1

DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

CENTER FOR DRUG EVALUATION AND RESEARCH

ONCOLOGIC DRUGS ADVISORY COMMITTEE

Wednesday, December 1, 2004

8:00 a.m.

Holiday Inn, Silver Spring

8777 Georgia Avenue

Silver Spring, Maryland

PARTICIPANTS

Victor Santana, M.D., Acting Chair (a.m. session)

Silvana Martino, D.O., Acting Chair (p.m. session)

Johanna Clifford, M.S., RN, Executive Secretary

MEMBERS

Otis W. Brawley, M.D.

Ronald M. Bukowski, M.D.

Bruce D. Cheson, M.D.

Stephen L. George, Ph.D.

Pamela J. Haylock, RN

Maha H.A. Hussain, M.D.

Silvana Martino, D.O.

Joanne E. Mortimer, M.D.

Michael C. Perry, M.D.

Gregory H. Reaman, M.D.

Maria Rodriguez, M.D.

INDUSTRY REPRESENTATIVE (NON-VOTING)

Samuel Maldonado, M.D.

CONSULTANTS (VOTING)

For clofarabine

Joanne Kurtzberg, M.D.

David Poplack, M.D.

Victor Santana, M.D.

Alan Wayne, M.D.

For Marqibo

Michael Bishop, M.D.

Wyndam Wilson, M.D., Ph.D.

PATIENT REPRESENTATIVES (VOTING)

Ruth Hoffman - for clofarabine

Susan Krivacic - for Marqibo

PARTICIPANTS (Continued)

FDA

Martin Cohen, M.D.

Ramzi Dagher, M.D.

Ann Farrell, M.D.

Maitreyee Hazarika, M.D.

Steven Hershfeld, M.D.

Richard Pazdur, M.D.

Robert Temple, M.D.

Grant Williams, M.D.

C O N T E N T S

Page

Morning Session

Call to Order:

Victor Santana, M.D. 6

Introductions 7

Conflict of Interest Statement

Johanna Clifford, M.S., RN 9

Opening Remarks

Richard Pazdur, M.D. 11

NDA-21-673 Clolar (clofarabine)

ILEX Products, Inc.

Sponsor Presentation

Introduction

Steve Weitman, M.D., Ph.D. 19

Pediatric Leukemia

Bob Arceci, M.D., Ph.D. 21

Clofarabine Pivotal Studies

Steve Weitman, M.D., Ph.D. 26

Clinician's Perspective

Steve Sallan, M.D. 49

Clofarabine Development Plan

Steve Weitman, M.D., Ph.D. 54

FDA Presentation

Martin Cohen, M.D. 60

Questions from the Committee 83

Open Public Hearing 139

Committee Discussion 163

C O N T E N T S (Continued)

Afternoon Session

Call to Order

Silvana Martino, D.O. 221

Introductions 221

Conflict of Interest Statement

Johanna Clifford, M.S., RN 223

Opening Remarks

Richard Pazdur, M.D. 226

NDA 21-600 Marqibo (vincristine sulfate

liposome injection)

Inex Pharmaceuticals

Sponsor Presentation

Introduction

Alexandra Mancini, MSc. 234

Overview

Fernando Cabanillas, M.D. 236

Pharmacology

Tom Madden, Ph.D. 245

Efficacy and Safety

Alexandra Mancini, MSc. 251

Clinical Benefit

Fernando Cabanillas, M.D. 284

FDA Presentation

Maitreyee Hazarika, M.D. 292

Questions from the Committee 309

Open Public Hearing 347

Committee Discussion 358

P R O C E E D I N G S

Call to Order

DR. SANTANA: This is a meeting of the

Oncology Drugs Advisory Committee for the FDA with

additional pediatric oncology representation this

morning because we are going to discuss a new drug

application for the proposed drug clofarabine.

With that brief introduction, I want to

make two comments. First of all, I really want to

keep the agenda on schedule. We will allow all the

presentations to occur both from the sponsor and

from the FDA, and then we will proceed with a

period of discussion and comments, have our break,

and then come back and deal with questions and the

advice to the Agency.

So, I really want to keep on schedule as

much as possible, because there is another schedule

this afternoon that the Committee has to abide to,

and I want to make sure that they get their

opportunity this afternoon, too.

Secondly, I want to go ahead and do a very

brief introduction of all the members of the

Committee, since we have additional pediatric

oncology representation today. So, if we could go

around the table starting with Dr. Poplack on the

left corner, introduce by name and affiliation.

Introductions

DR. POPLACK: David Poplack from Baylor

College of Medicine.

DR. KURTZBERG: Joanne Kurtzberg from Duke

University Medical Center.

DR. WAYNE: Alan Wayne from the National

Cancer Institute, Pediatric Oncology Branch.

MS. HOFFMAN: Ruth Hoffman, Patient Rep.,

Candlelighters Childhood Cancer Foundation.

DR. MALDONADO: Samuel Maldonado from

Johnson & Johnson, Industry Representative to this

Advisory Committee.

DR. GEORGE: Stephen George, Duke

University.

MS. HAYLOCK: Pamela Haylock, Oncology

Nurse, Consumer Representative.

DR. HUSSAIN: Maha Hussain, University of

Michigan.

DR. PERRY: Michael Perry, University of

Missouri, Ellis Fischel Cancer Center.

DR. MORTIMER: Joanne Mortimer, University

of California, San Diego.

DR. SANTANA: Victor Santana, St. Jude

Children's Research Hospital in Memphis.

MS. CLIFFORD: Johanna Clifford, Executive

Secretary to the ODAC.

DR. MARTINO: Silvana Martino from the

John Wayne Cancer Institute.

DR. BRAWLEY: Otis Brawley from Emory

University.

DR. CHESON: Bruce Cheson, Georgetown

University, Lombardi Comprehensive Cancer Center.

DR. BUKOWSKI: Ronald Bukowski, Cleveland

Clinic Foundation, Cleveland, Ohio.

DR. COHEN: Martin Cohen, FDA.

DR. DAGHER: Ramzi Dagher, Pediatric

Oncologist and team leader in the Division of

Oncology Drug Products, FDA.

DR. WILLIAMS: Grant Williams, FDA.

DR. PAZDUR: Richard Pazdur, FDA.

DR. SANTANA: With that, I will hand it

over to Johanna. She needs to read the Conflict of

Interest Statement.

Conflict of Interest Statement

MS. CLIFFORD: The following announcement

addresses the issue of conflict of interest and is

made a part of the record to preclude even the

appearance of such at this meeting.

Based on the submitted agenda and the

financial interests reported by the committee

participants, it has been determined that all

interests in firms regulated by the Center for Drug

Evaluation and Research present no potential for an

appearance of conflict of interest at this meeting

with the following exceptions:

Dr. Victor Santana has been granted a

waiver under 21 USC 355(n) for owning stock in a

competitor. The stock is valued from $5,001 to

$25,000. A waiver under 18 USC 208(b)(3) is not

required because of the de minimis exception

2640(b)(2) applies.

Dr. Stephen George has been granted a

waiver under 18 USC 208(b)(3) for serving as a

consultant to the competitor on an unrelated

matter. He receives less than $10,000 per year.

Ms. Ruth Hoffman has been granted a waiver

under 18 USC 208(b)(3) because her husband serves

as a consultant to two competitors on unrelated

matters. He receives less than $10,001 per year.

A copy of the 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.

We would also like to note that Dr. Samuel

Maldonado has been invited to participate as the

Non-Voting Industry Representative acting on behalf

of all regulated industry. Dr. Maldonado is

employed by Johnson & Johnson Pharmaceutical

Research and Development.

In the event that the discussions involve

any other products or firms not already on the

agenda for which am FDA participant has a financial

interest, the participants are aware of the need to

exclude themselves from such involvement and their

exclusion will be noted for the record.

With respect to all other participants, we

ask in the interest of fairness that they address

any current or previous financial involvement with

any firm whose products they may wish to comment

upon.

Thank you.

DR. SANTANA: Thank you, Johanna. Any

other Committee member want to disclose any

conflict at this moment?

[No response.]

DR. SANTANA: Thank you. We will go ahead

and have Dr. Pazdur give us the introduction from

the FDA perspective.

Opening Remarks

DR. PAZDUR: Good morning. The sponsor of

the application in this morning's session requests

marketing approval of clofarabine for the proposed

indication of the treatment of pediatric patients

with refractory or relapsed acute leukemia.

The presentations will focus on one

single-arm trial conducted in 35 patients with

relapsed, refractory AML and a second single-arm

trial in 49 patients with relapsed, refractory ALL.

A Phase I study was also conducted in 25 patients

with relapsed or refractory acute leukemia.

For the treatment of acute leukemia, the

Division has recommended the use of improved

survival or a complete response rate of a

sufficient magnitude and duration to ensure the

demonstration of clinical benefit.

Complete response rates of sufficient

duration are considered clinical benefit, because

they are usually associated with reductions in

infection rates and blood product transfusions, and

may be considered established surrogates for

survival in this disease.

Response duration is usually measured from

the time of initial response until documented tumor

progression. One problem encountered in this

application is the introduction of bone marrow

transplantation in patients who have received

clofarabine, but have not had documented disease

progression.

The addition of transplantation prior to

the documentation of disease progression confounds

any interpretation of clofarabine's response

duration. No consistent prospective criteria were

used to determine patient selection for

transplantation. Some patients went to

transplantation with only a clofarabine partial

response or even without a response in these

single-arm trials.

A clofarabine induction response may

simply indicate a chemosensitive-leukemia and the

patient might do as well with transplantation

without clofarabine induction.

In patients who did not go on to

transplantation and, hence, response duration, can

be measured. These response durations were

generally short and many of these responses were of

uncertain duration because they were not confirmed

by a repeat marrow aspirate.

These results are presented in the

preamble to your ODAC questions. In 35 patients

with AML, there were no complete responses, only 1

complete response without complete platelet

recovery, a so-called CRp, and 8 partial responses.

Of the 9 responding patients, 2 patients

did not go on transplantation prior to disease

progression. These patients had PR's. Their

response duration was short, 12 and 34 days.

Of the 49 patients with ALL, there were 6

CR's, 4 CRp's, and 5 PR's. In this population,

response duration was not confounded by

transplantation in only 9 patients. The 5 patients

with CR's had response durations of 43, 50, 82,

93+, and 160+ days. Only 3 of these 5 CR's had a

confirmed response. As in AML, PR's had a very

short duration of only 7, 16, and 21 days.

As stated previously, the Agency has

recommended a substantial complete response rate

and duration at endpoints for regular approval in

hematological malignancies denoting clinical

benefit.

In 1992, the accelerated approval

regulations allowed the use of additional endpoints

for the approval of drugs that are intended to

treat serious and life-threatening disease and that

either demonstrate improvement over available

therapy or provide therapy where none exists.

The FDA may grant accelerated approval

based on the effect of a surrogate endpoint that is

"reasonably likely" to predict clinical benefit.

A drug is approved under accelerated

approval on the condition that the manufacturer

conduct studies to verify and describe clinical

benefit. The regulations stated an expectation

that post-marketing studies would usually be

underway prior to accelerated approval, however,

this is not a requirement.

At a March 2003 ODAC meeting, the ODAC

reinforced the Agency's view that these

confirmatory trials should be ongoing at the time

of accelerated approval is granted. Approval with

subsequent commercial availability of the drug may

interfere with subsequent enrollment to the

confirmatory trial.

We, the Division, are asking your opinion

regarding the accelerated approval of clofarabine

based on the data presented. The ALL indication

should be considered separately from the AML

indication. There exists uncertainty regarding the

response duration because of the lack of subsequent

bone marrow biopsies to confirm a response and the

introduction of transplantation prior to the

documentation of disease progression.

Where durations can be measured, the

Division considers, with some exceptions, these

response durations to be limited. We have asked

the sponsor to present ongoing planned trials in

both pediatric and adult leukemia. Presently, we

have not identified any study that has been

designated as a confirmatory trial for the

subsequent demonstration of clofarabine clinical

benefit.

For our Division, this is the first time

we are considering a pediatric application for

accelerated approval. Pediatric drug development

in the treatment of pediatric malignancies differs

from adult drug development, therefore, we have

supplemented this ODAC membership with voting

members from the pediatric oncology community.

Pediatric drug development has been

blessed by an exceptionally high rate of patient

enrollment compared to enrollment in adult studies.

Great strides have been made in curing and

prolonging the survival of children in the past

decades. Most children, especially with the

diseases under consideration this morning, are

treated on protocols at referral centers rather

than in the community.

Your discussions should consider the

ramifications of accelerated approval for the

pediatric development of clofarabine. Approval of

the drug for a pediatric indication should not be

at a lesser standard than that expected for an

adult indication.

Approval decisions should be based on a

risk-benefit determination. A reasonable question

is whether the necessary information regarding this

risk-benefit relationship can be derived from a

single-arm study where the primary endpoint is

confounded by the introduction of a subsequent

therapy, specifically bone marrow transplantation.

Your decision regarding the approval

status of this drug should be based on the above

scientific decision, not simply a desire to provide

drug access to patients. Access to a yet approved

drug, especially with a limited patient population

encountered in these applications, can be

accomplished through additional registration trials

and expanded access programs.

We are interested in your discussion on

the impact of this drug's accelerated approval at

this time, and the timely completion of any

confirmatory trials in pediatric oncology. An

appropriate question is whether the drug approval

at this time, especially since the designated

confirmatory trial is not underway, may interfere

with the conduct and completion of confirmatory

trials.

Discussions may focus on whether approval

of this drug, with its response rate and

uncertainties regarding response duration, is

appropriate, or whether additional data should be

available before a definitive approval decision is

made.

Thank you.

DR. SANTANA: Thank you, Dr. Pazdur.

I want to note for the record that Dr.

Temple has joined the table. If you would briefly

introduce yourself, Dr. Temple.

DR. TEMPLE: Good morning, everyone, sorry

I am late. I am Dr. Robert Temple. I am the

office director of the office in which oncology

lives.

DR. SANTANA: Thanks. We will go ahead

with the sponsor presentations, and I would ask the

sponsor to go ahead and follow the schedule, and

after each speaker, the next speaker can get up and

follow with their presentation.

Thank you.

Dr. Weitman.

NDA 21-673, Clolar (clofarabine)

ILEX Products, Inc.

Sponsor Presentation

Introduction

DR. WEITMAN: Good morning. I am Dr.

Steve Weitman, Chief Medical Officer for ILEX

Oncology. I am also a pediatric oncologist.

I would like to start by thanking the ODAC

panel members, as well as the FDA today, for the

opportunity to be here today to present the results

of our clofarabine studies in pediatric patients

with acute leukemia.

[Slide.]

I would also like to start by just

recognizing some of the pediatric leukemia experts

that are here with us today, as well as the

investigators in a number of our trials that are

also here, that may help present and also answer

any questions that may come up during later parts

of this discussion today.

[Slide.]

Following my brief introduction, I am

going to introduce Dr. Robert Arceci, who is here

today to talk about pediatric leukemia and the need

for new treatments. I will then return to the

podium and talk about the results from our two

pivotal studies in pediatric patients with acute

leukemias.

Then, Dr. Steve Sallan will come up to the

podium. Again, he was not an investigator on any of

our studies, but will provide his perspective as a

pediatric oncologist and caregiver regarding what

clofarabine means to him.

Then, I will return, as Dr. Pazdur alluded

to, to talk about our plans moving forward with

clofarabine both in the pediatric population, as

well as in adults with acute leukemias.

With that, let me introduce Dr. Robert

Arceci.

Pediatric Leukemia: Need for New Treatment Options

DR. ARCECI: Thank you, Steve, and thank

you, Dr. Santana and the Committee for allowing us

to present these data. I am a pediatric oncologist

and was an investigator on the clofarabine trials.

[Slide.]

What I want to try to do is to give you an

overview of the situation that we deal with in

pediatric oncology particularly with regard to

pediatric acute leukemias.

Currently, treatments for newly diagnosed

patients with acute lymphoblastic and acute

myelogenous leukemia all use very aggressive

combination chemotherapies, quite intensive and

have been become increasingly intensive over the

past 10 years.

The overall survival for pediatric

patients with ALL and AML has improved

significantly, but over the past 5 to 10 years, has

started to approach plateaus. This is in spite of

a maximum intensification that we are using, so new

drugs clearly and new approaches are needed in this

group. Despite that intensification, 20 percent of

patients with acute lymphoblastic and possibly a

little more than 50 percent of patients with

myelogenous leukemia will have disease recurrence.

[Slide.]

Those numbers lead to the following

conclusion, that that is, relapsed acute leukemia

represents the third most common cancer that we

deal with in pediatric oncology, so although an

orphan disease, this is a major problem for those

of us treating patients in pediatric oncology.

[Slide.]

Now, the challenge to approach that group

is immense. We know that at relapse and even at

diagnosis, these relapsed leukemias represent a

very heterogeneous group of diseases. At the time

of relapse, they are usually multi-drug resistant,

and that resistance crosses most of our

conventional drugs, so it is a major problem and

multifactorial.

Dose intensification with combination

chemotherapies, as I mentioned to you, for newly

diagnosed patients often leads to much more heavily

treated patients than we did 10 or 15 or 20 years

ago, so this group of patients have highly

resistant disease, and they often have

co-morbidities in terms of organ toxicities when

they relapse.

In many respects, transplantation is not

only the best, it is possibly the only curative

therapy we have for these children. So, getting

them to transplant in a state of what we would

expect to be minimal residual disease is a vital

and important component of how pediatric

oncologists approach these children.

[Slide.]

This just shows you some data. Although

from 1997, the fact of the matter is the data have

not changed significantly since these data were

published. These are patients and their outcomes

who had relapsed and refractory pediatric leukemia

treated with chemotherapy only, and this is

primarily data from pediatric oncology group

studies, Phase II studies. In 2004, these results

are really no different.

Transplantation, I should note can improve

the outcome of those patients somewhat.

[Slide.]

Now, in spite of what I have told you, few

agents have been approved for pediatric leukemia.

The most commonly used agents that have been

approved, have been approved many years ago, 1950s

onward.

The development of new agents that are

well tolerated and more effective becomes an

increasingly important part of what we are trying

to accomplish in our field in pediatric oncology.

[Slide.]

So, to conclude, and I hope this gives you

a bit of a feeling for what we are dealing with, is

that relapsed leukemia is the third most common

cancer we deal with. Successful treatment for

relapsed and refractory pediatric leukemias remains

an enormous challenge for us.

These are children who often don't have

very much time because of the progression and rapid

rate of growth of their leukemias. Patients with

these multi-drug resistant leukemias also have

these co-morbidities because of prior

intensification therapies, therefore, we conclude

at least that we need well tolerated, new, and

effective agents to induce minimal residual disease

states, to get responses, so we can then move

towards a more curative approach.

Thank you very much.

Clofarabine Pivotal Studies

DR. WEITMAN: Thank you, Dr. Arceci.

What I would like to do now is tell a

little bit of a story, really a story of how we got

to this podium today to present the results of our

pediatric studies in acute leukemia.

I think, as most pharmaceutical companies

do, we started our adult Phase I study back in

1999, and that was followed shortly thereafter,

approximately 18 months later, by our Pediatric

Phase I study.

For those of you who are familiar with

pediatric oncology, I think most of us would

recognize it is becoming a more and more

traditional pathway in which new agents are

introduced into the pediatric environment.

The results that we saw of this study,

though, were very striking as regards to the Phase

I study, and particularly in a very heavily

pretreated population of patients and with an

acceptable profile.

The results of these studies, though, were

extremely impressive and really propelled the

studies and the program in pediatric oncology

forward at a much faster rate than otherwise would

be expected.

In addition, because of the fact that

there was a lack of other opportunities in other

protocols for these patients, because many of these

patients would not qualify for other studies, we

saw an increase in demand for access to this drug

by pediatric oncologists.

Because of this, we also opened up an

expanded access program that was focused almost

exclusively on the pediatric population.

[Slide.]

This next slide shows somewhat of a

timeline of our program to date, and if I can just

walk you through this. As you can see on the top

of the slide, the adult studies are in blue. On

the bottom of the timeline, the yellow represents

the pediatric studies.

As you can see here, again, the Phase I

study in adults was started back in 1999. This was

followed 18 months later by the pediatric Phase I

program. Subsequent to this, two adults Phase II

programs were started with a predominant focus on

patients with AML.

We then started our two pediatric

programs, one in AML and one in ALL in 2002.

Again, following that was the Phase I/Phase II

combination studies with clofarabine and ara-C in

adult patients, again predominantly with AML.

You can see, moving forward, our plan is

to do a Phase II study with clofarabine and ara-C

in patients with AML, but there has also been

interest in the same study in patients with ALL

through the Children's Oncology Group.

But I think this slide shows you again the

stepwise approach that we have taken where the

adult studies preceded the pediatric program, but

it was really the results of the Phase I study in

the highly refractory-resistant population of

patients that propelled our program in pediatrics

forward at a much faster rate than what otherwise

may have happened.

What I would like to do now is to talk a

little bit about that Phase I program because again

I think it was pivotal in our decision to move

forward with this program.

[Slide.]

As you can see here, and as Dr. Pazdur

noted, 25 patients were enrolled in this program.

Dose levels ranging from 11 to 70 mg/m

2 with MTD or

recommended Phase II dose being 52 mg/m

2 for 5

days. The dose-limiting toxicities of this study

were increases in bilirubin, as well as skin rash.

Now, most Phase I studies are not designed

to really characterize the response rate in this

population of patients. However, again, I have

conducted a number of studies in both solid tumors

and pediatric patients with leukemia, and what

really struck us was an over, really impressive

response that we are seeing in this patient

population. There were 5 CR's noted in this

patient population.

I should note that 4 of these CRs lasted

more than 50 weeks. One of these CRs was in a

patient with AML that had failed fludarabine and

ara-C before coming onto the clofarabine study.

Following this, this patient refused to go

on to transplant and stayed on clofarabine for 8

cycles, and then stayed in remission for 43 weeks

after clofarabine treatment. Again, that was a

patient with AML.

As you can see here, we also had 3

patients who had a PR, and 7 of these 25 patients

went on to a bone marrow transplant or stem cell

transplant.

[Slide.]

Because of these results, we again moved

forward with 2 studies, one in acute myelogenous

leukemia, the other one in ALL. The primary

endpoint for both of these studies was overall

response rate defined as complete remission, as

well as complete remission without full platelet

recovery.

A Fleming 2 stage design was used in these

studies. However, we began to see very early on in

this program that patients much more heavily

pretreated than we had ever anticipated began

coming on to study. In fact, most of the patients

that were being enrolled had undergone a bone

marrow transplant, in come cases 2 transplants,

before coming on study.

In addition, as you can see here, some of

the patients had up to 6 prior regimens before

being treated with clofarabine. In addition, one

other confounding factor, that has already been

alluded to, many of these patients were being taken

off study very quickly to move to transplant, and I

will talk about that a little bit later on.

Again, a lot of these patients were PR in

which they had 0 to 5 percent blasts, but did not

have full ANC recovery, so because of that, they

were still considered a PR, but were going to

transplant.

In addition, a lot of these patients did

not go on to second or third cycles of treatment

because they had a donor that was identified and

wanted to move to transplant.

Following discussion with our external

thought leaders, and well as the FDA, we decided to

expand these studies to get a better determination

of the response rate in this highly refractory

patient population.

What I would like to do now is just walk

you through a couple slides that I think again

highlight how heavily pretreated this patient

population is.

[Slide.]

This slide shows the number of unique

agents that these patients were exposed to prior to

coming onto the clofarabine study. As you can see

here on the left-hand axis, this is the number of

unique agents, and across the bottom is the patient

number.

This first slide is for patients with ALL.

This first patient had 12 unique agents before

coming on to clofarabine, and then had also a bone

marrow transplant.

[Slide.]

The next patient, patient 7, I just want

to highlight had 9 unique agents, a bone marrow

transplant that included total body irradiation.

As you can see here, for the rest of the patients,

again, this was an extremely heavily pretreated

patient population. In most cases, the patients

received anywhere from 9 to 12 unique agents.

Some of the patients received up to 16

unique agents before coming on the study. Many of

them also had bone marrow transplant and many of

them have also had total body irradiation.

[Slide.]

This patient here, that had 16 unique

prior agents, also had two prior transplants, as

well as total body irradiation as part of the

conditioning regimen for the transplant. This

patient went on to achieve a CRp after treatment

with clofarabine.

[Slide.]

This next slide again is a very similar

presentation of the number of unique agents that

these patients had been exposed to prior to coming

onto the clofarabine study.

As you can see here, again, your first

impression, or at least my first impression, was

the fact that these were patients who were less

heavily pretreated than those with ALL. In fact,

they were probably more heavily pretreated than the

patients with ALL.

Both sets of diseases, ALL and AML, the

patients had been exposed to a median of 3 prior

regimens before coming on study. It appears that

there are a fewer number of unique agents that

these patients were exposed to, where, in fact, I

think that represents two important findings.

Number one, that there are fewer options

for patients with AML particularly at the time of

relapse. In addition, many of these patients

received the same agent over and over and over

again during their treatment courses, so again,

they appear to be less heavily pretreated, but, in

fact, I think again they were just as heavily

pretreated and likely just did not have many

options as far as new agents to be used.

[Slide.]

What I would like to highlight is one

patient in particular, just because I think this

again shows how heavily pretreated this patient

population was, and it also points out a number of

other key findings.

This is a 4-year-old with AML that was, as

you can see here, treated with multiple regimens

before coming on to this study. This patient I

think was most notably exposed to a number of

nucleoside analogues including cytarabine,

thioguanine, gemcitabine, fludarabine before being

treated with clofarabine.

I think again it is important to note that

this patient received cytarabine 4 times in

different treatment regimens during the course of

his disease. Regimen 3, asparaginase/cytarabine,

this patient was refractory to that treatment, and

then went on to receive a multi-agent regimen,

again was refractory to that treatment. Then, went

on to receive cytarabine and idarubicin.

This patient did go into remission in

October of 2001, and again as it very commonly seen

when there is a donor available, this patient moved

very quickly to transplant. So, in December,

approximately two months later, this patient went

into transplant, being conditioned with TBI,

thiotepa, and fludarabine, and then underwent a

stem cell transplant.

This patient stayed in remission until

July of 2002, approximately 30 weeks from the very

start of cytarabine/idarubicin until relapsing. In

July, this patient underwent clofarabine treatment,

received 1 cycle, and then went into complete

remission.

This patient went on to receive 5

additional cycles of clofarabine before undergoing

a bone marrow transplant, and as data cutoff, this

patient is alive with no evidence of disease.

The period of remission from the start of

clofarabine to the date of cutoff was over 70

weeks, nearly twice as long as the period of

remission from the start of the cytarabine regimen

until relapse.

I think this slide also shows again the

fact that this patient had been exposed to a number

of nucleoside analogues in the past, as well as

cytarabine multiple times before coming onto the

clofarabine study, also had been proven to be

refractory to a number of the best agents that we

have available right now in pediatric AML.

[Slide.]

What I would like to do now is focus a

little bit on the efficacy results from our two

studies.

[Slide.]

Before I talk about the two studies, I do

want to highlight the database that was used in our

analysis. As was noted earlier by Dr. Pazdur, we

did analyze patients with ALL separately from those

with AML.

However, for safety analysis, we did

combine these two patient populations together, and

we also included the patients from the Phase I

study. So, that is why there is a little bit

difference in the patient numbers for the database

size for these two analyses.

[Slide.]

Key endpoints of the study were overall

response rate, again CR and CRp. It should be

noted that an Independent Review Panel reviewed all

patients enrolled in the study, and their

determination of response was used in all

determinations of efficacy moving forward.

In addition, at the time that they

determined that a response occurred, that was then

used to determine duration of remission.

We also looked at post-transplants,

survival, and obviously, the safety profile.

[Slide.]

Now, with regards to patients with ALL.

[Slide.]

Again, I will focus initially on the

efficacy results of this study.

Forty-nine patients were enrolled in this

study with a median age in years of 12, ranging

from 1 to 20. These patients received a median of

3 prior treatment regimens before coming onto the

study with a range of from 2 to 6.

Approximately, two-thirds of the patients

were refractory to their last therapy before coming

on study, and most of these last therapies again

were multi-agent regimens. Whenever there was an

evidence, for the most part, of palliative therapy,

such as oral etoposide, we would go back to the

previous regimen before that.

As you can see here, approximately a third

of the patients had bone marrow transplant with ALL

before coming onto this study.

[Slide.]

The Independent Review Panel found that 20

percent of the patients enrolled in the study had

either a complete response or complete response

without full platelet recovery with confidence

intervals ranging from 10 to 34 percent.

Patients with at least a PR had a 31

percent response rate, again with confidence

intervals ranging from 18 up to 45 percent.

The patients that were deemed refractory

to their most recent prior therapy had a response

rate of 17 percent. Again, this was for CR and CRp.

[Slide.]

This next slide shows a duration of

remission for patients enrolled in this study. As

you can see, for patients with at least a PR, the

duration of remission was approximately 10 weeks.

For patients with a CR and a CRp, the duration of

remission was 20.2 weeks.

[Slide.]

Of these patients, again, one of the

critical next steps in a curative approach is to

try to take these patients to transplant. As you

can see here, 14 percent of these patients went on

to transplant, 2 with CR, 2 with CRp, and 1 with

PR.

The median time to transplant was 32 days,

but as you can see here, again, they moved very

quickly to transplant in many cases, where in as

little as 16 days, the patient would go to

transplant, and this was after following remission

induction.

The median number of cycles before

transplant was 2, and 5 of 7 patients are alive

post-transplant.

[Slide.]

This slide represents the overall survival

of patients enrolled in the ALL study. The bottom

green line represents overall survival for patients

enrolled in the study of 11.7 weeks. If you look

at the top line, this is for patients with a CR or

a CRp where the overall survival was a little bit

over 1 year.

[Slide.]

Now, to move on to our studies in patients

with AML.

[Slide.]

Thirty-five patients were included in this

study. The median age was 12 with a range of from 2

to 22. Again, as I noted earlier, the number of

median regimens was 3 prior to coming onto this

study. Sixty-three percent of these patients were

refractory to prior therapy, and over half of these

patients had some form of bone marrow transplant

before coming onto this study.

[Slide.]

The Independent Review Panel found that 1

patient had a CRp and that there were 26 percent of

patients who had at least a PR. Again, as has been

noted, PR's are not typically viewed in hematologic

malignancies as a benefit, however, this PR allowed

many of these patients to move on to transplant

that may not have otherwise had that opportunity.

Four of the patients that were refractory

to prior therapy did have a PR. It should be noted

that 3 of these patients went on to transplant and

2 of these 4 patients are still in remission today

following clofarabine and transplant. One of these

patients had failed cladribine and idarubicin

before coming onto the study.

Again, this was a patient with AML,

received 1 cycle of clofarabine and went into

remission, and that he had zero percent blasts in

this bone marrow, his ANC was increasing, however,

it did not reach the threshold of 1,000, which was

needed for a CR. Before reaching that threshold,

the patient went on to transplant, and is now in

remission 58 weeks since undergoing clofarabine and

bone marrow transplant.

[Slide.]

This slide shows the duration of remission

for all patients that had either CR, CRp, or a PR

in the AML study. The median duration was 16.2

weeks.

[Slide.]

Again, one of the key endpoints for any

patient at this stage is to get to transplant,

particularly in patients with AML where a PR may be

more meaningful than certainly in patients with

ALL.

As you can see here, over a third of the

patients went on to transplant, 1 patient with a

CRp, 6 patients with a PR. As has been noted in

your briefing document and some of the questions

that you received, 2 of the patients had treatment

failure, still went on to transplant.

I think these 2 patients are worth noting

and explaining in a little bit more detail.

The first patient had 98 percent blasts at

study entry. After a cycle of clofarabine, this

patient dropped down to 2 percent blasts in the

bone marrow. However, when the IRP looked at that

patient's report and smear, they noted that there

were still some myelomonocytic cells present, so

they considered that patient a treatment failure.

However, this patient still went on to transplant

because the treating investigator felt that the 98

percent blast at entry down to 2 percent was

substantial cytoreduction to allow this patient to

go on to transplant.

The other treatment failure that is listed

on this slide is a patient with monosomy 7. Again,

I think most of the pediatric oncologists would

recognize that as a fairly resistant to leukemia.

This patient came on study with 68 percent blasts

at study entry, went down to zero percent blasts

after 1 treatment, and was proceeding to transplant

when this patient did, in fact, have a relapse.

However, because of that substantial cytoreduction

that was still present, and the fact that there was

a haplo-identical donor available, the treating

investigator still wanted to go to transplant, so

while both patients were deemed as treatment

failures by the IRP, there were still significant

cytoreduction and benefit afforded these patients,

so they could move on to transplant.

I do want to mention just a couple others

really quick. Again, the median time to transplant

for these patients was 38 days, again, as little as

21 days after remission induction, they would go to

transplant.

The median number of cycles was 2, but

particularly in AML, they were very interested in

going to transplant as soon as possible, because

this disease is so difficult to treat.

As you can see at the bottom of the slide,

7 of the 12 patients are alive post-transplant, and

4 of these patients are still in remission.

[Slide.]

This slide shows the overall survival

curve for patients with AML enrolled in this study.

As you can see, the bottom line is for all patients

where the median survival was 21 weeks. The top

line represents those patients who had at least a

PR where the median survival was 39 weeks.

So, at this point, in summary, as far as

the efficacy results, again, recurrent pediatric

acute leukemia is a substantial unmet medical need

especially for patients with AML where new agents

are desperately needed.

We saw impressive response rates for

clofarabine in pediatric patients with ALL and AML

that had become cross-resistant to most standardly

available agents. The duration of remission was

long enough and sufficient enough to allow these

patients the opportunities for those with donors to

be able to proceed to transplant. Long-term

survival was observed in patients with both ALL and

AML who responded to clofarabine.

[Slide.]

Now, I would like to touch on the

integrated safety analysis. Just as a reminder,

again, this was combined data from both patients

with ALL and AML into one database, as well as

those patients from the Phase I study.

[Slide.]

This slide shows all Grade 3 and Grade 4

adverse events that occurred in greater than 10

percent of the patient population regardless of

causality. As you can see here, the most common

Grade 3/Grade 4 adverse event was fever and

neutropenia. This was followed by nausea, fever,

epistaxis, hypotension, sepsis, and anorexia.

A couple of factors should be noted here.

Number one, most of these patients had been in

relapse many times, weeks, if not months, before

coming onto this study. If you look at the list of

concurrent conditions, many of these events were

present at the time of study entry.

[Slide.]

This slide shows the drug-related adverse

events as determined by the investigators. Again,

these are Grade 3 and Grade 4 events only, that

occurred in greater than 5 percent of the patient

population.

As you can see here again, fever

neutropenia was the most common event, nausea,

fever, diarrhea, neutropenia, vomiting, and

dermatitis. In almost all cases, Grade 3 was much

more common than Grade 4.

[Slide.]

We also looked at the laboratory

abnormalities that were observed in this study.

Again, this represents Grade 3 and Grade 4

hepatobiliary and renal abnormalities that were

observed.

By far the most common were elevations in

transaminases, both ALT and AST. In almost all

cases, these tended to occur very early,

approximately one week after starting drug, and

then would resolve over the next week or two back

to baseline. We also saw increases in bilirubin,

creatinine, and alkaline phosphatase, but again, in

almost all cases, Grade 3 was much more common than

Grade 4.

[Slide.]

Deaths during study were fairly equally

divided between those from disease progression, as

well as those from non-drug or drug related AEs. A

couple of factors really stand out when you look at

these patients.

Number one, they were extremely heavily

pretreated patients before coming on study, many

with a variety of concomitant conditions and on a

variety of different medications before coming on

study. Many of them also had persistent disease or

progressive disease, as well as bacterial and

fungal infections. We also saw a number of cases

of capillary leak in this patient population.

[Slide.]

In summary, again, this study was

conducted with extremely heavily pretreated

patients. Most of the adverse events were

consistent with the underlying leukemia, and the

events were not unexpected particularly for a

cytotoxic agent, and most adverse events were

reversible, and not again unexpected.

At this point, I would like to introduce

Dr. Steve Sallan. Again, he was not an

investigator on these studies, but he is here today

to provide his perspective on clofarabine as a

pediatric oncologist and caregiver.

Clinician's Perspective

DR. SALLAN: Thank you very much. Good

morning.

[Slide.]

My name is Stephen Sallan and I have been

a pediatric oncologist for over 30 years, and have

been treating this patient population nearly every

working day of my life during that time, and have

been very blessed, as have been all the other

members of the pediatric oncology community, to

watch huge success being made in the 20th century

and getting to a point where childhood acute

lymphoblastic leukemia is really cured in 75 to 80

percent of children using multidrug chemotherapy,

multidrug chemotherapy that you have all seen

already, all of which was developed before the

1970s.

In AML, the cure rates are in the 40 to 50

percent range, again principally with multidrug

chemotherapy and enhanced by bone marrow

transplantation. Clearly, for us, the successful

treatment of relapsed and refractory pediatric

leukemias is our major challenge.

[Slide.]

Shown on these curves is really a picture

of the success story that I have alluded to, and

mostly from my medical oncology colleagues at the

table, I would like to reiterate that while we are

justly proud of these accomplishments, if one looks

at the curves, these really end in the '90s.

