1
DEPARTMENT OF
HEALTH AND HUMAN SERVICES
FOOD AND DRUG
ADMINISTRATION
CENTER FOR DRUG
EVALUATION AND RESEARCH
ONCOLOGIC DRUGS
ADVISORY COMMITTEE
NDA
21-824
ZARNESTRA
(TIPIFARNIB)FILM COATED TABLETS
TIBOTEC THERAPEUTICS, A
DIVISION OF ORTHO BIOTECH,
LP PROPOSED
INDICATION FOR THE TREATMENT OF
ELDERLY PATIENTS WITH
NEWLY DIAGNOSED POOR-RISK
ACUTE MYELOID
LEUKEMIA (AML)
Thursday,
May 5, 2005
8:00
a.m.
5630
Fishers Land
Room
1066
Rockville, Maryland
2
P A R T I C
I P A N T S
ADVISORY COMMITTEE
REPRODUCTIVE HEALTH DRUGS
Silvana Martino, D.O. - CHAIR
Otis W. Brawley, M.D.
Ronald M. Bukowski, M.D.
Bruce D. Cheson, M.D.
Stephen L. George, Ph.D.
Pamela J. Haylock, RN [Industry
Representative]
Alexandra M. Levine, M.D.
Joanne E. Mortimer, M.D.
Michael C. Perry, M.D.
Gregory H. Reaman, M.D.
Johanna M. Clifford, M.S.,
RN, Executive Secretary
CONSULTANTS AND GUESTS
Consultants (voting)
Susan O'Brien, M.D.
Patient Representative
(voting):
Arthur Flatau
Acting Industry Representative
(non-voting):
Roger Porter
Guest Speaker (non-voting):
Frederick Appelbaum, M.D.
FDA PARTICIPANTS
Qin Ryan, M.D.
Ramzi Dagher, M.D.
Robert Justice
Richard Pazdur, M.D.
Robert Temple, M.D.
3
C O N T
E N T S
PAGE
Call to Order
Silvana Martino,
D.O., Chair 4
Introduction of
Committee 4
Conflict of Interest
Statement
Johanna
Clifford
6
Opening Remarks
Richard Pazdur,
M.D. 9
Sponsor Presentation -
Tibotec Therapeutics, Inc.
Introduction
Robert DeLap,
M.D., Ph.D. 15
AML in Elderly Patients
Richard Stone, M.D. 23
Clinical Data
Alain Thibault,
M.D. 36
Benefit/Risk
Alex Zukiwski,
M.D. 62
FDA Presentation: NDA 21-824, Zarnestra
Qin Ryan, M.D. 70
AML in Older Individuals
Frederick R. Appelbaum,
M.D., 90
Open Public Hearing 121
Questions from the
Committee 129
Discussion of the
Questions 181
4
1 P R O C E E D I N G S
2 Call to Order
3 DR. MARTINO: Good morning, ladies
and
4 gentlemen.
I'd like to begin this meeting.
5 The topic before us this morning
is the
6 drug Zarnestra, presented by Tibotec. And the
7 proposed indication is for the treatment of
elderly
8 patients with newly diagnosed poor-risk
myeloid
9 leukemia.
10 Introduction of Committee
11 The first order of business is I
would
12 like the members of the committee to
introduce
13 themselves.
And I'd like to start with Dr.
14 O'Brien, please.
15 I need you all to use your
microphones,
16
please.
17 DR. O'BRIEN: I'm from the leukemia
18 department at MD Anderson.
19 DR. CHESON: Bruce Cheson, head of
20 Hematology, Georgetown University Lombardi Cancer
21 Center.
22 DR. GEORGE: Steve George, Duke
University
5
1 Medical Center.
2 DR. BRAWLEY: Otis Brawley. I'm a medical
3 oncologist and epidemiologist at the Winship
Cancer
4 Institute of Emory University.
5 DR. MORTIMER: Joanne Mortimer,
medical
6 director, UCSD Morris Cancer Center.
7 MS. HAYLOCK: Pamela Haylock,
oncology
8 nurse and doctoral student at UTMB,
Galveston,
9 Texas.
10 MR. FLATAU: Arthur Flatau, I'm the
Patient
11 Representative.
12 DR. REAMAN: Greg Reaman, pediatric
13 oncologist, Children's Hospital, Washington,
D.C.;
14 George Washington University in the
Children's
15
Oncology Group.
16 DR. LEVINE: Alexandra Levine, head
of
17 hematology at University of Southern
California,
18 Norris Cancer Center.
19 DR. MARTINO: Silvana Martino, from
the
20 Angeles Clinic in Santa Monica, California,
main
21 medical oncologist.
22 MS. CLIFFORD: Johanna Clifford,
Executive
6
1 Secretary to the ODAC.
2 DR. PERRY: Michael Perry, medical
3 oncologist, University of Missouri Ellis
Fischel
4 Cancer Center, Columbia, Missouri.
5 DR. PORTER: Roger Porter, retired
both
6 from the NIH and Wyeth, now a consultant.
7 DR. BUKOWSKI: Ronald Bukowski,
medical
8 oncologist, Cleveland Clinic, Cleveland,
Ohio.
9 DR. DAGHER: Ramzi Dagher, Division
of
10 Oncology Drug Products, FDA.
11 DR. JUSTICE: Robert Justice,
Acting Deputy
12 Director, Oncology Drug Products, FDA.
13
DR. PAZDUR: Richard Pazdur, FDA.
14 DR. TEMPLE: Bob Temple, Director,
OD-1.
15 DR. MARTINO: Next, I'd like Ms.
Johanna
16 Clifford to read the conflict of interest
17
statements for the members of the panel, please.
18 Conflict of Interest
Statement
19 MS. CLIFFORD: The following announcement
20 addresses the issue of conflict of interest,
and is
21 made as part of the record to preclude even
the
22 appearance of such at this meeting.
23 Based on the submitted agenda all
24 financial interests reported by the committee
25 participants, it has been determined that
all
7
1 interest in firms regulated by the Center
for Drug
2 Evaluation and Research present no potential
for an
3 appearance of a conflict of interest, with
the
4 following exceptions.
5 In accordance 18 USC 208(b)(3),
full
6 waivers have been granted to the following
7 participants: Dr. Stephen George, for
consultant
8 for a competitor, which he received less
than
9 $10,001 per year; Dr. Ronald Bukowski, for
10 consulting with a competitor, which he
receives
11 less than $10,001 per year. Pamela Haylock has
12 been granted waivers under 208(b)(3) and 21
USC
13 505(n) for her spouse owning stock in a
competitor.
14
The stock is valued from $25,001
to $50,000.
15 A copy of the waiver statements
may be
16 obtained by submitting a written request to
the
17 agency's Freedom of Information Office, Room
12A-30
18
of the Parklawn Building.
19 In addition, we would like to not
that Dr.
8
1 Frederick Appelbaum, FDA's invited guest
speaker,
2 is participating as a representative of the
Fred
3 Hutchinson Cancer Research Center. He has no
4 financial interest in, or professional
relationship
5 with any of products of firms that could be
6 affected by the committee's discussions.
7 With respect to the FDA's invited
industry
8 representative, we would like to disclose
that Dr.
9 Roger Porter is participating in this
meeting as
10 the industry representative, acting on
behalf of
11 regulated industry. Dr. Porter is a private
12 consultant to industry.
13 In the event that the discussions
involve
14 any other products or firms not already on
the
15 agenda, for which an FDA participant has a
16 financial interest, the participants are
aware of
17 the need to exclude themselves from such
18 involvement, and their exclusion will be
noted for
19 the record.
20 With respect to all other
participants, we
21 ask, in the interest of fairness, that they
address
22 any current or previous financial
involvement with
9
1 any firms they may wish to comment upon.
2 Thank you.
3 DR. MARTINO: Thank you, Ms.
Clifford.
4 And, next, Dr. Richard Pazdur will provide
some
5 opening remarks to this meeting.
6 Opening Remarks
7 DR. PAZDUR: Thank you, Dr.
Martino.
8 First of all, I'd like to say
Feliz Cinco
9 de Mayo to everyone.
10 [Laughter.]
11 For everybody that's lived in
Texas, such
12 as Susan and myself--
13 VOICE: Muchas gracias.
14 DR. PAZDUR: De nada.
15 [Laughter.]
16 This is a big day in the State of
Texas.
17 And I just want to bring that up.
18 I have some enjoyable issues to do
to
19 start out here, and that is to thank two
members of
20 our committee that will be retiring. And I use
21 that word "retiring" in
quotations, because,
22 because of their expertise, we probably will
be
10
1 inviting them back--both to sit on ODAC
special
2 committees, as well as to serve as special
3 consultants to us during the process that we
4 evaluate drugs outside of the ODAC committee.
5 The two individuals that will be
leaving
6 the committee are Stephen George and Otis
Brawley.
7 Stephen George is, obviously, from
Duke
8 University in North Carolina, and has served
on the
9 ODAC Advisory Committee as the committee
10 statistician from July of 2001, to June of
2005.
11 In addition to his committee participation,
Dr.
12 George has visited the FDA and he's given
numerous
13 presentations to us on various statistical
issues,
14 and has served as a consultant to us on many
NDAs
15 and IND matters. So we really appreciate Dr.
16
George's efforts--on this committee and also behind
17 the doors here--in helping the FDA. And we look
18 forward to working with you, as you leave
the
19 committee, on a continuing basis.
20 So I have this beautiful plaque
here that
21 I'd like to present to you.
22 [Applause.]
23 Thank you very much.
24 DR. GEORGE: Thank you. Thank you.
25 DR. PAZDUR: And the gentleman that
is
11
1 sitting right next to him, Dr. Otis Brawley,
is
2 professor of hematology and oncology and
3 epidemiology, and is the Associate Director
for
4 Cancer Control at the Winship Cancer
Institute at
5 Emory University.
6 He's an international authority in
the
7 field of health disparities research and
prostate
8 cancer.
He's served on the committee from July of
9 2001 to 2005. Like Dr. George, Otis has been
10 involved with many of the behind-the-scenes
efforts
11 at the FDA; consulting with us on numerous
12 applications, considerations of Phase I and
Phase
13 II trial designs, and Phase III trial
designs also.
14
We really have appreciated Dr. Brawley's
15 work with us. And here, again--this is not to say
16 goodbye, but just as a transition to another
role
17 with us in the FDA. Thank you very much.
18 [Applause.]
19 Hasta luego, he said. Okay.
20 [Laughter.]
21 Recuerdos a todos--regards to
everyone.
22 I'm just going to make a few short
23 comments about this application, because I
think
24 that the speakers will probably address all
of the
25 salient points, and I think we could bring
up any
12
1 regulatory issues relatively succinctly
during our
2 presentations, and also in our discussions.
3 Basically, what we have here is a
4 single-arm study in a patient population
where
5 there has to be discussion that there is no
other
6 available therapy for this patient
population--or
7 the results are so impressive here that we
need to
8 consider the approval of the drug.
9 We're going to be asking basically
the
10 question: does this drug deserve full
approval?
11 Okay?
And one of the reasons why we're asking this
12 is that we have accepted basically a
situation
13 where complete response rates have equated
clinical
14 benefit.
Okay? We believe that this is an
15 established surrogate for survival. And, in
16 addition to that, we believe that
application in a
13
1 complete response rate would have a
reduction in
2 transfusion requirements and other
supportive care
3 products.
So this is meaningful clinical endpoint,
4 an established surrogate for clinical
benefit.
5 So, with that in mind, I'd like
you to
6 proceed with the discussions, and I'll turn
the
7 discussions--the presentations--over to Dr.
8 Martino.
9 DR. MARTINO: Rich, before I let
you sit
10
down, please, I just want to be sure that I'm clear
11 that, in fact, the application is seeking
for full
12 approval, and not accelerated approval.
13 DR. PAZDUR: Correct--yes.
14 DR. MARTINO: Because, when I
reviewed the
15 material, I came to this with one view, and
then
16 when the question was presented to me, I
realized
17 that it was full approval.
18 DR. PAZDUR: Correct.
19 DR. MARTINO: So I just wanted to
be sure
20 that that is what you mean.
21 DR. PAZDUR: And this is the area
that I
22 want to clarify here. The issue here is: we look
14
1 at complete response rates. And, here again,
2 that's not only the response rates but
response
3 rates of a sufficient magnitude--okay? So we look
4 at both of these parameters: the response
rate and
5 the magnitude. If that is sufficient, one should
6 conclude that this would be an established
7 surrogate for clinical benefit. Okay?
8 DR. PERRY: Do we have the option
of
9 recommending it for accelerated approval and
not
10 for full approval? Or is this simply "yes" or
11 "no."
12 DR. PAZDUR: Yes--we would
entertain any
13 topics or discussions regarding that.
14 DR. PERRY: So we've got three
options
15
then: yes, no--
16 DR. PAZDUR: Correct. But we'd like to
17 firsts discuss that endpoint of full
approval. And
18 then, if you want to deviate from that,
let's
19 please have a discussion of what. Okay?
20 DR. PERRY: Okay.
21 DR. MARTINO: Are there other
questions for
22 Dr. Pazdur at this point? Ladies and gentlemen?
15
1 Okay.
2 Turn the microphone on, please.
3 DR. PAZDUR: They're right there in
the
4 plastic box.
5 DR. MARTINO: A practical question, there.
6 Thank you.
7 At this point, I would like to
turn to the
8 Tibotec representatives to present their
data. And
9 Dr. DeLap, if you would please introduce
your
10 subsequent speakers, as well.
11 Sponsor Presentation - Tibotec
Therapeutics
12 DR. DeLAP: Thank you. Madam Chair,
13 members of the committee, colleagues and
14 guests--good morning. I'm Dr. Robert DeLap, Vice
15 President of Regulatory Affairs at Johnson
&
16 Johnson Pharmaceutical Research.
17 We are pleased to be here today to
present
18 data generated in National Cancer Institute
and
19 company-sponsored studies on the efficacy
and
20 safety of tipifarnib in poor-risk AML, and
our
21 application for approval of tipifarnib for
use in a
22 patient population that is not well served
by
16
1 existing AML therapies.
2 Our application proposes that
tipifarnib
3 will be indicated for the treatment of
elderly
4 patients with newly diagnosed, poor-risk
acute
5 myeloid leukemia, based on durable complete
6 remissions that were observed in the CTEP-20
study.
7 As noted in the agency's briefing
8 materials for today's meeting, complete
remissions
9 have been used as evidence of patient
benefit to
10
support approval of new treatments for AML. Thus,
11 the consideration today is the use of these
data to
12 support the approval of tipifarnib for this
13 indication.
14 Elderly patients with AML obtain less
15 benefit from the intensive induction
treatment
16 regimens used in younger patients, and the
risks of
17 severe treatment toxicities and
treatment-related
18
mortality rise with increasing age.
Since
19 risk-benefit considerations for
intensive-induction
20 treatment are often not favorable in these
21 patients, new treatments are clearly needed.
22 In the clinical
research to be discussed
17
1 today, tipifarnib has demonstrated
meaningful
2 efficacy in elderly patients as described in
our
3 proposed indication, with a
well-characterized
4 safety profile in outpatient treatment.
5 [Slide.]
6 Tipifarnib was originally synthesized
in
7 the Johnson & Johnson Pharmaceutical
Research
8 laboratories. Clinical investigations began with
9 solid tumor studies in 1997. Based on mutual
10 interest in this compound, the company and
the NCI
11 entered into a Cooperative Research and
Development
12 Agreement in 1999.
13 Pre-clinical evidence of activity
against
14 leukemias led to clinical research in AML,
with
15 initiation of the CTEP-1 Phase 1 study in
1999.
16 Complete clinical remissions observed in
CTEP-1 led
17 to further research, including the CTEP-20
Phase 2
18 study in myeloid malignancies.
19 Following consultations between
the
20 company, the National Cancer Institute and
the FDA,
21 CTEP-20 was subsequently amended and
expanded to
22 focus on evaluating the efficacy and safety
of
18
1 tipifarnib in elderly patients with newly
diagnosed
2 poor-risk AML.
3 Considering the unmet need in this
patient
4 population, tipifarnib has recently received
orphan
5 designation for AML, and has been granted
6 fast-track status by FDA.
7 The NDA for tipifarnib was
accepted into
8 FDA's Continuous Marketing Application-1
pilot
9 program, which has allowed for expedited
submission
10 and review of these data.
11 [Slide.]
12 This slide summarized the study
program
13 for tipifarnib in poor-risk AML.
14 Following the CTEP-1 study, the
INT-17
15 study evaluated tipifarnib in patients with
16
relapsed or refractory AML.
17 Today's discussions will focus on
the
18 CTEP-20 study, as this is the study that has
19 evaluated efficacy and safety in the patient
20 population of interest for today's
discussion.
21 The AML-301 study, in patients
greater
22 than 70 years of age with newly diagnosed
leukemia
19
1 is evaluating the effect of tipifarnib
versus best
2 supportive care. And that study is designed to
3 establish the magnitude of the anticipated
survival
4 benefit, and is actively enrolling patients.
5 The ongoing CTEP-50 study is an
iterative
6 trial being conducted under NCI supervision,
which
7 is evaluating alternative dosing regimens in
8 elderly patients with newly diagnosed
leukemia.
9 Finally, the AML development
program also
10 includes related studies, not shown on this
slide,
11 in maintenance of remission and use in
combination
12 with other agents. Those studies are briefly noted
13 in the company's background materials for
today's
14 meeting, but will not be included in our
15 presentation today, as they represent work
in
16 progress in other AML settings.
17 [Slide.]
18 The CTEP-20 study focused on
patients who
19 have generally not been represented in AML
clinical
20
trials, and are not well served by
21 intensive-induction chemotherapy
regimens. The
22 median age of the elderly patients enrolled
in this
20
1 study was 75. Most of the patients had antecedent
2 myelodysplastic syndromes; 49 percent had
3 unfavorable karyotypes; the remainder had
4 intermediate karyotypes. No patients with
5 favorable karyotypes were enrolled.
6 Overall, 90 percent of patients
had two or
7 more risk factors, considering age,
antecedent
8 myelodysplastic syndromes, unfavorable
karyotypes,
9 or evidence of organ dysfunction. These are
10 patients who would be expected to have poor
11 tolerance for standard AML treatments, and
would be
12 much less likely to benefit from existing
13 treatments.
14 [Slide.]
15 As you will see in today's
presentation,
16 tipifarnib demonstrates a favorable benefit
risk
17 for a more fragile patient population that
18 generally does not receive standard AML
therapy.
19 Evidence of meaningful clinical efficacy has
been
20 observed, with a 15 percent rate of durable
21 complete remissions in the planned analysis.
22 Treatment safety has been well
documented,
21
1 with more than 1,000 patients in monotherapy
2 studies.
This includes a total of 409 patients in
3 the AML studies, CTEP-20 and INT-17, as well
as
4 patients exposed at lower doses in solid
tumor
5 studies.
6
Tipifarnib produces predictable and
7 reversible myelosuppression with continued
daily
8 dosing, but it is myeloablative, and the
incidence
9 of life-threatening, not-hematologic
toxicities has
10 bene low.
11 Patients in the CTEP-20 study were
able to
12 spend much of their time outside of the
hospital.
13 Thus, tipifarnib can serve as an oral
out-patient
14
treatment for these patients.
15 [Slide.]
16 Our agenda today includes three
additional
17 presentations. In a few moments I will turn to Dr.
18 Richard Stone, from the Dana-Farber Cancer
19 Institute, who will discuss the problem of
AML in
20 elderly patients.
21 Dr. Alain Thibault will then
review the
22 clinical data provided in our new drug
application,
22
1 focusing on the CTEP-20 study, which has
provided
2 data on the efficacy and safety of
tipifarnib in
3 elderly poor-risk patients with newly
diagnosed
4 AML.
5 Finally, Dr. Alex Zukiwski will
summarize
6 benefits and risks of tipifarnib treatment
in this
7 patient population.
8 We are joined today by medical
experts to
9 contribute to the discussion, and to help
address
10 specific questions. These include Dr. Karp, who
11 served as principal investigator for the
CTEP-20
12 study; Drs. Sekeres and Stone, who have
special
13 expertise in AML and have experience with
the use
14 of tipifarnib; and Dr. Wright, from the
NCI's
15 Cancer Therapy Evaluation Program. Unfortunately,
16 Dr. Albitar from Nichols Institute, who
provided an
17 independent review of bone marrow slides in
the
18 CTEP-20 study, and had planned to be with us
today,
19
could not be here because of a family emergency.
20 This concludes my
introduction. I will
21 now turn the podium over to Dr. Richard
Stone who
22 will discuss the problem of AML in elderly
23
1 patients.
2 Thank you.
3 AML in Elderly Patients
4 DR. STONE: Dr. DeLap, thank you
very much.
5 Members of the panel, guests--I'll
be
6 spending the next few minutes discussing the
7 following topic: AML in the older-age
patient
8 represents a therapeutic area of significant
unmet
9 need.
And this is particular true for those
10 subjects who have an inferior prognosis
compared to
11 the average.
12 [Slide.]
13 Now, AML in the older population
is not
14 uncommon, and the number of cases will be
15 increasing over time. This is clearly a
16 biologically and therapeutically distinct
disease
17 compared to AML which may occur in younger
adults.
18 And the reasons for this distinctive
character are:
19 number one, it's an intrinsically resistant
disease
20 to chemotherapy; and number two, there are
markedly
21 inferior outcomes to available
chemotherapeutic
22 agents compared to younger adults.
23 Some subgroups of patients who are
older
24 adults with AML have a markedly worse than
the
25 average prognosis which, I think you'll see
in a
24
1 minute, is quite poor to start with. As such, many
2 patients who are older with
AML are not offered
3 and/or refuse the standard cytotoxic
induction and
4 post-remission therapy. As such, an efficacious,
5 relatively non-toxic approach would be
welcomed by
6 patients and leukemia doctor's alike.
7 [Slide.]
8 There are approximately 12,000 new
cases
9 of AML each year in this country. Approximately
10
9,000 people die of this disease.
11 In contrast to what might be the
case if
12 you look at a tertiary care cancer center,
the
13 median age of AML is at least 68. The incidence
14 increases markedly as people get older. For
15 example, if you're 50 years old you have a 1
in
16 50,000 chance of having AML. If you're 70, you
17 have a 1 in 7,000 chance.
18 [Slide.]
19 This next slide graphically
depicts the
25
1 marked increase in the incidence of AML that
occurs
2
as people get older. And it's
particularly
3 striking once you get to the sixth, and
4 particularly seventh, decade of life.
5 [Slide.]
6 Moreover, as I think everybody is
aware,
7 the demographics of our population are
changing to
8 the fact that we're getting to be older as a
9 country.
And, as such, the number of cases of AML
10 in this age cohort--or the number of cases
11 total--will be increasing over the next few
12 decades.
13 [Slide.]
14 Now, this slide depicts the
situation that
15 was true in the 1980s, when chemotherapy was
16 applied the same way to younger adults and
older
17 adults with AML. This is data taken from
18 cooperative group trials on both sides of
the
19
Atlantic, and this required the patient to get to a
20 center where they could get
chemotherapy. So, as
21 I'll come back to, it may not be
representative of
22 what really happens in the community.
23 Nonetheless, this
slide makes the
24 important point that the therapeutic outcome
in
25 older adults is much different than younger
adults.
26
1 For example, if you're over age 65, your
chance of
2 achieving complete remission is only 45
percent,
3 compared to 70 percent in younger adults
with AML.
4 If you achieve remission your chance for
staying in
5 remission is only about one in five,
compared to
6 about 45 percent of younger adults.
7 And what's particularly striking
to me as
8
an oncology and a leukemia doctor taking care of
9 these patients, is the treatment-related
mortality
10 rate is about one in four, and the early
death rate
11 is much lower in younger adults.
12 The overall
survival--about 10 percent,
13 walking in the door. And these are people that
14 could get chemotherapy--compared 1 in 30
younger
15 adults.
16 The median survival of older adults who
17 present with AML and go on clinical trials
is only
18 10 months--which I think we'd all agree is
not
19 someplace we'd like to be.
20 [Slide.]
21 There are two major reasons--as I
22 indicated--for this inferior outcome. The first
23 general category is decreased host
tolerance. Of
24 course, being older, these patients have a
higher
25 incidence of having co-morbid diseases such
as
27
1 diabetes and vascular disease. Perhaps because of
2
that, and for just general aging features, they
3 have a decreased ability to clear
chemotherapy,
4 which obviously could contribute to
increased
5 toxicity.
6 Thirdly, it's been shown in
multiple
7 studies that the ability to recover from
myelotoxic
8 chemotherapy is diminished in older adults,
and
9 they have a longer period of neutropenia and
10 thrombocytopenia--which leads to an enhanced
rate
11 of chemotherapy-induced complications.
12 [Slide.]
13 At least as important, if not more
so, is
14 the fact that the leukemias which arise in
older
15 adults are intrinsically resistant,
biologically.
16 This fact of increased intrinsic resistance
is
28
1
manifested by, or associated with, these features:
2 number one, there's an increased instance of
what's
3 called "unfavorable chromosomal
abnormalities" in
4 older patients, such as the loss of the long
arm of
5 the entire chromosome-5 or -7' problems at
11q23,
6 and complex cytogenetic abnormalities. These are
7 the same type of abnormalities that occur in
people
8 who have myelodysplastic syndrome, and/or
people
9 who had prior chemotherapy for other
cancers.
10 There's an increased incidence of
11 antecedent--either known or
suspected--hematologic
12 abnormalities, most particularly
myelodysplastic
13 syndrome.
14 There's an increased likelihood of
15 expression of genes which encode drug
resistance,
16 most notably the MDR-1 protein, which is a
17 chemotherapy efflux pump, as well as other
ones
18 like MRP, LRP, MSH-2.
19 [Slide.]
20 Now, this combination of
biological and
21 host factors, and the inferior outcomes, led
to a
22 slight change in the approach of cooperative
groups
29
1 in the 1990s, where separate clinical trials
were
2 designed for older adults compared to
younger
3 adults with AML.
4 This is a representative list of
trials
5 conducted in the 1990s in cooperative groups
in
6
Europe and in America. And even
though these
7 trials evaluated different novel therapeutic
8 strategies, such as the use of growth
factors,
9 different chemotherapeutic drugs, and
10
drug-resistance modulating drugs, the results are
11 very stereotyped from trial to trial. There are
12 other trials out there which I could have
picked,
13 such as the recently published MRC trial,
but that
14 was largely a little bit younger patient
15 population.
16 The median age is 68 in all the
trials.
17 The complete remission rate is about 40 to
50
18 percent.
And, again, the toxic death rate--and
19 this is the 1990s--is in the 20 percent
range
20 throughout all the trials. And, again, all the
21 trials--median survival, nine to 10 months.
22 And I want to stress one very
important
30
1 point: it's that these--the patients who
went on
2 these trials were, number one, deemed to be
3 chemotherapy candidates. They got to a center that
4 was participating in a cooperative group
trial.
