of management of

patients?

DR. ALEXANDER: Well, no. This was 54

centers, I believe, headache centers in the U.S.,

generally, and patients were instructed to take one

dose of MT100--this was open label--one dose for

the treatment of moderate to severe headache.

There was no second use of MT100 as rescue

medication, for example.

It was a safety study. Subjects were

asked to record adverse events after taking doses

within 24 hours and then, when they came back to

clinic, at 3-month intervals.

DR. HUGHES: But these patients were

taking it on a chronic intermittent basis.

DR. ALEXANDER: Well, I would call it--I

would like to make the distinction that we are

talking about episodic or PRN use because chronic

181

intermittent could also include the psychiatric use

of neuroleptics in the treatment of schizophrenia,

for example.

One thing I would point out is that the

episodic three-times-a-month use is probably very

different than chronic intermittent

time-on/time-off of a neuroleptic used in the

treatment of psychiatric conditions. These were

PRN intermittent. As Dr. Schapira has shown, the

average number of doses per month was 4 and the

median number of doses that each subject took over

the 12-month period was 22.

DR. HUGHES: So would you characterize the

study as mimicking what might occur in clinical

practice?

DR. ALEXANDER: Yes; exactly. It was a

clinical-practice type study, a real-world study,

just looking at the adverse-event collection over

the time enrolled.

DR. KIEBURTZ: Dr. Katz, you had a

question?

DR. KATZ: No; he answered my question.

182

DR. KIEBURTZ: Dr. Jeste.

DR. JESTE: Do you have a theory why

metoclopramide would work better in people without

nausea?

DR. ALEXANDER: Thank you, Dr. Jeste. We

believe, and we have phase 1 data to show, that the

addition of metoclopramide to naproxen sodium will

enhance the absorption of naproxen through

accelerating the--well, it brings the T-max closer

to the time of dose. We have phase 1 studies

showing that the T-max for naproxen is decreased by

about 30 minutes, on average.

That is Slide 25.

(Slide EB--25)

Just to illustrate this point further, is

that, in the phase 1 studies, when we gave naproxen

sodium 500 milligrams without metoclopramide, the

T-max was 72 minutes. We added, first, 8

milligrams of metoclopramide and the T-max dropped

to 57. With 16 milligrams, which is the dose in

MT100, the T-max was 44. So I think I am answering

your question about why it speeds the absorption

183

because it empties the stomach. Naproxen is only

absorbed in small bowel as are all the other

NSAIDs.

DR. JESTE: So you think it would work

differently in people without versus with nausea?

DR. ALEXANDER: I have no data for sure.

The theory, just to answer that, is that when

nausea is present, there is probably more gastric

stasis, and 16 milligrams of metoclopramide cannot

overcome the gastric stasis present.

There also may be--there is a theory that

dopamine has effects in migraine and maybe the

dopamine has an effect to increase nausea.

DR. KIEBURTZ: Dr. Sacco is next.

DR. SACCO: I had a question regarding

adverse experiences for someone from the company to

maybe comment. Looking at the FDA briefing

document, in the clinical review section, FDA

Clinical Review, there is an overview that puts

together all the adverse experiences. I would like

someone to just comment for me, as a stroke

neurologist, not a movement-disorder neurologist,

184

how best to be sure that the increased frequencies

of restlessness and feeling jittery that are noted

with MT100 and MT100 two tablets versus some of the

other groups wouldn't be possibly forme fruste or

partial TD-related symptoms.

This is on Page 103 of 128 in the FDA

document.

DR. KIEBURTZ: Your document may not be

entirely the same as theirs. But I am not sure. I

believe they would know the table.

DR. KATZ: Could you just repeat which

page, which document.

DR. SACCO: The document I have, which is

under a section called FDA Clinical Review, it is

Page 103 of 128 in the section, I guess, document

dated--Table 64. I have a table number. I guess

my question just is, again, getting back to that

original slide where I asked the clarification,

akathisia, restlessness, is possibly a form of

tardive dyskinesia. Being, again, a stroke

neurologist, I am just trying to understand and be

sure that the restlessness and feeling jittery

185

issues that are listed in that table that are

increased--I see 13 in the MT100 group and then,

actually, 12 or 4 percent in the MT100 two tablets

group, and I see feeling jittery, 13 and 4, also

there.

I am just trying to get a better handle,

maybe, from the company what some of those adverse

experiences were all about.

DR. ALEXANDER: Thank you, Dr. Sacco. I

am looking at Table 64, that the data across all

the treatment--each treatment arm. I would point

out that, if you look at restlessness with MT100,

there were 2400 subjects who were exposed to a

single dose. There were 13 cases of restlessness

there, less than 1 percent.

That was the same percentage relative to

the restlessness in naproxen sodium,

metoclopramide, sumatriptan and placebo. All were

less than 1 percent. You have pointed out that

there is a 4 percent incidence of restlessness with

the two-tablet dose of MT100. There were 313

subjects who received the two-tablet dose.

186

Pozen has not, did not, submit the NDA

asking for approval of a two-tablet dose of MT100.

So, therefore, the 4 percent higher rate of

restlessness may not be relevant to our discussions

today.

But I would want to say that these were

acute events that were seen and they did resolve.

The question of restlessness with the use of MT100

could, perhaps, be looked at in the 302 database

just to put this in perspective.

Let me have, please, Slide No. 19.

(Slide SA-19)

This is the study, the large study,

repeat-dose study, that we have spoken about. I

just think it might be beneficial to the committee

to see that the most common event in these 1,006

subjects was somnolence, dizziness, 7 percent.

Restlessness, overall, was 2 percent. Again, these

are 23,000 single doses of MT100 being treated,

being used by these 1,006 patients.

Anxiety is 4 percent and nervousness, 1

percent. Fatigue is a little more common.

187

Perhaps, that would help put this in context.

DR. KIEBURTZ: Thanks. Actually, Dr.

Temple and Katz, but we are keeping a list.

DR. TEMPLE: I am still curious about the

theory of why the people with no nausea might

benefit. You might even think that, if they have

delayed absorption, they would benefit more--you

might. I guess I couldn't understand what the

theoretical basis for expecting a benefit on the

sustained response and not on the acute response

would be. If you shortened T-max, it would seem

more likely to do the opposite, I would have

thought.

That is a question to the people who put

forth the theory, not to the committee.

DR. ALEXANDER: I'm sorry. I thought you

were posing that to committee. I think the

question would be what happens when migraine is

effectively treated by an analgesic agent. We

don't know the effective plasma dose of naproxen

that will start the process of analgesia. I don't

think we know that.

188

So we, in theory, if we can bring the time

of obtaining whatever that is closer to dosing and

start the process of analgesia and the effect on

neurovascular leakage or whatever it is that is

causing the migraine pain, then we can effect,

perhaps, a smaller two-hour improvement. But that

improvement then goes on after two hours to 24, and

we see less relapse. We see less need for

remedication and the pain response, as measured by

sustained pain response, may be there.

That is just a theory. But, certainly,

the quicker you treat the pain, perhaps the better

it will be at 24 hours.

DR. TEMPLE: Plausible. But it doesn't

treat it quicker. I mean, that is what the results

show.

DR. ALEXANDER: Remember I showed the SPID

data. SPID data is used in the analgesic division

as the most sensitive tool we have to measure the

duration, the onset, the amount of pain relief.

And we showed, in both studies, that, by two hours,

MT100 was statistically significantly better than

189

naproxen sodium using SPID scores.

So I would just say we have data that

makes it effective. Let's have Slide 83.

(Slide EB-83)

Just to answer this. The SPID scores in

Study 301. The pink line at the top is MT100. The

blue line is naproxen sodium. The green line is

metoclopramide. At one hour, MT100 is

significantly better than metoclopramide. At

one-and-a-half hours, naproxen sodium is better

than metoclopramide. At two hours, MT100 is

significantly better than naproxen sodium, p-value

0.044.

The same thing was seen in 304. The next

slide would be, I think, 84.

(Slide EB-84)

Importantly, in 304, this effect was seen

and one-and-a-half hours. By, by one-and-a-half

hours after dosing, you have separation of these

two treatments.

DR. KIEBURTZ: Dr. Welch.

DR. WELCH: I would like to pursue the

190

same issues that Dr. Temple is discussing. There

may be other explanations other than just simply

absorption from the gut. When you look at the PET

studies or MR studies, you find that basal ganglia

and a whole network of systems are activated during

an acute migraine attack. It could be that it is

interfering with that at some stage or other.

But I would really like you to address the

issue of the time from the onset of the migraine

attack to treatment as another possible

explanation. Nausea is very variable. Sometimes

it comes on in the middle of the attack and often

is associated with the severity of the pain or

severity of the attack.

Could it be, can you reassure us, in

making this differentiation of a 10 percent

difference between nausea and non-nausea and you

don't have a population that treated itself at a

different time, and if, for example, the group

without nausea treated themselves earlier than the

group with nausea, then you might expect a

differential benefit because you are interfering

191

earlier with the cascade of events involved in the

migraine attack. So there is another possible

explanation that may have nothing whatsoever to do

with absorption from the gut.

DR. KIEBURTZ: Let me just, to clarify,

your question is is there evidence regarding the

time that has passed between the onset of headache

and the self taking of the medication in those who

report nausea versus those who don't as a possible

potential explanation.

DR. WELCH: Correct.

DR. KIEBURTZ: Could you address that

specifically? Do you have that?

DR. ALEXANDER: First of all, Dr. Welch, I

don't have data to answer your question. We did

not measure the time from onset to treatment in

these studies. I think you are referring to early

treatment modalities and perhaps treating when the

pain is mild or moderate before nausea may develop

in a subject. We just don't have data to support

that.

DR. WELCH: It is not necessarily mild.

192

It is the duration--it is the time when you treat

it first.

DR. KIEBURTZ: Thank you. Dr. Jung.

DR. JUNG: Can we go back to Table 19.

DR. ALEXANDER: That is the overall safety

data from 302?

DR. JUNG: Can you tell me what the

difference is between restlessness, anxiety and

nervousness? Was there a specific way of

differentiating for the patients or for the

providers who submitted the reports?

DR. ALEXANDER: No; there wasn't. These

were characterization of the adverse-event reports

by the investigators who recorded the data in the

case-report form.

DR. KIEBURTZ: Are those WHO-ART coded?

Are these verbatim? Does anybody know?

DR. ALEXANDER: They do get coded through

the WHO-ART system, or CoStart system--I'm sorry;

the MedDRA system.

DR. KIEBURTZ: So these are MedDRA roll-up

terms?

193

DR. ALEXANDER: Yes.

DR. KIEBURTZ: So that means that these

are terms, the text was coded into these terms. To

be sure, you would have--this is not the actual

verbatim; correct?

DR. ALEXANDER: Well, it may be. I would

have to look--

DR. KIEBURTZ: It may or may not be.

DR. ALEXANDER: I would have to look at

the MedDRA Coding Guidelines to see if these are

verbatim. But there are some terms that, as you

say, do get coded to something else.

DR. JUNG: Does that mean that there is

clarity between the difference between anxiety,

nervousness--there isn't?

DR. ALEXANDER: No.

DR. JUNG: So if you take restlessness,

anxiety and nervousness and added all that up in

terms of adverse events, wouldn't that be pretty

close to the primary 108 that was reported

somnolence? If you have restlessness, anxiety and

nervousness, that is 50, 60--

194

DR. ALEXANDER: It is a percent. If you

look at the bottom of that column, 78 percent of

these subjects who were treated for up to a year,

at some time or another, reported some adverse

event. I don't think it is uncommon to have 78

percent of the subjects reporting something over

the course of a 12-month study.

DR. JUNG: I had a second question.

DR. KIEBURTZ: Please, go ahead.

DR. JUNG: Not related to the slide.

Going back to Dr. Fahn's comments earlier about the

significance of movement disorders on quality of

life, do we have--this is for the FDA staff--do we

have any data that looks at the incidence of other

adverse effects associated with other treatments

for migraines, so, for example, cardiovascular side

effects, G.I. side effects, just for comparison?

DR. KATZ: The rates of things like chest

discomfort and stuff can be read in the labeling

for all these drugs. They are not inconsiderable.

There is a fair rate in those things. But I don't

think we have them on the top of our head at the

195

moment. Anybody with a Blackberry can probably

find it.

DR. JUNG: I guess my point was that, if

we are trying to decide whether or not this drug is

worthwhile, it would be reasonable to compare that.

DR. KIEBURTZ: Dr. Porter?

DR. PORTER: Back to Dr. Schapira. On

your combinations on metoclopramide and analgesics

marketed in the U.K. and the 95,000 patients per

year that receive these doses which are twice that

of the MT dose, the question arises which is in a

similar vein to what other people have asked. On

the dystonia oculogyric crisis, for example, or

extrapyramidal disorders not specified, what is the

possibility that a substantial number of those, or

even some of those, could be tardive dyskinesias

and not just properly classified. That would make

a big difference, I think, in our interpretation of

this rather impressive database if we thought that

that was possible.