If one looks at what has happened in the

last decade, there has been incremental-only

increases, and as Dr. Arceci already alluded to,

these are approaching plateaus of about 80 percent

in ALL and somewhere between 45 to 50 percent in

AML.

[Slide.]

Now, there is no child on those curves who

has been cured who probably has not received 6 MP

and methotrexate if they had ALL, or cytarabine if

they had AML. Interestingly, when we look at these

data today, the question is what is the expectation

of a single active antileukemic agent, in this

case, against de novo ALL, so for this, we really

have to look at historic data, and I have adapted

this table from a textbook from 1974.

What you can see is that the workhorses in

these diseases, when they are tested against de

novo ALL as single agents, gave complete remissions

in this 20 to 25 to 30 percent range. I might also

say that the stringency of that definition of

complete remission, as you look at the old

literature, is highly variable.

So, when one sees a population of today's

relapsed, refractory patients, it is very difficult

to have any comparative population, and, in

addition, what impressed me when looking at the

clofarabine data, was that we saw a CR and CRp rate

that was really in the similar ballpark as

effective drugs are against de novo disease.

I think, although I am showing this for

ALL, you saw similar results with clofarabine

generating principally a CRp and PR's in that

population.

[Slide.]

What impressed me about the availability

of a new drug again as you have heard, in part, is

that 1 in 5 children with multiple drug resistant

ALL achieved a clinical response--sorry for the

CR--principally, the majority of those who went to

transplant had CR's or CRp's, and as you saw, some

PR's, as well, and similarly, for this relapsed,

refractory AML population, 1 in 3 children with

again multiple drug resistant disease was able to

come to transplant.

We strongly believe, as a community, that

transplantation is the curative therapeutic option

in the early 21st century for children with

drug-resistant childhood AML.

[Slide.]

So, in closing, what really impresses me,

as a clinician, is that we have a drug,

clofarabine, which is well tolerated, very

importantly, the absence of overlapping toxicities,

so that we can treat these children without

additive cardiac, renal, or other organ toxicities

permitting them to be good candidates when they get

into the transplant setting; that the drug provides

a clinical benefit, as shown in our responses, in a

very heterogeneous population, which right now, in

2004, is critically important.

I mean we are all focused in part on

targeted therapies, very few, none of which have

really come to this population, so the fact that we

have a drug that gives a clinical benefit in a very

heterogeneous population is extremely helpful, and

most importantly, for these children, there are no

meaningful alternatives.

I would say it is the last point that

really causes me to feel very positive and

enthusiastic and really desire to have something

that is well tolerated, that is beneficial, and is

available now, and that is why this data is

important to me.

Thank you very much.

Clofarabine Development Plan

DR. WEITMAN: Thank you, Dr. Sallan. We

feel that in some ways, we have embarked on really

a historical approach, an approach that is

different than what has been in the past when it

comes to pediatric oncology patients.

[Slide.]

The approach that we have taken

potentially is a new paradigm for getting access to

pediatric new agents into the pediatric community.

The sponsor, moving forward, commits to

several factors including the further development

of this agent in the pediatric oncology population.

Number 1. To continue to follow these

patients that are enrolled in these two studies for

long-term follow-up data. We also have a

commitment to move to less heavily pretreated

patients, both patients with ALL and AML, and to

ultimately, at some point, proceed to a randomized

study with clofarabine in newly diagnosed patients.

This commitment includes a very close

working relationship particularly with the

Children's Oncology Group, but also through CTEP.

[Slide.]

Just to highlight the program through the

Children's Oncology Group, we have two studies

moving forward there. One is in AML, and again

this is a combination study with ara-C and

clofarabine. This is actually a Phase II study in

patients with first relapsed AML.

The study chairs for this are Dr. Razzouk

from St. Jude, as well as Dr. Cooper from the

University of Alabama.

Again, in ALL, the parallel companion

study is a combination of cytoxan with clofarabine,

but there has also been interest from the

Children's Oncology Group to look at clofarabine in

combination, not only with etoposide, but also with

ara-C.

Again, this is a different population of

patients that we have studied in this submission.

In these follow-up studies, we are looking at less

heavily pretreated patients, in this case, second

relapsed patients.

Again, many of the patients enrolled in

our study would not have been eligible, and

certainly would not have been eligible for these

follow-up studies.

[Slide.]

To bring it full circle, our program in

the adult community is also moving forward,

although it has a little bit different focus. The

focus in the adult oncology community is

particularly focused on AML, because that is where

most of the activity has been seen.

Combination studies, particularly with

ara-C are moving forward, and there has been an

interest in the elderly population of patients

again because of some of the early pilot studies in

adults that have shown considerable activity in

this group of patients.

One study that I would like to highlight

is a CLO-141 study. Again, this was the

combination study of clofarabine with ara-C. This

study has met full accrual, but is still open.

Interim results of this study were just

recently published in Blood where the overall

response rate was 40 percent, again in a refractory

population of patients with leukemia, and the

overall response rate was defined as complete

remission and complete remission without full

platelet recovery.

Because of these results, our plans are to

move forward, and we have already discussed this in

brief with the FDA, as far as two potential

randomized studies, one with clofarabine and ara-C

in elderly patients newly diagnosed, as well as

clofarabine with ara-C in recurrent or refractory

adult patients with AML.

[Slide.]

At this point, I just want to return again

to the fact that we are here today presenting the

results of our pediatric studies in acute leukemia.

We found that clofarabine had an acceptable profile

particularly in this extremely heavily pretreated

population of patients, that impressive benefits

were observed including meaningful clinical

responses, such as CR, CRp, and even PR's and again

in this highly refractory patient population, that

allowed many of these patients to move on to

transplant.

As you can see here, 23 percent of the

patients were able to proceed to transplant, 14

percent of the patients with ALL, 34 percent of the

patients with AML. At data cutoff, 22 percent of

the patients with ALL, and 26 with AML are alive.

So, in conclusion, we believe that

clofarabine does meet an urgent unmet medical need

in a population of patients that frequently has not

been included in many other current protocol

opportunities, and the fact that activity has been

seen in a very highly resistant and refractory

group of patients.

[Slide.]

Again, I would like to thank the ODAC

panel members, as well as the FDA, today for the

opportunity to present the results of our pediatric

studies with clofarabine in patients with acute

leukemias.

Thank you.

DR. SANTANA: Thanks also to Drs. Weitman,

Sallan, and Arceci.

I want to recognize for the Committee, and

ask the individual to introduce himself by name and

affiliation, Dr. Hershfeld has joined the meeting.

DR. HERSHFELD: Steven Hershfeld, Food and

Drug Administration.

DR. SANTANA: Thank you.

With that, we will proceed with the FDA

presentation. Dr. Cohen, please.

FDA Presentation

DR. COHEN: Good morning. My name is

Martin Cohen, and the NDA being presented today is

No. 21-673. The study drug is clofarabine, which

structurally is chloro-fluoro-Ara-A. The sponsor

is ILEX Products, Incorporated.

Clofarabine is a second-generation purine

nucleoside analogue. It is a prodrug that must be

metabolized to its triphosphate conjugate by

deoxycytidine kinase within tumor cells.

Clofarabine has a greater affinity for this enzyme

than does other purine nucleoside analogues.

[Slide.]

The proposed indication for this NDA is

that clofarabine is indicated for the treatment of

pediatric patients 1 to 21 years old with

refractory or relapsed adult leukemias including

both pediatric acute myelogenous leukemia and acute

lymphoblastic leukemia.

[Slide.]

Regarding clofarabine dose and schedule, a

Phase I study in pediatric acute leukemia patients

indicated that when a daily times 5 schedule was

used, the selected dose was 52 mg/m

2. Clofarabine

treatment cycles are repeated every 2 to 6 weeks

following recovery to acceptable organ function.

[Slide.]

The pertinent clinical trials in this NDA

submission are summarized on this slide. There

were two, Phase II trials conducted by the sponsor,

one in pediatric AML, the other in pediatric ALL.

In addition, there was a pediatric Phase I study

conducted at M.D. Anderson Hospital.

[Slide.]

For both of the Phase II studies, the

primary efficacy objective was to determine the

complete response rate and the complete response

rate in the absence of platelet recovery, that is,

the CRp rate.

Secondary objectives were to document the

partial remission rate and also to document time to

event parameters including remission duration and

overall survival.

[Slide.]

Study inclusion criteria for both the AML

and ALL studies included an age less than or equal

to 21 and the presence of greater than or equal to

25 percent bone marrow blasts.

Eligible AML patients were in their first

or subsequent relapse and/or they were refractory,

having failed to achieve remission following one or

more different regimens.

Eligible ALL patients were in their second

or subsequent relapse and/or they were refractory,

having failed to achieve a remission following two

or more different regimens.

Patients had an ambulatory performance

status and had adequate bone marrow, liver, and

renal function.

[Slide.]

Response definitions are listed on this

slide. A complete response, or CR, required no

circulating blasts, no extramedullary disease, and

an M1 bone marrow defined as having less than 5

percent myeloblasts or lymphoblasts.

There also had to be recovery of

peripheral blood cell counts to a level of greater

than or equal to 100,000 platelets/microliter, and

an absolute neutrophil count greater than or equal

to 1,000/microliter.

A complete response in the absence of

platelet recovery meets all the criteria of a CR

except that the peripheral blood platelet count has

not recovered to 100,000/microliter.

A partial response is defined as no

circulating blasts along with an M2 bone marrow

defined as having 5 percent to 25 percent blasts

accompanied by the presence of normal progenitor

cells.

In addition, an M1 marrow without

peripheral blood count recovery would be classified

as a PR.

[Slide.]

A total of 18 Unites States sites

participated in the two, Phase II pediatric acute

leukemia studies. Thirteen of these sites enrolled

patients in the acute myelogenous leukemia study or

CLO-222, and 14 sites enrolled patients in the

acute lymphoblastic leukemia study, CLO-212. As

mentioned previously, an independent response

review panel was established to confirm response to

therapy for each patient. Independent pathology

review was also available.

[Slide.]

Demographics and Karnofsky Performance

Status of patients participating in the acute

myelogenous leukemia study are shown on this slide.

A total of 35 patients were enrolled and treated.

As indicated, the median age was 12, and ranged

between 1 and 22.

Approximately, one-third of patients were

female, and two-thirds were male. The majority of

patients were Caucasian. Despite the fact that

patients had relapsed and/or were refractory to one

or more prior regimens, performance status was good

with 89 percent of patients having a Karnofsky

Performance Status of 80 or better.

[Slide.]

Therapy is administered prior to entry

into the clofarabine AML study are listed on this

slide. The median number of prior induction

regimens was 3 with a range of 1 to 5. Five

patients received one prior regimens, 12 patients

received two, and the remaining 18 received three

or more prior regimens.

A total of 18 of the 35 patients, or 51

percent, received at least one transplant before

study entry, 13 of 35, or 37 percent, having

received one prior transplant and 5 of the 35, or

14 percent, having received two prior transplants.

[Slide.]

As previously indicated, the rate of

complete response and complete response without

platelet recovery were the primary efficacy

endpoints. Responses were determined by an

independent response review panel and confirmed by

FDA.

There were no complete responders and only

1 CRp. There were 8, or 23 percent, partial

responses. Seven of the 35 study patients were not

evaluable for reasons listed on the slide.

[Slide.]

This slide shows pediatric AML patients

who received a transplant after initial clofarabine

treatment. Because stem cell or marrow transplant

in pediatric AML may be associated with durable

remissions, there is pressure to proceed with a

transplant if a suitable donor is available.

In the clofarabine AML study, 12 of the 35

study patients underwent transplant including the 1

CRp patient, 6 of the PR patients, 2 of the 7

non-evaluable patients, and 2 of the 19 treatment

failures.

Transplants were performed after patients

had received 1 to 5 cycles of clofarabine

treatment. Because of the transplants, it was not

possible to determine duration of remission after

clofarabine treatments alone.

[Slide.]

This slide indicates some of the

difficulties encountered when evaluating this

application. As listed and as mentioned by Dr.

Pazdur earlier today, the traditional endpoints for

evaluating acute leukemia studies include the

confirmed complete response rate, complete response

duration, and overall survival.

Confounding factors in evaluating this NDA

submission were that some patients were

transplanted early, either before clofarabine

response could be confirmed or response duration

determined.

Further, some study patients received 1 or

more transplants prior to entering the clofarabine

study, and some did not. Whether these groups can

be compared must be considered today by the ODAC

Committee.

Because of the above difficulties, I chose

to evaluate an exploratory endpoint, namely, longer

time to progression with clofarabine treatment with

or without transplant, then, with the therapy

immediately preceding clofarabine, whether or not

it also included the transplant.

[Slide.]

This slide summarizes 4 pediatric AML

patients with a longer time to progression with

clofarabine plus transplant, then, with the therapy

that immediately preceded clofarabine.

One of these patients also had a longer

response duration to clofarabine plus transplant

than he had with his prior transplant.

Reviewing each of these patients

individually, Patient 14-03 had a time to

progression of 270 days for the treatment regimen

preceding clofarabine study entry. He had received

a prior transplant with a time to progression of

150 days. He received 5 cycles of clofarabine and

was a CRp. His clofarabine plus transplant time to

progression was 519+ days.

Patient 15-17 had a time to progression of

60 days for the treatment regimen preceding

clofarabine study entry. He had not received a

prior transplant. He received 1 cycle of

clofarabine and was a partial response. His

clofarabine plus transplant time to progression was

465+ days.

Patient 06-36 had a time to progression of

30 days each for the 2 treatment regimens preceding

clofarabine study entry. He had earlier received 2

prior transplants with response durations of 365

and 485 days, respectively. He received 5 courses

of clofarabine and was a partial response. His

clofarabine plus transplant time to progression was

130+ days.

Patient 14-31 had a time to progression of

30 days each for the 2 treatment regimens preceding

clofarabine study entry. She had not received a

prior transplant. She received 2 cycles of

clofarabine and was a partial response. Her

clofarabine plus transplant time to progression was

93+ days.

[Slide.]

Turning now to Study CLO-212, the acute

lymphoblastic leukemia study, this slide summarizes

the demographics and Karnofsky Performance Status

of participating patients.

A total of 49 patients were enrolled and

treated. As indicated, the median age was 12 and

ranged between 1 and 20. Approximately 40 percent

of patients were female and 60 percent male.

Hispanic and Caucasian patients comprised the bulk

of the study population.

Despite the fact that patients had

relapsed and all were refractory to 2 or more prior

regimens, performance status was good with 31

percent of patients having a Karnofsky Performance

Status of 100, and 39 percent a Karnofsky

Performance Status of 90-80.

[Slide.]

Therapies administered prior to entry into

the clofarabine ALL study are listed on this slide.

The median number of prior induction regimens was

3, with a range of 2 to 6. A total of 15 of 49

patients, or 31 percent, had received at least 1

transplant prior to study entry, 13 of the 49, or

27 percent, having received 1 prior transplant and

2 of 49, or 4 percent, having received 2 prior

transplants.

[Slide.]

Best response to therapy as judged by the

independent response review panel and confirmed by

the FDA for the pediatric ALL study is shown on

this slide. There were 6 complete responders and 4

complete responders in the absence of platelet

recovery. There were 5 partial responses and 8 of

the 49 study patients were not evaluable for

response.

[Slide.]

This slide shows pediatric ALL patients

who received a transplant after initial clofarabine

treatment. As previously noted in the AML study,

marrow transplant in pediatric ALL may be

associated with durable remissions, thus, marrow

transplant is often recommended if a suitable donor

is available.

In the clofarabine ALL study, 7 of the 49,

or 14 percent, of study patients underwent

transplant including 1 of the 6 CR's, 3 of the 4

CRp's, 2 of the 5 PR patients, and 1 patient who

was non-evaluable because of a poor quality bone

marrow.

Transplants were performed after 2 cycles

of clofarabine treatment in 5 patients and after 3

cycles of clofarabine treatment in 2 patients.

[Slide.]

In the pediatric ALL study, it was

possible to evaluate complete response duration in

patients who did not receive a transplant. As

previously indicated, 6 patients achieved a

complete response and 5 did not have a transplant.

The 2 CR's listed on this slide had a longer time

to progression with clofarabine treatment than was

achieved with their immediate prior therapies.

Another 2 of the 5 non-transplanted

complete responders remained in remission, but

follow-up is brief. As seen on this slide, Patient

07-18 had a time to progression of 60 and 30 days

for the 2 treatment regimens preceding clofarabine

study entry. She had not received a prior

transplant. She received 3 cycles of clofarabine

and was a complete response. Her clofarabine time

to progression was 143 days.

Patient 6-47 had a time to progression of

30 days each from the 2 treatment regimens

preceding clofarabine study entry. He had not

received a prior transplant. He received 2 cycles

of clofarabine and was a complete response. His

clofarabine time to progression was 76 days.

[Slide.]

As shown in this slide, 3 of the 4 CRp

patients, 2 with a transplant, 1 without, had a

longer time to progression with clofarabine with or

without transplant than with immediate prior

therapy.

Reviewing each patient individually,

Patient 09-24 had a time to progression of 120 days

for the treatment regimens preceding clofarabine

study entry. He had received a prior transplant

with a time to progression of 60 days. He received

3 cycles of clofarabine and was a CRp. His

clofarabine plus transplant time to progression was

259 days.

Patient 12-14 had a time to progression of

30 days for the treatment regimen preceding

clofarabine study entry. He had not received a

prior transplant. He received 2 cycles of

clofarabine, was a CRp. His clofarabine plus

transplant time to progression was 168+ days.

Patient 14-40 had time to progression of

30 days for the treatment regimen preceding

clofarabine study entry. He had not received a

prior transplant, nor did he receive a transplant

after clofarabine. He received 2 courses of

clofarabine, was a CRp. His clofarabine time to

progression was 64 days.

[Slide.]

Turning now to the supporting trial in

this application, the CLO Phase I study performed

at M.D. Anderson Cancer Center is summarized on

this slide.

In this study, patients 21 years or

younger with refractory leukemia or lymphoma, who

had a Zubrod Performance Status no greater than 2,

and who had adequate organ function were eligible

for enrollment.

There were 25 acute leukemia patients

entered, 17 with ALL, 8 with AML. Using M.D.

Anderson response criteria, complete response were

noted in 5 of the 25 patients 4 with ALL and 1 with

AML. In addition, there were 3 PR's.

ILEX convened an independent response

review panel to review the 5 M.D. Anderson complete

responses using the same modified COG review

criteria that were used in the sponsor's Phase II

studies. The independent response review panel

reclassified the 5 CR's to 2 CR's, both in ALL

patients, 1 CRp in the AML patient, and 2 PR's.

[Slide.]

Turning now to safety, this slide

summarizes clofarabine exposure by cycle. The

database includes 113 patients derived from the

sponsor's Phase II studies, the M.D. Anderson

study, and from adult trials that included

pediatric patients.

As indicated, all 113 patients received at

least 1 cycle of clofarabine, 68 received 2 cycles,

and 24 received at least 3 cycles. Dose reductions

were necessary as is expected for heavily

pretreated patient population.

[Slide.]

Some of the clinically more important

baseline conditions present in the 113

clofarabine-treated patients are listed on this

slide. Despite ambulatory performance status, the

patients appeared to be relatively fragile with CTC

Grade 3 to 4 baseline toxicities including

tachycardia, pyrexia, nausea, and anorexia.

[Slide.]

An overview of adverse event occurrence is

shown on this slide. Ninety-nine percent of

patients had 1 or more adverse events, and 83

percent had 1 or more serious adverse events. Four

percent of patients discontinued therapy because of

an AE. Fifty-three percent of patients had at

least a Grade 3 AE, 23 percent a Grade 4 AE, and 20

percent a Grade 5 AE.

[Slide.]

Frequent adverse events are summarized on

this slide. Gastrointestinal toxicity in the form

of vomiting, nausea, and diarrhea occurred

commonly. Grade 3-4 vomiting occurred in 10

percent, Grade 3-4 nausea in 16 percent, and Grade

3-4 diarrhea in 10 percent of patients. Grade 3-4

febrile neutropenia occurred in 58 percent.

Constitutional symptoms, such as headache,

pyrexia, rigors, fatigue, and anorexia were also

common and occurred in 30 percent to 48 percent of

patients.

[Slide.]

Other adverse events noted during

treatment are listed on this slide. Infections

were an important adverse event because of

prolonged immunosuppression and myelosuppression

from both current and prior therapies. SIRS, or

systemic inflammatory response syndrome, capillary

leak syndrome manifested by the rapid onset of

respiratory distress, hypotension, and multi-organ

failure occurred in 10 patients. It most often

occurred in conjunction with rapid tumor lysis.

Renal insufficiency was multifactorial in

etiology included nephrotoxic antibiotics,

hyperuricemia from tumor lysis, and hypovolemia and

hypotension. Hypotension was a component of SIRS,

but was also associated with sepsis and

dehydration.

Hepatobiliary toxicities were frequently

observed as the liver is a known target organ of

clofarabine toxicity. Approximately one-third to

one-half of study patients had Grade 3 elevations

of transaminases during study.

Left ventricular systolic dysfunction was

noted in 15 study patients. This might have been a

direct cardiotoxic effect of clofarabine as

clofarabine cardiac toxicity has been seen in

preclinical rat studies.

Numerous contributing factors were

present, however, including sepsis, prior

anthracyclines, and prior whole body radiation

therapy. Hand-foot syndrome was noted in 12

percent of treated patients.

[Slide.]

To summarize standard efficacy results for

the study population of relapsed, refractory

pediatric acute myelogenous leukemia patients,

there was 1 CRp, or 3 percent, and 8 PR's, or 23

percent, among the 35 treated patients.

Remission duration could not be determined

because 12 patients went on to transplant.

[Slide.]

To summarize exploratory efficacy results

for the study population of relapsed, refractory

acute myelogenous leukemia patients, 4 clofarabine

plus transplant AML pediatric patients had a longer

time to progression of clofarabine plus transplant

to time to progression of immediate prior therapy

with or without transplant.

One of 2 patients with a prior transplant

also had a longer time to progression with

clofarabine plus transplant than with his preceding

transplant. The other is too early to evaluate.

[Slide.]

To summarize the standard efficacy results

for the study population of relapsed, refractory

pediatric acute lymphoblastic leukemia patients,

there were 6 CR's, 4 CRp's, and 5 PR's among 49

treated patients.

Two CR's who did not have a transplant had

a longer time to progression than was achieved with

their immediate prior therapies. Another 2 of the

5 non-transplanted CR's remain in remission, but

follow-up is brief.

One CRp who did not have a transplant also

had a longer time to progression than was achieved

with his immediate prior therapy.

[Slide.]

To summarize exploratory efficacy results

of the study population of relapsed, refractory

pediatric acute lymphoblastic leukemia patients, 2

CRp patients had a longer time to progression with

clofarabine plus transplant than with the treatment

immediately preceding clofarabine.

One of the 2 above patients had a

pre-clofarabine transplant. This patient also had

a longer time to progression with clofarabine plus

transplant than with his previous transplant

regimen.

[Slide.]

As to safety conclusions, toxicity was as

expected for a heavily pretreated relapsed,

refractory acute leukemia population. Principal

toxicities were gastrointestinal including nausea,

vomiting, and diarrhea. As expected, there was

significant hematologic toxicity, fever and febrile

neutropenia, hepatobiliary toxicity, infections,

and renal toxicity.

SIRS, tumor lysis syndrome, multi-organ

failure, hypotension, renal insufficiency, and left

ventricular systolic dysfunction also occurred.

With attentive patient care, however, the drug

appeared to be tolerable.

[Slide.]

To conclude, I would like to come back to

a slide that I showed earlier. This slide

indicates some of the difficulties encountered when

evaluating this application.

As listed, the traditional endpoints for

evaluating acute leukemia studies include confirmed

complete response rate, complete response duration,

and overall survival.

Confounding factors in evaluating this NDA

submission were that some patients were

transplanted early either before clofarabine

response could be confirmed or response duration

determined.

Further, some study patients received one

or more transplants prior to entering the

clofarabine study, and some did not. Whether these

groups can be compared must be considered by ODAC.

Because of the above difficulties, I chose

to evaluate an exploratory endpoint, namely, longer

time to progression with clofarabine treatment with

or without transplant, then, with the therapy

immediately preceding clofarabine, whether or not

it also included the transplant. Whether this

appropriate must also be considered by ODAC.

Thank you for your attention.

DR. SANTANA: Thank you, Dr. Cohen, if you

would remain on the podium, there may be some

issues of clarification or questions for you.

I want to go ahead and open up the session

for questions, and I want to reiterate to the

Committee, it is primarily to ask clarifications or

questions related to the presentations, comments

that have been made, or anything that is in your

briefing document. I really don't want to get into

the issues or the questions for discussion later in

the morning.

Questions from the Committee

DR. SANTANA: I will go ahead and get

started with one question. I want to focus a

little bit on the issue of toxicity as it may help

me define clinical benefit for these patients.

When you looked at toxicity across the

board, then specifically at those patients that did

not go to transplant, but had either a CR or a PR,

did you notice any difference in toxicity rates,

particularly serious infections, hospitalizations,

that nature, that would help me assess whether

those patients that did not have transplant really

benefited?

DR. COHEN: Well, I think all of the

toxicity data was gathered before patients went to

transplant. It represented the effects of

clofarabine treatment--

DR. SANTANA: Could you point to patients

that were not transplanted? I believe there were 9

ALL patients and a few AML patients.

DR. COHEN: I did not break it down that

way, I did not look specifically.

DR. SANTANA: Maybe the sponsor can

comment on that later, if they have looked at the

data that way.

Kind of a follow up to that toxicity

issue, this theme of hepatobiliary toxicity and its

resolution, were you able to assess whether that

required dose modifications in subsequent courses

or even extending it a little bit, did it impact

those patients that went on to transplant?

DR. COHEN: Well, the protocol, depending

on the level of hepatobiliary toxicity, the

protocol called for dose modifications, and those

modifications were followed.

DR. SANTANA: Yes.

DR. MARTINO: A question to the sponsors.

Within the protocol, can I assume that there were

some guidelines as to when a patient might be taken

to transplant? How was that decision made, by whom

was it made, did the protocol restrict or simply

allow judgment truly to the individual clinician?

DR. WEITMAN: The protocol did allow

patients to go to transplant, and again, as in many

of these studies, if they had available donor, that

was pretty much the primary requirement. There

wasn't a requirement for number of cycles or

anything else before going to transplant.

DR. SANTANA: Dr. Cheson, did you have a

question?

DR. CHESON: Yes, actually, I have three

questions for the sponsor.

Number one, if you could please flash

Slide 26 of your slide kit, and while you are doing

that, let me ask my other one while he is taking

time to do this.

You included the Phase I patients in your

toxicity analysis and give a percent of toxicities.

Were those all the patients in the Phase I,

including those who had rather low therapeutic

doses or only those closer to the MTD or DLT?

DR. WEITMAN: It included all patients

including those at the lower dose.

DR. CHESON: So, is that really fair to

include the patients at the really low doses?

DR. WEITMAN: We thought it was. Well,

first of all, we also included patients at a higher

dose, as well. When you look at the numbers of

patients, 13 patients out of the 25 were at the

MTD, at the 52, so the majority, if not more, 51

percent were at that dose, plus there is additional

patients treated above that.

DR. PERRY: What proportion of responses

were seen in that group of people who got 52

milligrams or more?

DR. WEITMAN: The responses were spread

throughout from 30 to 40 to 52 mg/m

2. So, it was

spread I will say fairly equally, but certainly at

the 52 mg dose, that is where the majority of the

patients were treated at.

DR. CHESON: The second point. Looking at

these curves, you said the CR and CRp, which is the

upper curve, has a median of 20.2 weeks. The 50

percent dotted line seems to hit smack right around

12 weeks. Am I missing something here? Take that

50 percent line, you draw it across, it hits there.

DR. WEITMAN: Which curve are you looking

at?

DR. CHESON: The upper curve.

DR. WEITMAN: The yellow?

DR. CHESON: The yellow curve. The 50

percent seems to hit at about 12, and not 20.2.

DR. WEITMAN: If I can, I will ask Brett

Wacker to comment on this.

DR. WACKER: The way that the median is

calculated when it goes all the way across, it

takes the average from the first point where it

hits the 50 to the end of it, so that is why the

20.2 is in the middle of that interval.

DR. CHESON: So, we would need longer

follow-up to see what really happens to that.

Okay.

The third, this CRp thing that you have

got there, there have been several published, at

least in adult response criteria in AML, one in

1990 and the other about a decade or so later, in

neither of those was CRp included as a response

criterion, and, in fact, it was discouraged in the

most recent International Working Group as being

included because it really hadn't been validated as

being different from CR. In other words, there was

some evidence that CRp's, in fact, don't do as well

as CR's, and the recommendation of that

International Group was that CRp's not be included

with CR's.

Are there data from the pediatric studies

to suggest, to validate CRp as an endpoint? It

snuck in here with the gemtuzimab study, the

pivotal trial there, and it has kind of hung on

ever since, but the most recent response criteria

have clearly suggested that it not be used.

DR. WEITMAN: I think you will see that in

some of the pediatric leukemia studies going

forward, that CRp is becoming more of an endpoint

that they are looking at.

What we felt, at least looking at the

study, number one, was that patients with CRp, most

of these patients, realize came into the study

following bone marrow transplant and rarely had any

platelet recovery even coming into the study. So,

it was somewhat hard to know whether lack of

platelet recovery after clofarabine really was any

sign of lack of activity.

When we looked at the patients with

overall survival, there were small differences, but

again small numbers comparing CR with CRp. But I

think it is a valid point that just needs to be

continued to be followed particularly in the

pediatric population.

If I can--

DR. SANTANA: Sorry, go ahead.

DR. WEITMAN: I just wanted to get back to

your question earlier, Victor, about toxicities.

Clearly, those patients that received multiple

cycles, the toxicity that really was dose-limiting

at least as far as giving the subsequent cycles was

bone marrow suppression, and that particularly of

patients that have received 8, 10, 11 cycles of

treatment, that was predominantly what was

requiring a delay in treatment or dose reduction

was myelosuppression.

I don't know if that gets to your

question.

DR. SANTANA: Sort of. What I am trying

to assess is those patients that did not get

transplant, that remained on study drug because

they had either a CR or PR. That was particularly

the ALL population, so I believe you had 9

patients, and there were a few patients there that

were in the 100+ day kind of remission status,

which indicates to me that they must have received

a few cycles of therapy.

DR. WEITMAN: Correct.

DR. SANTANA: I was trying to get a sense

of what impact the therapy had on their infection,

hospitalization, liver toxicity, to get at this

issue whether they were truly benefiting or whether

the drug was giving them toxicity that they

otherwise considered difficult to manage.

DR. WEITMAN: Right. I think I will ask,

if I can, Dr. Steinherz, Peter Steinherz, to

comment on that, because he had probably the most

experience with these patients that went on

multiple cycles.

DR. STEINHERZ: Patients who received

multiple cycles of clofarabine had a near normal

quality of life. They did fairly frequently after a

cycle of chemotherapy, have a brief fever and

neutropenia admission, but the cycles were done

every 28 days, and other than the two, three day

hospitalizations, the rest of that time was at home

with full quality of life.

DR. SANTANA: Did any of the serious

infections that were reported in the briefing

document occur in those patients?

DR. STEINHERZ: Not once they achieved

remission. The infections that were serious were

really during remission induction.

DR. SANTANA: Dr. Hussain.

DR. HUSSAIN: I have a couple of

questions, please. I don't think I heard in any of

the slides, other than a description of statistics

being traditional, what were the actual efficacy

assumptions that would have been desirable that you

put in prospectively in the trial?

The second question is in comparable

settings, I assume transplant and the description

from the experts were that there are multiple drugs

available in the last 10 years, so I assume there

may have been other agents that might have been

tested in previously pretreated or heavily

pretreated patients, that made it to approval.

What were the efficacy or the

characteristics of those agents that led them to be

approved in terms of whatever criteria they were

looking at?

DR. COHEN: In terms of criteria for

approval, in those days, in the 1960s, 1970s, I

don't know for sure, but I would expect it would be

response rate.

DR. HUSSAIN: In the past five years, have

there been no drugs approved in this setting?

DR. COHEN: In pediatric acute leukemia, I

don't think so.

DR. HUSSAIN: Or 10 years?

DR. SANTANA: I don't think there has been

any pediatric leukemia drugs approved in more than

a decade.

DR. HUSSAIN: I am sorry, the second

question was dealing with the efficacy endpoint

that were prospectively put in the Phase II trials.

DR. WEITMAN: I will ask Dr. Tannen to

come up and comment on that, please.

DR. TANNEN: Can I have Slide 101, please.

This trial was designed based on using Fleming

2-stage design, and at that time, we have the data

available on the Phase I trial based on 25

patients, and the response rate at that time, they

were observing is 30 percent response rate. The 40

percent response rate was hypothesized for the

clofarabine treatment groups with the control group

of the 20 percent, a 2-fold increase with the

clofarabine.

So, Fleming design basically says that you

look at the data with the 20 patients and observe

the response rate. At that time, the response rate

was observed with ALL, about 20 percent, and the

activity was seen, and the criteria to move to

second stage was not met at this time, but based on

the advice from the investigators, as well as the

FDA, it was decided to move to the Stage 2.

Activity was seen at the first stage, and

you have to see what is the clinical significance

of 20 percent response, which was observed in this

trial, and as mentioned by Dr. Weitman, the number

of patients, about 34 percent in AML patients, went

to transplant.

So, the response rate is what we saw here,

the 20 percent, which we believe is clinically very

meaningful, and has to do with the patient

population that is very heavily treated. In the

AML patients, it has to do with the patient

management that is the key issue.

DR. WEITMAN: If I can, I would like to

ask Peter Adamson to step up, as well, and make a

comment about this.

DR. ADAMSON: So, without a reasonable

comparative database to look at, to say, well,

where should the bar be in ALL or AML, I personally

think the assumption of 40-20 was an incorrect

assumption, and it missed the mark by a large

measure.

As Steve Sallan had pointed out decades

ago, single-agent activity in newly diagnosed

patients are in the 20 to 40 percent range, and I

think you can count on two fingers the single-agent

activity in ALL above that.

So, setting the bar in this population at

40 percent, my personal view was way off the mark.

Where should it be? I think it is a more difficult

question, but maybe I can share with you very

briefly the Children's Oncology Group challenges in

the same population, so with a concurrent

population, the challenges that we are facing.

We have three, Phase I studies in ALL and

AML going on. For Phase I, as you know, a criteria

for evaluability, they have to make it through a

cycle of therapy, which is 28 days.

We are nearing the point where we are

going to have to abandon that design, because we

have about a 70 to 80 percent inevaluability rate

in Phase I. These patients are so heavily

pretreated and are so rapidly progressive, we can't

get Phase I data anymore in leukemia, because we

can't get them through four weeks of therapy.

So, that will set the stage saying the

mere fact that they got through a Phase I, to me,

was an indicator that there is something going on,

because we can't get through Phase I's right now.

Now, in Phase II, where are we putting the

bar? That is a moving target, and maybe I can take

this brief opportunity to come back to Dr. Pazdur's

initial introduction where CR, duration of response

and overall survival are the three criteria.

I certainly agree with those, but it is a

shifting paradigm, because I no longer think we are

going to be able to get duration of response data

in this population, and there are two reasons for

that.

In AML, it is very clear that the current

standard of care is that we whisk these patients to

transplant as soon as feasible, as soon as you have

had cytoreduction, they go to transplant.

So, as much as scientifically, we would

all like to see a duration of response that is not

confounded, we are not going to. We are not going

to do it in our trials, and I don't think industry

is going to do it, because the standard of care

won't allow it.

Now, in ALL, duration of response becomes

somewhat of a different challenge. Yes, they are

going to transplant, but I think Dr. Pazdur

correctly pointed out that a reasonably high

fraction of responders didn't have a repeat marrow

assessment.

This is where our scientific desires run

head to head with some ethical concerns, as well as

current standards.

As everyone or many of you know, in

pediatrics, there are additional safeguards

afforded, FDA, DHHS regulations, and for studies

that offer no potential for direct benefit, we

can't mandate those studies in children.

So, as much as we would like to see repeat

marrows done, the reality is once a child is in CR,

it is hard to describe what the benefit to the

child is to repeating a marrow later on.

The reality in ALL, for those of us who

take care of it, is it is no secret when a child

recurs. We don't need to look in their marrow. We

look in their marrow to confirm it, but it is not a

secret.

So, it becomes difficult once you have

documented a CR, to go ahead and confirm it later.

We absolutely would like to do that. The reality

on the front lines is that it becomes exceedingly

difficult to.

So, the criteria we have put forward as

far as CR's, duration of response and overall

survival, the CR's, we need to document. AML gets

confounded, ALL, less so, but duration of response

is going to be a pretty difficult standard right

now.

I know I have gone off on a tangent a bit,

but the design as far as 40-20, I think was

misplaced, and I would right now say for ALL, we

are tending to set designs in our studies at 25, so

the Fleming 2 stage at 25 with a 20 percent

response rate of interest.

DR. SANTANA: Dr. George, you had a

question and since this is kind of in this area of

discussion. Go ahead.

