Read VIRGINIA: text version

1

FOOD AND DRUG ADMINISTRATION ONCOLOGIC DRUGS ADVISORY COMMITTEE MEETING

AFTERNOON SESSION

Tuesday, September 1, 2009 Silver Spring Hilton Maryland Ballroom 8727 Colesville Road Silver Spring, Maryland

2

ADVISORY COMMITTEE: S. GAIL ECKHARDT, M.D., Professor and Division Head, Medical Oncology, University of Colorado at Denver NICOLE VESELY, PHARM.D., Designated Federal Official, Advisors and Consultants Staff (HFD-21), Center for Drug Evaluation and Research, Food and Drug Administration GREGORY CURT, M.D., U.S. Medical Science Lead, Emerging Products, AstraZeneca Oncology RALPH FREEDMAN, M.D., Ph.D., Clinical Professor, Department of Gynecologic Oncology, The University of Texas MICHAEL LINK, M.D., The Lydia J. Lee Professor of Pediatrics, Chief, Division of Hematology/Oncology, Stanford University School of Medicine GARY LYMAN, M.D., M.P.H, FRCP (Edin), Director, Health Services and Outcomes Research Program-Oncology, Duke University Medical Center VIRGINIA MASON, RN, Executive Director, Inflammatory Breast Cancer Research Foundation RONALD RICHARDSON, M.D., Consultant, Department of Medical Oncology, Mayo Clinic MIKKAEL SEKERES, M.D., M.S., Associate Professor of Medicine, Staff, Cleveland Clinic Taussig Cancer

3

Center, Department of Hematologic Oncology and Blood Disorders MARGARET TEMPERO, M.D., Deputy Director, UCSF Helen Diller Family, Comprehensive Cancer Center, University of California, San Francisco, UCSF Cancer Center WYNDHAM WILSON, M.D., Chief, Lymphoma Therapeutics Section, Metabolism Branch, Center for Cancer Research, National Cancer Institute

TEMPORARY VOTING MEMBERS: LODOVICO BALDUCCI, M.D., Program Leader, Senior Adult Oncology Program, Moffitt Cancer Center JOHN BARRETT, M.D., Section Chief: Stem Cell

Allotransplantation, Hematology Branch, National Heart Lung and Blood Institute, National Institutes of Health ARTHUR FLATAU, Ph.D., Patient Representative THOMAS FLEMING, Ph.D., Professor, Department of Biostatistics, University of Washington, Seattle

ALSO PRESENT: HUANYU CHEN, Ph.D., Statistical Reviewer - Drug Oncology, Division of Biostatistics 5, Office of Biostatistics, Officer of Translational Science, CDER, FDA

4

ALBERT DEISSEROTH, M.D., Medical Officer, DDOP, OODP, OND, CDER, FDA ROBERT KANE, M.D., Medical Officer, DDOP, OODP, OND, CDER, FDA RICHARD PAZDUR, M.D., Director, Office of Oncology Drug Products ROBERT JUSTICE, M.D., Director, Division of Drug Oncology Products ELIHU ESTEY, M.D., Guest Speaker, Member of Fred Hutchinson Cancer Research Center, Seattle, Professor of Medicine, Division of Hematology, University of Washington School of Medicine ALSO PRESENT: TANYA LEWIS, Vice President, Regulatory and Quality, Vion Pharmaceuticals, Inc. BOB LOWENBERG, M.D., Ph.D., Chairman, Department of Hematology, Erasmus University Medical Center ALAN BURNETT, M.D., Professor, Department of Hematology, University of Wales, College of Medicine ANN CAHILL, Vice President, Clinical Affairs, Vion Pharmaceuticals, Inc. FRANCIS GILES, M.D., Chief, Division of Hematology and Oncology, University of Texas, San Antonio GARY SCHILLER, M.D., Professor, Department of

5

Hematology and Oncology, UCLA School of Medicine

6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Products. DR. CHEN:

P R O C E E D I N G S (1:07 p.m.) DR. ECKHARDT: and convene this meeting. All right. I'd like to go ahead

As you know, ODAC is meeting

today to discuss the NDA for laromustine by Vion Pharmaceuticals proposed indication for remission induction therapy for patients 60 years or older with de novo poor risk AML. And what I'd like to do is to start with

introductions beginning on my left. DR. PAZDUR: Drug Products. DR. JUSTICE: Oncology Products. DR. DEISSEROTH: DR. KANE: Al Deisseroth, FDA. Robert Justice, Division of Drug Richard Pazdur, Office of Oncology

Robert Kane, Division of Drug Oncology

Huanyu Chen, statistician, FDA. John Barrett, Hematology Branch,

DR. BARRETT: NHLBI, NIH. DR. BALDUCCI: in Tampa, Florida. DR. TEMPERO:

Lodovico Balducci, H. Lee Moffitt

Margaret Tempero, University of

California, San Francisco. DR. LYMAN: Gary Lyman, Duke University. Ron Richardson, Mayo Clinic,

DR. RICHARDSON:

7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Rochester, Minnesota. MS. VESELY: official, ODAC. DR. ECKHARDT: Colorado. DR. WILSON: NCI. MS. MASON: Virginia Mason, I'm the consumer rep., Wyndham Wilson, medical oncologist, Gail Eckhardt, University of Nicole Vesely, designated federal

and I'm with the Inflammatory Breast Cancer Research Foundation. DR. FLATAU: representative. DR. FLEMING: Washington. DR. FREEDMAN: Cancer Center, Houston. DR. SEKERES: DR. CURT: Mikkael Sekeres, Cleveland Clinic. Ralph Freedman, M.D. Anderson Thomas Fleming, University of Arthur Flatau, the patient

Gregory Curt, industry representative. All right. Thank you. And now

DR. ECKHARDT:

we'll have a conflict of interest statement. MS. VESELY: And before the conflict of interest

statement for topics such as those being discussed at today's meeting there are often a variety of opinions, some of which are quite strongly held. Our goal is that today's

meeting will be a fair and open forum for discussion of

8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

these issues and that individuals can express their views without interruption. Thus, as a gentle reminder,

individuals will be allowed to speak into the record only if recognized by the Chair. meeting. In the spirit of the Federal Advisory Committee Act and the Government in the Sunshine Act, we ask that the advisory committee members take care that their conversations about the topic at hand take place in the open forum of the meeting. We are aware that members of the We look forward to a productive

media are anxious to speak with the FDA about these proceedings. However, FDA will refrain from discussing the

details of this meeting with the media until its conclusion. Also, the committee is reminded to please refrain from discussing the meeting topic during breaks or lunch. you. And now for the conflict of interest statement. The Food and Drug Administration is convening today's meeting of the Oncologic Drugs Advisory Committee under the authority of the Federal Advisory Committee Act of 1972. With the exception of the industry representative, all members and temporary voting members of the committee are special government employees or regular federal employees from other agencies and are subject to federal conflict of interest laws and regulations. Thank

9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

The following information on the status of this committee's compliance with federal ethics and conflict of interest laws covered by, but not limited to those found at 18 U.S.C. Section 208 and Section 712 of the Federal Food, Drug, and Cosmetic Act is being provided to participants in today's meeting and to the public. FDA has determined that members and temporary voting members of this committee are in compliance with federal ethics and conflict of interest laws. Under 18

U.S.C. Section 208, Congress has authorized FDA to grant waivers to special government employees and regular federal employees who have potential financial conflicts when it is determined that the agency's need for a particular individual's services outweighs his or her potential financial conflict of interest. Under Section 712 of the

FD&C Act, Congress has authorized FDA to grant waivers to special government employees and regular federal employees with potential financial conflicts when necessary to afford the committee essential expertise. Related to the discussions of today's meeting, members and temporary voting members of this committee have been screened for potential financial conflicts of interest of their own as well as those imputed to them including those of their spouses or minor children and for purposes of 18 U.S.C. Section 208 their employers. These interests may

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

include investments, consulting, expert witness testimony, contracts, grants, CRADAs, teaching, speaking, writing, patents and royalties, and primary employment. Today's agenda involves discussions of Vion Pharmaceutical's new drug application 022-489, proposed trade name Onrigin, laromustine injection, proposed indication for remission induction therapy for patients 60 years or older with de novo poor-risk acute myeloid leukemia. Laromustine injection, with the proposed trade

name Onrigin, has a proposed use for remission induction therapy for acute myeloid leukemia. This is an initial

approach to AML treatment designed to induce or bring about remission of leukemia in patients 60 years or older with de novo or first occurrence AML designated as poor risk or more likely to have a poor outcome. This topic is a particular

matter involving specific parties. Based on the agenda for today's meeting and all financial interests reported by the committee members and temporary voting members, no conflict of interest waivers have been issued in connection with this meeting. To ensure

transparency, we encourage all standing members and temporary voting members to disclose any public statements that they have made concerning the product at issue. With regard to FDA's guest speakers, the agency has determined that the information to be provided by this

11

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

guest speaker is essential.

The following interest is being

made public to allow the audience to objectively evaluate any presentation or comments made by the speaker. Dr. Elihu Estey has acknowledged that he is a coinvestigator for two studies funded by Genzyme. As a guest

speaker, Dr. Estey will not participate in the committee deliberations, nor will he vote. With respect to FDA's invited industry representative, we would like to disclose that Dr. Gregory Curt is participating in this meeting as a nonvoting industry representative acting on behalf of regulated industry. Dr. Curt's role at this meeting is to represent Dr.

industry in general and not any particular company. Curt is employed by AstraZeneca.

We would like to remind members and temporary voting members that if the discussions involve any other products or firms not already on the agenda for which an FDA participant has a personal or imputed financial interest, the participants need to exclude themselves from such involvement and their exclusion will be noted for the record. FDA encourages all other participants to advise the

committee of any financial relationships that they may have with any firms at issue. And we also just wanted to note that Dr. Estey will be available for questions during part of the afternoon

12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

session.

He will not be speaking as a guest speaker though. DR. ECKHARDT: DR. PAZDUR: Okay. Thank you. Dr. Pazdur.

Good afternoon.

This afternoon's

presentation will review the NDA for laromustine in the proposed indication for remission induction in patients 60 years or older with de novo poor-risk AML. Patients for

induction would be those who have any one of the following poor-risk features: age greater than 70, ECOG performance

status equal to 2, unfavorable cytogenetics, and cardiac, pulmonary or hepatic dysfunction. As noted in the morning comments, ODAC has discussed the value of randomized trials on numerous occasions. Randomized trials allow the use of additional In

endpoints rather than solely relying on response rates. addition to complete response rates, a randomized trial would allow an evaluation of important time-to-event

endpoints including progression-free survival and overall survival. Remember, the endpoint of overall survival is not

only an efficacy endpoint but is a critical endpoint in evaluating safety, a point that I will come back to later. An important caveat regarding randomization is that the randomization process allows us to address prognostic factors that we know about and, more importantly, prognostic factors that we do not know about. Several

meetings have been held over the past eight years between

13

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

the FDA and Vion regarding the development of laromustine. At the end of phase 1 meeting in June of 2004, the FDA expressed concern about Vion's proposal to combine hydroxyurea with laromustine as remission induction in the proposed CLI-033, phase 2, single-arm trial, quote, "Administration of both hydroxyurea and cloretazine or laromustine to all patients enrolled in this noncomparative study will make it difficult to evaluate the contribution of each drug to the treatment effect, especially given Vion's proposal that hydroxyurea could enhance treatment effect by inhibiting DNA repair," unquote. At the January 17, 2006, end-of-phase 2 meeting the FDA requested a randomized comparison of laromustine against standard therapy, quote, "We strongly encourage you to conduct a randomized study which can identify and isolate the safety and efficacy of laromustine. If you choose to

conduct and submit an NDA based on two single-arm studies, the adequacy of the CR rate and duration will be a review issue and will likely require discussion with ODAC." Data from two single-arm studies were submitted in support of the proposed indication. CLI-043 was designed to

test the efficacy and safety of laromustine in patients greater than 60 with AML who were prospectively enrolled to be poor risk. This trial involved laromustine induction and CLI-033 was originally designed

cytarabine consolidation.

14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

as a broad phase 2 trial which studied hydroxyurea given with laromustine as induction for both AML and MDS, previously untreated and relapsed. 55 patients greater than 60 with de novo AML were then retrospectively selected from the original CLI-033 trial and were deemed to be similar in their pre-treatment, poor-risk characteristics to patients on CLI-043 for the NDA efficacy analysis. The FDA review of these trials note that

the remission rate observed in the two single-arm trials is 28 percent in the CLI-043 trial and 29 percent in the CLI033 trial. Of importance is the duration of response is

less than 90 days in 38 percent of the responders in the 04 trial and 31 percent of responders in the 033 trial. The

leukemia-free survival was 174 days in the 043 trial and 111 days in the 033 trial. Overall survival was a median of 98

days in the 043 trial and 103 days in the 033 trial. An important toxicity that will need to be addressed in your deliberations is that of pulmonary toxicity. Pulmonary toxicity was in the top three causes of

all toxicity listings in the single-arm trials and caused 21 percent of treatment deaths in a phase 3 trial that will be discussed subsequently. Both acute and delayed pulmonary toxicities were documented in both the 043 and 037 trials. As previously

mentioned, an additional trial, a randomized trial

15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

designated as CLI-037 deserves attention in your discussions. The trial's objective was to study the effects

of adding laromustine to cytarabine on the response rate, overall survival, and toxicity in patients with relapsed AML who were 18 years or older. The therapy was a two-to-one

randomization between the combination of laromustine with cytarabine versus placebo with cytarabine. 19 percent of the 86 patients who received placebo and cytarabine achieved a complete response or a CRp. Whereas, 35 percent of the patients who achieved cytarabine with laromustine achieved a complete response or CRp. However, the midpoint review led Vion and the FDA to place the trial on hold due to excess mortality in the laromustine arm despite improvement in complete remission rate associated with the addition of laromustine to cytarabine. There was a greater than a threefold increase in adverse events leading to death on the laromustine-containing arm compared to the placebo-containing arm, 12.8 percent versus 40 percent. The causes of death on the laromustine-containing arm were primarily infections and pulmonary. Pulmonary

deaths comprised 21 percent of the deaths on the laromustine arm, whereas no pulmonary deaths were observed on the placebo arm. The list of pulmonary toxicities from this

randomized trial is similar to the listing in the two

16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

single-arm studies described previous. The same dose of laromustine was used in all trials including the randomized trial. The detrimental

effect on survival observed in this randomized trial enrolling a younger patient population of patients with AML, albeit in a more refractory setting and in combination with cytarabine, raises concerns about the use of this drug in an older population of patients proposed in this indication. I would like to remind the committee, as I pointed out in my opening remarks, is that overall survival is not only an efficacy endpoint, but is an important safety endpoint that can only be evaluated adequately in a randomized trial. The single-arm trial submitted with this Hence, we would like

application cannot address this issue.

the committee to discuss the single-arm trial findings in the context of this randomized trial demonstrating a detrimental survival effect despite an improvement in a complete response rate. Additional concerns that we would like ODAC advice and discussion include the following: the results of a CLI-

033 single-arm trial have not isolated the effects of laromustine from hydroxyurea on remission rate in 033. raises the question whether 033 can be considered a controlled trial. The results of the 043 single-arm trial This

have not isolated the effects of laromustine from cytarabine

17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

on leukemia-free survival and overall survival in the 043 trial. The data from the single-arm trials do not define the treatment effect of laromustine in the patient population proposed for the indication. The population of

poor-risk patients greater than 60 years old with AML who may benefit from laromustine is not well defined. Of the

responders, 29 percent of the patients in the 043 trial and 38 percent of the patients in the 033 trial had age greater than 70 years and/or performance status equal to 2 as the only risk factors. These patients may have qualified for Thank you. Thank you. And now

available standard induction therapy. DR. ECKHARDT: All right.