5 They had to meet the eligibility criteria
for these
6 trials--which varied from trial to trial,
there
7 were subtle differences. Most of these trials did
8 not allow people with secondary AML; that is
AML
9 that occurred myelodysplastic syndrome, or
after
10
prior chemotherapy to go on. Some
did. Some of
11 the trials had a lower boundary of age 60,
some 65,
12 some 55.
13 And so those are kind of the best
results
14 one can get with chemotherapy.
15 [Slide.]
16 The situation in the community is
17 certainly much worse. That's number one.
18 Number two is: if you look into
the
19
subgroup analysis of these trials, you can find
20 some important facts which suggest that you
can
21 identify patients that even have a worse
prognosis
22 than the average. For example, if you have prior
31
1 myelodysplastic syndrome, your chance of
remission
2 is 24 percent compared to 52 percent if you
don't.
3 If you have one of those poor cytogenetic
4 abnormalities that I mentioned, 21 percent
5 likelihood of remission, compared to 55
percent if
6 you don't. And these data were taken from
the SWOG
7
trial.
8 This data from the recently
published ECOG
9 trial says that if you're over age 70,
you've got a
10 29 percent of going into remission, compared
to 51
11 percent for those younger than age 70. However,
12 only 13 percent in this trial were above age
75.
13 Only 5 percent of those in the
Lowenberg
14 trial were above age 80, and in those people
the
15
chance for remission was only 14 percent.
16 So, number one, you can find bad
17 prognostic factors within these groups and,
number
18 two, the number of patients who are really
old who
19 go on these trials is low.
20 [Slide.]
21 Community data is shown in this
slide
22 which suggests that if you're over age 65
and you
32
1 present with AML, your median survival may
be only
2 in the several-month range, compared to the
another
3 10 months we saw for the people who went on
those
4 chemotherapy trials.
5 [Slide.]
6 So, given this sort of dismal
outcome with
7 chemotherapy, the value of chemotherapy in
this age
8 population is debated, particularly in those
who
9 have poor prognosis features.
10 There have only been a couple of
11 randomized studies--and these were done in
Europe
12 in the 1980s--which tried to compare early
13 aggressive chemo versus less intensive
approaches.
14 And they both showed a small increase in
survival
15 for early intensive chemo, but there was no
16 associated quality of life studies, and
cost, in
17 terms of up-front mortality was quite high.
18 These issues are reflected in the
National
19 Cancer Center Network guidelines--it's a
consensus
20
panel of AML experts--which acknowledges that
21 standard induction chemotherapy is an
option, but
22 clinical trial with either new agents or
biological
33
1 agents is the preferred approach even for
those
2 people who have good performance status who
are
3 over age 60 and present with AML.
4 [Slide.]
5 How do people make
this difficult decision
6 between a treatment which has a 25 percent
toxic
7 death rate and a low cure rate, versus
supportive
8 care?
It's a very difficult decision.
My
9 colleague Dr. Sekeres, when he as at the
10 Dana-Farber, tried to study this by a
prospective
11 patient-doctor questionnaire, and among the
12 findings from the study are that, number
one,
13 patients consistently inflate the chance of
cure,
14 compared to what is stated in medical record
15 predicted by the physician; number two,
despite
16 documentation in the medical record that the
17 doctors discussed multiple treatment options
with
18 the patients, the patients said, no, they
didn't
19 discuss this with me. That may be because there
20 really aren't too many options, and the
options
21 seem to be so stark that people feel they
don't
22 really have any.
23 [Slide.]
24 What happens in the
community? This slide
25 suggests the choices people make.
34
1 First of all, if you're in the
younger
2 cohort of the overall older cohort, you only
choose
3 chemotherapy about half the time. As you get
4 older, the chance of choosing chemotherapy
is quite
5 low.
6 What are the consequences of this
7
decision? Well, if you choose to
get chemotherapy,
8 you're going to spend significantly more
time in
9 the hospital than if you choose supportive
care.
10 What's even more important than
that is
11 you don't get any bang for the buck, because
the
12 percentage of time you spend in the
hospital,
13 compared to the total amount of time that
you have
14 left is still higher if you choose
chemotherapy.
15 So you don't seem to reap any benefit in
that
16 regard.
17 [Slide.]
18 So I think it's quite clear that
the
19 efficacy of standard chemotherapy is reduced
in the
35
1 older adult with AML compared to the younger
adult.
2 The therapy is poorly tolerated. There is clearly
3 a high therapy-related mortality rate. No trials
4 that I know about have really addressed the
quality
5 of life cost of chemotherapy.
6 If there is a small improvement in
7 survival with chemotherapy, it's quite
likely to be
8 offset by an increase in hospitalization and
other
9 negative QOL factors. And it's moot for many
10 patients, because two-thirds of patients who
are
11 older than age 65 don't choose chemotherapy
as
12 their primary treatment modality.
13 So non-chemotherapeutic
approaches,
14 besides supportive care, which may have
efficacy
15 and low toxicity, are badly needed.
16 [Slide.]
17 In summary, it's clear that AML in
the
18 older adult is a biologically and clinically
19
distinct entity. Even in the
so-called "best"
20 patients who go on chemotherapy trials, the
chance
21 for treatment-related death with induction
22 chemotherapy is actually greater than the
chance
36
1 for cure.
In those poor prognosis patients who
2 have--in the older part of this group--who
have
3 prior MDS, adverse cytogenics, the
advisability of
4 chemotherapy must be considered very low.
5 Patients and doctors are often
choosing a
6 non-intensive approach, but currently there
is
7 really no such therapy in this category
which
8 offers and appreciable chance for remission.
9 So I'd like to thank you for your
10 attention.
And I'll be happy introduce Alain
11
Thibault from Johnson &
Johnson Pharmaceutical
12 Research & Development to talk about
CTEP-20.
13 Clinical Data
14 DR. THIBAULT: Thank you, Dr.
Stone.
15 Good morning. My name is Alain Thibault.
16 I'm responsible for the clinical development
of
17 tipifarnib worldwide.
18 I will first describe key features
of
19 tipifarnib, then I will present the results
of
20 CTEP-20, which was a trial sponsored by the
Cancer
21 Therapy and Evaluation Program of the
National
22 Cancer Institute. These data are presented in
37
1 support of our application for the approval
of
2 tipifarnib in the treatment of newly
diagnosed
3 elderly patients with poor-risk AML.
4 [Slide.]
5 Tipifarnib is a selective and
competitive
6 inhibitor of the enzyme farnesyl
transferase. The
7 enzyme processes more than a hundred
proteins
8
intracellularly--some of which are shown here, and
9 many of which are involved in signaling
pathways
10 linked to the control of cell growth. Other
11 extensive research has been conducted. To this
12 date, the specific pathways associated with
the
13 anti-leukemic activity in AML are still the
subject
14 of ongoing research.
15 [Slide.]
16 Tipifarnib is an oral
treatment. So,
17 after oral administration, plasma and bone
marrow
18 concentrations rapidly exceed the IC50 of
AML cell
19 lines as determined in vitro. Tipifarnib is
20 metabolized in the liver by several
pathways. The
21 major metabolites are biologically inactive.
22 Tipifarnib is not a substrate from drug
efflux pump
38
1 encoded by the MDR-1 gene.
2 The diversity of metabolic routes
may
3 explain why tipifarnib has demonstrated a
low
4 potential for drug interactions in several
5 pharmacology studies that have been carried
out to
6 date.
7 [Slide.]
8 The recommended dosing regimen is
600
9 milligrams po BID, given for 21 days in
four-week
10 cycles.
This was established by a classical dose
11 escalating Phase I trial conducted in 34
patients
12 with poor-risk leukemias, most of whom had
AML.
13 The 21-day administration is
required to
14 maintain sustained inhibition of the target
so as
15 to maximize efficacy. And the seven-day rest
16 period is required to reduce the incidence
of
17 peripheral neuropathy, which had been
observed when
18 continuous, uninterrupted dosing was used.
19 The 600 milligram BID dose is
associated
20 with consistent farnesyl transferase
inhibition;
21
plasma and bone marrow concentrations that exceed
22 the IC50, and acceptable patient
tolerability.
23 [Slide.]
24 Let's now turn to the description
of the
25 study design. I'll first describe the rationale in
39
1 design, then I'll go over demographics,
indices,
2 characteristics. Then we'll go over efficacy and
3 safety.
4 CTEP-20 was part of a research
agreement
5 established between Johnson & Johnson
and the
6 National Cancer Institute. This was a single-arm
7
study. It was conducted at six
sites across the
8 United States, and Johns Hopkins University
was the
9 coordinating center.
10 From the very beginning, the aim
of the
11
investigators was to study tipifarnib in patients
12 with poor-risk myeloid neoplasms--I'll go
over them
13 in a few minutes. And this was a very critical
14 feature of the study from the very outset--that
15 study of patients with poor-risk myeloid
disorders.
16 [Slide.]
17 So--newly diagnosed patients, with
18 high-risk MDS--myelodysplastic
syndrome--chronic
19
myelomonocytic leukemia, and acute myeloid leukemia
40
1 were initially enrolled. Specifically regarding
2 AML, the diagnosis of AML was based on WHO
3 criteria.
No prior therapy for AML was
4 allowed--with the exception of hydroxyurea,
which
5 could be used to control counts prior to the
6 patient's entering the study.
7 With respect to age, the study
initially
8 enrolled patients age 18 to 65, with risk
factors
9 such as unfavorable cytogenetics, prior MDS,
or
10 prior exposure to chemotherapy. On the other hand,
11 patients aged 65 and above could enter the
study
12 with or without risk factors. These were the
13 initial age requirements for AML patients.
14 The patients had to have
approximate
15 status of 0 to 2 on the ECOG scale. And this was
16 chosen because it would enable them to
receive
17 treatment in the outpatient setting.
18 [Slide.]
19 After consultation and discussion
and
20 input from the FDA in July of 2003, the
study
21 protocol was amended: it was amended to
focus on a
22 more homogeneous population of elderly
patients who
41
1 had AML, who were not candidates to receive
2 intensive induction chemotherapy because the
3 associated risks were felt to outweigh the
4 benefits.
5 Therefore, after this amendment,
the age
6 requirements were raised. Patients aged 65
to 74
7 had to have a prior history of MDS, while
patients
8 75 years or older could enter the study in
the
9 absence of other risk factors.
10 So these are the patients that Dr.
Stone
11 described as commonly excluded from trials
that are
12 commonly reported in the literature.
13 [Slide.]
14 The primary endpoint of the study
was
15 complete remission, assessed by the investigators.
16 Duration of CR, partial remission,
17 hematological improvement, and overall
survival
18 were evaluated as secondary endpoints. Then the
19 safety profile was characterized.
20 [Slide.]
21 The study applied standard
morphologic
22 criteria of complete remission, which are
listed
42
1 here.
2 To achieve a complete remission, a
patient
3 had to demonstrate less that 5 percent
leukemic
4 blasts in the bone marrow; full recovery of
5 peripheral counts--without, obviously,
circulating
6 blasts or any evidence of extramedullary
AML.
7 [Slide.]
8 Partial responses were defined as:
9 complete recovery of counts in the presence
of
10 residual blasts--5 to 19 percent, following
at
11 least a 50 percent decline from baseline
values.
12 Then, hematology improvement was defined as
partial
13 recovery of peripheral counts, with the same
bone
14 marrow context.
15 So these responses can be viewed
as broad
16 indicators of anti-leukemic activity, even
though
17 their correlation with survival is not well
18 established.
19 [Slide.]
20 Treatment with tipifarnib as an
outpatient
21 monotherapy regimen--to ensure patient
safety all
22 patients were monitored weekly for
toxicity. All
43
1 patients had a bone marrow aspirate and
biopsy
2 performed prior to study entry. And this procedure
3 was repeated at the end of each treatment
cycle.
4 Unless they had dose-limiting
toxicity,
5 patients continued to receive tipifarnib in
a
6 cyclical fashion until disease progression
or
7 relapse.
8 The only exception concerns
patients who
9 achieved a complete remission. These patients had
10 the option to stop treatment after three
additional
11 cycles, and then be re-treated at the time
of
12 relapse.
And I will go over the outcome of this
13 maneuver, as well.
14 [Slide.]
15 In response to adverse events, the
16 investigators could individualize treatment
using
17 three strategies: either dose reductions,
treatment
18 interruptions within a cycle, or treatment
delays
19
between cycles.
20 Each cycle was to include a
minimum of 21
21 days of tipifarnib. The minimum rest period was
22 seven days, which could be extended by a
maximum of
44
1 35 days if needed. The total cycle duration could
2 not exceed 63 days.
3 [Slide.]
4 So let's go over the details of
the
5 population right now.
6 So, CTEP-20 started as a study of
7 tipifarnib in high-risk MDS, CMMD and
AML. Of the
8 171 patients that were enrolled, 158 were
9 considered to be poor-risk AML
patients. The 137
10 elderly poor-risk patients were strictly
defined in
11 the final protocol--as I outlined
earlier. And
12 from now on, the presentation will focus on
the 136
13 patients who were actually treated.
14 The 61 patients aged 65 to 74,
with prior
15 MDS; and the 75 patients aged 75 or more
make up a
16 unique population.
17 [Slide.]
18 The median age of the patient
population
19 is 75 years.
The male to female ratio is similar
20 to that of the general elderly AML
population, and
21
appears to reflect the higher incidence of MDS in
22 men.
23 The major of patients were
Caucasians--or
24 White.
Among the seven non-Caucasians, three were
25 African-Americans, two Asians, and two were
45
1 Hispanics.
2 Most patients were symptomatic,
either
3 from AML or from pre-existing
co-morbidities.
4 [Slide.]
5 Now, using the age-specific
incidence of
6 AML that Dr. Stone referred to earlier, if
we use
7 this as a comparator, CTEP-20 appears to be
more
8 representative of the general AML population
than
9 what is found in most other trials, in that
the
10 median age of the CTEP-20 patient was 75
years old.
11 And this exceeds that of the major
cooperative
12 group studies by approximately 10 years.
13 [Slide.]
14 Now, we know that many issues
underlie the
15 assessment of AML patients, especially in
the
16 elderly.
The chance of cure, the risk of toxicity,
17 the need for hospital stays all have an
impact on
18 treatment options that are offered to these
19 patients.
This has been reviewed by Dr. Stone.
20 What this slide lists are the
clinical
21 reasons given by the investigators--the six
22 principal investigators of this study--for
23 including the patients on the trial rather
than
24 administering intensive chemotherapy to
them.
25 Age and the presence of risk
factors were
46
1 most commonly invoked. Patient preference over
2 physician preference of experimental
treatment over
3 chemotherapy played a minor role.
4 [Slide.]
5 Now, the clinical rationales I've just
6 reviewed can be more objectively described
in terms
7 of the risk factors that are classically
associated
8 with poor outcome from chemotherapy So the
CTEP-20
9 patients had, by design, one of the highest
10 prevalence of risk factors ever reported in
the
11 literature.
Prior MDS was documented in 82 percent
12 of the patients. Forty-nine percent of the
13 patients harbored unfavorable cytogenetics,
such as
14 deletion of chromosome-5 or -7. The other patients
15 had intermediate karyotypes. Patients with
16 favorable karyotypes, such as inversion-16,
were
47
1 not enrolled on this study.
2 All in all, 41 percent--this
number is not
3 on the slide--had both MDS and unfavorable
4 cytogenetics.
5 Now, in terms of increased
morbidity
6 risks, 55 percent were age 75 or older--more
than
7 half; and 61 percent had evidence of organ
8 dysfunction.
This was manifested by two or more
9 active medical conditions other than the
AML, on
10 either history, physical examination or
laboratory
11 findings.
12 [Slide.]
13 Looking at the number of risk
factors per
14 patient is also very interesting: 44 percent
of the
15 patients had two risk factors; 35 percent
had
16 three; 11 percent had four of these risk
factors.
17 So, overall, 90 percent of the trial
population
18 entered with two or more risk factors on
this
19 trial.
20 [Slide.]
21 The full spectrum of leukemic
burden was
22 represented also in this study. The median bone
48
1 marrow blasts count at diagnosis was 46
percent.
2 All patients met the WHO criteria--as I
3 mentioned--except for one, who was
nevertheless
4 included because he was clearly evolving
from MDS.
5 This patient, who did not achieve a CR was
excluded
6 by the FDA.
But, for the sake of this
7 presentation, I'm reporting the results
based on
8 the investigator's assessment.
9 [Slide.]
10 The majority of the patients were
severely
11 myelosuppressed--as would be expected--prior
to
12 study entry.
The median neutrophil count was 636,
13 with 61 percent having Grade 3 or 4
14 myelosuppression at the time of entry. Fifty-six
15 percent of the patients had a similar degree
of
16 thrombocytopenia, with a median platelet
count
17 value of 41,500.
18 [Slide.]
19 So, in summary, the 136 patients
accrued
20 to this study represent a population with a
high
21 incidence of risk factors, and they
routinely do
22 poorly with available treatment.
23 So, it is in this context that I'd
like to
24 review the efficacy of tipifarnib for the
next few
25 minutes.
49
1 [Slide.]
2 Complete remissions were
documented by the
3 investigators at the research sites in 20
patients,
4 for a complete remission rate of 15
percent. The
5 associated 95 percent confidence interval
ranges
6 from 9 to 22 percent.
7 Partial remissions and hematologic
8 improvements were documented in 10 additional
9 patients.
So, in total, tipifarnib reduced the
10 leukemic burden of 30 patients, or 22
percent of
11 the study population.
12 [Slide.]
13 In general, the data shows
consistency
14 across risk groups. And this is what we illustrate
15 here on this slide.
16 For example, the complete
remission rate
17 in the 111 patients with prior MDS,
secondary AML,
18 was 16 percent; and the response rate in the
19 patients aged 75 years or more was 12
percent.
20 [Slide.]
21 The complete remissions were
documented at
22 four of the six sites. No single institution
23 accounts for the majority of cases.
24 [Slide.]
25 And as a specific quality control
feature
50
1 of the study, the complete remissions were
sought
2 to be confirmed by the investigators, even
though
3 the primary endpoint of the study was
achieving a
4 CR.
5 So what this slide shows here, is
that the
6 primary endpoint of the study was complete
7 remission, and that of the 20 patients who
achieved
8 a complete remission, 17 were confirmed by
repeat
9 bone marrow biopsy at one month. The three that
10 were not confirmed include one patient who
relapsed
11 early, one patient who died while in CR, and
one
12 patient who refused further follow-up with
bone
13 marrow biopsies--and I will go over them in
some
14 detail for you.
15 [Slide.]
16 Patient 318 was an 81-year-old man with 90
51
1 percent blasts at baseline. He achieved a complete
2 remission but had evidence of early relapse
by
3 peripheral counts on day 58.
4 Patient 336 was an 80-year-old man
who
5 achieved also a CR following one cycle of
6 tipifarnib at the recovery of counts at the end
of
7 the cycle.
Upon re-treatment with a second cycle,
8 he developed drug-induced
myelosuppression. This
9 was complicated by neutropenic fungal
sepsis, and
10 then the patient died in CR on day 67.
11 And, finally, patient 508 was a
12 79-year-old man with 90 percent blasts at
the time
13 of study entry. He entered a CR which was
14 established by bone marrow biopsy. He then refused
15 further bone marrow assessments, but was
maintained
16 in CR by continuous treatment for 121 days,
based
17 on peripheral counts.
18 [Slide.]
19 Another aspect of the quality
control
20 measure was an independent review which was
21 performed by Dr. Albitar. This reviewer was
22 blinded to patient outcome.
23 This review involved a retrospective
24 collection of the baseline diagnostic bone
marrow
25 slides from the 136 patients, and in
addition, the
52
1 key aspirate slides which were obtained from
2 patients on treatment. This included slides from
3 18 of the 20 patients who achieved a CR, so
that
4 the independent reviewer had access to a
large
5 proportion, although not all, of the CRs.
6 So his findings are summarized on
this
7 slide.
8 [Slide.]
9 As a result of his review, all 18
10 CRs--this is the first bullet--all 18 CRs
that were
11 available for him were agreed upon, for a
12 concordance rate of 100 percent.
13 The reviewer--and this is the
second
14 bullet--the reviewer also agreed that 15 of
the 16
15 confirmed CRs that were available for his
16 review--there were 17 by the investigators,
16
17 available.
And therefore, from his review of the
18 16, he verified that 15 were indeed
maintained at
19 one month.
20 The one disagreement is in the
patient on
21 whom there was agreement that he achieved a
CR, but
22 there was a disagreement as to the number of
blasts
23 in the follow-up period. The investigators
24 assessed as less than 5 percent, and the
25 independent reviewer assessed it variously
between
53
1 6 percent and 9 percent. Both agreed on the time
2 of relapse.
3 [Slide.]
4 Now, most patients who achieved a
complete
5 remission on this study had more than one
risk
6 factor.
These complete remissions, but also
7 partial remissions and heme improvements were
8 documented regardless of the number of risk
factors
9 present in any given patients. And so the overall
10 rates of anti-leukemic activity ranged from
19
11 percent to 29 percent, irrespective of the
number
12 of risk factors. There is no trend that can be
13 identified.
14 [Slide.]
15 On this Kaplan-Myer plot, the
x-axis is
16 labeled in days. The complete remissions were
54
1 durable.
They range from 33 to 376 days, with a
2 median duration of 220 days. This is slightly more
3 than seven months.
4 The 95 percent confidence interval
around
5 this is 154 up to 275 days.
6 Duration of complete remissions
was
7 calculated starting from the first
documentation of
8 CR until time of relapse. And the dots represent
9 patients that were censored at the time of
clinical
10 cut-off, or the time of death, if they were
still
11 in CR.
12 [Slide.]
13 Now, turning your attention to
survival,
14 on this Kaplan-Myer plot, and on the one
that will
15 follow, patients were censored at the time
of
16 clinical cut-off or last follow-up. 131 patients
17 have complete follow-up, and five patients
had
18 follow-up--at least partial follow-up.
19 The median survival of the
patients who
20 achieved a CR was 433 days. This is in excess of a
21 year.
Two patients were alive at two and three
22 years, respectively. And these data suggest that
55
1 complete remissions are indeed associated
with
2 survival benefit in this patient population.
3 This potential effect on survival
is being
4 investigated in a 301 trial that Dr. DeLap
5 described at the beginning of this
presentation.
6 [Slide.]
7 Finally, the overall survival of
the 136
8
patients is depicted here. The
median survival was
9 164 days, or approximately five to
five-and-a-half
10 months.
11 [Slide.]
12 The information collected by the
13 investigators concerned tipifarnib
administration,
14 obviously, but also treatment administered
after
15 failing the first course of treatment.
16 This slide shows the re-treatment
of
17 patients who achieved a CR--seven of the 20
18 patients who achieved a CR chose to stop
after
19 three cycles. They were re-treated with tipifarnib
20 at relapse.
One of these seven patients achieved a
21 second complete remission of similar
duration to
22 the first one: approximately six months.
23 So these data suggest that
patients may
24 remain sensitive to tipifarnib at the time
of
25 relapse.
56
1 [Slide.]
2 The use of chemotherapy after
tipifarnib
3 in the 136 patients was also
recorded. The
4 majority of the patients went on to receive
5 palliative care only.
6 Only 12 patients, most of whom
were less
7 than 75, were able to receive intensive
8 chemotherapy, usually anthra-cycline plus
Ara-C.
9 [Slide.]
10 So, in summary, tipifarnib is
active in
11 elderly, poor-risk AML. The CTEP investigators
12 documented a 15 percent complete remission
rate,
13 which is the primary endpoint of the study.
14 Most of the complete remissions
were
15 confirmed at one month by the investigators,
and
16 verified by the independent reviewer. Complete
17 remissions were durable, lasting 220 days,
or 7.2
18 months.
The median survival of patients with
19
complete remission was 433 days,
or 14.2 months.
20 [Slide.]
21 This enters the final section of
the
22 presentation, which concerns safety.
23 In terms of drug exposure, the
median
24 treatment cycle duration was 38 days. 47 percent
25 of the patients received two or more
cycles. So
57
1
this includes patients with complete remissions,
2 partial remissions, and hematologic
3 improvement--but also several patients who
4 maintained stable disease for s several
months.
5
[Slide.]
6 The need for treatment
interruptions or
7 delays explained the median cycle duration
of 38
8 days which I just presented. Those reductions were
9 implemented in 35 percent of the patients.
10 Reductions were implemented in approximately
half
11 of the patients who received two or more
cycles,
12 such as the patients who achieved a CR.
13 The most common reason for dose
reductions
14 were related to myelosuppression,
gastrointestinal,
15 CNS--and, rarely, renal or dermatological
signs and
16 symptoms.
17 Patient age did not appear to have
an
18 impact on the tolerability.
19 [Slide.]
20 So, as expected in a population
with AML,
21 the adverse events were common: 61 percent
of the
22 population experienced drug-related adverse
events.
23 These were mostly related to
myelosuppression--in
24 the background of, of course, severe
25 myelosuppression.
58
1 In contrast, only 10 percent of
the
2 patients were withdrawn for drug adverse
event.
3 And adverse events were also associated with
a low
4 mortality rate. Only nine patients in whom an
5 adverse event--of the nine patients,
actually, in
6 whom an adverse event led to death, only one
was
7 assessed by the investigators as related to
8 tipifarnib.
9 [Slide.]
10 To go into more details, a
majority of
11 patients experienced Grade 3 or 4
myelosuppression
12 by peripheral count assessments. Fewer patients
13 went on to develop clinical adverse events
in the
59
1 form of sepsis or bleeding complications.
2 And, as I mentioned, approximately
3 two-thirds had Grade 3 myelosuppression at
the time
4 of study entry, and so the overall incidence
has to
5 be interpreted in this context.
6 [Slide.]
7 Turning our attention to
non-hematologic
8 adverse events--that is, events excluding
9 myelosuppression and its complications--the
overall
10
rate of life-threatening or Grade 4 events was low:
11 2 percent.
Most events were reversible following
12 treatment interruption.
13 The GI tract was most commonly
involved.
14
Nausea, diarrhea, vomiting were most often mild to
15 moderate.
The incidence of drug-related mucositis
16 was only 3 percent. This is what this shows--5
17 percent, and 3 percent if we consider Grade
3 or 4
18 in severity.
19 Now, given that mucositis is a
major
20 contributor of morbidity and death in
21 myelosuppressed patients, this is probably
the most
22
important--the most important
safety advantage of
60
1 tipifarnib in this older population.
2 [Slide.]
3 Rare adverse events affecting the renal
4 and CNS systems were documented. Few reached Grade
5 3 or 4 severity. All these adverse events were
6 managed appropriately by protocol-defined
treatment
7 interruption and dose
reductions. Many appeared in
8 the context of sepsis, and were of short
duration.
9 Rapid ejaculation, the median duration of
CNS
10 events was two to three days.
11 [Slide.]
12 Very few patients--as we show
here--were
13 removed from the study because of these
events.
14 Indeed, the adverse events that
led to the
15
termination of treatment were varied.
No one AE
16 appears to be a major contributor. The most common
17 causes were elevation of serum creatinine
and skin
18 rash, both of which were reversible.
19 [Slide.]
20 This study did not have a specific
module
21 collecting data on quality of life as we
would
22 have.