DR. GEORGE: Well, it is related. I have

a question for the sponsor. In the duration of

response, and, for that matter, the overall

survival curves, it appears that the transplant

issue was handled by simply ignoring it. I mean the

analysis would have been exactly the same if no one

had gotten a transplant, is that correct? This is

the clarification question.

DR. WEITMAN: Well, again, we looked at

those patients that went on to transplant, correct,

and we included that in our duration of remission

and overall survival as part of that.

DR. GEORGE: Well, just to be clear, that

means that if someone did get a transplant, and

then later relapsed, I mean that time to relapse

was taken, whatever it is, just as is.

DR. WEITMAN: Correct.

DR. GEORGE: Another issue that just came

100

up, that I hadn't thought of before, was this issue

of confirming the relapse. I assume all those were

done with some marrow samples, so they were

confirmed.

DR. SANTANA: Let me go back to that

issue, because I got confused, too, reading the

documents. Are we talking about not confirming the

response with follow-up marrow, or not confirming

the relapse with a follow-up marrow?

Can the sponsor clarify that for me? Did

all patients have a confirmatory response for the

tagging of response, and it was that when they

recurred or progressed, a marrow was not repeated

to document that?

DR. WEITMAN: It depends, Victor. Most of

the cases did have repeat marrows done, but

clearly, if there were patients who had no evidence

of relapse, and all of a sudden developed, 20, 30,

40 percent blasts in their peripheral blood, then,

again, a lot of those patients did not have

confirmatory marrows done at that time, but at that

point, that was considered disease progression

101

without necessarily requiring a confirmed repeat

marrow.

DR. SANTANA: So, this issue of not doing

bone marrows was at the time of progression, is

that correct?

DR. WEITMAN: Correct.

DR. BRAWLEY: I didn't understand it that

way either. Can we clarify that?

DR. COHEN: The issue in terms of

confirmed marrows was not confirming the response,

that is, a second marrow was not done within a

month of the first marrow. It had nothing to do

with progression.

DR. BRAWLEY: This is very important. Can

you just slow down just a little bit. You did not

do a marrow to confirm complete response. You sort

of looked at the blood and accepted complete

response on some sort of religious basis, is that

right?

DR. WEITMAN: Otis, no, not exactly.

DR. BRAWLEY: Okay.

DR. WEITMAN: Confirmatory marrows were

102

done in most of these patients. Now, clearly, at

the time that that first marrow that was done that

showed remission, if that patient was in remission

and had a donor, they would go right to transplant.

Many of those patients would still get a

repeat marrow the day or so before transplant, but

again, most of those patients had, first of all,

had a marrow done to confirm response or to show

that they were in remission, and then they would go

either to transplant or would continue on the

study, and those patients that continued on study

without a transplant would get marrows as dictated

by the protocol.

DR. HUSSAIN: Is it possible to put a

slide about that, because I think right now I am

totally confused about what was confirmed, what was

not confirmed.

DR. SANTANA: Do you have a summary of

that?

DR. WEITMAN: Again, what I can just

highlight again is that for patients, I think there

is one slide that has the confirmatory marrows. It

103

should be in the briefing document, as well, that I

believe the FDA has supplied to you, as well.

DR. PAZDUR: If you take a look at the

questions for the Committee, it does go through on

Table 2 and Table 4, the differences, and these are

in patients that did not have transplants, because

we look at transplants as really confounding this

whole issue of what a response is.

You could see the data and a single

asterisk means the responses were not confirmed,

double asterisks, responses were confirmed. What

we are talking about here is a discrepancy of

whether these had a confirmatory bone marrow

biopsy.

I guess as the sponsor said, and perhaps

he could give this information because I don't have

it off the top of my head, if you take a look at

all of the patients, how many had a confirmatory, a

bone marrow at, for example, 4 weeks, because that

is what you are after, the CR's.

DR. WEITMAN: Again, just to reconfirm

that all patients had a marrow to document

104

remission. Now, as far as confirming remission,

the majority of those patients, and I don't have

the exact numbers, but the majority of those

patients--

DR. PAZDUR: That is the question. What

is the majority?

DR. WEITMAN: Let me get it.

DR. SANTANA: Ruth Hoffman had a comment

or a question?

MS. HOFFMAN: This is for the sponsor, as

well. The overall survival for ALL median was 11.7

weeks at the close of study, and for AML, it was 21

weeks. Because the study for a lot of these

patients was January or February of this year, this

is now 10, 11 months later, do you have the status

of the survivors, how many are still alive?

DR. WEITMAN: If I can show the follow-up

on some of the additional patients.

DR. SANTANA: Go ahead.

DR. WEITMAN: I will ask for the slide

that looks at follow-up patients and survival.

I would like to, if I can, show the

105

response follow-up first.

DR. SANTANA: Since you have the survival,

why don't you just do that first.

DR. WEITMAN: Okay. We will start here.

This is just a follow-up for the patients that were

on the study. Again, what was submitted in the NDA,

and again I will comment that the updated data has

not been reviewed by the FDA, and they have not

seen this follow-up data.

Essentially, all it shows is that the

response rates were similar with the additional

follow-up data or additional patients that have

been treated with clofarabine for both ALL and AML.

If I can show the survival slide. In

follow-up, this shows 22 percent of patients alive

with median follow-up of 28 weeks, and I believe

this is as of November, actually, this was the

beginning of November, September 30th of this year

with 18 percent patients alive with median

follow-up of 47 weeks.

DR. SANTANA: Thank you.

Dr. Maldonado, you had a question or a

106

comment?

DR. MALDONADO: Just a couple of

questions. This is just for my own clarity on the

biological plausibility of the results we are

seeing here. I hate to take you out of the

clinical, but I want to go back to the basics

seeing that the sponsor had been surprised by the

higher unexpected results.

What is known about the mechanism of

action of clofarabine, and why do you think this is

a different nucleoside analogue?

Pardon my ignorance, but it appears that

that is basically what has been the surprise for

the sponsor, and is that because the cells that

have more permutations to elicit the resistance? I

still don't understand much about the biology and

molecular biology of the drug here.

DR. WEITMAN: I will make two comments. I

will first start by asking Dr. Gandhi to come up

and talk about the preclinical mechanism of action.

DR. GANDHI: Clofarabine is very similar

to other nucleoside analogues with regard to its

107

metabolic aspect, that it does require

deoxycytidine kinase to get phosphorylated.

It resembles to some extent to cladribine

and to some extent to fludarabine. With

cladribine, it resembles regarding metabolic

aspect, that is, it is a favored substrate for

phosphorylation.

With regard to clofarabine, it resembles

that once it gets incorporated into DNA, it results

in chain termination, whereas, with cladribine,

there is no chain termination.

Does that answer your question regarding

the comparison with other nucleoside analogues?

DR. MALDONADO: So, you think that the

activity or the efficacy we are seeing in clinic is

because this is a chain terminator versus the other

comparisons?

DR. GANDHI: There are several other

factors. One, especially, in the clinical

pharmacology, which goes a little bit away from the

mechanism of action. When you look at the

clofarabine triphosphate in the leukemia cells

108

during therapy, it is very different from

cladribine triphosphate and fludarabine

triphosphate.

All these three nucleoside analogue

triphosphate, they live for a long time in the

indolent leukemia setting in the leukemic

lymphoblasts, but when you compare it for the acute

leukemia setting, fludarabine triphosphate half

life is 7 hours, cladribine triphosphate half life,

which hasn't been tested very well, in fact, we

have the guru right here for all the cladribine

studies, but it was about 7 hours half life, and

for clofarabine triphosphate, the half life is more

than 24 hours.

So, we think there is a benefit with

clofarabine because it remains for a long time in

the acute leukemia blasts.

DR. SANTANA: Dr. Pazdur, you had a

comment?

DR. PAZDUR: I wanted to follow up on

Maha's question about response rate and what is the

appropriate response rate here.

109

I want to stress that remember this study

was done as a Phase II trial, and one of the

concepts of a Phase II trial obviously is an

exploratory trial where you would get data and a

response rate which would indicate whether the drug

should be carried on further in drug development.

That was the original intent of when we talk about

what a Phase I study, Phase II, Phase III paradigm

is.

What we are talking about here, though, is

drug approval, and I want to stress that simply a

mere screening for drug activity does not

necessarily constitute the requisite information

that one should have for approval of a drug.

Remember, when we are talking about

approval of a drug under accelerated approval, you

have to be reasonably likely that this response

rate can predict clinical benefit.

Now, you know, we have in the past with

other diseases in adults, for example, lung cancer,

colon cancer, taken a look at single-arm trials in

very refractory diseases, in some of the

110

hematological malignancies where we were able to

take a look at duration.

If we can't take a look at duration, maybe

this paradigm of taking a look at very, very

refractory disease patients and trying to approve a

drug on that basis, is not the appropriate paradigm

to have in all clinical scenarios here, and I think

that is an important point.

Somebody could actually do one day a

randomized study for drug approval and surprise the

FDA.

DR. SANTANA: I will not comment on that.

Dr. Martino?

DR. MARTINO: Actually, I want to ask this

question of Dr. Arceci. It is a clinical

understanding that I need.

In these children, when you achieve a

partial remission with this agent or any agent,

what does that mean to you clinically? Is that

meaningful, to what degree is it meaningful, and

what does it mean in terms of their survival?

DR. ARCECI: In terms of the practical

111

side of what a partial remission means is that it

usually buys time, and usually, it improves quality

of life during that period of time. It also means

a chance to proceed to a more definitive therapy.

If a patient does not have at least a good

partial remission, it essentially buys that extra

time to finish marrow typing, finish insurance

approval for transplant, then, we don't get to

transplant.

So, a PR has enormous utility in our

business although it, of course, is not going to

result in a cure. Those patients with PR's are

going to relapse clearly and progress, but it gives

us a chance that if we can move that process

forward, we can often cure it, or at least in a

percentage of those patients can cure them.

Is that answering your question?

DR. MARTINO: You gave me a sense of what

I needed to hear. Thank you.

DR. SANTANA: Dr. Poplack.

DR. POPLACK: I just want to go back to

the point that Dr. Pazdur made two comments ago,

112

and that related to the ALL patients who were

judged to have responses.

In the data that we received, it is not

possible to determine which patients did or did not

have that confirmatory second marrow, and I think

it is critically important for us to get that

information. So, that is the first point.

The second one is--and that is on Table 2

of the questions that were given to us--the second

question I have for Dr. Weitman is, what would be

the plans--and I assume this is a fair question,

and it seemed like what you stated was a little

vague to me--for a confirmatory study were this,

for example, to be approved?

DR. WEITMAN: With regards to your first

question, my helpful colleagues here pointed out

that 9 of the 15 patients with ALL had a

confirmatory marrow, again a second marrow after

the initial marrow was done, so there was 9 of 15

for ALL.

Two of 9 for AML had a confirmatory

marrow. The remaining patients went on to

113

transplant or probably had either progressive

disease at the point, so again, 9 of 15 for ALL,

and 2 of 9 for AML.

DR. POPLACK: I think the information I

would like to get is which of those responses, if

we look at the patients who responded, which ones

did have or didn't have a subsequent marrow.

DR. WEITMAN: If I can then also touch

base on your second question, and that is, again,

our plans are to work with the Children's Oncology

Group towards a randomized study.

As you know, I think doing a randomized

study in this patient population is very difficult,

if not heroic, efforts to undertake, but our plans

would be to move forward with the cooperative

groups and at some point go into a randomized study

where clofarabine is added on to the backbone

therapy of either patients in first relapse or

newly diagnosed patients at the appropriate time,

where this drug was added onto the backbone,

looking for changes in survival or duration of

remission.

114

DR. SANTANA: I don't think I heard the

answer to the first question yet, and I think we

need to get an answer because it has come back now

three or four times in the last 10 minutes.

So, of the patients in ALL and AML, who

actually were coded as having a response, how many

of those had a subsequent marrow that confirmed

that response? Can somebody give us that

information?

DR. COHEN: If you look at the FDA review,

Table 11 on page 25 is the AML, and Table 23 on

page 33 is the ALL, and the seventh column down is

patients who had confirmatory marrows.

DR. SANTANA: Can you summarize it

publicly because I think some people don't have the

documents in front of them? I hate to put you in a

position, but--

DR. WAYNE: And also remind us of the page

and the table number, please.

DR. COHEN: This is Table 11, page 25, FDA

briefing document. It lists 9 responding AML

patients and 2 of the 9 had confirmatory marrows,

115

and the other 7 did not, of the responders, this is

just responders.

DR. SANTANA: So, 2 of the 9 AML patients

that were coded as having an initial response, had

a confirmatory marrow, is that correct?

DR. COHEN: That is correct.

DR. KURTZBERG: I still don't understand

if that means they had the first marrow that

confirmed their response, and then they did have a

second.

DR. COHEN: This is the second. The

confirmatory marrow is the second one.

DR. SANTANA: It's the follow-up bone

marrow.

DR. COHEN: And for ALL--

DR. PERRY: Excuse me. Just for clarity,

they had a bone marrow that met response criteria,

but only 2 of the 9 had a second bone marrow at

least a month later establishing the came criteria.

DR. SANTANA: That is my interpretation.

DR. COHEN: That is mine, too.

For ALL--

116

DR. SANTANA: That is Table 23 on page 33.

DR. COHEN: Right. There are 15 patients,

confirmatory marrows were 9, 9 of the 15 had

confirmatory marrows.

DR. SANTANA: Thank you.

DR. WEITMAN: Can I ask Dr. Arceci to

comment about Dr. Poplack's question regarding the

COG and plans moving forward?

DR. SANTANA: Yes, I think that is an

important point to clarify.

DR. PAZDUR: While he is coming up to the

table, I would also like to emphasize for Dr.

Poplack that the confirmatory study can be done in

an earlier stage of the disease. We have

repeatedly done this in adult disease where

somebody might get an approval in a very refractive

population of colon cancer patients and then the

confirmatory study is done in the same disease in

an earlier stage.

We have interpreted that as a way of

moving forward drugs and trying to escalate the

speed of drug development basically.

117

DR. ARCECI: I would just comment briefly

on David's point, and that is, a randomized study

of this drug in newly diagnosed patients, or even

in the relapse setting, is very unlikely to happen,

but the cooperative groups right now are, in fact,

moving towards combination trials, and both in AML

and ALL, those studies are moving very rapidly

through the cooperative group mechanism.

We have plans to then try to randomize

those combinations in the setting of newly

diagnosed patients. We are doing that now in

upfront studies with new agents and combining them

to see if we can offer an advantage over standard

therapy.

So, what will probably happen I suspect

with the combinations is that because of the

somewhat noncross-resistant toxicity profile of

clofarabine, in pediatrics, for instance,

anthracyclines are tremendously concerning agent,

and we would eventually love to be able to

introduce combinations that, in fact, don't have

those anthracyclines in cardiac exposures.

118

So, that is the plan right now, at least

in the leukemia trials. I think I will stop there.

DR. MARTINO: Can I ask a question. I

appreciate the word randomization, and we all get

excited over it, but I have yet to hear what

exactly you are thinking to randomize, to what?

DR. ARCECI: Amen. In the Phase III

trial. So, depending upon the results of those

combination studies, what we would very likely do,

for instance, in AML, is to introduce a combination

clofarabine/ara-C, if that combination looks as

promising as we hope.

We would then try to randomize that

against a standard course of intensification or

induction therapy or consolidation therapy, so we

would try to have one arm, for instance. What we

are using now, for instance, is MRC-based therapy

with the modification now--

DR. SANTANA: The adult oncologists maybe

don't recognize the acronym.

DR. ARCECI: I am sorry, the Medical

Research Council. So, we are basing 5 courses of

119

intensive therapy in AML based upon the Medical

Research Council out of Britain. In the U.S., we

are randomizing the addition of gemtuzimab

ozogamicin to that upfront trial.

One group will get several courses of

combination and the other one will get the standard

therapy. In the setting of a combination with

clofarabine and ara-C, it would be particularly

advantageous to try to have a group that would get

that standard therapy versus an introduction or

replacement of an anthracycline, for instance,

containing regimen or another regimen that would be

potentially more toxic, and see if we could get the

same or better outcome. That is the type of

randomization we are currently pondering.

DR. PAZDUR: Could I comment? We are very

concerned about this aspect, and that is why I

spent some time in my introductory comments on the

need for confirmatory studies. I would like to

remind the ODAC Committee that in March of 2003, we

spent an entire I think day and a half or two days

on this issue.

120

Basically, the regulations, although it is

not a requirement, stipulates that there is an

expectation that these trials should be ongoing.

We have not met with the sponsor regarding this

issue. There may be differences on what

cooperative groups need to do versus what our

appropriate studies that would meet a requirement

for a regulatory purpose, i.e., isolating the

effectiveness of the proposed drug.

Here again, we would have to discuss this

in detail with the sponsor. It is quite bothersome

that we are at this point of talking about approval

of the drug, and not having met with the sponsor.

This is something that you will have to

take into consideration regarding this drug. Other

points that I would like to bring up for further

discussion perhaps at the next hour is whether this

confirmatory study needs exclusively to be done in

pediatrics or could be looked at in an adult

indication. Here again, that is something for the

pediatricians to discuss.

I am very concerned about this. The

121

Division is very concerned about the lack of

ongoing studies at this time. Remember, these

studies are to be done with due diligence. One has

to question, if they haven't been done to date or

even initiated to date at due diligence, has the

sponsor demonstrated due diligence if they haven't

been even initiated at this point.

That is a consideration that I think you

need to bring into consideration, whether children

are best served with the approval of this drug at

this time or whether further study needs to be

done.

DR. SANTANA: As a follow-up to that,

because I think you leave us with a sense that we

need to discuss this, but it is, from my own view,

is it a procedural issue that hasn't occurred, you

know, give us some of the reasons why it hasn't

occurred.

DR. PAZDUR: Usually, the sponsor would

come in and discuss these trials with us throughout

the course, and the drug has been in development

for a lengthy period of time here. It is not that

122

this drug was developed just within the past year.

One of the things that we stressed at the

previous meeting in March is that there should be a

comprehensive development program for drugs, and

this would include perhaps an exploratory study,

however, we encourage early initiations of

randomized studies, not late, not, well, let's get

the drug approved and then let's talk about

randomized studies.

I would like to emphasize that a lot of

sponsors have been very responsible with this after

our meeting of bringing forth a single-arm study

and discussing a single-arm study, but also making

a commitment and having ongoing accrual to a

randomized Phase III study.

An example of this, for example, is

Velcade, where we approved the drug on a single-arm

study, however, 40 or 50 percent of the patients

were already randomized to a study. This is the

type of drug development program that we are

looking at, a real commitment to drug program.

DR. SANTANA: Yes, go ahead.

123

DR. PERRY: I understand exactly I think

what you said except for me there is a difference

between adult and pediatric oncology. In the adult

world, we have lots of patients who aren't going on

cooperative group trials. In the pediatric

community, 90 percent of patients are on

cooperative group trials.

So, it seems to me the sponsor can almost

only work with the cooperative groups, and work at

their pace. They can't very well dictate to the COG

and say do my study and do it now, and do it in

this direction, because I need further approval

from the FDA.

In the adult world, it is an entirely

different playground.

DR. PAZDUR: And specifically, I think

this is why we wanted to discuss this issue here.

DR. SANTANA: I think that is a very

important point. I think there are limitations in

terms of patient resources that pediatric oncology

has, and when we get into Phase II trials that

require a larger number of patients or we get into

124

Phase III trials, you know, we have to work with

the existing structure of the cooperative group to

get those studies done.

So, I think your point is very well taken

and the Committee needs to recognize that. I am

not here to defend either side, I am trying to be

impartial, but I think it's a reality of how the

process occurs in pediatric oncology.

DR. PAZDUR: And here again, Victor, this

is why we wanted a public discussion of this.

DR. TEMPLE: I think the crucial point is

that accelerated approval was thought of originally

as something that happens if you get surprised,

that is, you are doing ordinary development,

whatever that requires, but the results are so

impressive in the early studies that you reach the

reasoned judgment that it is better to make it

available before you have the final data and get

those data later.

For obvious reasons, this isn't really to

blame anybody, it has also become, to some degree,

an alternative path to approval where the only

125

thing that people think about at least sometimes is

accelerated approval, and they don't really spend a

lot of time thinking about what the whole program

is going to look like.

We recognize the realities of availability

and things like that. Sometimes people go to other

countries to do the trials where there might be

possibilities that aren't available here.

But I think the main point that Rick wants

to make, and I do, is that it should be playing a

part of a coherent plan, and we would be happy to

talk with them, we would be happy to talk with the

pediatric oncology groups, but it ought to be part

of the thinking process, and we are concerned that

sometimes it doesn't seem to be.

DR. SANTANA: Joanne.

DR. KURTZBERG: I have a question for the

sponsor along these same lines, if accelerated

approval is granted, and if trials were planned as

part of the arrangement, would the sponsor's

support both drug and data collections costs for

those trials even after drug approval?

126

DR. WEITMAN: I would just like to make a

couple comments also. Again, that is, at least

historically, if you look at what the sponsor has

done with Campath in the past, is that we have met

I guess requirement for that drug as far as doing a

post-commitment study. That study is fully accrued

and is completed, and we are in the last phases of

really data collection follow-up on that, so we do

have a track record of meeting that requirement.

As far as going forward, we are fully

willing to support these studies going forward,

certainly with drug, but also financially as needed

to make these studies come to completion.

I think, Joanne, you and I have both been

in the position where drug all of a sudden

disappeared as we have been trying to do these

studies in the pediatric oncology community, and I

think as a sponsor, we fully accept that

responsibility to provide drug, as well as

financial support going further for these studies.

Now, when you go through the cooperative

groups--and I think the point that was made was

127

very appropriate--doing these studies outside the

cooperative group, I think is extremely difficult.

In fact, at this stage, one of the reasons

why it was done early, because a lot of these

patients just weren't eligible for any studies that

were in the cooperative groups going forward. A

lot of these patients now would be, particularly as

we move into less heavily pretreated patient

populations.

So, I think going into that population now

is really where we need to be with this drug. The

support for this agent going forward clearly

depends on I will say a little bit on the

Children's Oncology Group. When I have met with

them and talked to them about just that point, how

to support these studies going forward, there was a

couple of options.

One, they actually asked if we would

provided some support for particularly outside

studies, correlative science studies, and so forth,

but a lot of that support would come through the

CTEP mechanism, and they actually preferred that

128

versus sponsor support, but we are willing to

support it either way depending on what works best

for the cooperative group, as well as to get this

agent moving forward.

DR. SANTANA: Dr. Wayne.

DR. WAYNE: So, though, we are asked to

consider the data presented in the application on

the pediatric trials in regard to the accelerated

approval, it might help to have additional adult

data that might be extrapolated. I was somewhat

pleased and surprised to see on page 112 of your

briefing document, in Table 46, 60 percent, 63

percent CR with cycle 1 in adults on a Phase II

trial with single agent.

Are there data that can be shared further

in that regard? It is the IST study, and no

further mention, or I didn't find any further

mention of that in the briefing document.

DR. WEITMAN: We have had obvious a

continued program, but mostly with the focus on

adult AML. There has been a few ALL patients

treated, but most of it is in the adult population

129

with AML.

The responses again, as I mentioned, in

the combination study, were about 40 percent.

Again, this was pretty much echoed in the

single-agent study, as well, in adult patients with

AML with a fairly high response rate. In fact,

some of the patients, the elderly patients have had

response rates in the 60 percent range with adult

AML previously untreated patients.

That is why the program in adults is

really moving in that direction more than in ALL,

because of the activity we have seen in that

disease subtype.

DR. WAYNE: Do you have more formal data

in regards to that AML trial? Clearly, the very

limited data you have shown us in regard to AML in

pediatrics, and that single-agent agent might be

helpful in regards to extrapolation in AML.

DR. WEITMAN: Well, I don't have a slide

with me today on that, but clearly, again, that has

shown activity. The response rate in the

combination was 40 percent CR and CRp. The adult

130

elderly patients was over 60 percent response rate

with AML. Against, that was in an IST STUDY, that

wasn't a pilot study.

I think across the board, that is where we

have seen the response essentially confirmed right

between 40 and 60 percent with AML.

DR. SANTANA: Dr. Brawley.

DR. BRAWLEY: Steve, I have two questions

really, if you could respond. One of the things

that some of us are concerned about, those of us

who are a little bit more orthodox in our design of

clinical trials, like to set the endpoint and then

not move the endpoint during the trial. Sometimes

we can be accused of tailoring the endpoint to fit

our data.

Can you just respond and assure us that

that hasn't happened here?

DR. WEITMAN: No, absolutely not, Otis,

that hasn't happened here.

DR. BRAWLEY: The next thing, and I don't

want to make any allegations or implications

against anyone, but out of a sense of fairness, can

131

your academic advisors disclose for us their

financial relationships with the company, who own

stock, who has taken honoraria, and that sort of

thing?

DR. SANTANA: Johanna, where is the

Executive Secretary? Can you help us with that?

Do we have--before anybody gets to the podium--do

we have any information from the consultants for

the company in terms of their conflict of interest?

DR. BRAWLEY: Also, is it even appropriate

to ask that question?

DR. SANTANA: It is appropriate publicly

to ask that question, because the consultants

should have cleared that.

There is a question on the table whether

the consultants for the sponsor have cleared issues

of conflict of interest. What is the process?

MS. CLIFFORD: We don't do that.

DR. SANTANA: We don't do that, okay.

DR. PAZDUR: They don't do that.

Obviously, it is up to the individual people that

have presented here if they would like to avail

132

that information.

DR. SANTANA: I personally have no issue

with that. If other members of the Committee feel

differently, this is the time to raise your

concern. I have no issue with the sponsors and

their consultants.

DR. TEMPLE: Just to be clear, I mean our

presumption is that the people who are advocates

for the company have an interest. So, you have

listened to them, and you listen to what they say,

but they are not neutral parties like the advisors

here.

DR. SANTANA: Dr. Mortimer.

DR. MORTIMER: I am just curious. The

cytogenetics, I know you had cytogenetics on one

patient, but what was the impact of cytogenetics

pre- and post?

A second question is after completion of

therapy, did patients receive additional treatment

afterwards, or all these patients stopped with the

study drug?

DR. WEITMAN: Your first question about

133

cytogenetics, I was very interested in that myself

and tried to take a look at it. As you can

imagine, in these patients that have gone through

five, six prior regimens including transplant and

total body irradiation, their cytogenetics ended up

running onto several lines, and trying to really

tease out anything in particular was very difficult

to do.

Certainly, a number of these patients did

have 922, monosomy 7, and some of those patients, I

think there is one of each did respond. But again,

looking at them, trying to make any correlation

between cytogenetics proved extremely difficult in

these multiply relapsed patients.

Your second question?

DR. MORTIMER: Treatment after completion.

DR. WEITMAN: The only patients that

received essentially treatment after study were

those that relapsed and went on to something else,

and that was the same before going to transplant,

as well. The only treatment they received before

transplant was a conditioning regimen before

134

transplant.

If I can, I would like to just comment on

Dr. Wayne's question about other results in the

adult population. The Phase II study results,

again, there was 31 patients with AML in that

patient population. Thirteen of those 31 had a CR

and 4 had a CRp, for a total of 55 percent response

rate in that patient population.

There were patients with ALL. That was 12

patients with ALL. Again, these are recurrent

disease patients. Twelve patients had ALL with 1

CR and 1 CRp. Again, because of that difference

with the propensity towards AML's, is where we are

moving with the program in adults.

DR. SANTANA: Dr. Hershfeld, I will give

you the last comment.

DR. HERSHFELD: Well, I will briefly just

comment on the data that Dr. Weitman just stated,

and if I may, I would like to ask a question too.

The Pediatric Subcommittee of this

Committee has been a big proponent of extrapolation

and has previously had discussions about data

135

between adult and pediatric leukemia, and had some

recommendations in that regard, but I should note

that the data that were just cited were not

submitted to the FDA, nor reviewed by the FDA,

therefore, that should be taken into account.

I wanted to ask briefly, if I could, a

risk-benefit question. That is, in risk-benefit,

the risk and the benefit are more or less separated

and then one integrates the two.

I wanted to ask Dr. Weitman if there was a

gain or change in either the risk or the benefit

that was gained with experience with the drug, that

is, patients that sought early on maybe had a

different risk or a different benefit than later

on.

Somewhat related to that, and this relates

to Dr. Brawley's question about the endpoint that

wasn't in the protocol, the transplant goal, the

CR/CRp sum in the ALL patients was 20 percent, and

yet half of those went on to transplant. The

CR/CRp in the AML was 3 percent, and yet 34 percent

of the patients went on to transplant.

136

Why the differences or what was the

criteria?

DR. WEITMAN: With regards to your first

question about sort of gaining experience with the

drug, clearly, at least my impression of seeing the

patients going on study, when the study first

opened, we were seeing very heavily pretreated

patients going on study.

As you alluded to, with more experience

with the drug, as people began to see this drug's

activity, they began to put better patients on

study. But we also began to I think get a better

idea of how to manage some of the problems.

Quite frankly, we weren't anticipating the

extent of tumor lysis that we were seeing. These

patients would many times come on with white counts

of 100,000, and within a few days drop their white

count to 0.1, 0.2.

So, again, having considerable tumor lysis

that probably wasn't recognized was going to be a

problem early on, was managed much better as the

study went further, with extra fluids and just

137

realizing that this drug has potent activity in

some of these patients.

I believe one of the patients of Dr.

Arceci, that he treated, actually had a uric acid

that went up to like 17 or 18. In my pediatric

experience, I have not seen many patients with a

uric acid of 17 or 18. Again, I think that just

reflects the activity of this drug.

But as time went on, and we began to

realize that, I think we were able to improve that

risk-benefit ratio by again extra fluids,

monitoring, and so forth, just as we learned as we

went forward.

With regards to your other question about

AML and why there was a higher number of those

patients going to transplant, it is a good

question. I am not sure I honestly have an answer

for it. It is what we did see.

Clearly, I think a lot of patients with

AML from the very early point, tried to identify a

donor whenever possible, probably much earlier than

patients with ALL, but again, there was more

138

interest in particularly patients with AML going to

transplant.

There is also more willingness in that

group to take patients in partial remission, where

their ANC may not have fully recovered, to go to

transplant, and that is probably not the same

situation with patients with ALL, where they want

to see a complete remission, solid remission more

likely than not than in patients with AML.

So, I think the threshold particularly for

patients with AML, because those patients do do

reasonably well even if they PR going to transplant

as opposed to patients with ALL.

If I can ask Dr. Arceci and Dr. Sallan

maybe to step up for a second.

DR. SALLAN: I just wanted to step up to

address Dr. Brawley's question. I have no interest

in the company whatsoever other than getting paid

for my time to review the data and to be here.

DR. ARCECI: I, too, think it is an

important question to clarify. My involvement has

been to participate in the study and my interest in

139

drug development, and they paid for my

transportation from Baltimore to Silver Spring.

[Laughter.]

DR. SANTANA: We won't ask how you got

here.

I think we will conclude this part of the

session. We are going to reconvene exactly at

10:45, because we do have a couple of individuals

that have signed up for public hearing.

Thank you.

[Break.]

Open Public Hearing

DR. SANTANA: I have a public paragraph

that I need to read for the record, so let me go

ahead and do that.

Both the Food and Drug Administration,

FDA, and the public believe in a transparent

process for information gathering and

decisionmaking. To ensure such transparency at the

open public hearing session of this Advisory

Committee meeting, the FDA believes that it is

important to understand the context of an

140

individual's presentation.

For this reason, the 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, with its product, and,

if known, its direct competitors.

For example, this financial information

may include the sponsor's payment of your travel,

your lodging, or other expenses in connection with

your attendance at this meeting.

Likewise, the FDA encourages you at the

beginning of your statement to advise this

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, I think we have a number of public

speakers. You have a list? Go ahead.

MS. CLIFFORD: We will start with Hal

Wilson.

141

DR. WILSON: My name is Dr. Hal Wilson.

The only relationship I have with ILEX is they flew

me here and allowed me to stay in a hotel, which is

nice, because it's a long walk from Phoenix,

Arizona.

The reason why I am here, first, I have a

little background about me. I am a board-certified

family practice physician and I run and own a

practice in Phoenix called Maxel [ph] Medical

Group. I also do a lot of emergency room work with

a group called Emergency Physicians Professional

Association.

I had the opportunity to have dinner last

night with a number of parents of children who

suffered from the diseases that have been discussed

today. I am not such a parent. The reason why I

am here is because I am an uncle, and, Dr. Weitman,

when you had your slides up with the description of

the patient population, I felt like you had a

little picture of my niece on every one of them as

far as being in multiple drugs and failed bone

marrow transplantation, and the whole thing, ended

142

up in M.D. Anderson, and went on clofarabine.

So, that is why I am here. As I was

thinking about what to say today, I thought it

might be a good idea to approach this situation

with a sense of appreciation. I am not here to

crunch numbers or toxicities, risks and benefits.

I just want to say thank you to ILEX and

Dr. Weitman for bringing this drug as far as it has

come along so far. In this particular instance, I

realize it's a case of one that had some benefit.

Dr. Weitman and staff, I appreciate the commitment

you have made to pediatric oncology.

A couple of things that came up in the

discussion today, which I thought was really

interesting, which was fascinating to me, was the

idea that not very many new pediatric drugs have

come down the pipeline for a number of years.

I also heard it said that relapsed

leukemic patients and pediatric oncology patients,

regarding those patients, one of the major

challenges for a pediatric oncologist is the

relapsed patient, and I really appreciate what you

143

do as far as dealing with those situations,

because, as everybody knows, by the time the

patient makes it to you, they don't have a lot of

options left.

So, basically, in conclusion, I would just

like to say I remember at the beginning of the

presentation, there was a discussion about

approving a drug based on scientific data versus

providing drug access to patients. In this

particular patient population, I think it is

important to remember that really, they don't have

any other options left.

Thank you.

DR. SANTANA: Thank you for your comments.

MS. CLIFFORD: Colleen McCarthy.

MS. McCARTHY: First, I would like to

disclose that my travel arrangements and my air

fare were paid for by ILEX Oncology.

My name is Colleen McCarthy and I am a

clinical research nurse at Children's Hospital in

Los Angeles, and I work with the Leukemia and

Lymphoma Program.

144

For the past 12 years, I have been a

pediatric oncology nurse working In a variety of

settings, inpatient, outpatient, bone marrow

transplant, nurse education, and now clinical

research, but always my focus has been pediatric

oncology.

I want to tell you a story about one of

our patients that we took care of. In August of

2002, we enrolled our first patient onto the

clofarabine trial. In total, we treated 8 children

on this study. One little boy, I would like to

tell you a quick story about.

We enrolled a 3 1/2-year-old, ALL relapsed

patient, who was diagnosed when he was 1 1/2 years

old. He had had 4 prior regimens before coming

onto this treatment. He had had a bone marrow

transplant included in that.

When he first started his therapy with

clofarabine, he was very irritable, very clingy on

mom. Mom would try and put him down, he wouldn't

walk. Mom was very, very upset that he wouldn't

act normal.

145

We gave him his clofarabine and lo and

behold, a week later, walking, feeling great,

eating, interacting with his siblings, feeling

great. Mom was very, very excited, and most

importantly, he was able to be discharged from the

hospital.

Now, he had been in the hospital for two

months prior to clofarabine, dealing with fevers,

et cetera, from toxicities from prior therapies.

The fact that mom got to go home was so wonderful

for she and her family.

This young child ended up getting 3 more

cycles of therapy, 2 of which were outpatient, so

he was home for about 3 months before he came in

for his last cycle, had a fever, and stayed in for

a few weeks before going home.

The reason I want to bring up this little

boy is because the mom shared with me, when we

started this drug, that she really wanted this drug

to work for as long as it could because she wanted

time with her young son.

The father was not in the house at the

146

time. He is returning to the house about a month

after the child had relapsed. She wanted to make

sure that there would be enough time for this young

boy to be able to be with his dad prior to him

passing away.

He did get 2 months with his father, and

the last thing the mom told me right after the

child did pass away, was the fact that after

getting the clofarabine, after cycle 3, the family

went to Disneyland together, mom, dad, little 1

1/2-year-old sibling, older cousin, and grandma,

and they had a normal day.

For that family to have that one normal

day, she was so thankful. After the fact, when the

child did pass away, she called me and said thank

you for letting my child be part of that study.

So, that is my little story about our 3

1/2-year-old.

As a nurse taking care of patients with

leukemia, when you first meet a family when they

are newly diagnosed, they will ask you what is

next, what should we expect, what treatment, and

147

all these wonderful questions, and those questions,

as a nurse, you are able to answer when the child

is newly diagnosed.

When that same family comes back to you a

few years later with relapsed or refractory and

asks you those same questions, it is very difficult

to answer those questions, because you are not sure

what is next, or what to look forward to, or what

they can expect.

So, I just wanted to point out those quick

little things. Thank you.

DR. SANTANA: We appreciate your comments.

Thank you.

MS. CLIFFORD: The next speakers are Tasha

and Steven Virostek.

MS. VIROSTEK: I am Tasha Virostek and I

do not have a financial relationship with ILEX.

They did, however, pay for dinner last night.

My son Jeffrey lost his battle on

September 25th, 2003, to acute myelogenous

leukemia. I am not here as a scientist, I am not

here as a doctor, but I am here as a mother who has

148

been deeply affected by this terrible disease.