we'll move on to the sponsor presentation. MS. LEWIS: Good afternoon. I am Tanya Lewis,

vice president of regulatory affairs at Vion Pharmaceuticals. We'd like to thank FDA and the members of

the committee for the opportunity to discuss our application with you and the benefit-risk of Onrigin in the treatment of elderly, poor-risk acute myeloid leukemia patients. We also

thank our investigators, patients, and their families who are most impacted by this disease and without whom this work would not have been possible. The data we will present support the use of Onrigin as single-agent remission induction therapy for

18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

patients 60 years or older with de novo poor-risk AML. There are currently no approved therapies and few treatment options in this patient population. Because of this unmet

medical need and because it's not feasible to conduct a phase 3 study in this patient population as we'll explain later in our presentation, Vion submitted this application under the FDA's accelerated approval regulations. Those

regulations allow approval based on a surrogate endpoint that is reasonably likely to predict clinical benefit. The

endpoint selected, complete remission, is an established surrogate endpoint for clinical benefit in patients with AML and has been used in past approvals of AML therapies. In addition to the data we will present today, Vion is committed to the ongoing clinical development of Onrigin for the treatment of AML and are in discussions with FDA regarding the design of a confirmatory phase 3 study. Onrigin was granted fast-track designation and the patient population was defined as patients over 60 years of age with poor-risk AML who have risk factors such as poor performance status, organ dysfunction, unfavorable cytogenetics, or antecedent myelodysplastic syndrome. As

noted in the fast-track designation, this group represents an unmet medical need due to their decreased median survival, higher chemotherapy-related mortality rates, and higher relapse rates in comparison to patients who are

19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

younger with better prognosis. Analysis of SEER-Medicare data underscores this urgent need for new treatments for these patients. In

nearly 3500 newly diagnosed AML patients aged 65 and older, the median survival was 2.4 months. But for those patients

who don't receive treatment for their disease, which is 66 percent of all elderly AML patients, median survival is just 1.7 months. The complexities of addressing this unmet medical need are reflected in the NCCN treatment guidelines where treatment choice is influenced by age, performance status, cytogenetics, and the presence of comorbidities at baseline. As highlighted here in yellow, poor-risk patients are recommended for low-intensity therapy, best supportive care, or clinical trial participation. And this is exactly the

patient population we saw in our trial. As demonstrated by the enrollment numbers in CLI043, the trial was in line with NCCN guidelines again, underscoring that these patients have few treatment options. 96 percent of patients enrolled in the trial would not have been considered for standard induction by NCCN treatment guidelines. It is important to note that this was a very

different patient population than the one described this morning. Patients in our trial had a greater prevalence of These patients were

risk factors, and they were sicker.

20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

prospectively defined at baseline by objective definitions of poor-risk AML features. Definitions that speak directly

to the poor health status of the patients enrolled. Here's our agenda for today's presentation which was designed to review key data and to specifically address concerns raised by the FDA's briefing document. First,

Professor Bob Löwenberg will discuss the treatment challenges associated with AML. discuss AML in the elderly. Professor Alan Burnett will

Ann Cahill will present the Dr. Gary Schiller will

Onrigin efficacy and safety data.

provide a clinical perspective on the benefit-risks for Onrigin. And finally, Ann will return for closing remarks. In addition to our presenters, we have the following external experts with us to answer questions: Dr.

Burton Dickey from M.D. Anderson; Dr. Frank Giles from the University of Texas, San Antonio; and Professor Robert Hills from Cardiff University. Before I turn the podium over to Dr. Löwenberg, I want to emphasize that as a company, we've let the data lead us. We've been diligent in defining a patient population We have We've

that would benefit from treatment with Onrigin.

been thorough in characterizing its safety profile.

developed labeling that clearly describes our patient population and advises physicians on how to use Onrigin safely.

21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

The totality of the data that will be presented today support the accelerated approval of Onrigin. turn to Professor Löwenberg. DR. LÖWENBERG: Good afternoon, thank you for the I now

opportunity to introduce some general principles of treatment of acute myeloid leukemia in 2009 first and then in connection with this focus on some particular aspects of laromustine. I'm professor of hematology at Erasmus

University in Rotterdam. This cartoon depicts the treatment approach to the patient with AML. First, remission induction chemotherapy

is applied, often a 7+3-type of chemo with the intent to induce a complete remission. In all the patients, this is

associated with considerable toxicity, infections, pulmonary toxicity, hematopoietic toxicity. for many patients. But it is the standard

Following this, in those attaining a

complete remission, post-induction consolidation chemotherapy is frequently offered with the purpose to prevent relapse. The post-induction regimen usually differs

from the induction treatment. With this common practice, AML in older patients has remained a devastating disease. As shown, for instance,

by the overall survival data from this reason lodged more than 800 patients phase 3 study in AML patients 60 years or older. There has been no progress in induction treatment.

22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

This study is representative of numerous studies dealing with comparisons of induction treatment. In this case, it

was a comparison of two dose levels of 7+3, again, with no effect on survival. Post-remission treatment may be administered in older patients who attain a complete remission. However, it

should be noted that its therapeutic value in this elderly age group has not been established. So, we often give it,

but we lack evidence whether it provides any therapeutic benefit. AML in older age has a bad outcome in general. So, this is beyond any

There is an abundance of literature.

debate indicating that all the patients have an even worse prognosis when they show unfavorable cytogenetics, increasing age, decreasing performance status, secondary AML, and organ dysfunction. And these poor risk factors are

more common among patients of older age. Thus, we are dealing with a huge problem in our daily reality. The AML incidence increases with age. Most

patients are over 60.

Currently available treatment options

are often withheld from the majority of elderly patients, and this is the significant, unmet clinical need that we deal with when we consider AML at older age including poor results, increased risk factors and the fact that treatment is often not offered to many patients. New agents are

needed to increase the therapeutic options for remission

23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

induction. In general then, two approaches can be pursued to develop these new agents for AML treatment in the elderly population. One approach is to integrate a new agent in This is a common

cytarabine-based backbone of chemotherapy. and realistic approach.

And one study is underway, and I

will very briefly refer to that in younger patients with laromustine. The second option is to substitute the The

cytarabine-based chemotherapy regimen with a new agent.

HOVON laromustine trial is an example of a combination study in younger AML patients between 18 and 65 years of age. This is a large international study that will assess the contribution of laromustine when added to 7+3 followed by post-remission treatment. This study was developed because of the remarkable anti-leukemic activity of laromustine. One of the most

impressive anti-leukemic activities recently seen in the context, we believe, of a reasonable toxicity profile. And

indeed, according to early data from this study with now 60 patients enrolled, it appears that laromustine in combination with 7+3 in younger patients is feasible. This

study will be important to further understand the benefit of laromustine combination treatment and the treatment of AML in general. To conclude, there is a significant unmet need in

24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

all AML patients at every age but especially in the elderly. Laromustine has been studied in the poor-risk older AML patients. These patients exhibit multiple risk factors

predicting for poor outcome in response to 7+3. Laromustine applied in a single dose, one time, has shown meaningful activity and importantly a manageable toxicity profile which makes it an attractive agent to be added to the AML treatment options currently available to older patients. Laromustine represents an important

treatment option for all patients with poor risk factors, a population in need of new remission induction treatments. And the efficacy shown in phase 2 studies should be confirmed in randomized phase 3 trials, hopefully in appropriate patients in combination with standard regimens. Thank you very much. Professor Burnett will now discuss the

unique needs of elderly AML patients. DR. BURNETT: Well, good afternoon, thank you.

I'm going to make a few remarks, much of which you've heard this morning. But with the first slide, just to remind

ourselves that this is a disease of the elderly, and with the demographic changes that we're going to see, particularly in western countries, the prevalence of the problem will increase in the coming decade or two. And so,

we have done poorly in terms of our, the attention we've paid to developing new treatments for a large proportion of

25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

these, in fact, all these elderly patients, next slide. This looks at the history of patients restricted only to patients in the MRC studies for the last 40 years, many hundreds of patients here showing that we are not getting better even if restricted to older, if we used patients restricted to just over 60, it wouldn't look a great deal different. This is a couple of implications. Secondly, this is

First of all, we're not getting better.

restricted to patients who had a number of regimes equivalent to 7+3 therapy and did not cater for the patients who did not or were not considered suitable candidates for whatever reason for 7+3. And I think this sort of experience can be reproduced by any collaborative group. And as Professor

Löwenberg has said, much of the reasoning that these patients do not get exposed to intensive chemotherapy, irrespective of the fact that it's unsatisfactory, is because of the aggregation of patient-related high-risk features and disease-related high-risk features that lead to some doubt as to whether even adequate intensive chemotherapy will do the patient much favor. And from the next slide we can see one of the worrying features here that may feed that perception, and that is the fact that if you just look at age and performance score, as you've already seen this morning, then

26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

there is a reasonable chance for older patients and poorrisk patients to be prematurely terminated or have premature mortality due to an intensive chemotherapy approach. So they show that the data's in early induction mortality is extensively related to age and performance score which are factors that are available to the clinician making a decision as to which approach of treatment they should follow. Several of our investigators over the years

said to us, "Look, you're not providing treatment approaches for the patients that we see in our daily practice." So, in

the next slide and as an attempt to remedy that some years ago, we targeted the population of patients who weren't in the original slide, i.e., the patients who were -- who had the diagnosis, but for various reasons were not subjected to 7+3 approach, this trial was AML-14. And the principal aim of this trial originally was to randomize patients where there was uncertainty to an intensive or a nonintensive approach. And with significant

numbers of patients, we could potentially work out in the randomized context, which factors pointed which patients in which direction. However, in the several hundred patients randomized, only 10 randomizations of this took place. So,

it's difficult to break down the prejudices or the logical reasons that physicians have or patients have about which

27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

treatment approach to carry forward.

But I don't think

there's any doubt that a population of patients exists who are not receiving currently available standard of care. This trial compared low-dose cytarabine to best supportive care, and we found as a result there was a significant benefit in providing low-dose ara-C overall, but this was limited only to the patients who achieved remission which was 18 percent and only to the patients who did not have an adverse karotype. So if a patient did not enter If patients have an

remission, there was no benefit.

adverse karotype, there was no benefit. This, of course, makes the problem of what is a control arm for the unfit or frail patient if one of the options is low-dose ara-C because it clearly is unlikely to work in patients with these two features, next slide. One of the things, however, we did learn was that 18 percent of patients who did enter remission had a respectable duration of remission, as you can see here. And

these patients did not have any further form of treatment. So, the median duration of remission was about 15 months for those who achieved remission. So, the lesson that we learn

is that progress in these patients must be related to an improved ability to achieve remission either because it makes the patient fit for something else, or in itself, like illustrated here, it improves the survival prospects to

28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

something more reasonable. MS. CAHILL: good afternoon.

Thank you.

Thank you, Professor Burnett, and

Onrigin is a rationally-designed alkylater As such, it

from the class of sulfonyl hydrazine drugs.

bears both similarities and improvements over other traditional alkylaters. The chlorethylating species

produced by Onrigin preferentially alkylates at the 06 position of guanine, the cross-link formed as the chlorethyl 06 guanine adduct is one of the most cytotoxic DNA lesions known. Potentially important clinical characteristics

include the drug's ability to cross the blood-brain barrier and the fact that it is not a substrate for multidrug resistance transporters. MDR is commonly overexpressed in

older AML patients and is responsible for the efflux pump implicated in the reduced activity of other AML agents. The clinical development plan for Onrigin and AML simultaneously pursued a dual track in combination regimens and in single-agent studies. Combination studies with ara-C

and other agents were conducted in patients with relapsed or refractory disease while single-agent studies were being conducted in the elderly, poor-risk population. In the setting of relapsed AML and based on the MTD determined in a phase 1 study of the combination of Onrigin and ara-C, a phase 3 randomized study was conducted. I'd like to take a few minutes now to discuss this phase 3

29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

combination trial for relapsed patients in more detail before we move on to the single-agent indication under discussion today. CLI-037 was a randomized phase 3 trial. Unlike

the indication under review today, the target population was relapsed AML patients, and Onrigin was administered in combination with three days of continuous infusion ara-C and compared to ara-C alone. At a planned interim anlaysis by

the DSMB at the midpoint of enrollment an imbalance in deaths was noted leading to suspension of accrual. Subsequent analysis of unblinded safety data conducted by an independent medical team revealed that the early increased mortality was related to the additive toxicities of two agents delivered at high dose, primarily in terms of myelosuppression leading to sepsis and infection, including pneumonia. Study modifications including a dose attenuation of Onrigin were proposed and reviewed by DSMB and FDA. And

FDA subsequently approved resumption of this phase 3 study. To augment awareness, we described the enhanced myelosuppression due to combination therapy in our proposed label. The safety and efficacy of combination therapy with

Onrigin has not yet been established, but trials are underway examining combination regimens. Let me now return to the single-agent development

30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

and the indication under review today.

In the development

of Onrigin as a single agent in AML the phase 1 trial identified 600 milligrams per meter squared as the MTD. dose-limiting toxicity of Onrigin was identified as prolonged myelosuppression. Subsequently, Onrigin was taken The

forward in single-agent development in AML where myelosuppression is a desired treatment effect and the relative lack of extramedullary toxicity might confer benefit over existing therapies. Based on the efficacy signal and tolerability observed in the phase 2 study, CLI-033, the initiation of a phase 3 study in this poor-risk patient population was deliberated with investigators as a next logical step. no acceptable comparator could be found for patients demonstrating these particular poor-risk features. Evidence But

that a randomization of myelosuppressive therapy with a goal of complete remission to low-intensity therapy or best supportive care is not feasible in this population, is best demonstrated by the MRC experience where less than 10 out of over 1400 patients entered such a randomization. In order to move forward and following discussion with the FDA, we looked to design a second phase 2 study to confirm the results of CLI-033 in a poor-risk elderly patient population. Given the subjectivity and prior

applications in elderly AML, Vion objectively and

31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

prospectively identified elderly AML patients with poor-risk disease features for inclusion in this trial. CLI-043 evaluated Onrigin at 600 milligrams per meter squared as a single-dose induction therapy. Patients

with bone marrow improvement but just shy of complete remission could receive a second induction. And patients

achieving complete remission received consolidation therapy with ara-C, 400 milligrams per meter squared for five days. The primary endpoint was overall response rate or ORR defined as complete remission plus complete remission with incomplete platelet count recovery. responses were independently reviewed. Diagnosis and

Patient-inclusion

criteria specified that patients have age greater than 60 and at least one additional risk factor predicting for poor outcome. These risk factors were prospectively identified.

At the time of enrollment, patients were assessed with cytogenetic karyotyping, comprehensive medical history, and scoring of their medical comorbidities. This allowed us to detail the full weight of poorrisk AML factors at baseline for the enrolled patient population. Risk factors assessed were age greater than 70,

poor performance status defined as ECOG performance status of 2, unfavorable cytogenetics, and important criterion identifying poor-risk AML patients, organ dysfunction which was defined at baseline according to the HCTCI comorbidity

32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

index.

The index has shown to be a prognostic indicator for

early death and survival in patients receiving induction therapy in AML. Objective eligibility criteria were designed to enroll a patient group for whom few treatment alternatives exist according to the NCCN treatment guidelines and resulted in the poor-risk patient population treated in this study. Specifically, 78 percent of the patients enrolled 41 percent of the patients had an 47 percent had unfavorable

were 70 years or older.

ECOG performance status of 2. cytogenetics.

In addition, 73 percent of the patients had And 77 percent

cardiac dysfunction at baseline.

demonstrated pulmonary comorbidity indicating the poor general health of these patients at baseline. In total, 96 percent of the 85 patients in CLI-043 had at least two risk factors. or more risk factors. 39 percent actually had four

This constellation of risk factors

may provide some insight as to why these patients were not deemed candidates for treatment with 7+3 by their physicians at presentation. And in this population of elderly AML

patients with poor-risk factors, a meaningful number were able to achieve complete remission. The independently

confirmed rate of overall response in CLI-043 was 32 percent. 85 percent of these remissions were achieved These

following a single, 60 minute infusion of Onrigin.