DR. SCHAPIRA: The identification of

dystonia or oculogyric crisis is a phenomenon

196

associated with the acute, generally speaking,

exposure to a neuroleptic specifically, in this

context, metoclopramide. Therefore, they are, I

believe, more easily distinguished in terms of an

acute dystonic episode from tardive dyskinesia.

Also, acute oculogyric crises or acute

dystonia, upon exposure, which may, for instance,

occur with the first dose ever of metoclopramide,

resolves spontaneously and I think can be more

easily, or relatively easily, distinguished from

tardive dyskinesia which, generally speaking,

involves a different type of movement.

Stan has already--I'm sorry; Dr. Fahn has

already alluded to the issue of persistent

dyskinesias as opposed to tardive dyskinesias. I

would make a distinction between those persistent,

brackets, tardive, brackets, dyskinesias and the

sort of acute oculogyric, acute dystonic, reactions

that we see with metoclopramide.

DR. PORTER: So you don't think that these

dystonia extrapyramidal disorders dyskinesia were

persistent in the same way that tardive dyskinesia

197

might be persistent.

DR. SCHAPIRA: Well, I can only speculate,

of course, because I don't have the original case

reports in front of me. But, as these have come

from medical practitioners, I would hope that at

least the great majority of them, the vast majority

of them, would relate specifically to an acute

dystonic reaction. Otherwise, they would have

been, I would hope, classified as something else.

DR. PORTER: They had at least the option

to check a box that was tardive dyskinesia?

DR. SCHAPIRA: They wouldn't, I believe,

have a box to check. They would have a space to

include what they thought.

DR. PORTER: Okay. Thank you very much.

DR. KIEBURTZ: Can I have four more people

I have no my list. Those are Drs. Goldstein,

Jeste, Sacco and Katz and I have a couple of

questions. So if we get through those, then we can

do more.

DR. GOLDSTEIN: First, just a few

questions for clarification just for me, I guess.

198

The data that you showed very quickly about the

difference in time in rate of onset, I guess, for

the three treatments that were just shown, the

statistics for that. I know there is this whole

debate about what is the right statistics to use

but, in that particular graph, you were showing

repeated measures over time. I was wanting to know

whether those were just uncorrected pairwise

statistical comparisons or whether that was

analysis of variance with repeated measures and

then post hoc tests.

DR. ALEXANDER: Dr. Goldstein, are you

speaking of the SPID scores?

DR. GOLDSTEIN: Yes.

DR. ALEXANDER: Could we have the SPID

score methodology? Just come on over, Susan.

Susan Spruill is the Senior Director of

Biostatistics at Pozen.

DR. SPRUILL: The answer to your question

is that it was an analysis of variance for each

time point and it was not corrected for multiple

time points.

199

DR. GOLDSTEIN: So it wasn't corrected.

Okay. The second question is, again, just a

general question about the trials that were done.

You guys must have measured or recorded concomitant

medications. Many patients with migraine are also

on prophylactic therapy as well as this rescue

therapy. This is basically a rescue therapy.

Were there differences in what types of

other medications these patients were on?

DR. ALEXANDER: There were exclusion

criteria if patients had changed their prophylactic

medication within several weeks to a month of

enrollment. We did not exclude patients that were

on prophylactic medications or anti-depressants,

anything that was used for the routine medical

treatment. There was no stratification of the

randomization by use of prophylactic medications.

DR. GOLDSTEIN: Thanks. The third

technical question is, again, related to this whole

issue of absorption. The preparation that you are

looking at here has the naproxen that is sort of

like buried inside this sort of shell. Is there a

200

difference in the bioavailability of the naproxen

alone in that preparation? You said you used that

in the controls also in the studies. Is there a

difference in bioavailability of that compared to

the usual preparation of naproxen if you would get

it within that particular formulation?

DR. ALEXANDER: I am really not qualified

to answer that with data. My impression is that

there is no difference.

DR. PORTER: The point, I guess, is that

what you are hypothesizing is that the

metoclopramide is improving the absorption of the

naproxen in this particular preparation because

that is the comparison you have. But you don't

know that, if you didn't have it in this particular

preparation, that there would or would not be a

difference.

DR. ALEXANDER: No; we used the same

blinded study medication that was naproxen in the

core with a placebo around it in these studies that

were conducted. So I would answer to say that it

seems that we have controlled for everything except

201

the metoclopramide component.

DR. KIEBURTZ: Let me just restate it. I

think your question is, in the naproxen alone, how

does that compare to a standard naproxen

preparation.

DR. GOLDSTEIN: Exactly.

DR. KIEBURTZ: Does the formulation here

somehow delay the naproxen so that the

metoclopramide then just speeds it back to normal?

DR. GOLDSTEIN: That is exactly right.

DR. ALEXANDER: Let me clarify. These two

are bioequivalent as far as the naproxen component.

I was not familiar with the study.

DR. KIEBURTZ: Dr. Jeste.

DR. JESTE: If people with nausea respond

differently to metoclopramide, that people without

nausea have greater therapeutic benefit, it is also

possible that they may have differential response

in terms of side effects. They might have more

side effects or less side effects than people with

nausea.

Do you have any evidence for that?

202

DR. ALEXANDER: That is one of the things

that we looked at very quickly after seeing this

efficacy difference was to see if there was any

apparent difference in adverse events. There are

none. Again, these were prespecified analysis for

the endpoint subgroups so we went back and looked

at the adverse events. There is no difference

between the two subgroups as far as adverse events.

DR. SACCO: It is a question for, I think,

Dr. Schapira maybe to comment on. We heard from

Dr. Matchar about this unmet need, about the number

of people suffering with migraine, the disability,

and how this drug could, perhaps, be an unmet need.

One of your slides talks about the fact

that this is available in the U.K. and some of your

numbers have been 95,000 patients possibly treated.

I guess what I am trying to get an understanding of

is in the U.K., using the U.K. as an example, the

sponsor provided in their document, Appendix 3,

which talks about prescription use in migraine

patients.

If anything, at least from my

203

interpretation of this table, the migraine

anti-emetic combinations, of which the main one is

this kind of drug out there, is dropping kind of

precipitously from November, 1999, about 112,000 to

79,000 in November of 2003.

When you look at the overall proportion of

migraine sufferers again estimated in these tables,

it is a little lower and dropping, so, just again,

a feel for, if we use the U.K. as an experience

where this drug was available, we recognize the

unmet needs that perhaps are possible in the U.S.,

can you comment on why this seems to be dropping

off as a migraine type of drug.

DR. SCHAPIRA: Yes. I think that the

guidelines that have incorporated the use of these

combinations into the acute treatment of migraine

is in response to a need to rationalize the acute

treatment of migraine.

First of all, I think many patients with

migraine never even go to a doctor. They just

treat a headache, themselves. Those that do go to

a doctor, I think are recommended to begin an

204

analgesic just to see if they have tried that and,

if they have and have not responded, then there is

a pressure, if you like, from the guidelines, the

Headache Society in the U.K., et cetera, to try and

build the patient's future management on an

appropriate logical step which is, if you like,

tailored to the patient because I think a

significant proportion of patients are being taken

from simple analgesic to triptan without this step

in between.

I think, to some extent, this is a

response to the introduction and marketing of

triptans in the U.K. What we have tried to do is

to actually insert this intermediate step between

those two.

DR. KIEBURTZ: Which one of you two--you

had your hand up before, or would you like Dr.

Bastings--

DR. KATZ: Eric has, I think, an

interesting question but I just have a few maybe

sort of random thoughts. The side-effect profile

or incidences that were shown for Study 302, that

205

is open-label data. It is hard, really, to

understand exactly what those numbers mean. There

is no concurrent control group.

I don't know. I didn't really understand

Dr. Alexander's answer when he said, well, you add

them all up, you get 78 percent of people reported

some sort of an adverse event, therefore, it is not

unexpected.

I think Dr. Jung's point is a very good

one. We are well familiar with the distinction

between lumpers and splitters when adverse events

are recorded and reported. Of course, it is hard

to do know what to make of the specific point here

about adding up all these things because we have no

idea, really, what those terms represent, which is

a point, I think, that has been made.

But, certainly, of course, it is possible

that the restlessness, anxiety, nervousness,

sleeplessness, whatever these things are, it is

possible that they are all akathisia or some

extrapyramidal symptom. It is hard to know, but I

think the point is a very good one that,

206

generically, these sorts of events are notoriously

poorly reported, not in this application

specifically, but across applications, and poorly

translated from verbatim reports to MedDRA or

CoStart.

So that is something to watch out for. It

is true individually they are less than what--but,

even if you look individually, they are all less

than 1 percent more or less across all the groups.

But, 0.9 percent is different, perhaps, then 0.2

percent. But if you just list them all as less

than 1 percent, they all look the same.

So that is just something to keep in mind.

Again, it is hard to know what to make of it

specifically in this case.

The only other thing I want to say is

there has been a lot of discussion about the

rationale of why it should be true that it works in

people who don't have nausea at baseline. Anything

is going to be somebody's interesting theory. It

is impossible to really know what the rationale

would be.

207

It would be nice to know what the

rationale was to sort of have a real solid

well-accepted rationale in hand because I do think,

as Dr. Temple said, that this was probably an

unexpected finding. It was probably expected to

work better in people who had nausea at baseline.

There is going to be a retrospective explanation,

but we are mostly concerned about whether or not it

is a real effect and, as one of our questions asks,

whether or not it is an appropriate, well-defined

population.

So just sort of generically, I think we

are probably less concerned with the rationale for

it than is it real and is it a population that we

can reliably identify so that we can write

labeling.

DR. KIEBURTZ: Why isn't as important as

if.

Dr. Bastings.

DR. BASTINGS: It is a question to Dr.

Schapira regarding these combinations available in

U.K. Is there any evidence from controlled studies

208

that metoclopramide provides any contribution to

efficacy for these preparations?

DR. SCHAPIRA: Thank you. I have not

actually looked at and don't have access to the

original data on Paramax or MigraMax. But I think

that Dr. Alexander may have in order to answer your

question.

DR. ALEXANDER: Your question is whether

or not there is data that this combination would be

effective in migraine.

DR. BASTINGS: No.

DR. KIEBURTZ: Let me restate the question

and make sure I understand it. The question of the

addition of metoclopramide to the other analgesic

like what we have been going around about today, in

a factorial kind of design, what is the evidence of

the additional benefit of metoclopramide to the

underlying analgesic.

Did I understand your question correctly,

Dr. Bastings?

DR. BASTINGS: Yes.

DR. ALEXANDER: There are limited studies.

209

One study I will show is twenty years old. That is

this slide here.

(Slide EB12)

It is a study by Tfelt-Hansen looking at

the preparation Migravess which Dr. Schapira noted

was aspirin plus metoclopramide. This was a

factorial study looking at Migravess, aspirin and

metoclopramide, versus aspirin alone versus

placebo.

You can see that this is a modest

improvement of only 2 percent for a two-hour pain

response rate in that study. A good small number

of subjects, basically 118 subjects, treating three

migraine attacks.

There have been a number of comparative

studies between these combinations and even with

triptans that have shown generally equivalent

results in the treatment of migraine.

DR. KIEBURTZ: Thank you.

DR. SMITH: I want a little clarification

on the estimated risk of the TD with the

intermittent use because Dr. Bastings mentioned, in

210

his talk, that there was an estimate of 380 cases

per million persons per year and then, in the FDA

review, there was an estimate quoted from the NDA

of 20 per million.

I am wondering what is the estimate that

the company has and what is the basis for their

estimate.

DR. ALEXANDER: Thank you for that

question because it needs clarification. Back in

2004, when Pozen was preparing for the meeting with

the FDA, the type A meeting, the critical-path

meeting that you heard about, we were looking at

the figure of 1 percent as the figure that was

given to us in the not-approvable letter as being

the upper limit of possibility. But is still a

figure of 1 percent.

That certainly is significant when you

consider tardive dyskinesia. So we undertook to

try to estimate how much lower the risk might

actually be. We looked at spontaneously reported

cases in the databases available. You have seen

the 87 cases or the 40 cases from the Shaffer

211

article.

We also looked at estimates of

prescriptions or usage over a year's period of time

and applied a multiplier for the spontaneous

reported rate of, say, 1 percent. If only 1

percent of all incident cases are reported, what

would the risk be? It is a conservative estimate.

In that way, we calculated the risk could

be as high as 0.002 percent or something like that,

just to make the distinction that it is much less

than 1 percent. That has gotten carried over into

this discussion today and I just want to point out

that it is based on estimates or assumptions. The

better data are the actual-use data that Dr.

Schapira has shown with the use of metoclopramide,

we think, in migraine in a large population in the

U.K..

DR. SMITH: Let me just make sure I

understand. You are basing the estimates, these

lower estimates, on reporting rates to

adverse-events reporting systems in the U.K. and

U.S.? Is that correct?