So, hospitalization data provides a valuable
61
1 perspective on the safety profile and the
patient
2 tolerability of this drug. And what it suggests is
3 that outpatient treatment of this elderly
4 population is feasible.
5 Up to 40 percent of the patients
received
6 their full course of treatment as
outpatients. The
7 majority of patients who were hospitalized
were
8 hospitalized only once or twice.
9 The median duration of combined
hospital
10 stay was 15 days--all hospitalizations. And this
11 represents 14 percent of the patients spent
on
12 study.
13 [Slide.]
14 Finally, few patients died from adverse
events on
15 this study.
No single cause appears to account for
16 the majority of cases, most of which appear
related
17 to a complication of AML.
18 In the opinion of the
investigators, only
19 one death from neutropenic fungal sepsis was
20 related to tipifarnib.
21 [Slide.]
22 So how can we best summarize
CTEP-20?
62
1 First of all, CTEP-20 was a study of patients
with
2 significant unmet medical need, determined
by the
3 advanced age, and the high prevalence of
risk
4 factors which are associated with poor
patient
5 outcome.
6 In this patient population,
tipifarnib
7 induced as a monotherapy a 15 percent
complete
8 remission rate that was both durable and
9 independent of risk factors.
10 The safety profile of tipifarnib allowed
11 outpatient treatment. This might be due to a very
12 low rate of life-threatening
non-hematological
13 toxicity, especially when one considers
mucositis.
14 So, consequently, the time spent
in
15 hospital represented a small fraction of the
total
16 study time: approximately 14 percent. Only one
17 death was attributable to tipifarnib.
18 This concludes my review. And Dr. Alex
19 Zukiwski will deliver the closing
presentation for
20 Johnson & Johnson.
21 Benefit/Risk
22 DR. ZUKIWSKI: Thank you, Dr. Thibault.
23 Good morning. I'm Alex Zukiwski from the
24 Oncology Development Group at Johnson &
Johnson
25 Pharmaceutical Research and
Development. I'm going
63
1 to conclude the presentation today with a
summary.
2 The proposed indication is:
tipifarnib is
3 indicated for the treatment of elderly
patients
4 with newly diagnosed poor-risk acute myeloid
5 leukemia.
The basis of this approval is complete
6 remissions-- an accepted efficacy endpoint
in AML
7 which has been shown to correlate with
overall
8 survival.
9 [Slide.]
10 It is evidence that elderly
patients with
11 poor-risk AML have limited therapeutic
options. As
12 noted in Dr. Stone's presentation, select
elderly
13 patients should be considered for
chemotherapy,
14 although older patients do not do as well as
15 younger patients. These patients who receive
16 induction chemotherapy were described as
having the
17 best ability to tolerate and benefit from
such
18 treatment.
19 In the United States,
approximately
64
1 two-thirds of the patients greater that 65
years of
2 age do not receive IV chemotherapy. This is due to
3 an unfavorable benefit-risk. These patients have
4 risk factors which predispose to decreased
5 efficacy, including antecedent hematological
6 disorders such as myelodysplastic syndrome,
and
7 also have unfavorable cytogenetics.
8 They also have factors which
predispose
9 them to increased toxicity, such as
co-morbidities
10 and compromised performance status.
11 As indicated in the NCCN
guidelines,
12 options available for this under served
patient
13 population include investigational studies,
14 low-intensity chemotherapy,
and--unfortunately for
15 many of these patients--it is simply best
16 supportive care.
17 [Slide.]
18 The patients enrolled in CTEP-20
were felt
19 not be well-served by standard induction
20 chemotherapy. And this represents a unique patient
21 population which is not well represented in
the
22 literature. For example, the CALGB and ECOG
studies
65
1 mentioned in Dr. Stone's presentation
enrolled
2 patients who were good candidates for
induction
3 chemotherapy, and many of these studies
excluded
4 those patients with prior myelodysplastic
syndrome.
5 The median age of the CTEP study
6 population was 75 years, and 90 percent of
the
7 CTEP-20 patients had two or more risk
factors which
8 predisposed them to decreased efficacy and
9 increased toxicity from conventional
anti-leukemic
10 therapies.
11 [Slide.]
12 The complete remission rate in
this very
13 difficult to treat patient
population was 15
14 percent by investigators' assessment. As per the
15 key academic investigators involved in the
CTEP
16 study, this response rate is meaningful in a
17 patient population that is often excluded
from AML
18 studies, and many times receives palliative
care
19 only.
20 Of the 20 complete remissions as
assessed
21 by the investigators, the independent
reviewer was
22 able to confirm and verify 15 complete
remissions
66
1 for quality control purposes. While the other five
2 cases could not be verified for a variety of
3 reasons, there is evidence of substantial
4 anti-leukemic activity in these five
patients.
5 The median duration of complete
remissions
6 was 220 days, and complete remissions were
observed
7 across all risk groups.
8 The exploratory analysis of
survival--as
9 outlined by Dr. Thibault--is encouraging,
and will
10 be further examined in an AML study
specifically
11 designed to evaluate a survival endpoint.
12 [Slide.]
13 The anti-leukemic activity observed
in the
14 CTEP-20 study has to be considered in the
overall
15 treatment context as presented by DR.
Stone. Even
16 in the "best" patients who can
receive induction
17 chemotherapy, individual risk factors which
were
18 very prominent in the CTEP-20 study have a
negative
19 impact on complete remission rates.
20 As shown in this slide, a
reduction of
21 complete remission rates by approximately
one-half
22 is observed in patients with prior
myelodysplastic
67
1 syndrome, unfavorable cytogenetics, or
advanced
2 age--with even the most effective
chemotherapeutic
3 agents.
4 The early death rates in these
trials,
5 which includes the more younger and fit
patient
6
populations, is approximately 20 to 25 percent,
7 with the majority of these deaths associated
with
8 bone marrow aplasia and severe mucositis.
9 [Slide.]
10 Now, to put the CTEP-20 study into
11 context.
12 These are patients which
experienced AML
13 investigators felt were not the best
candidates for
14 combination chemotherapy. They had an 83 percent
15 incidents of myelodysplastic syndrome; a 49
percent
16 incidence of unfavorable cytogenetics, and a
median
17 age of 75.
18 What would be the anticipated
complete
19 remission rate in this patient population
with
20 combination chemotherapy?
21 The anticipated early death rate
in this
22 patient population with combination chemotherapy
68
1 would most likely be at least double the 12
percent
2 early death rate observed in the CTEP-20
trial.
3 [Slide.]
4 In the patient population studied
in
5 CTEP-20, advanced age, with its associated
6 co-morbidity co-morbidities, and the
underlying
7 disease complicates any treatment efforts.
8 Despite this, a management and
predictable
9 safety profile was observed in the patients
treated
10 with tipifarnib. Adverse events were managed with
11 supportive care measures. As the scheduled
12 treatment was for 21 days, dose
13 modifications--including dose interruptions
and
14 dose reductions--could be quickly
implemented to
15 address any emerging toxicities.
16 Forty percent of the patients did
not
17 require hospitalization. And for those who were
18 hospitalized, the median duration was
approximately
19 15 days. The limited time spent in hospital is an
20 important feature of this outpatient
treatment.
21 Twelve percent of the patients
died on
22 study within 30 days of the first dose of
69
1 tipifarnib.
And only one treatment-related death
2 was reported by the investigators during the
study.
3 This data indicates that tipifarnib has been
shown
4 to safely treat a unique elderly patient
5 population.
6 [Slide.]
7 In conclusion, the application
under
8 review is focused on a difficult to treat
patient
9 population for whom an unmet medical need
exists,
10 and alternative treatments are required.
11 Approximately one in seven to one in nine of
the
12 patients treated with tipifarnib developed a
13 durable complete remission.
14 This novel targeted therapy is
15 administered as an oral tablet which allows
for
16 flexible outpatient dosing. As presented here
17 today, there is a positive benefit-risk
evidence
18 with tipifarnib treatment.
19 Approval of tipifarnib for this
orphan
20 indication will provide a new treatment for
elderly
21 patients with poor-risk AML. These patients are
22 currently not well served. Standard induction
70
1 chemotherapy has a high degree of toxicity
and a
2 lower degree of efficacy in this patient
3 population.
4 This concludes the sponsor's
presentation.
5 DR. MARTINO: Thank you.
6 Ladies and gentlemen, I will not
allow
7 questions until we have also allowed the FDA
and
8 Dr.
9 Fred Appelbaum to present. So those of you that
10 have questions please know that that will be
your
11 opportunity.
12 Dr. Ryan, representing the FDA,
will
13 present next.
14 FDA Presentation
15 NDA 21-824, Zarnestra
16 DR. RYAN: Good morning. I'm Qin Ryan,
17 medical officer in the Division of Oncology
Drug
18 Products.
I'm here today to present the main
19
findings from our review of Zarnestra NDA 21824.
20 [Slide.]
21 My presentation will cover the
relevant
22 regulatory background; clinical development
71
1 overview and proposed indication; CTEP-20
study
2 design, efficacy and safety findings.
3 [Slide.]
4 First, the relevant regulatory
background.
5 In oncology, clinical benefit has
been
6 defined as a longer life, a better life or
an
7 effect on an established surrogate for
either.
8 In acute leukemias, durable
complete
9 remission--CR--has been considered evidence
of
10 benefit.
11 In some cases where leukemia CRs
were of
12 uncertain duration, CR was considered a
surrogate
13 reasonably likely to predict clinical
benefit.
14 [Slide.]
15 Here are some examples. As first-line
16 indications for acute myeloid leukemia
--AML--both
17 idarubicin and daunorubicin were regular
approvals
18 based on demonstration of durable remissions
in
19 randomized trials. In addition, idarubicin also
20 demonstrated a survival advantage in two
21 comparative studies.
22 In the case of gemtuzumab,
accelerated
72
1 approval was granted for patients aged 60 or
older
2 with CD-33 positive disease who are not
candidates
3 for second-line cytotoxic chemotherapy. Approval
4 was based on a pooled complete remission
rate from
5
three single-arm studies. Of 277 patients enrolled
6 into these three studies, 157 were 60 years
or
7 older.
Although relaxed free survival was
8 evaluated, the ratio of remission could not
be
9
reliably ascertained, as 45 percent of patients who
10 achieved remission also received additional
11 anti-leukemic therapy.
12 [Slide.]
13 Next, I will discuss the clinical
14 background for this NDA.
15 Patients with newly-diagnosed AML,
if not
16 treated, will progress rapidly to a fatal
outcome.
17 The standard induction therapy for newly
diagnosed
18 AML, such as 3+7 regimen of cytarabine and
19 daunorubicin can be expected to achieve 60
to 75
20 percent complete remission. And the
21 treatment-related death rate, less than 10
to 20
22 percent in adult AML patients younger than
age 60.
23 [Slide.]
24 One follow-up study indicated that
30 to
25 40 percent of patients in this age group will
73
1 survive three years or more. However, clinical
2 outcome would depend on multiple factors.
3 Unfavorable outcome is usually associated
with
4 multiple poor-risk factors, such as poor
5 performance status, organ dysfunction, or
6 significant co-morbidity, older age,
unfavorable
7 karyotype, prior MDS, disease resistance
8 or patient intolerance to cytotoxic therapy.
9 Although the incidence of adult
AML
10 increases with age, the chance for patients
to
11 receive treatment decreases.
12 [Slide.]
13 Menzin, et al., analyzed a data
base of
14 over 2,600 AML patients aged 65 or older,
15 identified by Medicare claims. Overall, only 30
16
percent of those patients received some form of
17 chemotherapy. Asa shown here, the percentage of
18 patients receiving chemotherapy decreased
19 drastically with increasing age.
20 [Slide.]
21 Menzin, et al., also estimated
survival
22 for these patients. As indicated by the curve with
23 diamond points, the median survival for all
24 identified patients was two months, with a
two-year
25 survival of 6 percent.
74
1 [Slide.]
2 Compared to younger adults,
elderly AML
3 patients usually present with other risk
factors,
4 such as poor performance status, organ
dysfunction,
5 co-morbidity, unfavorable karyotype, prior
MDS, and
6 drug resistant disease, as mentioned
before. The
7 less tolerant to standard induction therapy
the
8 elderly poor-risk AML patients are, the more
likely
9 they are to be a therapeutic challenge.
10
[Slide.]
11 One review pointed out that
12 treatment-related mortality in elderly
patients
13 with poor-risk AML may be as high as 25
percent,
14 and the complete remission rate may be less
than 50
15 percent.
16 In a study of AML patients at
least 80
75
1 years of age, the mortality rate at one
month as 48
2 percent, and the complete remission rate was
less
3 than 30 percent.
4 Few elderly AML patients are
expected to
5 live free of disease after standard
cytotoxic
6 chemotherapy.
7 [Slide.]
8 Next, I will discuss the clinical
program.
9 Zarnestra is a farnesyl
transferase
10 inhibitor.
It is formulated in film coat 100 mg
11 tablets.
In the CTEP-20 study, Zarnestra was
12 tested as first-line AML induction
therapy. It was
13 administered hourly with food, at a dose of
600 mg
14
twice daily for 21 days, followed by a rest period
15 of seven to 42 days.
16 [Slide.]
17 Zarnestra is being proposed for
the
18 treatment of elderly patients with newly
diagnosed
19 poor-risk acute myeloid leukemia.
20 [Slide.]
21 Eleven studies relevant to safety
and dose
22 findings have been submitted in this
NDA. Among
76
1 those, the studies relevant to efficacy and
safety
2 in AML are summarized here. The most relevant
3 population for the proposed indication is a
4 subgroup of subjects in Study CTEP-20,
accounting
5 of 79 percent of CTEP-20 enrollment.
6 The studies INT-17 and CTEP-1 are
less
7 relevant to the proposed indication. Therefore we
8 will focus our discussion on CTEP 20.
9 I will now go over the CTEP-20
study
10 design in the next few slides.
11 [Slide.]
12 This open-label, single-arm,
multicenter
13 study which opened on October 10, 2001, was
14 originally designed to assess the efficacy
and
15 safety of Zarnestra in subjects with
previously
16
untreated, poor-risk hematologic malignancies.
17 After enrolling 110 patients, Amendment 6
was
18 implemented on September 16, 2003. The target
19 population as redefined as subjects either
75 years
20 or older with newly diagnosed AML, or 65 to
74
21 years of age with AML arising from prior
22 myelodysplastic syndrome.
23 [Slide.]
24 The original primary objective was
to
25 determine response rate, which included CR
and PR.
77
1 As mentioned, after the 6th Amendment, the
primary
2 objective changed to
determining the complete
3 remission rate in elderly subjects with
previously
4 untreated, poor-risk acute myeloid leukemia.
5 Secondary objectives were to determine
the
6 progression-free survival, overall survival,
7 duration of response, and safety profile.
8 [Slide.]
9 After Amendment 6, the major
inclusion
10
criteria can be described as follows: untreated
11 newly diagnosed AML patients, 75 years or
older, or
12 65 to 74 years of age with prior MDS.
13 Our eligible subjects should have
14
pathologic confirmation of AML showing equal or
15 more than 20 percent marrow or peripheral
blasts.
16 Patients must have an ECOG performance score
of
17 zero or one-- which was changed from the original
18 zero to two--with adequate renal and liver
19 function.
20 Patients with the following
conditions
21 were excluded: hypoleukocytosis despite
22 leukopheresis or hydroxyurea; acute
promyelocytic
23 leukemia; previous anti-leukemic
chemotherapy other
24 than hydroxyurea; disseminated intravascular
25 coagulation, or leukemia involvement of the
central
78
1 nervous system.
2 [Slide.]
3 Leukemia assessments were
conducted at
4 baseline and the end of each cycle, ranging
from 29
5 to 64 days.
This included medical history,
6 physical examination, bone marrow aspiration
and
7 biopsy, CBC and chemistry.
8 The criteria for response
assessment were
9 based on NCI-sponsored workshop on
definition of
10 diagnosis and response in acute myeloid
leukemia.
11 Per protocol, CR is defined as
marrow
12 showing less than 5 percent myeloblasts,
with
13 normal maturation of all cell lines; an ANC
of at
14 least 1,000 per micro liter; a platelet
count of
15 100,000 per micro liter; absence of blasts
in
16
peripheral blood; absence of identifiable leukemic
79
1 in the bone marrow; clearance of
disease-associated
2 cytogenetic abnormalities; and clearance of
any
3 previously existing extramedullary disease.
4 In addition, CR must be confirmed
four to
5 six weeks after initial documentation; at
least one
6 bone marrow biopsy should be performed to
confirm
7 CR.
8 Subjects who achieved a complete
remission
9 could receive additional Zarnestra treatment
until
10 disease progression, or receive up to three
11 additional cycles and stop.
12 Re-treatment with Zarnestra was
allowed.
13 Subjects with progressive disease at any
time
14 during the Zarnestra administration were
withdrawn
15 from the study. The first follow-up occurred 30
16 days after treatment termination for
subjects who
17 did not have a documented progression, or
had not
18 started subsequent therapy; every 90 days
after
19 documentation of progressive disease or
start of
20 subsequent therapy.
21 In the next few slides, I will
discuss the
22 CTEP-20 patient population efficacy data.
23 [Slide.]
24 Of the total 171 patients enrolled
in
25 CTEP-20, 158 of them had AML. At the time of
80
1 clinical cutoff, 157 AML subjects were
treated with
2 at least one cycle of Zarnestra. Of those, 136
3 were elderly subjects with poor-risk AML,
and are
4 most relevant to the proposed indication.
5 Please note that one of the
elderly
6 subjects with poor-risk AML was excluded
from FDA's
7 efficacy analysis, but not safety
analysis. The
8 reason for this exclusion was that this
patient's
9 baseline blast count was less than 20,000
per cubic
10 milliliter, as assessed by both the
investigator
11 and the sponsor's central review. This patient did
12 not respond to Zarnestra treatment.
13 This resulted in 156 evaluable
patients in
14 the all-treated AML population, and 135
patients in
15 the elderly poor-risk AML population.
16 Two thirds of subjects had a
performance
17 status of one, and a quarter of them had a
18 performance status of zero. There were
19 approximately 10 percent of patients who were
81
1 enrolled before Amendment 6, with a
performance
2 status of two. As per investigators, all 136
3 patients were ineligible for standard
chemotherapy
4 for at least one reason.
5 Of patients aged 75 or older, 96
percent
6 were considered ineligible due to age,
whereas in
7 the 65 to 74-year-old group, approximately
50
8 percent of patients had age or other risk
factors
9 as a reason for ineligibility.
10 [Slide.]
11 Based on the sponsor-provided
data, risk
12 factors in the CTEP-20 elderly poor-risk AML
13 population are summarized by category and
number.
14 Eighty-two percent of patients had
prior
15 MDS; more than half of the patients were
older than
16 75 years, or with poor organ function as
defined by
17 the sponsor; and two-thirds of patients had
18 unfavorable karyotypes; 44 percent and 35
percent
19 of patients had at least two or three of
these risk
20 factors, respectively. About 10 percent of
21 patients had either one or four risk
factors.
22 Complete responders were initially
82
1 assessed by the site investigators. The
sponsor
2 appointed an independent reviewer to
reassess the
3 complete remissions. FDA requested all available
4 bone marrow slides of CRs from the sponsor,
and
5 reviewed them with an FDA-appointed
hematology
6 consultant.
7 [Slide.]
8 The assessment of CR by the study
9 investigator, the independent reviewer, and
FDA are
10 summarized here. Of the three unconfirmed CRs, the
11 FDA and independent reviewer agreed with the
12 investigator that two could not be confirmed
due to
13 death and disease progression. FDA also agrees
14 with the independent reviewer that on CR by
15 investigator assessment was unconfirmed due
to
16 insufficient data. In addition, slides of two
17 subjects were not available for response
assessment
18 by the sponsor's independent reviewer or FDA
19 consultant.
20 [Slide.]
21 FDA agrees with the
sponsor-appointed
22 independent reviewer's assessment of
complete
83
1 remission; that is, 15 subjects were
confirmed
2 complete remission from the FDA-identified
elderly
3 poor-risk AML patient subgroup.
4 FDA assessment of the confirmed
complete
5 remission rate is 11.1 percent, with a 95
percent
6 confidence interval of 6.6 to 18 percent.
7 [Slide.]
8 Based on FDA exploratory subgroup
9 analysis, patients older than age 74 have a
lower
10 tendency to achieve complete remission than
do
11 patients age 65 to 74, with a rate of 6.7
percent
12 versus 16.4 percent, respectively.
13 In addition, of the 25 patients
older than
14 74 years with de novo AML who enrolled in
CTEP-20,
15 only one patient achieved complete remission
with
16 confirmation.
17 Less responders were seen in
subjects with
18 unfavorable karyotypes.
19 [Slide.]
20 The FDA has explored the duration
of
21 confirmed CR as a secondary endpoint of
study
22 CTEP-20.
Per protocol, no anti-leukemic therapy
84
1 other than Zarnestra was given to patients
who
2 achieved a response until after disease
progression
3
and removal from the study.
4 At the time of cut-off,
progression of
5 disease or death occurred in seven of 15
patients,
6 giving a median remission duration of 275
days,
7
with 95 percent confidence interval of 127 to 376
8 days.
9 Next, I will discuss the CTEP-20
safety
10 data.
11 [Slide.]
12 In CTEP-20, all of the elderly
poor-risk
13 AML patients received at least Zarnestra
during
14 first cycle; 47 percent of them were treated
for a
15 second cycle, and 20 percent received a
third
16 cycle.
17 The median duration of these
cycles was 36
18 to 38 days.
The mean intensity for the first cycle
19 was 749.4 mg per day, which is approximately
63
20 percent of the planned 1,200 mg per day
dose.
21 The calculated dose intensity may
not
22 reflect true drug exposure, since the
Zarnestra
85
1
exposure measurements were primarily based on the
2 pharmacy dispensation record. A patient medication
3 diary was not planned for CTEP-20.
4 [Slide.]
5 Ninety-eight percent of CTEP-20
subjects
6 experienced adverse events. The most frequently
7 reported non-hematological adverse events
were
8 diarrhea, fatigue, nausea, skin rash, fever,
9 anorexia, constipation, vomiting and
dyspnea.
10 Dizziness, and ataxia or abnormal gait were the
11 most frequently seen nervous system adverse
events.
12 In addition, confusion was the most commonly
seen
13
psychiatric adverse event. This
may be related to
14 age and hospitalization.
15 The most common dermatological and
16 infection-related adverse event were
neutropenia,
17
with or without fever; purpurea, thrombocytopenia,
18 anemia, bacterial infection, candida and
other
19 fungal infections.
20 The most frequent metabolic
disturbances
21 were increased creatinine, hypokalemia, and
22 hyponatremia.
23 Of 136 subjects, 21, 47 and 56
subjects
24 had at least one adverse event leading to
treatment
25 termination, dose reduction, and temporary
86
1 interruption of Zarnestra,
respectively. The top
2 three adverse events leading to changing
treatment
3
were neutropenia, increased creatinine and rash.
4 FDA agrees with the sponsor that
113
5 subjects, or 83 percent of the elderly
poor-risk
6 AML group, experienced Grade 3 or 4 adverse
events.
7 The most frequent treatment-emergent Grade 3
or 4
8 adverse events were secondary to
myelosuppression,
9 including neutropenia, with or without
fever,
10 infection, thrombocytopenia, and anemia.
11 Other frequent severe adverse
events were
12 fatigue, rash, dyspnea, confusion, diarrhea,
and
13 hypokalemia.
14 [Slide.]
15 Thirty-one of the 136 elderly
poor-risk
16 AML subjects in CTEP-20 died, either within
30 days
17 of treatment termination, or within 30 days
of
18 receiving the first dose of medication. The death
19
rate within 30 days of the first dose was 12
87
1 percent. Based on the sponsor-provided data,
we
2 verified the sponsor's summary and agree
that 19 of
3 31 deaths were due to disease progression,
and nine
4 of them were due to adverse events. The deaths due
5 to adverse events were 7 percent with causes
such
6 as cardiac failure and various infections.
7 One death due to adverse event was
thought
8 to be drug-related by the investigator. This was a
9 patient who had a neutropenic fever, fungal
10
infection, and renal dysfunction.
11 There were three of 31 deaths
attributed
12 to adverse events or progression of disease
on
13 subsequent treatment, after patients
progressed
14
from Zarnestra, which the sponsor categorized as
15 "other" cause of death.
16 [Slide.]
17 Eighty-one subjects, or 60 percent
of the
18 total 136 patients, were hospitalized during
the
19 study.
Fourteen percent of the subjects received
20 Zarnestra treatment in the outpatient
setting
21 completely.
22 The median total duration of
88
1 hospitalization was 15 days. Ten subjects required
2 at least three hospitalizations during the
study
3 period.
4 [Slide.]
5 In summary, durable complete
remission has
6 been accepted as an endpoint supportive of
regular
7 approval in AML. Zarnestra efficacy findings
8 should be considered in the context of a
poor-risk
9 AML population and the toxicity profile
observed.
10 Although the remission rate does not compare
11 favorably with that reported with cytotoxic
12
therapy, the one-month's mortality and treatment
13 related date rate of 12 percent and 7
percent,
14 respectively, compare favorably with the
greater
15 than 25 percent treatment-related death rate
16 reported in the literature for patients age
60 or
17 older, with or without other risk factors,
who had
18 adequate organ function to receive
chemotherapy.
19 We will have one question for the
20 committee to discuss: does the risk-benefit
21 analysis support regular approval of
Zarnestra for
22 the treatment of elderly patients with AML?
23 Thank you very much for your
attention.
24 DR. MARTINO: Thank you, Dr. Ryan.
25 At this point, ladies and
gentlemen, we
89
1
have one additional speaker, and that is Dr. Fred
2 Appelbaum, who will present via
videoconference at
3 a quarter to the hour.
4 So, at this point, I'm going to
give you
5 about 15, 20 minutes of a break, and we'll be
back
6 here at 10:45 for his video
presentation. We will
7 take questions subsequently.
8 [Off the record.]
9 DR. MARTINO: Back on the record.
10 The next presentation is Dr. Fred
11 Appelbaum.
He's going to speak to us via
12 videoconference. He is the Director of Clinical
13 Research at Fred Hutchinson Cancer
Center. And I
14 don't know if we are actually ready to go.
15 Ms. Clifford? Are we ready to go with Dr.
16 Appelbaum?
17 MS. CLIFFORD: I think so.
18 DR. MARTINO: I someone going to
clue him
19 in?
Or shall Dr. Martino simply say: "Action."
20 AML in Older Individuals
21 [Via videoconferencing]
22 DR. APPELBAUM: I couldn't hear
anything
23 you were saying. Can you see my slides?
24 VOICE: [Off mike.] [Inaudible.]
25 You won't be able to follow the
pointer,
90
1 is that right?
2 Well, thank you for inviting me to
say a
3 few words about AML in older
individuals. I was
4 asked by the FDA just to provide a general
5 background.
I am not speaking particularly about
6 this product or any of the information that
was
7 presented about the product, but rather just
a
8 global picture of AML in the elderly. Some of this
9 was probably already gone over this morning,
so I
10 will be relatively brief.