33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

important remission results differ from those presented by FDA because they include independently verified samples that were requested by FDA and Vion but submitted after the FDA cut-off date. Leukemia-free survival is an indicator of the durability of response. is six months. The median leukemia-free survival

The Kaplan-Meier estimate for leukemia-free

survival is 53 percent at six months, and 28 percent at 12 months. The current AML treatment paradigm in younger

patients is to administer consolidation therapy as soon as possible following a complete remission in order to prevent relapse. Consolidation and induction are not separated in

the determination of leukemia-free survival in leukemia trials. The data shown supports that for most of the

elderly, poor-risk patients treated with Onrigin the remissions achieved were durable and not merely transient. Similar results are apparent when looking at overall survival. The median overall survival for patients

achieving a complete remission with Onrigin is 12.4 months. These overall survival results highlight the significance of complete remission as it is associated with a meaningful survival in the majority of Onrigin responders. These

results are especially meaningful when considering the 1.7 month survival among elderly AML patients not receiving induction therapy.

34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Turning to review of the Onrigin safety profile, a total of 818 patients have been exposed to Onrigin at any dose. The studies of Onrigin at 600 milligrams per meter

squared provide a safety dataset of 277 patients with hematologic malignancies primarily AML. a median age of 71 years. These patients have

The safety profile is typical for

a myelosuppressive agent with expected complications of leukopenia, anemia, and thrombocytopenia. In addition,

there is a common and well-characterized infusion reaction of patients which is mild, transient and easily managed. Finally, consistent with approved alkylaters with a similar mechanism of action, Onrigin appears to cause a pattern of delayed, noninfectious pneumonitis in approximately 9 percent of patients. As with all cytotoxic

agents used in the treatment of leukemia, myelosuppression is both a measure of toxicity and required for efficacy. Both measured adverse events and abnormal laboratory measures demonstrate that myelosuppression and its complications are experienced by nearly all leukemia patients treated with Onrigin. The severity of these events

are important indicators of bone marrow suppression and underscore the need of aggressive supportive care during the period of recovery following induction treatment. Neutrophil and platelet nators typically occur around day 15 with recovery around day 30. All pulmonary

35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

events and grades 3 through 5 events related to pulmonary function are shown on this slide. Pulmonary adverse events

are commonly recorded -- reported in elderly patients undergoing myelosuppressive treatment for leukemia, particularly in the setting of pneumonia and sepsis and may be accompanied by symptoms such as dyspnea, pleural effusion, and hypoxia at all grades. Grades 3 to 5

pulmonary adverse events reported ranged from 1 to 11 percent as shown on this slide. We also examined the incidents of pulmonary adverse events in AML patients who presented with underlying pulmonary comorbidities and who were treated on two Onrigin phase II trials. Not unexpectedly, the reported incidents

of pulmonary adverse events of all grades is increased with the exception of pleural effusion. Whether this reflects a

drug effect or exacerbation of baseline disease is difficult to separate. Furthermore, many of these symptoms such as

dyspnea, cough, hypoxia, and pulmonary edema frequently coexist with leukemia or pneumonia during the period of post-induction myelosuppression. As with other alkylating agents, the occurrence of a delayed noninfectious pneumonitis related to Onrigin has been reported. After the initial case report of pneumonitis

and following the advice from an advisory panel, a careful case-by-case review from 277 patients was conducted for the

36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

manifestation of a delayed, noninfectious pneumonitis.

This

review resulted in the identification of 25 patients or 9 percent who demonstrated a delayed pattern of ground glass, bilateral infiltrates by chest X-ray and clinical symptoms of dyspnea, cough, and hypoxia. Approximately half of these To

patients received steroids and half of those responded. augment awareness, this pneumonitis is described in the proposed Onrigin label.

Patients who were treated with Onrigin for their AML had a risk of induction mortality that is comparable to that observed in other administered induction regimens. 14 percent rate of all-cause 30 day mortality is an important finding given the age and the poor-risk status of these patients. Causes of death that occur during the The

induction period are shown in this table and reflect the impact of myelosuppressive therapy on the bone marrow and in the setting of persistent leukemia. The induction mortality

is similar when looking at either the total 277 patient safety set or patients with comorbidities and other risk factors enrolled to CLI-043. Now that I've walked you through the efficacy and safety data, let me summarize a few key points. The safety

profile is primarily characterized by myelosuppression. There is limited extramedullary toxicity and a mortality rate that is consistent with other agents used as induction

37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

treatment for AML.

The positive benefit-risk profile is We know that

generated by the 32 percent response rate.

complete remission is requisite for survival and many elderly AML patients responding to Onrigin demonstrate this survival benefit. I'd now like to ask Dr. Schiller to present his clinical perspective. DR. SCHILLER: Good afternoon, and thank you for

the opportunity to speak to you and especially to speak to a former patient on the panel. I will now discuss how the

data presented for laromustine directly relate to the unmet needs of a distinct, elderly poor-risk population of patients with acute myeloid leukemia represented in my clinical practice. The challenges of managing AML in the

elderly are that this population presents with both distinct patient-related characteristics and distinct disease-related characteristics which make the disease difficult to manage. In AML patients over the age of 60 there is a high frequency of comorbid medical conditions and concomitant organ dysfunction. Independent of the patient

characteristics are disease-related features that make acute myeloid leukemia in the elderly different. Among elderly

patients, AML is more frequently associated with drugresistance efflux mechanisms, adverse karotype, and adverse molecular features even in the setting of normal karotype.

38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Both directly influence patient outcomes and both influence the selection of therapy. Yet, when confronting this unmet

need of managing adverse pre-treatment characteristics and adverse disease biology, few agents are available. Here we see a timeline of agents approved for the treatment of acute myeloid leukemia. date back to the 1950s. The anthracyclines

Cytarabine was approved in 1969.

And for wont of anything better, continuous infusion over seven days with cytarabine plus three doses of an anthracycline has become the mainstay of therapy for all patients with AML regardless of patient-related factors, adverse karotype, and biologic subtypes. Although, high-dose cytarabine became popular in the late 1980s, it is not available for management of elderly people due to a high-risk of treatment-related toxicities, specifically, cerebellar toxicity, but also hepatic toxicity. Gemtuzumab was approved in 2000 for a Its

small subset of elderly AML patients at first relapse. use as an alternative induction therapy has not been established. If we look at the type of patients who were

treated with laromustine, we see that they represent the poor prognosis group of elderly patients with acute myeloid leukemia, most are over age 70. Almost half have poor

performance status due to a variety of underlying, comorbid

39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

medical conditions.

Nearly half of the patients had disease

characterized by unfavorable cytogenetics identifying a high-risk disease biology of AML in an elderly population. Most had cardiac disease or pulmonary dysfunction. These represent patients with adverse patient and diseasespecific characteristics. For many of these features,

conventional cytarabine and anthracycline induction would not be considered. Unlike conventional chemotherapy which

has been poorly studied in this population and requires seven days of continuous cytarabine infusion and three doses of a potentially cardio toxic anthracycline. With

laromustine patients were treated with a single dose administered over 60 minutes. We administered laromustine

in both the in-patient and out-patient settings given the single-dose, single-treatment schedule. With laromustine remission was achieved in a variety of older AML patients with adverse patient-related features such as advanced age, poor performance status, cardiac and pulmonary dysfunction, and adverse disease features. Among patients with multiple risk factors which

constituted the majority of laromustine patients presented this afternoon, remissions were achieved across all risk factors, patient-related as well as leukemia-related. Thus,

in poor-risk patients not typically treated with cytarabine and anthracycline, laromustine provided broad response.

40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Looking at the actuarial survival curve achieved on the laromustine trial, credible long-term survival is demonstrated. Following remission, a variety of postThe laromustine study did

remission strategies were used.

include a recommended post-remission consolidation regimen. But basically, once remission is achieved, there is no established, consolidation regimen for the management of elderly patients. is 35 percent. The six-month survival for all patients

And the 12-month survival is 20 percent.

And for those who achieved remission with a single dose of laromustine, 52 percent of patients were alive at one year demonstrating the efficacy of this treatment. This slide presents one of my patients who was treated with laromustine in October 2008. consent to present his history. He has given his

Ten months after remission,

Mr. M, I will call him you'll meet him shortly, is with us today at this meeting. was 68. At the time of presentation, Mr. M

Typical of many older patients with AML, he had an

ECOG performance status of 1 and several accompanying risk factors including cardiac comorbidity, pulmonary comorbidity, and leukemia biology that was characterized by complex karotype. His cardiac comorbidity consisted of a history of coronary artery disease requiring stenting, a history of myocardial infarction, and pulmonary comorbidity consisted

41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

of pericardial effusions and worsening dyspnea. presented with leukopenia.

He

And his bone marrow was He received a single On day 29

extensively infiltrated with blasts.

dose of laromustine infused over one hour.

cytogenetic and morphologic complete remission were confirmed. Mr. M received consolidation chemotherapy with And he continues to be free

cytarabine on day 60 and 111. of leukemia today.

As my concluding thoughts to this advisory committee, I think that the data are compelling because they show that a single drug, given as a single dose can produce remission in patients with both adverse patient-related and adverse disease-related variables. Many of the patients

that were treated on this trial would not easily have been treated with conventional infusional cytarabine and three days of anthracycline, particularly those with underlying cardiac disease for whom treatment with an anthracycline is inconceivable. The thought of randomizing such patients to

a trial which included an anthracycline would be unacceptable. The major point is that laromustine achieved remission rates in the same range as the SWOG study with continuous infusion cytarabine and three days of anthracycline. However, with a single dose given once over

one hour in a population not typically treated with

42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

cytarabine and anthracycline, unlike our experience with continuous infusion cytarabine and anthracycline, the toxicity profile of laromustine was well-defined in this poor-risk patient population. term survival were achieved. MS. CAHILL: And both remission and longThank you. The data

Thank you, Dr. Schiller.

presented today demonstrate the ability of Onrigin to induce complete remissions in elderly AML patients who by virtue of very specific features of leukemia biology and medical history have a poor prognosis. This is a different patient

population than has been previously discussed today. The constellation of their risk factors may account for some of the reasons that two-thirds of all elderly AML patients receive no treatment for their disease. These can be captured in labeling discussion. Once the

decision has been made that an older AML patient is not a candidate for 7+3, there are virtually no treatment options for induction of a complete remission. The NCCN treatment guidelines for these AML patients identify clinical trial participation, lowintensity therapy, or best supportive care as the only treatment options for patients included in the proposed indication for Onrigin. Complete remissions are not And

achievable with best supportive care in any patient.

complete remissions are not achievable for patients with

43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

unfavorable cytogenetics.

96 percent of patients enrolled

to CLI-043 would not have been considered for standard induction according to the NCCN algorithm. In closing, the presentation today addressed the concerns raised by FDA. Vion is committed to phase III

trial execution in the appropriate AML patient population. This is evidenced by the ongoing phase III trial with HOVON and our commitment to reinitiate the phase III trial and first relapse AML. The data demonstrate the efficacy and

safety of Onrigin in this untreated population providing a positive benefit-risk profile with single agent induction treatment. Patients with this picture of poor health status and other risk factors have not previously been studied to this degree nor identified. But we have objectively and

prospectively selected for a poor-risk patient population with multiple risk factors which represent an unmet medical need consistent with the FDA fast-track designation for Onrigin in 2005. The complete remission rate of 32 percent and the median overall survival of 12.4 months for responding patients is of sufficient magnitude, clinically meaningful and acceptable as a surrogate likely to predict clinical benefit. As such, Onrigin is an important therapeutic

option for the treatment of elderly patients with poor-risk

44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

disease.

Thank you so much for your attention. DR. ECKHARDT: All right. Thank you. We're going

to take a couple of minutes to change out cable before the FDA presentation. (Recess) DR. ECKHARDT: continue. Okay. Take your seats. We're going to

Next up is the FDA presentation. Good afternoon. My name is

DR. DEISSEROTH: Albert Deisseroth.

I am a medical officer at the FDA who The results

will report to you on a review of NDA-22489.

which I will discuss are based on Vion's submissions and datasets received through June 30, 2009. The members of the

FDA team which reviewed NDA-22489 are listed on this slide. My presentation will cover the following topics: proposed indication, prior therapy approvals, regulatory history, phase II single-arm studies, phase III randomized trial, pulmonary toxicity, and risk-benefit analysis. will now discuss the proposed indication. Laromustine is proposed for the remission induction in patients 60 years or older with de novo poorrisk AML. Patients for the indication would have any one of age greater than 70; ECOG I

the following poor-risk features:

performance status 2; unfavorable cytogenetics; and cardiac, pulmonary, or hepatic dysfunction. Some of these poor-risk

features such as organ dysfunction and unfavorable

45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

cytogenetics may lead to a decision not to use standard intensive induction therapy whereas age greater than 70 and PS 2 are less uniformly utilized. The median age of onset of AML is 69 years. Less

than one-third of elderly patients with AML receive therapy in the United States. AML in the elderly is less responsive

to therapy, and finally, therapy-outcome in elderly patients with AML depends on pre-treatment patient characteristics, therapy characteristics, and disease characteristics including cytogenetics and molecular signatures. I will now discuss previous FDA decisions regarding approval of drugs used in AML induction regimens. Standard intensive induction therapy for AML is composed of three days of anthracycline combined with seven days of cytarabine. Since 1975 FDA approvals for drugs used in this

induction program for previously untreated de novo AML which is known as 3+7 have included randomized trials. In the

year 2000, the FDA approval for Mylotarg for patients 60 years or older with relapsed AML was based on three singlearm trials. In May of 2005 ODAC voted against approval of an application based on a single-arm trial for tipifarnib for previously untreated elderly patients with de novo or secondary AML. Important issues were the low response rate

and toxicity, the fact that many of the responders were

46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

eligible for and did well on subsequent intensive induction therapy, and that there was no randomized comparison of tipifarnib with standard induction in the population proposed for the indication. Next I will discuss regulatory history. Several

meetings were held over the past eight years between the FDA and Vion about laromustine. Notably, no agreement was

reached for a randomized confirmatory trial, nor accelerated approval. At the end of phase I meeting in June of 2004,

the FDA expressed concern about Vion's proposal to combine hydroxyurea with laromustine. Administration of both

hydroxyurea and cloretazine or laromustine to all patients enrolled in this noncomparative study will make it difficult to evaluate the contribution of each drug to the treatment effect especially given Vion's proposal that hydroxyurea could enhance treatment effect by inhibiting DNA repair. At the January 17, 2006, end-of-phase II meeting the FDA requested a randomized comparison of laromustine against standard therapy. We strongly encourage you to

conduct a randomized study which can identify and isolate the safety and efficacy of cloretazine or laromustine. If

you choose to conduct and submit an NDA based on two singlearm studies, the adequacy of the CR rate and duration will be a review issue and will likely require discussion with ODAC.