212

DR. ALEXANDER: Let me just specifically

address the 0.038 percent risk that Dr. Bastings

more recently used in this. That is based on an

estimate of a risk that was the most conservative

that we could come up with based on the data.

DR. SMITH: Okay. Thank you.

DR. KIEBURTZ: I think we will talk about

that some more, too, when we address the specific

questions. I want to ask one question and then we

are going to have to close, and this might go to

Dr. Jinnah, if he is still here, or, if you don't

know the answer.

I believe Dr. Jung brought this up in

part. In stimulant-induced sterotopy or models of

tardive dyskinesia, are there animal models, and

this was alluded to in at least one presentation,

as to whether single or a few doses, or

intermittent doses, could induce tardive

dyskinesias in animal models as opposed to chronic

use. Can you speak to this?

DR. JINNAH: I can. I am familiar with

that literature. It is going to be difficult to

213

translate that literature which is mostly derived

from rodents to humans because the types of drugs

and the temporal course of drug delivery in rodents

to create these movements that are labeled as

analogous to TD are quite different.

So whether or not we should be even

calling it TD is at question. So I am not sure

that answering the question is going to help.

DR. KIEBURTZ: Could you answer it anyway?

DR. JINNAH: Sure.

DR. KIEBURTZ: Here is my question. I

understand there are animal models. And I

understand that animal models don't represent the

human disease. But I think it would be useful to

know whether or not tardive dyskinesia--or those

movements have been induced with intermittent or

few exposures are they are required to have chronic

exposures for them to develop in models.

It doesn't necessarily, therefore, mean

that the human experience follows that.

DR. JINNAH: Correct.

DR. KIEBURTZ: Let me put it a different

214

way. My understanding is those models indicate

that intermittent or few exposures can induce those

changes and reduce the threshold for

stimulant-induced stereotopies. Am I wrong?

DR. JINNAH: No. That's correct. Either

chronic or intermittent administration, as few as

just one or two doses of the drugs, can induce

these movements that are called tardive dyskinesia.

I should temper that statement with the observation

that Reserpine is one of those drugs and Reserpine

does not cause tardive dyskinesia in humans, to my

knowledge.

So what the movements are, I think, is

still really a key part of that answer.

DR. KIEBURTZ: So the specificity and the

meaning of these models is unclear.

DR. JINNAH: I think that is true. But,

perhaps, Dr. Jeste can comment on the issue of

using neuroleptics intermittently in psychiatry

because there is some data showing that drug

holidays of neuroleptics and intermittent use of

neuroleptics in human patients does, in fact,

215

increase the risk of tardive dyskinesia.

But, again, it is not PRN or intermittent

use as we have been talking about. It is more

chronic for weeks, months, maybe years with a

holiday and then chronic again. So it is a little

bit different but the same idea.

DR. JESTE: That actually applies to

people with schizophrenia or other psychiatric

disorders when you do long-term treatment. People

who get the treatment intermittently seem to be a

high risk of developing tardive dyskinesia. The

theory is that has something to do with kindling,

that if you administer something continuously,

there is development of tolerance where it is

intermediate. Intermittent administration may lead

to kindling and increase the risk of TD.

DR. KIEBURTZ: Thank you.

We are going to stop the discussion period

at this point. Just to reiterate, to the point of

being boring, not to continue any discussions of

the presentations not in the public forum.

We will start with the Open Public Hearing

216

next. All registered Open Public Hearing speakers,

please register at the Registration Desk. 1

o'clock is when we will start again.

Let me just thank all the participants

from the sponsor, from the FDA and from the

committee for bearing with the way we ran things.

We are adjourned until 1:00.

(Whereupon, at 12:00 p.m., the proceedings

were recessed to be resumed at 1:00 p.m.)

217

A F T E R N O O N P R O C E E D I N G S

(1:00 p.m.)

Open Public Hearing

DR. KIEBURTZ: The first part of the

afternoon is the Open Public Hearing. I have a

statement to read here.

Both the Food and Drug Administration and

the public believe in a transparent process for

information gathering and decision making. To

insure such transparency at the Open Public Hearing

session of the Advisory Committee meeting, FDA

believes that it is important to understand the

context of an individual's presentation. For this

reason, FDA encourages you, the Open Public Hearing

speaker, at the beginning of your written or oral

statement, to advise the committee of any financial

relationship that you may have with the sponsor,

its products and, if known, its direct competitors.

For example, this financial information

may include a company's or group's payment of your

travel, lodging or other expenses in connection

with your attendance at the meeting. Likewise, FDA

218

encourages you at the beginning of your statement

to advise the committee if you do not have any such

financial relationships.

If you choose not to address this issue of

financial relationships at the beginning of your

statement, it will not preclude you from speaking.

So far, we only have one registered

speaker for the Open Public Hearing and that is Dr.

Cynthia McCormick who can come on up to the podium

and address the committee.

DR. McCORMICK: Thank you. Good

afternoon, Dr. Kieburtz and members of the PCNS

Advisory Committee, Dr. Katz, Dr. Temple.

First let me start with the disclosures

and also thank you for giving me the opportunity

today to address the committee as a member of the

public, which is a new experience for me, and to

share with you my thoughts as you proceed into the

deliberations for today.

So, in terms of disclosure, since leaving

the FDA, I have worked as an independent consultant

to regulated industry including upwards of probably

219

close to 70 companies by now, many of which are in

the area of pain therapeutics including a few

companies developing drugs targeted to treat

migraine. Among these, I have served as a

consultant to Pozen.

Today, however, I am speaking on my own

behalf and my comments are not driven by my

industry consultation but rather by own personal

experience. The division has heard these, I think,

in some manner but probably not in this context.

By introduction, I am a neurologist and a

former federal employee, 17 years a federal

employee, most of which was spent at the FDA,

initially as a reviewer in the Division of

Neuropharmacologic Drug Products, and the remainder

as the Division Director in Anesthetic Critical

Care and Addiction Drug Products where I had the

responsibility, among other things, for therapies

to treat various painful conditions.

As such, I have been in the position of

approving drugs before. I know that it is not

always straightforward and there is a great deal to

220

consider. I also know that there is a fair amount

of discretion in making these decisions and that,

perhaps, makes the decision-making even more

difficult.

I think I have to say that both approval

decisions and not-approval decisions can have

important consequences for patients. I believe in

the FDA process. I believe in the mission of the

FDA and I have enormous confidence in the FDA.

With all of its oversight systems of checks and

balances, high ethical standards, it does, really,

a remarkable job in protecting the public health,

both in not approving drugs that are not effective

and safe and also in approving drugs that are. The

job isn't without its struggles. The recent public

criticisms that FDA has had to endure which, in my

opinion, are unwarranted, make it even more

difficult.

But I am speaking to you today not as a

former regulator but as a chronic migraine patient.

I know what drugs are available, both for

prophylaxis and for PRN treatment of acute migraine

221

attacks and I have tried most of them. I know the

side effects of the medications and I have been

able to make informed decisions because the FDA

process works and because product labeling is

usually fair and very complete regarding the risks

of medications.

So, two and a half years ago, while I was

serving as FDA's Director of Anesthetic Drug

Products, I suffered a serious, potentially

life-threatening, adverse event from an approved

migraine drug which I took on the day before an

advisory committee meeting, much like this one,

although the topic was prescription drug abuse and

oxycoton which probably explains the headache.

I found myself in the intensive-care unit

and I have permanent sequelae from the event. But

the thing is that it left me with no realistic

treatment options for migraine and I, as many

others, consider this a disorder that can be very

incapacitating at times.

So consider for a minute what is currently

approved for the treatment of migraine. Most of

222

the approved products are associated with vascular

risk, as you might expect, including significant

cardiovascular risk. These include the ergot

alkaloids, the triptans and, up until spring of

this year, one of the selective cox-2 inhibitors no

longer on the market.

So, for patients who have suffered a

myocardial infarction, or who have cardiovascular

risks, these medications really aren't an option.

So what is left? Not very much. I think we heard,

when we saw the slide with three things on it this

morning, well, eliminate one column. So,

over-the-counter and off-label medications. Those

are my choices right now. So, for people who have

no options, any therapeutic gain is significant.

So, when you consider how to deal with

what is an acceptable risk in this population,

consider the following: the existing medications

have risks that are not trivial. Patients who

suffer from migraines tend to know a lot about

their medications and truthful labeling is

absolutely critical and something they expect from

223

the FDA and something the FDA expects from

themselves, as well. Without it, they can't make

informed choices about their treatment.

To put into perspective the kind of risks

that some patients are willing to take with

migraine to get relief, take a minute to consider

how bad the disorder can be. It took me several

years after the first triptan approval before I was

willing to take one.

I listened to the adverse-event profile

unfold during the NDA review because I was in the

division that reviewed that drug. So when I

finally took one, I did it with my eyes wide open,

with absolute full knowledge and understanding that

there was a cardiovascular risk and what that risk

might be. And I was willing to take that risk,

obviously, and I did.

But my headaches were so debilitating at

that time, and so incapacitating and really

impaired my ability to work, that I was willing to

take that risk. So that is how bad it can be.

Here we are with an armamentarium that

224

includes three, now, actually, two classes of drugs

that are off-limits for a segment of the population

like me, and few options. Right now, you are

discussing a drug that has a different kind of side

effect. It has been implicated with the potential

for having a risk of tardive dyskinesia with

chronic administration.

I am a neurologist. I know what tardive

dyskinesia is. I have seen it. It is a bad

condition. It can be very debilitating. So the

question is is that worse than an MI? I don't

know. That is what I am asking myself. I think

each patient has to make that decision themselves

as to what risks are worse than others and what

they are willing to take.

It might be, if there were some benefit

that I could gain from that drug. But I can

absolutely guarantee that, if I didn't get any

benefit from the drug on the first administration,

I wouldn't be taking it more than once. That is

one thing that I think we heard this morning

about--that is pretty characteristic of people with

225

migraine. We self-select our medications. You are

not very likely to accumulate a cadre of patients

who are continually deriving no benefit and only

risk. So that is something to put into the

equation.

My message to the committee and to Dr.

Katz is the following: please keep an open mind.

Migraine can be incapacitating. For a segment of

the population, the drugs that are currently

approved don't exist. While it would be ideal to

have criteria that would give us the perfect drug,

patients like me don't expect perfection and would

be very happy with any therapeutic effect so long

as the labeling truthfully reflects the risks as

well as they can be known.

Thank you.

DR. KIEBURTZ: Thank you.

Is there anyone else who would like to

address the committee during the Open Public

Hearing? Not hearing any, thanking Dr. McCormick

for her statement, we will close the Open Public

Hearing phase of this meeting.

226

Committee Discussion and Response to FDA Questions

DR. KIEBURTZ: We will now, as a group,

deliberate the questions that were posed.

Committee members, just so you are familiar, I

believe, just from my seating chart, that Dr.

Porter is the only non-voting member.

So, on the five questions, we have to

individually vote our opinion about the questions

posed. Some of them are multipart so it is going

to be a little difficult. We can discuss the

question and then we need to come to an answer. At

the time you give your answer, or your vote, you

can get some clarification as to why you are voting

that way.

This will be a chance for the agency to

get a clarification of how we are thinking.

Here is the first question. Take it as a

given that this has been estimated as an annual

incidence--that is, the number of new cases of

tardive dyskinesia--at this dose for use of up to

six doses per month, so up to 72 doses per year,

380 cases, new cases, of tardive dyskinesia per

227

million patients per year.

Do you think this is a reasonable

estimate? Dr. Fahn.

DR. FAHN: I would like to start. I did

some calculations on the margins of the slides that

were handed out. The risk for developing--the

incidence for developing tardive dyskinesia has

been calculated by John Kane and his colleagues at

Long Island Jewish who has studying the

epidemiological of tardive dyskinesia for decades.

He calculated the risk to be about 5

percent per year, so a person would take--now,

these were neuroleptics, but these were

dopamine-receptor blocking agents which, of course,

metoclopramide is also. The calculation was that 5

percent per year, that if you took it for four

years, then you would have a 20 percent risk of

getting tardive dyskinesia and so forth.

Multiply 5 percent per year times 72 days

divided by 360 days, I come up with a 1 percent

risk, not 0.38 percent risk. Now, granted that the

drugs that were used in the psychiatric population

228

are daily doses continuously, so that might have

been the risk for 72 days in a continuum. What

would happen if it was discontinuous? I think this

is the big unknown. I don't think anybody, as we

heard all day so far--that no one really knows what

these intermittent or periodic risks are and it may

be less.

The intermittency, when you stop--for

psychiatric patients, of course, and then restart

it, the risk seems to go up. But, giving it for

migraine is one thing. So I think this a level

that is probably too low. I think it, at least

from these calculations, is going to be probably

closer to the 1 percent risk.

If you want me to address the second part

of that question, I would be glad to do that.

DR. KIEBURTZ: Let's just stick with this.

I mean, we can have a little bit of general

discussion about whether we think it is reasonable

or not, and then we are going to have to just each

vote about that question.