11 [Slide.]
12 The problem of AML
in the elderly is a
13 substantial one because the disease, as you
can see
14 from the first slide, in terms of incidence,
goes
15 up quite rapidly once patients are in their
sixth
16 decade.
Before that, it is relatively uncommon.
91
1 But once patients pass the age of 50, the
incidence
2
of AML goes up quite markedly.
3 The problem of AML in the
elderly--or the
4 older individual--is different than AML in
the
5 younger individual for two primary
reasons. First,
6
the patients are different and, secondly, the
7 disease is different.
8 In terms of patients being
9 different--well, older patients are
older. And,
10 with that comes an increased incidence
11 co-morbidities and decreased performance
status.
12 [Slide.]
13 This next slide shows you a
typical
14 picture of patients that were entered onto a
15 protocol for patients with AML above age 55,
using
16 a standard induction regimen of daunomycin
and
17 Ara-C.
18 As you can see from the
performance status
19 and the age, patients under age 60, a
relatively
20 small proportion of them--about 8
percent--will
21 have a very poor performance status of 3 or
22 greater; whereas once patients are over age
75,
92
1 that incidence at least doubles.
2 Yet, even on those over age 75, at
least
3 60 percent--on this Southwest Oncology Group
Study,
4 which was the last one performed--had a
performance
5 status of zero or one, indicating relatively
good
6 performance.
7 These statistics almost certainly
8
understate the problem of decreased performance
9 status in the elderly, because these are
patients
10 that the doctors chose to enter onto the
trial.
11 And it may be that, particularly among the
elderly,
12 there's a substantial proportion who were
not
13 entered onto the trial because of decreased
14 performance status.
15 I know of no easy--or no way, in
fact, at
16 all--that we can get data which truly
reflects the
17 performance status on a population basis of
the
18 elderly patients with AML. This is almost
19 certainly an underestimate of the difficulty.
20 [Slide.]
21 The next slide shows, I think, a
very
22 important principle that hasn't been talked
93
1 about--or it hasn't been published, to my
2 knowledge, quite in this way--and that is
the very
3 great interaction between performance status
and
4 age.
So that this looks at patients, again,
5 treated with a standard induct of
Daunomycin-45 x 3
6 and Ara-C, and a chance of dying within the
first
7 month of being entered onto study, based on
the
8 performance status and age.
9 So that older patients--that is,
those
10 that are over age 70, have a relatively low
chance
11 of dying within the first 30 days if they
have a
12 good performance status--only 9 percent,
which is
13 not substantially different than what you'd
see in
14 patients even under age 50.
15 Yet once patients get older, and
their
16 performance status goes done, then you see a
marked
17 interaction.
So that if you're both over age 70
18 and have a performance status of 3, the
chance of
19 dying within the first 30 days is 62
percent, which
20 is very, obviously, substantial. So there's this
21 interaction which is--both of performance
status,
22 because the younger patients don't have that
high
94
1 chance of dying, even with a poor
performance
2 standard.
That's only 17 percent. So you
can see
3 there's this interaction with both age and
4 performance status, which are cumulative.
5 [Slide.]
6 This, of course, reflects what
would
7 happen with the overall complete response
rate--I'm
8 sorry, this slide simply shows what I just
told
9 you, but in a different graphic form. That is
10 performance status and age being cumulative
in
11 terms of the chance of early death, with 62
12 percent--that's in the back right, versus a
very
13 low chance, in the front left, if you are
young and
14 have a good performance status.
15 [Slide.]
16 This clearly reflects--on complete
17
response rates, which are shown in the next slide,
18 so that if you're over age 70 treated with
standard
19 chemotherapy and have a good performance
status,
20 you have a relatively good chance--50 percent
or
21 greater--of getting a complete
response. But if
22 you're older and have a poor performance
status, it
95
1 deteriorates to, in this study, 29 percent.
2 Whereas if you're younger and have a poor
3 performance status, the interaction doesn't
seem to
4 be as great.
The numbers in these individual cells
5
get fairly small with a total n of 500.
6 So, the first way that AML in the
elderly
7 or the older patient differs is that the
patients
8 are older, they tend to have a poor
performance
9 status, and a poor performance status is a
clear
10 bad prognostic factor for getting a CR and
for
11 early death.
12 The second way that AML in the
older
13 patient differs is that is biologically is
14 different from AML in the younger
patient--in
15 general; not in each specific case, but as a
16 population it differs.
17 AML in the older patient is more
often
18 preceded by myelodysplasia. It is a less
19 proliferative disease. It's more frequently
20 associated with unfavorable cytogenetics,
and it
21 more often expresses multidrug resistence.
22 And I'll show you data about each
of
96
1 these.
2 First, as far as "preceded by
3 myelodysplasia"--and here one has to be
careful
4 about the definitions. If one takes a strict
5 definition of having a documented
hematologic
6 disorder preceding the diagnosis of AML by
at least
7 three months, generally this is seen in
about 15 to
8 18 percent of patients who are over age 55,
and
9 seen in about half that number in patients
who are
10 less than age 55.
11 [Slide.]
12 As far as "less
proliferative" is
13 concerned, that's shown on the slide you now
have
14 in front of you. And this is a large number of
15
patients--about 900 patients on three sequential
16 studies--and it just shows that the white
count, at
17 diagnosis for AML, in patients who were less
than
18 age 55 is about 17,000, but it goes down to
about
19 12,000 when you're over age 75. And the percent
20 blasts if you're less than age 55 tends to
be
21 higher, at 39 percent; but if you're over
age 75 it
22 drops to about 26 percent--not marked
differences,
97
1 but there is a biologic difference here
which
2 suggests that AML in the very old patient
tends to
3
be a less highly proliferative disease.
4 [Slide.]
5 The next slide shows the marked
difference
6 in cytogenetics as patients age in AML.
7 The blue bars at the very top are
the
8 percent of AML patients with unfavorable
9 cytogenetics, according to age. So that if you're
10 less than age 55, about 21 percent of
patients will
11 have unfavorable cytogenetics, using the
SWOG--or
12 Southwest Oncology Group--criteria for
cytogenetic
13 risk group--which is very similar, but not
quite
14 identical, to the MRC's definition.
15
If patients are over age 75, the incidence
16 of unfavorable cytogenetics increases
markedly from
17 21 percent to 52 percent; and, conversely,
the dark
18 bar at the bottom, the percent with
favorable
19 cytogenetics--that's 8;21, inversion 16, or
15, 17,
20 then that drops from 20 percent in patients
who are
21 less than age 56, to only 4 percent if
you're over
22 age 75.
23 The particular cytogenetic
abnormalities
24 which are seen more frequently as one ages
are
25 shown on the next slide.
98
1 [Slide.]
2 And they are of marked increase in
the
3 incidence in the loss of -5, or part -5, on
the
4 long arm, or loss of -7 or part of 7 on the
long
5
arm, where either one of these was seen in either 6
6 to 8 percent in patients age less than 56,
but if
7 you're over age 75, you see this in a three
or
8 four-fold higher incidence, with 26 or 22
percent,
9 respectively.
10 Similarly, a loss of the short arm
of 17
11 is seen in only 2 percent less than age 56,
and
12 greater than 11 percent in those age greater
than
13
75. And all those p-values you
can see are .0001.
14 Conversely, as I've already
mentioned,
15 inversion 16, and 8;21s drop as you get
older,
16 among the favorable cytogenetic risk group.
17 It is, I think, just
intellectually
18 interesting to speculate why we see this
particular
19 differences with age. We don't, in fact, know the
99
1 answer.
2 Some have argued that AML in the
elderly
3 is more the result of multiple cytogenetic
or
4 mutational [technical difficulty] a
myelodysplasia
5 with a subsequent alternations.
6 An alternative hypothesis is that
our stem
7 cells get older, and getting a leukemia stem
cell
8 is a bad thing to do and the disease,
therefore
9 behaves differently in the elderly. And there is
10 some data for each of those arguments.
11 [Slide.]
12 MRK is one way of measuring
multidrug
13 resistance.
It may not be the best. It's a
14 simple, reproducible one. But whether one uses MRK
15 staining, or one uses actual drug efflux,
which is
16 cyclosporin inhibitable, in either way that
you
17 measure it, you will find that, as patients
age,
18 there is a higher incidence of multidrug
19 resistance.
20 In this--which included about 600
21 patients--if you were less than age 56,
about 33
22 percent would have a bright MRK
staining. If you
100
1 were over age 75, it's about twice that
number.
2 [Slide.]
3 Now, if you take all of these
possible
4 changes, and you look at the likelihood of
getting
5 a complete response, in univariate analysis--this
6 is a study that we did in the Southwest
Oncology
7 Group--in univariate analysis you can see
that
8 these make a big difference in outcome. AML, if
9 you have a secondary AML--that is, secondary
to
10 primary myelodysplasia; this isn't treatment
11 related, which is another kind of secondary
AML.
12 But this is secondary to primary
13 myelodysplasia--the CR rates are half as
much as if
14 you have a de novo disease, if you are CD34
15 positive in some, but not all studies, in
this one,
16 that seemed to give a lower incidence of
complete
17 response rates.
18 As I obviously mentioned, MRK
staining--if
19 you have a bright, you have half as much
chance of
20 getting a CR as if you're negative.
21 Unfavorable cytogenetics--again,
about
22 half as much a chance of getting CRs as if
you have
101
1 intermediate or favorable. And functional drug
2 efflux, like MRK staining, also predicts for
CR
3 rates.
4 [Slide.]
5 Now, this is in univariate
analysis. When
6 we looked in multivariate analysis, we found
three
7 factors-- and that's shown on the next slide
that
8 predicted for complete response rate: and
that is
9 whether you had secondary AML; if you had
10 unfavorable cytogenetics, and then if you
had MDR1
11 expression--either by functional drug efflux
or MRK
12 staining.
And each one of these were independently
13 significant in this multivariate analysis.
14 And if you had all three factors
present,
15 you had almost no chance of getting a CR;
that is,
16 if you had secondary AML, with unfavorable
17 cytogenetics and MDR1 expression, the chance
of CR
18
was 11 percent. But even if you're over age 65, if
19 you had none of the three factors, you'd
have an 81
20 percent chance of getting a complete
response--much
21 like you would expect in a younger patient
with
22 AML.
23 [Slide.]
24 Now, we're not the only ones that
have
25 seen this.
This has been seen by others. One
of
102
1 the largest studies of treatment of AML was
the MRC
2 AML 11 Study. It looked at three different
3 preparative regimens: a classic
daunomycin-Ara-C
4 and 6TG regimen, versus the daunomycin-Ara-C
VP-16
5 regimen, versus a mitoxantrone-Ara-C
regimen--and
6 then also randomized to giving G-CSF after
7 induction.
And they also had a randomization and
8 consolidation of two, versus
six, cycles.
9 The study itself isn't what
I"ll be
10 talking very much about, since none of these
11 factors had a major impact on outcome. But I
12 would--just to look at this
study-- which was
13 published in Blood in 2001--for the general
14 principles of treatment of AML. And, as I said,
15 this was a large study--it will show on the
next
16 slide--with over 1,300 patients. And they had ages
17 between 56 and up to above 90 years old.
18 [Slide.]
19 As I said, in the Southwest
Oncology
103
1 Group, we see about a 15 to 18 percent
incidence of
2 secondary AML, which is exactly what the MRC
study
3 saw.
And they had a 4 percent incidence of
4 treatment-related disease. Ours was a bit higher
5 in SWOG--about 6 percent. But these are very
6 typical numbers for AML in the older
individual.
7 [Slide.]
8 In their study, they had a 62
percent
9 complete response rate. They had a 7 percent death
10 in incomplete response. They had a relapse rate of
11 78 percent, and they had disease-free
survival at
12 five years of 15 percent.
13 These results are fairly typical
of most
14 studies of chemotherapy for patients over
age 55.
15 As I have tried to make the
point--and
16 will make it again--
17 [Slide.]
18 --oh, this shows you the outcome
by
19 treatment group. And, as you can see, there is
20 essentially no difference among the three
groups,
21 with a median survival that is between nine
and 11
22 months, and a five-year survival which is
down
104
1 around 15 percent on these studies.
2 Now, the point I've tried to make
3 repetitively is that AML in older patients
is a
4 very heterogeneous disease.
5 [Slide.]
6 And this shows you--the next slide
shows
7 you--what the MRC found. And they found
8 essentially the same thing that we had
previously
9 reported in SWOG--and other have reported,
as
10 well--and that is that cytogenetics, age,
whether
11 you have primary or secondary disease,
performance
12 status and--in their hands--also white count
at
13
diagnosis were powerful predictors.
And you can
14 see, with a large study of over 1,300
patients, how
15 powerful these are. Those p-values are 10-14 or
16 10-6, or 10-7--so that's a lot zeros, suggesting
17 that these are very powerful
predictors. So that
18 unfavorable cytogenetics--both for complete
19 response rate and overall survival--is a bad
fact;
20 having a high white count likewise gives you
a low
21 chance of CR or overall survival. As you age,
22 things get worse. If you have secondary disease
105
1
and if you have a poor performance status--the same
2 things that we had previously reported.
3 [Slide.]
4 So, in summary then, using
conventional
5 chemotherapy, if you take a large cohort of
6 patients that are over age 60--or 55,
depending on
7 the particular study--complete response
rates of 50
8 to 60 percent can be expected; a median
survival of
9
9 months can be also expected in this cohort of
10 patients; and perhaps 10 to 15 percent may
continue
11 to be alive after four to five years.
12 However--and this is the most
important
13 point--the patient and disease-related
factors very
14 greatly, among this population, and heavily
15 influence treatment outcome. In my mind,
16 particularly important in considering any
patient
17 group are age, performance status, primary
versus
18 secondary presentation, cytogenetics and the
MDR
19 status.
20 And these facts have been
relatively
21 unchanged over the last several decades
because our
22 therapies for AML in the elderly haven't
changed
106
1 very much over the last several
decades. And there
2 clearly is a need for new and effective
agents for
3 this patient population.
4 I'd be happy to answer any
questions about
5 this relatively brief and perhaps
superficial
6 presentation.
7 Thank you.
8 DR. MARTINO: Fred, Thank you. And ladies
9 and gentlemen, it is my understanding that
we are
10 able to handle some questions to Dr.
11 Appelbaum--from a technical
perspective. So, if
12 there are questions to him directly, this
would be
13 your opportunity.
14 Dr. Mortimer?
15 DR. MORTIMER: Yes,
Fred--this is Joan
16 Mortimer.
I wonder if you could just make a
17 comment on the role of growth factors. I presume
18 that since you didn't talk about it in the
MRC
19 trial that there is no advantage or
disadvantage to
20 the use of growth factors?
21 DR. APPELBAUM: In the MRC trial
there was
22 no advantage or disadvantage for the use of
growth
107
1 factors.
There have been--as you know, probably
2 better than anyone on the planet, Joan--I
think at
3 least a dozen studies of the use of growth
factors
4 after chemotherapy for AML in older
individuals.
5 And the vast majority of those
studies
6 show that the use of growth factors shorten
the
7
duration of neutropenia quite consistently by five
8 to seven days. They are more variable in whether
9 that shortening of neutropenia changes the
risk of
10 significant infection; and only, as I'm
aware, two
11 studies showed a change in survival or
complete
12 response rate. The majority of them showed no
13 effect on CR rates or survival, but did show
an
14 important shortening of the period of
15 pancytopenia--that's after induction. And then
16 after consolidation, the results are a
little more
17 consistent that you shorten neutropenia and
18 decrease the incidence of infections--again,
with
19 no change in survival.
20 DR. MARTINO: Dr. Brawley, you're
next.
21 DR. BRAWLEY: Yes, Otis Brawley
here.
22 Dr. Appelbaum, in the studies that
you
108
1 presented, and all the data that we've
reviewed, it
2 seems like the goal is always to get a
complete
3
remission, as if the complete remission is a
4 surrogate for patient benefit in terms of
increased
5 survival.
6 Has anyone ever tried any studies
that
7 look at the possibility of a drug that might
sort
8 of suppress a smoldering factor, where the
goal is
9 not complete remission but suppression of
the
10 leukemia, and perhaps try to determine of
that
11 actually increases survival--especially in
an
12 older, sickly population?
13 DR. APPELBAUM: That's an excellent
14 question, Dr. Brawley. And the lessons that had
15 been learned in acute leukemia in younger
patients,
16 where it's a much more proliferative
disease, have
17 sort of given the indication that you really
have
18 to get a complete response if patients are
going to
19 survive for any length of
time, because the disease
20 is so very proliferative.
21 Now, on the other hand, if you
take the
22 totally other tack of looking at
myelodysplasia as
109
1 being a sort of symbol for a less aggressive
2 disease--we now have data that a drug which
doesn't
3 get complete response rates, so that it
4 phibezacytadine by perhaps a slowing--some
people
5 can argue why phibezacytadine works. Some people
6 believe it's a differentiation factor. Other
7 people believe that it actually is working
as a
8 cytotoxic.
But, without getting complete
9 responses, it appears that it may prolong
survival.
10 Years ago, the way that many
11 patients--very old patients with AML--were
treated
12 was with much less aggressive therapy using
oral
13 6MP, or using daily or weekly VP16, trying
to keep
14 the cap on things. And I believe that there were
15
some suggestions--not proven in randomized trials
16 ever, but suggestions that there was
improvement in
17 survival.
18 You are, I thin, correct that it
is
19 possible that there will be drugs--maybe
many
20 drugs--that could cause differentiation,
slowing
21 down the proliferation, and be of some
benefit to
22 the patient without getting a complete
response.
23 But the lessons from the more
aggressive
24 disease is that generally those effects are
25 temporary and not as lasting as physicians
would
110
1 like.
2 I'm not sure if that answers the
question
3 adequately.
4 DR. BRAWLEY: Actually, it answers
it
5 wonderfully.
6
The reason I asked the question is I look
7 at our data on Zarnestra, and I wonder what
a lower
8 dose of Zarnestra for a longer period of
time,
9 without an endpoint for complete response
would do
10 in terms of patient benefit.
11 Do you think I'm going down the
wrong path
12 to say that that's something that perhaps we
ought
13 to be looking at with a farnesyl transferase
14 inhibitor?
15 DR. APPELBAUM: I think that it's
going to
16 be very patient-dependant. I believe that there
17 are--well, I know there are AMLs in older
patients
18
which look just like AMLs in younger patients,
19 which are explosive diseases, rapidly
111
1 proliferative, and I think it's less likely
that
2 we're going to be able to keep very good
control of
3 the disease without establishing a complete
4 response.
5 There are other patients who are
older who
6
[technical difficulty] very similar to
7 myelodysplasia and, in fact, getting rid
of--it's
8 much less proliferative and, you know, there
are
9 cases, in fact, of erythroleukemia in the
elderly
10 where simply giving red cells causes blasts
to
11 decrease, and the average survival in some
of those
12 patients, even without aggressive
chemotherapy, can
13 be six to nine months. So it's very
14 patient-specific.
15 DR. MARTINO: Dr. Cheson?
16 DR. CHESON: Hi, Fred. I have two
17 questions for you.
18 The first one follows on what Otis
was
19 sort of asking you. A lot of the data we have on
20 survival of AML are from olden times, when
we were
21 young.
And there's a drug that was pretty good at
22 controlling the disease: hydroxyurea. And now that
112
1 we have better antibiotics, better
supportive care,
2 better growth factors, etcetera, one wonders
how
3 you might be able to control this disease
using a
4 collection of those agents and do very
nicely
5 without--you know, less expensive, etcetera.
6 The second question relates to the
fact
7 that there have been a number of randomized
trials
8 in the past of chemotherapy versus
supportive care
9 in elderly patients with AML. And I was wondering
10 if you could comment on those?
11 DR. APPELBAUM: Well, the first
[technical
12 difficulty] interesting one, and I agree
that, with
13 better support care, with better
antibiotics,
14 possible with better hematopoietic growth
factors,
15 we may be able to eke out some increased
survival
16 of months or even longer without getting
complete
17 responses, [technical difficulty] responses,
or
18 just diminished blast percentage.
19 The quality of life of those
patients--I
20 think most of us who care for a lot of AML
21 patients--as you do--isn't great, but
[technical
22
difficulty] trips to hospital, a lot of transfusion
113
1 support--[technical difficulty]--
2 [Communication lost.]
3 DR. MARTINO: Can anyone help us at
this
4 point?
5 DR. CHESON: We lost you, Fred.
6 DR. MARTINO: Fred, if you can hear
us, we
7 cannot hear you. So, stand by, please.
8 [Pause.]
9 [Communication re-established.]
10 DR. APPELBAUM: As far as the
randomized
11 trials of support care, there have not--well,
maybe
12 you can inform me about any large, modern
13 randomized trial of supportive care versus
14 chemotherapy for AML in the elderly. The
ones that
15 I'm aware of were don--were quite small, and
done
16 decades ago, without current supportive care
17 measures.
18 I'm not aware of any that have
been done
19 in the last 20 years.
20 DR. CHESON: Right, but the results
of
21 chemotherapy versus supportive care in those
22 studies--some of which were modestly sized,
114
1
including the ones from Europe--suggest that a
2 chemo really probably wasn't of much
benefit. But,
3 again, that's before supportive care, growth
4 factors, etcetera.
5 DR. APPELBAUM: Yes, and those are a
couple
6 of decades old. And what they did show is that
7 those--if you take a population of patients,
you're
8 exactly right, the patients that got a CR
9
benefitted--or at least appeared to benefit--but
10 the overall population did not. But, again, those
11 are several decades old, they were done in
the
12 '70s, without current supportive care measures.
13 DR. MARTINO: Are there other
questions?
14 Yes?
15 MR. FLATAU: Dr. Appelbaum, I was
wondering
16 if you could comment on survival with
supportive
17 care in older patients with
de novo AML, versus AML
18 that arises from [Off mike.] [Inaudible.]
19 DR. APPELBAUM: If I understand the
20 question, you're asking about what is the survival
21 with supportive care alone in AML in the
older
22 patient compared to the younger
patient. Is that
115
1 the question?
2 MR. FLATAU: No, I'm wondering: de
novo AML
3 in older patients on supportive care, versus
AML
4 that arises from prior myelodysplasia.
5 DR. APPELBAUM: Oh.
6 Historically, the
median survival for de
7 novo AML on supportive care is about two
months
8 when patients are given supportive care
alone.
9 With older individuals given just supportive--now,
10 it depends on what you mean by
"supportive care."
11 It's been a long time since people just got
12 supportive care, with using 6MP as a single
agent,
13 or VP16 as a single agent. There were some studies
14 published in the '70s which suggested
survivals of
15 four to five months, using single-agent
16 chemotherapy in older patients with AML.
17 Most of those studies did not, at
that
18 time, break down the groups between AML that
was de
19 novo versus AML that was secondary to a
prior
20 myelodysplasia. I'm not aware of any data which
21 shows that with supportive care there is any
22 difference in the survival of de novo AML
versus
116
1 AML that arises from a prior myelodysplasia. All
2 the data that I'm aware of simply says that
AMLs
3 that evolve from a prior myelodysplasia are
much
4 less sensitive to chemotherapy, but not that
the
5 pace of disease with supportive care alone
differs.
6 Now, there may be data out there
that I'm
7 unaware of, certainly. And if anyone there knows
8 such data, I'd be interested to be educated
about
9
it.
10 DR. MARTINO: Dr. Temple.
11 DR. TEMPLE: It's worth mentioning
that the
12 trial here didn't show improved survival
compared
13 to a control, either. It showed that some
14 individuals seemed to do well.
15 A lot of data went by fast--but
tell me if
16 my impression here is correct. It seemed to me, as
17 you were going through the decreasing
response
18 rates and survival in people with
progressively
19 worse baseline status, you still were at a
point
20 where the response rates--complete response
21 rates--were in the 20s, and that was really
with
22 the very bad people. But for most of the people
117
1 with modest dysfunction, it was higher than
that.
2
And that's still higher than we've seen here.
3 And your estimates of five-year
survival
4 of 15 percent seem well greater than in the
5 population that was studied with Zarnestra now.
6 That's not fair, it's not the same
population--I
7 realize that.
8 But I wonder if you'd comment on
that.
9 And the thrust of my question is: it seems
to me
10
that this drug makes sense only for people in whom
11 you've unequivocally decided you don't want
to try
12 chemotherapy--or, conceivably, you do this
before
13 you try chemotherapy and see if you're lucky.
14 Can you comment generally on that?
15 DR. APPELBAUM: Yes. I apologize if I
16 presented the data too quickly.
17 DR. TEMPLE: That wasn't a
complaint. It
18 just was a lot.
19 DR. APPELBAUM: [Laughs.]
20 If you take a single risk factor,
like
21 cytogenetics, or like prior myelodysplasia,
or like
22 a very high white count, or poor performance
118
1 status--each one of those profoundly impact
the
2 overall CR rate. And you are correct that, when
3 you look at those univariately, you will get
CR
4 rates in patients over age 60, using
standard
5 daunomycin-Ara-C, that will be in the 25 to
30
6 percent range in general. If you start combining
7 multiple poor-risk factors, such as having
prior
8 myelodysplasia plus having unfavorable
cytogenetics
9 plus having multidrug resistance, then you
can
10 select some very bad populations where you
would
11 expect CR rates of even less than 10
percent, if
12 you combine multiple poor-risk factors.
13 But if you look at the general
population,
14 the CR rates are 50 percent in general; 25
to 30
15 percent with single bad prognosis; and then,
as I
16 stated, less than perhaps 10 to 11 percent,
if you
17 have multiple bad risk factors.
18 The CR rates are reflected in, then,
19 improved survival. Those that get CRs tend to
20 survive for nine to 12 months, whereas those
who do
21 not have very short survivals. And you are
22
correct: in the overall population--and the MRC was
119
1 the largest, that's a population of almost
1,400
2 patients--there was about a 15 percent
survival at
3 five years.
4 But, again, that is a population
that
5 includes patients who are in their late 50s,
60s
6 and 70s--all the way up to 90. It isn't just
7 patients over age 70.
8 DR. MARTINO: Are there other
questions?
9 [No response.]
10 If not, Fred, I'll take the last
question.
11 This is Slivana Martino.
12 What are you thinking of a patient
where
13 you're going to provide supportive
care? What
14 actual drugs, in the chemotherapy arena, do
you
15 think of in that setting? And why do you think of
16 those?
17 DR. APPELBAUM: If there's a
patient who
18 either, because of patient choice does not
want
19 aggressive chemotherapy, or I do not think
that
20 chemotherapy is warranted because of very
poor
21 performance status, then it depends on what
the
22 initial problem is. So that there can be patients
120
1 who have AML, who have a very rapidly rising
white
2 count, where the goal is to avoid
leukostasis, and
3 you have to use a chemotherapeutic agent to
try and
4 support them. And in those circumstances we use a
5 number of different drugs: hydroxyurea, as
Bruce
6 mentioned, is a classic one. Single-agent VP16
7 sometimes can work; single-agent topotecan
is
8 sometimes used. There, the goal is to just
9 decrease the white count.