47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

I will now discuss the two single-arm, phase II studies submitted by Vion to support the proposed indication of laromustine and poor-risk AML. CLI-043 tested the

efficacy and safety of laromustine and cytarabine in patients greater than 60 with AML who were prospectively enrolled to be poor risk. CLI-033 was originally a broad-

based phase II trial of hydroxyurea and laromustine as induction therapy for both AML and MDS. Later 55 patients

greater than 60 years of age with de novo AML were then retrospectively selected from the original trial who were deemed to be similar to the poor-risk patients on CLI-043. I will now discuss efficacy results of CLI-043. CLI-043 was designed to address the applicant's indication population of patients with poor-risk AML. criteria were: The eligibility

patients 60 years or greater with de novo

AML; no history of chemotherapy; no antecedent hematological disorders; and poor risk due to any one of the following pre-treatment characteristics age greater than 70, PS equal to 2, unfavorable cytogenetics, and cardiac, lung, or liver dysfunction. Greater than 70 percent of the patients had cardiac or pulmonary pre-treatment comorbidities and were 70 years or older. Less than half of the patients had PS equal Four percent of the

to 2 or unfavorable cytogenetics.

patients had just one poor-risk factor, while 21 percent of

48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

the patients had two poor-risk factors, and more than 30 percent of the patients had either three or four poor-risk factors. The design of the therapy is shown on this slide. The first induction treatment was laromustine at 600 milligrams per meter squared. A second laromustine

induction was given to 14 patients of the total population. Cytarabine at 400 milligrams per meter squared daily for five days was given as the first consolidation to 18 patients in the total population. was given to four of the patients. The primary efficacy endpoint was a percent of patients with CR or CRp as documented by independent review. Complete remission, CR, was defined as bone marrow blasts less than 5 percent, ANC greater than 1,000 and platelets greater than 100,000 per cubic millimeter. CRp was defined A second consolidation

as blasts less than 5 percent, ANC greater than 1,000 per cubic millimeter, and platelets greater than 20,000 per cubic millimeter without platelet transfusions. to CR and CRp as response or remission. Secondary endpoints included leukemia-free survival, overall survival, and toxicity. Leukemia-free I may refer

survival, which is the time from the onset of remission to death or relapse for the responders only, is used by Vion to estimate remission duration. 24 patients or 28 percent

49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

achieved a complete response or a CRp by independent review and 32 percent by investigator review. 29 percent of the responders by independent review and 33 percent by investigator review had age greater than 70 and/or performance status 2 as the only poor-risk features. Therefore, some of these patients may have been

candidates for intensive induction therapy. The data on this slide shows that some of these remissions were brief. 38 percent of the remissions by

independent review and 41 percent by investigator's review were less than 90 days. The remissions of less than 90 days Remissions ranged from 1 to 581

are listed on this slide. days.

And the median leukemia-free survival calculated by

Vion was 174 days. The treatment given to responders is listed on this slide. 16 of the 24 responders by independent review

received one cycle of cytarabine after laromustine induction. This makes it difficult to judge the relative

contributions of laromustine versus cytarabine to duration of remissions. The median overall survival for all patients Survival

treated on the trial was 98 days or three months.

may also be a result of both laromustine and cytarabine. Our efficacy conclusions are listed on this slide. 28 percent of the patients achieved a CR or CRp by independent review. 38 percent of the remissions by

50

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

independent review and 41 percent by investigator review were less than 90 days. 29 percent of the responders by

independent review and 33 percent of the responders by investigator review may have been eligible for standard induction therapy. Since the majority of the responders

also received cytarabine, the contribution of laromustine to remission duration and overall survival is not defined by this trial. I will now discuss safety data for CLI-043. 15

percent of patients died by day 30 after the first induction and 22 percent by day 42. Pulmonary toxicity was in the top

three causes of early deaths and grade 5 adverse events. Pulmonary toxicities were also among the three leading causes for grade 3 or grade 4 adverse events and SAEs. This

table taken directly from the Vion study report itemizes individual toxicities in CLI-043 for infections, cytopenias, and pulmonary toxicities. Note that MedDRA includes

pneumonias under the "infections" category in this report. Our safety conclusions for CLI-043 are pulmonary toxicity was in the top three causes listed for early deaths, grade 3 to 5 adverse events and SAEs. And these

pulmonary toxicities included acute respiratory failure, ARDS, pulmonary effusions, dyspnea, hypoxia, and pulmonary edema. I will now discuss the efficacy results of CLI-

51

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

033.

CLI-033 is another phase II single-arm study provided This trial studied the

to support the proposed indication.

effect of hydroxyurea combined with laromustine on response rate leukemia-free survival, overall survival, and toxicity. CLI-033 was originally a broad-based phase II trial with the following eligibility criteria: age greater than 18 for

relapsed AML or MDS, age greater than 60 for untreated de novo AML or MDS, performance status 0 to 2, and no active cardiac diseases. The design of therapy is listed on this slide. The first induction was hydroxyurea at 30 milligrams per kilo orally every 12 hours for six doses and one dose of laromustine at 600 milligrams per meter squared. Patients

could receive a maximum of two induction cycles of hydroxyurea with laromustine at 600 milligrams per meter squared plus one consolidation therapy of laromustine alone at 400 milligrams per meter squared. The first problem with CLI-033 is that hydroxyurea was used with laromustine in 15 of the 16 responders defined by independent review. Thus, the protocol does not isolate

the effect of laromustine on remission rate or on other treatment outcomes. Vion has stated that although short-

term administration of hydroxyurea alone is unlikely to induce remissions, its effects on DNA synthesis are predicted to inhibit DNA repair and enhance anti-tumor

52

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

activity of laromustine. 67 percent of the responders received hydroxyurea even though their initial white cell count was less than 10,000. Following completion of CLI-033, Vion attempted the

retrospective selection of patients for the NDA efficacy analysis who were similar to those patients on CLI-043 in terms of their poor-risk, pre-treatment features. The steps

involved are listed on this slide: reclassification from the FAB to the WHO criteria, identification of patients with de novo AML by WHO criteria, and selection of patients with any one of the following poor-risk features: age greater than

70; PS equal to 2; unfavorable cytogenetics; heart, lung, or liver comorbidities. The second problem with CLI-033 is that the patients selected retrospectively from CLI-033 to be similar to the patients in CLI-043 in terms of pre-treatment, poorrisk features, turned out to be lower risk than those on CLI-043. 42 percent of the patients selected from CLI-033

had pre-treatment cardiac comorbidities, whereas 73 percent had cardiac comorbidities in CLI-043. 29 percent of

patients selected from CLI-033 had pre-treatment pulmonary comorbidities, whereas 77 percent had pulmonary comorbidities on CLI-043. The percentage of responders by independent review was 29 percent and 44 percent by investigator review. A

53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

third problem with CLI-043 is that many responders were eligible for standard induction. 29 percent of the patients 38 percent of the

in CLI-033 had just one poor-risk factor.

responders by independent review and 50 percent of the responders by investigator review had age greater than 70 and/or PS 2 as their only poor-risk features. As mentioned

multiple times earlier today, the presence of either one of these two pre-treatment features, age greater than 70 or PS 2, does not necessarily exclude patients from intensive induction therapy. The fourth problem with CLI-033 is that many of the remissions were less than 90 days. 31 percent by

independent review and 38 percent by investigator review. The remissions less than 90 days are displayed on this slide. Remissions ranged from 11 to 981 days. And the

median leukemia-free survival as calculated by Vion was 111 days. The median overall survival for all patients treated

on CLI-033 was 103 days or 3.5 months. Our efficacy conclusions are listed on this slide. Only one of the 16 responders received laromustine alone. Thus, the results do not isolate the effect of laromustine alone on the treatment outcomes. 38 percent of the

responders by independent review and 50 percent by investigator review may have been eligible for standard induction therapy. 31 percent of remissions by independent

54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

review and 38 by investigator review were less than 90 days. Patients on CLI-033 had a lower-risk profile than those on CLI-043. I will now discuss the integrated safety analysis. The patients selected for the safety analysis were 60 years or above, had either AML or MDS and all had received at least one intravenous treatment with laromustine at the 600milligrams-per-meter-squared dose. 15 percent of patients

died within 30 days after the first laromustine induction. And 25 percent died by 42 days. Pulmonary toxicities were

in the top three causes of early deaths by 42 days and of grade 5 treatment-emergent adverse events. Pulmonary

toxicities were in the top three causes of grade 3 or grade 4 treatment-emergent adverse events and SAEs. This table is taken directly from the NDA. itemizes individual conditions leading to pulmonary toxicities which were in the top three causes of SAEs. The It

pulmonary toxicities included dyspnea, respiratory failure, effusions, hypoxia, ARDS, pulmonary edema, and pulmonary hemorrhage. Our safety conclusions are listed on this slide. Pulmonary toxicity along with infections in leukemia were the top three causes of early deaths grade 3 to 5 treatmentemergent adverse events and SAEs. And the pulmonary

toxicities which you have seen on the previous slide,

55

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

included respiratory failure, ARDS, pulmonary edema, and pulmonary hemorrhage. I will now move to a discussion of the phase III randomized trial, CLI-037. This is a phase III perspective,

randomized, double-blind placebo-controlled trial of laromustine with cytarabine versus placebo with cytarabine. The goal of this trial was to study the effect of adding laromustine to cytarabine on the response rate, overall survival, and toxicity in patients with relapsed AML who were 18 years or older. This involved a two-to-one

randomization between laromustine with cytarabine versus placebo with cytarabine. Induction therapy consisted of three days of cytarabine at 1.5 grams per meter squared per day along with a single dose of laromustine at 600 milligrams per meter squared on day 2 randomized against the same three-day schedule of cytarabine with a placebo on day 2. 268

patients were entered, 263 were treated, 177 on the combination of cytarabine plus laromustine, and 86 on the cytarabine-plus-placebo arm. The addition of laromustine to

cytarabine increased the response rate from 19 percent on the placebo arm to 35 percent on the laromustine arm. However, the midpoint review led Vion and the FDA to place the trial on hold due to excess mortality in the laromustine arm as shown on the next slide.

56

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 slide.

This slide provided by Vion depicts the overall survival with laromustine and cytarabine, the bottom curve, and the placebo and cytarabine, the top curve. It appears

as if the excess deaths with laromustine have offset the effect of the increased remission rates on overall survival. The deaths on the two arms of the trial are presented on this slide. There was a greater than three-

fold increase in deaths from 13 percent on the placebo arm to 40 percent on the laromustine arm. These deaths were This

primarily due to infections and pulmonary toxicity. table is taken directly from the NDA.

It shows that 21

percent of the treatment deaths on the laromustine arm were associated with pulmonary toxicities, whereas notably, there were no pulmonary treatment deaths on the placebo arm in the absence of laromustine. The pulmonary toxicities leading to treatment deaths on the laromustine arm, included hypoxia, ARDS, respiratory failure, respiratory distress, and pulmonary alveolar hemorrhage. These pulmonary toxicities from CLI-

037 are similar to those seen on the CLI-033 and 043 singlearm trials in which laromustine was given at the same dose as in the CLI-037 trial. The conclusions from CLI-037 are listed on this The addition of laromustine to cytarabine produced

an unacceptable increase in pulmonary and infectious deaths.

57

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

The pulmonary toxicities seen in CLI-037 were similar to those in the two single-arm trials. The inferior results in

CLI-037 with a younger population raised concern about results that could occur in the older population proposed by Vion for the indication. this concern. A single-arm study cannot resolve

A randomized trial is necessary.

I will now discuss the report and findings of the pulmonary advisory group in impaneled by Vion in July of 2008 to review the pulmonary toxicity of laromustine. Vion

convened an expert pulmonary advisory panel which met in July of 2008 to review memoranda from medical monitors of trial CLI-043, the single-arm phase II trial we have discussed, and the phase III trial, CLI-037, which randomized laromustine versus the placebo. These experts

concluded that laromustine resembles BCNU in causing direct toxicity to the lung. The expert panel reported the

following cases of both early and delayed pulmonary toxicity associated with the administration of laromustine in protocol CLI-043 and CLI-037. Acute pulmonary infusion

reactions, one patient; acute respiratory failure and ARDS with bilateral infiltrates hours to 30 days after laromustine, 35 patients; subacute symptomatic infiltrates greater than 30 days after laromustine, six patients; delayed symptomatic infiltrates greater than 60 days after laromustine, four patients; and noncardiogenic pulmonary

58

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

edema, five patients. I will now proceed to a discussion of risk-benefit analysis of the use of laromustine in patients greater than 60 years of age with AML. The remission rate is 28 percent

in CLI-043 and 29 percent in CLI-033 by independent review. 29 percent of the responders in CLI-043 and 38 percent of the responders in CLI-033 had age greater than 70 and/or performance status 2 as the only pre-treatment, poor-risk factors. These patients may therefore have qualified for

standard induction therapy. The remission duration is less than 90 days in 38 percent of the responding patients in CLI-043 and 31 percent in CLI-033. The median leukemia-free survival was reported Overall survival

as 174 days in 043 and 111 days in 033.

was a median of 98 days in 043 and 103 days in 033. Pulmonary toxicity was in the top three causes of all toxicity listings in single-arm trials and caused 21 percent of treatment deaths on the laromustine arm of the phase III trial. Both early and delayed pulmonary toxicities were Infections and pulmonary adverse events

documented.

accounted for 90 percent of the deaths in the laromustine arm of the randomized study. Whether the population of poor-risk patients greater than 60 with de novo AML proposed by Vion for the laromustine indication will be at higher risk of toxicity on

59

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

laromustine is not known at this time.

This question can Our

only be answered by a randomized phase III trial. concerns are summarized on this slide.

The results of CLI-

043 have not isolated the effects of laromustine from cytarabine on remission duration, leukemia-free survival, and overall survival. The results of the CLI-033 trial have

not isolated the effects of laromustine from hydroxyurea on remission rate. The data from the single-arm trials do not define the treatment effect of laromustine in the patient population proposed for the indication. The population of

poor-risk patients age greater than 60 with AML who may benefit from laromustine is not well defined. In light of

the trial results from CLI-037, to determine whether laromustine would result in a favorable risk-benefit ratio in the population of patients proposed by Vion for the laromustine indication, will require a randomized, controlled trial. question. A single-arm study cannot resolve this

This completes the presentation of the FDA

analysis of NDA-22489. DR. ECKHARDT: now until 3:00 p.m. (Recess) DR. ECKHARDT: Question part of this session, Thank you. We will take a break

questions can go to either the FDA or the sponsor, and I

60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

believe that Dr. Estey is here as well to take any questions. Oh, nevermind regarding Dr. Estey. So, I would

like to start out with a question to the sponsor and it may be buried somewhere in the data that we have, but I was curious to hear a little bit more about the baseline pulmonary dysfunction of patients that had more serious pulmonary complications during the study with regards to -if someone could address that please. MS. CAHILL: sure if this is on. everywhere. Okay. Certainly, slide up, please. I can look I'm not

Oh, there we go.

This is a slide of the patients Again, I want to

presenting with pulmonary dysfunction.

remind people that this is not subjectively taken off medical history, but these are pulmonary dysfunction defined according to comorbidity indices, refers to a group of 140 patients on which this comorbidity index was scored. And

you'll see here the numbers of those patients reporting with -- presenting with pulmonary dysfunction as opposed to those patients who do not report with pulmonary dysfunction and the differences in their pulmonary adverse events as outlined here. These, of course, were at any time during the study including presenting with infection, et cetera, just as reported AEs. DR. ECKHARDT: Right. But when you look at

61

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

pulmonary deaths, how does that relate to this? MS. CAHILL: pulmonary. comment. DR. DICKEY: Well, the numbers of pulmonary Sure. We'll put up the grade 5

Slide up, please, and I'll ask Dr. Dickey to

adverse events are not surprising, and I'm not alarmed by them. They're similar to what we see at M.D. Anderson in My slide up, please.

patients treated with different drugs.