Jeffrey was diagnosed with AML at the age

of 2 1/2. At the time, we knew we were dealing

with a terminal illness, but we had hope that with

the best doctors, the best medicines that were

available, and the best facilities for him, we

would conquer this disease.

Jeffrey received his induction of

chemotherapy beginning in November of 2001. When

he did not receive a remission, a protocol was

abandoned, and a new regimen was sought. What I

found out today is that many of these drugs are 30,

40 years old.

Christmas of 2001, he spent in the

hospital fighting infections, affected by rigors

from medicines given to wipe out fungal infections,

bacterial infections, and viral infections. His

body was too weak to fight infections and also too

weak to combat the side effects of the drugs. It

was depressing and frightening.

Miraculously, he did pull through and he

came home to celebrate his third birthday. His

149

next round of chemotherapy also left him weak and

susceptible to infection.

In March of 2002, we began treatment which

led us to our first bone marrow transplant. This

was our only option, and luckily, we did have a

donor, his sister who was 5 at the time. She was a

perfect match.

This was an extremely difficult process on

Jeffrey because he received even more toxic

chemotherapies which compromised his immune system

and again left him vulnerable to infection. The

chemo made his appetite diminish, prompting the

need for fluids and nutrition via his double lumen

port. Mucositis became prevalent. The intense

pain that accompanied this was unbearable at times

and had to be managed by morphine, which, in turn,

caused him to be lethargic.

This treatment kept Jeffrey from doing

what a 3-year-old should be doing, playing,

learning, laughing, enjoying being a child. As a

parent, it was agonizing to watch your child

struggle with the vary basic functions of life. No

150

one should have to endure so much.

This process also brought many anxieties.

We worried about the short- and long-term effects

of the chemotherapies on his body, liver damage,

damage of the heart and the lungs. We worried

about his future physical and psychological

development, and we worried about the effect on his

spirit. We worried about graft versus host

disease.

Jeffrey's diagnosis with cancer and his

treatments had other consequences on our family.

Being his full-time caregiver, I had to rely on my

extended family to take care of my daughters that

were at home. They were too young to understand

and my absence caused resentment and anxiety

amongst our young children.

Jeffrey's diagnosis brought a lot to our

family, but luckily, when he came home we were able

to have 9 months of remission. During that time,

our lives began to resemble, quote, unquote, a

"normal" life, and in October of 2002, Jeffrey

experienced a Make a Wish trip to Sea World to meet

151

Shamu.

But around his 4th birthday, he relapsed.

We returned to Children's Hospital where we faced

limited choices. We could redo the bone marrow

transplant or we could do nothing. We decided to

fight the cancer. Once again, we spent weeks upon

weeks in the hospital, confined to our small room.

Visits from family and friends were prohibited.

The second bone marrow transplant was

especially difficult on my daughter Megan. With

much trepidation, she received daily shots to

increase the number of stem cells in her peripheral

blood. Placing the catheter in her artery was a

painful experience that she still remembers today.

She then endured a long and difficult extraction

process.

Despite using the best doctors and

treatments available, Jeffrey's cancer returned

just a few weeks after his treatment. Our options

were limited. We had exhausted all known

treatments and he was not eligible for any drug

trials at this time. So, our last-ditch effort was

152

to give Jeffrey additional stem cells combined with

interleukin-2 in the hopes that the good cells

would outnumber the cancer cells.

At this time, our main goal was to

maintain Jeffrey's quality of life. We wanted him

to remain with us at home, so that he could play

with his sisters, he could be a little boy and play

with his friends and visit public places. We

wanted to enjoy each day that we were together and

we were frustrated by the lack of treatment

options.

Jeffrey did experience some pleasures like

a trip to the beach before the cancer took over.

His last weeks with us, however, were filled with

intense pain and suffering. The cancer took his

life.

We know that Jeffrey's story is not an

isolated case. Since the time of his death, our

family has personally known more than half a dozen

individuals who have relapsed and many have since

died.

The professionals in this room have the

153

knowledge, the resources, and the influence to make

a difference in the lives of children with cancer.

I encourage you to conduct your work with the

utmost diligence and expedience to bring new

treatments which will combat leukemia more

effectively, with fewer side effects, that allows a

higher quality of life. So many lives are

depending on you.

DR. SANTANA: Thank you so much for your

comments. I know it is difficult for you to talk

today, but I think it is a very honorable way to

remember your son to share your story with us.

Thank you.

MS. CLIFFORD: The next speaker is Ms.

Nadia Hendry.

MS. MAROUN-HENDRY: Good morning. Thank

you for this opportunity to share my experience

with you.

As I was listening earlier, I want to

clarify that Dr. Steinherz can confirm we have had

many bone marrow follow-ups, and thank you to the

doctors who quickly do that for us parents who are

154

waiting anxiously and chase them down the hall to

get those good and not so good results.

I consider it an honor, not only to speak

on behalf of my son, but hopefully, my remarks will

represent countless children and their families

whose lives, hopes, and future have been derailed

by the horrors of childhood cancer.

Throughout my son Matthew's 4-year ordeal,

struggling with the ravages of leukemia, I came to

see myself as something of an expert on the topic,

but please, make no mistake, I am here only as a

mother, more specifically, I am here as a

heartbroken mother, having lost our beloved son on

March 25th, 2003.

When Matthew was only 3 years old, filled

with wonder and the joy of life, we were informed

that he had cancer, telling you on that day,

September 9th, 1999, our world stopped and lost its

sunlight does not begin to express how profound our

grief was from that day on.

But as so many other parents have known

before me, we had to take our grief and turn it

155

into hope and action. We had to become educated.

Despite having been plummeted into a world of

protocols, side effects, frequent hospitalizations

of over 500 days in the hospital, new

relationships, financial and family turmoil,

somehow with the grace of God, supportive family

and friends, excellent medical care, and faith in

the promise of medical research, we managed never

to lose hope until Matthew drew his last breath.

During his illness, Matthew rose to the

challenge and rigors of numerous protocols and

treatment, including a bone marrow transplant and

subsequent boost. Each course of treatment renewed

our hope that this would be the last.

From the point of diagnosing to his

passing 3 1/2 years later, he only was free of any

drugs or treatment for two months of his life.

Matthew relapsed two months after his first bone

marrow transplant. Prior to his relapsing, he was

diagnosed with Epstein Barr Virus, a possible side

effect that we were aware of to transplant.

After 9 months of treatment at Memorial

156

Sloan-Kettering, preceded by 2 years at Albany

Medical Center, we were devastated with the lack of

options. There was nothing, it was dismal.

Matthew's doctor, Dr. Peter Steinherz,

offered to include him in a clinical trial using

clofarabine. We were given all the necessary

information and eagerly grasped at this one last

opportunity to save Matthew's life.

Remission followed shortly thereafter the

clofarabine was starting, allowing us to go ahead

with the second transplant. This was our last

chance. He stayed in remission and was 100 percent

donor cells, but oddly, we couldn't figure out what

was going on. He was having seizures and numerous

lesions to the brain.

As the seizures became more numerous and

more severe, a decision to treat for a possible

fungal infection was made. After weeks of

treatment with no improvement, it became obvious

that this was not a fungal infection.

By this time, Matthew's speech was

impaired, his thought processes sluggish, he was

157

unable to walk. The severity of his weakened

condition was cause for several hospitalizations

and increased medication.

Matthew's little body was broken and

weary, but never did he lose his spirit. He had a

strong sense of self. Our darling little boy, this

dear old soul told me it was time to face my fears,

and he asked me that I promise there would be no

more hospital.

After years of hope, we had to now turn

our love towards acceptance and making his last

months comfortable and peaceful. We were faced

with what no parent should ever have to deal with,

arranging for the end of our son's life.

It was through his strength of spirit that

we requested an autopsy to determine the cause of

death since he had been on the clinical trial. A

month after Matthew's passing, we learned that he

was clear of leukemia upon time of death, and it

was the Epstein Barr virus that caused his death.

Therefore, we knew the clofarabine has worked.

We had to know, we had to know for

158

ourselves, and we had to know for the other

Matthews that this treatment had been effective.

We were happy to know that the clofarabine had been

effective and could only wonder, with very heavy

hearts, what if he had gotten this sooner.

Matthew only had 6 years on this earth. I

will have the rest of my life to wonder what if.

I failed to mention my trip was paid by

ILEX and dinner and my room, and I thank them for

giving us the opportunity to have an extra 8 months

with my beautiful son. Without that, we would not

have had that time. We were living in New York

City for over 9 months. Silly me, I thought the

chance that we would be home in 2 months and

everything would fine, and 9 months later, all

Matthew wanted to do was go home and be with his

dog and his brother and the rest of his family, and

he was able to do that because he took part in

this. It gave us time, and I don't think time can

be measured, and it is priceless.

Thank you.

DR. SANTANA: Thank you for sharing your

159

story with us.

MS. CLIFFORD: The next speaker is Jan

Manlapaz.

MR. MANLAPAZ: Good morning, ladies and

gentlemen. I am Jan Manlapaz. I come here as a

parent. We originally came from the Philippines.

I have a son, who is right now on my left side, who

is an AML patient.

He got multiple relapses, 5 relapses, 1

bone marrow transplant relapse, and I am speaking

on a good tune because this is our bible right now,

got multiple relapse and one bone marrow

transplant, and on July of 2002, the doctors

advised us that he going to live only for around

three months, and with the decision and advice of

Dr. Steinherz, they introduced clofarabine, and

they use it for him. He is now in 22 months in

remission, and we would like to thank you, thank

ILEX for this medicine, and thank all the people

who are in this research.

I want to give the microphone to my son

just for a simple note to say thank you.

160

Thank you for all the doctors for making

me well and thank you, Dr. Steinherz, for giving me

some medicine. Good morning.

[Applause.]

MR. MANLAPAZ: Thank you very much.

DR. SANTANA: I don't think anybody can

beat that comment. So, next.

MS. CLIFFORD: Mr. Dahlman.

MR. DAHLMAN: Thank you. George Dahlman.

I have no interest in ILEX. I didn't go

to dinner last night, and I live in the Washington

area.

My name is George Dahlman. I am the Vice

President of Public Policy for the Leukemia and

Lymphoma Society of the nation's second largest

voluntary cancer organization and the world's

largest dedicated to blood cancers, but probably

more importantly, I am the father of a childhood

cancer survivor, or now, more accurately, an

obnoxious teenager.

In the first capacity, I represent a lot

of patients and family members, and we lobby for

161

government funding, we fund our own research. We

run patient services programs, and we lobby for

better care.

Now, we all know that the leukemia

survival rate has improved dramatically over the

last 30 years. It has gone completely upside-down.

It used to be that survival rates were more like 20

percent, and now they have flipped around, they are

more like 80 percent.

Most of the credit for that dramatic

change is really with the physicians, many of you

pediatric oncologists who have tweaked and changed

the protocols and had the vision and creativity to

really make those differences.

Now, even with that progress, though,

there are no real cures yet as long as there are

kids that still have to go through this, and that

is the cold fact that parents have to face and many

parents have to tragically experience like those

here today.

There is still too many kids who don't

make it, and as a parent who had to look at that

162

prospect, I can tell you there is nothing more

frightening than the thought of losing your own

child. I have known a lot of people like this that

have gone through that.

I would like to say in the memory of those

children and for their parents, and all those

patients yet to come, we still have to keep

tweaking the protocols. That is why I think this

hearing and this drug is such an important

milestone.

That is what clofarabine offers. As we

all know, this is the first one initially labelled

for pediatric leukemia in over a decade, and that

is really a tragedy. It is a deficiency that

should not be allowed to continue.

I am proud to say that my organization and

others that we work with are very actively involved

with trying to develop incentives, public policy

incentives and corporate incentives that will

address that issue.

But now, as we all know, the progress in

this is incremental, and one in which the FDA has

163

traditionally accommodated with even the most

modest improvements in outcome, and professionals,

like yourself, need every weapon in your arsenal to

make those incremental improvements, and

clofarabine offers that kind of incremental

improvement.

Thanks for your attention.

DR. SANTANA: Thank you.

Committee Discussion

We will go ahead and start our questions

and further discussions, but before we do that, as

I heard the previous discussion, I think there were

two central themes that I think would be important

to try to reach some resolution or at least some

further discussion before we go into the questions,

because I think they will be pertinent for

answering those questions in the best way that we

can.

One has to do with this issue of bone

marrow transplantation and how does that impact

both of these studies and the indication that the

sponsor is requesting.

164

To me, what I would like some discussion,

and I am going to ask Joanne Kurtzberg, who is a

bone marrow transplant specialist, to help us with

this issue, is to give us a sense of what is the

practice of bone marrow transplantations with the

pediatric patients that have refractory or

recurrent leukemias, in what settings are

transplants clearly indicated, and what additional

value does treatment prior to transplantation offer

to these patient populations.

I think that will be very pertinent as we

go to trying to establish this issue of clinical

benefit for this drug.

Joanne, can you tackle that one for me?

DR. KURTZBERG: Yes, sir. Well, first of

all, we have to separate the discussion into a

discussion about ALL and AML, because the answer is

really not the same.

In children with ALL, transplant is really

not effective unless they are in remission at the

time of transplant, not partial remission, not

remission with low platelet, real remission, and in

165

all the diseases that we transplant, the success

rates for curative therapy are still the lowest in

children with ALL, because the disease appears to

be more resistant and maybe there is less of a

graft versus leukemia effect.

So, this drug would be helpful or any drug

would be helpful if it could take a child with

resistant disease and put them into a true

remission, but the idea of putting them into a

partial remission, I don't think is operative here,

because results in children in partial remission,

95+ percent of children relapse, if not all.

One thing that was said earlier that I

need to take issue with is that you can certainly

have a child with ALL walk in the door, have normal

to near normal blood count, and be in relapse, and

you cannot tell from the peripheral blood count

necessarily that that child is in remission or

relapse.

As a transplanter, we get many children

referred to us who relapse between the time they

left home and the time they arrive at the

166

transplant center, and when you do that marrow,

they have got 20, 25, 30 percent blasts, and you

don't know that from their peripheral counts or

their physical exam or any clinical parameter.

As much as you want to cure every child

that walks in your door, if you take that child to

transplant, you are not going to cure that child.

So, I think rigor in that setting is really

important, and a drug would be valuable in that

setting if it produces a true, durable remission.

In AML, the story is different. Children

with AML, transplanted in relapse, still have

anywhere between a 25 and 40 percent cure rate

depending on the nature of their disease, the

nature of the transplant, the type of donor, and

the type of prep regimen.

So, you could argue in AML that it is not

essential at all to get a child in remission.

Having a drug that cytoreduces may be valuable to

prevent some toxicity, but I am not sure it really

contributes in a heavily pretreated, multiply

relapsed patient to overall outcome.

167

I think it would be an interesting

question to study, but I don't see anything in this

data that helps us know if it really provided an

increased cure post-transplant compared to a child

that was transplanted in frank relapse.

DR. SANTANA: Just to summarize very

straightforward like I like to be, in the setting

of relapsed, refractory ALL, there may be some

added value of getting those patients into

remission prior to transplantation, because those

are the patients that ultimately we think do well

with bone marrow transplant, but in the setting of

a lot of disease in the ALL setting, further

transplantation doesn't really add anything.

Am I correct in that simplistic view?

DR. KURTZBERG: I think so. Patients with

ALL benefit from transplant if they are in a true

remission at the time of transplant.

DR. SANTANA: And then for the AML

patients, it is debatable whether cytoreduction

really improves their ultimate outcome with

subsequent therapy as aggressive as

168

transplantation, is that correct?

DR. KURTZBERG: Yes.

DR. SANTANA: Comment about that? Dr.

Hussain.

DR. HUSSAIN: Just to ask the doctor

again, because I think I heard a little bit

different, and that is, it is a complete remission

that is what is important, not any remission, and

the second question is, is that a confirmation of

that remission is crucial because you could be in a

complete remission today and walk out of the door,

and that remission is not maintained, which I think

is central to what we are talking about here.

DR. KURTZBERG: In ALL, yes--

DR. HUSSAIN: That is my question.

DR. KURTZBERG: --remission is important,

and a child who is transplanted not in remission

has an overwhelming probability of relapse, and

generally does not benefit from the transplant,

and, yes, confirmation is important.

DR. HUSSAIN: I am sorry, but we are

talking about a complete remission or any

169

remission?

DR. KURTZBERG: Complete remission.

DR. HUSSAIN: Complete remission.

DR. KURTZBERG: As defined in 1970.

DR. HUSSAIN: So, complete remission and a

confirmed complete remission is crucial.

DR. KURTZBERG: Yes.

DR. RODRIGUEZ: In AML, do I understand

you to say that cytoreduction isn't of any value in

terms of preparing the child for transplant?

DR. KURTZBERG: It can be of value

particularly if the child has a very high tumor

burden, because it can minimize toxicity, but to be

completely honest, there are many ways that you can

do that without achieving a remission. You can use

hydroxyurea and BP-16. You can use ara-C, which

may not put the child in remission, but will have

some lytic effect on the blasts.

So, you know, I don't think this would be

the only option in that setting.

DR. SANTANA: Any other questions on that

point?

170

DR. MARTINO: So, it occurs to me that

perhaps there are two patient populations. There

are patients for whom a transplant is a possibility

in the sense that there is someone who can donate

the marrow, and then there are patients for whom

that appears to be not such a possibility, and the

goals that are striking as being somewhat different

if I had to design a trial, and one of the problems

that I keep struggling with in the data that has

been presented to us is were the studies designed

in a manner that allowed me to answer the questions

that they are suggesting I am able to answer.

I am really struggling with the fact that

to me, there really are two populations, there are

two endpoint potentials here. One is judging is

this good therapy prior to, or concurrent with, or

somehow related to who gets to go to transplant,

and for whom is that a valuable thing to do.

Then, I have patients who really aren't

people going to transplant where I might want to

ask a somewhat different question. Yes, ma'am.

DR. KURTZBERG: I agree with you, but I

171

think the strategy for any new drug for ALL

shouldn't really have transplant as an endpoint.

I mean I think transplant came as an

endpoint because you are dealing with desperate

situations where that is the only possible cure,

and that is kind of what everybody looks to when

they are in that situation, but I think you are

absolutely right, the drug ought to be tested in

first relapse or second relapse patients, and there

are many of those unfortunately, with an endpoint

of response, and it is an additional facet, with

additional therapy planned after that based on the

standard of care. It might be transplant, it might

be more aggressive chemotherapy depending on

duration of first remission and many other things.

But the transplant endpoint really

confuses the whole situation, and I think it just

came to play because of the clinical situation

people are in when their child relapses multiple

times, not because that would be an appropriate

endpoint for proving a drug is good for leukemia.

DR. MARTINO: But what I feel I need to

172

judge here is whether the data, as presented,

allows me to answer the question that the study

states it was designed to answer, and I still

remain confused because of this transplant issue,

realizing its clinical value and applicability, but

it doesn't quite allow me to judge what I view as

the question posed in the study.

DR. SANTANA: Dr. Temple.

DR. TEMPLE: If I understood you, you are

saying that at least in ALL, the question is

complete response rate properly documented, and

some people get transplanted, some won't, but you

are saying that is not the most important thing, it

is the complete response rate, which here is in the

neighborhood of 12 percent or something like that,

subject to subsequent debate.

But in AML, that is not as clear from the

sound of what you said. That is not as clear

because getting the count down, which you guys say

could be done in a lot of ways may be of benefit,

so what is the right test there?

DR. KURTZBERG: I think honestly, if I

173

were doing this, transplant wouldn't be in the

discussion, and I would do a standard Phase II

trial where the endpoint was response. I think

that the drug probably does have activity, I am not

questioning that, but I think testing it in first

relapses in upfront window, looking at response at

a month, you know, and squaring that, clearly is a

good thing to do.

DR. TEMPLE: So, in those cases, people

might be willing to wait before they--I mean in the

AML cases here, they were already transplanted

before you got a chance to look at how long the

response lasted, so you never could find out. Do

you think in earlier disease, that sense of urgency

would be less, and you could actually get that

data?

DR. KURTZBERG: Yes, I do, and I also

think you could get better data because the

patients are generally less resistant at that

point.

DR. TEMPLE: Okay.

DR. KURTZBERG: I think the response

174

post-transplant is not interpretable. There are so

many kinds of transplants, kinds of prep regimens,

kinds of donors, and covariants that are

confounding, that you can't make a--

DR. TEMPLE: So, specifically, for these,

there are at least a reasonable number of people

with ALL whose response was tracked without a

transplant intervening, so you got some idea of how

long it lasted.

In AML, there were very few who managed to

do that, because the all got transplanted pretty

early if they looked even reasonably good, so you

don't really have duration data, is that correct?

DR. KURTZBERG: Yes, and part of that is

the hidden message that most people don't

transplant children with ALL if they are not in

remission, and these patients weren't in remission,

so not so many of them were transplanted.

DR. TEMPLE: Right. They behave just like

you say they would.

DR. KURTZBERG: But in AML, people do

transplant in relapse.

175

DR. TEMPLE: So, just to be sure I

understand it-- and I am sorry to struggle, I don't

do this for a living--in ALL, in these relatively

advanced and refractory patients, in ALL, you do

get a chance to see, in a small population, what

the complete response rate is, and it is not

confounded by early transplant that messes up

finding out how long it lasts. In AML, it is more

of a problem here because you don't.

Would that be true, is that what you are

saying?

DR. KURTZBERG: The way the data was

presented--

DR. TEMPLE: I am asking you what you are

saying.

DR. KURTZBERG: I think you could, but I

am not sure it happened in this case.

DR. TEMPLE: No, I mean as conducted here.

DR. KURTZBERG: Yes.

DR. SANTANA: Dr. Bukowski, do you have a

question or a comment?

DR. BUKOWSKI: Just a clarification and

176

maybe you can comment on this. Essentially, what

you are saying then is the population in second

relapse and beyond in both diseases is not an

appropriate one to evaluate a new drug in when

using the endpoint of complete response, because of

the confounder of transplant, is that correct?

DR. KURTZBERG: No. I think because in

many situations, you will take a patient in first

relapse to transplant. So, it is not the

transplant--I think transplant is confusing the

whole story. I think that drugs will have a better

chance of being fairly tested in patients in first

relapse because they are less resistant and less

heavily pretreated, and you are more likely to see

activity.

But it could be tested in second, third

relapse, but the endpoint should be response, not

whether they go to transplant and how they survive

after transplant.

DR. BUKOWSKI: So, is it possible, in the

pediatric group, to hold the transplant, to look

for the complete response? I mean it sounds like

177

this wasn't--at least it wasn't possible in this

particular setting.

DR. KURTZBERG: No, I mean complete

response after a first course or maybe two courses

at the most, and, yes, that is possible. First of

all, you wouldn't take the patient with ALL to

transplant without a complete response, and the AML

patient, it is a clinical judgment, but if they to

on a trial like this, they have a month or two to

be assessed.

DR. SANTANA: Dr. Maldonado, did you have

a comment?

DR. MALDONADO: It is not related to

transplant, it is actually a question that I have

been struggling with.

DR. SANTANA: Could you hold onto it then,

because I want to finish this discussion as one of

the points that I wanted to get clarity on.

DR. MALDONADO: Okay.

DR. SANTANA: Did you have a comment on

it, Dr. Poplack, on this issue we are discussing

right now, before we go to the next one?

178

DR. POPLACK: Yes, it's a question. I

think, obviously, one of the scenarios that is

being put forth here is that the value of this

agent is that it can get patients to transplant by

putting them into remission.

So, I would like to ask Joanne perhaps to

educate the group in terms of from her experience,

what percentage of patients who come to transplant

don't get into remission and don't get to

transplant, and what does she see as a potential

for this type of agent in that circumstance.

DR. KURTZBERG: That is a complicated

question because it really varies based on the

practice of the treating oncologist and their bias

for or against transplant and when they refer.

I am a transplanter, so my bias is that

any child who relapses should be referred for

transplant when they achieve their second

remission, but many oncologists don't practice that

way, and they believe a child should go through

multiple relapses before they prove they need to

take the risk that is associated with transplant.

179

So, that confounds the answer to your question.

I think, you know, a child with ALL on

standard therapy who relapses has about a 70

percent chance of achieving a second remission with

standard therapy, and if that child didn't achieve

a remission with standard therapy, and took a new

drug and had a 30 percent chance of achieving a CR,

that would be, to me, a valuable activity in that

drug.

If you take a child who has had 5 relapses

and expose them to any new drug, the chances they

are going to achieve a remission, no matter how

active the drug is, is much, much less. I don't

know how to put a number on it, but that is a

clinical practice issue, not whether the drug is

active issue.

For AML, the chances of achieving a second

remission after relapse, they are about 50, 60

percent with standard therapy, but again it depends

on therapy relapse, off therapy relapse, how close

to recent therapy, so there are a lot of

confounding variables.

180

My point is that I think this has been too

ambitious in taking down the transplant road is

just too complicated, and what you really want to

know, in childhood leukemia, is do you get a

durable response, not that lasts, to me, it doesn't

matter as a transplanter, if it would last two

months, six months, 12 months. It matters to me

the child is really in remission and that you have

the time to get them to transplant, which requires

another two to four weeks.

DR. SANTANA: So, you would define the

clinical benefit for the ALL population as reaching

that goal?

DR. KURTZBERG: Right.

DR. SANTANA: Of getting a remission of

some reasonable duration to allow you then to get

something more definitive.

DR. KURTZBERG: Right.

DR. SANTANA: You would define that as a

clinical benefit.

DR. KURTZBERG: Right.

DR. SANTANA: Ms. Hoffman, and then Dr.

181

Perry.

MS. HOFFMAN: Can you clarify, for the AML

population, I guess the benefit of allo- versus

auto-transplant, and if there is clinical benefit

to taking these AML patients using an

auto-transplant?

DR. KURTZBERG: This is my opinion, but

there is very little benefit to auto-transplant at

all, and if you have a multiply relapsed patient,

there is even less benefit. Not only do you have a

higher risk of relapse, which is 80, 90 percent,

but you also have a much higher risk of secondary

malignancies. Long term, the results in those

patients using auto cells are terrible.

So, I think allo-transplant is where you

need to go in all of these leukemias.

DR. SANTANA: Dr. Perry.

DR. PERRY: I think we have heard from the

patients' description and the parents' descriptions

that these are desperate people, and when any drug,

whether it is this drug or another, produces some

benefit, they see the transplant as the next

182

opportunity for a cure. They are not so naive as

to think that this drug is the miracle drug that is

going to get them disease-free forever.

So, when they see the opportunity for a

transplant, whether the odds are great or small,

they are going to leap upon that, and present

company excluded, most transplanters basically are

hammers, and they see the world as a nail, and if

there is a patient, there is a indication.

DR. KURTZBERG: That is a new one for me.

DR. PERRY: I don't think we can fault the

company for the actions of the patients and the

treating physicians who see a response and then

say, great, now is our chance to leap onto a

transplant, which might be curative. I don't think

the company, in this circumstance, can control

that.

So, I don't think the company designed the

trial to say this drug is a bridge to transplant.

They designed the trial to see does this drug work

and to what degree, and I think to some degree it

works. The question is does it work enough.

183

DR. SANTANA: Dr. Maldonado.

DR. MALDONADO: Now, it is about

transplant. The question is, I mean transplant was

not an endpoint for these trials, and by hearing

the experts, I believe that there may be different

standards for transplant. It doesn't appear to be

a single standard. This is a multiple center

trial, so there may have been a standards that

occurred in the trial.

It is just like hospitalization. I mean

hospitalization is another endpoint that should not

be used as endpoint unless there is criteria for

that, and I haven't seen that criteria outline in

this trial of when a transplant occurs.

More than a transplant, it should be a set

of criteria that the patient meets, and in that

case, it will be complete response. But was that

transplant an endpoint or not?

DR. SANTANA: No, no, that is the

confounding issue, that transplant was never part

of these trials, but what happened in real practice

is, like you have heard comments around the table

184

and from different experts, is that this drug in a

way was used as a bridge for those patients that,

at least in the ALL population that were having a

response and made it onto transplant, and for the

AMLs, everybody decided which patients got

transplant, which ones didn't.

So, it was not a study endpoint and we

should not fault the sponsor for that, because they

didn't have any control over that. That was just

the practice of patients and physicians who

participated in the trial.

DR. MALDONADO: But at the same time, if

that is the reality, why the studies were not

designed with that in mind and set up that as an

endpoint, so measure it in some way?

DR. SANTANA: I can't answer that

obviously, that is a historical issue of why the

studies were not designed that day, but the data we

have is the data that we have.

I want to put this discussion to an end

because we have got to move. I just raised the

issue because to me, if this agent gets approved

185

under the accelerated approval, we have to discuss

the issue of the complete responses and the

durability of those, and then the additional

clinical benefit that those responses provide.

So, I wanted to have the discussion, so

that at least we get some sense from the ALL

population and from the AML population, the

differences in those depending how the discussion

of the questions goes on further.

I am going to stop the discussion here for

that. One more point that I want to discuss before

going to the questions, which is very relevant, and

I am going to ask Dr. Poplack and Dr. Hirschfeld to

comment on, and it is an issue that was raised

earlier by Dr. Pazdur in terms of pediatric drug

development and the timing of studies and how what

we are discussing with this agent today is relevant

to that bigger picture, because I think we are

going to have to address that in the context of any

new drug that is developed for kids that comes to

this Committee.

So, David, can you give me some insight

186

into that issue, and then I will ask Steve to

comment, too?

DR. POPLACK: I think it is pretty obvious

to all that this is the worst situation, the worst

scenario in which to try and do a study of any new

agent. I think that Joanne put it quite well that

one would much prefer to do a study on a new agent

like this at an earlier stage in the clinical

courses of the patients.

That being said, it is not so easy. The

whole concept of therapeutic windows is a whole

theme unto itself in terms of the pros and cons of

that, are the patients really available given the

practice, et cetera.

It is a real conundrum and it requires

superb cooperation frankly, between the group, the

Children's Oncology Group, the sponsors, the FDA,

to make these types of things happen. I think this

isn't the best situation, there is no question

about it.

DR. SANTANA: Steve, would you follow up

on that, please.

187

DR. HERSHFELD: Yes. I will, as I have

for most of my life, agreed with Dr. Poplack in

that it is a multi-dimensional issue and that it

requires coordination and integration of all the

involved parties due to the relative rarity of the

diseases and the lack of resources.

I would just note that when these two,

single-arm studies were developed and they were

designed with response rate as the endpoints, much

as Dr. Kurtzberg had recommended, it was during the

time of transmission of when the Pediatric Oncology

Group and the Children's Cancer Group were merging

into the Children's Oncology Group, and there was

issues of the availability of protocols and

logistical issues about setting up collaborations,

so the sponsor, in essence, wanting to develop a

drug in that time frame, during that historical

period, had relatively few options in terms of what

would be feasible.

DR. SANTANA: Dr. Temple.

DR. TEMPLE: I want to beg the Chair's

indulgence because I thought I understood what Dr.

188

Kurtzberg was saying, but the questions that come

up make me uncertain whether I did. Just one more

time.

My understanding was that you think a

complete response, let's ignore whether these all

were, and whether the rate was high enough for the

moment, is meaningful actually whether or not

people decide to do a transplant--this is in

ALL--whether or not people decide to do a

transplant, so that the transplant situation in

some sense here is not relevant if you once

documented a complete response however you have to

do that, for ALL.

That makes ALL sound simpler to me in this

case, and the only issues would be whether a 10

percent or 12 percent response rate is good enough

and whether the responses were adequately

documented. Those are the usual kinds of questions

we face.

So, is that correct? So, it doesn't

really matter that they didn't specific transplant

or they were confounded by the community behavior

189

or anything in ALL. Is that true?

DR. KURTZBERG: Yes, it's true, and I

would say the same thing for AML, and what I am

saying is that I don't think this duration of

response question for pediatric cancer is important

in this kind of population.

DR. TEMPLE: The AML, I guess I thought

many of the documentation of the response was

truncated by the transplant, so that is perhaps a

separate problem in AML.

DR. KURTZBERG: But there still was an

initial response.

DR. TEMPLE: All right. Fine. Thank you.

Let's go ahead and deal with the

questions.

I am just going to make an introductory

reading of the first paragraphs and then I think we

have the summary page at the end.

DR. MALDONADO: Dr. Santana, I just have a

question for clarity, not related to the questions.

DR. SANTANA: Go ahead.

DR. MALDONADO: I have been hearing how

190

difficult these studies are to define or to be

done. However, when I saw the presentation, I

believe Dr. Arceci or somebody on behalf of the

sponsor.

We are talking about the third most common

cancer in patients and looking at the numbers, the

numbers are not small, so where is the difficulty

other than the COG might not be acting fast enough.

I am just not understanding what the difficulty of

the feasibility of these studies may be. I am

talking about the confirmatory studies.

DR. SANTANA: Part of it is study

availability of when the Phase I or Phase II

studies are really available. I mean you heard an

example earlier today of a comment from COG how

they only have three, Phase I studies open right

now, and that studies close and open constantly,

you know, depending on the number of patients that

are available, so there is a lot more patients out

there than there are studies.

Now, the issue of the studies and how well

designed the studies are and which centers they

191

occur, that is a whole separate discussion. I can

tell you that I think the COG and CTEP and other

major institutions have been advocates to getting

studies open, so that these patients can

participate in, so independent of the law we had

during the cooperative group merger, I think that

is an impetus from the Pediatric Oncology Committee

to do studies and to get these patients on studies.

Ultimately, the decision is made about the

parents, whether they participate or not, so you

may always have 500 patients, but parents make the

final decision whether they want to participate or

not.

DR. MALDONADO: The reason I ask that is

because if the drug becomes available, then, it

will be even more difficult, because then the

sponsor will need to compete with its own drug

being available, so patients will tend not to

enroll in trials, because it is more difficult for

them to comply with the trials than to just get the

drug that is available.

DR. SANTANA: I think a general comment to

192

that, and certainly Dr. Poplack and Dr. Kurtzberg,

as pediatric oncologists can add to that, that has

never been an issue in pediatric oncology. Most

people, even though the drugs are commercial, will

try to design new studies to continue to study

drugs, and patients participate in them.

So, I think the approval of a drug and

making it commercially available in pediatric

oncology has not had the difference in practice

that it may have in adult oncology, because we

still stride to get patients on studies and design

new studies to continue to ask the questions that

are still unresolved.

So, I don't think that ultimately impacts

whether patients participate or not.

So, let's go ahead and get started. So,

this is for NDA 21-673, clofarabine from ILEX

Products, Inc., for the proposed indication of

treatment of pediatric patients 1 to 21 years old

with refractory or relapsed acute leukemia.

The was one, Phase II study conducted in

35 patients with relapsed or refractory AML, at

193

least one relapse or primary refractory. There was

an additional Phase II study conducted in 49

patients with relapsed or refractory ALL, defined

as a combination of at least two relapses or

refractory induction attempts.

Then, in addition, there was a Phase I

study conducted in 25 patients with

relapsed/refractory acute leukemias, a mixture of

ALL's and AML's. Response assessments are

presented in Tables 1, 2, 3, 4, and 5. In both of

these studies, the primary endpoints as defined by

the protocols or the studies, and the ones relevant

to possible approval of the drug, are response rate

and response duration.

In the absence of a reasonable rate of

durable complete responses, which has been

considered clinical benefit in some previous

applications and which did not occur in these

studies, clofarabine can be considered only for

accelerated approval under Subpart H. This would

be based on the conclusion that the responses seen

are reasonably likely to predict a clinical

194

benefit.

Table 1 summarizes the best responses for

ALL, 6 complete responses, 12.2 percent response

rate, and then 4 additional patients with CRp,

complete response without platelet recovery, an

additional 8.2 percent of the patients in that

category.

Table No. 2 briefly summarizes the

response duration for ALL patients that did not go

on to transplant. My interpretation of these

columns is that there were a number of patients who

had the response confirmed with a confirmatory

marrow 3 or 4 weeks later, and those are the ones

in the third column, the 43 and 50 for the CR, and

then there were 3 additional patients who did have

their response subsequently confirmed, and that

response duration was 82, 93+, and 160+ days.

Similar data is presented in Table 3 for

the AML best response in which there were zero

complete responses, for a percentage of zero, and

then there was 1 complete response without total

platelet recovery of 2.9 percent.

195

Similar data in terms of response duration

for AML patients that did not go on to transplant,

which were very few, zero for the CR, zero for the

CRp's, and only 2 PR's that were not transplanted,

had response confirmed by a subsequent marrow.

Table 5 briefly summarizes the response

rates that were seen in the Phase 1 study in which

there was documentation of 2 CR's in ALL out of 17

potential subjects and 1 CRp in the AML subgroup

out of 8 potential subjects.

So, let's go on to the question. I think

you have the introductory paragraph there, so I

won't read it again since it is projected in the

video, and you have it in front of you. I will

give you a minute to read it, and then we will get

started with the questions.

I just want to mention to the Committee

that Dr. Maldonado, as the industry representative,

is a nonvoting member, but everybody else on the

table is voting, and we do appreciate your vote.

The first question is although the

protocol required responses to be confirmed at

196

least three weeks later, this was often not done.

