1

DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

CENTER FOR DRUG EVALUATION AND RESEARCH

DERMATOLOGIC AND OPHTHALMIC DRUGS

ADVISORY COMMITTEE

Thursday, May 23, 2002

8:30 a.m.

Kennedy Ballroom

Holiday Inn

8777 Georgia Avenue

Silver Spring, Maryland

PARTICIPANTS

Lynn A. Drake, Acting Chair

Karen M. Templeton-Somers, Executive Secretary

MEMBERS

Elizabeth A. Abel, M.D.

Roselyn E. Epps, M.D.

Robert Katz, M.D.

Lloyd E. King, Jr., M.D., Ph.D.

Paula Knudson (Consumer Representative)

Sharon S. Raimer, M.D.

Ming T. Tan, Ph.D.

CONSULTANTS (VOTING)

Warwick L. Morison, M.B., B.S., M.D.,

M.R.C.P.

Seth R. Stevens, M.D.

J. Richard Taylor, M.D.

GUEST (NON-VOTING)

Robert Swerlick, M.D.

CBER, FDA

Ezio Bonvini, M.D.

Louis Marzella, Ph.D.

Jay Seigel, M.D.

Karen Weiss, M.D.

C O N T E N T S

Call to Order and Opening Remarks:

Lynn A. Drake, Acting Chair 4

Introductions 5

Conflict of Interest Statement:

Karen M. Templeton-Somers, Ph.D. 7

BLA 125036, alefacept, Biogen, Inc.

Introduction:

Ezio Bonvini, M.D. 19

Sponsor Presentation, Biogen, Inc.

Overview:

Burt Adelman, M.D. 14

Clinical Experience:

Akshay K. Vaishnaw., M.D., Ph.D. 28

Clinical Safety:

Goria Vigliani, M.D. 59

Alefacept Risk Benefit Profile:

Mark Lebwohl, M.D. 77

FDA Presentation

Louis Marzella, M.D., Ph.D. 90

Questions from the Committee 126

Open Public Hearing

Gail M. Zimmerman 202

Diane Lewis 205

Maryellen Crawford 210

Sean Morton 212

Alan Menter, M.D. 215

Committee Discussion and Vote 222

1 P R O C E E D I N G S

2 Call to Order and Opening Remarks

3 DR. DRAKE: Hello. My name is Lynn Drake.

4 I am from Harvard Medical School, the Massachusetts

5 General Hospital. I am pleased to be the chair of

6 this meeting.

7 The first thing I would like to do is open

8 the meeting. This is the Dermatologic and

9 Ophthalmologic Drugs Advisory Committee. First of

10 all, I would like to welcome all the members of the

11 committee. As you know, you had fairly extensive

12 briefing documents. You have had to take a lot of

13 your personal time to review all this and take your

14 time to come here today. We are so appreciative

15 that you have given that volunteer time to help

16 review the product before us today.

17 I would like to thank the FDA staff, the

18 whole team. The briefing documents were actually

19 very well done. They were concise. They were easy

20 to read and it was clear that effort had been put

21 into it. So I do want to thank the whole FDA staff

22 and team for giving us such a nice group of

23 documents to work from. The preparation was

24 obvious.

25 I would also like to thank the sponsor for

1 bringing forward a new drug. You know, we have

2 patients with bad disease and we are always

3 appreciative that you take the time to try to

4 develop a new drug that will help our patients. So

5 we are very grateful to you for bringing forth this

6 new drug.

7 I also would like to welcome all the

8 guests who are here today. I think public interest

9 in the proceedings in important and significant and

10 so we are grateful. I am particularly pleased that

11 we have some documented participants in the open

12 public hearing. That is delightful to see because

13 we don't always have that and that kind of input

14 just makes us do our job better.

15 So, having said all that, the first person

16 I would like to introduce is Dr. Karen Templeton-Somers, my

17 Executive Officer for this. She has

18 done a yeoman's amount of work. You can't imagine.

19 Karen, I would like to thank you very much in

20 advance for all the work you have done and all the

21 help you are going to give me today. She keeps me

22 out of trouble. In case you guys don't know what

23 she does, her primary job is to keep me out of

24 trouble from here on out.