In thinking about the pulmonary complications of AML, it's helpful to break these into acute events, which we've somewhat arbitrarily cut off at 30 days, and chronic events. Now, reviewing a database of 3600 patients with leukemia and drawing upon that database was first interrogated by Elihu Estey and more recently by Bill Weirday (ph) and Jorge Cortez about 25 to 35 percent of patients during the acute period, develop infiltrates. And about half of those we

estimate are due to pneumonia secondary to the myelosuppression and then the other half to pulmonary edema usually noncardiogenic, a synonym would be ARDS. And here the thinking as to the etiology is it's a secondary effect from the tumor lysis with the release of endogenous inflammatory factors like ATP urate, HMGB1, hecoch proteins and so on and so forth. So, these are the

events that we're talking about mostly with that early morbidity and mortality. Now, down on the bottom is an

62

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

event that is specific to the use of laromustine in AML and a rate of 5 to 10 percent of late infiltrates was identified which is typical for an alkylating agent and about half of those responded to steroids. So I think it's important to recognize which events we're talking about and to split those out. But we

see a very high rate of pulmonary events at M.D. Anderson not using laromustine, similar to what's reported here. DR. ECKHARDT: DR. LINK: follow up on that. Dr. Link.

I have a couple questions, but I'd So what's your take on a randomized

trial where you have, you know, the other agents, certainly cytarabine, and then you have the comparator group where you have to stop the study because of pulmonary toxicity? was interested, I didn't hear much about that. your take on that? actually? MS. CAHILL: Well, we pulled the slide up. I just So I

So, what's

And what's happening with that study

want to clarify that this study was suspended to enrollment for an induction mortality difference, not necessarily ascribed to pulmonary toxicity. confuse the answer. DR. DICKEY: But I think what I'm saying is that So, I just wanted to not

the early events, the early pulmonary events are not due to a direct toxic effect, extramedullary toxic effect of the

63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

drug.

It's a secondary effect.

So, the prolonged

myelosuppression is what accounts for the very high rate of pneumonia. And the pulmonary edema, ARDS picture, the

bilateral early pulmonary infiltrates go along with effective tumor killing. So, when you have the high rate of

induction, you also have a high rate of cytolysis that leads to ARDS secondarily. And this is very similar to the traumatic ARDS that's been described, you know, for a long time. DR. LINK: Well, let me ask it a different way.

So, there's a bunch of very short remissions, you know, like one day. Now, one-day remission I interpret to mean nobody

can relapse, I mean, nobody accepts having a bone marrow done two days in a row. So, I assume it was a remission

death, meaning a kid, excuse me from my bias, patient enters remission but died within the next couple of days from an ongoing toxicities. How many of those were pulmonary

toxicities which were ongoing, you know, sort of right at the 30-day point? Because I think the assessment of That was the

remission took place at about 30-some days. median time to neutrophil recoveries. you're making your assessment. MS. CAHILL:

That's when I assume

I think I can clarify that data.

The

leukemia-free survival range to which I think you're referring is 3 days to 571 days. But it's leukemia-free

64

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

survival.

So it can be stopped or leukemia-free survival is So that largely may pertain to a

also measured at relapse.

patient who had evidence of peripheral blood relapse three days after a bone marrow suggesting that they were in CR. It does not reflect -- and it doesn't pertain to death or toxicity. DR. LINK: No, I understand that. But it's pretty

hard to have a normal bone marrow and three days later have peripheral leukemia. I mean, it doesn't, you know, maybe in

Burkitt's lymphoma, but even there it doesn't happen very frequently. So, the presumption that I would make taking

care of leukemia patients a lot is that those leukemia-free survival, you know, where the curve drops, must be remission deaths or deaths from something other than recurrent leukemia. MS. CAHILL: Slide up, please. This is the list

of deaths within 30 days of induction.

It doesn't

specifically address, but it captures -- because most of the responses on the CRs were in the 30-35 day window. We also

need to -- so, you're asking about a death from a -DR. LINK: This is death during induction. This

is, you know, anybody who takes care of leukemia sees that list. I'm talking about people who are in remission which

meant that they had a reasonable, by definition, have to have a reasonable functional status, normal peripheral blood

65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

counts, and normal bone marrow, or less than 5 percent blasts in the bone marrow. So having achieved that, you

know, it's very hard to die the next day of leukemia. DR. SCHILLER: Dr. Link, if you'll allow me to

address that because I'm a leukemia doctor, just as you are. It's just that my practice is confined to older patients, and you have younger patients. You have to consider the When I see

pre-treatment characteristics of these patients.

a new patient with AML who's over age 60, the first decision point is whether they're eligible for myelosuppressive therapy or not. If they're so frail as Dr. Balducci used

that expression this morning, and so infirm as to not even consider an attempt to achieve CR, then they go on to hydroxyurea or supportive care or low-dose cytarabine. But if they wish to go for myelosuppressive therapy in order to achieve a CR, which after all is the cornerstone upon which you build the structure of long-term survival, you have to recognize that many of them will have serious adverse pre-treatment characteristics making them ineligible for anthracycline, for example, underlying heart disease. They would have been excluded from SWOG trials or

underlying pulmonary disease that makes it difficult for them to be treated. But it is conceivable that one could

achieve CR and then notwithstanding their underlying organ dysfunction, they expire even in CR, a short period into the

66

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

complete remission. Notwithstanding that, this drug produced complete remission again in patients for whom cytarabine and anthracycline could not have been given. the cornerstone, again, of this process. DR. LINK: they died of. toxicity? I'm with you. I just want to know what And I think that's

So in other words, did they die of pulmonary

In other words, that's all I'm trying to find

out, was that where the pulmonary toxicity was effecting the survival. I'm trying to get what is it that killed them. MS. CAHILL: Sure. Slide up here. And in this slide I

DR. SCHILLER:

guess you can see among the 27 patients who achieved response, one patient received additional treatment, seven patients relapsed prior to their death. during remission. question. MS. CAHILL: Slide up to the next slide. Okay. Slide up. 10 patients died

So that might be the core answer to your

DR. SCHILLER: MS. CAHILL:

So if the question was where does

pulmonary fall into that, I think that that's outlined in these causes on this list. DR. DICKEY: Can I try again, maybe? The rate of

acute pulmonary events is high but no different from what we see using regimens that don't involve laromustine. So we

67

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

see a lot of bilateral acute infiltrates that we attribute to an ARDS syndrome from the tumor lysis and then a lot of pneumonia in acute leukemia. So I don't think there is any unique pulmonary toxicity during the early stage. I think that this drug is No question there is a

very nontoxic acutely on the lungs.

late pulmonary toxicity, this pneumonitis which is seen with all nitrosoureas, alkylating agents, and different part of the picture. DR. ECKHADT: DR. TEMPERO: Dr. Tempero. I just don't quite understand your Was it

definition of underlying pulmonary dysfunction.

based on something objective like pulmonary function studies or was it history or clinical history? MS. CAHILL: Yes. On the criteria and definitions

of all the comorbidities were objective and measured at baseline according to the HCTCI cormorbidity index. And

I'll ask Dr. Giles to speak directly to that indexing scoring of comorbidities. DR. GILES: Thank you. Our use prospectively of

that hematopoietic stem cell transplant comorbidity index was in a way to try and get this ODAC out of the dilemma that's pretty obvious. Based on the experience of ODAC with I and my colleagues think And

Zarnestra two things were clear.

we know people we cannot give 3+7 to when we see them.

68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

that's the great, great, great majority of people such as we're discussing today. The second problem, however, is an objective audience can't just take our word for it that we know something when we see it. As a regulatory group, they're We faced precisely

entitled to more objective evidence.

this dilemma when we came out of the first of the phase II studies that Dr. Deisseroth summarized. In other words, we

had a new agent, minimal extramedullary toxicity, no mucositis, no cardiac toxicity, no renal toxicity, we said where is the best area to further explore this drug. We then looked around to say comorbidity is the big, unquantifiable factor in all of the discussions. And

the elderly are very heavily biased against because of that absence. Our attention was drawn to the Saarer (ph) score

because as you know, most of the patients who went into that transplant dataset had myeloid leukemias. The score was

proven to be predictive for transplant-related mortality and overall survival. So we then took the score and did two very important things. We looked at the M.D. Anderson database

to see does this score predict for anything important in people over 60 given the standard regimen for younger patients. It will be no surprise to some of you, but just

to put it in context, we had to review a 17-year M.D.

69

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Anderson database to find 170 patients who'd received 3+7 and to retrospectively assess the value of this scoring system. The scoring system, I think the publication was

alluded to this morning, showed that in this retrospective analysis the worst score you had, the earlier you died, and the less your chance of getting a complete remissions. So, we had now found a score. It had been As you

developed in an independent leukemia population. know, it's been validated in MDS since.

We then took that

score, looked at the tiny minority who get I.A. and said, "Well, what happens there? Yes, it looks useful."

We then went back to the data on the first study and said well on that study, on the first population who got Onrigin did we have comorbidity patients as defined by the score who got Saarer. The answer was "yes." And that

brought us to where we are today.

We said, "Fine."

Comorbidity captures an utterly distinct, if overlapping, aspect that puts people off giving combinations cytotoxic therapy to the elderly. it. So we use, slide up please, we used these specific factors from the HCTCI prospectively and said, "This is what you must have if you're to go on this study." eligibility, of course, is one thing. Now We've got to prospectively define

But as you all know, I must say, I was

what's likely different is who goes on.

70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

both enthused and a little bit terrified by everybody's enthusiasm to put 70 percent of the patients on that 043 study. Over 60 patients of the 85 had either severe or

moderate pulmonary or cardiac toxicity. And if I could show the slide please again, which I think is very important, and addresses also the earlier question about pulmonary toxicity, even in patients who had prospectively defined severe pulmonary or cardiac toxicity, we were able to maintain a CR rate. So, it always behooves This is the sort

us and as Ms. Cahill -- slide up, please.

of pulmonary comorbidity that as prospectively defined. These are patients in whom the overall objective response rate was over 30 percent, and they are 60 of the overall 85 patients. But again, just slide up this one please, I think

this speaks to our attempts to prospectively define these patients. It is pretty clear, for example, if you look at the group with cardiac dysfunction, none of us are going to give those patients an anthracycline. None of us wants to

give those patients low-dose ara-C where they have zero chance of getting a CR as shown by the MRC data. So, we are

very excited about the therapeutic option of having Onrigin available for these patients. There is no quibble

whatsoever with those of us very involved in all of this that a randomized study should and will be done. The

71

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

commitment is there already. But the issue is should the pressing rationale to have a randomized study in every possible circumstance, should that triumph over the need where a population has no standard of care other than one that is futile? DR. ECKHARDT: Dr. Richardson. I have a couple of questions.

DR. RICHARDSON:

First of all, with regard to the pulmonary toxicity issue, it's been said that about half these patients were steroid responsive. And I guess I'm curious whether this means this

was reversible totally in this group, and were any of these patients able to go on to further therapy? And then I'd

like to ask Dr. Schiller what he thinks the likelihood of cardiac toxicity is for anthracycline in somebody who has some of the cardiac criteria in the 043 study, for example, the ejection fraction less than 50 percent, it seems to me looks at a pretty good performance status group, prior history of one or more stents being placed in this group if somebody's got a reasonably good ejection fraction with that type of cardiac history, I wonder what their likelihood of cardiac toxicity is and is this the same magnitude as a 9 percent fatal pulmonary toxicity. MS. CAHILL: So, why don't we take those two

questions in turn starting with the steroid responsiveness to the noninfectious pneumonitis and then if it's all right,

72

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

we'll address the question of cardiac. DR. DICKEY: I haven't reviewed the cases one by

one, the laromustine patients who experienced the late pneumonitis, but typically, just to tell you in general, patients who experience an alkylation-related late pneumonitis that does respond to steroids typically that clears and patients can go on and, you know, do well. Now, they are at risk of increased subsequent pulmonary toxicity from things like radiation or additional alkylating agents or other pulmonary toxic agents like taxains (ph). So we would avoid them in those patients in

general, but I haven't gone case by case through the Vion cases to know whether those patients got additional subsequent therapy, so I can't comment on that. MS. CAHILL: I'll put a slide up, please. This

just gives a little bit more detail of the other therapies of supportive care antibiotics and oxygen that patients might have received and how that influenced their symptoms. Six of 12 patients receiving steroids improved, patients with this pattern of pneumonitis who did not do well did so mostly because of their leukemia and complications of pneumonitis, I'm sorry, complications of pneumonia and infections were the primary reasons of cause of death in these patients. There was, additionally, cases of acute

respiratory failure in that setting.

73

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Slide up, please. of death in the 16 patients. DR. DICKEY:

This is the list of the cause

One of the critical things when this So, now

event occurs is early recognition and intervention.

that the syndrome has been well-defined for this particular alkylating agent, it's important to make patients and caregivers aware, and Vion has proposed to include that in the label, which I think is certainly the right thing to do. MS. CAHILL: Again, and I also just wanted to just

point out the shift I made from 043 to when we talk about the toxicities I'll move to the entire 277 patient dataset getting 600 milligrams per meter squared. And now Dr.

Schiller, perhaps you would address the cardiac complications and contraindications. DR. SCHILLER: Dr. Richardson, when I see an older

patient with acute myeloid leukemia in clinic who elects to receive myelosuppressive therapy in an attempt to achieve complete remission, I am constrained by our data with cytarabine and anthracycline. The SWOG trials with which I

have a little bit of familiarity because we were a SWOG site, restricted access to continuous infusion cytarabine and anthracycline on the basis of left ventricular ejection fraction under 50 percent which is a pretty tight criterion, what they called significant arrhythmia, not clearly defined, and angina.

74

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

And so if I have patient who's had a prior myocardial infarction and has had a stent, then again I'm constrained given the limited information we have, and slide up will show you the SWOG database. You see that very few

patients on 9031 which is one of the trials in the elderly using cytarabine and anthracycline had increased performance status or advanced age. And there's no discussion in the

Blood paper in 2007, for example, on the number of patients who had any degree of prior cardiac history. And so my

assumption is that patients with cardiac disease cannot receive an anthracycline. an anthracycline. And I'm hard-pressed to give them

So I think that answers the question.

The second thing is on the CLI-043 there were overlapping risks. And so, for example, my patient who is

here today had not only a history of the coronary stenting and myocardial infarction but also had bilateral pleural effusions and pulmonary disease. And so I think that there

might be a slide that Vion can show that identifies that the risk factors on this study, slide up, please, in terms of those with cardiac comorbidity, usually had extensive cardiac comorbidity with multiple high-risk features that would have, again, rendered them ineligible to the so-called standard of cytarabine and three days of anthracycline. I hope I answered your question, sir. DR. RICHARDSON: I guess you answered it about as So

75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

well as one can without any information on that. DR. ECKHARDT: DR. FLEMING: Dr. Fleming. I'd like to go back to the adverse

events, the infectious deaths, and the pulmonary deaths. The 037 trial does provide a randomized comparison and reflects 15 pulmonary deaths and 49 infectious deaths, 8.5 percent and 28 percent respectively. If we saw that without

a control we might say, "Well, this is partly natural history." Oh, the randomized control says there were no There were only three

pulmonary deaths in the control arm. infectious deaths in the control arm.

So then we go forward to the 043 study data. Things went pretty quickly here, but I thought the case was being made that it's different in this setting and that we're not seeing early deaths while pulmonary-related deaths, there were six. That's 7 percent. That's almost

the same rate that we were seeing in the 037 trial. Neutropenic sepsis deaths, there were five. that were infectious or pulmonary were 28 and 13 respectively. The data that was put up really quickly for SAEs

pulmonary deaths by no pulmonary dysfunction, pulmonary dysfunction, didn't seem to differentiate between those two categories meaning that they weren't all attributed to the category of people who had pulmonary function, dysfunction at baseline. The overall numbers of patients that had SAEs

76

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

that were infectious in their 277-person metanalysis were 77, 24 pneumonias, there were 12 patients that had -- well, in fact there were 77 patients that had pulmonary grade 3 to 5 adverse events. So, it seems as though if start from the randomized trial, there is an incredibly strong signal for an increase in infectious deaths, pulmonary deaths. And

then when we go through and track through the data that we have from the 043 trial and 033 trial and then from the metanalysis, it doesn't seem to give us a sense that it's not relevant and apparent there as well. So then we look at the efficacy side and what we see are some CRs. They are interestingly relatively shorter

duration, though, for remission, 5.7 months and only 3.6 months. And then FDA had a slide, it didn't get shown in

the oral presentation, but a slide talking about what was the duration of response in the 037 trial when you look against the active comparator. I don't suppose you have

that on your fingertips to show? But it's in the briefing -- FDA briefing document on page 29. And it shows quite a lot shorter duration of And we

response for the responders in the laromustine arm.

had some questions before about what's happening when you're in response and are there deaths that are occurring that it could be toxic-related. And it all hangs together. It

77

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

makes you wonder.