Do you consider an unconfirmed response useful for

considering drug effect?

We will start with you, Dr. Poplack.

DR. POPLACK: It's not confirmed.

DR. SANTANA: Is that a yes or a no?

DR. POPLACK: That's a no.

DR. SANTANA: Anything else you want to

add?

DR. POPLACK: No.

DR. SANTANA: Dr. Kurtzberg.

DR. KURTZBERG: No.

DR. SANTANA: Dr. Wayne.

DR. WAYNE: The people that know me know

there is no such thing as a one-word answer, so I

just have to say I think the data presented, the

testimonials we see today suggest to us that there

is activity, and the question do you consider an

unconfirmed response useful for considering drug

effect, I think the responses we have heard about

and have seen are drug effect.

But it throws a bomb into the china shop

197

of drug approval, but I have to vote yes, I think

that what we have seen or heard suggests drug

effect.

It doesn't meet the standard that has been

set for drug approval in my reading of the

literature, but the question is, as I read it,

unconfirmed response useful for considering drug

effect, I think yes. It doesn't necessarily meet

the bar that we are asked to prove or disprove in

that regard.

DR. SANTANA: Let me clarify that, and

maybe the FDA can help me with that. I interpreted

that as an extension that it is not drug effect,

but drug effect documented as a response.

DR. PAZDUR: As I stated before, we are

not looking at the approval process as a screening

process, so this has to be a meaningful effect

reasonably likely to predict clinical benefit.

DR. SANTANA: Like the response.

DR. TEMPLE: Although that would have to

do with the number of the responses, too, but in

this case, we are just asking about the nature of

198

the response.

DR. PAZDUR: It isn't just the screening

process.

DR. WAYNE: So, if you have a truly

refractory patient population, a response rate, in

my view, in fact, supports likely benefit.

DR. KURTZBERG: But that is a question

that is confirmed.

DR. TEMPLE: You are saying even in your

view, even an unconfirmed response, going with the

numbers that are up here, does that for you.

DR. WAYNE: So, this is where we talk

science and data, and split hairs. When I have

distinguished colleagues who are pediatric

oncologists, and loving parents who say that their

children had responses and we see data that they

were responses, that drug is active in a refractory

patient population.

So, that doesn't equate to the sorts of

endpoints you are asking to be met, but I believe,

in a refractory patient population, that those

responses reflect activity of that drug in that

199

disease, and therefore, is likely in a more

systematic study to, in fact, translate to clinical

benefit, but I can't prove that.

DR. TEMPLE: Right. I have no brief for

what the answer is, but Dr. Kurtzberg at various

points said there were certain kinds of activity

that really don't tell you anything good is going

to happen, and there are other kinds that do.

DR. WAYNE: But I would just argue that if

you have a drug that kills leukemia cells, that

that is activity. How you apply that activity in a

systematic way to be clinical benefit is a separate

question, but I am just responding literally to

this question. I do believe that unconfirmed

response in shades of gray could, in fact, be

considered drug effect.

DR. TEMPLE: In the end, though, we are

going to be asking whether you think the evidence

of activity that is seen is such that there is

likely to be a clinical benefit. So, in the end,

you have to get to that question.

DR. SANTANA: We will get to that, but we

200

have to wrestle with the question at hand, and I

think you already voted, so we will move forward.

Dr. Rodriguez.

DR. RODRIGUEZ: Yes.

DR. SANTANA: Ms. Hoffman.

MS. HOFFMAN: Yes.

DR. SANTANA: Dr. George.

DR. GEORGE: I have a question before I

can answer this. I am assuming this unconfirmed

response means there was an initial marrow that was

clear, that was a complete response, say, but it

just wasn't confirmed later.

DR. SANTANA: Subsequently. That is the

discussion we had earlier this morning.

DR. POPLACK: The question is quite

ambiguous obviously, and you are talking about--and

I didn't do too well on multiple choice tests,

which is obvious--but there are many ways to look

at this, and I think the real issue is whether one

considers an unconfirmed response useful.

You are talking about an unconfirmed

complete response? What are you really talking

201

about?

DR. KURTZBERG: Do you mean the first

marrow showed remission, and you didn't have a

second one three weeks later? I didn't answer it

that way.

DR. SANTANA: I think what they are trying

to ask us is, for all those patients who did have a

bone marrow that assessed a response, but

subsequently, those patients did not have another

marrow, how do you interpret that information with

the lack of an additional subsequent follow-up

marrow? Does that unconfirmed response still tell

you that there was drug effect? That is what they

are asking.

DR. TEMPLE: That is because the protocol

said you were going to confirm them, but we heard

why that was in many cases difficult.

DR. POPLACK: Obviously, if the first

marrow confirms a response, then, that is evidence

of drug activity.

DR. TEMPLE: Confirm here means another

marrow.

202

DR. SANTANA: Confirm means here that they

did that third bone marrow, if you want to put it

that way. They did the diagnostic relapse marrow,

they did a marrow after they gave drug to assess a

response, and then my understanding is the protocol

required that there should be another follow-up

marrow. That is the sequence of events, and

obviously, as we heard earlier, there were a number

of patients who had a response with the second

marrow, who did not get that third marrow.

DR. PAZDUR: Here again, by drug effect,

we mean reasonably likely to predict clinical

benefit.

DR. KURTZBERG: Then, I change my vote to

yes.

DR. SANTANA: So, let us start off with

that clarification because I want to make sure

people know what they are voting on.

Let's start with Dr. Poplack.

DR. POPLACK: So, it's a yes given that

change.

DR. SANTANA: Dr. Kurtzberg.

203

DR. KURTZBERG: Yes.

DR. SANTANA: Dr. Wayne.

DR. WAYNE: Yes, I didn't change my

answer.

DR. SANTANA: We are starting all over.

We erased everything.

Dr. Rodriguez.

DR. RODRIGUEZ: Same answer as previously,

yes.

DR. SANTANA: Ms. Hoffman.

MS. HOFFMAN: Yes.

DR. SANTANA: Dr. George.

DR. GEORGE: I am glad I brought that up.

Yes, I would say.

DR. SANTANA: MS. Haylock.

MS. HAYLOCK: Yes.

DR. SANTANA: Dr. Hussain.

DR. HUSSAIN: Yes.

DR. SANTANA: Dr. Perry.

DR. PERRY: Yes.

DR. SANTANA: Dr. Mortimer.

DR. MORTIMER: Yes.

204

DR. SANTANA: Dr. Santana. Yes.

Dr. Martino.

DR. MARTINO: No.

DR. SANTANA: Dr. Brawley.

DR. BRAWLEY: No.

DR. SANTANA: Dr. Cheson.

DR. CHESON: I would have to say yes,

there is evidence of drug effect.

DR. SANTANA: Dr. Bukowski.

DR. BUKOWSKI: No.

DR. SANTANA: Can you tally the votes for

me? Twelve Yes and 3 No.

Question No. 2. Transplantation,

especially in AML patients, was common in the data

that was presented. Although it is possible that

response to clofarabine encouraged physicians to

consider transplant when they otherwise would not

have, there is no way to know this and there were

no criteria for transplantation in the protocols.

Some patients went to transplantation without a

clofarabine response.

Do the transplantation data contribute to

205

the assessment of the effectiveness of clofarabine?

We need to discuss each of the two diseases

separately. So, let's vote on ALL first.

Dr. Poplack.

DR. POPLACK: For the reasons that Dr.

Kurtzberg stated, I would say no.

DR. SANTANA: Dr. Kurtzberg.

DR. KURTZBERG: No.

DR. SANTANA: Dr. Wayne.

DR. WAYNE: No.

DR. SANTANA: Dr. Rodriguez.

DR. RODRIGUEZ: No.

DR. SANTANA: Ms. Hoffman.

MS. HOFFMAN: No.

DR. SANTANA: Dr. George.

DR. GEORGE: No.

DR. SANTANA: Ms. Haylock.

MS. HAYLOCK: No.

DR. SANTANA: Dr. Hussain.

DR. HUSSAIN: Yes.

DR. SANTANA: Dr. Perry.

DR. PERRY: Yes.

206

DR. SANTANA: Dr. Mortimer.

DR. MORTIMER: Yes.

DR. SANTANA: Dr. Santana. Yes.

Dr. Martino.

DR. MARTINO: No.

DR. SANTANA: Dr. Brawley.

DR. BRAWLEY: Yes.

DR. SANTANA: Dr. Cheson.

DR. CHESON: No.

DR. SANTANA: Dr. Bukowski.

DR. BUKOWSKI: No.

DR. SANTANA: Can I get a tally of votes

for ALL.

Ten No, 5 Yes.

So, the same question. Do the

transplantation data contribute to the assessment

of the effectiveness of clofarabine in AML?

Dr. Poplack.

DR. POPLACK: No.

DR. SANTANA: Dr. Kurtzberg.

DR. KURTZBERG: No.

DR. SANTANA: Dr. Wayne.

207

DR. WAYNE: No.

DR. SANTANA: Dr. Rodriguez.

DR. RODRIGUEZ: No.

DR. SANTANA: Ms. Hoffman.

MS. HOFFMAN: No.

DR. SANTANA: Dr. George.

DR. GEORGE: No.

DR. SANTANA: Ms. Haylock.

MS. HAYLOCK: No.

DR. SANTANA: Dr. Hussain.

DR. HUSSAIN: Yes.

DR. SANTANA: Dr. Perry.

DR. PERRY: Yes.

DR. SANTANA: Dr. Mortimer.

DR. MORTIMER: No.

DR. SANTANA: Dr. Santana. No.

DR. MARTINO: No.

DR. BRAWLEY: No.

DR. CHESON: No.

DR. BUKOWSKI: No.

DR. SANTANA: Can I get a tally of the

votes for AML. Thirteen No, 2 Yes.

208

As noted, in the refractory acute

leukemias, the FDA has considered a good complete

response with complete responses of good duration

to represent clinical benefit.

There was clearly no substantial CR rate

in AML, and in ALL, only 2 non-transplanted

patients had a response duration of at least 3

months. The partial response duration in AML is

not assessable in many responders because they had

early transplantation. There is somewhat more

information in ALL, and that refers back to Table

No. 2, if I remember correctly.

So, the third question is does the ODAC

believe that the clofarabine complete response rate

with available response duration data is reasonably

likely to predict a clinical benefit in ALL?

Dr. Poplack.

DR. POPLACK: I guess I would ask, at the

same level? Isn't that the key? At the same level

that was indicated in this trial, is that what you

are getting at?

DR. PAZDUR: This is the drug approval

209

question, reasonably likely to predict a clinical

benefit, what would be used for accelerated

approval.

DR. SANTANA: So, the question is, for

accelerated approval in ALL, this question

addresses the issue that the available response

duration is reasonably likely, the CR rate and the

duration of response in those patients is likely to

provide some assessment or some idea of clinical

benefit.

DR. TEMPLE: The question, of course,

highlights the lack of knowledge about duration in

a number of cases, but, of course, that is because

they got transplanted, and maybe that is not a

problem. So, that is the question.

DR. POPLACK: The issue of duration of

response, I think is actually quite important here,

or lack of follow up, because there are 3 patients

who had very short responses, who relapsed during

that time period, who didn't have follow up, so you

don't know the length of that duration. Those are

the ones of 43, 50, and 82 in 82 days, if I am

210

correct. We didn't have any follow up on those.

DR. SANTANA: No, no, the 82-day did have

a follow up, am I correct, if I interpret the

columns? Yes, they had a confirmed marrow.

So, referring to Table 2, Dr. Poplack, the

last column, as I interpret it, is the 3 patients

with ALL that are CR, who did have a confirmed

marrow, so obviously, one patient had a confirmed

marrow that lasted 82 days, and the other patients

had a confirmed marrow at 93, and that patient we

think is still in remission, and patient 160 had a

confirmed marrow at 160 and is still in remission.

That is how I interpret that table.

DR. POPLACK: So, 43 and 50 that we have

no information on.

DR. SANTANA: Forty-three and 50 were the

short remissions that were not confirmed with a

subsequent marrow.

DR. MARTINO: I need clarification.

DR. SANTANA: Yes.

DR. MARTINO: I need clarification on this

issue. Dr. Pazdur, do I understand that the point

211

to this question is whether the data, as presented,

is sufficient, not that it shows a whiff of

activity, but is sufficient activity to merit

accelerated approval, is that the question you want

answered?

DR. PAZDUR: Yes, it is. That is why I

said this is the approval question. We are asking,

with the data that you have seen in ALL, is the

drug approvable, should the drug be approved under

accelerated approval conditions.

DR. WAYNE: Can we have more discussion at

this juncture about implications of one way or the

other, or do you want us just to vote based on this

question and call it a day?

DR. PAZDUR: What would be your

discussion? We have been discussing this, and, you

know, there is ambiguities in this application

obviously.

DR. WAYNE: So, I think between the

acknowledgment that there is activity, and there is

acknowledgment that more systematic studies are

required, is a big chasm, and the question is, is

212

accelerated approval the best mechanism to, in

fact, prove (a) activity, in (b) systematic

studies, or will accelerated approval complicate

that.

Most importantly, are we, as a committee,

do we have enough knowledge about that key

question, is what next.

DR. TEMPLE: In some ways, that isn't

really the primary question. We do worry about

whether the confirmatory data will get done, but

the point of accelerated approval is that for

diseases with no current treatment, we are prepared

to accept a different kind of evidence of

effectiveness, that is, you don't have survival

data, you don't have mature any kind of data, but

you may have evidence of an activity that convinces

you that there really will be a benefit when you do

the rest of the studies.

That is a very complicated judgment. The

answer is usually not obvious. It has something to

do with how high the rate is. I mean if you only

saw one response, you probably wouldn't find that

213

convincing.

If you found 40 percent, you probably

would, and somewhere in between, we line out, but

it isn't really about whether the best way to get

the data--I mean the best way to get the data is to

plan it all out, have properly done studies, and

look at them early and say, oh, well, fine, I am

ready to approve it now, and then I have got the

data in hand.

Well, we try to get people to do that, and

as Rick says, more and more people are, that is not

our problem here. We don't have that.

DR. PAZDUR: With the available data that

you have here, is what you are seeing here

reasonably likely to predict clinical benefit for

these patients.

DR. TEMPLE: For a group of patients who

don't have any other choice. That is what the

point of it all was.

DR. SANTANA: Yes, Dr. Martino.

DR. MARTINO: I would like to just restate

your words. The way I understand the charge of

214

this committee is that if we vote yes, then, that

basically means that this drug, as of today,

tomorrow, or whenever, becomes available to any

physician who has such a patient. It is common use

in this setting at least, and then beyond this

setting, that we are actually approving.

It isn't that, in fact, perhaps someone

will then show us more data in the future.

DR. PAZDUR: Correct. You have to go on

what is presented here today and make the

scientific decision, a clinical decision on a

risk-benefit situation on the data that you have at

hand on the basis of this single-arm trial.

DR. TEMPLE: And assume that we will, in

one way or another, along with the company,

actually get the rest of the data.

DR. SANTANA: Yes, I was going to add

that.

DR. TEMPLE: We all know that has been a

problem, we are being forceful in insisting on

seeing the protocols prior to approval. We have,

in some cases, urged that there be enrollment in

215

some of the other studies.

We are worried about that, and we didn't

put that question to you. Feel free to comment,

but you should be assuming that we will get the

ultimate data.

DR. SANTANA: I was going to say an

accelerated approval, my understanding, and the

Agency can correct me, is that there is a firm

commitment that the Agency has regulatory oversight

if it doesn't happen that those additional studies

be done and be presented.

DR. PAZDUR: Correct.

DR. TEMPLE: And usually, pretty well

honored, but there have been troubles, I mean I

don't want to hide that from anybody.

DR. SANTANA: Dr. Poplack.

DR. POPLACK: No, you clarified it for me.

DR. SANTANA: So, are we ready to go back

to this question? So, for ALL--I won't repeat the

question--Dr. Poplack, for ALL?

DR. POPLACK: Yes.

DR. KURTZBERG: Yes.

216

DR. WAYNE: Yes.

DR. RODRIGUEZ: Yes.

MS. HOFFMAN: Yes.

DR. GEORGE: No.

MS. HAYLOCK: Yes.

DR. HUSSAIN: No.

DR. PERRY: Yes.

DR. MORTIMER: Yes.

DR. SANTANA: Yes.

DR. MARTINO: No.

DR. BRAWLEY: No.

DR. CHESON: No.

DR. BUKOWSKI: No.

DR. SANTANA: Can I get a tally?

Nine Yes, 6 No.

DR. SANTANA: Question No. 4 is the same

crux of the matter, now with the AML population.

Does the ODAC believe that the 2 clofarabine

complete responses p's, (1 in Phase I and 1 in

Phase II) are reasonably likely to predict a

clinical benefit in AML?

DR. CHESON: Point of clarification.

217

There was not a CR, it was a CRp.

DR. SANTANA: Can you clarify the acronym

that you have put on this question?

DR. PAZDUR: It is CRp, I believe.

DR. CHESON: CRp, it is not a true CR.

DR. PAZDUR: Correct.

DR. SANTANA: Thank you for the

clarification.

Dr. Poplack.

DR. POPLACK: Just again to clarify. Does

this mean that we think that it has activity or

that it doesn't have activity?

DR. PAZDUR: It is not activity. It is a

clinical decision that you have to make based on

available data, that it is reasonably likely that

activity in these patients would represent clinical

benefit in the future.

DR. POPLACK: Clinical benefit meaning

they are going to live longer--

DR. PAZDUR: The survival, the patient

getting some benefit from it.

DR. SANTANA: I think there is two

218

questions. I think since we have a little bit of

time, I will go ahead and let Dr. Wayne, if you

have a clarification, not a long discussion.

DR. WAYNE: So, again, I just want to

point out that this question literally asks about

only the data presented in the pediatric trial.

DR. SANTANA: Correct. We have not seen

any adult data presented systematically by the

sponsor or the FDA, period.

Ms. Hoffman.

MS. HOFFMAN: And this impacts on labeling

in terms of prescribed use for ALL and AML.

DR. PAZDUR: Correct. If the drug is

approved only in ALL, the indication will only be

in ALL.

DR. SANTANA: Okay. Dr. Poplack.

Can you please identify yourself before

you vote, so I don't have to repeat the names.

Thanks, Dr. Poplack.

DR. POPLACK: No.

DR. KURTZBERG: No.

DR. WAYNE: No.

219

DR. RODRIGUEZ: No.

MS. HOFFMAN: Yes.

DR. GEORGE: No.

MS. HAYLOCK: No.

DR. HUSSAIN: No.

DR. PERRY: No.

DR. MORTIMER: No.

DR. SANTANA: No.

DR. MARTINO: No.

DR. BRAWLEY: No.

DR. CHESON: No.

DR. BUKOWSKI: No.

DR. SANTANA: So, there should have been

14 No and 1 Yes.

Does the Agency wish to discuss any

additional information to help you, or give you

further advice?

DR. PAZDUR: Not that I can think of at

this time.

Could we just make an announcement,

because we have to be here, back sharp, at what

time?

220

DR. SANTANA: 12:45.

DR. PAZDUR: 12:45, we will start right at

12:45 whether you are here or not, because many

people have planes.

DR. SANTANA: Thanks, everybody, for your

participation.

[Whereupon, at 12:07 p.m., the proceedings

were recessed, to be resumed at 12:45 p.m.]

221

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

[12:50 p.m.]

Call to Order

DR. MARTINO: If everyone would take their

seats within the next minute or two, I would like

to get started.

I have a plane to catch at 4:30, I need to

leave here at 4:30, and I mean to do that, so I

remind all of you that have something to say, to

please be clear and succinct. I can be fairly

forceful if need be.

Introductions

MS. CLIFFORD: We are going to start with

Dr. Wilson, if you would like to go ahead and

introduce yourself and your affiliation, please.

DR. WILSON: My name is Wyndam Wilson. I

am in the Experimental Transplantation and

Immunology Branch at the National Cancer Institute.

DR. BISHOP: Michael Bishop, Experimental

Transplantation Branch, National Cancer Institute.

MS. KRIVACIC: Susan Krivacic, Austin,

Texas, Patient Rep, non-Hodgkin's lymphoma

222

survivor.

DR. MALDONADO: Samuel Maldonado from

Johnson & Johnson. I am here as the Industry

Representative to this advisory Committee.

DR. GEORGE: Stephen George, Duke

University.

MS. HAYLOCK: Pamela Haylock, Oncology

Nurse, Consumer Representative.

DR. HUSSAIN: Maha Hussain, Medical

Oncology, University of Michigan.

DR. PERRY: Michael Perry, Medical

Oncology, University of Missouri, Ellis Fischel

Cancer Center.

DR. MORTIMER: Joanne Mortimer, Moores

UCSD Cancer Center.

DR. MARTINO: Silvana Martino, Medical

Oncology, from the John Wayne Cancer Institute.

MS. CLIFFORD: Johanna Clifford, Executive

Secretary to the ODAC.

DR. REAMAN: Gregory Reaman, Pediatric

Oncology, Children's Oncology Group, George

Washington University.

223

DR. BRAWLEY: Otis Brawley, Medical

Oncology/Hematology from Emory University.

DR. CHESON: Bruce Cheson, Head of

Hematology, Georgetown University, Lombardi

Comprehensive Cancer Center.

DR. BUKOWSKI: Ron Bukowski, Cleveland

Clinic, Medical Oncology.

DR. HAZARIKA: Maitreyee Hazarika, Medical

Officer, FDA.

DR. FARRELL: Ann Farrell, Clinical Team

Leader, FDA.

DR. PAZDUR: Richard Pazdur, FDA.

DR. WILLIAMS: Grant Williams, FDA.

DR. MARTINO: Thank you. Ms. Clifford

will now read the Conflict of Interest Statement

for this group.

Conflict of Interest Statement

MS. CLIFFORD: The following announcement

addresses the issue of conflict of interest and is

made a part of the record to preclude even the

appearance of such at this meeting.

Based on the submitted agenda and all

224

financial interests reported by the Committee

participants, it has been determined that all

interests in firms regulated by the Center for Drug

Evaluation and Research present no potential for an

appearance of conflict of interest at this meeting

with the following exceptions:

Dr. Stephen George has been granted a

waiver under 18 USC 208(b)(3) for serving as a

consultant to a competitor on an unrelated matter.

He receives less than $10,000 per year.

Dr. Michael Perry [technical interruption]

for owning stock in a competitor valued between

$5,001 to $25,000. A waiver under 18 USC 208(b)(3)

is not required because the de minimis exception

2640.202(b)(2) applies.

Dr. Ronald Bukowski has been granted a

waiver under 18 USC 208(b)(3) for lecturing for a

competitor on an unrelated matter. He receives

between $5,001 to $10,000.

Dr. Otis Brawley has been granted a waiver

under 18 USC 208(b)(3) for consulting with a

competitor on an unrelated matter. He receives

225

less than $10,001 a year.

A copy of the 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.

We would also like to disclose that Dr.

Samuel Maldonado has been invited to participate as

the Non-Voting Industry Representative acting on

behalf of all regulated industry. Dr. Maldonado is

employed by Johnson & Johnson Pharmaceutical

Research and Development.

In the event that the discussions involve

any other products or firms not already on the

agenda for which am FDA participant has a financial

interest, the participants are aware of the need to

exclude themselves from such involvement and their

exclusion will be noted for the record.

With respect to all other participants, we

ask in the interest of fairness that they address

any current or previous financial involvement with

any firm whose products they may wish to comment

upon.

226

Thank you.

DR. MARTINO: Is there anyone on the

Committee that needs to make a statement in terms

of conflict?

[No response.]

DR. MARTINO: Thank you.

Dr. Pazdur will now make some introductory

comments.

Opening Remarks

DR. PAZDUR: Thank you, Dr. Martino.

This afternoon's session focuses on the

marketing application of vincristine sulfate

liposome for the treatment of patients with

aggressive relapsed non-Hodgkin's lymphoma treated

with at least two combination chemotherapy

regimens. The sponsor is seeking accelerated

approval for this agent.

Since there are members of the Committee

who did not attend this morning's session, I would

like to reiterate several comments made earlier

regarding accelerated approval and then comment on

issues specific to this application.

227

The demonstration of clinical benefit is

required to achieve full approval. In oncology,

the demonstration of clinical benefit has usually

been an improvement in overall survival or the

amelioration of disease-related symptoms.

In 1992, the accelerated approval

regulations allowed the use of additional endpoints

for the approval of drugs that are intended to

treat serious and life-threatening diseases. These

drugs may either demonstrate an advantage over

available therapy or provide therapy where none

exists.

The FDA may grant accelerated approval

based on the effect of a surrogate endpoint that is

"reasonably likely" to predict clinical benefit.

A drug is approved under the accelerated

approval rule on the condition that the

manufacturer conduct studies to verify and describe

the clinical benefit. The regulations stated an

expectation that post-marketing studies would

usually be underway prior to accelerated approval,

but this is not a requirement.

228

At a March 2003 ODAC meeting, the ODAC

reinforced the Agency's view that these

confirmatory trials should be ongoing at the time

accelerated approval is granted. Approval with

subsequent commercial availability of the drug may

interfere with the enrollment of a confirmatory

study.

Accelerated approval had been based on

objective response rates with adequate duration in

single-arm trials in patients with refractory

disease. Since an agent must demonstrate an

advantage over available therapy or provide therapy

where none exists, we are asking you to consider if

available therapy exists for the indication under

consideration.

If available therapy exists, a randomized

trial comparing the investigational drug to an

available therapy arm would generally be needed to

demonstrate superiority. Available therapy usually

consists of drugs that are indicated in drug

labeling for the treatment of a specific disease,

however, in oncology, where drugs are frequently

229

used in non-approved indications as single agents

or in combinations, available therapy may

constitute therapy substantiated by "compelling"

literature evidence of efficacy.

There is no regulatory definition of the

word "compelling," hence, we are asking your

opinion regarding this word.

We are not asking you to reach a consensus

on a specific therapy. Available therapy may be a

single drug or it may be a combination regimen.

Available therapy may be several regimens or drugs.

Where there may be a lack of consensus regarding a

single specific treatment, the Agency has even

recommended using several regimens or drugs as a

treatment arm with the stipulation that superiority

is demonstrated by the investigational drug.

The primary endpoint of this single-arm

trial is response rate. The interpretation of

response rate is complex. We have emphasized that

the persuasiveness of the results of a single-arm

trial to support accelerated approval hinges on the

magnitude and the duration of responses observed in

230

trials.

In aggressive lymphomas, we have

emphasized to sponsors the importance of complete

responses with adequate and well-defined durations

as an endpoint for drug approval. In selected

hematological malignancies where partial responses

are observed, we, at the FDA, have been impressed

with substantial response durations, and these have

led to approval.

For example, the recently approved Velcade

for the treatment of refractory multiple myeloma

had a median response duration in excess of one

year. Similarly, the median duration of disease

control for fludarabine in CLL was in excess of one

year.

Since the vast majority of responses noted

in this application are partial responses with

uncertain durations, 13 out of 30 responders did

not even have a single repeat scan or progress

before a repeat scan could be performed, we are

asking your opinion regarding this endpoint.

Remember, this endpoint in this study must be

231

reasonably likely to predict clinical benefit.

In the morning session, we highlighted the

need for confirmatory trials to be ongoing at the

time of drug approval. This is again not a

requirement, however, the ODAC has supported our

viewpoint that accelerated approval trials be part

of a comprehensive drug development plan with early

initiation of confirmatory trials prior to drug

approval.

To date, a confirmatory trial for VSLI has

neither been started, nor agreed upon with the FDA.

In your deliberations, discussion must focus on

this aspect and the impact that any approval would

have on the conduct and the completion of any

confirmatory trial.

I would like to emphasize that accelerated

approval is not simply a screening process for drug

activity. Mere demonstration of a nominal activity

is insufficient for accelerated approval. Response

rate and duration must provide convincing evidence

that the magnitude and duration of responses are

"reasonably likely to predict clinical benefit."

232

This response rate and duration may vary

from disease to disease. The accepted response

rate in refractory metastatic colon cancer may have

little bearing on the response rate accepted for

refractory aggressive lymphomas, hence, we are

asking your clinical judgment in this disease

setting.

There must be confidence in any

recommendation that a drug approved under

accelerated approval represent a benefit over

available therapy or provide therapy where none

exists.

In making a regulatory decision, you must

be able to accurately assess a risk-benefit

relationship. You must be able to have confidence

in the benefit of the drug in relationship to its

toxicity.

As stated in the morning session, your

decision regarding the approval status of a drug

should be based on a clinical risk-benefit

decision, not simply a desire to provide drug

access to patients.

233

Access to a yet to be approved drug can be

accomplished through registration trials or

expanded access programs.

Seeking drug approval exclusively with a

single-arm trial is an inherently risky venture,

hence, we have strongly urged sponsors to consider

the single-arm trials part of a comprehensive drug

development plan that incorporates the early

initiation of randomized trials to define clinical

benefit.

Although single-arm trials are less

expensive, less complex to conduct, involve fewer

patients, and are performed more rapidly than

randomized trials, they frequently do not provide

the information required by physicians and patients

to make rational therapeutic decisions.

If results are robust in a single-arm

trial, then, everyone wins. Most importantly,

patients receive needed therapies earlier. If

results are nominal, resulting in ambiguity,

regarding a risk-benefit decision, randomized

trials will be needed to accurately assess the

234

drug. Unfortunately, this may delay drug approval.

In comparison to single-arm trials,

randomized trials provide the opportunity to

examine additional endpoints, such as survival,

time to progression, and even symptom benefits, and

also provide us an opportunity to more accurately

characterize adverse events.

I hope these opening comments will focus

your deliberations. Thank you.

DR. MARTINO: Thank you, Dr. Pazdur.

At this point, I would like to invite Inex

Pharmaceuticals to introduce themselves and present

their data to the Committee, please.

NDA 21-000, Marqibo (vincristine sulfate

liposomes injection)

Inex Pharmaceuticals Corp.

Sponsor Presentation

Introduction

MS. MANCINI: Good afternoon, Madam

Chairman, members of the Advisory Committee, and

the FDA. My name is Alexandra Mancini, and I am

the Senior Vice President of Clinical and

235

Regulatory Affairs at Inex Pharmaceuticals.

[Slide.

On behalf of Inex and our co-development

partner, Enzon Pharmaceuticals, I would like to

thank you for this opportunity to discuss our NDA

for Marqibo, which is vincristine sulfate liposomes

injection, or what I will call VSLI for short.

The indication we are seeking is for the

treatment of patients with aggressive non-Hodgkin's

lymphoma previously treated with at least two

combination chemotherapy regimens.

[Slide.

We have several consultants with us this

afternoon, two lymphoma experts, a Dr. Fernando

Cabanillas from M.D. Anderson, and Dr. Jane Winter

from Northwestern University.

[Slide.

As well as experts in lymphoma pathology,

Dr. Randy Gascoyne; Radiology, Dr. Scott Gazelle

and Dr. Sandra Chica.

[Slide.

Neurotoxicity, Dr. Shayne Gad;

236

Pharmacokinetics, Dr. Jean-Marie Houle; and

Statistics, Mr. Louis Gura.

[Slide.

We are very pleased to have Dr. Fernando

Cabanillas with us today. Dr. Cabanillas is a

Clinical Professor of Medicine at the M.D. Anderson

Cancer Center, as well as the Medical Director of

Auxilio Mutuo Cancer Center in Puerto Rico.

He was the Chairman of the Lymphoma

Myeloma Department at M.D. Anderson for

approximately 20 years, where he led the

development of many of the most commonly used

chemotherapy regimens for patients with relapsed

lymphoma including MIME, DHAP, and ESHAP. He has

contributed to over 300 original publications in

lymphoma and over 50 book chapters, and he was a

member of the International Workshop that developed

response criteria for lymphoma.

Dr. Cabanillas will now provide a disease

overview aggressive NHL and the unmet medical need.

Overview

DR. CABANILLAS: Thank you, Ms. Mancini.

237

Good afternoon. I would like to give you a brief

overview of non-Hodgkin's lymphoma over the next

few minutes. These lymphomas are broadly

classified into either aggressive or indolent

histologies.

The most common category is the

aggressive, which constitutes approximately 35 to

40 percent of all lymphomas. This group is

relatively homogeneous since it is made up of

essentially diffuse large cell lymphoma and

peripheral T-cell lymphomas.

DLCL, however, can frequently present with

divergent histologies, which means that there are

other areas which contain an indolent cell type.

The treatment, however, is driven by the most

aggressive histology, and the response is measured

in the same way.

At relapse, most patients with aggressive

histologies usually die within a few months.

In contrast, the indolent lymphomas, when

they release, can survive for years.

Finally, the response rate and duration of

238

response of aggressive lymphomas are lower than for

the indolent cell types. Thus, it should not be

surprising that it has been easier to demonstrate

efficacy in indolent lymphomas and consequently,

four agents have recently been approved in the

U.S.A., whereas, there have not been any new agents

approved for aggressive lymphomas in the last 17

years.

[Slide.

First-line treatment of aggressive

lymphoma consists of rituximab plus CHOP. This

combination will cure approximately 50 percent of

patients. Management for patients younger than 65

consists of a standard dose regimen followed by

high-dose chemotherapy and autologous stem cell

transplant if they relapse.

However, only those patients who respond

to standard dose salvage therapy go on to receive

transplant. Patients older than 65 usually are not

considered candidates for transplant. In addition,

there are other reasons for which transplant is not

feasible. Less than 10 percent of such cases can

239

be cured, and their median survival is only six

months.

Finally, with each relapse, the response

rate, as well as the duration, drop considerably.

[Slide.

At the time of third-line therapy, the

situation is even worse. This is not an uncommon

problem and the prevalence in the U.S.A. is 10 to

15,000 cases. There is no established therapy for

patients in second or more relapse.

To make matters worse, their bone marrow

is frequently compromised by prior therapy, thus,

leaving us with few treatment options. These

patients frequently are symptomatic and reduction

in their tumor burden will lead to symptom

improvement.

The results are dismal. Complete

responses are rarely achieved, and survival is very

short, as you will see in the next slide.

[Slide.

This graph depicts the survival of

patients treated at third line or more. This is

240

derived from a study we performed at M.D. Anderson

during the pre-transplant era and the pre-rituximab

era using an ifosfamide/etoposide based regimen

known as MIME.

After one year, 75 percent of the patients

were dead, and at two years, virtually all except 4

percent were dead. This is a highly lethal

disorder. As you will see soon, the patient

population we have treated with VSLI is similar to

the one presented on this slide.

[Slide.

At this point, I would like to discuss the

available literature data on single agents for

relapsed aggressive lymphoma. The FDA briefing

document contains a table with a large number of

papers quoting response rates with single agents

and combination regimens for lymphoma.

As you can see from this table, many of

those response rates are in the range of 40 percent

and as high as 69 percent, which probably strikes

you as unusual for a single agent at third relapse.

At this point, I would like to put this

241

into perspective. Many of these papers were not

adequate for comparison to the VSLI studies for

several reasons. For example, the first five drugs

listed here were tested in less than 10 patients

with the correct histology.

One of the drugs, oxaliplatin, listed as

having a response rate of 24 percent, was tested in

a population which included indolent lymphomas, as

well as aggressive, but the response rate was not

separated for these two histologies.

Finally, the number of prior therapies was

not comparable for the last three agents on the

slide.

[Slide.

This slide is a summary of the previous

slide, and in the first three columns, I have

listed the reasons why the studies are not

comparable to the VSLI studies. There are only 2

of these 11 single agent papers which actually can

be compared to the VSLI studies.

Moving down to combination therapies, you

see that the situation is very similar, and only 5

242

of the 35 papers quoted are comparable.

[Slide.

The two single agent papers that can

actually be compared to VSLI are Rituxan studies by

Rothe and Tobinai. We have added another important

study by Coiffier, which was not included in the

FDA document.

As you can see in the first row of data,

the median number of prior regimens was 2 in all of

these papers, but a substantial percentage of

patients actually were treated at first relapse,

and those are not comparable to the VSLI patients.

As you can conclude from this slide, even

in a population with less relapses than ours, the

response rate is less than 40 percent, and the time

to progression is only 2 to 4 months.

[Slide.

These are the five combination regimen

papers that are adequate for comparison to VSLI.

The top half represents those with a median number

of prior regimens equal to 2, while the bottom have

a median number of 3, which is similar to the VSLI

243

studies.

The response rates range from 39 to 65

percent, but as you can see from the bottom half,

the CR rate is very low at third or more relapse.

The top three regimens are not commonly used

salvage combinations, and neither is ifosfamide,

hydroxyurea, and etoposide shown in the last row

and all of them are certainly more toxic than VSLI,

as you will see later on, that, even though a

popular regimen, is not commonly used to third

relapse because of its serious myelosuppressive

toxicity, which I am sure those of you who have

used it are familiar with.

In summary, the regimens shown here are

not used commonly by the oncology community with

the exception of DHAP, which is used mostly at

second line prior to autologous transplant. It is

highly toxic and thus rarely used in patients at

third or more relapse.

[Slide.

By now, you probably have realized that

there exists several unmet clinical needs in the

244

management of aggressive relapsed lymphomas. Many

patients do not qualify for aggressive combination

regimens or have failed autologous transplantation,

and they need some other alternative.