10 Now, there are other patients
where the
11 white count isn't running up that quickly, and
the
12 real problem may be that they have no
granulocytes,
13 and there, trying to eke out a few
granulocytes
14 even with a G-CSF, or trying to use
something like
15 5-azacytadine, hoping that you're going to
get some
16 differentiation, may be useful.
17 In other cases, erythroleukemia,
for
18 example, you can get quite a bit of
satisfactory
19 responses simply by giving red cell
transfusions,
20 which diminish the red cell blast count in
the bone
21 marrow, and help the problem.
22 So the problems in those
individuals are
121
1 different.
Some, the threat to life is--or quality
2 of survival--is acytopenia, in other it is
the
3 rapidly growing white count.
4 In none of the cases, though, is the
5 outcome very satisfactory. They're very difficult
6 cases when patients have very poor
performance
7 status and are in their 80s or late 70s and
have
8
leukemia. There are not a lot of
good options for
9 such individuals.
10 DR. MARTINO: Thank you. And, seeing no
11 further questions, we appreciate your
presentation.
12
Thank you.
13 DR. APPELBAUM: Sure. Thank you.
14 Open Public Hearing
15 DR. MARTINO: The next portion of
the
16 meeting is the open public hearing. And we do have
17 one person who has signed up to address the
group.
18 But before I allow them to speak I need to
read
19 something to you.
20 Both the Food and Drug Administration
and
21 the public believe in a transparent process
for
22 information gathering and
decision-making. To
122
1 ensure such transparency at the open public
hearing
2 sessions of the advisory committee meeting,
FDA
3 believes that it is important to understand
the
4 context of an individual's presentation.
5 For this reason, FDA encourages
you, the
6 open public hearing speaker, at the
beginning of
7 your written or oral statement, to advise
the
8 committee of any financial relationship that
you
9 may have with the sponsor, its product and,
if
10 known, its direct competitors; for example,
this
11 financial information may include the
sponsor's
12 payment of your travel, lodging or other
expenses
13 in connection with your attendance at this
meeting.
14 Likewise, FDA encourages you, at
the
15 beginning of your statement, to advise the
16 committee if you do not have any such
financial
17 relationships.
18 If you choose not to address this
issue of
19 financial relationship at the beginning of
your
20 statement, it will not preclude you from
speaking.
21 MS. CLIFFORD: We have one
speaker. His
22 name is Roland McPherson.
23 If you can take the mike in the
back,
24 please.
25 MR. McPHERSON: Is it on now? Okay--thank
123
1 you.
2 My name is Roland McPherson. I live in a
3 suburb of Cleveland, Ohio. I have not been given
4 any financial remuneration--didn't even buy
stock
5 in Johnson & Johnson as I probably
should have
6 done.
But that's all right.
7 [Laughter.]
8 I was diagnosed with
myelodysplasia in
9 January of 2001. I had some symptoms maybe six
10 weeks before. I found myself out of breath on the
11 tennis court, which wasn't me. And so I was
12 treated for myelodysplasia, primarily with
Procrit
13 for some time. And then I needed a few
14 transfusions. And then, in 2002, the year after I
15 got this, I had a therapy of arsenic
trioxide and
16 thalidomide.
And that worked fine for about two
17 months, and then it stopped working.
18 I'm not quite sure about the years
on this
19 thing, but don't worry about that. But then I was
124
1 not completely pleased with the cancer
center I was
2 going to, so I looked around at different ones.
3 And finally I found a Dr. Mary Laughlin at
the
4 University Hospital in Cleveland that I
seemed to
5 hit it off with. First of all, I liked her. The
6 first time I met her she said, "Well,
first of all,
7 it's the Hippocratic oath: do no harm. And I agree
8 with that."
9 And by that time, the bone marrow
tests
10 showed that I really was in the low range of
AML.
11 And so she talked about all these induction
12 therapies, and the problem was, at my
age--as
13 you've heard and probably already knew
anyway--the
14 side effects can be very bad. And so she said,
15 "Well, let's just try maintenance for a
while."
16 And so I simply had transfusions
as
17 needed--which wasn't very frequently, but I
needed
18 some.
And then, of course, I had to take decirol
19 because of the iron buildup. And, at the same
20 time, I began looking at the NCI clinical
trials
21 on-line to find things, and I would bring
them in
22 to Dr. Laughlin, and she'd say, "Well,
this still
125
1 isn't right."
2 So--let's see, in early 2004--I've
3 forgotten the exact month--I hit this
Zarnestra
4 trial.
It looked good to me. And I
brought it in,
5 and she said--I don't remember what she
said, but I
6 would have said, "Bingo." Because she said, "This
7 looks like the thing to do." And she called Dr.
8 Karp at Johns Hopkins, and said, "Well,
this is one
9 to try."
10 So I did come here in April of
2004 and
11 had a meeting. And then maybe at the end of April
12 of 2004 we started the first Zarnestra. And we
13 went through three sessions of three weeks
each,
14 with a couple weeks off between. And at the end of
15 that period--in fact, before it was all
done--Dr.
16 Karp informed me that my blast count was
down to
17 zero.
18 And so this continued to be good
until
19 another six or eight months. Now, I had some side
20 effects, but they weren't severe. I would get a
21 little tired, and sometimes I had trouble
sleeping
22 and a few other things, but nothing severe
and it
126
1 didn't last very long.
2 And, let's see--oh, I might point
out that
3 I was very fortunate when I found this study
that
4 there were five places in this country doing
it,
5 and my brother lives in Montgomery County
Maryland,
6 so I had a place to stay. And I also learned later
7 that Dr. Karp in Johns Hopkins was the lead
in this
8 study, anyway, so that was fortunate.
9 So, about December of last year,
when I
10 came here for a day of testing, Dr. Karp
informed
11 me that my blasts were up now--I think the
number
12 was like 30 percent. So in January we started
13 another series, and at the end of that
series, my
14 blasts were down to--I may not have the
exact
15
number, but I think they were 6 percent.
But my
16 blood counts were still kind of low, hadn't
really
17 come up.
And then I had another series in March.
18 And at that time, the blasts were still
down. And
19 I'm not sure--I have some notes, but I
didn't want
20 to bring them up here.
21 But I think some time in the last
few
22 months, my blast count got down to zero, but
my
127
1 blood counts have not come up--especially
the
2 neutrophil is the big thing to worry about.
3 So here I am. I guess I didn't mention
4 that I'm 77 years old, so I certainly fit in
with
5 this group of elderly patients. And well, that's
6 sort of a short summary.
7 I don't know if anybody has any
questions.
8 I'm happy to answer anything to the best of
my
9 ability.
But that's sort of where I am..
10 DR. MARTINO: Thank you, Mr.
McPherson.
11 And we're all very happy to have you
here. Thank
12 you.
13 MR. McPHERSON: Thank you for the
14 opportunity.
15 DR. MARTINO: Okay, at this point,
shall we
16 do lunch next, and reserve the
questions? Because
17 it's 11:30.
18 DR. PAZDUR: I can give you the
option of
19 working through lunch.
20 DR. MARTINO: Well, realize that we
have
21 questions.
We're going to have discussion.
And
22 I'm not sure that--how many of you are
interested
128
1 in lunch?
2 [Laughter.]
3 Because we're not likely to finish
in an
4 hour or two.
So that's my concern here, is that
5 you may not be interested in lunch right
now, but
6
you will be an hour from now, and then people are
7 going to get antsy. And I'd rather not have people
8 in a semi-distressed circumstance--okay?--as
9 they're deliberating on this drug.
10 DR. PAZDUR: We can take lunch.
11 DR. MARTINO: So I think I'd rather
do
12 lunch, and maybe we can come back a bit
earlier?
13 Is that reasonable? Okay?
14 Now, before I give up the mike, where
does
15 one have lunch in this fine, fine
institution and
16 building?
17 DR. PAZDUR: You'd have to ask
Johanna
18 about that.
19 DR. MARTINO: The place across the street.
20 I'm not sure why she doesn't want to say it
21 publicly.
22 DR. PAZDUR: She doesn't know the
name of
129
1 it.
2 DR. MARTINO: But, at any rate--so,
does
3 that mean that--should I give everyone the
entire
4 time, is what I'm asking here, Johanna.
5 Okay. So shall we plan on coming back at
6 maybe 12:30-ish? Is that okay?
7 All right, let's aim for return at
12:30,
8 and starting shortly thereafter.
9 [Off the record.]
10 DR. MARTINO: Ladies and gentlemen,
I'd
11 like you all to take your seats and we will
resume
12 our proceedings.
13 Ladies and gentleman--and Dr. Temple
14 included.
15 [Laughter.]
16 Ha-ha.
17 Questions from the Committee
18 DR. MARTINO: The next portion of our
19 meeting are questions to either the
pharmaceutical
20 company or to the FDA.
21 And I think Dr. Perry already is
champing
22 at the bit.
23 DR. PERRY: Yes, I have two
questions for
24 the sponsor, if I may.
25 If you look at Slide 53 in your
130
1
presentation, the complete remission rate by site,
2 I'm struck by the fact that Cornell, which
put on
3 16 percent of the patients, and Georgia,
which put
4 on two percent of the patients, had absolutely
no
5 complete responders.
6 So I want to know whether there's
7 something wrong in some portions of New
8 York--because if you go to Rochester, you do
9 better?
Or whether there's something else
10 involved?
Eighteen percent of the patients had no
11 responses at those two sites. That seems to me to
12 be more than a statistical aberrancy.
13 DR. DeLAP: We did look for any
reason that
14 there might be a different result at
different
15 sites.
We did not find a reason to account for
16 that, in terms of demographics, risk
factors.
17 We can discuss further the
statistics of
18 it, if you'd like. We feel that the overall result
19 of the study--the overall response rate--15
percent
131
1 for the study, the planned analysis as
described in
2 the study, I think that's our best estimate,
and
3 that includes zero from some sites and like
25 from
4 other sites.
But the net is 15.
5 DR. PERRY: Were they putting--you
don't
6 know whether they were putting in sicker
patients
7 in some other ways?
8 Basically, the question I'm asking: has
9 somebody looked at that?
10 DR. DeLAP: Yes, and we can show
you some
11 data.
12 DR. THIBAULT: Alain Thibault,
clinical.
13 We have--while we're waiting for
our slide
14 to come up--yes, bring the slide up. No, that is
15 not it.
16 While we're waiting for the slide
to come
17 up--we have looked at all sites. There is 15
18 percent CR rate total. The rate of response varies
19 from zero to 25.
20 At Cornell, this is a site that
accrued 22
21 patients.
As far as we can tell, in terms of
22 median age, in terms of number of risk
factors, in
132
1 terms of incidence of prior MDS, these
patients
2 were all similar.
3 We've looked at baseline factors,
as well,
4 such as blast count, and we have found
nothing
5 there, either.
6 And so what we are left to
postulate: we
7 noticed the difference, but we're left to
postulate
8 is that we are looking what you would call a
9 statistical aberration, or an effect of
subgroup
10 analysis.
11 DR. PERRY: Okay.
12 DR. THIBAULT: What I should
mention--not
13 to belabor it too much--is at this site
there were
14 patients who achieved partial remissions and
15
hematological improvements that were maintained for
16 several months. And therefore that means these
17 patients had a reduction in tumor burden,
but not
18 quite the recovery of peripheral counts you
would
19 expect.
20 DR. PERRY: Okay. I have a second
21 question--and maybe you're the person to
ask.
22 Several times during the
presentation the
133
1 point has been made that these patients have
to
2 take a pill; you know, that this is a great
drug,
3 they only have to take a pill. By my count, they
4 take 12 pills a day; six 100 mg tables in
the
5 morning and six in the evening.
6 So my question is: how did you
ensure that
7 you actually got compliance in the people
who had
8 to take a lot of pills?
9 DR. DeLAP: Yes--I'll ask Dr.
Zukiwski,
10 actually, to address that.
11 DR. ZUKIWSKI: As outlined in the
FDA
12
presentation, detailed compliance diaries were not
13 kept at all sites. However, we're confident that
14 the patients did take their medication as
15 prescribed.
16 There was dispensing logs in the
pharmacy.
17 The drug was dispensed in blister packs of
seven.
18 It was dispensed for the entire 21-day
period.
19 The patients, we believe, were
highly
20 motivated coming into the study. They were
21 assessed weekly by the research personnel or
the
22 physicians, and we would have anticipated
that if
134
1 the patients were having a problem taking
their
2 drugs it would have been picked up at that
time and
3 would have been reported.
4 In addition, one of the other
studies that
5 Dr. DeLap has outlined--INT-17--has a very
similar
6 safety profile to what was observed in
CTEP-20.
7 And in the INT-17 study we kept detailed
patient
8 diaries, and the safety profile is the
same. So
9 we're relatively confident that the patients
did
10 take the drug.
11 But I will ask Dr. Karp to give
her
12 experience, as the PI, regarding the
compliance
13 issue.
14 DR. KARP: Thank you very
much. I'm Judy
15 Karp.
I'm the principle investigator of this
16 study.
17 AS Dr. Zukiwski said, this was a
very
18 motivated group of patients. And my patients would
19 never lie to me.
20 [Laughter.]
21 DR. PERRY: Right.
22 DR. KARP: [Laughs.] If you were one of my
135
1 patients, you wouldn't lie to me.
2 [Laughter.]
3 I wouldn't want that, anyway.
4 I think we have some clinical
evidence. I
5 mean, people's counts went down--not
spontaneously.
6 The other piece of evidence we
have is
7 that we looked at the ability of Zarnestra
to
8 inhibit farnesyl transferase activity at two
9 sources.
One was in the bone marrow, and the other
10 was in buccal mucosa. And the buccal mucosal
11 studies, 92 percent of the patients tested,
we were
12 able to demonstrate enzyme inhibition. So I think
13 that that's evidence that the patients
actually
14 took the drugs.
15 DR. PERRY: Well, I'm sure they
took some
16 of the drugs. My question was whether they took
17 all of the drug. That, I think, is a question
18 that's a little more difficult to answer.
19 Thank you.
20 DR. MARTINO: Dr. Mortimer?
21 DR. MORTIMER: I have two questions
for the
22 sponsor.
One goes back to the question about
136
1 growth factors, because I think the
literature of
2 growth factors in leukemia in the elderly is
3 somewhat confounded. There are positive trials.
4 And I wonder what the use of growth
5 factors was-- whether it was dictated on the
6 trial--unless I missed it in the information
we
7 were give.
8 DR. DeLAP: There was very little
use of
9 growth factor in the study. And I'll ask Dr.
10 Thibault for the exact number.
11 DR. THIBAULT: Yes, we've actually
12 collected information on specific growth
factors.
13 Twelve patients took growth factors,
overall, in
14 this study.
And it induced an improvement in the
15 ANC count in five of them. So the vast
16 majority--120, roughly, patients--did not
use any
17 growth factors.
18 DR. MORTIMER: The second question
I have
19 actually relates to the patient
population. And I
20 think you're to be congratulated for investing
in a
21 trial like this, in the geriatric
population.
22 But the patients were not ideally
to be
137
1 candidates for chemotherapy, and yet 10
2 subsequently got chemotherapy--six of whom
actually
3 responded.
And I think that kind of confounds the
4 results a bit, that there were six CRs --and
none
5
of whom had responded to tipifarnib before getting
6 salvage therapy.
7 DR. DeLAP: I'll ask Dr. Thibault
to
8 comment on the use of subsequent
chemotherapy.
9 DR. THIBAULT: Yes, we've collected
very
10 detailed information on the subsequent use
of
11 chemotherapy.
12 May I have the slide up, please.
13 [Slide.]
14 Of the 136 patients who were
treated, the
15 majority went on to receive palliative
care. This
16 is the subsequent treatment that the 136
patients
17 received.
Seven were re-treated with tipifarnib,
18 and one of them achieved a complete
remission.
19 Seventeen patients received single-agent
20 chemotherapy, and none of them achieved any
21 remission to that type of treatment. Twelve
22 received combination chemotherapy--Ara-C
with
138
1 anthra-cycline at the usual dose.
2 The patients who received
anthra-cycline
3 in combination chemotherapy included two who
had
4 responded to tipifarnib first. Out of the 12, only
5 two were 75 or older, and they tended to
have a
6 lower blast count at the time of
re-treatment in
7 the range of about 30 percent, based on
blast count
8 when they entered tipifarnib was about 45
percent.
9 So this is re-treatment of a very
small
10
minority of patients. And we think it reflects on
11 the fact that the vast majority of the
patients
12 were felt by the investigators not to be
adequate
13 candidates for intensive chemo.
14 DR. MARTINO: Dr. Levine.
15 DR. LEVINE: I have several
questions.
16 The first relates to
myelodysplasia. You
17 gave us information on pathologic re-review,
or
18 central review of the CR. But who diagnosed the
19 myelodysplasia, and who confirmed the
20 myelodysplasia? So that was my first question.
21 DR. DeLAP: The antecedent myelodysplasia
22 at the time of entry into the study, that
was not
139
1 something that was part of the pathologic
review,
2 that was part of the patient's history. So, again,
3 that's not part of the central review.
4 DR. LEVINE: Do we have any idea of
the
5 length, the duration of myelodysplasia prior
to
6 study?
The number of transfusions that were given
7 prior to study--you know, during
myelodysplasia?
8 And then afterward? Is there any information
9 there?
10 DR. DeLAP: We do not have a
presentation
11 on that.
Perhaps Dr. Karp can give us some insight
12 into the range of prior dysplasia that was
present
13 in these patients?
14 DR. KARP: Well, the minimum
duration, per
15 definition: one month of documented
antecedent
16 hematologic disorder. We had patients whose
17 myelodysplasia--as per a patient who spoke
to us
18 today--had a duration of myelodysplasia
prior to
19 transformation of up to three years.
20 I can't tell you what the median
is. But
21 it was a broad range.
22 DR. LEVINE: Another question--I'm
just
140
1 trying to get at the poor prognosis. It was the
2 same as Dr. Mortimer: were these really
3 poor-prognosis cases?
4 So one of your definitions of poor
5 prognosis was organ dysfunction. So could you
6 define "organ dysfunction?" And it had to be organ
7 dysfunction within the confines of what was
allowed
8 onto protocol--right?
9 So then my next question, beyond
that is:
10 if you exclude organ dysfunction, if you
just look
11 at factors of age and cytogenetics, and
12 myelodysplasia, what percentage of the
patients had
13 one of those three, forgetting the organ
14 dysfunction?
15 DR. DeLAP: I'll ask Dr. Thibault
to start
16
with this.
17 DR. THIBAULT: The proportion of
patients
18 who had therefore three risk factors, if we
don't
19 count for the proportion [sic], was
high. We would
20 have 82 percent of them having MDS. We will have
21 49 percent of them having prior
myelodysplasia. We
22 will have 41 percent of them having both.
23 And then 55 percent of the
patients were
24
75 and above.
25 DR. LEVINE: So, in other words, 41
percent
141
1 had both myelodysplasia and the cytogenetic
2
abnormalities.
3 DR. THIBAULT: Yes. The majority--as you
4 recall, the inclusion criteria were for 65.
5 DR. LEVINE: Okay. Another question: going
6 back a little bit to the compliance, one of
the
7 major non-hematologic toxicities was nausea
and
8 vomiting.
Do you think that might have influenced
9 the number of pills that these people
took? Or how
10
do you deal with that? You're
going to need to
11 know whether they really are taking this
pill or
12 not.
Or 600 BID, or what they're doing.
13 DR. DeLAP: Yes.
Nausea and vomiting was
14 a relatively minor problem for most
patients, but
15 we'll ask Dr. DePorre to go through the
data.
16 DR. DePORRE: DePorre--clinical.
17 Indeed, patients had nausea and
vomiting.
18 However, in the vast majority of these
patients,
19 that was Grade 1 and 2. It resolved without any
142
1 intervention, and it was of short duration.
2 DR. LEVINE: So, in other
words--that was
3 another question: so they are not on
antiemetics
4 for all this time.
5 DR. DePORRE: Correct.
6 DR. LEVINE: And with continued
use, the
7 nausea goes away? Is that what happens?
8 DR. DePORRE: Yes. Correct.
9 DR. LEVINE: Okay.
10 One more: the
confirmed complete
11 remission--so that was one month after the
initial
12 CR.
Those confirmations were on drug?
On
13 continued drug?
14 DR. DePORRE: Yes, patients who were taking
15 drug and had a confirmed complete remission
at one
16 month would obviously continue drug.
17 DR. LEVINE: Okay.
18 DR. MARTINO: I'd like to ask the
next
19 question.
20 I am aware that there is a Phase
III trial
21 ongoing that deals with this agent in this
patient
22 population.
At some point, someone needs to
143
1 explain that trial and help me to understand
where
2 it is, where it isn't, and what the
objectives are.
3 DR. DeLAP: Are you referring to the
4 AML-301 trial? Yes.
5 Dr. Zukiwski?
6 DR. ZUKIWSKI: Yes. Could we have a slide
7 up, please?
8 [Slide.]
9 Well, let me just take you through
some of
10 the background on the AML-301 trial.
11 This is a randomized trial of
tipifarnib
12 versus best supportive care. It is being conducted
13 outside of the United States. The trial is active
14 and ongoing.
15 There is a total of 90 sites that
will be
16 activated; approximately 67 sites have been
opened
17 to date, and the accrual is continuing to
ramp up.
18 The overview is here. The eligibility:
19 patients greater than or equal to 70 years
of age.
20 This is a consensus that was reached with
the
21 ex-U.S. investigators. There was a lot of
22 discussion on what the actual patient
population
144
1
could be. And this was where the
consensus
2 agreement was reached.
3 These are patients with initial
AML. The
4 patients--we'll be collecting karyotype
5 categories-- favorable, unfavorable,
intermediate.
6 We'll be collecting data on prior
myelodysplastic
7 syndrome.
There will be a stratification built in.
8 There is no interim analysis. There is an IDMC to
9 give us guidance on whether or
not to continue with
10 the trial.
11 We anticipate that this trial will
have
12 data, so we will be able to submit to the
Food and
13
Drug Administration sometime in
2007.
14 DR. PERRY: Excuse me. I have a hard time
15 believing that's the only eligibility
16 criteria--just leukemia and greater than
70? No
17 end-organ function tests? Nothing else?
18 DR. ZUKIWSKI: I can go through the
detail,
19 but this is just the overview. If you'd like--
20 DR. PERRY: Well, it doesn't give
us a very
21
good idea of how severely ill these patients are.
22 They just have leukemia.
23 DR. ZUKIWSKI: Can we go on to the
next
24 slide, please? Next slide.
25 [Slide.]
145
1 Okay, I think it's on Slide 9.
2 [Slide.]
3 Basically, the ECOG is 0 to 2,
4
newly-diagnosed, not willing to receive induction
5 chemotherapy. So this is a conversation that will
6 take place between the treating physician
and the
7 patients--that determination will be made on
8 whether or not the perceived risk is worth
the
9 possible benefit in those patient
populations.
10 And, obviously, the
thoroughly-described
11 study, where we will be--the informed
consent does
12 describe the "best supportive
care" elements. And
13 so it does list other options for patients.
14 DR. TEMPLE: About that, can you
say
15 something about what the consent form is
going to
16 say in that study?
17 DR. ZUKIWSKI: Dr. Temple, I'd have
to get
18 the exact informed consent. I'll get back to you
19 on that one.
I don't have it off the top of my
146
1 head.
2 But it's important to note that
this study
3 is being done to determine the magnitude of
the
4 survival advantage which we believe will
take
5 place.
So, 306 patients--we'll have that data
6 sometime in 2007.
7 DR. MARTINO: How many patients are
in the
8 trial at this point, Doctor?
9 DR. ZUKIWSKI: Currently, as the
accrual is
10 ramping up, there is approximately 88 or 90
11 patients.
I got an update in the last day or so.
12 DR. MARTINO: Okay.
13 DR. ZUKIWSKI: So I'm not certain
of the
14 exact figure.
15 DR. MARTINO: And how much of this
is
16 happening in this country versus elsewhere?
17 DR. ZUKIWSKI: This is entirely
ex-U.S.
18 DR. MARTINO: It's entirely--
19 DR. ZUKIWSKI: It's not being
conducted
20 within the United States.
21 DR. MARTINO: Thank you.
22 Dr. Cheson?
23 DR. CHESON: Yes, first a comment:
I would
24 like to thank you for changing the
definition of
25
elderly from 60 to 65.
147
1 [Laughter.]
2 Then I have two questions left,
since
3 everyone's been asking all the other ones I
wrote
4 down.
5 One of the major points that would
be
6 attractive for this drug is the fact that it
often
7 doesn't require hospitalization. Your figure is
8 that 40 percent of patients did not require
9 hospitalization.
10 How many of that 40 percent were
11 responders?
In other words, is it just the
12 patients who got too sick and they decide
it's not
13 worth putting them in the hospital? Or are these
14 patients who were actually benefitting from
the
15 therapy, getting their complete--and perhaps
16 partial--responses, and still able to stay
out of
17 the hospital?
18 DR. DeLAP: Part of the benefit,
actually,
19 of the complete response is improvement in
148
1 disease-related morbidity and the ability to
stay
2 out of the hospital. So we do see that patients
3 who have a response do better, in terms of--
4 DR. CHESON: Do you have numbers?
5 DR. DeLAP: yes. Dr. Thibault will show
6 you.
7 DR. THIBAULT: Yes, may have the
slide up,
8
please?
9 [Slide.]
10 This is a slide that focuses on
patients
11 who achieve a CR. On the left-hand side are
12 variables of interest: hospitalization,
transfusion
13
of platelets and red blood cells; use of growth
14 factors.
15 And on the middle column you have
the
16 hospitalization rate and rate of other
17 interventions--in those CR patients prior to
18 reaching their CR. The time to CR is 43 days.
19 Then in the extreme right column
you have
20 the use of health care resources, once the
patient
21 was in CR.
22 So what you can see is that once
the
149
1 patient achieved a CR, only six patients
required
2 further admission, some of them for
treatment of
3 ancillary disorders. But the majority of
4 patients--15 of the 20--were hospitalized.
5 The duration of these
hospitalizations is
6 12 days in the CR patient.
7 DR. CHESON: Okay--so
three-quarters of the
8 patients who did respond, did require
9 hospitalization initially. And then are those the
10 same ones--those further six--
11 DR. THIBAULT: Yes.
12 DR. CHESON: I mean, those six are
the--
13 DR. THIBAULT: Are these six a
subset of--
14 DR. CHESON: Or does it make it all
20
15 required hospitalization at one time?
16 DR. THIBAULT: I will have to
verify this.
17 My recollection is these six belong to the
15.
18 DR. CHESON: Okay.
19 My next question: when we devise a
20 response criteria, we recognize, as most
people
21 do--not an original idea--that complete
remission
22 is the most important response category. And we
150
1 recognize, and put in the paper that partial
2 response was primarily useful to identify
the
3 activity of a new agent, not necessarily
saying
4 anything about the efficacy of the new
agent.
5 You have a bunch of patients who
have
6 partial responses, and so-called
"hematologic
7
improvement" which was in the MDS response criteria
8 is not really a response criteria in AML.