Not all responses are the same.

But it

all, again, points to the fact that we have on-target effects, and we have off-target effects. influence outcome. And they both

And really understanding what's disease-

related because we're hearing a lot from the sponsor now about things being disease-related is really extraordinarily difficult to determine without a proper randomized control. And when we have one in 037 there's a striking discordance between a doubling in the response rate. yet survival going in the opposite direction mediated through off-target effects that aren't being captured by what you're looking at when you're looking at responders living longer than nonresponders. So, in essence, that one figure you're saying isn't available, but it's worth at least looking at briefly on page 29 of the briefing document. DR. DEISSEROTH: survival analysis. median duration. DR. ECKHARDT: DR. WILSON: Dr. Wilson. Yeah. That was a leukemia-free And

And it showed a very short duration,

So, I actually had some very similar

comments, but I just kind of wanted to maybe have a overview perspective. I think we're faced with a situation where we

have a number of trials that show activity, but they're confounded by the use of other agents. And so it's very

78

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

difficult to really be able to sort out how much this new drug is adding to it. And then we have this randomized

study where it's randomized against, well, the ara-C or araC plus placebo where I think very predictable findings are coming out. This class of agents, nitrousureas, are very, very old. They are well known to cause pulmonary problems. I

guess as somebody who is a -- who really deals with chemotherapy agents, I would say they really stand out at the top of the list among alkylaters that cause pulmonary toxicity. I disagree with one comment which is they cause pulmonary toxicity like other alkylaters. as a class, at the top of the list. They really are,

They're also a class of And so while

agents that tend to be quite stem-cell toxic.

we want to kill the putative leukemia stem cell, we need to have the normal cells come back as well. And they are

notorious for hitting normal stem cells delaying recovery. And it's that delay in recovery which has not been addressed here but may play some role in the number of infectious deaths. And then finally, I think that ara-C is another drug that causes pulmonary toxicity. And I would be

concerned that when this agent is used with, not simultaneously with ara-C, but with ara-C following it,

79

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

perhaps over multiple cycles, that we will then begin to see an emergence of more pulmonary toxicity. So all my comments aren't to say this isn't an agent which has interest. But to me how effective it is

versus how safe it is is still unknown because of the absence of a randomized study except for the one we have here where it was clearly worse making me very concerned about the relative risk-benefits. MS. CAHILL: those comments in turn. Sir, I'd like to be able to address And again, trying to remember the

differentiation between a phase III in a first-relapse setting with an average age patient of 60 and a CR of 1 of nine months, but still allowed us to characterize the pulmonary profile at the 600 milligram dose when used in combination with ara-C. So, before I ask Dr. Dickey to comment on the pulmonary problems which have been very well characterized, I'd like to take the two shorter points and ask Dr. Giles if he would comment on the use of consolidation as confounding results. DR. GILES: Well, and I guess my first comment

would be, it's a confounding factor I wish I got to use more often. I mean, I think all this talk about toxicity is

losing sight of one of the key descriptors that we all use to communicate with each other about toxicity in AML and

80

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

that is the early death rate, all causes, on this patient population of 85 patients, no other drug involved, no hydrea. The early death rate was 14 percent. That is

about less than half that we had documented in 3+7 patients with multiple risk factors in the retrospective analysis. And is very, very comfortably in the median of what we would even see in considerably younger patients. So the second thing that may be contributing to some unfortunate confusion, the randomized study, we ran into excess myelosuppression, the chief reason that people had excess mortality was excess myelosuppression that probably as you say reflects stem cell toxicity, which is one of the positive aspects of this drug and one of the reasons that it probably gives us over 30 percent CR rate where we might not otherwise get it. So I think the prospective randomized studies will indeed sort out the interaction between Onrigin and ara-C. But I do want to be very clear that that interaction at the

pulmonary or the marrow level, is utterly distinct from the Onrigin, I think we can all agree single-agent data in which the most objective fact in front of you is a 14 percent allcauses early mortality traded against an over 30 percent objective response. I wish it were possible to induce anyone in AML

81

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

without any induction-related deaths, but that's -- we're not quite there yet. And as regards the questions about any

of the classes of agents, we heard discussed this morning the weaknesses of nucleoside analogs and myelosuppression in the liver, perhaps on the kidneys, anthracycline's cardiac toxicity. There's no question that outside of the marrow

the lungs are going to be the vulnerable point as you said, particularly of any agent in this class. But I think that

can easily be dealt with by prospective analysis. And I don't want this figure of a 9 percent mortality, 9 percent of people were found to perhaps, in retrospect, have something like delayed pulmonary toxicity. As you saw, half of those responded quite well to steroids. And in point of fact, unfortunately, the others, because no CR was expected in this circumstance, were let loose without adequate warning, without the sort of warning about if you get these symptoms come see me or give me steroids, the sort of thing that you would expect in a responsible label. So, I would ask you to -- I think it's important that we distinguish between a combination regimen in which we clearly got the dose wrong, and single-agent data in which we have a very reassuring response rate for a regrettable, but considerably ordinary, 14 percent early death rate. MS. CAHILL: Because I think the pulmonary AEs in

82

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

report are well characterized and I don't want to leave confusion on that, I'll ask Dr. Dickey to speak. I will

briefly interrupt speaking to the stem cell and recovery that the median nater (ph) is at day 15 and the median day of recovery is at day 30, so some similarities and some differences to other induction regimens. But please, Dr.

Dickey, regarding the characterization of pulmonary AEs. DR. DICKEY: So the two points you made, the one

about the nitrosoureas having well-known pulmonary toxicity, I agree completely. But that's very long term. And I don't

think that there's any good evidence that they have any direct acute pulmonary toxicity. So, we would put that in a

later stage, and, of course, it can go out to as late 20 years we can see progressive pulmonary toxicity very well described, I agree with you completely since the days of BCNU. Now going to the acute toxicity, and as you say, ara-C certainly causes acute pulmonary adverse events. I don't believe that those are direct. But

So, in a nonleukemic

animal model, there's essentially no toxicity from therapeutic doses of cytosine arabinoside. So the effects

are indirect and, again, probably related to the tumor lysis possibly to some endotoxin leak from the gut. But, you

know, a little bit of edema in most places in the body besides the brain and the lungs doesn't cause a tremendous

83

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

amount of symptoms.

A little bit of fluid in the lungs

causes a lot of symptoms. So, when you have the acute tumor lysis and a little bit of fluid leakage into the lungs, that's ALI/ARDS. So I think that those are indirect, you know, nondirectly toxic, pulmonary toxic events. DR. WILSON: will stop. Let me just follow up, and then I What I was going to say is, I

Are you okay?

just think that it is important not to think that the -that -- to say that the toxicities of the nitrosoureas are long term and therefore these patients aren't going to live long term, therefore they're not going to be an issue, I don't think really necessarily applies. We don't know this.

But if you even -- if you give ara-C after these drugs, it is very possible that they may enhance the toxicity of the ara-C. So I just think, we don't know, that's the problem. And I think we just don't

There's a lot of signals here. know.

But I don't think you can separate the two even if

they're physically separated until you have an adequate randomized study. DR. ECKHARDT: Chair's prerogative. had a question. Okay. I'm going to take the Dr. Sekeres, you

We're moving on now.

And I just would like to make sure that we

focus with questions that relate to specific data in either presentation because we're getting into some debates over

84

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

things that we really don't have the data at hand to clarify. Thank you. DR. SEKERES: also a leukemia doc. Okay. Thank you, Dr. Eckhardt. I'm

So we have a surfeit of leukemia docs When you have an older adult These folks have

here in the room right now.

with AML, it's a very difficult situation.

the potential for being cured, but that potential is between 5 and 10 percent, five year leukemia-free survival. We often offer these folks either basically supportive care, which can include hydrea, some sort of midrange therapy, and I actually believe Dr. Burnett has set the bar for how we define mid-range therapy with low-dose cytarabine or alternatively a DNA methyltransferase inhibitor such as azacitidine or decitabine or we offer aggressive intensive chemotherapy. I'm having difficulty

getting my head around this drug in this trial of how I would compare it to one of those three treatment approaches. On the one hand, the sponsors are edging into comparing this drug to intensive chemotherapy. On the other hand, they're

deliberate in presenting Dr. Burnett's data with low-dose cytarabine. Can the sponsors comment on how we should best compare this regimen? MS. CAHILL: That is an excellent comment and so

helpful because in developing a new myelosuppressive drug

85

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

and an option for a patient population not currently being treated, you run into this conundrum of what are you. I

remember people saying to us, "You're no better than 7+3." And we were thinking that that's a good thing in patients who can't get it and in such a large percentage of patients not getting anything. There's no question that this drug is

myelosuppressive, there's no question it induces complete remissions. It's not chemo light. It's just chemotherapy

that can be tolerated and delivered in a very poor-risk patient population. Slide up, please. So I think to address your

question, and I guess, and what I'm saying is, I'm acknowledging that there are different buckets of treatment. If someone would say to me, "Are you saying that no 70 year old can get 3+7," that is not what we're saying, that's not who the trial enrolled. There is a patient population that

is older, that carry with them leukemia biology risks and comorbid risks which we objectively identified and have demonstrated around this patient population. And that is

the -- it is the kind of chemotherapy is, and it is a chemotherapeutic option for this type of patient. DR. SEKERES: second. I want to readdress that for a

And I'm going to echo some of what Dr. Estey spoke The majority of patients in this

about this morning.

country are over the age of 68, if you believe SEER data.

86

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

So age of 70 or older to me means this is a typical older AML population. ECOG performance status of 2, we can argue Cytogenetics here are defined

because it's subjective.

differently than they were in the MRC, and perhaps in California having a stent for your cardiac disease is unusual, but I can tell you that in Ohio it pretty much defines my patient population. So I'm taking issue with your definition of a poor-risk older AML population. And I'm going to ask my

question again, to which of those three treatment approaches would you compare this regimen? MS. CAHILL: Again, I would have to say it has

features of a myelosuppression regimen for patients who currently don't have therapeutic options. So this patient

population, to me, represents that group of patients for whom the decision to get 3+7 has already been made. And I

think that the characteristics that we show really identify for that. Not only does that identify them as patients

where the physician's already decided this patient couldn't get 7+3, the data suggests that because only six of the patients went on to get a 7+3 regimen later. And only one

of those patients responded underscoring that these aren't the 7+3 patients. So then if you ask me, "Are they the no-treatment patients, or are they the low-dose ara-C patients," to that

87

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

I would say, you know, we don't know -- I don't think they're the no-treatment patients because they have the 32 percent complete remission rate. Are they the hydroxyurea

best supportive care patient, I don't think that either because they are able to achieve a complete remission in the face of an acceptable induction mortality. Perhaps Dr. Giles can really give a better picture of who these patients are. DR. GILES: Well, I totally, obviously, sympathize Absent

with the question, and let me answer very directly.

the Onrigin data, biologically if I had to choose one of the three, it's the low-dose ara-C would be the comparator. other words, you clearly cannot give patients with this degree of comorbidity complex anthracycline ara-C regimens. So biologically, if I had to choose, it would, of course, be low-dose ara-C. But then we get into the small matter of ethics. You and I face this everyday. If we were to go into a In

patient and randomize knowing the Onrigin data that now exists and knowing the low-dose ara-C data for minus five, minus seven poor performance patients that we saw earlier, most of our patients, as you know, we were able to turn around and say, "Well, would you go on the study," or, "what would you do? Would you put, whatever, whoever you care

about on the study?"

88

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Okay.

I think most of us would have grave, grave difficulty asking a patient, I want you to go on one arm that may have toxicities, we need clarification, certainly it has toxicities if combined within appropriate doses of other agents, but if I give it to you as a single agent, I have a large credible database in which you have a 30 percent chance of getting a CR, and if you get a CR I can do something with that. On the other hand, I can give you another agent which is a higher early death rate because the chief cause of death is leukemia, and it has no chance of giving you a CR. MS. CAHILL: question -DR. ECKHARDT: Dr. Balducci. DR. BALDUCCI: Well, first, a brief comment All right. We're going to move on. To completely answer your

concerning frailty because this is something that will come again and again hopefully on ODAC. specific definition in geriatrics. Frailty has become a And there are some

specific criteria established by the cardiovascular health studies out to establish that. And generally are not

patients who are already infirm, are patients who may be coming firm from a small trauma. Otherwise, I just wanted to ask a couple of

89

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

questions concerning the study.

First of all, why did you

not consider people with previous hematological diseases among the high-risk patients? The second, what -- why --

whereas the Saarer classification is certainly a very legitimate classification for comorbidity, I didn't see any trial to have a comorbidity index done out of that to certify the patient. For example, a patient with congestive

heart failure clearly is a much more serious condition than the patient who had a stent put in five years earlier. So,

to put these things together is like comparing apples and oranges. And third, could I see the confidence interval in the response rate of patients with poor cytogenetic abnormality? Because the response rate there was fairly

low, was 18 percent. MS. CAHILL: I'm going to try to move quickly and Number one, CLI-043 was

take those questions in turn.

designed without antecedent hematologic disorders because it was a logical progression of the first phase II study. We

prospectively wanted to identify in a group of patients that identified to have the greatest benefit. And the complete

response rate in the first phase II was somewhat lower. That was debated among the physicians at the design of the trial because antecedent hematologic patients have such a poor prognosis. They wanted them to include it, but we felt

90

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

like it was best to show -- to make sure we were with a homogenous patient population. You probably already -- oh, sorry. DR. BALDUCCI: For the comorbidity index, in other

words, all comorbidities are not the same, chronic hepatitis is not the same thing as liver failure. And heart failure

is not the same thing as history as coronary artery disease. MS. CAHILL: Absolutely correct. And the beauty

of the HCTCI was weighting in those and our patients were weighted so that they weren't all considered together. There were no patients enrolled with just a single cardiac comorbidity, and it counted the same. Those were weighted.

Dr. Giles, could you comment on the HCTCI scoring? DR. GILES: Well, I'm delighted to because we have

put a lot of effort into this because we want this large group of people who are neglected, we want them back in the therapeutic fold. In fact, if you look at the distribution

of severe and moderate, between the cardiac and pulmonary adverse events, in fact, the CR rate was maintained in the 30 percent range which I found very encouraging. So, we did

not have to get into the issue even were we to say, "Well, I think some of the low moderates are not that bad," that did not actually affect the overall results. The second thing is, it was the robustness of the database that Dr. Saarer and the others had that made us use

91

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

this specifically, one, generated predominantly in myeloid malignancies. Two, we had access to the fact that although

it had been generated with both a development and a training set in Seattle, the Anderson had validated in their transplants, as you know, and this was reproducible. It was

also reproducible for two things that are very germane in elderly AML, your chances of surviving the procedure, which in some ways is equivalent. And that was the main driver.