Some of the characteristics of such

patients are listed here. In the past, patients

with compromised marrow function were treated with

rituximab, which is not myelosuppressive, however,

that is no longer an alternative because by the

time they get to third line treatment, most of

these patients have already been exposed multiple

times to rituximab and are resistant to it.

In summary, there is no compelling

literature evidence for available therapy after

second relapse, and there is a great need for an

agent that can provide clinical and meaningful

benefit without excessive toxicity, because at this

point, we are dealing mostly with palliation, and

we should not induce severe toxicity if we are to

effectively palliate them.

[Slide.

I would like now to introduce Dr. Tom

245

Madden from Inex, who will discuss the pharmacology

of VSLI.

Pharmacology

DR. MADDEN: Thank you, Dr. Cabanillas.

[Slide.

As the name indicates, VSLI is a liposomal

formulation of vincristine sulfate where the drug

is encapsulated within the aqueous interior of

small liposomes. These are composed of

sphingomyelin and cholesterol, and this lipid

composition provides a highly stable bilayer with

relatively low permeability.

[Slide.

The intention with VSLI is to increase

tumor exposure to vincristine, and this is achieved

through two mechanisms, first, by providing higher

drug levels at the tumor sites, and secondly, by

providing a mechanism to provide a sustained

duration of exposure. The following slides will

illustrate these two mechanisms.

[Slide.

Following intravenous administration of

246

VSLI, the liposomes, being particular carriers,

tend to remain within the blood compartment as they

are not able to readily extravasate across the

continuous endothelial lining present in most

normal blood vessels, and this can be illustrated

using a window chamber model, which allows

visualization of florescently labeled VSLI.

The panel on the left-hand side here shows

normal blood vessels, and you can see that the

liposomes are constrained within the vessels with

very little accumulation seen in the interstitial

spaces.

Within tumor including lymphomas, however,

the neovasculature tends to be leaky, exhibiting

pores or discontinuities. The liposomes in VSLI

are of an appropriate size to allow extravasation

through these pores with subsequent accumulation in

the interstitial spaces.

Again, we can illustrate this using the

window chamber model.

As you will see when I start the video,

the blood vessels within the tumor vasculature tend

247

to be smaller, and you will see the architecture is

highly chaotic. You will also see that

extravasation of the liposomes is apparent and

appreciable accumulation occurs in the interstitial

spaces.

The period of time followed in both videos

is the first 60 minutes after VSLI administration.

The VSLI also accumulates preferentially in the

tissues and organs of the mononuclear phagocyte

system, such as the liver, spleen, and lymph nodes.

These are, of course, also sites of lymphoma

involvement.

As you are well aware, vincristine is a

cell-cycle-specific agent inhibiting mitosis.

Studies using lymphoma isolates from patients have

shown that only a small fraction of cells, less

than 5 percent, are actually in the sensitive G2M

phase at any point in time. Therefore, the

duration of vincristine exposure is critically

important in terms of its activity.

[Slide.

This is illustrated on the slide shown

248

here. As you can see, the fraction of tumor cells

surviving is greatly reduced as the exposure time

to vincristine is increased from 1 hour to 24

hours, and I should note that the fraction of

viable or surviving cells shown here is represented

on a log scale.

[Slide.

The rate of vincristine release from VSLI

is therefore a critical factor in terms of its

activity, and has been characterized in several

nonclinical studies. Vincristine release occurs by

passive diffusion across the liposomal bilayer, and

similar behavior is seen in the mouse, rats, and in

the dog.

Shown here for the rat is the rate of

vincristine release in plasma. As you can see,

release is slow and sustained with approximately 50

percent of the drug released by 24 hours and

essentially complete release seen by 72 hours.

[Slide.

These changes in the pharmacokinetics of

vincristine resulting from liposomal encapsulation

249

would be expected to increase antitumor activity.

This was shown in several studies, comparing VSLI

with conventional vincristine against a range of

human and murine tumor models.

The study illustrated here compares

activity in the Namalwa lymphoma model. As you can

see, vincristine is active in this model, but VSLI

shows much increased antitumor activity.

I should note that this increased activity

is seen when VSLI is given at the same dose as

conventional vincristine, as will be discussed

during the clinical presentation. Patients are

administered VSLI at approximately twice the

intensity typically used with conventional

vincristine.

[Slide.

The pharmacokinetics of VSLI have been

characterized in patients, and the slide shown here

illustrates plasma vincristine levels following

administration of VSLI to published data for

conventional vincristine.

As you can see, much higher plasma drug

250

levels are achieved for VSLI, and these are

maintained for a considerable period of time.

Again, you will note that the drug levels shown on

the y axis are represented on a log scale.

At almost all time points, plasma drug

levels are approximately 2 orders of magnitude

higher for VSLI compared to conventional

vincristine.

In summary, VSLI provides increased

exposure to vincristine through higher tumor drug

levels and also an extended duration of exposure.

In nonclinical studies, this results in increased

antitumor activity compared to conventional

vincristine.

I haven't presented any data on the safety

evaluations conducted for VSLI. These show that it

elicits the same toxicities seen with conventional

vincristine, and importantly, no new toxicities

were observed.

I will now pass the podium back to

Alexandra Mancini who will present the clinical

results.

251

Clinical Efficacy and Safety

MS. MANCINI: Thank you, Tom.

[Slide.

There are two trials in the NDA that

provide efficacy data in the indicated population.

We have supportive data from the first study we did

in aggressive NHL, our Phase IIa study.

This was an investigator-sponsored trial

at M.D. Anderson Cancer Center, which was a

broad-based protocol including lymphoma and

leukemia, and there were 92 patients with relapsed

aggressive NHL in that study.

Our primary Phase IIb study was an

international multicenter study that enrolled 119

patients at 42 centers. These are the two largest

trials reported for patients with multiply relapsed

aggressive NHL.

Both were single-arm studies and both used

similar response criteria. They provided

consistent efficacy results in a total of 211

patients, therefore, for brevity, I will focus on

the pivotal study only.

252

[Slide.

A larger safety database was provided that

includes 53 patients.

Our clinical development in aggressive NHL

is ongoing using combination regimens. We have had

several discussions with the FDA, three meetings in

fact, regarding our proposed randomized controlled

trial that would serve as a post-approval

commitment trial.

We received comments from the FDA as part

of our special protocol assessment that we

submitted, and the revised protocol will be

resubmitted shortly. The study will start within a

few months.

[Slide.

Before I present the results of our

studies, I will address several issues raised by

the FDA in their review that relate to the adequacy

of the pivotal trial and its conduct. One must

first be assured that the results are reliable

before embarking on their interpretation.

[Slide.

253

The Agency has raised concerns about the

low number of eligible patients attributed to the

following reasons: numerous protocol amendments

and exemptions, a low histologic eligibility rate,

and incomplete staging in 19 percent of patients.

Additionally, there have been concerns

raised about the Independent Review Panel for

efficacy evaluations, specifically with respect to

the wording of the response criteria, operations of

the core imaging lab, and amendments to the IRP

charter.

I would like to provide further

clarification on these points for your

consideration.

[Slide.

It is true that there were 9 versions of

the protocol, but it is important to note that the

study did not begin enrollment until Version 5. As

shown on this slide, almost all patients were

enrolled under two versions of the protocol,

Versions 8 and 9. Therefore, there were really

only four amendments after the trial started.

254

[Slide.

The initial protocol used for patient

enrollment required that patients had a CR or CRu

to first line chemotherapy, and they also had to

have achieved at least a PR to their last line.

We were unable to enroll patients to this

narrowly defined population. The trial was open

for 6 months, and we had 4 patients enrolled.

Therefore, we amended the protocol. We deleted the

requirement for a complete response at first line

and allowed patients in if they had only at least a

minor response.

We also removed any requirement for a

response to their last therapy, and in doing so, we

now defined a poorer prognosis population.

[Slide.

To further enhance patient enrollment, we

added some additional histologic categories shown

here, but as you can see from the number of

patients, this did not result in significant

additional enrollment.

Once again, this defined a poorer

255

prognosis population, and there were no further

changes to eligibility criteria. Importantly, with

each amendment, the FDA agreed that the trial

population was a suitable population to support an

accelerated approval.

[Slide.

The last amendment changed the schedule to

obtain CTs 4 weeks after the first documentation of

response instead of the original 8 weeks, and we

further clarified the wording that these

confirmatory CT scans should be obtained instead of

must be obtained as we realized the original

wording might be misconstrued as a requirement for

confirmation of response.

This was not the intent, as it is not a

part of the International Workshop criteria. This

protocol amendment was implemented after

approximately half the patients in the trial were

enrolled and before Inex was unblinded to the study

data.

We remained blinded to the data until all

patients were enrolled.

256

[Slide.

We did allow medically justified

exemptions for some patients, but we were always

careful to not allow exemptions that would have

enhanced the apparent VSLI response rate, and we

are prepared to discuss specific exemptions during

the question period.

The net effect of the exemptions was

enrollment of a patient population with a poorer

prognosis.

[Slide.

Correct histologic diagnosis of

non-Hodgkin's lymphoma is problematic in clinical

practice. For this reason, we included a

retrospective central pathology review to confirm

eligibility. Nineteen percent of the patients were

deemed histologically ineligible by the Central

Review, and these were mostly indolent lymphomas.

The FDA excluded an additional 7 patients

described as probably eligible by the Central

Review. Dr. Randy Gascoyne was the lead

pathologist in our Central Review. He is available

257

to discuss these 7 cases during the question period

and why they should be considered eligible.

It is important to clarify that these

histologically ineligible patients were not

protocol violations or due to the amendments.

Enrollment eligibility was determined by the site

pathology assessment.

[Slide.

This slide provides a summary of

histologic eligibility for several large studies in

aggressive NHL. The rate seem in our study, 81

percent, is very consistent with what has been

reported in the literature.

Furthermore, the studies listed here were

all conducted in newly diagnosed patients where one

often has a larger biopsy specimen available for

review.

[Slide.

Now, on this next slide, I believe that

the comments I was prepared to make are perhaps no

longer valid. We have looked ahead at the slides

that the FDA is to present, thank you for giving

258

those to us in advance, and I think that some of

these numbers won't match anymore, so please excuse

that.

But what we were going to comment here was

that in our opinion, some of the categories here

listed as reasons for exclusion are not categories

that we feel should be used for exclusion from an

efficacy analysis.

[Slide.

There was 1 patient who did not have

complete CTs at study entry, and, of course, in an

eligible patients analysis, they should be

excluded.

Having had bone marrow biopsy earlier or

having missing LDH at baseline is not a reason, we

feel, those missed data points should not be a

reason to justify exclusion from an efficacy

analysis.

Negative bone marrow and normal LDH are

required, of course, at the time one declares a

complete response whether or not they were abnormal

at baseline. They are a part of the criteria for a

259

partial response.

The other reason noted here, which was

missing neurologic exams at baseline, I don't

believe that that was a reason, and FDA will

clarify that in their presentation, so I believe

that they have removed that category.

[Slide.

Concerns have been raised that would

suggest that the International Workshop criteria

were not followed. I would like to clarify that

the wording in the protocol did not, in fact,

reduce the strictness of the criteria. As

excellent as the International Workshop criteria

are in providing appropriate guidance for the

response determination in this very complex disease

setting, there are some situations where the

criteria are ambiguous or silent.

The wording clarifications in the protocol

were undertaken to uphold the rigor of these

criteria and to ensure consistent interpretation in

this multicenter environment.

[Slide.

260

Before the study was started, we met with

the FDA to discuss the protocol Version 5, which

contained the clarified wording, and the FDA agreed

with the protocol wording and no changes were ever

made to that.

With respect to the internal operations of

the core imaging lab, the FDA noted correctly that

the manual in the NDA was dated one year after the

reviews of images began. We had a manual in place

before the reviews began, but it was our oversight

that this earlier version was not included in the

NDA, but we have now provided that to the FDA for

their review.

We can summarize, though, that were no

changes to the lab procedures during the entire IRP

process.

[Slide.

With respect to the amendments to the

Independent Review Panel charter, there were no

changes to the conduct of the IRP radiology or

oncology reviews. Most of the amendments were in

place before the reviews began.

261

A few radiology clarifications were

requested by the IRP radiologist, Dr. Scott

Gazelle, for situations not previously anticipated,

and he is available during the question period to

discuss the details.

All images were read in chronologic

sequence and they were locked after review, and no

changes were permitted.

[Slide.

To summarize the conclusions regarding the

study conduct issues, the protocol amendments and

exemptions neither contributed to ineligibility,

nor favored a positive outcome with VSLI. They, in

fact, defined a population with a poorer prognosis.

The histologic eligibility of 81 percent

in this study is comparable to what is reported in

the literature. Only 8 percent of patients were

ineligible for efficacy evaluation due to protocol

violations.

Lastly, the Independent Review Panel

process was well conducted, therefore, we are

confident that we are providing for your

262

consideration today a well-defined and reliable

assessment of objective response in the indicated

population and that our pivotal study meets the

criteria for an adequate and well-controlled trial.

[Slide.

I would like to now turn to the

presentation of the pivotal study results.

[Slide.

As agreed with the FDA, some of the key

eligibility criteria in this trial were as follows.

First of all, this was a population of patients

with aggressive NHL that was either from first

diagnosis de novo or transformed, and it was

required that they had at least received 2 prior

combination regimens, one of which had to contain

an anthracycline.

We required only a minor response to

first-line therapy, which is usually CHOP therapy

containing vincristine.

[Slide.

The additional criteria on this slide

define a population that would not be eligible for

263

many clinical trials in that we did not have a

maximum on the number of prior regimens patients

could have, there was no requirement for response

to prior salvage therapies, no upper limit on age,

ECOG performance up to level 3 was accepted, a

Grade 2 neuropathy was permitted, and as VSLI is

hematologically well tolerated, we were able to

allow enrollment of patients with low granulocyte

or platelet counts, who would not be able to take

standard chemotherapeutic agents.

Some of these criteria, particularly the

first two, allowed a somewhat heterogeneous

population to be enrolled, but this is the

population for whom we are seeking an indication.

[Slide.

VSLI was given as a monotherapy regimen in

this study. It was given as 2 mg/m

2 without any

dose capping, 1 hour infusion every 2 weeks. The

protocol specified 12 cycles maximum with the

intent to go 2 cycles after a complete response.

What is highlighted now in yellow are

points of differentiation from the conventional

264

vincristine dosing schedule. Conventional

vincristine is given at 1.4 mg/m

2 often with dose

capping at a flat dose of 2 mg, and it is usually

repeated every 3 weeks, so by implementing these

changes, we are able to at least double the dose

intensity of vincristine.

[Slide.

The efficacy endpoints in this trial were

the traditional oncology endpoints with objective

response as the primary endpoint. The primary

population for analysis was the intent-to-treat

population as defined in the protocol and the stats

plan.

The secondary population was based on

patients who met the key inclusion criteria and who

were evaluated. That is what we call the

"per-protocol" population, and we identified 77

patients who met those criteria.

[Slide.

The assessments of response were

determined using the International Workshop

Criteria. A key point here was that 6 indicator or

265

target lesions were to be measured carefully

throughout the study, and all other disease was

assessed qualitatively.

According to these criteria, response does

not require subsequent confirmation.

As this was a single-arm study, it was

important to use an Independent Review Panel for

the primary efficacy assessment. The IRP was

blinded to the site's opinion of response, and they

also independently chose 6 indicator lesions.

[Slide.

So, who did we enroll in this study?

[Slide.

In the interest of time, I will discuss

only a few characteristics of the study population.

The extent of prior therapy was a significant

predictor of response in this study. The protocol

required a minimum of 2 prior regimens, and we see

that 19 percent of the patients in the study had

exactly that amount of prior therapy, but the vast

majority of these patients had much more prior

therapy.

266

The mean was 3.8, the median was 3, so

therefore VSLI was being given predominantly as a

fourth- and fifth- line therapy to this population.

[Slide.

Another important perspective that

provides context for the interpretation of our

results is how these patients responded to their

previous therapies. When they were at their

first-line therapy, the overall response rate was

92 percent, 50 percent complete response.

By their second line, the response rate

had dropped down considerably to 41 percent, and we

were now seeing only 20 percent CRs. At their last

therapy, which could have been second line, but was

usually much further along than second line, the

response rate was down to 35 percent with only 13

percent complete response.

I would like to emphasize that

three-quarters of the patients were receiving a

combination regimen as their last line. Therefore,

these data demonstrate that the population in this

trial had disease that had become very difficult to

267

treat.

The median duration of response was 8

months at first-line therapy and had decreased to 5

months with salvage regimens.

[Slide.

Patients were further characterized with

respect to the sensitivity or resistance of their

disease to their last qualifying therapy.

Two-thirds of the patients in this trial had

resistant disease, and half of the patients had

truly refractory disease, meaning that they did not

respond to their last therapy.

Another 17 percent had early relapses

within 3 months. Therefore, only one-third of the

population had sensitive disease with responses

lasting more than 3 months to their last therapy.

[Slide.

Turning now to the efficacy data.

[Slide.

The results are very similar in the

intent-to-treat and per-protocol populations. For

brevity, I will focus on the intent-to-treat

268

population, which was the primary efficacy

analysis. The per-protocol analyses have been

provided for your review in Appendix B of the

briefing document.

The primary efficacy endpoint was the

objective response rate as assessed by the IRP, and

these are the results on this slide. The IRP

determined that 30 patients were responders. That

is 25 percent. We had 8 patients with a CR or CRu,

so these were predominantly partial responses in

this trial.

Additionally, one-quarter of the patients

had stable disease with VSLI therapy.

[Slide.

Now, on this slide, I provided comparison

of 3 different analyses. The first column is the

intent-to-treat data that I just shared with you.

The second column is the per-protocol

population that we defined of 77 patients, and the

response rate is very similar, as is the

distribution being predominantly PRs, but still

maintaining some complete responses in about a

269

quarter of the patients with this disease

stabilization.

In the far right column, I have presented

the FDA eligible analysis, as was shared in their

original posted briefing document. I believe they

will be updating this analysis today, but this

analysis again showed a very similar response rate

at 22 percent with about 4 percent complete

response.

[Slide.

A key focus of today's deliberations is

whether the objective response rate observed in

this study is likely to predict clinical benefit.

We looked at other available data for the

responding patients as we were requested by the FDA

to prepare patient benefit summaries to facilitate

their review.

[Slide.

We determined that there was evidence of

some symptomatic improvement associated with

objective response. There were 8 patients who were

determined to be complete responses or CRu by the

270

IRP. Three of those patients were actually

asymptomatic at study entry, but the remaining 5

patients who were symptomatic either had resolution

of their symptoms or an improvement in their ECOG

performance status.

Of the 22 patients called partial

responses, 15 of them had improvements in symptoms

or ECOG performance status.

I should mention that there was no formal

symptom efficacy endpoint in this trial, so this is

a summary of baselines signs and symptoms or that

resolved after VSLI treatment. Other evidence of

possible clinical benefit was presented in the

briefing document.

[Slide.

Turning now to the secondary efficacy

endpoints, the time-to-event endpoints.

[Slide.

The first one I would like to discuss is

duration of response. This was analyzed using

Kaplan-Meier method, the usual method, and it is

done only, of course, for the 30 responders. The

271

median was not reached in our analysis, but we were

very close at 52 percent probability at the last

event of progression or approximately 3 months.

The FDA analysis appears to have used a

slightly different definition that included

withdrawal due to neurotoxicity as a progression

event, but this analysis provides a similar median

estimate of about 2 1/2 months.

[Slide.

The analysis of time to progression is, of

course, conducted on all 119 patients. The median

here was estimated again to be about 3 months, but,

of course, this was heavily influenced by the

majority of patients in the trial who did not

respond to VSLI treatment, so an additional

analysis was done for the responding patients.

For this subgroup, the estimated median

time to progression was not reached, but once

again, we were approaching the median at 45 percent

probability. So, we were at about 4 months.

[Slide.

The overall survival is shown in this

272

Kaplan-Meier curve. The median was 6.7 months, and

at 2 years, we have 25 percent of the patients

still alive.

[Slide.

The protocol and stats plan prespecified

several subgroup analyses, and there were two

factors that were determined to be statistically

significant predictors of objective response.

[Slide.

With 2 significant predictors of response,

the most informative presentation is for the 4

subgroups, as shown on this slide. The response

rate for patients who had 2 prior regimens was 46

percent, and for the patients who had more than 2

prior regimens, which was a much larger proportion

of the patients on the trial, the response rate was

lower, at 20 percent.

Within each of these 2 big groups, if we

would further subdivide by whether they had

sensitive or resistant disease, we see the impact

of that factor, as well.

So, we have quite a range of response

273

rates here, from the lowest being 15 percent, to

the highest being 64 percent, and what we can

conclude from this analysis is that the overall

response rate of 25 percent, that we saw in our

study, was very much the result of the relative

number of patients in these 4 subgroups that were

enrolled in the study.

Had we had more patients in this subgroup,

the poorest prognosis subgroup of more than 2 prior

regimens with resistant disease, we would have had

a lower response rate closer to the 15 percent.

Had we had more patients with sensitive

disease in the protocol, we would have been in the

30 to 40, 50 percent response rate.

One criticism of this analysis is that

some of the subgroups are very small.

[Slide.

So, on this slide, we are now showing the

Phase IIa supportive study alongside the data that

I just showed you. The denominators here are for

the combined studies, and what is impressive is

that the results for the 2 studies are very

274

consistent for the 4 subgroups. These 2 predictors

of outcomes were consistent in both trials.

Therefore, one can be reassured that the

estimates for these subgroups are reasonable.

[Slide.

This slide summarizes some other subgroup

analyses that are noteworthy as they did not

identify significant predictors of response. The

response rate was not any lower in patients who had

previously undergone autologous stem cell

transplant, nor was the response rate affected by

age. The median age was 60 years.

Twenty-eight patients on this study were

older than 70 years, and this subgroup had a 36

percent response rate, so at least as good as the

total population.

Patients who have relapsed after

transplant and those who are elderly are

particularly in need of an effective therapy that

is minimally myelosuppressive.

[Slide.

Returning now to the question of whether

275

VSLI provides meaningful therapeutic benefit over

existing treatments, we have the rituximab

publications presented earlier by Dr. Cabanillas on

this slide in the first three columns, and now I

have listed the results of our Phase IIb study in

the two right columns.

The furthest right column has the results

for the intent-to-treat population, which had a

median of 3 prior regimens, therefore, to provide a

better comparator I have pulled out the subgroup of

patients who had 2 prior regimens as that is more

comparable to what is in the literature citations.

But it is important to note again that

these 3 publications for rituximab did include a

substantive proportion of patients at first

relapse, which we did not have in our study.

Even with a slightly more favor population

enrolled in these studies, our response rate, which

is shown along this line here, a response rate of

46 percent compares favorably to what was shown

with rituximab.

Duration of response was not reported in

276

the publications, but on time to progression at the

bottom of the table, the VSLI is in the same range.

[Slide.

We can also compare to the rate of

response demonstrated with rituximab for the

patients in our study. There were 20 patients who

had rituximab as a single agent therapy as their

last therapy before coming into the study.

The response rate to rituximab was 25

percent, and there were no complete responses.

With single-agent VSLI then, as their next

therapy, the response rate achieved was 40 percent,

and we did see some complete responses in these

same patients. Therefore, we are seeing a somewhat

higher rate of response.

[Slide.

I would like to turn now to the safety

data presentation.

[Slide.

The mean number of cycles of VSLI

administered was 4.6 with a median of 4, and the

dose intensity was very close to the target of 1,

277

indicating that very few dose reductions occurred

on the study. I should say that was 1 mg/m

2/week.

That was our target at dose intensity.

[Slide.

Fourteen percent of the patients withdrew

due to associated adverse events, which was mostly

neuropathy. It was 13 percent for neuropathy.

There were no treatment-associated deaths.

[Slide.

The dose-limiting toxicity of conventional

vincristine and of VSLI is, of course, neuropathy.

[Slide.

This slide summarizes the number of prior

regimens that the patients had that contained

neurotoxic agents. Eighty-six percent of the

patients had at least 2 prior regimens that

contained a neurotoxic agent. Therefore, it is no

surprise that 85 percent of the patients entered

the study with some level of neurologic deficit.

[Slide.

Now, on this slide, I am showing the worst

grade of neuropathy on study for patients grouped

278

by their grade of neuropathy at study entry. We

did allow up to Grade 2, as I mentioned.

These data are the worst values for any of

the 5 neuropathy symptoms that we tracked, as shown

in the title. The scoring system here is the NCI

CTC scale, which goes from zero, which is normal,

to Grade 4 at the worst end.

For patients who entered this study with

Grade 1 neuropathy, one-quarter of the patients had

no change on study, and about half of them worsened

by one grade to a Grade 2 level, one-quarter went

to a Grade 3 level.

For patients who entered this study at

Grade 2, approximately 40 percent of them did not

have any worsening on the study, but half of the

patients did worsen one grade to Grade 3. As would

be expected, patients who entered the study with

worse neuropathy had a higher chance of developing

Grade 3 neuropathy.

[Slide.

Numbness in the hands was the symptom that

was most adversely affected on this study, and this

279

plot shows the mean change from baseline to Cycle 6

for this parameter.

We observed a gradual cumulative increase

in hand numbness, reaching a peak which was less

than one grade, and again, this is the NCI grading

system that goes to 4.

This peak was reached after 5 cycles, and

the data beyond Cycle 6 don't show any further

increase. We are down to fewer numbers of patients

on the study at that time.

[Slide.

Her analysis was done to estimate the dose

that would result in Grade 3 or 4 neuropathy if all

patients continued to be dosed. The Kaplan-Meier

method was used for estimation, and all five

symptoms were included. Reaching Grade 3 in any

one of these symptoms was called an event in this

analysis.

One-third of the patients on the study

developed a Grade 3 or 4 neuropathy, but as you can

see, they were almost all Grade 3 neuropathies.

There were only 3 patients who went to Grade 4.

280

The estimated median cumulative dose to

achieve this was 21.2 mg/m

2, which is approximately

11 doses of VSLI. This is equivalent to

approximately 15 doses of conventional vincristine.

This is a lot of vincristine for patients who have

received previous neurotoxic agents, and it speaks

to the safety from the liposomal encapsulation.

[Slide.

With the 25 percent response rate, an

important consideration is the risk exposure for

patients who will not respond to VSLI therapy.

This slide summarizes the magnitude of

treatment-emergent worsening in neuropathy from no

change up to 3 grades. This are all zeros here.

This is presented separately from responders versus

non-responders. So these are the grade changes from

baseline to worse value, the treatment-emergent

changes.

As one would expect, the responding

patients who received more drug did achieve bigger

changes in their neuropathy, mostly 1 or 2 grade

changes, however.

281

In the nonresponding patients, 55 percent

of the patients had no change on study, and

one-quarter had a one grade worsening. This group

includes stable disease patients, and which

includes their minor responses, not meeting the

definition of a partial response, and some of our

stable disease patients actually were treated for a

very long time.

So, I have pulled out a smaller group here

in the bottom row, which are the patients who had

rapid progression, and we see that 69 percent of

the patients in that group had no change at all on

study and 19 percent had a one grade worsening.

Because of the gradual development of

neuropathy, we are able to avoid significant

toxicity in patients who will not benefit from this

therapy.

[Slide.

It is also of interest to compare the

timing of the antitumor effect to the timing of the

neuropathy, and when we prepared the patient

summaries, we observed that the antitumor activity

282

was evident very early in those patients who would

be declared responders, usually within the first

two weeks of off the first injection of the drug.

There was evidence of symptomatic

improvement, reduced palpable adenopathy or

decreased LDH long before we were doing CTs. As

shown earlier, the development of neuropathy is

gradual and predictable, therefore, in contrast to

most other drugs, with VSLI, the physician and

patient can make an informed treatment decision

before significant toxicity develops.

[Slide.

With respect to hematologic abnormalities,

at study entry, we see here at study entry,

approximately 80 percent of the patients had some

level of anemia, 40 percent have thrombocytopenia.

One-third of the patients would not have been

eligible for standard chemotherapeutic agents that

were myelosuppressive based on the low neutrophil

or platelet counts as defined here.

[Slide.

This slide now summarizes the treatment

283

emergent grade changes from baseline to worst grade

on study for these three hematologic parameters.

As shown in this column, about half of the patients

for any particular parameter, about half of the

patients had no change on study.

Most of the changes were in the 1 grade

category. Neutrophils was the parameter that had

the most change, and we see 20 percent of the

patients having a 3 or 4 grade change, and

approximately half of those, so 10 percent of the

patients is where it was considered to be treatment

related.

With respect to the worse level of

neutropenia reported on study, 8 percent of

patients had Grade 4 neutropenia and 3 percent had

febrile neutropenia. This occurred in the setting

of only 2 percent prophylactic filgrastim usage,

so this study provides a good estimate of the VSLI

effect on neutrophils.

Grade 4 thrombocytopenia occurred in only

one patient and 6 percent of patients received

platelet transfusions.

284

Based on all of these data, we conclude

that VSLI was hematologically well tolerated in

this study.

Dr. Cabanillas will now describe patients

who achieved net clinical benefit in this study.

Clinical Benefit

[Slide.

DR. CABANILLAS: The FDA has requested

that the company prepare patient benefit summaries

to facilitate the review of clinical benefit.

There were 38 patients considered to be

responders by either the IRP or the investigator,

and summaries were prepared for all of those

patients. In addition, there were 5 patients with

minor responses who had evidence of clinical

benefit.

Therefore, a total of 43 individual

patient benefit-risk assessments were prepared.

[Slide.

Of the 43 patients analyzed for clinical

benefit, there were 41 who actually manifested

evidence of benefit. I will discuss the findings

285

for the first two categories noted in yellow here,

specifically, symptom improvements and patients who

went on to receive stem cell transplants.

Other categories of clinical benefit

included durable responses and better outcomes than

previously achieved. Some of these will be

demonstrated in a few case studies.

[Slide.

This slide summarizes two of the clinical

benefits, improvement in symptoms and improvement

in performance status.

Twenty of the 43 cases had improvement in

either B symptoms or some other symptom related to

lymphoma, 13 patients experienced improvement in

performance status. A total of 26 patients had

improvement in one of these two categories.

Another clinical benefit we consider

important is being able to provide these patients

with the opportunity to receive a stem cell

transplant.

[Slide.

Six patients were able to receive

286

transplants after the VSLI study. Both

responsiveness to VSLI and maintenance of a good

performance status enabled these patients to

receive their transplant. Five of these 6 patients

are actually alive, 1 with disease, and 4 with no

evidence of disease for two to three years.

[Slide.

In the pivotal study, we observed some

very striking outcomes in several patients, and I

wish I could review each one of them, however, in

the interest of time, we will only be able to go

over three of these patients, but we have also some

patients who will be testifying today.

The first case is a 56-year-old lady with

primary mediastinal DLCL and associated symptoms.

Her response to prior therapy had consisted of a

brief PR to CHOP. Upon relapse, she was treated

with ESHAP and RICE without response to either.

She achieved a CRu of 1 year after 20 cycles of

VSLI. Her toxicity consisted of only one episode

of Grade 4 neutropenia.

What is striking about this case is that

287

she was able to obtain a CRu after being refractory

to 3 combination regimens including CHOP. She

benefited also from resolution of B symptoms and

anemia.

[Slide.

This 76-year-old lady with DLCL and IPI of

3 presented with multiple pulmonary metastases and

thrombocytopenia of 72,000. Her prior therapy

included CHOP and subsequently Cytoxan, VP16,

Rituxan.

After 8 courses of VSLI, she obtained a PR

of 8+ months, and her platelets normalized. Her

tolerance to VSLI was excellent with no Grade 3 or

4 toxicities.

The pulmonary metastases improved, but

never disappeared completely. The residual

lesions, however, have not changed for 2 1/2 years,

suggesting that they are scars, rather than

lymphoma.

This patient benefited from a long

chemotherapy-free interval still ongoing at 27+

months, which is a longer remission than she ever

288

experienced with any of her prior therapies.

[Slide.

This third case is a 47-year-old male with

advanced DLCL and a mediastinal mass plus bone

marrow involvement. Front-line therapy with CHOP

failed to induce a response, and second-line

therapy with RICE also failed.

After 8 cycles of VSLI, he achieved a PR

according to the IRP or a CR according to

investigator without serious toxicity. He is alive

with no evidence of disease after 30+ months and

has not required any subsequent therapy.

Therefore, the investigator's assessment

of complete remission was correct.

Most striking about this case is the

achievement of a CR in the context of primary and

secondary refractory to chemotherapy. The B

symptoms and anemia also resolved.

[Slide.

I would like now to turn our attention to

our conclusions about the benefits versus the risks

of VSLI.

289

[Slide.

Our patient population consisted of

patients treated with a median of 3 prior regimens,

which translated into 4th and 5th line therapy for

most of these patients.

One-third of our patients had received a

prior autologous transplant and one-third had low

blood counts, which would have made them ineligible

for a treatment with a myelosuppressive agent.

Half were refractory to the last

qualifying therapy and one-quarter were older than

70 years. Finally, two-thirds had an elevated LDH

at time of entry.

In summary the prognostic factors

associated with this population are extremely

adverse.

[Slide.

In this study, we have 25 percent overall

response rate. In patients treated on second

relapse, however, the response rate was better at

46 percent. We consider this response rate to be

clinically important for this population.

290

In fact, I consider this drug to be the

most active single agent I have used since we

tested a phosphamide back in the 1970s

non-Hodgkin's lymphoma. The objective responses

usually translated into symptomatic improvement.

The median response duration was

approximately 3 months and time to progression was

approximately 4 months. This is in the setting of

an overall median survival in the range of 7

months.

[Slide.

Regarding the risks, we can summarize them

by stating that neuropathy was the dose-limiting

toxicity. Its development is gradual and

cumulative, and only 13 percent withdrew because of

neuropathy.

Compared to other agents, VSLI is well

tolerated with a low incidence of severe

myelotoxicity and hospitalizations. In addition,

nausea and vomiting, as well as alopecia are

infrequent.

[Slide.

291

An important point is that the improvement

in symptoms and the antitumor response occur early

with VSLI. This allows for informed treatment

decisions before serious neuropathy develops.

In essence, there is a favorable

benefit-risk profile for this population, which has

not standard therapy options.

[Slide.

So, why do we need VSLI? It is an

effective and well-tolerated agent for patients on

third relapse or later. It offers a therapeutic

alternative for patients who do not qualify for

aggressive combination regimens or who have

relapsed after transplant, as well as for those

with a compromised marrow function.

VSLI resulted in benefits for 1 out of

every 4 patients with minimal toxicity.

Thank you for your attention during the

presentation. We would be happy to answer your

questions at anytime.

DR. MARTINO: Thank you.

At this point, I would like the FDA to

292

present their evaluation of this data.

FDA Presentation

DR. HAZARIKA: Good afternoon. My name is

Maitreyee Hazarika, medical reviewer.

This NDA submission is for Marqibo, which

is vincristine sulfate liposome injection.

[Slide.

The indication is for the treatment of

patients with aggressive non-Hodgkin's lymphoma

previously treated with at least two prior

combination chemotherapies.

[Slide.

This presentation will go through the

regulatory issues with this application. Study

CA99002 is the major trial submitted. Study

DM97-162 is the supportive study. The FDA analysis

of the efficacy and safety will be discussed

followed by the summary and the issues for ODAC.

[Slide.

The regulatory issues for this application

includes accelerated approval, available therapy,

endpoints, adequate and well-controlled trials, and

293

confirmatory trial requirements.

[Slide.

Accelerated approval is granted by the

Agency if a drug appears to provide a benefit over

available therapy, and the benefit is determined by

the drug's effect on the surrogate endpoint deemed

reasonably likely to predict clinical benefit.

[Slide.

Because accelerated approval requires an

advantage over available therapy, the definition of

this term is critical. Available therapy should be

interpreted as therapy that is reflected in the

approved labeling of regulated products.

[Slide.

There are exceptions where a safe and

effective therapy for a disease exists, but it is

not approved for that particular use by the FDA.

In oncology, treatments that are not

labeled for use but is supported by compelling

literature can be considered available therapy.

The ODAC members will need to use their

expertise on what constitutes available therapy for

294

aggressive non-Hodgkin's lymphoma.

[Slide.

Initially, FDA-approved drugs were based

primarily on clinical data and review of

literature. Most of these drugs listed here are

used as part of a combination in relapsed patients.

[Slide.

Within the past 15 years, FDA has approved

four biologic agents mainly for the treatment of

low grade follicular non-Hodgkin's lymphoma. This

is a different indication from the one being

discussed here today. Approvals were based on

single-arm or randomized, controlled trials. The

first three approvals were based on response rates

and duration. Intron approval was based on a

longer progression-free interval and median

survival.

[Slide.

To determine whether vincristine sulfate

liposome meets the criteria for accelerated

approval, the ODAC and the FDA must consider not

only approved drugs, but also the published

295

literature. There are many known approved

combination therapies used for relapsed aggressive

non-Hodgkin's lymphoma, which also includes

high-dose chemotherapy with stem cell

transplantation.

Some examples are shown here. The

combinations used have overall response rates of 30

to 88 percent with complete responses of 18 to 53

percent.

[Slide.