9 Do you have evidence of clinical
benefit
10 in patients who got partial responses? Stable
11 disease, hematologic improvement--what is
the
12 evidence that these patients really did
experience
13 benefit, other than the complete responders?
14 It would be nice if, you know--even
if you
15 didn't get a complete response, if you could
16 demonstrate decreased transfusion
requirements,
17 decreased infections, decreased
hospitalizations
18 compared to previously--etcetera, etcetera,
19 etcetera, etcetera.
20 DR. DeLAP: We do have some
analysis that
21 suggests some benefit in the partial
response, the
22 hematologic improvement subset. It's not as well
151
1 established, obviously, as complete
response.
2 I'll ask Dr. Zukiwski to--
3 DR. CHESON: And when we did come
up with a
4 hematologic improvement, we also had
"major" and
5 "minor"--which we'll probably get
rid of at some
6 point.
7 So, how important was the hematologic
8 improvement?
In other words, if the hemoglobin
9 went from 6 grams to 7 grams, it's probably
not a
10 big deal.
If it went from 6 grams to 10 grams,
11 then maybe it's something worth talking
about.
12 DR. ZUKIWSKI: Okay.
13 Unfortunately, Dr. Cheson, we did
not cut
14 the PR or the hematological improvement as
we did
15 with the CRs to show you hospitalizations
and other
16 utilization of resources, etcetera. What we did do
17 is we did do an exploratory analysis,
looking at
18 survival, for our internal purposes, and we
found
19 it to be very encouraging, looking at the CR
20 patients, the PR patients and those
non-responders.
21 So, for our internal planning
purposes for
22 the 301 study, it gives us a lot of
encouragement
152
1 that there will be benefit even derived from
the
2 patients who have achieved a PR.
3 DR. CHESON: It would
be--hopefully--to the
4 advantage of the drug if you did have those
5 numbers.
Because if you could demonstrate
6 improvement in these parameters in patients
who
7 didn't get a CR, it makes the drug a lot
more
8 interesting.
9 DR. ZUKIWSKI: Right. And this slide just
10 flashed up.
11 [Slide.]
12 This slide has to be taken for
exactly
13 what it is: this is an exploratory analysis,
what
14 would be called "classic responder,
non-responder."
15 No firm conclusions can be drawn. The only purpose
16 we had done this is to look for internal
data. And
17 there is a trend that gives us some
indication that
18 PR patients may have an improved
survival--but no
19 conclusions can be drawn from this data at
all.
20 DR. MARTINO: Dr. George?
21 DR. GEORGE: I have a number
questions in
22 sort of different areas. One is a follow-up on the
153
1 details on the confirmatory trial--the
AML-301, I
2 think it's called.
3 What kind of difference in
survival are
4 you thinking--what are you looking for there? What
5 is the study powered to pick up?
6 DR. DeLAP: The study is powered to
pick up
7 a 50 percent increase, in terms of the
hazard
8 ratio, from approximately three months to
9 approximately four-and-a-half months.
10 DR. GEORGE: We can come back to
some of
11 that in discussion, but some of that has to
do with
12 what kind of difference you might have
expected
13 based on the potential benefits--satisfiable
based
14 on the response rates. But that's--we'll come back
15 to that in discussion, I think.
16 Is there any quality of life study
17 included in that confirmatory trial? The
18 randomized trial?
19 DR. DeLAP: We're not doing the
formal
20 quality of life-type questionnaires that are
21 sometimes used in that trial.
22 DR. GEORGE: Okay. Another area had a
154
1 little question about--maybe for the sponsor
or the
2 FDA.
It has to do with the discrepancy between the
3 CR assessment--rate assessment.
4 If I'm understanding it right, out
of
5 those 20 cases, there were three that the FDA
6 didn't count because they weren't confirmed
by the
7 usual definition. Now, those are for different
8 reasons: one died early. And maybe you could
9 address that.
10
DR. RYAN: one patient
died early, before
11 confirmation. Another patient progressed, died on
12 confirmation. The third one does not have
13 sufficient bone marrow data to support a
confirmed
14 CR.
15 DR. GEORGE: But that third one was
a
16 long-term survivor.
17 DR. RYAN: Yes. Then there were two
18 patients who had complete hematological
response,
19 who do not have bone marrow slide to support
CR.
20 DR. GEORGE: Okay. I'm just--
21 DR. DeLAP: Yes, there were a
couple of
22 cases where the slides were lost in transit,
155
1 basically, and could not be reviewed by the
2 independent reviewer. And for that reason they
3 were not included in the 15 cases that could
be
4 confirmed and verified.
5 But it's important to note that
the slides
6 for those patients that were lost in transit
had
7 been reviewed at the original
institution. So we
8 do have readings.
9 DR. GEORGE: Okay. The last point is a
10 more generic question, I guess, and I know
you
11 won't be able to answer this directly.
12 But one of the things that most
impressive
13 from all these presentations for me is that
this
14 indication--this particular medical
condition is a
15 very serious one--but that it's highly
16 heterogeneous. And that's one of the issues,
of
17 course, that's going to come up in
discussion of
18 any single-arm trial.
19 But, what Fred Appelbaum presented,
if you
20 looked at the cases with different risk
factors,
21 from the worst to the best was really
dramatic. I
22 mean, the ones with three risk factors had
an
156
1 estimated response rate of 11 percent. Remarkably,
2 that's the same thing that the FDA came up
with on
3 their study.
And the best cases had an 81 percent
4 response rate.
5 Now, these are patients that, of
course,
6 might have been different in other
ways. They did
7 receive chemotherapy, for example. And that
8 relates to my question.
9 Is there any way--do you have any
feel, or
10 any way to know this--what percent of the
patients
11 on the CTEP-20 study actually might have
received
12 chemotherapy, as opposed to the eligibility
13 criteria was sort of "not able to
willing"
14 basically, to receive chemotherapy, which--I
can
15 understand what you're getting at there, but
there
16
is a difference.
17 If you're trying to relate, you
know, what
18 you observed to what might be obtained with
19 chemotherapy, it's kind of hard for me to
figure
20 out how many of those patients actually
might have
21 received chemotherapy, and might have been
on the
22 study, as opposed to those that just really
were
157
1 medically unfit for chemotherapy.
2 DR. DeLAP: Yes, I'll ask Dr. Stone
to
3 comment on that. The other side of the equation of
4 risk-benefit, of course, is that in addition
to a
5 low response rate we're looking, with
combination
6 chemotherapy, with a high treatment-related
and
7 other mortality in the first 30 days, which
is 25
8 percent or more.
9 So that's--again, in terms of the
10 risk-benefit, I think it's rather clear why
11 patients would not get the combination
12 chemotherapy.
13 But I'll ask Dr. Stone to comment
further.
14 DR. STONE: Well, Dr. George, I
assume I
15 understand what you're saying: it would be
nice to
16 know exactly how these people would have
done had
17
they had chemotherapy run into their veins, or had
18 they had supportive care, perhaps.
19 All we can do is try to extract
that type
20 of data from the cooperative group
trials--which
21
may not be analogous because those patients were
22 getting chemotherapy. But even if you do that, you
158
1 can see from the data that Fred showed, that
I
2 showed, that if you just look at prior
3 myelodysplasia, which characterized a great
deal of
4 these patients on CTEP-20, the remission
rate would
5 be about 20 percent. And the mortality rate is
6 going to be at least as high as the average
7 mortality rate that was in those trials,
which was
8 about 25 percent.
9 Probably the decrease in the
remission
10 rate in people with those poor prognostic
factors
11 is, in part, due to resistance manifested by
a lack
12 of going into remission and, secondly, of
higher
13 mortality rate. Hard to tease out, but I think, in
14 looking at this as a leukemia doctor, I
would say
15 that the CR rate had chemotherapy been run
into
16 their veins, would probably be in the
magnitude we
17 discussed in the slides: in the 10 percent
range.
18 Probably most of these people were
really
19 like the triple-losers that Dr. Appelbaum
20 mentioned.
If they had had MDR, they might have
21 been positive, and that would be about 11
percent.
22 So--I mean, that's the best I can
give you
159
1 on
that. The mortality would be very high,
2 reflecting the, probably, 90 percent who
didn't go
3 into remission, probably half would have
died.
4 DR. MARTINO: Dr. Brawley?
5
DR. BRAWLEY: Thank you.
6 Can you bring back up the slide of
7 hospitalizations that Dr. Thibault was
talking
8 about?
I need to try to understand that a little
9 better.
10 DR. DeLAP: This is the slide about
11 hospitalizations for the patients that had
12 remissions?
13 DR. BRAWLEY: Right. Right.
14 The question is--
15 [Slide.]
16 --yes, there we go.
17 The period of time prior to CR,
versus the
18 period of time during remission--you know,
if one
19
were to take the margin and say that "during
20 remission" was seven days, and
"prior to CR" was
21 six months--this slide doesn't make any
sense.
22 So what I'm wondering is: do you
have a
160
1 similar slide that compares hospitalizations
after
2 treatment for individuals who did not have a
3 remission, versus individuals who did have a
4 remission?
5 DR. DeLAP: One thing I'd like to
comment,
6 in line with your question.
7 DR. BRAWLEY: Sure.
8 DR. DeLAP: And then we'll show
that slide.
9 The "prior to CR," these
patients entered
10 CR after one to two cycles of
treatment. So that
11 interval is limited. And then the "during
12
remission," obviously the median duration of
13 remission was several months, so that
interval is
14 actually quite long. So that's another thing you
15 have to think about when you look at these
16
percentages.
17 DR. BRAWLEY: That helps me to
understand
18 this slide and actually think of it as valid
data.
19 Thank you.
20 Do you have the other data, as
well?
21 DR. DeLAP: I'll ask Dr. Thibault.
22 DR. THIBAULT: I'll show you very
similar
161
1 data.
Next slide up, please.
2 [Slide.]
3 This is the same slide with an
extra
4 column added. On the extreme right you are seeing
5 the hospitalization rate for the patients
who never
6 achieved a CR, and you see that the
hospitalization
7 rate if you don't achieve a CR is similar to
the
8 hospitalization rate of the CR patients
prior to
9 their entering a CR. Then after they enter the CR,
10 then they seem to derive benefit in terms of
reduce
11 hospitalization.
12 DR. BRAWLEY: Show this slide first
next
13 time.
14 [Laughter.]
15 DR. THIBAULT: We're hoping we
won't have
16 to come back again.
17 [Laughter.]
18 DR. MARTINO: Dr. Brawley, anything
else
19 you'd like to contribute at this time?
20 [No response.]
21 Are there other questions?
22 Dr. Bukowski.
23 DR. BUKOWSKI: Can you help me
understand
24 the rationale that was used to select the
dose that
25 has been used in this particular study: the
600 mg
162
1 BID?
2 You mentioned that in the Phase I
trial,
3 there was some hint of activity, and the
slide
4 mentioned that you achieved the IC 50 in
vivo with
5 that dose.
Can you just clarify? I mean,
were
6 there responses at lower doses in the Phase
I
7 trial?
And have you had any experience with other
8 dosing levels?
9 DR. THIBAULT: Well, we have our
most
10
extensive experience at 600 dose.
11 If you want me to take you
through: in the
12 Phase I trial, the doses were 100, 300, 600,
13 900--1,200 was not tolerable. There was one
14
remission at 300--sorry, one remission at 100, one
15 remission at 600. This is a very limited data set:
16 approximately six patients per dose level.
17 Most of our experience is, of course,
at
18 600.
The reason we chose this dose was because at
19 600 we are certain that the farnesyl
transferase
163
1 enzyme is fully inhibited. We are also covering
2 the range of IC 50s--of the different cell
lines
3 that have been tested. There's about a 10-fold
4 difference in IC 50 when you look at the
different
5
blinds. And so we're covering
that at 600 for a
6 longer duration of the treatment period than
we
7 would if we were at 300.
8 And then, finally, when we looked
at the
9
toxicity of this, we knew that we would go into
10 Phase II with a dose that would require
about a
11 third of the patients to be
dose-reduced--which is
12 kind of the definition of the MTD. And so the data
13 supports that approach.
14 We do not know whether treating
these
15 patients, right off the bat, at 300 would
yield the
16 same results. But we do know that we treat them at
17 600, and then we adjust the dose, we get the
18 results that we're showing today.
19 DR. BUKOWSKI: So the 600 is the
MTD from
20 the Phase I trial?
21 DR. THIBAULT: Yes, in the definition
we've
22 used, "MTD" is the dose level
below the one that
164
1 causes more than a third of the patients to
be in
2 TLT.
So this is the recommended Phase II dose.
3 DR. MARTINO: Dr. Porter?
4 DR. PORTER: Just a clarification
on your
5 Slide 67.
You have "AEs leading to deaths," and
6 then "Drug-related." I realize this is the
7 investigator's opinion, but since you only
have one
8 drug, and it's an AE from that drug, you
really
9 have 10 drug-related deaths--to read that
table
10 correctly.
11 Is that a correct interpretation?
12 DR. DeLAP: Can we have that slide
up,
13 please?
14 [Slide.]
15 DR. DePORRE: Let me first clarify
one
16 point: it is that the "1" and the
"9" should not be
17 added.
The "1" was part of the "9." So it--
18 DR. PORTER: Isn't total. Okay.
19 DR. DePORRE: --is total, nine patients
20 who had a death that was attributed to an
AE.
21 And I agree with you that it's
very
22 difficult--not to say almost impossible--to
discern
165
1 whether the death was due to the AE, or due
to the
2 underlying disease.
3 DR. PORTER: Well, this is the way
you
4 classified it, though: you said, "AEs
leading to
5 death."
6 DR. DePORRE: Correct. But that is--the
7 drug relatedness is difficult to discern
between
8 disease and drug. And the one patient was
9 attributed--the AE was attributed very
clearly by
10 the investigator to the drug. For the other eight,
11 there are other confounding factors, that
is, in
12 the context of the progression of disease.
13 But an absolute certainty to
discern
14 between the two is not possible.
15 DR. PORTER: Okay
16 DR. MARTINO: Are there other
questions?
17 DR. DeLAP: I would come back again
to the
18 overall mortality, which was 12 percent in
the
19 first 30 days with this drug. And, as we've heard
20 from Dr. Stone, that's well below the
expectation.
21 So, for this patient population,
with the
22 natural progression of their leukemia, we
feel that
166
1 tipifarnib is not contributing substantially
to an
2 early mortality issue. We think that deaths we're
3 seeing in the first 30 days reflect the
natural
4 history of the disease, by and large.
5 DR. MARTINO: Dr. Temple?
6 DR. TEMPLE: A study, INT-17, was
7 mentioned.
That was in people who had been
8
previously treated. If I recall,
the response rate
9 was virtually nil in that--something like 3
out of
10 250, or something like that? Did I read that
11 aright?
12 DR. DeLAP: The response rate is
low.
13 DR. TEMPLE: So, I just want to be
clear:
14 you have no intent that people who've been
15 previously treated be treated with this
16 drug--correct?
17 DR. DeLAP: We do not have
risk-benefit
18 data to support treatment of previously
treated
19 patients.
20 DR. TEMPLE: Okay--but I mean, that
might
21 even be something we would ask to go into
labeling,
22 if the drug were otherwise okay. Because the rate
167
1 was so low.
2 DR. DeLAP: Our intent would be
that the
3 patients for whom we have the established
4 risk-benefit, where we can show the durable
5 remissions and the good outcomes, that those
are
6
the ones who should get the drug--yes.
7 DR. TEMPLE: Okay, just one other
question.
8 I mean, obviously, as everybody's
been
9 hinting, the response rate here is quite
low. So,
10 presumably, people who are candidates for
11 regulation chemotherapy, you'd want them to
get it.
12 Is this also a drug that someone
who's at
13 the margin might give a whack at, and then
use
14 chemotherapy afterward, if nothing came of
it here?
15 Is that part of your thinking, too? You know, take
16 the easy one and if it doesn't work, go on
to real
17 chemotherapy?
18 DR. DeLAP: Yes, it's difficult to
address
19 exactly that issue. I think that--again, the
20 recommendation in labeling would be to stick
as
21 closely as we can to the risk-benefit that
we've
22 established, and treat those patients.
23 But I'll ask Dr. Sekeres, who can
talk to
24 us about how the decision is made as to
25 chemotherapy or no.
168
1 DR. SEKERES: Hi, I'm in
clinic. It's a
2 great question you ask.
3 And this is a disease that I think
of as
4 the worst solid tumor you can imagine;
people
5 coming in with State IV disease, and then
add on
6 top of that, poor prognostic factors.
7 So we don't have a lot of room to
work
8 with when we talk with patients. A typical
9 conversation I'll have is: "I can offer
you options
10 A, B or C.
A is supportive care--"--I like to call
11 it "aggressive supportive care" so
patients don't
12 think I'm giving up on them. B is some sort of
13 mid-range therapy. And C is remission-induction
14 therapy.
15 My approach, for better or worse, whether
16 this works is will aggressive chemotherapy,
to look
17 somebody in the eye and say, "The
chance I can get
18 you into remission is, for all comers over
60,
19 somewhere between 40 and 55 percent; for a
169
1 70-year-old, or 80-year-old, that will go
down to
2 30 percent.
And with worse risk factors, as low as
3
15 percent, and a treatment-related mortality of 20
4 to 25 percent.
5 And you've heard that message over
and
6 over again.
7 Option B is mid-range therapy. Now, what
8 I have to offer in mid-range therapy are
clinical
9 trials or Hydrea. And I can't look somebody in the
10 eye and say, "I can offer you a
complete remission"
11 with either of these approaches.
12 And I leave the decision up to the
13 patient.
Now, I may weight it a little bit, but I
14 give them real percentages, if at all
possible.
15 And then whatever decision they make, I
support it.
16 If a person were to choose option
B--if
17 that option happened to include a farnesyl
18 transferase inhibitor such as tipifarnib, I
would
19
support them through it. If they were then to
20 progress, I'd probably re-address the issue
with
21 them.
22 DR. MARTINO: Are there any other
170
1 questions?
2 Yes?
3 MR. FLATAU: I'd just like to ask
the
4 question: if you're looking a patient in the
eye
5 and you're going to treat him with
tipifarnib,
6 you're going to say: "You have a 15
percent chance
7 of getting into remission." And what about the 85
8 percent of the people? What are you going to tell
9
them?
10 DR. DeLAP: After Dr. Sekeres, I'd
also
11 like Dr. Karp to come up and say, because
she has,
12 obviously, the direct experience in the
protocol.
13 DR. SEKERES: It's not easy to look
14 somebody in the eye and give them those
15 percentages.
It really isn't.
16 Fifteen percent isn't zero
percent. So I
17 would say to that person: "There's a 15
percent
18 chance you could go into a complete
remission."
19 I'd try to be clear that complete remission
is not
20 the equivalent of cure. If I have data to support
21 the fact that that person may live eight or
10
22 months longer than with no other therapy,
then I'll
171
1 make that statement also, and I'll say to
that
2 person: "If you don't happen to go into
a
3 remission, and your leukemia remains, we'll
discuss
4 what your options are after that."
5 That may involve remission-induction
6 therapy, if that's what a person wants. It may
7 involve aggressive supportive care. Or it may
8 involve something like Hydrea if they want
9 something in the middle.
10 MR. FLATAU: But, I guess the point
is: you
11 can't tell them that this is better--if
they're in
12 that 85 percent, that this is better than
best
13 supportive care, because you don't have any
data.
14 DR. SEKERES: I don't know where they'll
15 fall when I first have that discussion. So I just
16 give them the percentages as I just gave
them to
17 you.
And then if their disease comes back, I
18 wouldn't be disingenuous about what I'd say
in the
19 future: "You have relapsed
leukemia. This is
20 something we need to worry about. We can offer
21 things that can manage your disease, but we
can't
22 reliably offer you something that will
improve your
172
1 survival."
2 DR. DeLAP: I think the question
about the
3 patients who don't have a complete
remission, and
4 what to say to them--I think there is some
other
5 observations perhaps Dr. Karp can give us
about
6 things that were seen in the trial. Although,
7 again, the complete remissions are the ones
that
8 have the established link to patient
benefits.
9 DR. KARP: Well, first of all, not
to be
10 too brutal, but this is a fatal
disease. So, if
11 one is fortunate enough to be in that 15
percent
12 who will achieve a complete remission, the
benefit
13 of that is very palpable.
14 Now, for the other 85 percent,
there are
15 patients who do get partial remissions,
there are
16 patients who get hematologic
improvement--which is
17 not as concrete a definition. And then there are
18 patients--in fact, I believe 30 percent of
our
19 patients--had something that was akin to
"stable
20 disease." And you can say, well, what is "stable
21 disease?"
22 These patients did not
progress. And, in
173
1 fact, these patients didn't quite quality
for bona
2 fide palpable improvement, yet they were
alive;
3 they felt relatively well; they were out of
the
4 hospital; they were living at home. And there are
5 a number of patients who went on for months
like
6 that.
7 I have a gentleman 76 years old--I
seem to
8 specialize in older men--
9 [Laughter.]
10 --[whispers] especially if they're
11
wealthy.
12 [Laughter.]
13 But he's a 76-year-old gentleman
who had
14 myelodysplasia for about a year, and then
15 transformed into AML, with virtual replacement
of
16 his bone marrow by blasts, and no functional
17 platelet production, no functional red-cell
18 production.
19 This patient has been on Zarnestra
for one
20 year.
He has been taking Zarnestra for 21 out of
21 every 28 to 42 days for one year. He does not have
22 a "partial remission" by standard
criteria, and yet
174
1 he has not required a platelet transfusion
for
2 eight months. He has not required a red-cell
3 transfusion for eight months.
4 So, when Dr. Sekeres was talking I
said to
5 myself, "Well, what would I tell my
patient? I
6 think I'd have him call Dr. Sekeres, because
he
7 said it so eloquently."
8 And so it is a very painful thing
to say
9 to someone that "I don't have anything
for you."
10 And that may--you know, when we think about
11 benefit, it may not be such a terrible thing
for
12 there to be a benefit for the caregivers, as
well.
13 Because there are so many patients--so many
elderly
14 patients--for whom intensive chemotherapy is
15 clearly not the answer. And it is painful to
16 either say to them, "Well, we can do
this. We can
17 do anything.
But we stand a very good chance of
18 killing you in the process--or certainly
doing you
19 no help." Or, on the other hand, "We don't have
20
anything that we can do. We can
turn a couple of
21 dials."
22 Having another option, I think, is
175
1 extremely important for everyone concerned
in this
2 partnership: the patient, the physician, the
nurse,
3 all the family members.
4 DR. MARTINO: Anyone else have any
5 questions?
6 [No response.]
7 Seeing no further questions--Dr.
Pazdur, I
8 think you wanted to address the committee
before we
9 go into the discussion phase?
10 DR. PAZDUR: Let me just clarify
some of
11 the comments that I made earlier.
12 Originally, when we asked the
question, we
13 were asking a regular approval. And I prefaced my
14
comments that the agency had looked at complete
15 remissions with sufficient duration--I
should
16 emphasize--as equaling clinical benefit.
17 But I think it's important when we
discuss
18 this, we cannot just look at the endpoint
"complete
19 remissions" without looking at the
magnitude of
20 change on this endpoint. Because, obviously, a
21 complete remission rate of 80 percent, with
a
22 durable duration, is much different than
something
176
1 with an 11 percent duration.
2 So I think we have to address that
3 question is: is this an established--with
the
4 endpoint of complete remission, with the
magnitude
5 that is associated with this, is this an
6 established surrogate for clinical benefit;
in
7 other words, will it equal an improvement in
8 survival of these patients, if the entire
patient
9 population was followed out? Or would people want
10 to discuss whether we should look at this as
a
11 surrogate reasonably likely to predict
clinical
12 benefit; i.e., looking at an accelerated
approval
13 situation?
Okay?
14 And let me go into a little
further
15 description about accelerated approval,
since there
16 are some people that have not been on the
committee
17 before.
18 For accelerated approval, we're
looking at
19 life-threatening diseases--which this
obviously is.
20 The surrogate has to be reasonably likely to
21 predict clinical benefit. You have the surrogate:
22 it's a complete remission rate of x
177
1 percent--whether one wants to believe it's
11
2 percent, or 15 percent; whether one wants to
put
3 into this equation the partial responses, or
the
4 disease stabilization. That's something that you
5 have to consider.
6 Generally, in acute leukemia we've
looked
7 at complete responses--or complete
8 remissions--alone, and not these other
evidence of
9 activity.
10 But, in your mind, is that
reasonably
11 likely to predict clinical benefit? Okay?
12 The other caveat to accelerated
13 approval--and this is a very important
question
14 that deserves more discussion here--is that
the
15 therapy has to be an improvement over available
16 therapy.
17 Now, that improvement can be an
18 improvement in terms of toxicity, and it
could be
19 an improvement in terms of efficacy--and
I'll get
20 to that later. But I think one of the major issues
21 here is the concept of available
therapy. And that
22 has to have an impact on what population are
you
178
1 talking about.
2 When we originally discussed this
3 application and bringing in a single-arm
study, we
4 were told that this patient population would
not be
5 treated by hematologists or by medical
oncologists,
6 because these people generally are not
treated;
7 they're too old, there's a consensus,
relatively
8 among oncologists that these individuals
that are
9 greater than 75, or greater than 65 with
MDS, would
10 not constitute a patient population that
would
11 receive standard induction therapy. And, hence, we
12 believed that going ahead with a single-arm
trial
13 seemed reasonable in that situation.
14 It's interesting, however, that
there are
15 individuals that were on this study that
went on
16 to, obviously, standard induction
therapy. So I
17 think one has to define, you know, what is
the
18 population.
19 Clearly, the indication that has
been
20 presented here by the company--the treatment
in
21 elderly populations--is not something that
we would
22 consider.
At the least we would--or at the
179
1
most--we would consider really defining the
2 population as studied; i.e., patients that
are
3 greater than 75, or patients that are
greater than
4 65 with MDS--and perhaps even put in a
caveat "that
5 cannot tolerate chemotherapy."
6 But, here again, I think that
requires,
7 again, some discussion by the
committee. Because
8 what is this population? Should they receive
9 standard induction therapy? Are we defining the
10 population correctly here, by the inclusion
11 criteria that the sponsor put into this
single-arm
12 study?
13 When we originally discussed this
14 application with the sponsor, we did discuss
15 accelerated approval, and that's why this
ongoing
16 study with a supportive care arm is
ongoing. One
17 may question why is this study going on
18 predominantly in Europe? One of the issues was
19 that if this drug received accelerated
approval,
20 this would have a significant impact on the
accrual
21
of patients onto this trial.
Obviously, if the
22 drug is not approved under accelerated
approval, we
180
1 could re-look and re-examine the sites of
study.
2 I think when we do the questions
and
3 actually come to the voting, and as you
ponder this
4 application, there are three decisions that
we will
5 be asking you to make.
6 One, should this be regular
approval? In
7 other words, does this 15 or 11 percent
complete
8 response rate equal clinical benefit? Is this an
9 established surrogate for clinical
benefit? Or,
10 should this be considered for accelerated
approval?