As Ann has just said and as you know, they had done a beautiful job of dealing with the issue you've raised, skewed all the weighings towards things that were actually likely to kill you with the conditioning regimen in the case of transplant or induction in the case of 3+7. we felt it was it most appropriate one available. are working on one. There's another European one. But I would draw your So,

The MRC I'm sure

we'll end up in a consensus.

attention to the fact that people have looked at this in high-risk MDS, which biologically as we all know is pretty similar to these patients, and it works there also and is predictive of overall survival and one's likelihood of perishing early in therapy. If you look at the totality of that data, and I present you with 85 patients, over 70 percent of which have either or both of severe or moderate cardiac or pulmonary toxicity, you've a 14 percent early death rate. I think

92

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

that's not something that we've ever encountered before. It's something we deliberately set out to demonstrate and it very exciting for these patients. DR. ECKHARDT: question? DR. BALDUCCI: The confidence interval in the Okay. And then your third

response rate to in-patients with -- who are cytogenetics, if you have it. MS. CAHILL: We'll pull that slide up.

Unfavorable cytogenetics as a subgroup, you're right, did less well in this study as presented there. slide up, I can see it, but you can't. DR. BALDUCCI: DR. ECKHARDT: DR. LINK: Okay. Thanks. Dr. Link. Oh, I'm sorry,

All right.

I just have a real quick question about

the drug related to other alkylating agents, so cyclophosphamide has some activity in AML, and I'm just curious to know, you know, is this one much better, is there any data to suggest that this is better than, sort of more conventional, earlier alkylating agents? MS. CAHILL: Yes, I'll ask Dr. Giles to also

comment -- or Dr. Löwenberg, I'm sorry, because I'm often reminded that we weren't the first with an alkylater in AML. DR. LÖWENBERG: Maybe this question can bring us I mean, we are

back to the reality of daily practice.

93

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

really discussing patients that are denied chemotherapy today. So we get very good remissions. We have discussions

about alternative drugs, could they receive alternative agents, they don't receive it. We have discussions about

what would happen if cytarabine would be given after this. Perhaps you are aware of the fact that all randomized studies that have been done, post-remission, with cytarabine have been negative. There's no proven added value. It

doesn't make sense to do it.

I think the point that is made

here is let's take advantage -- let us take the gain generated by this drug when it is applied as a single agent in one single injection. And if these patients achieve a CR, the best option is probably to go to a transplant. So, any other

drug, I don't think is consideration at the moment for these patients. DR. LINK: That's a great answer, but it wasn't My question is about

answering my question.

cyclophosphamide, not cytarabine. DR. LÖWENBERG: Cyclophosphamide -- Dr. Link, I

think cyclophosphamide is also not something one would consider for these patients based on the available evidence. These poor-risk patients -- no data to support the use of cyclophosphamide. DR. ECKHARDT: Thank you. Dr. Curt.

94

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

DR. CURT:

I have a question for the sponsor, and

that is how significant is the pulmonary signal if patients with underlying pulmonary morbidities are excluded from the analysis? What I'm trying to get at is, how significant is

the risk if we look at the patients along the lines of your proposed label? MS. CAHILL: I'm sorry, if I restate that, you

want to know how significant -DR. CURT: Yeah, how significant is the pulmonary

signal if you exclude the patients with underlying pulmonary morbidities? MS. CAHILL: So we probably can get that from the Okay. So,

pulmonary AEs by dysfunction or no dysfunction? we'll pull that slide up.

Again, those patients haven't

been excluded because they had the response rate in the same range. Slide up, please. The left-hand column is those Dr. Dickey, a

patients without pulmonary dysfunction. comment? DR. DICKEY:

Well, the comment has been made that

we don't know precisely, we don't have the data from a headon comparison. But speaking from the experience at M.D.

Anderson that I mentioned to you before, these rates look similar. DR. ECKHARDT: DR. FREEDMAN: Dr. Freedman. First of all, I'm not a leukemia

95

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

physician, but I was struck by the disparity between the responses in the survival in the randomized trial. understand this trial was reopened in early 2008. FDA allowed it to continue. And I And the

What was it that the FDA must

have felt satisfied that the trial could continue at that point? DR. PAZDUR: The dose was lowered. Dr. Sekeres, last question. Yeah, thanks again, Dr. Eckhardt.

DR. ECKHARDT: DR. SEKERES:

Getting back to something I started with earlier, if we're only able to cure 5 to 10 percent of these patients, then ultimately anything we give is going to palliative in 90 to 95 percent of these patients. So, can you tell me what

percentage of these patients were hospitalized and whether you have any data on patient-reported outcomes? MS. CAHILL: reported outcomes. hospitalized. We have limited data on patient-

The majority of patients were

The treatment regimen does not require And somewhere in the order of 10 to 12

hospitalization.

percent of our patients were able to receive Onrigin induction as an out-patient. an in-patient disease. However, AML, as you know, is

And the hospitalization numbers,

limited that we have in the first six weeks, look about the same across the literature as to other AML regimens. Sorry, I'm not answering it?

96

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

DR. SEKERES:

No, you are.

But the numbers that

you're giving me in the time to neutropenia and then count recovery sounds a lot more like remission induction therapy than it does low-dose cytarabine which can routinely be given as an outpatient with CR rates of 18 percent. DR. SCHILLER: MS. CAHILL: Well, this, if I may interrupt?

Please. This regimen can be given either as And I certainly did give it to

DR. SCHILLER:

in-patient or out-patient.

patients in the out-patient setting awaiting their neutropenia and then hospitalizing them. The 18 percent

remission rate that you quote with low-dose cytarabine does not apply, of course, to patients with adverse cytogenetics. And Dr. Burnett showed us with that treatment, almost nobody would achieve remission. And the palliative treatment arm

would also include prolonged hospitalization for transfusion support and the management of infections with a survival estimate about 1 or 1.7 months. So the palliative arm is

also one complicated by prolonged hospitalization. DR. SEKERES: Just one retort, and that is that as

we both know I think, Dr. Schiller, the definition of adverse cytogenetics, I believe varies between the MRC study and this one. DR. ECKHARDT: public hearing. We're going to move on to the open

97

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

MS. VESELY:

Both the Food and Drug

Administration and the public believe in a transparent process for information gathering and decision making. To

ensure such transparency at the open public hearing session of the advisory committee meeting, FDA believes that it is important to understand the context of an individual's presentation. For this reason, FDA encourages you, the

open public hearing speaker that at the beginning of your written or oral statement to advise the committee of any financial relationship that you may have with the sponsor, its product and if known, its direct competitors. For example, this financial information may include the sponsor's payment of your travel, lodging or other expenses in connection with your attendance at the meeting. Likewise, FDA encourages you at the beginning of

your statement to advise the committee if you do not have any such financial relationships. If you choose not to

address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking. The FDA and this committee place great importance in the open public hearing process. The insights and

98

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

comments provided can help the agency and this committee in their consideration of the issues before them. That said, in many instances and for many topics, there will be a variety of opinions. One of our goals

today is for this open public hearing to be conducted in a fair and open way where every participant is listened to carefully and treated with dignity courtesy and respect. Therefore, please speak only when recognized by the chair. Thank you for your cooperation. five minutes each. DR. ECKHARDT: Muller. MR. MULLER: and gentlemen. Thank you, good afternoon, ladies Thank you. First, Mr. Ronald And each speaker will have

Wow, elderly and at risk, that was me a When

year ago when I fell into Dr. Schiller's case study.

I entered the UCLA Medical Center I was suffering from pneumonia, I was suffering from a heart attack, and was diagnosed with AML leukemia. And we went and discussed with Dr. Schiller and my chances were pretty dog-gone bleak. And they always say

that one picture is worth ten thousand words, I'd like to present myself as that picture today because I was in very,

99

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

very bad shape with limited chances of survival. We did the Onrigin with one-hour, one-shot deal. And it got me into remission. And we had subsequent And I feel like I'm

consolidation treatments two times.

here today because of the drug that the Vion Pharmaceutical Company provided to me. And I just, again, want to thank

all the people that were involved in my healing and in my treatment. folks. one. Dr. Schiller and his staff are world-class

Dr. Schiller has taken an interest in me from day I'm very proud to be part of his

He's here today.

patient list. I also want to thank my wife of over 51 years for standing by me also because the most frightening words you can ever hear is, "You've got cancer, and you not only have a cancer, you have AML, and you don't really have a lot of chances and a lot of hope." But I think Vion And I just want Thank you

Pharmaceutical Company gave me that hope. to thank them. very much. DR. ECKHARDT: Jody Simon. MR. SIMON: All right.

And that's all I have to say.

Thank you.

Next is

I have nothing to disclose.

Good

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

afternoon, everyone.

My name is Jody Simon.

I'm the CEO

and one of the founders of the Geriatric Oncology Consortium. The GOC is a not-for-profit clinical research

and educational network put into place for the purpose of furthering our understanding of how best to treat the older person with cancer. As a healthcare community tasked with

treating all patients, we have been more than remiss in our focus on this group that constitutes more than 60 percent of all cancer patients, six out of every 10 cancer diagnosis is given to a person over the age of 65 to restate. Only in the last five years or about has there been any movement in this direction albeit nominal. My

father was a professional wrestler who died on this date, actually, this date today 15 years ago in 1994 from treatment complications of AML. He was 61, had lousy

cytogenetics, poor prognostic indicators, his therapeutic options were extremely limited. road since then. We've really made no end-

Perhaps if the products he was treated

with were studied in older adults, his complications could have been avoided, expected, maybe managed better. I felt obligated to come here today and to state

101

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

that we must be much more judicious in our approach to cancer in the elderly both within the grassroots treatment of this patient population by all oncologists, hematologists, but more importantly, the level of the regulatory bodies that choose to approve or not approve new products or uses of these products. The FDA mandates that products be tested in the populations in which they are to be used, yet the average age of a clinical trial participant in oncology is in the mid-50s with no comorbidities. This clearly does not match

the population where these products are normally used. Should we not have phase II studies specific for the older population to establish whether new drugs have the same efficacy and safety in these older adults as in younger individuals. Thank you for giving me the opportunity to

speak before you, GOC looks forward to working with ODAC and the FDA in this regard in the future. DR. ECKHARDT: MS. MEYER: Thank you. Thank you.

Ashley Meyer.

Hello, my name is Ashley Meyer, and

I'm very honored to be invited by Vion to be here today to tell my story and support the importance of medical research. Thank you so much for this opportunity.

102

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

November 11, 2001, was a day I'll never forget. I was diagnosed with acute myeloid leukemia at the age of 22. Life as I knew it changed, and this is where my story I was immediately hospitalized to undergo After two months in the hospital,

begins.

induction chemotherapy.

two rounds of chemotherapy, and one relapse, I was released knowing that the next step along my journey would be a transplant. The first attempt was an antalogous stem cell transplant in April 2002. After a series of ups and downs The type of

in my blood counts I relapsed in August.

leukemia I had required a full allogeneic bone marrow transplant from a donor. My family immediately researched

the number one bone marrow transplant facility, and we are so thankful for M.D. Anderson Cancer Center in Houston, Texas. They have been a lifesaver. On October 16, 2002, my sister gave me the greatest gift of all. Courtney. She gave me life. Thank you so much,

This second transplant allowed me to stay in I had started back to

remission for a little over a year.

work part time, and life was beginning to get back to what some people may call normal. I quickly learned from me that On November 11, 2003, two

life was not meant to be normal.

103

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

years to the day after I was originally diagnosed we are on routine check up to M.D. Anderson when we discovered the leukemia had returned. Not knowing what the future may

hold, we relied on faith and lots and lots of prayers. The answer to our prayers came when the leukemia doctors told us they had a new chemotherapy treatment they were going to try. They called it V.N.P., very new product.

This new product now called Onrigin gave us renewed hope. The original plan was to use Onrigin to get me into remission before attempting a third transplant. we know that this was all I would need. round of chemo I would ever receive. Little did

This was the last

I am so blessed to

have remained cancer free since that treatment in November of 2003. To make a very long story short, in April 2004, I received my third and last mini-transplant without any additional chemo. I am very proud, blessed, humbled, and

thankful to be here today cancer free and five years in remission. Without Onrigin and the grace of God I would not There were a few

be standing here today before all of you.

complications that were a result of my immune system being knocked down so many times. After the very first round of chemo in 2001, I received a fungal infection which eventually involved three strokes, two craniotomies, partial loss of my bones, and --

104

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

I'm sorry -- and a few other minor bumps in the road. that last round of chemo, life has been a series of adventures reaching new milestones every day.

Since

After the strokes we were looking into a seeingeye dog because I could not read the giant E on the eye chart. But miraculously my vision has almost completely And now I am able to drive again without any

returned.

accidents I might add. In 2006 after wobbling for two years with a cane and crutches, the doctors decided to do a complete bilateral hip replacement surgery to repair the damage done to my bones. After a long round of physical therapy, a couple of

scares, and support of family and friends the dreams continue. This past October my dad was able to walk me down

the aisle at my own wedding, and it was like a fairy tale. This past May I celebrated my 30

th

birthday.

And

this past weekend I was able to climb those stairs at Lincoln Memorial. Both of these accomplishments seemed Throughout my

almost impossible just a few years back.

journey my family and I have realized the true meaning of carpe diem, to seize the day. We realize you need to

celebrate each and every moment because you don't know what tomorrow is going to bring. Now as I mentioned before we still do have some bumps in the road where we've had some emergency room

105

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

visits.

But that gives us a continuous reminder of just how If you've ever spent time in a waiting Next week

precious life is.

room in a cancer clinic, you know what I mean. we'll be back at M.D. Anderson.

And I must say I'm very

much looking forward to our visit, being able to share a success story with them. I can't wait to tell them about my

visit this week in hopes that researchers are continually working on new treatments to find a cure. We have met many people along our journey. And

unfortunately some of them are here today only in spirit. My heart goes out to their families. And I am so thankful

everyday that God has spared my family the heartache that many of my friends have experienced. It is my hope that

some day every family will be able to be told, "Don't worry. We have a plan." As I mentioned when I started this story, there is a time for everything. This is my time to count my This is my time to

blessings for the gift of new life.

thank God for giving God -- for giving the doctors, researchers, and medical staff the talent to save lives, and for this we are eternally grateful. Speaking to you today and actually having the words come out right has truly been a dream come true. Thank you so much for giving me this opportunity to pay it forward in hopes of making other families' dreams come true

106

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

as well.

Everybody deserves that chance. DR. ECKHARDT: Thank you. And our last speaker is

Alexander Zelinsky. MR. ZELINSKY: Hello, my name is Alex Zelinsky. I

was diagnosed with leukemia in September 8, 2003. see my family doctor, I thought I had the flu.

I went to

They did a

blood test, and they came back with leukemia that evening. They sent me to Indianapolis where I seen Dr. Cooper. given 7+3, aracet-C with anthracycline, and Mylotarg. all failed. They found a bone marrow donor for me. And I was I was It

blessed enough to meet John O'Bryan (ph) who gave me a bone marrow transplant on December 16, 2003. Everything was

going good until about March when I went in for another bone marrow test. And my doctor said inquiring minds want to

know and they did a test, and it came back positive with leukemia again. He gave me a 10 percent chance of recovery, and he wanted to know if -- I asked him if there was anything that I could do. He said they were looking for people that would And I thought that

be willing to take experimental drugs. was a real good idea.