Many single agents are also used that are

not specifically labeled for non-Hodgkin's lymphoma

indication. Based on published literature, overall

response rates varied from 37 to 69 percent and

complete responses from 13 to 33 percent.

In both the combination and the single

agent reports shown here, these may not be the

exact population. Patients may have received

between 1 to 3 prior therapies and may have mixed

histologies.

Nevertheless, we will be asking the

Committee whether any of these constitute available

296

therapy.

[Slide.

Previous recommendations have been to use

complete response as the endpoint in this disease.

Two important questions should be in the

Committee's mind for this application. Should FDA

consider partial responses to be reasonably likely

to predict for clinical benefit in relapsed,

aggressive non-Hodgkin's lymphoma? If so, would

responses of the magnitude and duration seen in

this study predict for clinical benefit?

[Slide.

There are a few key regulatory points to

consider for the present application. The study

should use a design that permits a valid comparison

with a control to provide a quantitative assessment

of drug effect.

The method of selection of subjects should

provide adequate assurance that they have the

disease being studied.

The methods of assessment of subjects'

response should be well-defined and reliable.

297

[Slide.

Since 1999, the Agency has had around 20

meetings with the sponsor on various issues. In

those meetings, FDA has advised the sponsor that

response duration was of interest. FDA advised the

sponsor for the need for a confirmatory trial, and

FDA has emphasized the endpoint of durable complete

responses.

[Slide.

At the March 2003 ODAC meeting, the

Committee reinforced the Agency's recommendation

that the postmarketing studies be ongoing at the

time of accelerated approval. The FDA expects that

confirmatory studies to demonstrate that treatment

with the drug is associated with clinical benefit

will usually be underway at the time of accelerated

approval, although that is not a specific

requirement.

[Slide.

Study CA99002 is the major trial

submitted. It is a multicenter, open-label,

single-arm Phase II study with a primary endpoint

298

of response rate, which included complete response,

complete response unconfirmed, which is a complete

response with a residue in mass, and partial

responses.

119 patients were enrolled. VSLI was

given at 2 mg/m 2

intravenously over 1 hour every 2

weeks.

[Slide.

Patients had relapsed, aggressive

non-Hodgkin's lymphoma, who had received at least 2

prior combination therapies including 1 prior

anthracycline-based therapy.

[Slide.

Histologies included aggressive de novo

and transformed lymphomas, and included diffuse

large B-cell lymphoma and peripheral T-cell and

anaplastic lymphomas.

[Slide.

According to the Central Pathology Review,

only 75 percent patient histologies were identified

as definite eligible, 30 patients were probable

eligible or ineligible. The majority of the

299

ineligible patients had low grade histology on

biopsy. Two patients had slides missing for

Central Pathology to review.

FDA evaluated only the definitely eligible

patients. Definite eligible have the histology for

the indication proposed for VSLI.

[Slide.

These are the reasons for exclusion from

the FDA efficacy analysis. There were 30 patients

with histology not definitely eligible by Central

Pathology Review. Eight patients did not have

baseline indicator lesions measurable by the

independent radiologist.

A durable disease was defined as at least

1 bidimensionally measurable lesion with clearly

defined margins that were greater than 2

centimeters in the largest dimension by CT scan or

physical examination.

Five patients did not receive 2 or more

prior combination chemotherapy from the time of

diagnosis of transformation. Two patients did not

have a washout period of 4 weeks, and 12 patients

300

did not have complete baseline staging with CT scan

or bone marrow biopsies, or had bone marrow

biopsies done more than 2 months prior to the study

drug administration.

That is a total of 40 percent patients

were considered ineligible for the trial by the

FDA.

The briefing document gives slightly

different numbers, but the results do not change

significantly.

[Slide.

There were other study conduct issues,

such as bone marrow biopsies done between 3 to 8

weeks prior to entry. Missing full set of CT scans

at one or more visits, missing scans for tracking

the disease, and missing baseline neurological

examinations. These patients were included in our

efficacy analysis as we felt that they did not

impact on the response rate.

[Slide.

The FDA analysis used 61 percent of the

enrolled patients who met the critical eligibility

301

criteria. That is, they had relapsed, aggressive

non-Hodgkin's lymphoma, had received 2 or more

prior combination chemotherapies, including 1 prior

anthracycline-based therapy, and had required

baseline scans and bone marrow biopsies.

[Slide.

The response criteria used was based on

the International Workshop to standardize response

or non-Hodgkin's lymphoma. There were 4 categories

based on physical examination - lymph nodes, lymph

node mass, and bone marrow biopsy, complete

response, complete response unconfirmed, which

included a residue mass or indeterminate bone

marrow, partial response, and relapse or

progression.

The normal lymph node size was based on

the abnormal node size at diagnosis. A lymph node

greater than 1 cm was considered compatible with

involvement by non-Hodgkin's lymphoma.

These criteria do not require response

confirmation. Most international cooperative

groups require a confirmatory evaluation for the

302

response classification. Although it was agreed to

use these criteria, FDA emphasized that duration of

response must be examined and described.

[Slide.

The sponsor made some modifications to the

International Workshop, such as defining the normal

lymph node size and nodal mass size to be 1.5 cm,

and using indicator lesions that were a minimum

size of 2 cm in at least one dimension.

The FDA analysis of response used the

sponsor modifications.

[Slide.

Prior to any patient enrollment,

amendments included a statement which required

response confirmation by repeat assessment at 4

weeks following the first documentation of

response.

[Slide.

The sponsor's analysis of response rate

included patients with tumor size reduction

documented on at least 1 occasion.

In addition to the same response rate, the

303

FDA analysis was also done on confirmed response

rate where tumor size reduction was confirmed at

least 4 weeks later.

[Slide.

On analyzing response rate documented on 1

occasion, the sponsor found 4 complete responses, 4

complete responses unconfirmed, which to remind you

are complete responses with a residue of mass, and

22 partial responses, for an overall response rate

of 30 patients or 25 percent.

The FDA analysis on the evaluable patients

found 1 complete response, 3 complete response

unconfirmed, and 11 partial responses, for a total

response rate of 21 percent.

[Slide.

The FDA also did an analysis on confirmed

response rates with a greater than or equal to

4-week confirmation, and found zero complete

responses, 2 complete responses unconfirmed, and 9

partial responses,, for a total response rate of 15

percent.

[Slide.

304

In the sponsor's analysis, the duration of

response has been estimated using the Kaplan-Meier

procedure. Patients who did not have documented

progression were censored at the time of treatment

cessation, and the sponsor's analysis of duration

of response, 67 percent of the responders were

censored.

Any attempts to interpret a response

duration where two-thirds of the patients are

censored is of questionable value. The sponsor's

median duration of response was reported to be

greater than 85 days. In the FDA analysis of 11

confirmed responders, the median duration was 85

days.

[Slide.

In the sponsor's analysis, out of the 20

censored patients, reasons for treatment cessation

included neuropathy, relapse, underwent bone marrow

transformation, completed study, withdrew consent,

thrombocytopenia, and unknown reason.

Forty-three percent of the responders in

the sponsor's analysis did not have repeat scans or

305

physical examination or progressed before a repeat

scan was done. Thirty percent of patients

discontinued within 30 days of initial response.

[Slide.

Patients completed a median of 4 cycles of

therapy. The dose intensity was 96 percent of

planned. The most common cause of dose delay was

due to neuropathy followed by hematologic toxicity.

Neuropathy was also the most common cause

of dose reductions. Thirteen percent of the dose

reductions were at least 0.24 mg/m2.

[Slide.

The commonest Grade 3 or Grade 4 adverse

events were peripheral neuropathies, both sensory

and motor, which occurred in 60 percent of

patients, followed by myelosuppression in 45

percent of patients. Other adverse events were

fatigue and constipation.

[Slide.

Study DM97-162 was submitted as supportive

evidence. It was a single-center, open-label,

single-arm study in patient with relapsed lymphoma

306

and acute lymphoblastic leukemia.

The primary endpoint was response rate.

132 patients were enrolled, 116 had a diagnosis of

lymphoma, of which 97 patients had aggressive

lymphoma.

[Slide.

There was no independent review of

pathology or radiology. Selected CT scans were

reviewed retrospectively. There was incomplete

documentation of bidimensional measurements.

Case report forms were not used

prospectively. Standardized response criteria for

non-Hodgkin's lymphoma was not used.

Therefore, the use of the study for

support is questionable.

[Slide.

The sponsor reported the response rate in

the aggressive non-Hodgkin's lymphoma population as

29 percent. There was nor duration of response

assessed.

[Slide.

In summary, the submission is multicenter,

307

single-arm Phase II study in patients with

relapsed, aggressive non-Hodgkin's lymphoma

submitted for accelerated approval based on the

endpoint of response rate.

In the major study, 72 patients met the

critical eligibility criteria.

[Slide.

The FDA analysis found the response rate

documented on at least 1 occasion to be 20.8

percent with 1.4 percent complete responses. The

confirmed response rate was 15.3 percent with zero

confirmed complete responses.

The FDA analysis contains only patients

who have aggressive relapsed histologies.

[Slide.

The study conduct raises doubt regarding

the method of assessment of response. The duration

was short and not adequately evaluated. The use of

the supportive study is questionable for support.

There is no confirmatory trial underway.

[Slide.

We bring this application to the ODAC

308

Committee for the consideration of several issues.

The first issue is the available therapies for

relapsed, aggressive non-Hodgkin's lymphoma.

Drugs considered under accelerated

approval must demonstrate an advantage over

available therapy. The Committee needs to consider

not only the magnitude of the response rate, but

the data which indicates that this response rate is

comprised primarily of partial responses.

The Agency believes that the duration of

any response rate must be considered in assessing

the potential clinical relevance of any claimed

benefit.

Finally, the Committee should consider if

the sponsor has demonstrated in this single-arm

trial that VSLI represents an improvement over

available therapy, keeping in mind the activity of

multiple drugs and combinations in aggressive

non-Hodgkin's lymphoma.

[Slide.

This is the review team for the

application.

309

Thank you for your attention.

DR. MARTINO: Thank you. At this point,

ladies and gentlemen, we will move on to questions

from the Committee to either the sponsor or the

FDA, and also please know that we have chosen to

not allow you a break after the questions as

several of us need to leave, and I want to give the

group the opportunity to at least have their

questions asked and answered.

Dr. George.

Questions from the Committee

DR. GEORGE: I have a question. Maybe

it's for Dr. Pazdur, but it has to do with

accelerated approval again. The more I learn about

it, the more subtle it seems.

This has to do I think with this

application because of the available therapy issue,

the issue of appear to provide some improvement

over available therapy.

This is not, of course, the same thing as

proving it has any advantage over available

therapy, and it just has to do with accelerated

310

approval, because if you were to get full approval

later, you don't have to prove that it is superior

or in any way compared to available therapy.

So, I just want to be sure that that is

correct, and maybe some clarification on what would

be meant by "appear to be superior" or an advantage

over available therapy.

DR. PAZDUR: Let me over the regular

approval is for clinical benefit, and therefore, we

are looking at different endpoints. We would be

looking at a survival benefit. We would be looking

at a very durable CR rate here.

But for accelerated approval, we do want

an improvement over available therapies. Remember,

the whole purpose of accelerated approval, which

came to us from really the AIDS arena, you know,

decades ago, or about 15 years ago, was an effort

by the Agency to get out therapies that were novel,

that were an improvement over available therapy.

So, it isn't appears to be, it should be,

in our interpretation, and provides is the correct

word--thank you, Grant--an improvement over

311

available therapy here.

Now, just from a historical perspective,

you know, our first approval or one of our first

approvals, getting back to this available therapy,

was in colon cancer, a disease that I am very

familiar with, and many of you here are, too, for

irinotecan.

There, in 1995, there was no irinotecan,

there was no oxaliplatin, there was no Avastin,

there was no Erbitux, nothing like that existed.

All you had was 5FU and leucovorin, and that was

it.

Now, it was very easy to answer that

question, because nothing existed and there was

complete consensus in the community that any

therapy was going to be better than nothing

basically.

That is why we felt initially very

comfortable going ahead with this. Similarly, in

lung cancer, where you had patients progress on a

platinum-containing combination, on a

Taxotere-containing combination, we feel very

312

comfortable looking at single agent ERISA, and we

approved the drug.

In therapies where there has been, you

know, activity in available therapy, for example

multiple myeloma, when we approved Velcade, for

example, we had durations of responses that were

over a year, so it was very obvious to us that you

didn't need a randomized trial here.

That is a basic, trying to get away from

the nuances of drug regulation, what we are looking

for in a sense is should a randomized study be

done, and that is the clinical issue here, is there

enough evidence that you have from the literature

that you feel that there is compelling evidence

that there is available therapy that would warrant

a randomized study.

The issue also is "compelling" is a very

vague word, it is like beauty, it's like sexy, it's

in the eyes of the beholder. So, it is very

difficult to establish that, and when we looked at

the literature, we found it very confusing, and

that is why we decided we would ask your opinion

313

regarding this, because when you go back, it is

very hard unless you did a actual meta-analysis of

all of these reports to find out exactly what

therapies patients had, et cetera.

So, it is an area of ambiguity in a sense,

what is available therapy.

DR. MARTINO: Dr. Hussain.

DR. HUSSAIN: I had a question regarding

the comments about symptomatic improvement. Were

the data on symptoms collected prospectively and

systematically? I am really curious. I deal with

solid tumors and we would consider a 10 percent

high-grade neuropathy as an unacceptable rate of

neuropathy.

So, I am curious about your comments about

how this drug is so well tolerated, yet, you have a

60 percent Grade 3 or 4 neuropathy, which does not

really speak for a good quality of life in a

setting of this kind of disease.

MS. MANCINI: Yes, I will address your

first question regarding how the symptoms were

collected. Can I clarify, are you asking about the

314

symptoms of neuropathy or the symptoms of disease?

DR. HUSSAIN: I believe the doctor was

asked about quantifying clinical benefit, and we

saw a slide that showed there were improvements in

performance status and symptom improvement, and so

my question is, did you actually go back and look

in the charts and see that the doctor mentioned

something, or there was systematic collection of

the symptoms prospectively.

MS. MANCINI: I can answer that question

then. As I commented, we had no prospective

symptom endpoint in this trial, so therefore, these

are soft data. What we have is what was reported

as baseline signs and symptoms, so as part of

collecting adverse event data. We would watch what

happened to that as the patients were treated, so

you could see as it gets better basically, and as

it is resolved.

So, it is not the same level of evidence

we would normally have for a formal efficacy

symptom endpoint, but the reason we summarized that

data was we were asked to look to see was there any

315

evidence to support that the patients who got

partial responses particularly had other clinical

benefit.

So, it is soft data. It is the level of

data, however, that you would experience in your

clinical practice, the patient complains of tumor

pain, neck pain, et cetera, and then it goes away,

so it is at that level.

DR. PAZDUR: If I could just add to that

because I think I was the person that instigated

this question. We don't put any credibility or

very little credibility in a unblinded single-arm

trial as far as interpreting symptoms.

One of the reasons I just asked this, was

there any kind of supplemental information that

could give some credibility to a response rate,

because a response is, one, you know, even when we

talk about a complete response rate, is it a single

mass going down versus multiple masses, is there

something that might bolster this up, but the level

of proof that we would require to make any labeling

claims about that in any product would not come

316

from a single-arm, unblinded trial.

MS. MANCINI: If i could return to your

second comment, then, I would just like to clarify.

If we could back in the primary presentation,

please, to the slide--excuse me while I flip to it,

to the safety presentation--let's go to Slide 71,

please.

I may have gone quickly over this in the

presentation. There are two factors to consider

when we look at the tolerability of our drug. One

is how did the patients come into the study and

what happened from there.

So, just as a point of perspective, 86

percent of them had had at least 2 prior regimens

containing a neurotoxic agent, so 85 percent came

in with some level of neuropathy.

The next slide, please, 72. This is the

analysis that shows what happened to them by their

worst grade at study entry. So, when you look at

the people who get to Grade 3, they are mostly

coming from patients who started some neuropathy.

So, the incremental toxicity that they

317

have endured is what we are trying to summarize

here. So, for patients who came in with a Grade 2

neuropathy, half of them did worsen to Grade 3.

That is a one-grade change. Had they not come in

with Grade 2, they wouldn't have gone there

probably. So, in patients who came with no

neuropathy, we don't see that high a proportion

going to Grade 3.

I can also share with you the perspective

of if we go to, again in the primary, it is just a

couple of slides further, Slide 75. I think this

is the data that you were referring to--Slide 74, I

am sorry, my eyes are not good.

It is correct to say that a third of the

patients on this study got to Grade 3 or 4

neuropathy, and mostly Grade 3, but as I mentioned

earlier, the withdrawal rate due to neuropathy was

only 13 percent, therefore, Grade 3 neuropathy was

not a reason to stop treating in many patients.

I do have some additional data if you

would like to know what happened to these 37

patients, what the treatment decisions were. Would

318

you be interested in that? Okay.

If we go to Slide 739, please, in the

backups. This is now going to summarize all 37

patients who ended up with a Grade 3 or 4

neuropathy on study.

The first point is that 10 of those

patients had a simultaneous declaration of

progressive disease, therefore, there was no

further treatment decision to be made for those

patients.

So, of the 27 remaining patients, 11

withdrew for neuropathy or withdrew consent, so to

be conservative, we are taking the 7 and the 4, and

saying 11 patients probably withdrew due to

neuropathy, but 16 patients continued with VSLI

therapy. So, just over half of them chose to

continue despite getting to a Grade 3.

The next slide, 740, please. This shows

you for those patients who continued, for the 16

patients, 8 of them had 1 additional dose, and I

have a footnote here to say why did they stop, it

is not because they weren't tolerating it again, it

319

is because 7 of them had progressive disease after

the next cycle.

So, the reason to stop treating in this

study was almost always disease progression. There

are patients who were at Grade 3, but continued,

got 3 or 4 more additional cycles.

So, we conclude that although we did want

to present the total Grade 3/4 neuropathy as being

clinically important, it did not always trigger a

decision to stop treating.

DR. MARTINO: Dr. Wilson.

DR. WILSON: I would like to get some

better idea about how the patients were evaluated.

Of course, PR and CR per se, In the absence of a

meaningful length, is not really worth much. The

overall length of response is somewhere in the

3-month arena.

What did the protocol state in terms of

how often patients had CT scans done? The

documentation says that CTs were required on study

every 4 weeks, but in those patients in whom

therapy was actually stopped, were CT scans also

320

done every 4 weeks, and for those who didn't have

it done every 4 weeks, was this factored in, and

were those patients censored?

MS. MANCINI: Thank you for that question.

That gives me an opportunity to clarify a point in

that in our duration of response analysis, if I

could just have Slide 266, please, just to

demonstrate this point.

The duration of response was calculated in

a classic manner in our study. Start of treatment

is over here on the left, and then the first formal

evaluation of response was after about 6 to 8 weeks

on study.

Then, response was considered to be

continuing until you either had documented

progression or you were lost to follow-up, the

patient was lost to follow-up while still in

response. This happened very seldom.

So, if a patient stopped being treated in

here because of neuropathy, that did not matter.

We continued gathering the CTs. Now, the schedule

for CTs was originally every 8 weeks in the

321

protocol. We then amended it after about half of

the patients had been enrolled to say after the

first evidence of response, please get another set

of CTs 4 weeks later, and then we are back to the

every 8-week schedule.

DR. WILSON: So, I think it is fair to

say that with a median of approximately 3 months,

that many of these patients, if you were only

obtaining scans every 8 weeks, the true median, had

you been doing it on a more frequent basis, might

have even been shorter?

MS. MANCINI: I will ask Dr. Cabanillas to

comment on that in a moment. Many of the

progressions were declared based on the clinical

evidence. We had physical examinations, clinical

visits every 2 weeks on the study.

DR. CABANILLAS: I would like to also show

the Kaplan-Meier curve, Slide No. 270, please,

because it is true that many of these responses

were short, but as you will see from this slide,

those are these patients that relapse early, but

there are a number of patients that actually did

322

enjoy longer disease-free survival.

This is an IRP review. The IRP met at one

point, and they don't keep on meeting to update the

curve, so we don't have obviously a prolonged

duration here, but we do have some patients, and I

mentioned a few of them, and we have some that are

here today that will also show you that they were

actually some of these patients that had very long

remissions.

So, it is true, that if you look at these

patients that they had short remissions, but we

also have to keep in mind the other side of the

coin, which are the ones that did have long

remissions.

The other point that I would like to make

is that, as you know, Wyndam, there has been some

controversy regarding the use of CT scans in

lymphoma, and there is one study in which 39

patients were evaluated with the CT scans to

determine whether they relapse or not. Actually,

those were 39 patients who relapsed, I am sorry,

those were 39 patients who relapsed, and only 2 of

323

them were picked up by CTs, by routine CTs.

Most of them were picked up because the

patients developed symptoms or the LDH went up, or

physical examination showed the abnormality, so

even though intuitively, you might think that, yes,

doing more frequency, these might have detected

earlier relapses, I think that in real life, that

is not really what is happening.

DR. PAZDUR: Fernando, could you put that

slide back up, and tell us how many patients were

censored there or still on study?

MS. MANCINI: Yes, that's Slide 270. This

is the IRP analysis, and the circles are the

censored data. In the IRP analysis, there were

two-thirds of the patients as Maitreyee presented.

DR. PAZDUR: Two-thirds were censored.

MS. MANCINI: In this analysis, yes.

DR. PAZDUR: And did the rest progress?

MS. MANCINI: Yes, or continued.

DR. PAZDUR: How many continued?

MS. MANCINI: I am sorry, I am saying it

the wrong way, excuse me. Censored is continuing,

324

the response is continuing at the last time that

they were evaluated. When they are censored, there

is no data for the patient beyond that point, but

they have not been called a failure, they have not

been called a progression.

DR. WILLIAMS: But many of those patients

will never get another evaluation, right?

MS. MANCINI: That's correct.

DR. WILLIAMS: How many patients actually

still are on study and could contribute to ongoing

data versus how many are censored because of

neuropathy, or they went off study, et cetera?

MS. MANCINI: This is an Independent

Review Panel determination. The study is over, so

there is no further data coming, but I can show you

Slide 275, please. Because we have continued to

follow, as part of the survival update, those

patients who were still in response, and the reason

the other curve had this line here was because of

the censoring, we all understand that.

This is now the investigator duration of

response, which again the median was not different,

325

but you do see that there were some patients that

had longer durations.

DR. MARTINO: Dr. Perry.

DR. PERRY: This is for the sponsor. I am

just sort of a country doctor who doesn't do a lot

of lymphomas, but it seems to me that this turbo

vincristine is not going to be fairly compared

against a lot of other agents.

What happens if you go back and look at

vincristine itself, what kinds of responses do we

see in duration of responses if we go back to, say,

how in 1972, I think that is the kind of comparison

I would like to see. I think you are making a bad

comparison for yourself if you are not going back

to the original, because I can't imagine this drug

is going to have, from what you presented, is going

to have much of an impact as a single agent. It is

going to be used in combination to replace

vincristine.

So, if that is the case, does it do

compared to vincristine?

MS. MANCINI: That is very difficult to

326

do. The vincristine is a very old drug and the

literature from that time does not provide us the

data we would need to be comparing durations of

response.

Even the types of disease, as it is

described back in the 1960s, it is impossible to

compare to some of that old work. We have been

able to compare on a safety basis better, but not

very well on the efficacy basis.

We do have perspective, there was work

done in the early days by Dr. Don Jackson with

infusional vincristine, attempting to achieve what

we achieve with the liposome in terms of the

continued exposure, and we certainly can compare

our information to that. I will ask Dr. Cabanillas

to do that. There is no good efficacy comparison,

unfortunately.

DR. CABANILLAS: While you find it, let me

introduce a topic. Dr. Jackson made a study with

infusion of vincristine as a single agent a long

time ago, and he had some interesting findings.

He treated 25 patients with a variety of

327

NHL types. He used a 5-day continuous infusion at

0.25/m 2 after a bolus of 0.5,

and he repeated the

courses every 3 weeks.

The interesting thing is that he had 48

percent neurotoxicity, and the GI toxicity was the

most serious toxicity, which included severe ileus.

Hematological toxicity was minimal. Of course,

during those days, there were not too many agents

that you could offer to the patients, so many of

these patients were actually being treated at first

relapse, so it is not strictly comparable to what

we are showing. We are showing really 4th and 5th

line treatment.

He showed that there was an ability to

deliver high cumulative dose of vincristine with

the low GI toxicity with VSLI. That's his

impression when we showed him our data, and he

thought that the absence of ileus was highly

unusual because that was a very common complication

when he used the high dose infusion of vincristine.

Also, the ability to proceed to transplant

following monotherapy, he found also to be highly

328

unusual.

Now, I think that is a very important

question.

DR. PERRY: Slide 805, response rate.

DR. CABANILLAS: That is the problem. The

response rate, it's a mixture of different

histologies treated at first relapse, so it is

really impossible to compare with ours.

Do you remember exactly what the response

rate was?

MS. MANCINI: It is not literature that we

can compare on efficacy other than we did meet with

him to get his own impressions, and that is what we

are sharing with you here.

I think the best data that we can share,

that is the only data that we can share that is

head to head comparison, is actually preclinical

data. There is no efficacy publications of

singe-agent vincristine. It is always used in

combination.

If you would like we can show you some

preclinical data.

329

DR. MARTINO: Is there someone in the

Committee who actually has the answer to that

question?

DR. WILSON: Actually, they are right that

it was a mixed group, but keep in mind, at least

for the intent-to-treat study here, it was a mixed

group, as well. I believe one-quarter of the

patients were not felt to have a de novo large cell

or a transformed large cell.

If I recall that paper right, the response

rate was around 30 percent, and I also want to say

that one of the things that he did in that paper

was to take patients who he considered to be a

failure using bolus vincristine, and I just want to

reiterate that I think the patient population in

the study was very, very different from what we are

dealing with here, but I do think that it did give

the first hint early on that perhaps changing

exposure schedule may give you a better therapy.

DR. MARTINO: Thank you.

DR. CHESON: We just got the Jackson

abstract up.

330

MS. MANCINI: Thank you.

DR. CHESON: Nodular PL, diffuse

histiocytic, duration of response up to 16.4

months, median 4.4. Complete response and

histiocytic lymphoma and partial in 8 patients, so

9 out of 25 responded, 36 percent. You said 30, I

was over saying 40, so 40 percent response rate

lasting a median of 4.4 months with a variety of

histologies including diffuse large B-cell, diffuse

mixed, diffuse everything.

MS. MANCINI: Did they comment on other

significant predictors of the number of prior

therapies they have had or refractoriness of the

disease?

DR. CHESON: This is just the abstract.

All had received prior vincristine by conventional

bolus. It doesn't say.

MS. MANCINI: That is the difficulty we

have with interpreting the data. In our own

analyses, these were very significant predictors of

outcome, and therefore, it is very hard, as we have

all seen, it is very hard to compare to the

331

literature.

DR. CHESON: But even your best patient

group, it was only around the same response rate.

DR. CABANILLAS: It was 46 percent

response rate on patients treated after second

relapse. These are not first relapse patients.

The minimum acceptable for this study, the VSLI,

was that the patient had to have at least 2

relapses, so these are not really comparable, but

the 46 percent, I think actually compares favorably

with anything in the literature including Rituxan.

I think that it is important to point out

that vincristine is a cell-cycle activation as you

all know, and that is why, even after long exposure

to vincristine, it might actually result in a

higher response rate.

I think that might also explain why

Jackson's results are somewhat better than you

would expect, and also the fact that some patients

were refractory to vincristine, responded to it,

and I think that is what we are seeing also. We

had some patients that were clearly refractory to

332

CHOP, and yet they were able to go on to respond

even, some even with a complete remission.

DR. MARTINO: Thank you.

Dr. Brawley, do you still have a question?

DR. BRAWLEY: Yes. Again, at risk of

being politically incorrect, and without making any

allegations against anyone, would the academic

presenters be willing to disclose any potential

interest they might have in the company including

such things as stock ownership, honoraria, or are

they salaried by the company?

I am also interested in did the company

salary the investigators at the various sites that

put people onto the trial.

DR. CABANILLAS: I think that is actually

a very politically correct question. Let me

explain my interest in the company. When we did

the Phase IIa study at M.D. Anderson, Dr. Ceres

[ph] was the PI on that study, and he realized that

being able to give the drug every 2 weeks at a high

dose constituted a novel way of delivery

vincristine, so in the name, representing M.D.

333

Anderson, he applied for a use patent, which has

been issued to M.D. Anderson, and which we will

also share with M.D. Anderson. I am not a salaried

employee of the company, and, of course, I do

charge consulting fees.

MS. MANCINI: None of the investigators on

our trials are salaried.

DR. CHESON: You don't know about stock

ownership, I suspect.

DR. CABANILLAS: I don't have any stocks,

and I don't have any stock options either.

MS. MANCINI: Inex is a small

pharmaceutical company in Canada. We are not

traded in the U.S. We are on the Canadian stock

exchange, Toronto stock exchange, and people are

free to buy stock if they wish. We do not track

their participation in the Canadian market.

DR. PAZDUR: I would just like to

stipulate that submitted to the IND was the

financial disclosure, which did list Dr. Ceres and

Dr. Cabanillas, I think it was that the patent was

going to be allowed, or something like this, plus

334

the royalties to be paid to M.D. Anderson.

We did go back and check the informed

consent at M.D. Anderson, and it did, in our

estimation, provide adequate explanation and

adequate patient protection to the patients that

were enrolled on the study.

DR. MARTINO: Thank you.

Dr. Reaman.

DR. REAMAN: I just have a question about

the retrospective histopathology review. You cite

the incidence of ineligible patients, about 19 or

20 percent, to be similar to previously reported

studies, some of which were actually done 12, 20

years ago, none of which looked to demonstrate

efficacy of a single agent in a specific disease.

Can you explain why the retrospective

review was performed and why those ineligible

patients were then included?

MS. MANCINI: Yes, I will begin briefly

and then I would like to ask Dr. Gascoyne to

comment on what is the current finding in lymphoma.

Just to begin, why did we include a

335

retrospective review. We understood that getting a

correct histologic diagnosis in clinical practice,

in routine clinical practice, it is difficult to

get right, and Dr. Gascoyne will speak to that more

eloquently than I can.

Therefore, we built into the protocol that

there would be a retrospective review. It was not

the basis for allowing a patient to be enrolled

because we could not get a real-time pathology

assessment. This was an international trial. So,

we use the site pathology to allow patients to be

enrolled, and they were allowed to be treated.

Even if they were subsequently determined to be

histologically ineligible, we did not withhold

treatment for them.

I should also comment that all of the

patients were considered to be histologically

eligible by the treating sites, so what we are

dealing with here is the current situation

lymphoma, and I would like to ask Dr. Gascoyne to

comment further.

DR. GASCOYNE: The first comment I would

336

make is I am slightly biased, because as a

pathologist, and someone who has been doing

lymphoma for 20 years exclusively, I personally

think it all starts and end there with an accurate

diagnosis.

We spent the last 20 years trying to

determine that lymphoma is not one disease, it is

about 35 diseases, and I think to continue to mix

apples and oranges in any kind of trial is a

mistake. So, I quite frankly applaud them for

asking us to be involved in a study where we review

the pathology.

I think the issue of path exclusion speaks

to an issue that is even bigger than this trial,

and it is a problem with what is going on out there

in the community and the acceptance of expertise.

So, I think all of these things need to be

looked at. We excluded cases that we felt were

low-grade lymphoma, and I could sit up here and go

through the reasons why that occurs. I noticed

that the FDA was not willing to accept 7 cases that

we have looked at very seriously, and I will tell

337

you that in current trials, I am currently, and

have been for 5 years, the co-chair of the Eastern

Cooperative Oncology Group from the point of view

of lymphoma although I am a Canadian who lives on

the West Coast, which is a bit of a funny mix.

The kind of criteria we apply to those

ongoing current types of trials are what we applied

here. So, I don't think we were any more or less

rigorous as we applied those criteria. The cases

that we accepted were based on FNAs and needle core

biopsies, follow-up biopsies, but you have to

remember that those were not diagnostic biopsies in

patients with newly diagnosed lymphoma.

We are talking about patients in a

multiply relapse setting. To even get a few cells

at the end of a fine need aspirate is lucky. We

use that information in combination with growth

fraction and other types of ancillary studies in

order to arrive at a diagnosis.

Do I have confidence in those 7 patients

that you decided to exclude, that we left in, the

answer is yes, I would confidently stand here or

338

anywhere else and say that those patients deserved

to be in the trial, and to have been included as

being consistent with the aggressive histology

lymphoma.

DR. WILLIAMS: Can I just ask why you used

the word "probably" then instead of "definitely?"

DR. GASCOYNE: Because in studies like

that, we don't have architecture basically, so you

are talking about cytological detail, and when we

had to combine that, a few of those cases, if you

go back and look on the additional data, we went

back and were able to actually retrieve the

original pathology.

So, they didn't look like discordant even

based on an aspirate, and thus we re-reviewed some

of those diagnoses. I remember one of the cases in

particular we had said it's a needle, it's kind of

crushed, it is hard to interpret, but it was

accompanied by a growth fraction of 80 percent.

When we got the older pathology and there was

primary mediastinal large B cell lymphoma, we were

willing to accept that, in fact, that represented

339

aggressive disease.

That is not what I would call out of line

with current practice. In fact, those kind of

techniques are being used much more frequently here

in the United States than they are in the country

that I come from.

DR. MARTINO: Dr. Cheson.

DR. CHESON: Well, now that I am on, I

have got two questions, one for Randy. So, you

used other than morphology and growth fraction, I

presume. You immunophenotyping and all that stuff,

as well.

DR. GASCOYNE: In the actual diagnoses, we

had information that came from the hospital of

origin, so what we were trying to determine was

their eligibility as aggressive lymphoma, so in

many of the instances, of course, we had phenotype

available, either flow cytometric or paraffin

section immunostains.

But the particular example I was citing,

we actually were aware that there was a growth

fraction, and that was provided in terms of a Ki-67

340

stain. That was labeling at 80 percent. That is

not the labeling presented that one sees in

so-called indolent lymphoma, and I think most

people in practice nowadays would accept that as a

reasonable conclusion based on that material, that

that, in fact, was aggressive lymphoma.

But the ability or the desire here to use

fine needle aspiration and needle core biopsies, as

you know, is a serious matter, but you have to keep

it in the context that we are not dealing with

diagnostic biopsies, where it also is used far too

frequently.

DR. CHESON: My other question is more for

the sponsor, and I guess I can guess what the

answer is. Do you have any information on whether

these patients were treated with other things after

liposomal vincristine and what the responses were

to subsequent therapies, and what these other

therapies were?

MS. MANCINI: No, unfortunately, once they

were off study, we don't have the downstream

therapy. We do know of those 6 cases that went on

341

to stem cell transplant, we do know the outcome for

those patients, and that is what Dr. Cabanillas has

presented.

We collected that data as part of the

survival update, but we do not have all the other

therapies.

DR. MARTINO: Dr. Bishop.

DR. BISHOP: Related to what your last

statement was, the stem cell transplant patients,

and going back to your time to progression slide,

am I correct that there are only 5 patients on that

slide who were greater than 4 months time to

progression?

MS. MANCINI: No, no, there is more than

DR. BISHOP: Those were just the censored.

MS. MANCINI: Those were censored, yes,

there is definitely more. The line was flat at

about 50 percent. We were right at the median

there.

DR. BISHOP: So, you are estimating about

10 patients?

342

MS. MANCINI: Fifteen.

DR. BISHOP: Among those, does that

include the transplant patients?

MS. MANCINI: Transplant patients would

have been censored in that analysis at the time

they went to transplant.

DR. MARTINO: Dr. Wilson.

DR. WILSON: This is a question for both

Randy, as well as the sponsor. It is a little

unusual, in my experience, to do an intent-to-treat

for a Phase II study. A number of cooperative

group studies were shown in which it was shown that

only 80 percent of the patients were eligible based

on histology, and yet when those Phase II studies

are usually published, those 20 percent are

actually not included, so I wasn't quite sure why

that was done like that.

But specifically, Randy, I, too, feel that

we shouldn't be fixing apples and oranges, and I

think that the response to an agent like this, in a

low grade lymphoma, is going to be very different

from a de novo large cell.

343

I guess my question is, number one, do you

really think it is fair to be mixing de novo large

cell with those patients who have histologically

transformed low grade, both, number one, because

the biology of the histologically transformed is

very, very different.

Even though it may look like a high grade

lymphoma, it certainly may not clinically act like

one, and finally, you have some nodes that are low

grade and some that are high grade, so it is very

difficult to know if it's a low grade node

shrinking or the high grade.

DR. GASCOYNE: I can address the last part

of the question. I mean you certainly know

yourself that when you are treating patients and

one site blows up, it wouldn't be uncommon for you

to do a needle aspirate or a needle core at that

site, and make a determination that that previous

lymphoma has, in fact, transformed. We don't

biopsy all the sites and all the patients.

From the old data from the seventies, from

the stage and laparotomy data that came out of

344

Stanford and other centers, we know that there are

discordant histologies. I would agree with you

those biologies are different.

I wasn't involved in the design of that

part of the study. I was simply asked to apply

vigorous criteria to a histologic central path

review and deem which patients were eligible based

on the criteria.