Dr. Jeste, did you--

229

DR. JESTE: Actually, unrelated in a way,

but one general point I want to make. Since the

morning, there has been a lot of discussion of how

do we define tardive dyskinesia. I would argue

that there is, actually, a standard definition in

the DSM IV for tardive dyskinesia. By definition,

as Dr. Jinnah said, that is later occurring. The

word "tardive" means late occurring.

So, if it occurs in less than a month, it

should not be tardive dyskinesia. That doesn't

mean it cannot be persistent. So we need to

separate out tardive dyskinesia from persistent

dyskinesia. Acute dyskinesia can be persistent but

we should not call it tardive dyskinesia. I think

it is somewhat of an oxymoron to say tardive

dyskinesia that develops after one day of

treatment, of any drug, for that matter.

The DSM IV definition is minimum 90 days

except for older people where it is 30 days or

more. One can use different definitions for

possible different types but, again, in terms of

standard nomenclature, I think we should stick to

230

that definition. That would be my recommendation.

DR. KIEBURTZ: Dr. Porter.

DR. PORTER: Yes. An epileptologist

should never tackle a movement-disorders expert in

his own field, but, Stan, I think that you have

gone way too high.

I think that, if it were really 1 percent

for this particular drug in migraine, we would be

seeing 95 patients every year showing up in the

U.K. yellow-card system. While I think maybe the

U.K. yellow-card system is flawed, I think that it

is not nearly that flawed, especially when the dose

given in the U.K. is more than twice the dose as

proposed for the tablet here in the U.S.

So I think it is--actually, I think that

the company's estimate is high and I think your

estimate is very high.

DR. SMITH: I want to comment, because my

area is pharmacoepidemiology, and I feel that these

databases are kind of being maligned and misused.

I want to really make sure that that is understood,

that the use of reports and divided the number of

231

cases is, first of all, not what these databases

are meant for. It is a gross misuse of them and it

is a very dangerous misuse, particularly in this

case. If anything, it is probably the lowest

ballpark estimate of what the risk might be and it

is not even a good estimate at that.

One of the things we are pretty much, I

think, in agreement on without even voting is that

it is not a simple disease to diagnose, that there

is variation in what makes a diagnosis of TD, that

part of the diagnosis criteria does include, in

some places, the length of therapy which would

preclude people from reporting something it if

was--as TD, if it was after just one or two uses,

so we don't even know. So the reporting rate for

that would be substantially lower than what we

think.

The drug has been on the market a long

time. It is labeled for this so it is not

unexpected. So estimating the reporting rate is

just--it is just a guess. It is like throwing an

arrow. We don't know if it is 1 percent or

232

10 percent or 1 millionth of a percent. It is a

gross unknown and I think this is something we

really have to think about whether or not these

numbers are based on anything that is reasonable.

DR. LENAERTS: When we consider the risk

of tardive dyskinesia or other movement disorders,

would it be appropriate, so that is a question to

anybody, or a comment, to consider the relative

risk of migrainers where, on the one hand, we have

probably a relatively younger population. But, on

the other hand, we have significantly more women

with a 3-to-2 ratio of women to men, and also the

fact that the coadministration of tricyclic

anti-depressants and the SSRIs which we have seen

can give its own risk is particularly common in

that population of patients.

If I take the numbers that you showed, on

67 cases of TD, 14 also took anti-depressants which

between 20 and 25 percent. Is it appropriate to

consider the special population that migrainers

represent in that context?

DR. KIEBURTZ: Sure. I think we should.

233

Dr. Green.

DR. GREEN: I think that dopamine

hypersensitivity is right at the heart of migraine

pathophysiology and, therefore, it is not clear to

me that any data we get from the general population

rather than the migraine population is relevant to

this question.

I don't know if that impacts negatively,

positively, or not at all. But this is a condition

of dopamine hypersensitivity. But also the other

comment is I agree with several comments that there

is a huge need for a medication that is effective

in the treatment of migraine that is not

vasoactive, a desperate need for that.

DR. KIEBURTZ: Dr. Hughes.

DR. HUGHES: I try to explore the issue of

the safety study that the company had done this

morning. I think--you know, you have got 300

subjects followed for a year and 1,000 followed for

a few months. I think it would be hard to conceive

of the idea that the true rate would be 1 percent

or more, given that they didn't really see any

234

events or they saw zero events according to the

definition that they were using.

So I would hazard a guess that it is

somewhat below 1 percent. But how much below, who

knows?

DR. KIEBURTZ: One last comment.

DR. JESTE: I really think it is fair to

say we do not know the risk and we will not know

the risk unless and until we do a longitudinal

prospective study in which rigorous measures are

used to find out whether the patient has TD and one

has to use a scale like involuntary movements

scale, use some fixed criteria.

There are a number of confounds in

deciding the incidence of tardive dyskinesia. For

example, the drugs that produce TD are also the

drugs that are suppress it. So, if we increase the

dose, you may not see TD.

It also depends on what other medications

the patients are taking. If they happen to be on

dopaminergic drugs, for example, that might

precipitate TD. If you look at the history of the

235

neuroleptics in psychiatry, it is worth remembering

that it took more than 25 years before the field

accepted that there was such a thing as tardive

dyskinesia.

People argued that it was a symptom of

schizophrenia for many years. The only time that

the incidence became clear is when a number of

people did longitudinal prospective studies using

that. So I really don't think anybody can make a

claim for any kind of--whether it is lower or

higher, it doesn't matter. We just don't know the

risk at this stage.

DR. KIEBURTZ: Okay. So we are going to

vote now. I am going to go around the table, which

is, this is the proposed estimate, 0.038 percent,

as an annual incidence. Do you think it is a

reasonable estimate. We don't have a definition of

reasonable, but let me operationalize that. If we

are proposing 0.04, there is some range around that

from 0.01 to--an order of magnitude around that.

But you have a sense of where the estimate

is. Go ahead. Dr. Green? It is either yes, no or

236

abstain. But you can make a comment.

DR. GREEN: The answer is no.

DR. KIEBURTZ: Dr. Jeste.

DR. JESTE: I don't think, really, there

are any data to say this is a reasonable estimate

at this stage.

DR. KIEBURTZ: So the answer is no.

DR. JESTE: The answer is no.

DR. SMITH: No.

DR. WELCH: No.

DR. LENAERTS: No.

DR. FAHN: No.

DR. JUNG: No.

DR. GOLDSTEIN: No.

DR. SACCO: No.

DR. KOSKI: No.

DR. HUGHES: No.

DR. KIEBURTZ: I will say yes because I

think we don't know but, from the evidence, we have

some reasonable bounds, as Dr. Hughes alluded to,

that we certainly don't think it is 1.0 or higher.

It is somewhere under that. We don't have enough

237

evidence to conclude it is zero, but what is in the

reasonable range of between zero and 1.0 and I

think this is reasonable.

Dr. Temple.

DR. TEMPLE: Could I ask whether the

people who said no think it might be lower, higher

or just can't say.

DR. KIEBURTZ: I am going to do a show of

hands. How many people think you just really can't

say. There is not enough evidence to say, and that

is why you voted not. (Show of hands.)

DR. KIEBURTZ: That is seven.

DR. TEMPLE: I interpret that as meaning

they think it could actually be higher. Would that

be--

DR. KIEBURTZ: Just don't know. Dr. Fahn,

do you think it is higher?

DR. FAHN: I think it is higher.

DR. KIEBURTZ: Dr. Lenaerts does too.

Three people think-- does anybody think it is lower

than this estimate?

(No response.)

238

DR. KIEBURTZ: Okay. Is that helpful to

you?

DR. TEMPLE: Yes.

DR. KIEBURTZ: I am just going to reverse

the order a little bit here in the next set of

questions because I think one moots the other. So

let's just go to the third question here. Is any

risk of tardive dyskinesia acceptable for a

migraine population? If we conclude that there is

none, the middle question, is not--so, just

logically, I would like to go with that question.

I know Anuja is going to be mad at me for going out

of order.

Any discussion about that?

DR. GOLDSTEIN: It is a takeoff on one of

Dr. Matchar's slides, actually, where he said 79

percent of sufferers showed an interest in trying a

novel product with similar efficacy but fewer

adverse events than existing migraine medications.

The point is that this is obviously

another imponderable. We heard testimony that some

people might take that risk, whatever that risk

239

could be defined. But it is possible, through

decision analysis, to get some bounds for what the

people might--what patients might be able, or

willing, to accept as a risk for having a permanent

or even a transient sequela.

So, right now, we are just--we are making

this up as we go along. But there is a way to

actually get data to be able to address that point.

What I would encourage would be to try to formally

get that type of data. The studies are not that

difficult to do.

DR. KIEBURTZ: Dr. Koski.

DR. KOSKI: No; not right now.

DR. KIEBURTZ: Dr. Fahn.

DR. FAHN: I think the risk of getting

tardive dyskinesia on the first dose is extremely,

extremely, rare and even the second dose and so

forth until you start taking multiple doses. So,

when Dr. McCormick made her presentation about

migraine sufferers would not continue a drug if it

wasn't working, that shed a new light on me. I

mean, that is an experience--they are not going to

240

keep on taking this drug.

Therefore, I think the risk, then, is not

for the people who it doesn't help continuing to

take it. The risk, then, becomes those in which

the drug helps and then the drug, as they continue

to take it, the risk keeps going up every time they

take it.

But, because a drug is beneficial to those

patients, then I think the risk is probably worth

it for those people. Therefore, even a 1 percent

risk in a year, if it helps the patient, that may

be something the patient has to decide and they

know what that risk is and would take it.

So, therefore, I think the answer to this

is yes. Risk is acceptable for the migraine

population at least that gets a response to the

drug.

DR. KIEBURTZ: Dr. Welch.

DR. WELCH: My answer to this would be yes

as well. It depends on the migraine population and

the way the drug is used. For migraine patients

with severe disability that haven't responded to

241

anything else, or there are contraindications to

using such a medication, then this would be a risk

worth taking. So my vote would be yes.

DR. KIEBURTZ: Dr. Green. You don't have

to say about your vote yet. We are just doing

general discussion. But you can if you would like.

DR. GREEN: Again, in this special

population where vasoactive migraine therapies are

not an option, it is wonderful--it would be very

useful to all of us to have additional options. I

would disagree, however, with the comment that

migrainers don't overuse, or wouldn't be likely to

overuse, something that is ineffective. In fact,

it has been my impression that migrainers are less

likely to overuse something that is fully effective

and less likely to overuse something that is

completely ineffective and the things in the middle

which give partial relief are the things which put

people at risk for redosing.

DR. KIEBURTZ: Dr. Temple.

DR. TEMPLE: Because you skipped to the

third question, there is no context here about the

242

magnitude of the benefit from the thing that puts

you at risk. This drug is going to work on people.

It is obvious. Naproxen works in migraine just

like ibuprofen and aspirin work in migraine. So

there isn't any question it is going to work.

The question that was flagged there is how

much is it worth to you to get the added benefit of

the metoclopramide. That should be kept in mind, I

think, even in this part, though.

DR. KIEBURTZ: I think it is pretty

obvious we are going to get to the third question

because it doesn't look like there is consensus.

Dr. Koski.

DR. KOSKI: I really would like to second

that. In the sense that patients, because of the

fact that they can get a response from naproxen,

are going to get something, or at least a

significant portion of patients will get something

in response to the drug.

The question is is there something

additional that is taking place with the

metoclopramide addition. Because of that reason, I

243

sort of disagree with Dr. McCormick. I think that

this would not be a one-time event, that this would

be something that people might continue to take.

Certainly patients I have seen, some of

them, quite frankly, think if one tablet works

okay, maybe they should take two and then,

occasionally, maybe I have something--maybe it is

coming on and I am always told to take something

just as soon as I possibly can. I will take

another one then.

So I think there is a great deal of

potential for that type of abuse.

DR. KIEBURTZ: Let me just clarify the

question, here, too, which is not a

product-specific question. Is any risk of tardive

dyskinesia acceptable for a migraine population?

It is a question that is independent of benefit.

That is why I went there first. But it is a very

abstract question, as I read it. But maybe I am

reading it wrong.

Dr. Jung.

DR. JUNG: Speaking as the consumer

244

representative, my concern is that there would be

adequate patient education information about what

actually tardive dyskinesia is, given the fact

that, in this room of neurologists, we can't seem

to agree how you actually make a diagnosis. How do

you anticipate that, in your office, you are going

to be able to explain to a patient or be able to

show a video, for example, such as was shown

earlier, of what a significant risk is.

I think most people can understand the

risk of what a heart attack or a stroke is. But to

describe tardive dyskinesia and have the worst-case

scenario be apparent is not clear to me, that we

can actually do that in the clinical setting.

DR. KIEBURTZ: Thanks.

Dr. Welch, did you have a comment?

DR. WELCH: No; I was going to address--

DR. KIEBURTZ: Okay. Dr. Sacco.