So I got some Cloretazine with tThank you. All right. Thank you.

cells, and it's worked. DR. ECKHARDT: MS. VESEY:

The open public hearing portion of

107

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

this meeting has now concluded, and we will no longer take comments from the audience. The committee will now turn

its attention to address the task at hand, the careful consideration of the data before the committee as well as the public comments. DR. ECKHARDT: discussion. Okay. So now we're into our own So

And we'll be having the voting after that.

willing to take any comments, I just want to start out by saying I guess I'm, you know, there's a little bit of a dilemma here. I think there's some similar concerns with With regards to one

regards to population selection.

population that appears to have risk factors that really just include a PS of 2 and an age over 70, you know, I think the clear question there is whether or not this is a population where you would want to compare to more intensive therapy. And then you have the other population which I think is much more -- it's reasonably well-defined in terms of, in particular, comorbidities and enhanced or poor-risk patients. And certainly that, I think more than anything, The concern there is

you can see is an area of need.

really whether or not we have the risk-benefit profile that

108

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

really supports the drug in that population. So, in my view, you know, it's a dilemma with regards to again defining the population and having done that, whether or not a single-arm study supports the indication in that population and whether or not, in fact, the risk-benefit is in the patient's favor. comments? DR. FLATAU: I apologize in advance, this is a Other

little bit of a rant, but I mean, it's clear today, we've seen that AML is a terrible disease, and AML in older patients is a, you know, a really horrible disease. treatments we have are far from ideal. The

And, you know,

treatments that might potentially cure someone or produce a long-term remission have truly frightening toxicities and treatment-related deaths. And what we seem to be told over and over again, not just by this sponsor but by others is that anything is better than what we have now. see that. And I guess I don't really

And I think that it's pretty clear that although

laromustine might actually have some benefits for patients with AML that it really in no significant way changes the outlook for this disease.

109

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

DR. ECKHARDT: DR. LINK:

Dr. Link.

I had a question about the questions, So, there are times when

so I guess this is for the FDA.

you say that it's approved, but with sort of conditions or, you know, attached, in terms of close monitoring of what happens, who gets it, where it's distributed and those sorts of things. And I'm wondering if that can be put in

as an amendment or, you know, in other words, under special circumstances? DR. PAZDUR: You have to demonstrate safety and

efficacy before we do that in the intended population. Yes. The indication can reflect the study population that

was studied. DR. LINK: There's a comment on page 8 of the FDA

thing that says, I was just curious to what you really meant by this, is the broader question of whether the agency should be relying on complete response rate or an improvement in complete response rate as a surrogate endpoint for approval in the setting of acute leukemia. You know, so you sort of throw that out, and you didn't say anything about that any more. table here? But is that really on the

110

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

DR. KANE:

I think that our concern that we were

trying to express on the basis of the 037 study is the fact that there appeared to be a discordance between an improvement in the CR rate and yet an inferior survival for the total population. DR. LINK: And simply -Most leukemia studies show that

getting into remission though leads to a longer survival and better quality of life, is that not accepted yet or -DR. PAZDUR: Yes, it's accepted. But on the

other hand, what we were trying to, I think, phrase there and perhaps we didn't do it well, is that one has to take a look at the other flip side of that and that is the toxicity of the agent. Could I raise a question? Because I'd like some

discussion on this, and this is one of the perplexing things that have come into my mind. And I really need to The

have a clear discussion of the committee by this.

sponsor has, whether we talk about accelerated approval, regular approval, or whatever, the sponsor has the obligation to demonstrate safety and efficacy. It's not

the role of the FDA to prove that the drug is unsafe, that's for health supplements, et cetera. But they have

111

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

the obligation to prove that the drug is safe and effective. Usually we've always been discussing the efficacy side of this question, whether efficacy has been demonstrated. The problem that we have here is that

survival is the most important efficacy endpoint, but it's also the most important safety endpoint. forget about this. And frequently we

As I stated in my opening comments, it

can only be evaluated in randomized studies, okay, that clearly pits one patient group against another and takes a careful look at this in a very prospective manner, randomizing the various variables that would lead to differences in prognosis. We only have one randomized trial here which shows an increased death. And when I take a look at those

survival curves, throughout the entire course of those two survival curves, the addition of this drug, the combination is always inferior to the placebo, it's always below. I'm

not saying there's statistical significance there, but it's always below that. You could always talk about But, you know, attributions of Generally in unblinded,

attributions of death.

death is a very tricky issue.

112

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

randomized trials we don't talk about this.

And even when

we have blinded trials we have a great difficulty in looking at attribution of death. unblinded, single-arm trial. Now, it would like to be discussed that this all due just to the combination of these two drugs. there's three possibilities here. But Here we have an

One, it could be due to

chance alone that the survival curve for the addition of laromustine to cytarabine is there, that probably isn't the case. It could be due to the combination. It could be due

to the drug itself, laromustine. possibilities here.

Those are three

In the absence of any other randomized data, I guess the question that I have for you, which is one that ultimately we have to grapple with, has safety been demonstrated on the most important safety endpoint, and that is overall survival. DR. ECKHARDT: DR. WILSON: Dr. Wilson. And I want

So, I think that is key.

to go back to the class of agents because this class among chemotherapy agents really stands out. It is a class that

is well known to have long-term effects both in terms of

113

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

its effects on stem cells and, therefore, more myelosuppression later on, even if you come in with a different agent and long-term effects on the pulmonary parenchyma as well. And that's what got me concerned is that this is a unique drug class where these kinds of long-term effects are known, and I am concerned, given the randomized study, that they may be operable, even in a sequential study such as the ones that have been put forth. And in the absence

of a randomized study, I still have a question in my mind as to whether or not the survival may be adversely impacted by this agent. I'm not saying it would. But this is a

very unique drug class. DR. ECKHARDT: Well, I think in particular where

you've got a population that has underlying risk factors for development of toxicity, it's very difficult in a nonrandomized manner to really ascribe exactly what the attributions of these toxicities are. I think it's a clear

case for why you need to do randomized studies to really dissect out exactly what is related to underlying disease, complications of treatment versus a new therapeutic. Dr. Sekeres.

114

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

DR. SEKERES:

I think the crux of the question I was

you're asking is what the patient's goals are.

trying to get at this, perhaps not elegantly enough, with the dividing into three potential treatment arms. So, if a

patient wants a chance of cure and wouldn't ordinarily go for remission induction chemo, then these toxicities, I'm actually comfortable with as a leukemia doc. If the patient is going not for cure but for prolongation of survival, then it's incumbent upon us, us as an oncology community, to define what's an acceptable prolongation of survival. Now, Dr. Löwenberg and in addition, Dr. Tilly (ph) have performed the only two prospective randomized control trials, randomizing patients with -- older adults with AML to either intensive chemotherapy or low-dose cytarabine or at best supportive care, and that was back in 1989 and 1990. The prolongation of survival in the only

one of those that had a significant difference was 10 weeks. Interestingly, if we kind of go through the data that's been presented today where we have a median survival of 3.2 months overall and potentially a prolongation of

115

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

survival of a couple months, maybe even more than that, you could say, "Gee, it seems as if they're about meeting that bar." What I'm struggling with is have they gone beyond that bar, there aren't any patient-reported outcomes, so we don't know if patients actually receive some salutary benefit to this. And I'm still not convinced that this

population represents a poorer risk population than I would treat in Cleveland with induction chemo. DR. PAZDUR: The crux of my question has to deal

with can we be doing more harm than good. DR. ECKHARDT: DR. FLEMING: Dr. Fleming. That's a critical question. And if

our principal measure of good is a response, then we're using, in essence, as a surrogate endpoint for ultimately having a patient -- having evidence that the patient is living longer or that we're dealing effectively with disease-related symptoms. And showing that responders live

longer than nonresponders is not beginning to be a validation of response as representing treatment effect on clinical endpoints. What you need is an array of trials that

116

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

consistently show the treatment effect on response is accurately predicting treatment effect on survival. 037 shows exactly the opposite for this very agent. Short of that, what you need is evidence that indicates that you have an extremely strong effect of major magnitude and duration so the agency has at times used a complete response measure when it's a very high rate for a very long duration. 28 percent. In this setting, the CR, CRp rate is And

And only 18 percent unfavorable cytogenetics

the duration is 5.7 months under the second trial, 3.6 months. So those are more modest overall levels. Those

wouldn't qualify for what I would be saying would be profound. The other issue is those do not begin to capture the off-target effects. And that's part of what you're

getting at properly, Dr. Pazdur, is that you have to balance the on-target effects and the off-target effects. And there's no way to know what those off-target effects are. And it's made worse in this setting because the

single-arm trial is in a population where the FDA has shown 29 to 38 percent of these patients may well have qualified for other effective background therapy. And this agent is

117

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

being studied in 043 with cytarabine as consolidation and HU in 033. therapies. But the biggest concern for me is this uncertainty about risk. Now, Dr. Giles was getting excited So it's also being confounded by other

about the fact that there was a 14 percent, 30 day mortality, yet the median survival is only 3.2 months and 3.4 months. And so the data aren't pointing to the fact

that the lower 30 day is translating into a meaningfully prolonged survival. And in the absence of a control, how

do we possibly argue that that's in fact better than what we would have expected had we used an appropriate alternative standard. But the essence of this is you have the 037 trial. It's not in exactly the same setting. But the 037

trial is providing very strong evidence of off-target effects that are affecting mortality and that in that trial are offsetting the doubling in response rate. Now does that apply to the setting of 043 and 044? It's not entirely certain that it applies. And yet

the evidence indicates that you have background rates of these infections and these pulmonary events that are fatal

118

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

that are not that dissimilar to 037.

And then when you

look -- we keep showing slide 44 without saying what slide 44 is saying is when you're looking at these pulmonary deaths, for example, by those with no pulmonary dysfunction against pulmonary dysfunction, they're not all lying in the category of patients with pulmonary dysfunction at baseline. And so there isn't a clear explanation. This

seems to be extraordinarily, strongly arguing for a proper randomized control to understand what is the true nature of the aggregation of on-target effects and off-target effects relative to a proper control where you can also understand laromustine's contribution to that. DR. ECKHARDT: Dr. Richardson. This morning Dr. Flatau asked And I think we're seeing a We see data

DR. RICHARDSON:

how low can the standards go.

little bit further evolution on that scale.

from a single-arm study of 85 patients with supporting data that were developed by data mining some of the old studies. You look at a remission rate of about 30 percent, half of these, less than six months in duration. And about a third

of these, I think you could measure with your watch, less

119

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

than 90 days in duration. I think one can argue that lung toxicity is simply part of the treatment complications of these patients, but I think the randomized study that Dr. Fleming has described in detail strongly points the finger at laromustine as a cause of pulmonary toxicity in this group. And I wonder why one would want to use a drug like this in this population in the first place, those with pulmonary problems. And here I'm thinking of another pulmonary toxic drug that is used in my practice, I see a number of folks with germ cell malignancies, and we use bleomycin, which is a drug that has known pulmonary toxicity. Most oncologists

treating this particular type of malignancy would not use this drug in patients older than age 50 because of increased risk of pulmonary problems. So, I think if this drug is going to have a role in the treatment of AML, it should be on the basis of some adequate data looking at some type of randomized controlled study in this population. DR. ECKHARDT: DR. BALDUCCI: Dr. Balducci. I, of course, share most of the

120

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

concerns that were exposed during this discussion.

I would

like to make two points, however, first of all, Dr. Fleming pointed out that the median survival was 3.2 months. However, I was impressed by the fact that one-year survival was almost 30 percent was 28 percent. Let's remember that

there are several different positions in the survival where we -- where if we were looking at the median survival, we probably would have never had chop because originally chop of the response rate of about 40 percent. But of the 40

percent are responded possibly 60 or 70 percent were cured. So, I think that, you know, we have to look at the complex of the issue in general. The second thing is

these, you know, I am disturbed by the discussion about the bar and how low we put the bar because we are dealing with a very serious population of patients that has never been addressed before. And I want to compliment Vion and the

company before, Genzyme, to try to address this problem that has virtually never been addressed before. I mean, I -- if you see somebody who is drowning, you are not asking whether the lifesaver was certified, but I know many agents, you throw the white lifesaver, that's the first thing that you are trying to do now, not all of

121

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

these patients are drowning for sure.

But I think we have

to have -- I feel very strongly as I am trying to point out that we have to have a mechanism to study these patients. Maybe this was not the best mechanism. to lower the bar. Thank you. Dr. Barrett. Well, I agree a lot with what Dr. But I think we have

DR. ECKHARDT: DR. BARRETT:

Fleming said that really the 037 trial really gives you pause for thought that albeit in slightly different situation it would be very hard to go ahead with extending something more like the 043 without doing a randomized study, but perhaps moving away from the pulmonary toxicity. The other thing that concerns me is the short median survival because clearly, the rationale if you're using an agent like this would be to get rid of stem cells. And if you are going to see an effect of that, you would expect, perhaps, to see similar remission rates to what you might expect, but that the patients would stay in remission longer because you got rid of the residual stem cells that are responsible for relapse. We don't really see that in this study. So, you

know, I'm concerned, therefore, by the efficacy of this

122

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

agent in addition to the much-discussed pulmonary complications. DR. ECKHARDT: DR. FLATAU: All right. It seems to me -- yes.

I'd just like to commend Vion and I mean,

also Genzyme for addressing this population.

clearly, it's an unmet need, but it seems to me that we're taking the choices between someone who's drowning and do we throw them a big rock or a little rock, you know, we're not helping them. And you know the survival is still, the And

long-term survival is, you know, 5 to 10 percent. that's still dismal. DR. ECKHARDT: Thank you.

I think at this point

we'll go on with the vote.

And I'll go ahead and read it.

Similar to the previous one, again, this is a randomized -should a randomized study defining the efficacy and safety of laromustine in the population proposed for the indication be completed prior to approval of laromustine? We had a little bit of confusion on the last vote. So, let me define, "yes" is basically saying that a "No" means that

randomized study does need to be done.

you're satisfied with the level evidence presented without a randomized study.

123

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

So, we'll go ahead and take the vote. (Pause for voting) DR. ECKHARDT: Okay. I'll read the results. 13

"yes," 0 "no," 0 "abstained."

And I'm assuming everybody So, what we'll do now is

got their vote correct this time.

go around starting -- we'll start on my right this time, stating your name and briefly why you voted the way you did. DR. SEKERES: Yeah, Mikkael Sekeres, as I already

said, I had difficulty defining how I would clinically use this drug and whether this population really differed from patients I would treat with induction chemo. DR. FREEDMAN: Ralph Freedman, my concern is

about the toxicity in the 037 trial, that is not resolved in my mind. And I think that data is important for us to

go further with other trials with this drug. DR. FLEMING: Thomas Fleming, I voted "yes" for

reasons that I've already articulated. DR. FLATAU: Arthur Flatau, I voted "yes." I

just don't see the evidence being sufficient to support approval or support the drug efficacy. MS. MASON: Virginia Mason, same things as have

124

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

been said.

Thank you. DR. WILSON: Wyndham Wilson, same things I've

said before.

This is a drug class you have to be very

cautious with, and I think it requires a randomized study. DR. ECKHARDT: Gail Eckhardt, I voted "yes"

really based both on the durability as well as the safety issues. DR. RICHARDSON: Ron Richardson, I voted "yes"

for the same reasons outlined by Dr. Eckhardt. DR. LYMAN: Gary Lyman, I voted "yes." I think

this says -- the discussions illustrated exactly why a randomized control trial is really needed in this setting to assess both the efficacy side and the safety or lack thereof of a potential therapy in this setting. DR. TEMPERO: I voted "yes." And I really can't

add anything more to what's already been expressed by others. DR. LINK: same reasons. Michael Link, I voted "yes" for the

Although, it should be stated that the drug

is of interest, it would be nice to figure out for sure that the safety signal is not an issue. DR. BALDUCCI: Lodovico Balducci, for me it was a

125

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

very difficult vote because I could see the merit of the drugs. But I voted "yes" because I really could not find

enough justification in terms of efficacy and safety to go ahead and recommend approval of the drug, although, as I said, I compliment the companies that are getting involved in these studies, and I hope that that will not discourage future studies of this important issue. DR. BARRETT: John Barrett, yes, safety, And perhaps in a

efficacy, as everybody has said.

randomized study it would be nice to see really good, solid baseline pulmonary function data because one would hope that maybe a group of patients that really would benefit from this interesting agent. DR. ECKHARDT: meeting is adjourned. (Whereupon, at 4:31 p.m., the meeting was adjourned.) All right. Thank you, this

Information

VIRGINIA:

125 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

26330


You might also be interested in

BETA
untitled
Original