I think we applied that in a uniform way,

the same way I do for ECOG and the same way I do

for British Columbia, and so I don't have any

problem, and I am feeling comfortable about those

data.

DR. WILSON: So when you did score

somebody as a histologically transformed low grade,

they would have been essentially what we now term

as follicular Grade 3B?

DR. GASCOYNE: No. We tried to determine

in any setting in which we had tissue, of course,

we wanted to know architecture, so we wouldn't have

accepted that as transformation, and specifically

that is in the protocol that actually Grade 3B

345

would not have been included.

So, these are patients that we had to have

evidence that they were diffuse large B. It is

quite true that there were some fine needle

aspirates and some needle cores in there, in which

we attempt in some situations to actually do some

additional immunostains in the way of looking at

FTCs, et cetera, to know whether there is any

underlying follicular elements.

So, I am happy that those cases that we

looked at, particularly in regards to looking at if

they are only cytologies, looking a proliferation

rate, that I don't think Grade 3B is associated

with an 80 percent proliferation site, so I think

we can feel relatively comfortable that those were

probably cases of diffuse large B cell lymphoma

that represented transformations.

DR. CABANILLAS: I want to make a comment

regarding your question also, Wyndam, because you

are absolutely right, that the low grade or

indolent lymphomas respond differently to VSLI, but

the way they respond differently is they respond

346

less well because when we did the Phase IIa study

at M.D. Anderson, we found out rapidly that the low

grades were not really responding well, and that is

the reason why the next trial was done exclusively

with aggressive lymphomas.

It makes sense from a scientific

standpoint you would expect that they would respond

lower, because their S phase is lower, so they are

turning over less rapidly and exposing them to a

long duration of concentration of VSLI might not

really make any difference.

DR. WILSON: So, that would be consistent

with the Taxol data, as well, where it is also

hitting the microtubules where the higher grades

seem to do better with it.

MS. MANCINI: I would like to add just to

clarify, you asked a question about the

intent-to-treat analysis and why we presented that,

and if I could go to Slide 429, please. I would

like to also, while we are pulling up that slide,

comment that we did look specifically in response

to a question from FDA at the discordant lymphoma

347

cases where they had a mixed presentation to see if

we could conclude which parts, which types of

histologies were responding, and it was always

clear-cut. The patient either didn't respond and

therefore it was not an issue, or it was such a

dramatic response that all of the disease was

responding consistently.

This slide shows you, on the far right,

was the intent-to-treat analysis that was in the

main presentation. This now shows the per-protocol

population, and the ORR is the 27 percent, which is

very similar to the ITT.

Now, if we talk about those patients who

were histologically ineligible, based on Randy's

review, it was 23 cases, the response rate was

still the same in this trial. People who were

excluded from our per-protocol population for other

reasons had a slightly lower response rate.

Open Public Hearing

DR. MARTINO: Are there other questions?

If not, we will move on to the next part of the

program, which is the open public hearing. Before

348

we do that, I need to read a statement.

Both the Food and Drug Administration and

the public believe in a transparent process for

information gathering and decisionmaking. To

ensure such transparency at the open public hearing

session of this Advisory Committee meeting, the FDA

believes that it is important to understand the

context of an individual's presentation.

For this reason, the 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 had with the sponsor, with its product,

and, if known, its direct competitors.

For example, this financial information

may include the sponsor's payment of your travel,

lodging, or other expenses in connection with your

attendance at today's meeting.

Likewise, the FDA encourages you at the

beginning of your statement to advise the Committee

if you do not have any such financial relationship.

If you choose not to address this issue of

349

financial relationship at the beginning of your

statement, it will not preclude you from speaking.

Ms. Clifford will now announce the ladies

and gentlemen who have asked to speak.

MS. CLIFFORD: When I call your name, if

you would please approach the microphone in the

back in the audience, please.

Our first speaker is Helen Smith.

MS. SMITH: Good afternoon. My name is

Helen Smith from Greenbrae, California near San

Francisco.

I would like to thank you for the

opportunity to speak at this meeting. I would also

like to thank Inex Pharmaceuticals for covering my

costs to attend this meeting.

I am a patient of Dr. Jennifer Lucas,

Merin [ph] Oncology where I am being followed for

my non-Hodgkin's lymphoma. I was diagnosed with

non-Hodgkin's lymphoma in 1999. I remember it was

in June when I first noticed I became tired very

easily. I was usually very active.

After my diagnosis in December '99, I

350

participated in a cancer vaccine that was being

conducted by Dr. Levy at Stanford University

Medical Institute. They took one of my lymph nodes

to make the vaccine and then I started a course of

CHOP.

At the end of CHOP, I was given the

vaccine. My cancer did not completely go away.

Later, I had a second round of CHOP. The CHOP did

not make me sick, but I did lose my hair and wore a

wig for a while and felt very tired.

They also gave me medicine to prevent my

blood count from going too low. My CT scans

indicated that I had a lymphoma in my chest and in

October of 2001, I agreed to begin the trial with

VSLI.

I was treated with VSLI during the first

part of 2002. I had 8 cycles of treatment and

during my first treatment I had a fever, but after

that I was fine for the later injections. I did

get some tingling in my fingers and numbness in my

feet, and had difficulty with buttons.

Today, two and a half years later, I am

351

functioning quite well, although I do get tired

easily. I have not had any more cancer treatment.

I have not missed any of my tenpin bowling

sessions, and I use a heavy 14-pound ball, so my

fingers are fine. My only problem is my bowling

average has fallen from 150 to 121.

Thank you.

DR. MARTINO: Thank you.

Next, please.

MS. CLIFFORD: Our next speaker is

Virginia McCormick.

MS. McCORMICK: Good afternoon. My name

is Virginia McCormick and I am from Sparta,

Tennessee, and I think you for the opportunity to

speak at this meeting today. I thank Inex

Pharmaceuticals for covering my costs to attend

this meeting and for all the help that they have

given me.

I am a patient of Dr. Deng [ph] at the

M.D. Anderson Cancer Center in Houston, Texas,

where I am followed for my non-Hodgkin's lymphoma.

Let me begin by telling you that I was

352

diagnosed with non-Hodgkin's lymphoma in 1999. The

first year I took CHOP. The chemo made me weak and

nauseous at time, and, of course, I lost my hair.

This took several months because most of the time

my white cells would get too low and they couldn't

do the chemo, and I would have to go back home and

wait a week and then go back.

When I finally finished the treatment,

this was about six months later, the lymphoma had

returned, and I took other treatment, such as

Rituskin [ph], and this didn't work. During this

time, I had several bone marrow biopsies to see if

it had spread to my bones, and I was thankful that

it had not.

At this point, they wanted to do a bone

marrow stem cell transplant using my own cells, and

my stem cells were harvested and frozen, and I went

into the hospital, and for the first week I took

high-dose chemo and then they gave me my stem cells

back. Again, I was very weak and sick at times.

I was in the hospital for over a month and

within a few months after this, the lymphoma was

353

back again. I asked my doctors if there was

anything else that could be done there. I was told

that it would be too dangerous to do further

treatment because it could do some damage to my

main organs or possibly death.

At this time, my husband and I decided to

go to M.D. Anderson Cancer Center in Houston,

Texas. We just refused to accept the alternative.

At M.D. Anderson, my doctor, Richard Champlin [ph]

told me I could do a bone marrow transplant using a

donor.

So, I asked what my options were, and he

told me that they were having 30 percent survival

rate. So, I told him this was better than what I

had, because I didn't have any.

So, out of 6 of the blood donors, half of

them were a perfect match. So, Dr. Champlin

introduced me to Dr. Deng, who gave me a new

treatment, an anticancer drug called liposomal

vincristine or VSLI. This helped me a great deal

and I was able to receive my bone marrow

transplant.

354

The VSLI was an easy treatment for me to

take. I went into the hospital for one-hour

treatments, which they would watch me for a couple

hours afterwards, and then I could leave the

hospital.

The VSLI made my fingertips numb and the

bottom of my feet were numb, but my hands are okay

now, and my feel are getting better. The VSLI,

with it, I didn't feel weak or sick or anything,

and my hair didn't come out. Since this treatment,

I feel normal again, I feel better than I have in 6

years.

I can go to church, I sing in the choir.

I do my own shopping, and I play with my first

little grandson, who is 3 years old, and which I

might not have been able to do if it had not been

for VSLI.

When I was asked by Dr. Deng if I would be

interested in going to Washington, D.C. to speak at

this meeting, I thought, wow, this is an

opportunity of a lifetime to help so many people,

like it has helped me.

355

So, I am a living example that VSLI does

work and I just praise God that I have been

cancer-free for over two and a half years.

Thank you.

DR. MARTINO: Thank you.

Our next speaker, please.

MS. CLIFFORD: Barbara Cruse.

MS. CRUSE: Hi. My name is Barbara Cruse

and I live in Sugarland, Texas, outside of Houston.

I would like to thank Inex for inviting me

to speak at this meeting, and they have covered our

expenses to be able to come.

My cancer journey began in June of 1997

when I found a lump. I was diagnosed with Stage I

large B cell aggressive lymphoma. I had surgery to

remove the tumor, then 6 rounds of CHOP and 25

treatments of radiation.

On Christmas Eve 1999, my doctor told me

that I had a recurrence of the lymphoma. in

January of 2000, I began treatment at M.D. Anderson

Hospital for the lymphoma preparing for a bone

marrow stem cell transplant using my own stem

356

cells.

I received 3 rounds of chemotherapy

preparing to rid my body of the lymphoma. I was

admitted into the hospital in May for 7 days of

intense chemotherapy to prepare for my transplant.

I spent 5 and a half weeks in the hospital.

Fifteen months later I had a relapse

again. Dr. Fayad, my lymphoma doctor, at M.D.

Anderson was researching an exciting new trial,

which was the VSLI Phase II clinical trial. I was

accepted into the trial and began treatment in

September.

The treatments were done in an infusion

suite at the hospital and lasted 4 hours per

session. I had no side effects during the first 3

treatments. The good news, after 4 treatments, I

was restaged and I had a 96 percent reduction of my

tumor. The bad news, severe neuropathy in my hands

and feet.

In November I had 2 more treatments

preparing me for my second bone marrow stem cell

transplant using my brother's stem cells. I was

357

admitted to the hospital on December 13th and

received 3 days of chemo and then had my

transplant. I did so well with this transplant

that in two and a half weeks, I was released from

the hospital.

As I mentioned earlier, the down side of

the VSLI is the neuropathy. I was not able to

drive for a year and I needed help with daily

activities. I had an EMG to determine how much

nerve damage I had and to see if there was anything

that could be done to help me.

There is no drug that works for

neuropathy. I was told that all the previous

chemotherapy treatments that I had had, had

contributed to my neuropathy.

Finally, in March of this year, I began

going to a doctor and receiving acupuncture of the

neuropathy, which I have seen remarkable

improvement. This summer when I had my checkup

with Dr. DeLema, my bone marrow doctor, we talked

about my treatment options and if I had it to do

over, would I choose the VSLI, and I told him

358

honestly yes, I would choose it even with the

neuropathy.

In my cancer journey, I am one of the

lucky patients who survived with the help of this

very important drug VSLI. Thank you for this

opportunity to share my story.

DR. MARTINO: Thank you.

Are there any final questions from the

panel to either the FDA or the sponsor?

Seeing none, I would like to ask Dr.

Pazdur, do you mind if I give the group a 5-minute

break only, because Dr. Martino needs one? Thank

you. It is 5 minutes, however, ladies and

gentlemen.

[Break.]

Committee Discussion

DR. MARTINO: The final portion of this

meeting is the discussion of the Committee itself,

and that needs to be focused to questions that have

been posed to the Committee from the FDA that

relates to this application.

We have four questions, each of which will

359

require a vote at the end of the discussion that

pertains to the specific question. When we vote, I

will ask each of you to state your name as well as

your vote each time, please.

The first question I will read to you.

Review of the oncology literature suggests

that there are single agents and multiple agent

therapies capable of producing substantial response

rates, including reasonable CR rates, in relapsed,

aggressive non-Hodgkin's lymphoma.

Does the Committee believe that these

therapies constitute available therapy for

relapsed, aggressive non-Hodgkin's lymphoma

previously treated with at least two combination

chemotherapy regimens?

Dr. Cheson, I am going to ask you actually

if you would speak to this issue. Are there prior

therapies, either single-agent or multiple-agent,

that you feel would be an alternative?

DR. CHESON: As you know and as was shown,

there are very few drugs that have been recently

approved for the treatment of lymphoma, but there

360

are a whole bunch of other drugs out there, which

are used regularly in a variety of histologies of

lymphoma, some which have been around for a long

time and others which are relatively new.

I think that there are clearly a variety

of other single agent--it depends on whether you go

along with the company's date or the Agency's

assessment of the data, but there are quite a

number of other drugs out there that can give you

response rates in the range of 30 percent or so

lasting three or so months, be it etoposide, be it

gallium nitrate, be it a variety of other drugs.

There are some new drugs out there. Even

the radioimmunotherapeutics, since there are two, I

won't go and mention any names, there is only one

that has been used for aggressive lymphoma, and

that has been associated with a 43 percent response

rate in patients who have failed a median of two to

three prior regimens.

So, yes, there are a number of other

agents out there with different safety profiles,

some which appear to be, you know, like the

361

radioimmunotherapeutics have more marrow

suppression, don't have neuropathy, so they are

limited, but there are other drugs out there, and

it is quite a list of them that can give you 20 to

30 percent response rates.

DR. MARTINO: Are there other thoughts on

this issue? Are there alternative therapies for

these patients, and might those alternatives

constitute a control against which a randomized

trial might be done?

Dr. Wilson, do you want to speak to that,

please.

DR. WILSON: Well, I guess I would just

reiterate what Bruce said, that is that there

really is a list of agents out there, both in this

range. I think it important to note that the

population of patients has a huge impact.

I think we all know that. I do laud the

group for having broken the patients down, but as

it stands, the response rate of this drug seems

very much in the middle of the pack for many other

agents.

362

In terms of what would you compare this

to, I think there are numbers of agents you could,

and I guess I would like to think about that before

commenting on how to do a comparative study.

DR. MARTINO: Do you feel that, in fact,

if you were asked to choose a comparator, that you

could come up with one? I am not asking you to

tell me your choice, I simply want an answer to

whether you think that you could.

DR. WILSON: Well, if you wanted to come

up with a single agent, I think the answer is yes.

I think that, as Bruce points out, there are agents

with different toxicity profiles. I think one drug

that comes to mind that one could think about would

be etoposide.

The down side to that is that a

hematologic toxicity can be limiting for it, but I

think that if you didn't take tremendously

pretreated patients, that I don't think hematologic

toxicity would be limiting for it, and I think that

that would weigh off against the neurological

toxicity associated with a Vinca alkaloid.

363

DR. MARTINO: Dr. Pazdur, do you want to

comment?

DR. PAZDUR: One does not have to specify

a single drug here, and, in fact, as I stated in my

opening comments, there have been situations where

pharmaceutical sponsors have had varying

combinations or various drugs, and in a randomized

study, you could randomize to a treatment arm,

might have several treatments, kind of a treatment

de jure, as long as there were agreement by the

investigators that that was something that they

would consider a reliable and reputable treatment.

The stipulation is you would have to win

against that treatment arm.

DR. MARTINO: Are there other comments to

this question? If not, we will now turn to the

vote, and we will start with Bukowski on my right,

please. Please state your name and your vote, the

question being: Do we believe that there are

alternatives for this patient population?

DR. BUKOWSKI: So, the name is Bukowski,

and the answer is yes.

364

DR. CHESON: The answer is yes.

DR. BRAWLEY: Yes.

DR. REAMAN: Yes.

DR. MARTINO: Yes.

DR. MORTIMER: Yes.

DR. PERRY: Yes.

DR. HUSSAIN: Yes.

MS. HAYLOCK: Yes.

DR. GEORGE: Yes.

MS. KRIVACIC: Yes with a stipulation that

I don't know if there is potent, if you will, or I

guess show what has been shown here today, so there

is a lot of conflicting data between the FDA's

information and the sponsor's, so I am a bit

conflicted with this question, as well.

DR. BISHOP: Yes, with the stipulation I

can't think of an outstanding agent for patients

with a lack of hematologic reserve.

DR. WILSON: Yes.

DR. MARTINO: Our total is a unanimous

yes, although we have two members who have some

uncertainties.

365

The second question is: Previously, the

Agency has stated that the primary relevant

endpoint for aggressive non-Hodgkin's lymphoma were

rate of durable complete response and survival.

Partial responses were not considered predictive of

clinical benefit.

In this setting of relapsed, aggressive

non-Hodgkin's lymphoma, does the Committee agree

that durable CR should generally be the primary

endpoint for approval?

So, the issue again is CR's versus PR's,

or lesser degrees of response.

I would like to hear some comments on this

question, please. Dr. Perry.

DR. PERRY: I think part of the problem

here is the definition of complete response. There

are complete responses and there are complete

responses. It depends on how far you want to go.

Do you do a warm autopsy to biopsy everything

within the abdomen or do you simply say a CT scan

is evidence enough, or a PET scan or a gallium

scan, and what number of biopsies and how many PCRs

366

do you do.

Just saying "complete response" in these

days doesn't mean as much as it used to when we had

more primitive techniques and were smarter. We

knew much more than we do now.

So, I think there are partial responses

that are probably now the equivalent of old

complete responses, and I think some of them are

helpful.

DR. WILSON: I do think that PRs are a

relevant endpoint, but I think a PR in the absence

of a duration means little in a disease like this.

So, I know the question is written based on

previous ways in which this data was looked at, but

I think that a PR that lasts for a reasonable

length of time is a reasonable endpoint, but PR

alone, I do not.

DR. MARTINO: I share that feeling

completely. I don't think PR, in and of its own, I

mean a PR can be extremely fleeting and doesn't

always correlate with anything related to the

patient's behavior or the patient's well being. It

367

is often an x-ray event and primarily makes the

doctor feel good to be able to talk into a room to

say you have a response, and then not have to do

much more than that.

So, I completely agree that the issue of

durability and/or reduction of symptoms probably

are much more meaningful events.

Dr. Cheson.

DR. CHESON: As we discussed in the

response criteria paper, PRs in the setting of

relapsed and refractory disease are interesting to

identify drugs with some activity to pursue

further, but in and of themselves, I agree with you

that unless they have some durability, they are

rather meaningless, and I guess Dr. Perry has

stepped out, but not only are all CR's not all

CR's, but with the advent of PET technology, for

example, a lot of PR's turn out to be CR's, so we

are getting more sensitive measures of this, but a

true PR in the relapsed, refractory setting

generally bodes ill, and probably wouldn't even be

transplanted by many centers, these with a good

368

conscience.

DR. MARTINO: Dr. Pazdur, did you want to

comment?

DR. PAZDUR: I just wanted to reiterate

something that Bruce said. Here again, we are not

talking about drug screening. We are talking about

drug approval, and these are not necessarily the

same thing or they should not be the same thing

obviously.

So, there has to be a different level that

one is saying that they are going to accept between

something that is of interest to take to another

step in another development, and then saying, well,

this drug is ready for prime time here for general

use with all of the ramifications that that has

associated with it.

DR. MARTINO: For me, this is really a

problem that I have with this entire accelerated

approval process. I have sat on this committee for

about three years now, and it almost occurs to me

that we are looking for what is the least amount of

data to be convincing, and I think that is the

369

wrong approach, but that is what I see that we do,

especially with accelerated approval, is what is

the least amount that you can show me, to which I

will then give you a reward for that.

I actually think that as a medical

community, we have to rethink what our objectives

are and what our purpose are. They should be much

grander than that, and I think you are either

trying to shut me up or you want to say something.

DR. PAZDUR: The only thing I have to say,

Silvana, is go, girl, go.

[Laughter.]

DR. MARTINO: But in all--

DR. PAZDUR: Let me finish my comment,

though. That was just starting there.

When we have a meeting, such as this, we

have a litany of sponsors that come in and pose the

question to us. What is the lowest response rate

that you will take? What is the fewest number of

patients that you will take? And, in fact, we

actually have a euphemism regarding these meetings,

and it is called the "How long can you go?"

370

I think that that really represents a

clarification, and this is one of the reasons why

we have been discussing this and emphasizing the

accelerated approval commitments that these

patients have.

The purpose of accelerated approval was

not accelerated drug company profits. It was

accelerated access to people that had desperate

illnesses, that needed the therapies, and we were

allowing basically a surrogate to be used to get

these therapies out early to these patients that

needed it.

It wasn't a license to do less, less,

less, and less to a point now that we may be

getting companies that are coming in, well, what is

the lowest. It shouldn't be what is the lowest.

It is what is a sufficient amount to give patients

and physicians a real understanding of what their

drug will do.

Granted, we realize that there is a need

to get these drugs out, but we also have to have a

data package that we can understand and will make

371

labeling a strong process here. That is why our

commitment really is to get these trials ongoing,

these confirmatory trials, so it is very important

to us.

DR. MARTINO: The fear that I personally

have, as I have treated patients over the past 25

years, is that we, as a medical community, and that

"we" does include the pharmaceutical industry,

really in my observation have aimed for a lower and

lower behavior of drug, and in that process, if we

keep rewarding such behavior, we will see more and

more of it.

There is nothing new in the universe.

That is the way life works. So, we do have to

separate what our responsibilities are and to whom

are these responsibilities.

Dr. Bukowski.

DR. BUKOWSKI: When I think about this, I

think about the issue of unmet need as the main

factor that sort of leads me to think about how and

whether an agent should be considered for approval

in a particular area, and I think that to be

372

somewhat foremost in our minds.

I mean it may not be necessarily the issue

of how low can you go, but is there anything else

available in the area that can be utilized, and I

think that is very, very important, because

clearly, there are many situations where there are

unmet needs, where new agents may well have a very

minimal or modest response rate or modest activity,

but still these may be useful, and I think the

issue is, is getting those agents out to patients

in a very timely basis, with subsequently then

doing the appropriate studies to demonstrate the

clinical benefit associated with the agent.

DR. PAZDUR: Ron, that is specifically why

have the better than available therapy or an

improvement, or a situation where therapies do not

exist, but it has to be a real clinical situation,

it cannot be a contrived situation.

A couple of years ago we had a company

that wanted to develop a drug for leukemic patients

on a respirator, and the reason why patients were

on the respirator was because they received the

373

drug on the NDA, which was kind of ridiculous.

So, it has to be a real situation, a

really clinically relevant situation, not a

contrived situation.

DR. MARTINO: Dr. Brawley.

DR. BRAWLEY: Dr. Martino and Dr. Pazdur

are to be praised for their speaking of truth this

afternoon. I just want to add one thing. When we

teach our graduate students the development of

drugs, we teach them things about like how one can

look at data and actually think there is benefit,

but when one probes further, one finds that there

is not a benefit to that drug, there is actually a

net harm to the drug, and this is actually

frequently the reason why we need randomized

clinical trials.

I personally have been burned by clinical

studies that ultimately showed that beta carotene

increased the risk of lung cancer in smokers, and

did not decrease it, and I was one of the people

who said it's just a vitamin, how can it be

harmful. There are numerous examples in the

374

medical literature.

Now, we have heard today from some

patients who told us about some significant

toxicities that they are living with, and now we

have to make a decision is this drug beneficial

given those significant toxicities, and I think we

have not heard about some significant toxicities.

I will finish by saying I am very worried,

while I look at the FDA data versus the company's

data, I am very worried, not that there were

individuals who called responses because they

wanted to make money, I am worried that there are

doctors out there who saw patients who called

responses because they really wanted to see

responses in their patient. They were hoping

against all hope that they could do best for their

patient, but we have to remember there are some

significant discrepancies in data here.

DR. WILSON: We are looking at No. 2, and

we are asking is a PR of a certain length a

reasonable endpoint, but I think we should all

recognize that this is a unique circumstance. This

375

drug is known to be active. This is a drug that is

active, and the activity level is well within that

which has been seen before when it has been given

as either a single agent or as a continuous

infusion.

So, I think the bar is perhaps a little

bit even different than is it active. I think the

question is, is it active in a safer way, is it

active in a way that gives it significant additive

value.

DR. MARTINO: Other comments? Yes, Dr.

Hussain.

DR. HUSSAIN: It is just basically what

everybody said. I guess as doctors, you look at

things and you try to be objective, and to me,

objective with a patient benefit means you either

make them live longer or live better, and to look

at a scan, and a scan that goes down from a 6 cm

mass to a 3 cm mass, I would like to ask the

lymphoma doctors, has there been any precedent in

any drug, in lymphoma, in these kinds of settings

where a PR actually translated into a meaningful

376

thing as in patients, when tested prospectively, as

in patients living longer, quality of life

improved, or any direct benefit other than an image

benefit.

DR. CHESON: Well, there is that old

statistical conundrum of the

responder/nonresponder, and the CR's always do

better that the PR's, and the PR's always seem to

do better than the nonresponders. However, as we

all know, there are flaws in that sort of analysis.

Having some response confers some benefit,

but the magnitude of the response, you know, and it

depends what kind of response, was there associated

relieving of an obstruction, was there associated

decrease in symptoms going along with it, but as

you have said before, just shrinking something by

50 percent is not necessarily going to translate

into a meaningful clinical benefit unless there is

some durability of this and there is some

associated clinical benefit along with it.

DR. MARTINO: I think at this point I

would like to call the question to a vote. Again,

377

the important words to this question are does the

committee agree that a durable CR should generally

be the primary endpoint for approval, not PR, CR,

and the word durable.

We will start with Dr. Bukowski on my

right, please.

DR. BUKOWSKI: Yes.

DR. CHESON: Yes.

DR. BRAWLEY: Yes.

DR. REAMAN: Yes.

DR. MARTINO: Yes.

DR. MORTIMER: Yes.

DR. PERRY: Yes.

DR. HUSSAIN: Yes.

MS. HAYLOCK: Yes.

DR. GEORGE: Yes.

MS. KRIVACIC: No.

DR. BISHOP: Yes.

DR. WILSON: Yes.

DR. MARTINO: The total is 12 Yes, 1 No.

The next related question. Would high

rates of PR and long PR duration be reasonably

378

likely to predict clinical benefit, and thus

potentially support an accelerated application? If

so, please describe the PR rate and duration that

would be convincing.

Who would like to speak to that, please?

DR. CHESON: If I can reiterate something

I said before, about PR being a PR, we have a paper

in press in JCO in which we integrated PET scanning

into the International Workshop Criteria, and what

you see is that initially, there is a modest

difference--although this was in upfront patients

with large cell lymphoma--there was a modest

difference between the time to progression or

progression-free survival between the CR's and the

PR's.

Once you throw in PET scanning, the

difference becomes absolutely enormous, so it

depends how you measure these PR's. If you have a

PR that, indeed, is really a CR in disguise, then,

lo and behold, they are going to do quite well, and

so I think you have to look at this in the context

of what are you calling a PR, how are you defining

379

a PR, because that is going to make a big

difference.

DR. MARTINO: I personally would be leery

of answering the question with a number and a

duration, because I think the issue then, for me,

has to do what is the quality of life during that

time.

A PR that achieves improved quality of

life for some length of time is valuable, but it

would have to be accompanied by some true

measurement of quality of life for reduction in

symptoms. In and of its own, it would not impress

me very much.

DR. CHESON: And not just substitution of

one set of symptoms for another.

DR. PAZDUR: Perhaps this question is not

a voting question, but more of a discussion

question.

DR. REAMAN: I think the other discussion

issue is whether or not there are alternatives. I

mean if there are, in fact, other options, should

we really be discussing this in any great detail.

380

DR. BISHOP: The only thing I counter the

Chairperson's comments is that you can apply those

same criteria to a complete response. Yes, we

found that when you want to have a complete

response, and if you don't have quality of life and

everything else, then, should you count a complete

response that way, so I really don't think that is

a fair criteria.

I mean the only things that we have to go

on is improved survival, yes, we would all like

quality of life, but there is other things that as

we heard testify, that people are willing to live

with neuropathy, for one, and yet if that gives

them opportunity to be with family and friends, I

just don't think that is a fair criteria.

So, the duration is difficult to define,

but I don't think that is a fair criteria that you

have to have quality of life to go along with it.

DR. MARTINO: I think your point is

extremely well taken, and for me, what it reminds

me of is that even a CR may not be that meaningful,

which again gets to this issue of what is the point

381

of accelerated approval for me.

DR. PAZDUR: Here again, I think one of

the answers, and maybe when we wrote this question

it perhaps needed a bit of clarification, is there

any PR rate in duration that one would accept.

I think perhaps in this situation, one may

have to take a look at a randomized study if one is

even going to contemplate this, or is it a

situation, in a single-arm trial, that one would

accept a PR rate in a very refractory situation.

DR. WILSON: Without getting into numbers,

I mean I personally think yes, that there are

numbers of PR's, and there is durations that I

think would convince most people that that would be

accompanied by clinical benefit assuming that you

didn't have collection of quality of life issues.

So, I think the answer personally to No. 3

is yes, I think reasonable people can sit down and

hammer out where those numbers should lie. It is

probably beyond the scope of today.

DR. MARTINO: The problem that I see with

this is that when one looks at response rate, and

382

practically all tumors, what you primarily deal

with is PR, and CR rates are few and far between,

and that is a fact of life for all of us that deal

with oncology.

So, invariably, these applications do come

down to not arguing over is the CR rate high

enough, but rather it really comes down to the

issue of PR, whether we accept PR's as valuable or

not, because 90 percent of the time, that is

actually what you are getting pretty much in any

application that I have seen brought to this

committee.

DR. PERRY: Could I move we table the

question?

DR. MARTINO: Rick, are you comfortable

that you have heard enough?

DR. WILLIAMS: It is a basic question for

us, because we will have sponsors come to us, and

let's say there is no available therapy in a

situation, you know, the lymphoma situation, let's

say that is true, they will ask can we do a

single-arm study, and if we get a high enough

383

response rate, might we get accelerated approval.

Now, I don't have a sense. We have always

categorically said CR's are nothing, but we have

primarily been looking at tumors in an earlier

setting, we never really had this question, but now

we are getting the question a lot.

So, I think it is either the possibility,

as you suggest, Wyndam, I think that perhaps, I

mean you never know, a very high PR rate with a

long duration, it is conceivable, or perhaps it is

not, we will say what we have said in the past.

That is, you know, really I am going to

look at CR's, because you certainly can evaluate in

a single-arm study, the PR rate and the PR

duration, so there is no problem evaluating it, and

you may well say that this is way beyond what you

can do with anything out there, but the question is

do we think it is reasonably likely to predict

benefit. That is the essential question we would

have to ask.

DR. PERRY: I think the question differs

upon what tumor you are talking about. For

384

melanoma, we would take just about anything. For

Hodgkin's disease, we want very high standards

indeed. I think that is not a question we can

settle here this afternoon in this committee, at

this particular time, when we are trying to discuss

another drug. That is why I move to table, because,

as I understand it, it is not suitable for

discussion.

DR. PAZDUR: As far as endpoints, we will

be discussing endpoints in other meetings as far as

our ongoing endpoint project, we are going to be

having a hematology symposium on this perhaps this

year, so we will table that.

DR. MARTINO: Dr. Brawley.

DR. BRAWLEY: I would make a plea for PR

with quality of life criteria, and that probably

means that you are going to have to end up looking

at a randomized study, but if I were to see PR's

and better quality of life in a particular drug A

versus the leading drug in the treatment of that

disease, I would vote for it.

DR. MARTINO: Dr. Reaman.

385

DR. REAMAN: I would also like to clarify

from statements that were made earlier, about PR's

being indicative of activity and useful in

screening, we are only talking about PR being an

acceptable endpoint for accelerated approval,

correct, with a guarantee or with a plan for

definitive studies in place.

DR. MARTINO: Dr. Bukowski.

DR. BUKOWSKI: But clearly, there are PR's

that have long duration, and I think we have to

keep that in mind, where we will see refractory

settings where PR's develop that are of long

duration, and that has to be a consideration in our

deliberations.

I think we need to consider all these

alternatives.

DR. PAZDUR: Ron, in my comments

initially, for example, with Velcade, we had a

duration of a year, so we were quite happy with

that. You know, you didn't have to do a randomized

study here. But in the context of the disease that

we are talking about here, with the multiple drugs

386

that are available, to have somebody come in with a

single-arm trial with just PR, unless it was some

eye-popping results, I would discourage people, and

I think the tone of this whole conversation has

been a randomized study, so you have available

therapy even if you have to develop the drug in a

combination regimen before you first take it out

into a randomized setting might need to be done.

DR. BUKOWSKI: I agree with that, Rick,

and I purposely said durable as the modifier here,

and I think that has to be a part of it if you are

looking at a single-arm study. Clearly, the

randomized trial is the best way to do this.

DR. MARTINO: I think maybe we can sort of

summarize by saying that the Committee doesn't have

a strong feeling that response rate alone, without

some other bit of meaningful information, that that

alone is not probably adequate for an accelerated

approval.

Dr. George.

DR. GEORGE: A couple of comments. One is

I think it has been implied or even stated, and

387

something I agree with, you can't prove clinical

benefit from studies like this, these single arms,

so that you have to do something else.

The point I would like to make is you can

design such studies, so you could come up with

numbers here if you make certain assumptions, such

as, for example, as you do in ordinary studies, you

hypothesize certain differences you want to pick

up.

You could do the same thing here if you

said the only possible benefit is going to be in

those that have responses of some kind, and then

you can take it from there and say, well, if that

is true, then, what kind of response rates would

you have to have to even have a chance of finding a

clinical benefit in studies of a certain size.

In other words, a complete response or a

partial response, has to be at least a certain

level or you will never pick it up with a

reasonable size study.

There are ways to get at these numbers,

and if will depend on the disease, of course, but

388

it is not just sitting around the table saying,

well, what kind of duration of response or

percentage of responders could lead us to a

favorable conclusion. You could actually put

numbers on it.

DR. PAZDUR: No. 4 is the approval

question.

DR. MARTINO: So, we will move on to that

last one.

It reads: Do the partial responses at the

rate seen and for the duration reported for this

agent predict clinical benefit in relapsed,

aggressive non-Hodgkin's lymphoma?

Do you want to expound at all on what

clinical benefit means? I am reminding you of

earlier today when we had an issue of what is the

meaning of this, what is the exercise at hand here.

DR. PAZDUR: I think since we are talking

about accelerated approval, this question should be

reasonably likely to predict clinical benefit.

Here again, this is a predictor of clinical

benefit. Clinical benefit has been something that

389

is tangible to the patient, an improvement in

survival, as Maha said, an improvement in

disease-related symptoms, something tangible to the

patient.

So, what we are asking is does the partial

response rate and the duration, with all the

problems that we have discussed with this, is this

reasonably likely to predict clinical benefit.

DR. BRAWLEY: You are not asking is the

drug active? You are asking a different question.

DR. PAZDUR: Correct. This is an approval

question. We assume that the drug has some

response rates here. There is no argument with

anybody on that. This is an approval question.

Is it reasonably likely, with the data

that you saw, does this predict clinical benefit,

i.e., an improvement in survival, disease-related

symptoms, et cetera, something tangible to the

patient?

DR. BISHOP: Does that response include

comparison to currently available treatments?

DR. PAZDUR: We have already answered that

390

in a sense.

DR. BISHOP: No, I look at this question

different than that. This is asking does this have

benefit. You said this is in regard to approval,

and your opening minutes comments for accelerated

approval strictly assigned a demonstration of

clinical benefit, and you listed the four things -

survival, amelioration of symptoms, advantage over

available treatments.

DR. PAZDUR: That was advantage over

available treatments is what you need to show for--

DR. BISHOP: So, that is a definition of

clinical benefit in this question.

DR. PAZDUR: It is a requisite for

accelerated approval.

DR. WILLIAMS: You have already basically,

by answering No. 1 the way you did, probably ruled

out any possibility of approval if we went along

with that advice, but you would also have to answer

No. 4.

DR. MARTINO: Can I just make it very

simple? I think what the question is about is with

391

what we have heard today, and the discussions that

we have undergone today, do you believe that there

is enough substantial data to give approval to this

drug, so that it is available for someone to use

tomorrow.

DR. PAZDUR: Accelerated approval.

DR. MARTINO: It is that issue. Do you

think the data is good enough that you now want the

world to have it tomorrow, or do you think the data

is not of such magnitude. It is an issue of

magnitude, not is there a whiff of response. That

is the question.

Again, we cannot confuse the issue of is

there any activity, is there any value. It is not

the minimum requirement here. We cannot be aiming

for what is the lowest. That cannot be our goal

here. If it is, I am done with this group as of

this moment if that is our goal.

Dr. Bukowski, you are up first.

DR. BUKOWSKI: No.

DR. CHESON: No.

DR. BRAWLEY: No.

392

DR. REAMAN: No.

DR. MARTINO: No.

DR. MORTIMER: No.

DR. PERRY: No.

DR. HUSSAIN: No.

MS. HAYLOCK: No.

DR. GEORGE: No.

MS. KRIVACIC: No.

DR. BISHOP: No.

DR. WILSON: No.

DR. MARTINO: The vote is unanimous to

Question No. 4. It is No.

DR. PAZDUR: Thank you for your time and

interest.

[Whereupon, at 4:03 p.m., the meeting was

concluded.]

- - -