11 And if it's for accelerated approval, you
have to
12 feel that this endpoint of complete response
rate
13 and other--of supporting evidence--is
reasonably
14 likely to predict clinical benefit. In other
15 words, what you're seeing here, do you, in
your
16 clinical judgment, think that this confirmatory
17 trial that they're doing is going to turn
out
18 positive?
Okay?
19 And then you have to also ask
yourself: is
20 this therapy, in the patient population
they're
21
defining, an improvement over available therapy.
22 And, again, I said that improvement can be
an
181
1 improvement in toxicity evaluation, or in
efficacy.
2 And obviously, in this situation, most
people, I
3 think, would look at the safety issues and
the
4 safety benefit. But then one has to weigh that,
5 potentially, against: are you giving up some
6 efficacy, potentially, in this patient
7 population--and make a risk-benefit
decision.
8 So you have two decisions.
9 And there's a third decision
here. And
10 that third decision would be to say: "I
don't like
11 full approval. I don't like accelerated approval.
12 And perhaps we should wait for the
completion of
13 the
ongoing randomized study looking at this drug
14 against best supportive care."
15 All three of these decisions are
on the
16 table.
And I'll turn it over to Silvana.
17 DR. MARTINO: Thank you.
18 Discussion of the Questions
19 DR. MARTINO: I would like to take the
20 questions in a certain order, and it's the
order
21 that Rick has suggested.
22 So what I first would like a
discussion on
182
1 is the first question, which I will read to
you:
2
Does the risk-benefit analysis support regular
3 approval of Zarnestra for the first-line
treatment
4 of AML patients aged 65 or older, with prior
MDS,
5 or age 75 and older.
6
So it is the regular
approval concept that
7 I want your thoughts on first.
8 Dr. Cheson, I'm going to start
with you.
9 What are your thoughts on this?
10 DR. CHESON: I didn't volunteer,
but I'll
11 do it.
12 [Laughter.]
13 DR. MARTINO: Yes, that's right.
14 DR. CHESON: I'm fairly
uncomfortable with
15 giving full approval for this drug for this
16 indication, for all the reasons that have
already
17 been dealt with.
18 It's a heterogeneous population of
19 patients, some of which--we do treat
patients 65
20 years or older with aggressive therapy, and
some of
21 them do rather well. I've just it to somebody.
22 And given that there is such concern about
that,
183
1 and the fact that there is a randomized
trial going
2 on, I--in all good conscience--couldn't
support
3 that.
4 The data are--there are some answers to my
5 questions which I still haven't heard. The data
6 aren't available. The patients are heterogeneous.
7 There are other therapies out there. And
8
supportive care has gotten better.
And I'd like to
9 see the results of that Phase III
trial. I don't
10 feel comfortable with this.
11 DR. MARTINO: Dr. Perry.
12 DR. PERRY: First, I'd like to thank the
13 sponsor--
14 DR. MARTINO: That's for you, Dr.
Cheson.
15 DR. CHESON: Oh, there were a bunch
of
16 questions regarding the benefitted patients
who got
17 less than a CR--things that you said you
didn't
18 have the data on yet.
19 DR. DeLAP: We have worked to get a
little
20 more of that data, and we could still show
it if
21 you're interested.
22 DR. CHESON: I'd like to see it,
but I'm
184
1 not sure it will change my feeling about it.
2 DR. THIBAULT: Well, of the 10
patients who
3 did not achieve a CR, there were three
patients who
4 achieved a partial remission. All these patients
5 had a 50 percent reduction in blast count,
down to
6 5 to 19 percent. Their recovery of peripheral
7 counts was an ANC to 3.4 thousand, and a
platelet
8 count of 240,000.
9 For the patients who achieved
hematologic
10 improvement--there were seven of
them--again, in
11 the bone marrow there was a 50 percent
decrease in
12 the blast count to a range of 5 to 19
percent. The
13 recovery of counts was not complete: ANC was
at
14 2,000, but the platelet count recovery was
at
15 75,000.
16 In terms of reduction of
transfusion, we
17 see it during the duration of that benefit,
which
18 averages three months.
19 DR. MARTINO: Dr. Perry.
20 DR. PERRY: First I'd like to thank
the
21 sponsors for a very concise and very well
organized
22 presentation. You guys can take this show on the
185
1 road any time, I think.
2 Secondly, I'd like to echo Bruce's
3 comments.
I would feel uncomfortable with
4 approval--with regular approval. I would like to
5 discuss further the issue of accelerated
approval.
6 I was not on the committee when
Iressa
7 came before this committee, and I think it
was
8 my--probably, although it as not attributed
to
9 him-- my new best friend Dr. Brawley, who
pointed
10 out that Iressa's approval was perhaps a
mistake.
11 I would feel more confident if we
gave
12 this drug accelerated approval, and then
waited the
13 results of the 301 study.
14 DR. MARTINO: Now, I still want
this group
15 to deal with this issue of regular approval
before
16 we move on to anything else.
17 Dr. Levine?
18 DR. LEVINE: So my frustrations
relate to:
19 number one, I really, really think that it
would
20 have been helpful to have very strong
clinical
21 benefit data; not just talking, but what are
the
22 transfusions before and after, in all of
these.
186
1 That would be extremely helpful to know.
2 The second thing--I can't help it,
but it
3
frustrates me, because MDS is what you're hanging
4 your hat on.
And it would make me very much more
5 comfortable to know that I know that that is
MDS,
6 and it's not somebody who, for a month
before the
7 diagnosis of AML, you know, was a little
anemic.
8 You know, maybe that's not really MDS. I would
9 feel much better.
10 And so when I'm looking at this,
I'm just
11 looking at--I can look at age, over 75, and
I can
12 look at the poor chromosomes. You know, those are
13 very strong.
But it depletes the larger study.
14 So I would not be comfortable at
this
15 point in regular approval.
16 Dr. Bukowski.
17 DR. BUKOWSKI: Yes, I'd like to
echo those
18 comments.
19 I get the impression--I used to
take care
20 of acute leukemia when I was younger, under
the age
21 of 50.
22 [Laughter.]
23 I'm impressed, Bruce, that you
still do
24 this.
That's good.
25 But, anyway, this seems to be a
very
187
1 heterogeneous disease. Not one entity. We're
2 asked to make a judgment on Phase II data in
a very
3 different group of patients. There may be subsets
4 of individuals here.
5 Therefore, I think, in reality,
I'd like
6 to see more data before voting for regular
approval
7 at this point in time. I just am bothered by the
8 heterogeneity of the population that may
reflect
9 some of the results. So this concerns me somewhat.
10 DR. MARTINO: Dr. O'Brien, you've
been very
11 quiet.
But I do want to hear your voice.
12 DR. O'BRIEN: I think that the clear
13 problem, in my mind, with the data is again
how
14 well defined the population is.
15 Certainly, within AML, this is a
high-risk
16 group, in terms of achieving a remission and
17 induction mortality, because they're older
and
18 because they have poor cytogenetics, and
prior
19 myelodysplastic syndrome. But, from a patient
188
1 point of view, they're not really that bad.
2 So what am I trying to say by
that? I
3 looked in our data base before I came
here. We
4 have about a thousand patients treated over
the age
5
of 65. And I actually pulled out these same
6 eligibility criteria: over 75, or 65 to 74
with MDS
7 and performance data zero to one. And I looked at
8 the abnormal cytogenetics, were about 45
9
percent--so very similar to this trial, of 49
10 percent.
11 And then I only looked at regimens
that
12 included Ara-C and anthra-cycline. And the CR rate
13 was about 38 percent, and the induction
mortality
14 was 18 percent.
15 Now, the first thing anybody can
say is:
16 well, those patients were well enough to get
to
17 M.D. Anderson, and that selects them. And that's
18 absolutely true. But these patients were also well
19 enough to get to Stanford or Hopkins or
Cornell.
20 In fact, you heard that Mr. McPherson came
from
21 Ohio to Hopkins to get treated.
22 So I think in that point of view
they were
189
1 similar.
This is not a seek-out trial.
2
I think the biggest factor which affects
3 whether the patient and the physician decide
4 whether they're going to get chemotherapy or
not in
5 this older age group is the performance status.
6 But these are all performance status
zero-to-one
7 patients that are walking in the door. It's way
8 different if a patient comes in the door in
a
9 wheelchair.
But that's not the population that we
10 have defined here.
11 And it also excluded patients with
high
12 white counts over the age [sic] of 30,000,
which I
13 didn't do when I looked at my analysis. And they
14 have clearly a higher induction mortality
and a
15 lower CR rate.
16 So, I think if I was a patient I
would
17 rather get a CR on Zarnestra than chemo--it
is
18
easier. They spend less time in
the
19 hospital--there's no question.
20 My concern is that there may have
been
21 patients in the study who really would have
gotten
22 a CR with chemotherapy. And I think that was
190
1 alluded to before, when 10 of the patients
went on
2 to get chemo, and six got a remission.
3 So I think the drug clearly has
activity,
4 but I think the heterogeneity of the
population in
5 this trial, and the fact that I am hard
pressed to
6 believe that really all of these patients
would not
7 have been offered chemotherapy if this trial
wasn't
8 available.
9 Now, the randomized trial may make
that a
10 different story, because there the people
know
11 they're not going to get anything. So both the
12 patient and the doctor are clearly going to
agree,
13 in that randomized trial, that they're not
going to
14 take anything.
15 I don't think that that's
necessarily the
16 case here: that these patients wouldn't have
gotten
17 chemo and might not have benefitted from it.
18 DR. MARTINO: Dr. Brawley.
19 DR. BRAWLEY: you know, this really
is a
20 lot like Iressa all over again; even the 11
percent
21 response rate.
22 What we have here is a drug that
clearly
191
1 has activity. Some people do very well on it when
2 they respond. I think the slide that Dr. Thibault
3 showed after I asked him is very compelling
for
4 some type of approval.
5 But the reality of the situation
is: if we
6 approve this drug generally, we're going to
be
7 telling a company--we're going to put a
company in
8 a terrible conflict of interest
situation. We're
9 going to tell them that your market is all
people
10 over a certain age with AML, or all people
over a
11 certain age with MDS and then AML--but go
out and
12 find the 10 or 11 percent of people in that
13 population that you really should be selling
the
14 drug to.
15 So their job, after approval, will
be to
16 decrease--yes, actually Dr. Cheson, in great
17 wisdom, said this when we, as a group, voted
for
18 Iressa.
So I'm very torn.
19 There is clearly activity. This drug
20 clearly helps some people. We clearly need to
21 figure out a little bit better who among the
22 patients would be best served by getting
this drug,
192
1 and who among patients would be best served
by
2 going to something else, and not wasting
their time
3 and money on this drug.
4 I would love to see--that all
being said,
5 I would like to see some type of approval of
this
6 drug.
I don't know if full approval is the right
7 approval.
And I accept the fact that accelerated
8
still creates this problem that I just talked
9 about.
10 DR. PAZDUR: For clarification:
since
11 people are mentioning Iressa here and there,
Iressa
12 did have accelerated approval--so everybody
13 understands that. Because there's some people on
14 the committee that may have not been
present.
15 DR. MARTINO: Are there other
comments?
16 MR. FLATAU: Yes.
17 DR. MARTINO: You're my blind side,
and I
18 do apologize.
19 MR. FLATAU: Yes, I guess I'm not
really
20 comfortable with full approval. I mean, this is a
21 disease in these patients that, with or
without
22 Zarnestra, is uniformly fatal. And, at best, we
193
1 can hope to extend their life with a
reasonable
2 quality of life. And 85 percent of the patients,
3 we have really no idea whether their life is
4 extended or shortened or its the same, and
how
5 their quality of life is.
6 And without that, I don't know how
people
7 can be said--you know: here, you have a
choice of
8 this drug, or not.
9 I mean, there is standard therapy
now.
10 Standard therapy is, I guess, supportive
care. And
11 maybe they'll do better on supportive care
than on
12 this drug.
And we don't really know.
13 DR. MARTINO: Dr. George.
14 DR. GEORGE: Just a general
comment:
15 there's a longstanding issue of single-arm
Phase II
16 trials' being problematic, particularly in
an area
17 where the heterogeneity issue is really
important.
18 There are some Phase II trials that can be
done in
19 more homogeneous populations, where you have
a
20 little better feel for the patient
population. But
21 in this setting, not only is the disease
very
22 heterogeneous, but the way the patients got
on this
194
1 study is still, I think, contributes to the
2 heterogeneity that makes--all of that makes
for
3 sort of unquantifiable kinds of
uncertainties that,
4 to me, raise to the level of not--just to
address
5 this issue--not meeting the full-approval
criteria.
6 DR. MARTINO: I have a bigger
problem with
7 this whole issue. I think you're all being
8 remarkably kind.
9 I do think this drug does
something to
10 some people.
I'm just sitting here being made very
11 uncomfortable by the suggestion that
something that
12 works 10 to 15 percent of the time--and in
this
13 regard, I'd like to give the company the
benefit:
14 let's make it 15--okay?
15 But as the gentleman on my left
keeps
16 reminding all of you--and I'll remind you
that he
17 represents the real world, he's our patient
18 representative--someone's expecting me to
walk into
19 a patient's room and to say, "85 to 90
percent of
20 the time this thing ain't gonna work. Do you
21 really want it?" Okay--that's what you're asking
22 of me as a physician. Okay?
23 Are we really that excited about a
15
24 percent CR rate? I have a much more global problem
25 here.
195
1 DR. PAZDUR: Could I address
something
2 here?
And that is: what accelerated approval is.
3 Accelerated approval is not
"approval
4 light."
Okay? [Laughs.] I want to make that point.
5 This is an approval of a drug,
with
6 marketing of that drug--okay? Companies can charge
7
for that. Companies can
advertise. There are some
8 advertising restrictions. There are some
9 restrictions on that advertising, as far as
being
10 cleared by the FDA. There are provisions in that
11 accelerated approval, Subpart H, that can
allow the
12 FDA to withdraw the drug--and we may be
getting
13 into that issue in the future.
14 The other areas are post-marketing
15 studies; that we have more teeth to really
make
16 sure that they are done, and to look at, and
to
17 address.
18 But it is a marketing approval of
the drug
19 here.
And I think we have to understand this.
196
1 This is not "approval light;" some
issue of--and
2 you have to consider, it is an approval of a
drug
3 here.
And as such--let me finish, Bruce--you have
4 basically two criteria that you have to meet
here:
5 one, that the surrogate is reasonably likely
to
6 predict clinical benefit. And you really have to
7 ask in your mind while you're contemplating
the
8 decision here: given this data here, is this
9 reasonably likely to predict clinical
10 benefit--okay?
11 Number two: the issue of basically
being
12 better than available therapy, and what is
the
13 available therapy in this population? And how do
14 you compare it? Is it a homogeneous population
15 that Dr. George has been alluding to, that
would
16 give you confidence? And how can we define that
17 population?
18 DR. MARTINO: Dr. Cheson?
19 DR. CHESON: While I strongly
support the
20 spirit of the accelerated approval--you know
what
21 I'm going to say--when you have drugs--well,
you
22 know, we can make up a name and call it
"the Iressa
197
1 principle," since we seem to be
throwing that
2 around--when one has a drug with three
negative
3 Phase III trials, and it hasn't been taken
from the
4 market, it gives us a great deal of
discomfort when
5 we have another drug, with one Phase III
trial out
6 there, and if we give this accelerated
approval--I
7
know we're not there yet, I'm sorry, Madam
8 Chairman, but I can't resist this--and that
trial
9 is negative, then what's going to happen?
10 So, again, exactly the same
scenario--and,
11 as you know, I was one of the three
unpopular
12 people back then who voted against
Iressa. It just
13 makes me uncomfortable that there is sort of
an
14 edentulous position about the accelerated
approval
15 process.
I would have felt a lot more
16 comfortable--we tend to be a little
historical
17 here.
We have been reacting to a number of things
18 over again.
It has hurt some drugs and it has
19 helped others.
20 But it makes me much less
comfortable to
21 do that in the context of what happens with
that
22 compound.
Because I have no faith that if this is
198
1 a dead-negative trial, that the drug is not
going
2 to be continued to be sold.
3 And I'm sorry for that position,
but I'm
4 talking it, and I'm going to stand by it.
5 MR. FLATAU: How many drugs have
gotten
6 accelerated approval and have subsequently
been
7 withdrawn?
8 DR. CHESON: None. None.
That was the
9 point.
Zero. Thank for--if I didn't make
that
10 clear.
That's the one who really should have been
11 out of there--clearly.
12 DR. MARTINO: But I just want to
remind the
13 group--and I do appreciate that there is
history
14 here.
Our decisions need to be based on the merits
15 of this drug. Okay?
16 So, again, I want to take you back
to
17 regular approval.
18 Are there any additional comments
before I
19 take a vote on the question of regular
approval?
20 [No response.]
21 If there are none, then I will
start the
22 voting.
And in that process, first please identify
199
1
yourself, and then give us a "yes" or "no" vote.
2 The industry representative does
not get a
3 vote.
4 DR. O'BRIEN: Susan O'Brien No.
5 DR. CHESON: Bruce Cheson. No.
6 DR. GEORGE: Steve George. No.
7 DR. BRAWLEY: Brawley. No.
8 DR. MORTIMER: Mortimer. No.
9 MS. HAYLOCK: Haylock. No.
10 MR. FLATAU: Arthur Flatau. No.
11 DR. LEVINE: Levine. No.
12 DR. MARTINO: Martino. No.
13 DR. PERRY: Michael Perry. No.
14 DR. BUKOWSKI: Bukowski. No.
15 DR. MARTINO: Eleven
"nos."
16 Now I will turn your thoughts to
do we
17 want to give this agent accelerated
approval?
18 Who wants to speak to that,
please?
19 Dr. Perry.
20 DR. PERRY: I'd like to make a
motion that
21 we consider accelerated approval, please.
22 DR. MARTINO: Tell me that
again? I just
200
1 want to make sure I heard you correctly.
2 DR. PERRY: I said I'd like to make a
3 motion that we consider accelerated
approval.
4 DR. MARTINO: So moved.
5 DR. PERRY: I'm trying to follow
Robert's
6 Orders here.
I'm sorry.
7 DR. MARTINO: All right.
8 DR. BRAWLEY: Yes, I'll second
that.
9 DR. MARTINO: Thank you. I accept the
10 first and the second.
11 Now, I'd like some discussion.
12
[Laughter.]
13 DR. PERRY: All right. I'll start out
14 then.
15 I think this drug have activity in
a group
16 of patients that don't have very viable
options.
17 It's true, there are some patients who did
get
18 chemotherapy afterwards, but I can't escape
the
19 thought that maybe they improved on this
drug to
20 the point that they could be considered for
21 chemotherapy when they wouldn't have--they
and
22 their doctors wouldn't have considered it
before.
23 And I think this is a niche
drug. This is
24 not going to be a drug for thousands and
thousands
25 of people.
It's going to be for a small patient
201
1 population that, in my mind, has no other
option,
2 other than supportive care.
3 And I think with 301 in the
background,
4 already underway, we'll know at some point
in the
5 near future whether or not it really does
have a
6
significant survival advantage.
And at that time,
7 if it doesn't, it can be taken off the
market.
8 But, in the meantime, I think
there's a
9 potential to help people.
10 DR. MARTINO: Dr. Brawley.
11 DR. PAZDUR: Could I please
emphasize:
12 please concentrate. We have to make sure people
13 understand the two principles here: that
this is a
14
surrogate reasonably likely to predict clinical
15 benefit, and you believe that this is an
16 improvement over available therapy--that in
your
17 voting and your considerations, those two
18
conditions are met.
19 DR. PERRY: Did my statements
confuse in
202
1 that regard?
2 VOICE: Yes.
3 DR. BRAWLEY: Can we have the
4 hospitalization data back up here
again? My
5 favorite slide.
6 VOICE: The second one.
7 DR. BRAWLEY: The second one. The second
8 one.
9 [Slide.]
10 I think that this slide actually
is
11 evidence--since the people who responded
stayed out
12 of the hospital a lot more, I think that
this slide
13 is evidence that there is some
clinical--yes, I
14 think that this slide is some evidence of a
15 clinical benefit: the fact that people who
had a CR
16
had less hospitalizations that--
17 VOICE: [Off mike.] [Inaudible.]
18 DR. BRAWLEY: Well, that's
prior. During
19 remission, it's 30 percent versus 67 percent
for
20
those who did not have a CR.
21 VOICE: [Off mike.] [Inaudible.]
22 DR. BRAWLEY: Say again?
23 [Multiple speakers off mike.]
[Inaudible.]
24 DR. BRAWLEY: You can have multiple
25 hospitalizations, is the issue.
203
1 [Multiple speakers off mike.]
[Inaudible.]
2 DR. MARTINO: Otis, get to your
point,
3 please.
4 DR. BRAWLEY: Okay--I'm sorry. I'm trying
5 to interpret the slide again.
6 I do believe that there is a
clinical
7 benefit to this drug. I, however, think that there
8 is a sub-population--we talked about the
9 heterogeneity of AML. I suspect that there is a
10 sub-population of individuals with AML who
are
11 sensitive to farnesyl transferase
inhibition. And
12 I really do think that the science, and the
13 population, would be best served if we could
14 ultimately figure out who that population is
before
15 they're given this drug.
16 You know, this is a great deal
like
17 tamoxifen, breast cancer, before we had
identified
18 the estrogen receptor--finding the
individuals who
19 have the molecular or genetic marker that
indicates
204
1 that they will respond to farnesyl
transferase
2
inhibition.
3 My great problem is: should we
allow
4 patients who have AML, who are in a bad
situation
5 the crap-shoot--and a 10 to 15 response rate
is a
6 crap shoot--for response before we get to
the point
7 that we can figure out who truly benefits
from the
8 drug?
9 That being said--and I may very
well be
10 making the same mistake I made with
Iressa--that
11 being said, I would favor accelerated
approval.
12 DR. CHESON: can you put that slide
back up
13 again--Otis' slide? I have another question about
14 it. Maybe
you said this before and I didn't get
15 it.
16 DR. THIBAULT: Yes, may I have the
slide
17 up, please?
18 [Slide.]
19 DR. CHESON: Does he right-hand
column,
20 which is those patients who didn't get the
CR--
21 DR. THIBAULT: Yes?
22 DR. CHESON: At what point in
time--was
205
1 that before, during, after they got the
drug?
2 DR. THIBAULT: yes, I can go over
that for
3 you in detail.
4 The patients who are non-CR
patients were
5 hospitalized at any time on study.
6 The patients who achieved a CR,
many of
7 whom--most of them were hospitalized for the
8 initial duration of the CR, and then the
time after
9 they achieved a CR, then they remained
10 hospitalization-free.
11 DR. CHESON: But I still don't
understand
12 the right-hand column. Again, back to the 67
13 percent thing. I'm not sure--
14 DR. THIBAULT: So, the non-CR--
15 DR. CHESON: --there's not a better way to
16 analyze it.
17 DR. THIBAULT: No, the non-CR
patients, 66
18 of them, for 67 percent of the total--
19 DR. CHESON: That's not 67 percent
of the
20 total.
21 VOICE: [Off mike.] [Inaudible.]
22
DR. THIBAULT: 66 patients out of 116--yes,
206
1 that's right--of the patients were
hospitalized.
2 And this is at any time during the study
treatment.
3 And these patients spent a greater
proportion of
4 their time in the hospital, definitely, than
the
5 patients who achieved the CR.
6 DR. CHESON: But if that's at any
time
7 during their treatment, you could say the
same
8 thing for the patients who got a CR: 75
percent of
9 them were hospitalized at some time during
their
10 treatment.
11 So that right-hand column doesn't
help me
12 any.
I'd like to know whether that was before,
13 during or after they got the Zarnestra
therapy.
14 DR. THIBAULT: Our data
includes--may I
15 have the next slide up? There's a fifth column.
16 [Slide.]
17 No, there's four bullets.
18 Fourteen of the 20 patients were
never
19
hospitalized while in CR. Six
were--so--
20 DR. CHESON: Which means six of 20
21 were--okay.
22 DR. THIBAULT: Six were
hospitalized at
207
1 least once.
The time spent in hospital is 21 days,
2 and the range is six to 21 days.
3 VOICE: [Off mike.] [Inaudible.]
4 DR. THIBAULT: 12 days--it's the
5 double-language.
6 DR. CHESON: Dyslexia.
7 DR. THIBAULT: Then the percent of
the
8 remission time spent in hospital is 8 percent.
9 The reason for hospitalizations:
five
10 patients had drug-induced myelosuppression,
because
11 Zarnestra is myelosuppresive; and one
patient was
12 hospitalized for unrelated co-morbidity, for
which
13 he required surgery--GI surgery.
14 DR. CHESON: Next slide. It's the one
15 that's going to show me what I want to see.
16 [Laughter.]
17 DR. THIBAULT: So we do not have
this data
18 for the patients who did not achieve a
CR. What we
19 do know is that they spent approximately
twice as
20 long in the hospital.
21 DR. CHESON: Well, I don't see
those data.
22 I see--you're telling me twice as long. I don't
208
1 see numbers.
2 DR. THIBAULT: I will try to get
that for
3 you.
4 DR. MARTINO: Can I--I'm sorry,
people. I
5 need to remind this group what our
responsibility
6 is here.
It is not to decide whether this drug
7 does something, anything--"hopefully,
please:
8 something, anything." It is whether it is good
9 enough for us to give it accelerated
approval.
10 Accelerated approval means to me that
11 something is not barely visible in its
activity.
12 There has to be something that's bordering
on the
13 exciting, people. God, I wish someone would
14
impress me today.
15 DR. CHESON: That was precisely
what I was
16 trying to get at, Madam Chairman.
17 DR. STONE: Could I have the floor,
or is
18 that not possible?
19 DR. MARTINO: You may have the
floor at the
20 moment.
21 DR. STONE: Thank you very much. I
just
22 want to make one small comment about the
remission
209
1 rate, and what it means to talk to a
patient--in
2 light of what Dr. Sekeres said and, in
another
3 respect, what you said.
4 Although a 15 remission rate is
low
5 numerically, in a patient who, faced with
the
6 option of--even if you take Dr. O'Brien's
very
7 well-characterized MDSM data, has an 18 to
20
8 percent chance of dying in the hospital, I
would
9 submit these patients have a much higher
chance of
10 dying in the hospital were they to get
11 chemotherapy.
12 That has clinical meaning to a patient.
13 If you're sitting with them in the room, as
Dr.
14 Sekeres said, it would represent an option
which I
15 believe many would take, and I think it
would be
16
reasonable for them to do so, having treated many
17 people and seeing them die in mucositis and
of
18 sepsis due to chemotherapy, which does not
offer a
19 long survival benefit.
20 I'm not saying it's for everybody,
but
21 there are clearly patients out there who
would take
22 a 15 percent chance. It is a crap shoot, and we
210
1 should learn the science of it. But there are
2 patients out there today who would take such
a
3 chance to go into remission, with the
knowledge
4 that they had a lower chance to leave the
hospital.
5 I talk to them--they go into the hospital,
some of
6 them will not come home. This offers them a higher
7 chance to go home and see their family.