25 The first thing I would like to do so that

1 everybody knows who everybody is, I would like to

2 go around the table, have the committee members

3 introduce themselves sand your affiliation. I

4 would like to start with Dr. Swerlick.

5 One of the rules--we have these ridiculous

6 rules here. We have to speak into the mike.

7 Introduction of Committee

8 DR. SWERLICK: Robert Swerlick. I am an

9 Associate Professor of Dermatology at Emory

10 University.

11 DR. TAYLOR: Richard Taylor. I am

12 Professor at the University of Miami and Chief of

13 Dermatology at the Miami V.A. Hospital.

14 DR. ABEL: Elizabeth Abel. I am Clinical

15 Professor of Dermatology at Stanford in California

16 and in private practice in Mountain View.

17 MS. KNUDSON: I am Paula Knudson. I am

18 the IRB Coordinator for the University of Texas

19 Health Science Center in Houston.

20 DR. STEVENS: I am Seth Stevens. I am

21 from University Hospitals of Cleveland. I am Chief

22 of Dermatology at the Cleveland V.A. and at Case

23 Western Reserve University.

24 DR. KATZ: I am Robert Katz, in the

25 private practice of dermatology in Rockville,

1 Maryland, Clinical Associate Professor of

2 Dermatology at Georgetown University Hospital.

3 DR. TEMPLETON-SOMERS: Karen Somers,

4 Executive Secretary to the committee, FDA.

5 DR. MORISON: Lloyd Morison, Professor of

6 Dermatology at Johns Hopkins University.

7 DR. EPPS: Dr. Roselyn Epps, Chief of the

8 Division of Dermatology, Children's National

9 Medical Center which is affiliated with George

10 Washington University.

11 DR. KING: Lloyd King, Chief of

12 Dermatology at Vanderbilt University and at the

13 National V.A.

14 DR. TAN: Ming Tan, Associate Member of

15 Biostatistics, St. Jude Children's Research

16 Hospital.

17 DR. RAIMER: I'm Sharon Raimer, Chairman

18 of Dermatology at the University of Texas in

19 Galveston.

20 DR. BONVINI: I am Ezio Bonvini, Division

21 of Monoclonal Antibodies, Center for Biologics.

22 DR. MARZELLA: I am Louis Marzella,

23 Division of Clinical Trials in the Center for

24 Biologics.

25 DR. WEISS: Karen Weiss, Division of

1 Clinical Trials, Center for Biologics.

2 DR. DRAKE: Terrific. Next, I would like

3 to ask Dr. Somers to please inform us about our

4 conflict of interest statement.

5 Conflict of Interest Statement

6 DR. TEMPLETON-SOMERS: The following

7 announcement addresses conflict of interest with

8 regard to this meeting and is made a part of the

9 record to preclude even the appearance of such at

10 the meeting.

11 Based on the submitted agenda for the

12 meeting and all financial interests reported by the

13 committee participants, it has been determined that

14 all interests in firms regulated by the Center for

15 Drug Evaluation and Research present no potential

16 for an appearance of a conflict of interest at this

17 meeting with the following exceptions.

18 Dr. Ming Tan has been granted waivers

19 under 18 U.S.C. 208(b)(3) and 595(n)(4) of the FDA

20 Modernization Act for his ownership of stock in a

21 competitor. The stock is valued at between $5,001

22 to $25,000. Dr. J. Richard Taylor has been granted

23 waivers under 28 U.S.C. 208(b)(1) and 505(n)(4) of

24 the FDA Modernization Act for his employer's

25 contract with a competing firm. The value of the

1 contract is less than $100,000 per year.

2 These waivers permit Dr. Tan and Dr.

3 Taylor to participate in the committee's

4 deliberations and vote considering Biologic License

5 Application Submission Tracking Number 125036,

6 Amevive, alefacept, sponsored by Biogen,

7 Incorporated.

8 A copy of these waive statement may be

9 obtained by submitting a written request to the

10 agency's Freedom of Information Office, Room 12A30

11 of the Parklawn Building.

12 With respect to FDA's invited guest, Dr.

13 Robert Swerlick has a reported interest that we

14 believe should be made public to allow the

15 participants to objectively evaluate his comments.

16 Dr. Swerlick has a financial interest in Immunex

17 and Enbrel.