DR. SACCO: I think we have other data

from other patient populations that obviously have

accepted the risk of medical metoclopramide for

GERD. So I think it depends on probably the amount

245

of risk. Any risk is really tricky if it is 0.0001

and the benefit. So it is a risk/benefit ratio

but, in my mind, the fact that other patient

populations with pain somewhere else, including

gastroesophageal reflux, have accepted the risk of

the medicine that is out there, then any risk may

be acceptable. But it depends on the amount.

DR. KIEBURTZ: Dr. Smith.

DR. SMITH: I want to agree that just to

say is there risk without thinking about the

benefit, it just can't be separated unless you are

talking about cure and a one in a million risk.

Then maybe the answer could be yes. But I agree

that you have to think about what is the added

benefit when you know you are going to add this

risk beyond what is already out there.

DR. KIEBURTZ: Maybe I misread this third

question. Maybe it is meant to be contextualized

on the second question and not stand alone. Is

that true?

DR. KATZ: No; I think you read it

correctly. We wanted to find out if it is just

246

drug independent. But, of course, obviously, the

amount of risk you are willing to tolerate will

depend on what you think the benefit is. That is

obviously critical. But we just wanted to sort of

get a baseline because, as you say, it moots the

rest of it if everybody thinks no risk is

acceptable, no level of risk.

DR. KIEBURTZ: So let's vote the question

because I think we know the answer to this. Is any

risk of tardive dyskinesia acceptable for the

migraine population? We will go this way this time

to mix it up. I say yes.

DR. HUGHES: Yes.

DR. KOSKI: Yes.

DR. SACCO: Yes.

DR. GOLDSTEIN: Yes, with just the

additional comment that, again, there is a way to

get the information so that we are not

guesstimating. We can get this.

DR. JUNG: With some hesitation, yes.

DR. FAHN: Yes.

DR. LENAERTS: Yes, too, but for the same

247

remark. It depends on the efficacy. It is hard to

take out of context.

DR. WELCH: Yes.

DR. SMITH: A qualified yes.

DR. JESTE: Yes.

DR. GREEN: Yes.

DR. KIEBURTZ: Back to Question 2 which is

a little more complex. I think what we are to do

here is to assume that the risk of MT100 is 0.04

percent, let's say.

DR. KATZ: That is what the question asks.

But it turns out most people don't think that that

is a reasonable risk. Some people think it is

higher--or a reasonable estimate of the risk. So I

am not sure we want to sort of pin our risk/benefit

consideration on some risk estimate that nobody

believes.

Maybe the best thing to do is just sort

of--I mean, everybody has probably got their own

personal view of what they think the risk is. You

can't tell. I think maybe we should try to talk

about what sort of a--let's assume the benefit part

248

of this question, the 5 to 10 percent contribution

of the metoclopramide and then sort of talk about

what sort of risk you think would be acceptable.

That is really more or less the question,

and whether or not we think we are at the point yet

where we think that that is the risk.

DR. KIEBURTZ: Okay.

DR. KATZ: In other words, I don't want to

link it to a number which everybody has already

decided doesn't mean anything.

DR. KIEBURTZ: The way that we were

discussing the last question it got posed

around--several people volunteered it depends on

how much benefit you are seeing. So let's turn it

around and say, in the factorial designs, leaving

aside the question of statistical significance, the

proportion of individuals as reported as the

percentage of individuals who had sustained pain

relief at 24 hours differed in the MT100 group

versus the naproxen group by between 3--let's say

between 4 and 6 percent.

So between 4 and 6 percent additional

249

subjects had sustained pain-free response.

DR. BASTINGS: It is not pain-free. It is

just sustained pain response.

DR. KIEBURTZ: I mean sustained pain

response. Thank you for the clarification. We are

assuming that is true, that is the true effect, for

the purposes of this hypothetical discussion.

DR. BASTINGS: You also need to take into

consideration that there would be no benefit at two

hours, which is--

DR. KIEBURTZ: So that is the

parenthetical statement below which is, if you

looked at the comparison of MT100 to naproxen at

the two-hour pain response, those differences were

not statistically significant. They were

numerically different but they were not

statistically--Dr. Bastings' presentation showed

that in that, in the one study, quite a number of

people had a lower relapse rate and that added up

to the total difference between the groups.

In the second study he presented, the

majority of it was actually in the two-hour

250

response and there wasn't that much difference in

the relapse rate. But the numeric difference, even

at two hours, was 3.1, I think, if I remember

correctly.

So the comparisons at two-hour pain

response are not statistically significant. So,

overall, the effect is 4 to 6 percent but the

two-hour comparison of pain response is not

significantly different.

Have I framed the question correctly? We

are changing on the fly here.

DR. KATZ: In our hands, the way we have

analyzed it, the studies which we believe were what

the protocol said, the contribution of

metoclopramide for a sustained response, which was

the primary outcome, is not statistically

significant overall. It only seems to be nominally

significant when you look at the non-nausea

baseline subgroup which is partly where this 5 to

10 percent comes from.

There is a 10 percent contribution of

metoclopramide in the non-nausea at baseline

251

subgroup which, again, is normally significant in

two studies. We don't think that that is true

replication but, in the overall group, we don't

think that the contribution, even for sustained

response, is statistically significant

DR. KIEBURTZ: I wasn't suggesting that

you did.

DR. KATZ: Okay.

DR. KIEBURTZ: Everyone has got this

definition here. So the question is what amount of

risk would be reasonable for that amount of

clinical benefit. So we are not saying is that

risk reasonable. We are saying what is reasonable

with that amount of benefit, if you can follow the

hypothetical situation.

DR. SACCO: I just want to be sure. It

sounds like, in the hypothetical situation, we are

giving the compound more benefit than what is seen,

5 to 10 percent, the way this question is phrased,

and I understand where you may be getting with

not-nausea is to--and possibly an exaggeration of

the benefit.

252

DR. KIEBURTZ: Okay.

DR. SACCO: If we take what we got, which

was 4 to 6 overall, this question is not phrased as

just in those without nausea. So you are giving it

the benefit--

DR. KATZ: No; right. It is not. But the

range, 5 to 10 percent, includes--so the estimate

that was in the overall population and the maximum

difference that you saw in the no nausea. We could

talk about whether or not a study ought to be done

to replicate the no-nausea subgroup of whether

another study or studies should be done in the

overall, but we just want to know, is there some

well-defined population in whom the

contribution--to be determined what that is--if the

contribution of metoclopramide is in this range, as

a ballpark, because that is where the estimates

have sort of fallen out. Are the risks, whatever

they are, acceptable?

So I think of this range. We could figure

out which population we think that range applies to

down the road.

253

DR. TEMPLE: Our assumption is that there

is not going to be a bright line between 5 and 10

percent. They are both fairly modest. One is more

modest than the other, but that is a reasonable

ballpark to think about. If there is a bright line

that you have, say so.

DR. KIEBURTZ: Dr. Welch and then Dr.

Fahn.

DR. WELCH: Well, let's try and instill

some practical clinical reality into the question.

Surely what the question is saying is if there is a

small subpopulation, and it would be small at this

level of 10 percent improvement, that responds to

metoclopramide plus naproxen that doesn't respond

to naproxen and that that can be significantly

proven, and if you have answered yes to the risk of

tardive dyskinesia acceptable for a migraine

population, the answer would have to be yes for the

second part because you have said that you will use

this drug, or certainly you would use

metoclopramide, in a particular population that has

benefit that can't take anything else.

254

So the issue here is, have you really

proven that metoclopramide plus naproxen is, in

fact, statistically significantly different in a

subpopulation. I would suggest that it has not

been proven that 5 to 10 improvement is in the

nausea versus no-nausea, and we are going to talk

about that later. But, to me, that is not a stable

measure as yet.

But, if it was, there is one in ten

patients who responds to naproxen plus

metoclopramide that wouldn't respond to naproxen,

then you would use naproxen because there is

nothing else yes, the answer would have to be yes.

DR. FAHN: That is the point I was

essentially going to make and that is what I was

going to ask the FDA, if this drug were available

on the market, if it was restricted to those who

were, on naproxen alone, failed and then they still

had that 5 to 10 percent chance of getting better

with this combination, then I think the answer is

yes.

If it is just going to be open and anybody

255

can take the drug without even trying naproxen

alone first, then the risk for tardive dyskinesia

is too great for me. So I assume there will be

some kind of restriction that they have to be

tested first with naproxen alone because,

otherwise, I think the risk would be too great.

But if it is restricted that way, then I say, this

is a worthwhile risk.

DR. KIEBURTZ: Let me just refocus things

again. By pursuing this hypothetical question, we

are not agreeing that there is a benefit of MT100

over naproxen. We are just saying hypothetically,

if it were so. We are not agreeing that we think

that that has been shown. So I just want to make

that one statement.

Two is I don't think we can assume that we

know how the restrictions are going to be. That is

not how this conversation, I don't think--

DR. FAHN: I think the point is if more

people will be exposed to this, then more people

are going to get tardive dyskinesia. But if the

exposure is limited to those people in which this

256

drug may be the only thing left, then it is a

reasonable risk.

DR. KIEBURTZ: Dr. Porter.

DR. PORTER: I completely agree. If you

look at the history of cancer chemotherapy of these

kinds of incremental improvements are what have

made a lot of cancers much less a threat than they

used to be. I think that we would have to take

these little bites when we can get them.

I agree fully with the thesis that this is

a drug that, A, has to be proven. I am not sure

about this business about whether or not the nausea

is or is not related to its efficacy. I agree with

that. Certainly, there will be some labeling, I

hope, that would be relatively strong relative to

the risk we have been talking about.

But I think to reject the drug when it

might help some people who have gone through a

parade of migraine drugs would be unfair to them.

DR. JESTE: I am not a migraine researcher

but, based on my experience with psychiatric

patients, just a couple of points here. The one is

257

that the risk of tardive dyskinesia varies from one

patient group to another. Clearly, elderly

patients, the risk is 30 percent incidence with

conventional neuroleptics compared to 5 percent in

younger adults. So something like that will have

to be taken into account.

Secondly, all tardive dyskinesia is the

same. It can be pretty mild which is not a

problem, or it can be pretty serious when it does

become a problem. So it depends, really, on what

kind of tardive dyskinesia and what population. So

those will need to be taken into account in

whatever final decision is made.

DR. KIEBURTZ: Dr. Bastings.

DR. BASTINGS: I would also like people to

comment, if possible, on the fact that there would

be no effect on two hours but yet there would be an

effect demonstrated in a sustained endpoint. The

sustained endpoint is a measure which includes

two-hour pain response. There is a possibility to

have a significant effect on sustained response

without actually having an effect on the two-hour

258

pain response which is a particularity.

I would like to have some idea, in that

context, if that is an acceptable outcome measure.

DR. GOLDSTEIN: Again, just to interject a

little clinical thought that, let's say, if you

were a patient and you were having a migraine and

you took this and in two hours you weren't having

any benefit. Chances are you might, then, take

something else. You might just take naproxen

alone, for example, or some other nonsteroidal

alone. So the question I would have, as these

trials go on, from a practical standpoint, would be

is that kind of additional dosing factored in to

the 24-hour improvement because, if you factor that

in as well, then the clinical benefit might even be

less or, in that subpopulation of people who don't

respond to that second added drug, the benefit

might even be greater.

It is hard to know without that

information. They may have it but I don't know--we

didn't see the data that way. But that, I think,

would address the question that you are asking

259

here. Nothing at two hours; then is there

sustained relief. I think you have to see what

else was done at that two-hour standpoint to know

that.

DR. KIEBURTZ: That would be a relapse if

you had to take anything.

DR. GOLDSTEIN: Yes.

DR. TEMPLE: In these studies, as

conducted, and acknowledging that SPID and TOTPAR

might give you slightly different impressions, but

nobody would know the difference between this drug

and naproxen as near as we can tell at two hours.

So I think the question being raised, and it is

part of it, is when you think of somebody with a

terrible migraine that won't go away, you think of

the early response mostly.

Mostly that is just naive and we should

have been thinking about the sustained response all

the time. But part of the question is is this

modest effect on that endpoint and not on the

two-hour endpoint worth a certain amount of risk of

TD. We are not saying no, but we want to gain some

260

impressions of what you think.

DR. KIEBURTZ: This is a very tricky

question and we are modifying it en route. But the

trickiness of the question lies in how big the

benefit is and the fact that the benefit accrues on

a synthetic measure, one that incorporates both

immediate response and relapse into a single

measure when you have benefit on that but no

benefit on the immediate measure, if I am--

DR. KATZ: That is right with regard to

the contribution of the metoclopramide. It is

obviously complicated because you could construct

maybe a better 24-hour sustained-response in which

you say--right now it is just you are considered a

responder if you responded at two and throughout

the next 22 hours. But, at two, the drug might not

be better than placebo so you wouldn't win on the

typical thing.

But you could construct a

sustained-response outcome which says you have to

beat placebo at two hours and--it is a two-part

outcome--and you have to win on the next 22 hours

261

as well. So you could do that. But it is

complicated because it is possible that the drug,

itself, MT100 could win on that particular outcome

but the metoclopramide doesn't contribute to that

particular outcome. So you have to see what would

happen if you did that.