8
DR. MARTINO: I'm not
arguing the point
9 that there's some activity here. I'm not arguing
10 that point.
But I suspect Laetril has a response
11 rate, too.
Okay?
12 DR. STONE: I doubt it's--
13 DR. MARTINO: It's an issue of
relative
14 merit.
And that's the issue before us.
15 DR. STONE: That's for you to
decide.
16 DR. PAZDUR: Let's go back to the
17 regulations--okay?
18 It's a surrogate endpoint
reasonably
19 likely to predict clinical benefit. This response
20 rate that you are seeing, and the data that
you
21 have seen here, do you think that is
reasonably
22 likely to predict clinical benefit?
23 To put it in laymen's terms: this
trial
24 that you see here, you in a year or two
years are
25 going to see a supportive care trial. Is that
211
1 trial--if you were looking at it now, with
the data
2 that you have in hand--that data, does that
say
3 that that trial is going to be reasonably
likely to
4 be positive?
That's the issue here.
5 DR. MARTINO: Dr. Bukowski, I think
you're
6
up next.
7 DR. BUKOWSKI: I don't know that we
can
8 answer that question. Exactly--
9 DR. PAZDUR: That's a clinical
judgment.
10 DR. BUKOWSKI: I understand. That's a
11 clinical judgment question.
12 I think the data are
interesting. Whether
13 you call them exciting or not, I think we
could
14 debate that point. But they certainly are
15 interesting.
You have a subset of individuals for
16 whom there is no therapy--presumably.
They're not
17 eligible for chemotherapy. They won't take it.
18 They will--at least a small percentage of
19 them--have improvement with this agent.
20 Now, I think the science of this
issue is
21 complex, likely because it's pathways, it's
not
22 going to be simple to work out. There are many,
23 many divergent paths that lead from farnesyl
24 transferase.
So it's going to take some time to
25 work that out.
212
1 But, certainly, I'm convinced that
there
2 is some benefit to a subset of individuals
with
3 this particular drug. They do have hematological
4 remission, they do improve.
5 I don't know for certain whether
the
6 entire population will improve. If partial
7 remission, or partial responses at all have
any
8 bearing on this--which they may or may not,
then
9 certainly it's a possibility. I think giving the
10 drug the benefit of the doubt, in my mind,
is
11 something that is worthy of consideration.
12 DR. MARTINO: You can give it the
benefit
13 of the doubt by allowing another study to
take
14 place.
It's a matter of how you choose to give
15 benefit of doubt.
16 DR. BUKOWSKI: Absolutely. But then we're
213
1 ignoring some of the aspects that were
discussed
2 with regard to patients and their needs for
these
3 drugs.
Okay--should we look at that?
Because
4 that's not part of the accelerated
approval. We
5 have to think more in the scientific
vein. But
6 that enters into our deliberations,
obviously.
7 DR. MARTINO: Dr. George.
8 DR. GEORGE: I'd like to address
some
9 issues with respect to accelerated
approval--in
10 particular, what it means when you say that
11 something is "reasonably likely to
predict clinical
12 benefit."
13 So a number of questions have to
be
14 addressed.
First of all, what is the population
15
you're talking about? I believe,
in this case, it
16 would be those that are not "willing or
able"--to
17 use the language that was used earlier--to
receive
18 chemotherapy. And one of the problems that I
19 struggle with a little bit is: "not
willing" is a
20 little different than "not
able." So those are two
21 different things. And I don't know--you can define
22 "not willing"--you just ask. "Not able" is a
214
1 little more amorphous and creates the
2 heterogeneity.
3 But putting that aside somewhat,
let's
4 just assume we can agree on what that
population
5 is, what would it mean to say you have some
6 evidence that is "reasonably likely to
predict
7 clinical benefit?"
8 Now, I assume, Dr. Martino, that
you would
9 think it was exciting if it did predict--if
you did
10 end up proving clinical benefit, that's what
you
11 might be excited about. Not this evidence. So
12 this is just asking the question of whether
it's
13 "reasonably likely" to predict
that.
14 Now, what's going on is a study of
300 or
15 so patients that is designed to pick up a
pretty
16 large survival difference. I didn't--it would be
17 difficult to go through all the different
18 things--scenarios, different modeling you
could do
19 to try to see whether the evidence from this
Phase
20 II trial might indeed predict that kind of
outcome.
21 I'm a little skeptical of that myself, so
I'm a
22 little worried about that: that even though
I think
215
1 I might say, in the abstract, this is
reasonably
2 likely to predict clinical benefit, I would
3 probably say it's not reasonably likely to
predict
4 clinical benefit of that magnitude. And that
5 worries me some, just because of the
practicalities
6 of the size of the patient population, and
the
7 ability to do these studies.
8 But I'm afraid that's a reality
that we
9 might have to live with. That is, we could very
10 well be in the situation, if we did approve
this
11 for accelerated approval, of finding out
later that
12 it didn't work in that trial that was
supposedly
13 the confirmatory trial.
14 And that relates to another issue:
what is
15 the endpoint. I mean, when you say "predict
16 clinical benefit," we've said before,
at the last
17 December, I guess, in a leukemia study that
curable
18 complete response is, in fact, a clinical
benefit
19 in
itself.
20 Here we're talking a little
differently.
21 We're not quite believing that. I mean, you have
22 to be
careful, I guess, because you have clinical
216
1 benefit, and I think people often confuse
this:
2 clinical benefit in some patients. But if you give
3 a therapy, it's pretty clear that some patients
4 benefitted from this. But that's true of a lot of
5 things.
And is that enough to give approval?
And
6 I think the answer has generally been: no,
not by
7 itself.
8
And so, clearly, some patients benefit.
A
9 large number may not benefit and, in fact,
may be
10 harmed from certain therapies.
11 So what is the benefit here? It would
12
have to be overall survival then--I think, in this
13 population.
And the comparison group is best
14 supportive care. So it's exactly the trial you
15 have going on. I'm worried that it's a little
16 small.
17 There may be another clinical
benefit that
18 they're not addressing at all in that
19 trial--because I asked the question
earlier--and
20 that's quality of life; that it's entirely
possible
21 that it does not prolong survival in any
meaningful
22 way, but distinctly improves quality of
life. A
217
1 possibility, but you have to look for it and
2 carefully design those studies.
3 So, I'm just throwing all this out
as
4 information, but I think it's relevant to the
issue
5 of whether this really is--we have a
surrogate
6 "reasonably likely to predict clinical
benefit."
7 MR. FLATAU: You know, actually,
I'm fairly
8 excited about a 15 percent response rate in
these
9 patients with this--you know, with the
10 non-toxicity. But I think it's incorrect to say
11 that there's no care. I mean, I'll these patients,
12 whether they get Zarnestra, or whether they
get
13 high-dose chemotherapy, or whether get
supportive
14 care are going to die from the disease or
from its
15 treatment.
I mean, it's uniformly fatal. And
it's
16
not at all clear to me that survival--the overall
17 survival of patients--is higher in this over
18 standard therapy, which would be best
supportive
19 care.
20 You know, if this was maybe
curative, that
21 would be different. But it's not.
And I don't see
22 any data that compares those. I have no data to
218
1 say whether it's better or not.
2 And I think that the trial that
this--the
3 AML 301 trial has to be done and we have to
see the
4 results of that.
5 DR. MARTINO: Yes?
6 DR. PORTER: Well, I'd like to just
say
7 that I think that there is a strong signal
here,
8 and that it is a surrogate. I think that there is
9 a measure here that is valid.
10 I think that the idea that
patients
11 wouldn't be interested in a one change in
eight, or
12 one chance in 10 of improving is
erroneous. Most
13 patients will accept that if they're
desperate.
14 And, as a matter of fact, it's not just in
15 oncology.
It's true across the board in other
16 diseases, too: when you have only a slim
chance
17
that you can actually benefit,
will you take that
18 chance?
And the answer is yes.
19 And I think that accelerated
20 approval--personally, although I can't
vote--is the
21 right direction for this drug, because I
think that
22 it sends a message to the pharmaceutical
world that
219
1 we will accept data that even only 15
percent of
2 the patients will improve, but we want you
to
3 continue to fight. But we do want to see that
4 second trial.
5 And we do need the FDA to have
more power
6
to pull these drugs back if they don't work. And
7 there's nothing I can do about that.
8 But I think given the variables we
have,
9 accelerated approval is the right direction.
10 DR. MARTINO: Dr. O'Brien.
11 DR. O'BRIEN: I think that it keeps
coming
12 back to the patient population. If I really
13 believed--which I don't--that everybody in
this
14 trial would not respond to chemo, and/or
wouldn't
15 have gotten chemo--so it truly would be a
16 comparator of zero, because they're not
going to
17 get a response, or they're not going to get
18 treated--I would be perfectly happy with 15
19 percent.
20 I don't believe that you can get
that out
21 of this trial. Because, as I just told you, I
22 think there are patients who would have been
220
1 treated with chemo. We saw that there were
2 patients that went on to get treated with chemo
3 and, in fact, six of the 10 achieved a
complete
4 remission.
5 My concern would be: now you have
a
6 pill--which is always attractive to
patients--they
7 don't get chemo up front. So you might want to
8 argue: well, they get a pill if they--why
not take
9 the easy route, ad was asked before.
10 Well, what you have to keep in
mind is
11
that patients with AML who aren't in remission get
12 sick and have a declining performance status
the
13 longer they walk around without getting a
14 remission.
So potentially delaying chemo in
15
someone with an excellent performance status, who
16 might go into complete remission, would be a
very
17 powerful negative, in fact.
18 So, I can't get around the fact
that--the
19
heterogeneity in the patient population, and the
20 fact that, by definition, they all had a
good
21 performance status, I can't put that 15
percent
22 into a perspective that it's 15 percent of
what
221
1 would otherwise be zero.
2 DR. PAZDUR: Could I address that
issue?
3 Because it brings up a very important
regulatory
4 issue that affects both regular approval and
5 accelerated approval.
6 There's two conditions for
approval of an
7 NDA: obviously, the demonstration of safety
and
8 efficacy is one; and then the other one is
that the
9 clinical trial information has to provide
10 sufficient labeling information. In other words,
11 we have to be able to identify a population
that
12 the drug works in.
13 And, here again, that has some
caveats
14 with accelerated approval. And, here again, the
15 second aspect is that this has to be better
than
16 available therapy--okay?
17 So you have to define a
population. Now,
18 in the sponsor's indication, it just says
"in the
19 elderly--"--I forgot exactly what it
was, but a
20 relatively wide indication in elderly
patients.
21 That clearly is
inappropriate. If we
22 would--I said we would define the patient
222
1 population that was studied in the study:
greater
2 than 75, or greater than 65 with MDS. In that
3 population, does that capture what you're
getting,
4 or is that still inadequate? Can we provide
5 appropriate labeling that would prevent
somebody
6 from being given this drug if, truly, they
were a
7 candidate for standard therapy. And I think that
8
is an important question that one needs to answer.
9 Or, is the committee simply
willing to
10 say: well, we'll put in the indication that
the
11 physician should make that determination
that this
12 is appropriate for patients who cannot
tolerate
13 induction therapy.
14 But I think you're hitting on a
very
15 important issue here, and that is: labeling
of an
16
appropriate population--that's required by the law:
17 and, secondly, if you do look at accelerated
18 approval, is that population well enough
19 established here that you can say that this therapy
20 is better than available therapy--either in
terms
21 of efficacy or safety?
22 DR. MARTINO: Dr. Cheson?
23 DR. CHESON: Well, to me, 15
percent in
24 this patient group is important. But you raise--in
25 fact, you've really convinced me about what
to do
223
1 here.
2 The fact is, we don't know what this
3 patient population is. And I think there are three
4 of us at this table--maybe four--do you
treat
5 leukemia?--there are three of us who still
treat
6
leukemia--one, two, three--and the three of us are
7 very uncomfortable with the identification
of this
8 patient population.
9 Susan has data which suggests--and
there
10
were recent papers elsewhere which suggested up to
11 40 percent--there was a paper in Cancer last
12 year--up to 40 percent of patients over the
age of
13 70 or so will respond to seven-and-three kind
of
14 therapy.
15 Alexandra's uncomfortable on the
prior
16 MDS.
I'm also uncomfortable about the accelerated
17 approval process--but be that as it
may. I think
18 that we don't have a good grasp on the
patient
19 population that we--I couldn't draft a
labeling
224
1 indication for this based on what we've
shown.
2 I'm still uncomfortable with Otis'
3 favorite slide. That right column doesn't convince
4 me of anything.
5 If I thought that it was very
clear that
6 there was a big difference in all these
features,
7 then I'd be a little more comfortable with
an
8 accelerated approval. But the fact is: I don't
9 really know what patients are going to
respond,
10 what patients aren't going to respond. It makes it
11 very difficult for me to say: "Okay, we
will
12 approve this accelerated approval for this
group of
13 patients."
14 I just treated a 72-year-old with
AML with
15 standard seven-and-three, and he's out of
the
16 hospital, and he's in remission. You know, if he
17 would have said, "Gee, I don't want chemotherapy,
I
18 want that pill," then he'd have been a
candidate
19 for this study. But instead he may do better with
20 what he got--in the long run, even though
his
21 short-term events were a little more
complicated.
22 I'm just confused as to who I
would say is
225
1 going to respond to this. And, again, it gets back
2 to other drugs and patients who we didn't
know how
3 to predict a response, and we ended up in
trouble.
4 DR. MARTINO: Dr. Levine.
5 DR. LEVINE: Well, to be honest,
this
6 decision is a very difficult one for
me. Because,
7 on the other hand, I agree with Bruce: I
think 15
8 percent CR in a very difficult population is
9 something to approve, basically. That number
10 doesn't blow me away.
11 If I look at the unfavorable
karyotypes,
12 on page 52--so that one seems to me pretty
13 objective.
And if, in fact, we were going to do
14 this on an accelerated basis and say
"greater than
15 the age of 65 with unfavorable
cytogenetics," I
16 think there are some data to indicate 14
percent in
17 their counting.
18 I would never approve this without
the
19 other trial there. There has to be that other
20 trial looking at survival benefit. But I have
21 another question to the company, if I might,
and
22 that is: the lack of quality of life
instrument--I
226
1 can't believe you did that. And the lack of
2 adherence data--because it would have helped
you.
3 I mean, I'm believing that these patients
didn't
4 take the medicines. And I've written articles
5 about this.
I've had grants about this.
Patients
6
don't take the medicines.
7 And all I'm trying to say to you
is: I'll
8 bet you that the responses were not on this
dose.
9 And maybe if it was on the optimal dose
there would
10
have been better responses.
11 So my only point is: on this 301
trial, is
12 there a very careful adherence
instrument? And is
13 there a very careful clinical benefit? Because if
14
there isn't, then we're wasting time again.
15 DR. THIBAULT: On the 301 study,
which is
16 an international study, we are trying to
focus on
17 the quality of life measures that are the
most
18 reliable in the population for a short
period of
19 time.
So we will have detailed transfusion data.
20 We will have very detailed hospitalization
data.
21 We will have very detailed safety
data--because
22 we've used the knowledge from CTEP-20 to
design
227
1 301.
2 The CTEP-20 trial was what it was,
but the
3 301 study will be the study that will bring
this
4 additional information.
5 DR. DeLAP: Yes, if I could add
just one
6 thing: if there are concerns about the
statistical
7 power, this is a trial that has been looked
at
8 extensively, and the patient number and the
design
9 were discussed with FDA. But if it helps to
10 revisit the sample size, obviously we can
still do
11 that and redesign the statistical plan.
12 DR. LEVINE: Are there adherence
data that
13 are going to be--
14 DR. THIBAULT: Yes. May I address, on
15 behalf of CTEP, the compliance monitoring in
this
16 study?
17 All patients on this study had
drug
18 dispensed under supervision from the
research
19 pharmacy at each site. The patients had
20 instructions on how to use the
medication. The
21 medication was packaged in a way to
facilitate the
22 accounting of the pills.
23 The patients also had to bring
pills back
24 at each visit. We do not have the exact records of
25 what happened at the visits, because this
was--but
228
1 from what we saw when we reviewed the data,
these
2 discussions occurred--these pills were
counted.
3 And so this happened--remember
that each
4 patient was reviewed every week by either
the
5 investigator or the health care giver. And we do
6 know which dose they started on. We do know when
7 they were dose-reduced. We do know when there was
8 a delay in treatment.
9 What we do not know is the details
of
10 whether it's 20 pills or 18 pills that were
11 returned at the end of a cycle. But remember,
12 also, that each cycle counted 21 days of
treatment,
13 and there was 63 days in total to get this
done.
14 So, in the end, they had to finish the
pills.
15 So the patients took 21-day
equivalent of
16 tipifarnib in each cycle, unless there was
17 dose-limiting toxicity or an untoward even.
18 DR. LEVINE: Well, just for the--I
mean, if
19 you have data on pill counts, that would
really
229
1 have been nice to have seen it.
2 DR. THIBAULT: We do not have the
data on
3 pill counts, but we do know it was
thoroughly
4 performed by the NCI investigators.
5 DR. LEVINE: But it doesn't help if
we
6 don't have the information.
7 I have another question:
"best supportive
8
care." Is "best
supportive care" in Europe the
9 same as "beset supportive care" in
the U.S.? Will
10 they have growth factor support? Will they--you
11 know, what exactly will be that arm?
12 DR. THIBAULT: The arm of the 301
study is
13 "best supportive care." That includes, of course,
14 transfusion, antibiotics, growth factor
support--if
15 needed.
I remind you that very few needed it for
16 CTEP 20.
And then some patients can also use
17 hydroxyurea.
Hydroxyurea is included in that
18 control arm.
19 DR. LEVINE: And the supportive
care plus
20
the drug in the Zarnestra arm.
21 DR. THIBAULT: Oh, the Zarnestra
arm will
22 be the same supportive care, except for
230
1 hydroxyurea, obviously.
2 DR. LEVINE: Right. Okay, and then I'd
3 just like to make one other little point
which is:
4 if, in fact--I mean, I guess I was impressed
by the
5 fact that six out of the 10 patients who did
go on
6 to chemotherapy in fact had a CR, indicating
two
7 things--indicating that they probably would
have
8 had the CR without the Zarnestra, but also
9 indicating that the Zarnestra pre-treatment
didn't
10 stop those Crs. So there's just something to think
11 about for the future.
12 DR. MARTINO: Dr. Perry.
13 DR. PERRY: I'd like to go back to
the
14 second page of the background document that
we got
15 from the FDA, and reflect on the fact that
the
16 complete remission rate we're talking about
here is
17 11 or 15 percent--to be optimistic. With
18 chemotherapy, it's 30 to 50 percent which,
in my
19 experience, is pretty exaggerated for this
20 population group. My little trick is to take the
21 patient's age from 100, and that gives you
the
22 chances of response rate.
23 There will be exceptions. Dr. Cheson's
24 athlete--marathon runner, or whatever he
was--to
25 qualify for chemotherapy. If you look down at
231
1 treatment-related deaths, 7 percent, greater
than
2 25 percent.
We've talked all about response rates.
3 We haven't talked about treatment-related
deaths,
4 Madam Chairman. And one-month mortality: 12
5 percent versus 30 to 48 percent--to me is a
clear
6
difference in efficacy.
7 Bruce will have the opportunity to
give
8 his patient chemotherapy if he and he or she
9 decide.
If this drug is approved, he'll also have
10 the opportunity to give this drug to
patients who
11 are not candidates for chemotherapy--or, for
12 whatever reason, don't want to do it. You know, if
13 you tell somebody they're going to get
14
chemotherapy, they're going to be in the hospital
15 for six weeks. For a lot of people, that's
16 something they can't tolerate. And I think,
17 whether it's right or whether it's wrong,
that's
18 not our decision, and we can't second-guess
them.
19 So I don't have a difficult with
people
232
1 who are unwilling to take this
chemotherapy. I'm
2 not going to try to micro-manage their
lives. I
3 have a hard enough time with my own.
4 So I think this is a very good
5 alternative--although albeit limited in its
6 efficiency, it's not all that toxic.
7 DR. MARTINO: I think how you
present
8 things to patients has a great influence on
what
9 they choose to do or not to do. You know, we're
10 all sort of pretending like somehow patients
make
11 all the decisions. They certainly make many of the
12 decisions.
But you and I know it's a definite
13
interaction.
14 Dr. Mortimer.
15 DR. MORTIMER: I guess my problem
remains
16 the issue of the primary endpoint being
complete
17 response.
And if we use, again, the numbers by the
18 sponsor, 20 of 136, and if we compare it to
the
19 chemo, 6 of 136 minus how many died in the
first
20 month, I'm not sure there's a
difference. And
21 that's why I have difficulty thinking of it
as
22 accelerated approval.
23 DR. MARTINO: Yes, Doctor?
24 DR. DAGHER: Just for
clarification, since
25 you mentioned this table. And I know everybody
233
1 recognizes, since we've had all these
discussions
2 and speakers who addressed what the elderly
3 population means, just to notice for the
record
4 that: we really had a tough time finding a
5 completely comparable population. So the
6 chemotherapy column in that table--as we say
in the
7 text above that--really reflects those 60
and older
8 who had chemotherapy. So it doesn't address
9 whether or not they had poor performance
status.
10 It's irrespective of whether they had MDS,
11 etcetera.
12 So, just for the record.
13 DR. PERRY: If they had poor
performance
14 status, they wouldn't have been offered
15 chemotherapy. So this is the best--this is the
16 best you can expect. And we're comparing the best
17 to a group that's not as good.
18 DR. MARTINO: Are there any other
comments
19 or discussions related to potential accelerated
234
1 approval for this agent?
2 Yes, Doctor?
3 DR. GEORGE: A question for Dr.
Pazdur,
4
maybe, and the logic of this, in terms of
5 regulatory affairs: if, say, the accelerated
6 approval were not to be granted, and then
this
7 trial would go--in either case, this trial
is going
8
to go on--I assume. I hope it
goes on to
9 completion.
10 If they didn't have accelerated
approval,
11 they couldn't do the marketing or anything
of the
12 agent during this period while we're waiting
for
13 this other trial to mature.
14 DR. PAZDUR: The drug would not be
on the
15 market.
16 DR. GEORGE: Not be on the
market. And so
17 if that trial, then, were positive enough,
then
18 presumably they could come in for full
approval--
19 DR. PAZDUR: Yes.
20 DR. GEORGE: --based on overall survival.
21 Now, if it's negative, it creates
an
22 interesting situation. You'd have--it will be
235
1 negative at the end. Whereas, if we do have
2 accelerated approval now, and the trial is
3 negative--I guess what I'm asking: is there
any
4 change in the agency's position on
withdrawal of
5 agents after--for indications after--
6 DR. PAZDUR: Well, Bruce had
mentioned that
7 in, obviously, we are in active discussion
with
8 several pharmaceutical companies regarding
their
9 accelerated approval. There are areas of
10 confidentiality, obviously, that I cannot
broach at
11 this time.
But we are intensely looking at this.
12 There are issues. We are totally aware of
13
the situation. And let me just
say that this is an
14 evolving issue here that you will see some
15 resolution on. And I'm not saying either way what
16 this will go as. But I think it's important that
17 the agency is interested in looking at these
18 commitments.
This is a process and evolution.
I
19 would not make any decision based on kind of
a
20 tea-leaf reading of what we may and may not
do for
21 any specific drug.
22 DR. GEORGE: Well, the reason I
brought it
236
1 up was it just--
2 DR. PAZDUR: The regulations, as
they are--
3 DR. GEORGE: I understand.
4 DR. PAZDUR: --and assume that we will
5 abide by those regulations.
6
DR. GEORGE: But the conundrum sort of is
7 the effects of the decision now would
potentially
8 have a completely different impact--I mean,
the
9 results of this Phase III trial will have a
10 completely different impact later.
11 I mean, if they don't have
approval now
12 and the Phase III trial turns out to be not
13 positive enough, then they won't have
approval
14 period.
15 DR. PAZDUR: Correct.
16 DR. GEORGE: Whereas, the other
way, they
17 would have approval, and they may not have
it
18 withdrawn, even if that were negative.
[Laughs.] So
19 that's an interesting logical--
20 DR. PAZDUR: But I think you should
look at
21 it like if that provision is on the books
and it
22 can be exercised. Past performance does not
237
1 predict future performance.
2 DR. MARTINO: Can I ask a question
of the
3 company? Let us assume for the moment that you
4 don't get approval of this drug today in the
5 accelerated arena, that strikes me that that
would
6 then leave you open to expand your Phase III
trial
7
into the U.S., which right now you've chosen not to
8 do in this country.
9 Would that be your intent? Would that be
10 your thought?
11 DR. DeLAP: Again, we're already
committed
12 to, I think--what is the total number of
13 center?--90 centers globally, and I think
we'll be
14 able to achieve the sample size that we
currently
15 plan or, again, if we need to revisit the
sample
16 size, even a somewhat expanded sample size
quite
17 efficiently with the centers we have.
18 And again, we're quite confident
that the
19 results of that trial, given the reliability
of
20 complete remission as a surrogate in the
experience
21 in AML up to now, we're quite confident in
the
22 results of that trial.
23 DR. MARTINO: Thank you.
24 Dr. Bukowski?
25 DR. BUKOWSKI: Bob, what's your
estimate of
238
1 the trial duration for accrual? You must have that
2 information.
3 DR. ZUKIWSKI: We anticipate that
the
4 accrual should be finished in the first
quarter of
5 2006, right around that time--right around
the end
6 of the year, first quarter 2006.
7 DR. DeLAP: Operationally speaking,
it
8 would take some time if we decided to expand
it
9 further, with IRBs and everything else. So it's
10 probably sensible for us just to continue
with
11 that.
12 We're certainly intent on
following
13 through on this trial, regardless. But, again, we
14 think that given that it will be about two
years
15 before we're able to present data from that
trial,
16 we think that given the durable complete
remissions
17 and the evidence of benefit and safety profile,
18 that it is something that we'd like to make
19 available for patients at this time.
20 DR. MARTINO: Are there any other
comments?
21 If not, I'll take a vote.
22 And, again, the question is:
accelerated
23 approval for this agent. And again, we'll start
24 with Dr. O'Brien.
25 Please state your name, and your
answer.
239
1 DR. O'BRIEN: O'Brien. No.
2 DR. CHESON: Cheson. No.
3 DR. GEORGE: George.
No.
4 DR. BRAWLEY: Brawley. Yes.
5 DR. MORTIMER: Mortimer. No.
6 MS. HAYLOCK: Haylock. No.
7 MR. FLATAU: Flatau.
No.
8 DR. LEVINE: Levine. Yes.
9 DR. MARTINO: Martino. No.
10 DR. PERRY: Perry. Yes.
11 DR. BUKOWSKI: Bukowski. Yes.
12 DR. MARTINO: And we have a total:
seven
13 no, four yes answers.
14 Rick, do you have any other
questions you
15 want to deal with?
16 DR. PAZDUR: No.
17 DR. MARTINO: That being the case,
this
18 meeting is adjourned. Thank you.
19 [Whereupon, at 2:24 p.m., the
meeting was
20 adjourned.]
21 - - -