18 In the event that the discussions involve

19 any other products or firms not already on the

20 agenda for which an FDA participant has a financial

21 interest, the participants are aware of the need to

22 exclude themselves from such involvement and

23 exclusion will be noted for the record.

24 With respect to all other participants, we

25 ask in the interest of fairness that they address

1 any current or previous financial involvement with

2 any firm whose products they may wish to comment

3 upon.

4 Thank you

5 DR. DRAKE: Thank you, Dr. Somers.

6 We have a very packed agenda today. There

7 is a lot of information to be imparted. I will ask

8 the presenters to please stick to your allotted

9 time. If you go over, I will probably have to try

10 to signal you in some capacity because I want to

11 make sure we have plenty of time at the end for the

12 really important stuff.

13 I would also remind the committee that

14 brevity is wonderful and I will try to remember

15 that same rule, myself. So if we can keep

16 everything as concise as possible, we will move

17 through the agenda and accomplish everything.

18 With that, let's start. I think the first

19 presenter is Dr. Bonvini from the Division of

20 Monoclonal Antibodies, Office of Therapeutics

21 Research and Review.

22 Dr. Bonvini, welcome.

23 BLA 125036, alefacept, Biogen, Incidence.

24 Introduction

25 DR. BONVINI: Good morning.

1 [Slide.]

2 Madame Chairman, distinguished members of

3 the advisory committee, ladies and gentlemen, good

4 morning.

5 On behalf of the Center for Biologics, I

6 would like to thank you for your participation in

7 today's discussion of alefacept for the treatment

8 of chronic plaque psoriasis.

9 [Slide.]

10 My duty today in the next few minutes is

11 to introduce you to the BLA review committee and

12 introduce the molecular entity under discussion and

13 provide a brief immunological background for the

14 discussion of the clinical data for alefacept. I

15 am Ezio Bonvini and I serve as the Chairman and the

16 product review for alefacept.

17 The clinical review was the responsibility

18 of Lou Marzella and Electra Papadopoulos.

19 Pharmacologic and toxicology review were performed

20 by Laureen Black and David Green. The statistical

21 review was performed by Chao Wang. Bioresearch

22 monitoring supervision was under the responsibility

23 of Jose Tavarezpagan. Establishing and

24 manufacturing review for alefacept was alefacept

25 was the responsibility of Chiang Syin and Carol

1 Rehkopt. I would like to acknowledge the excellent

2 regulatory management of Beverly Connor and Lori

3 Tull.

4 [Slide.]

5 The molecule for today's discussion is

6 alefacept, also known as Amevive and also

7 identified in a number of publications as LFA3Tip.

8 Alefacept is a fusion protein comprising the human

9 LFA molecule fused with the human IgG-1 FC portion.

10 This molecule dimerizes through the disulfate bond

11 mediated via the IgG portion of the molecule.

12 As a background to introduce the

13 immunosuppressive mechanism of alefacept, I will

14 briefly review how T-cell activation occurs.

15 [Slide.]

16 The activation of T-lymphocyte is a

17 complex mechanism that is centered on the

18 recognition by the clonotypic T-cell receptor of

19 antigen. Now, that doesn't occur in soluble form

20 and the recognition by the T-cell receptor occurs

21 in the context of the major histocompatibility

22 complex of antigen-presenting cells. In addition

23 to the clonal T-PIC receptor, the interaction is

24 assisted by an invariant component, the CD8 or CD4

25 which interacts with the MHC Class 1 or 2

1 respectively.

2 The interaction with the T-cell receptor

3 and the antigen is a low affinity. For a stable

4 association to occur, other molecules intervene and

5 these are called accessory molecules. A critical

6 accessory molecule for the interaction of T-cell

7 with antigen presenter cells is C28 on the surface

8 of T-cells which interact with B7.1 and B7.2.

9 [Slide.]

10 But, additional molecules are also

11 involved in mediating this interaction and they

12 include a number additional molecules among which

13 LFA3 is one which interacts with CD2 on the surface

14 of T-cells.

15 Now the combination of signal via the T-cell

16 receptor and the costimulatory molecules lead

17 to a productive response resulting in lymphokine

18 secretion such as IL2, interferon, and a number of

19 chemotactic lymphokines such as IL8 which lead the

20 T-cell expansion and may be involved in the

21 proinflammatory process underlying the disease

22 under consideration with kerotinocyte proliferation

23 and differentiation.