DR. TEMPLE: There is no doubt that the

whole drug has an effect at two hours. Nobody is

doubting that. It is all about the contribution of

the metoclopramide.

DR. KATZ: Right. So the question is do

you think the metoclopramide has to contribute for

the two-hour point as well as the sustained portion

of it or is it okay if the drug overall wins at two

hours on this new sustained but the metoclopramide

only contributes to a part of it.

DR. WELCH: Again, I don't believe that it

has been established yet that there is an efficacy

so it is an assumption. So now you are talking

about what kind of efficacy measures you would like

to see. I would like to see a two-hour pain-free

sustained. I think that is the rigorous measure

262

and I think that, to show a difference between

MT100 and naproxen, pain free at two hours.

DR. KIEBURTZ: Dr. Bastings.

DR. BASTINGS: I would like also to point

out that this does not even include the issue of

the effect on relapse or use of rescue medication

such as in the factorial studies of the NDA, there

was no significant difference for relapse or use or

rescue medication either. It is just the combined

effect which gives you an effect.

DR. KIEBURTZ: So if you split the

combination thing of immediate response and relapse

rate, neither of the individual components was

significant. But when you combined them, that was

the only measure in which you achieved

significance.

We are going to vote this question. What

question are we voting? Let's vote this question

which is, if MT100 were to carry the same risk,

would such a risk be acceptable if the only

contribution of the metoclopramide is a 5 to 10

percent improvement on the sustained headache

263

relief with no improvement on the two-hour

endpoint.

DR. WELCH: Can you just qualify this

question a little bit because the question is in

the question. If you put with no effect on a

two-hour endpoint, I would have to say no. But if

you really said that there is a 5 to 10 percent

improvement on sustained headache relief as defined

by two hours pain-free and then no relapse, then

the answer would have to be yes.

DR. PORTER: 5 percent to 10 percent of

patients?

DR. WELCH: Yes; 5 percent to 10 percent

of patients. It is not 5 to 10 percent--which is

critical here because it is extra patients who can

respond who would not have responded as opposed to

5 or 10 percent more pain relief. It is very

critical.

DR. KIEBURTZ: Let me just clarify. The

question is what the question is, not as you

modified. I know that it not the question you want

to answer. But this is the question.

264

So, Dr. Green.

DR. TEMPLE: Just one thing. But it is

all about the contribution of metoclopramide.

DR. KIEBURTZ: Right.

DR. TEMPLE: Nobody doubts that the whole

drug works in two hours.

DR. KIEBURTZ: It is the contribution of

metoclopramide, 5 to 10 percent, on the sustained

measure with no effect on the two-hour measure

which is not pain-free. It is just pain relief.

DR. GREEN: Okay. Then I would say no.

DR. JESTE: No.

DR. SMITH: No.

DR. WELCH: No.

DR. LENAERTS: No. But I would like to

make a comment, if I may, afterwards.

DR. KIEBURTZ: Go ahead. Comment now,

please.

DR. LENAERTS: Of course, I just want to

clarify that, but we are really not talking about

taking people who have not responded to naproxen

alone because that would be--it gave the

265

impression, sometimes, in the discussion as if we

were heading that way.

The other thing is even though the outcome

of how they do at two hours and how they do between

two and 24 are both important. But, as far as my

experience and my reading on the subject, it is

clear that the weight of the two-hour response is

much more significant than that of the two to

24-hour period.

DR. FAHN: I got a little confused now in

the discussions. I thought that this was the added

benefit of having metoclopramide on top of

naproxen, you get another 5 to 10 percent of

patients getting better with less headache. So, on

that basis, I think that is yes.

DR. JUNG: No.

DR. GOLDSTEIN: No, not as this question

was written.

DR. SACCO: I am still confused about the

question. But I am going to say yes because I am,

first of all, reading it that if MT100 were to

carry the same risk, and I am reading that risk of

266

TD as being 0.038 percent--am I correct in my

assumption?

DR. KIEBURTZ: Yes.

DR. SACCO: Which is, I think, lower

than--

DR. KATZ: The way I would think of it, or

the way I think we would want you to think about

it, is assume, again, the benefit portion of this

is true, the 5 to 10 percent, not at two hours but

on sustained--assume all that is true. But we have

already determined that nobody believes 0.038. So

why would we want to link it in a risk/benefit sort

of consideration.

I would say whatever you think the risk is

now. Do you think the risk--

DR. SACCO: See, that's--it is a

risk/benefit question.

DR. KATZ: It is. But these are the data

we have. So we have to make a decision based on

these data. You might decide, well, since we don't

know what the risk is, I can't say yes to this, or,

since I believe that the risk is pretty low,

267

whatever I personally think it is, it would be

worth that risk. You have to make an individual

decision based on what you believe the risk data

are.

They are what they are. We can't do

anything about that.

DR. SACCO: Because I was looking at these

questions as true hypotheticals and reading them

very literally, it sounds like, then--my answer to

this is conditional on what I believe the risk of

TD is.

DR. KATZ: Absolutely. I think is how you

have to answer it.

DR. SACCO: So then I revise my answer and

I would say no.

DR. KATZ: Let me just say I think

everybody should address the question that way and

if the previous no's were assuming 0.038, maybe we

need to--

DR. KIEBURTZ: We will go back. Dr.

Koski?

DR. KOSKI: I think specifically, since

268

one of the things that I was concerned about had to

do with something that Lily actually mentioned and

that is if you split up all of these extrapyramidal

disorders, you know, whether they come on rapidly,

as long as they are persistent, I think they need

to be grouped together because that is going to

have some impact on the patient. So my response

would be no.

DR. KIEBURTZ: Dr. Hughes.

DR. HUGHES: I would respond yes,

particularly if it was at the upper end of the

range, towards 10 percent.

DR. KIEBURTZ: My vote is no.

Still no?

DR. GREEN: No.

DR. KIEBURTZ: Still no?

DR. JESTE: No.

DR. KIEBURTZ: Still no?

DR. SMITH: No.

DR. KIEBURTZ: Still no?

DR. WELCH: No.

DR. KIEBURTZ: Still no?

269

DR. LENAERTS: No.

DR. KIEBURTZ: Yes still?

DR. FAHN: Yes.

DR. KIEBURTZ: Still no?

DR. JUNG: No.

DR. GOLDSTEIN: No, again with there being

no effect at the two-hour endpoint because the

reason for that was that I think those patients

would take another rescue medicine where we have

heard that there is, then, probably no benefit.

DR. KIEBURTZ: Did you get the information

you wanted from the discussion and the voting?

DR. TEMPLE: What I hear is that, assuming

that benefit as described, not much or nothing at

two hours and something at 24 and longer, and in

light of what you will each individually think the

risk of TD or persistent dyskinesia might be, you

are saying no.

DR. KIEBURTZ: I believe that is what you

heard.

DR. WELCH: The reason being is that if

you don't have a difference at two hours, then you

270

can't prove that it is a metoclopramide effect and,

therefore, you shouldn't be putting the patient at

risk.

DR. TEMPLE: Why can't you attribute the

overall benefit at 24 hours to the metoclopramide?

That is the only difference between the two groups.

DR. WELCH: There may be other factors in

between which you can't stratify for.

DR. TEMPLE: Okay. But presume it is a

properly randomized trial and everybody is treated

exactly the same way in all other respects, the

usual concerns one would always have, I guess we

thought if they could actually win in a persuasive

way, you would have to attribute it to the

metoclopramide and the main question was how

valuable is that in light of a certain amount of

risk.

But I guess we thought that if they did

the study properly in one, you would attribute the

difference between metoclopramide-naproxen and

naproxen to the metoclopramide. But I think we

understood what people were saying.

271

DR. KIEBURTZ: Okay. To summarize. The

vote on Question 1 was one yes and 11 no's with no

abstentions. Question 2 is two yeses and 10 no's

with no abstentions. It was all yeses on the third

part of Question 1.

I think that is the hard part. We'll see.

We are going to go on to Question 2; Is there

sufficient evidence that the chronic intermittent

administration of metoclopramide does not carry a

risk of tardive dyskinesia. If we don't have any

discussion, we can just vote it. So we are just

going to vote it because I don't see anybody who

wanted to discuss it.

Is there sufficient evidence that chronic

intermittent administration of metoclopramide does

not carry a risk of tardive dyskinesia? I note no.

DR. HUGHES: No.

DR. KOSKI: No.

DR. SACCO: No.

DR. GOLDSTEIN: No.

DR. JUNG: No.

DR. FAHN: No.

272

DR. LENAERTS: No.

DR. WELCH: No.

DR. SMITH: No.

DR. JESTE: No.

DR. GREEN: No.

DR. KIEBURTZ: That is unanimous. There

is a subpart question. Is it possible to define a

maximum recommended number of monthly doses of

MT100 to avoid the risk of tardive dyskinesia? Is

there discussion on this, or is there preparation

to vote? Looks like we ready to vote. Dr. Green,

we will start at your end.

DR. JESTE: It is possible to define the

risk but not on the basis of the data that are

already there since one can do a longitudinal

prospective study and then define that.

DR. KIEBURTZ: Okay. So the answer for

now is--

DR. BASTINGS: We mean on the basis of

existing data.

DR. GREEN: No.

DR. JESTE: No.

273

DR. SMITH: No.

DR. WELCH: No.

DR. LENAERTS: No.

DR. FAHN: No.

DR. JUNG: No.

DR. GOLDSTEIN: No.

DR. SACCO: No.

DR. KOSKI: No.

DR. HUGHES: No.

DR. KIEBURTZ: I also vote no. That was

unanimous no's on the two parts of Question 2.

See, that was easier.

Question 3; do you believe that, based on

the existing data on medication-overuse headache,

there is evidence that a proportion of patients

prescribed MT100 will likely take a number of

monthly doses higher than the recommended amount?

Discussion on that question?

DR. GOLDSTEIN: The question again, the

first phrase there, is the existing data on

medication-overuse headache. I think the data on

medication-overuse headache is different than we

274

might think for patients who take more than they

are supposed to take because it may not lead to

medication-overuse headache.

I think, again, we can answer the question

the way it is written, but I want to find out

exactly what they had in mind because, if the

question is, do you think there is a proportion of

patients likely to take more than the recommended

dose, that is a different question than the first

phrase.

DR. KATZ: We were just trying to link it

to previous evidence of overuse of other migraine

treatments. But yes; we are interested to know

whether or not you think people are going to take

more than they are supposed to.

DR. KIEBURTZ: Right. And not so much

linked to the character or anything about what

causes medication overuse headache, just the

question of--is that clear?

DR. LENAERTS: It would be better stated

maybe, on medication-overuse in headache.

DR. KIEBURTZ: Okay. That is the intent

275

of the question. Discussion on that question? Is

it my turn to start?

Is there evidence that a proportion of

patients are likely to take a number higher than

recommended? Yes.

DR. HUGHES: Yes.

DR. KOSKI: Yes.

DR. SACCO: Yes

DR. GOLDSTEIN: Yes.

DR. JUNG: Yes.

DR. FAHN: Yes.

DR. LENAERTS: Yes.

DR. WELCH: Yes.

DR. SMITH: Yes.

DR. JESTE: Yes.

DR. GREEN: Yes.

DR. KIEBURTZ: Very good. Thank you. So

it is unanimous that it is likely that they will

take more than the number recommended.

So we are changing focus here entirely no.

This one is a little bit trickier question. All

currently approved acute treatments of migraine are

276

indicated without restriction regarding the

presence or absence of nausea at baseline.

Given that patients may have nausea at

some attacks and no nausea at others, given that,

does an indication limited to the subpopulation of

migraine patients with no nausea at baseline

represent a clinically meaningful and acceptable

indication?

Discussion on that? Dr. Green.

DR. GREEN: Well, given the fact that it

is so counter-intuitive, I really believe that both

doctors and patients will erroneously give the

wrong advice.

DR. KIEBURTZ: Dr. Koski.

DR. KOSKI: I think that patients do

intermittently have nausea at various times

throughout the course of their headache. I think,

because they are given a drug, they wouldn't

exclusively use it in terms of the periods when

they didn't have any nausea. So just even

accepting the data as it is, I think there is a

problem with the way a patient would use it.

277

DR. KIEBURTZ: Dr. Porter.

DR. PORTER: I agree. I think that this

drug will be used in patients with or without

nausea if it is marketed.

DR. WELCH: Nausea is such a variable. As

we have heard, this drug will be used in people

with or without nausea without question.

DR. KIEBURTZ: Further discussion? Dr.

Green, how do you vote on the question; is it a

clinically meaningful and acceptable indication.

DR. GREEN: No.

DR. JESTE: I am not a migraine researcher

so I think I should just abstain from answering

that question.

DR. KIEBURTZ: Fair enough.

DR. SMITH: I would do the same.

DR. KIEBURTZ: Abstain.

DR. WELCH: No.

DR. LENAERTS: No.

DR. FAHN: No.

DR. JUNG: No.

DR. GOLDSTEIN: No.

278

DR. SACCO: No.

DR. KOSKI: No.

DR. HUGHES: I abstain as well.

DR. KIEBURTZ: No.

So three abstentions, nine no's as to

whether this is a clinically meaningful and

acceptable indication. Any questions from your end

of the table? Okay.

The fifth question; In a new clinical

study, if the sponsor shows prospectively in a new

clinical study in migraine patients with no nausea

at baseline a significant contribution of

metoclopramide on sustained headache relief of 5 to

10 percent--this harkens back to Question 1--5 to

10 percent of patients, a proportion of

people--again, this is a proportion of the enrolled

subjects who have this response which is sustained

headache pain relief, so that is 5 to 10 percent of

the patients, no benefit at the two-hour pain

research, no contribution of metoclopramide at a

two-hour pain response, no contribution of

metoclopramide on the relapse rate or

279

rescue-medication use in the two- to 24-hour

period--so, again, this is a 5 to 10 percent

benefit on the synthetic measure but not a

significant impact on either of the elements that

make up the synthetic measure--would the

demonstrated benefit outweigh the risks related to

tardive dyskinesia?

This is, again, those effects in a group

that was enrolled with no nausea at baseline; would

this have demonstrated a benefit outweighing the

risks related to tardive dyskinesia.

Discussion of this? Dr. Sacco.

DR. SACCO: Again, I think the only thing

I would add here, we have a significant

contribution. When it says no contribution in the

second and third bullet there, I am assuming we are

saying no significant contribution, but, again,

there may be a trend in those two, that when you

add them up, you have a significant contribution.

DR. KIEBURTZ: I would say that is an

appropriate clarification. Again, I think that we

should clarify that the risk of tardive dyskinesia

280

is whatever in your heart you think that is since

we don't have evidence in that regard--heart, mind.

I am not trying to be flip about it. It is what

your own best estimation is.

DR. GOLDSTEIN: Clarification. It seems

that this question is a compound of all the other

questions. So we have answered, for each of these

components, in one way. I don't see what question

we are answering that is different than what we

have already answered.

DR. KIEBURTZ: Okay. But we are going to

answer it because that is our job.

DR. GOLDSTEIN: Okay.

DR. KATZ: We will just see if you are

being consistent when you answer this one.

DR. FAHN: It is a different question.

DR. KATZ: It is a test.

DR. GOLDSTEIN: I just want to know, are

you trying to get--I want to answer the question

that you are asking.

DR. KATZ: No, no; it puts everything

together. Then, of course, the thing we haven't

281

talked about is the very last thing. But, no; you

are right.

DR. KIEBURTZ: Dr. Fahn.

DR. FAHN: I read this question as

different from the first question, Question No. 1,

is that here we are saying on contribution of

metoclopramide on relapse rate or rescue medication

used in that two- to 24-hour period which wasn't in

the previous question. So this is now saying, in

the long run, you are still going to have to take

rescue medicine just as often, and so forth. So,

to me, that changes the equation quite differently.

It adds another element which the FDA wants us to

answer and I think this now adds that other element

we have to look at and think about.

DR. KIEBURTZ: There was not a

specification before about the no significant

contribution on the relapse rate.

Dr. Porter.

DR. PORTER: My problem with this question

is that it assumes that this no nausea at baseline

is, in fact, something that we really believe the

282

company has shown. In my understanding, this is

pretty much a post hoc analysis. We are not

absolutely sure that they can reproduce this.

I think that, to ask this question assumes

that they are going to look for this small

subpopulation which, in fact, might be a mistake

because, if they are wrong on that, then their drug

might be good for migraine as whole. I think that

we ought to ask the question of ourselves as well,

do we think that this is a subpopulation that

company should consider as a population that they

want to chase because it is a high-risk event. It

cuts down the total number of patients available,

et cetera.

I am not sure that we are there yet.

DR. KATZ: This question just sort of took

as a given that if this was a real reliably

identifiable or appropriate subgroup and they

were--and the first part says, and they were to

prospectively show it again, would all of these

things apply.

I think you have already said that you

283

don't think that the no nausea at baseline is an

appropriate subgroup in which to develop the drug.

I think that is more or less how I interpreted the

last vote, or whichever one it was.

So I think this question does, more or

less, incorporate several of the questions you have

already voted on and it would seem as if the answer

to this question would be obvious. I suppose we

could ask, if this is what you saw in the entire

population; in other words, they did a study both

with and without--you know, the typical population,

as they have already done, and basically saw the

same results that they have already seen twice,

would it be acceptable.

But I think you have already answered that

question as well. I think you have answered that

they have to show something in two hours. Anyway,

I actually think it has already been answered, but

if you want to vote on it, you can.

DR. KIEBURTZ: Dr. Temple, did you want to

say something?

DR. TEMPLE: Not to get into too much of

284

study design things, if the company wanted to do

further studies to document and say and an effect

at two hours and really thought that the no-nausea

population was the right place, they could still do

a study in the mixed population, make the primary

endpoint the effect of the no-nausea population and

get data on the other.

DR. KIEBURTZ: Absolutely agree.

DR. TEMPLE: That is all for later.

DR. WELCH: I guess it depends on what the

5 to 10 percent is. If it is used here as what do

you think would be a clinically persuasive

difference, independent of what the data has been

on nausea, that is a different question that really

addresses the second question about additional

data.

DR. KIEBURTZ: Just staying on the first

question, I think this is a pretty straightforward

one to vote which I think the comments to date, we

know what the answer is going to be. It is

essentially a contribution with a new feature about

specifying a nonsignificant contribution of the

285

relapse period.

Is it my turn to start first? Would the

demonstrated benefit outweigh the risks related to

tardive dyskinesia. I would vote no.

DR. HUGHES: I vote no.

DR. KOSKI: No.

DR. SACCO: No.

DR. GOLDSTEIN: No.

DR. JUNG: No.

DR. FAHN: No.

DR. LENAERTS: No.

DR. WELCH: No.

DR. SMITH: No.

DR. JESTE: No.

DR. GREEN: No.

DR. KIEBURTZ: So unanimous on the no.

If not, I guess our response was presaged

by considering if we said no, if we vote no to that

first question, what additional data, or desired

primary outcome measure, or desired effect on

sustained relief, could provide evidence of safety

and efficacy?

286

What I understand here is we can go back

up to these bullet points--and now is an

opportunity for us, although I am thinking we are

not going to take three hours of opportunity to do

this, to flesh out some of the things we have been

alluding to regarding whether two hours is

important. I think we can comment here even about

the nausea or not.

So I think a question which has already

been posed by Dr. Katz and Dr. Bastings is whether

we would like to see a benefit of the two-hour

response. Dr. Welch also mentioned a pain-free and

a sustained pain-free response. Maybe we can have

some discussion on this point.

Do you want to comment at all, Dr. Welch,

not to pick on you.

DR. WELCH: I think that having said that

we accept that this combination of medication could

be used for certain patients, that it behooves us

to ask for the most rigorous endpoints in any

clinical trial that we can. The most rigorous is a

two-hour pain-free sustained over 24 hours.

287

What patients really want is to be

pain-free as early as possible. That is really

what the patient ideal is. So, for that reason, I

would ask for that particular data given that we do

accept that, with this risk in a limited

population, that this could go ahead.

DR. KIEBURTZ: Dr. Katz and Dr. Porter.

DR. KATZ: I just have a question because,

obviously, we have not adopted as required for the

approval of the typical, if there is a typical,

acute migraine treatment, the pain-free at two

hours. We have had mild or no pain. I just want

to sort of flesh this out a little bit more. Do

you think it should be pain free at two hours

specifically because of the potential risk or do

you think that is sort of a generic requirement?

DR. WELCH: I think it is ideal endpoint

that the community know us for, the headache

community, the IHS regulations or guidelines. That

is the first thing. But I think it is the most

rigorous endpoint that you can get and I think,

because of the risk, that we deserve to see the

288

most rigorous endpoint tested, if that answers your

question.

DR. KATZ: Yes; it does.

DR. KIEBURTZ: Dr. Green.

DR. GREEN: I, like Dr. Welch, was

surprised that endpoint, the primary endpoint, to

the study wasn't two-hour pain something, anyway,

that certainly we would like to see trials going

forward with the primary endpoint being, hopefully,

two-hour pain free because, among other things,

that probably even predicts recurrence rates and

certainly, when we talked about drug overuse, if

someone is pain free, they are very unlikely to

redose.

DR. KIEBURTZ: So the two-hour time point

is an important one as assessment of efficacy.

DR. GREEN: Right.

DR. KIEBURTZ: Of those, the pain-free

response is perhaps, even more informative than the

pain response?

DR. GREEN: Actually, there is another

one. There is migraine-free which is pain-free,

289

photophobia-free, phonophobia-free,

nausea-vomiting free, which is even a tougher one.

DR. KIEBURTZ: You can escalate the

characteristics of what the two-hour endpoint is

but that is an important time point at which to

demonstrate the marginal benefit of--I don't mean

marginal, small, but the additional benefit of

metoclopramide over naproxen.

DR. GREEN: Definitely.

DR. KIEBURTZ: Dr. Fahn.

DR. FAHN: There are three conditions in

this area and one of them that we have been talking

about is the two-hour one. But I would say if any

two of these would be sufficient in my mind that

metoclopramide was superior to naproxen alone; that

is, even if the two-hour wasn't any more superior

pain-free than naproxen alone but they got better

at the end of the day and they didn't have any

relapses and they felt better, that would be--those

two would be fine with me.

So that is what I would look at. One

alone is probably not sufficient for the risk but

290

two out of the three would be sufficient for the

risk in my book.

DR. KIEBURTZ: So either--I am just trying

to think.

DR. FAHN: The relapse rate was one of the

risks. The two-hour time point is another one of

the outcomes. The sustained headache relief.

DR. KIEBURTZ: I don't think you could get

the components without getting the composite, but I

get your point. Two out of the three rather than

just one out of the three which is the circumstance

now.

DR. SMITH: A little off the endpoint

question, but I would like to see that combination

be better than a higher-dose naproxen is one

question because you could increase the dosage and

potentially have the same type of effect for

naproxen alone. The other thing is how would it

work among people that failed naproxen alone would

be another interesting question because, again, you

are talking about a very unique risk and so there

probably would be a drug that would not be a

291

first-line so how does it work in people that fail.

Would you see that same response?

DR. KIEBURTZ: So one is an issue of

comparators, not just to the naproxen dose

intrinsic in MT100 but to a higher dose as a

comparator.

DR. SMITH: Exactly.

DR. KIEBURTZ: The other question is one

of selection criteria in terms of people who

failed.

DR. WELCH: A specific trial in naproxen

failures. It has been done for triptan failures.

Then you might just ease up a little bit on the

rigor of your endpoint.

DR. TEMPLE: To do that trial right, you

have to randomize those people who, by history or

failures, back to this product again and naproxen.

So you are enriching it for naproxen failures.

That design can work if the effect is decent,

large.

DR. PORTER: I think you have to have a

little flexibility for the company. I think our

292

very rigid two hour must be pain-free might be a

hurdle too high. I would urge you to think about a

little flexibility even though you want to set the

standard high for a drug that is not as safe as

some.

DR. WELCH: But the company doesn't want

to be set a lower standard if the standard for the

whole headache specialty becomes that rigorous.

DR. PORTER: How many drugs do you have

that meet that standard?

DR. KIEBURTZ: Let's just not get into a--

DR. PORTER: Okay.

DR. KIEBURTZ: Please. Dr. Goldstein, you

had a comment?

DR. GOLDSTEIN: Yes. The other point,

this compound question that I think still needs to

be considered is is the addition of the rescue

medication--the question is, if you are taking care

of a patient again, that you want them to be able

to take a drug and have relief, be able to get back

with their work and not be having to take

additional drugs. If they require additional

293

drugs, I think that needs to be factored into the

equation.

The second point is, as you design a

trial, what the appropriate comparator group is. I

guess the question that you have from the

regulatory standpoint is the only question that you

are really looking at, is whether the

metoclopramide component adds to the naproxen but

now how the combination would compare with other

approved agents.

DR. KATZ: That's the first question,

absolutely, is the combination policy of both

components making a contribution. The question of

whether or not the actual product, how that

compares to other products, is usually not one that

we consider from a regulatory point of view except

when you are worried that the product might be more

dangerous than everything else that is out there.

There have been extraordinarily rare

occasions when you actually have to show that your

product actually beats something else out there in

order to get approved because it is so dangerous.

294

So, yes; in some sense, we always think about that

in the back of our minds but we almost never

require it because it just doesn't--the

circumstances don't support it.

DR. KIEBURTZ: Dr. Sacco.

DR. SACCO: I just wanted to go back to

the endpoint question again. I am just trying to

understand why the migraine community is basically

saying that outcome for clinical trials regarding

migraine drugs should be changed. From my

understanding and what I have learned today and

what I am reading is that pain research, none,

pain-free, or mild pain has always been the

endpoint.

Is it because we are hearing the risk of

this drug or do you think, in general, for every

migraine drug to come, the new endpoints should be

pain-free. If so, I am just trying to understand

why, because I am okay with none or mild pain, but

I am continuing to hear you want it changed to be

pain-free.

Am I not reading that right?

295

DR. WELCH: The issue is how do you define

mild. What is mild? It is very subjective. So

the most rigorous one would be pain-free. The

outcome from here on in, I believe, by the

community would be to use that.

There are other endpoints that you could

choose, but I would be afraid, addressing the issue

for the company is, if someone else came up with

another drug with a more rigorous endpoint and they

were given a more flexible one, that they wouldn't

compete well with it.

But I think if you certainly do a trial

against naproxen, itself, in naproxen failures, you

could let up on your endpoint, again qualifying the

study.

DR. KIEBURTZ: Dr. Hughes, you had a

comment?

DR. HUGHES: I guess the main comment I

have is I would like to see the other side of the

equation and get a bit more formal information

about the risk of TD in this population. I would

hate to see an efficacy trial with whatever

296

endpoint, or similar endpoints, as has been done

with a 5 percent difference and still have

considerable uncertainty about the risks of TD in

this population.

DR. KIEBURTZ: So addressing the issue of

what additional data, you would like more safety

data.

DR. HUGHES: Formally obtained; yes.

DR. JESTE: I think there is an

opportunity to look at safety in a broad

perspective, not just tardive dyskinesia but also

the various extrapyramidal symptoms, Parkinsonism,

microdystonia, akathisia. This should be done.

There are number of examples in which this has been

done using standardized rating scales for each of

these on a regular basis and continuing that for

six months to a year at the very least because

really one cannot--100 patients keep on the drug

for one month is not the same as 100 patient

months. You really need patients to be treated for

one year before you can say anything about the

incidence of tardive dyskinesia let alone its

297

persistence.

So what I would strongly recommend in

terms of suggested studies would be longer-term

follow-up studies, standardized rating scales for

Parkinsonianism, akathisia, dystonia and tardive

dyskinesia.

DR. KIEBURTZ: Dr. Katz and then Dr.

Lenaerts.

DR. KATZ: Maybe a little discussion on

this point would be worthwhile because these sort

of long-term large safety studies have been done

when you are trying to sort of cap the risk or

say--you might think about, well, I think 0.1

percent risk might be acceptable given this sort of

benefit on an appropriate outcome, so you would

have to figure out how many people that would be.

You may talk about thousands of patients

followed for a year. I am just wondering whether

or not people, again, postulating this sort of an

effect on an appropriate outcome, whether or not

people think that that would be something that

would be necessary, even if they were to show

298

effectiveness at this level.

DR. JESTE: My feeling is that if we did

not suspect tardive dyskinesia with this drug, then

I don't think one would ask for longer-term

studies. But here there are grounds to think that

the drug would be associated with some risk of

tardive dyskinesia which we don't know. It could

be minimal. But, still, there is clearly some

risk.

I think that it behooves on us, then, to

have at least some long-term data because some

patients will, indeed, be using the drug. And the

findings may turn out, may show that, actually, the

risk is very low, in which case it will be helpful.

I do think that long-term studies will be warranted

in a case like this.

DR. KIEBURTZ: Dr. Lenaerts.

DR. LENAERTS: Would the number needed to

treat shed light into this and, in that case, is my

assumption right if we have, say, in 304 study,

rounded out 32 percent versus 28 percent, would it

be a number needed to treat of eight? Would that

299

be correct? Would that be useful to look into

that?

DR. KIEBURTZ: I am not sure that that is

the right number.

DR. LENAERTS: That wouldn't apply?

DR. KIEBURTZ: For number needed to treat.

But, anyway. I am not sure.

DR. LENAERTS: I am just throwing in the

question.

DR. KIEBURTZ: Okay. Dr. Porter, you had

a question?

DR. PORTER: Just a comment and that is if

the risk is, in fact, relatively low in this 0.038

percent category, if you can have any kind of power

at all to measure this, it is going to take

thousands and thousands of patients, as you already

mentioned, unless you are lucky, or unlucky, and

you happen to stumble over one early. But you

could do 2,000, 3,000, 4,000 patients and still not

see it and it could still be there.

DR. KATZ: Right; you could decide up

front what sort of a risk you would be willing to

300

live with. In other words, not greater than 0.1

percent.

DR. PORTER: What kind of power you

wanted.

DR. KATZ: Right; and then you just figure

out how many people. But, yes; we would have to

know what people thought would be an acceptable

risk and then you work backwards and you figure out

how many people you need to follow. But, yes; it

could be, depending on what sort of a risk would be

acceptable, it would be thousands of people. Sure.

I am trying to figure out whether that is

something you think we ought to do and pick a risk

to cap it at, or--

DR. WELCH: That would be extraordinarily

difficult and extraordinarily expensive in the

migraine population. I think the real issue, that

what we are dealing with here, is that if MT100

comes to acceptance and general prescription, that

all of us know that, in a population of migraine

patients, or chronic-headache patients, whatever

that makes up, that they will be taking this drug

301

on a chronic basis, a long-term basis. It is very

difficult to actually reproduce, at any clinical

trial, an interaction of someone who has the

pathogenic factors of a chronic headache, perhaps

with dopamine hypersensitivity and the interaction

of using the drug.

So it is an almost impossible thing to do,

I would think. But the issue is, given that we

know that there will be a subgroup of patients who

will take this medication inappropriately, that we

really must be sure that it is effective for a

subgroup of patients who really need it.

Answering the very eloquent appeal for

pain relief in migraine, which is very real because

migraine patients will, even if some of them know

that they are at major risk for vascular insult,

will take a triptan and take that risk. We do know

that you can use metoclopramide and Naprosyn

separately. In fact, before the triptans, I am old

enough here to be a pre-triptan-era prescriber, it

was not uncommon for me to combine these two

medications in patients that didn't respond to

302

Naprosyn alone.

So the real reason why we must ask for

extreme rigor, I think, with this particular drug

is knowing that there will be a population that may

overuse this drug with the risk of tardive

dyskinesia. But to reproduce that in any clinical

trial, I would think it is just impossible.

DR. KIEBURTZ: I would ask questions in a

slightly different direction because we have not,

in this discussion, gone this way. Just say we

don't get more evidence about safety, that we are

unable to better estimate with precision then

something that we think is between 0.038 and 1.0.

Say we can't figure that out. Is there a magnitude

of benefit that is greater than this number here,

since we voted no to this, that would make people,

in that situation of relative ignorance, change

that to a yes.

So, if there was a 20 percent--the

difference between the naproxen and MT100 group was

20 percent--would that be enough to overwhelm

whatever the concerns are about the safety. I

303

think that is another question that we are being

asked about. Or would, no matter what the

magnitude of the benefit, those safety concerns

would be persistent.

I don't see any clear thoughts on that

one.

DR. FAHN: I would just go back to what we

talked about before that I feel to be absolutely

safe, patients would have to show they don't

respond at all to naproxen alone but now would

respond to the combination drug if there is a

combination drug on the market.

As Michael just said, that before there

was a combination drug, you could still prescribe

metoclopramide if you think it is going to be

added. You try naproxen alone and if it didn't

work, okay, we are going to try the two together.

You will see. Then, if it doesn't work, then you

drop it. If it does work, then there is a certain

risk that you tell the patient there is going to be

and you can take this combination if you are

willing to accept that risk.

304

I think that is the kind of question I

would like to see directed if we were going to look

for--in other words, it has to has some restriction

on a combination drug that you wouldn't have to

have if they were prescribed separately

because--normally, when they are prescribed

separately, it is an extra step.

Too many doctors will say, well, just take

this combination. It gives you both the advantages

of naproxen alone plus the metoclopramide. I think

that makes the risk too dangerous.

DR. KIEBURTZ: So, to paraphrase, the

approach, rather than answering by a bigger

benefit or quantifying the risk, is modifying the

path to which you get to that drug.

DR. FAHN: I think that is extremely

important how you give that path because there

shouldn't be--if you are going to take any risk,

you ought to make sure that risk is worth taking

and, therefore, that you are getting benefit that

you wouldn't have had otherwise.

For that particular person, if they fall

305

within that 5 or 10 percent, 20 percent, or

whatever number you want to give a category, why,

for that person, they are getting a great benefit

and they are willing to take that risk. I think

that is okay.

DR. KIEBURTZ: Are you getting answers?

Is this discussion germane to the--

DR. KATZ: Yes. There are a number of

ideas that we will have to think about but I think

they are all relevant.

DR. KIEBURTZ: Okay. This is not a voting

question because it is, what additional data would

you like. So it is not a yes or no. We can't give

a yes or no here. Would someone like to make

additional comments? I see two. Dr. Jeste.

DR. JESTE: Going back to safety, I am not

suggesting that we need to have huge studies to

find out the incidence of tardive dyskinesia. I

think one can use that normal involuntary movement

scale as outcome for the safety purposes as major.

So what you are looking at is really percent

increase in the M score. Say, the mean score,

306

let's say, goes up from 0.5 to 1.5. Those patients

will still not meet the criteria for tardive

dyskinesia because you need a minimum score of 2,

minimum at least 1. So that can be done in a few

hundred patients. So it is really not the number

of patients that is critical. It is the length of

the study that will be important. That can be done

in as many patients as are needed for just looking

at outcome.

DR. KIEBURTZ: Dr. Goldstein.

DR. GOLDSTEIN: Again, just from a

clinical standpoint, I think the real question is

whether the patient taking X medicine as opposed to

Y is going to get on with their life later, be able

to get back to their life sooner rather than later.

If you just improve pain but they are

still debilitated because of the concomitant

symptoms, you haven't really done much. So, as a

composite measure, I think it would be good to have

some measure of migraine-related disability, and I

know such scales exist, to be considered as one of

the endpoints.

307

That is what we are really trying to do

here. The other thing to consider in terms of some

of these trade-off issues--I said it before, but,

in terms of additional data, that data can be

obtained so we are not sort of guessing these

things. We can get finite data to support what

risk people with migraine would be willing to take

of having these sequelae to be pain-free or without

disability at whatever given high proportion.

So, again, I think that those data would

be very helpful to inform both the company's

decisions as to how to go and for the FDA as well.

DR. KIEBURTZ: To summarize Question 5, on

the first part of it, we voted no uniformly on the

way it was categorized and then discussed ways

where data might be helpful in addressing the

question. One is more safety data. Two is the

two-hour endpoint seems to be important as to

whether that needs to be something beyond the

traditional pain response is something that is

debated and, also, we talked about ways in which

subjects or patients might access the medication

308

through specific failure of other interventions

prior to exposure to this intervention.

I am going to briefly summarize our

discussion. Unfortunately, your chair has failed

my two-hour response to my migraine intervention so

my thinking is a little cloudy. So you will have

to bear with me. In addition to Dr. McCormick,

there are other people who have trouble responding

to their migraine medications.

We concluded that the current estimate of

tardive dyskinesia following exposure to

metoclopramide, particularly in this setting, is

not a reasonable estimate. We talked some more

about how we might get some better estimates of

that.

Both in the initial question and the

second question, that the amount of benefit

demonstrated so far, without saying whether that is

significant or not, is not sufficient given the

perceived risk in the absence of concrete data, the

perceived risk given the absence of benefit at

two-hour endpoints.

309

We don't think that there is enough

evidence that there is no risk of tardive

dyskinesia with chronic intermittent administration

nor can we identify a dose that would be below the

risk, a number of doses that would confer no risk.

We think it is likely that, no matter how the drug

is labeled and approved, that people will take it

more than whatever the recommended dosage is if

there is a limitation on the number of dosages, and

did not have support for the idea of individuals

with nausea at baseline being an identifiable group

of a clinically meaningful and acceptable

indication for the treatment of migraine.

I think that summarizes our discussion.

DR. PORTER: Could I just add one comment.

I think there is an uncertainty about whether or

not the nausea or no-nausea populations have been

demonstrated here. The company should not mislead

itself by a subgroup analysis post hoc that

necessarily takes them down the wrong path.

DR. KIEBURTZ: I think, then, unless there

are some comments from Dr. Katz.

310

DR. KATZ: No. I just want to thank

everybody. I think it is particularly difficult to

discuss these matters and come up with decisions

and advice based on so little data, at least on

parts of this. So I really appreciate it.

Also, I just learned that Anuja, this is

her last meeting with the committee. So I want to

thank her publicly for all her work. And good luck

in the future. We are sorry to see you go.

DR. KIEBURTZ: You keep your secrets well,

like everyone else here at this agency. Thanks to

the committee members for coming, for staying on

point, to the sponsor for their presentations and

their responsiveness to our questions.

We will adjourn the meeting now.

(Whereupon, at 2:40 p.m., the meeting was

adjourned.)

- - -