24 [Slide.]

25 Alefacept can interfere with this

1 mechanism in the context of this complex

2 interaction by either scavenging the physiologic

3 interaction of LFA3 with CD2, by itself engaging CD2

4 Now, in addition to this competitive

5 mechanism which occur at affinities relatively low

6 and similar to those involving the interaction of

7 endogenous LFA3 with CD2, another mechanism is

8 involved and that is the redirection of a second

9 class of cells, the macrophages and NK cells, via

10 engagement of the Fc receptor through the Fc

11 component of the alefacept fusion protein. This

12 delivers a signal which induces activation of NK

13 cells which delivers a lethal hit.

15 The susceptibility to NK-mediated lysis of

16 the cells may be different

17 depending on the subtype of cells under

18 consideration.

19 While the exact mechanism of the

20 susceptibility of T-cells to alefacept-mediated

21 lysis is not fully understood, the T-cell depletion

22 induced by alefacept and its potential for

23 competition with endogenous LFA3-CD2 interaction

24 are central to our discussion of the clinical

25 activity of alefacept and will be touched upon by

1 Dr. Marzella and Biogen in their review.

2 [Slide.]

3 CD2 is expressed prevalently on T-lymphocytes and

4 there is expression on NK cells. B-lymphocytes are largely

5 negative for CD2 expression

6 with only some precursors in the bone marrow being

7 positive.

8 [Slide.]

9 The concludes my brief introduction on the

10 immunological background. I need to remind this

11 committee that we are still addressing some

12 outstanding issues pertaining to the manufacturing

13 of alefacept that remain to be resolved. The

14 agency and Biogen are working close together and

15 are trying to address this issue in a timely

16 fashion.

17 I think I stuck to my time. This

18 concludes my presentation. I could take questions

19 or just give the podium to Biogen.

20 DR. DRAKE: I think you did a great job.

21 Do any of the committee members have a pertinent

22 question about the presentation? I'm sure we will

23 have some later. Thank you, sir.

24 DR. BONVINI: Okay.

25 DR. DRAKE: I think we have a latecomer to

1 the meeting, but we are delighted. Dr. Seigel, I

2 presume?

3 DR. SEIGEL: Yes.

4 DR. DRAKE: Welcome. We are delighted to

5 have you here.

6 DR. SEIGEL: Thank you. Pleased to be

7 here.

8 DR. DRAKE: I had just complimented you

9 and your team for a very nice presentation of the

10 documents. We are very grateful when it is so well

11 done.

12 DR. SEIGEL: Thank you very much.

13 DR. DRAKE: Moving forward, now it is time

14 for the sponsor which is Biogen for their

15 presentations. I believe the overview will be

16 given by Dr. Adelman.

17 Sponsor Presentation, Biogen, Inc.

18 Introduction

19 DR. ADELMAN: Thank you, Madame

20 Chairwoman. Good morning, members of the panel,

21 colleagues from CBER and members of the audience.

22 [Slide.]

23 My name is Burt Adelman. I am the

24 Executive Vice President of Research and

25 Development at Biogen. Much of our research

1 efforts at Biogen are focused on understanding

2 autoimmunity and developing therapeutic strategies

3 to treat autoimmune diseases. Today, as a result

4 of these efforts, we are pleased to be here to

5 discuss alefacept, a new agent that we have

6 developed for the treatment of chronic plaque

7 psoriasis.

8 [Slide.]

9 Our presentation will focus on data that

10 we believe supports the following indication.

11 Alefacept is indicated for the treatment of

12 patients with chronic plaque psoriasis who are

13 candidates for systemic or phototherapy. Alefacept

14 is a parenteral agent and we recommend a dosing

15 regimen as listed here, once per week dosing for 12

16 weeks.

17 The drug can be administered either as a

18 7.5 milligram intravenous bolus injection once a

19 week or a 15 milligram intramuscular injection once

20 a week. Repeat courses can be given after a 12-week rest

21 period.

22 [Slide.]

23 Our agenda this morning is listed here. I

24 will provide a brief overview of the product. Dr.

25 Akshay Vaishnaw, Medical Director at Biogen, will

1 talk about the clinical efficacy of the alefacept

2 and describe the pharmacodynamics. Dr. Gloria

3 Vigliani, Vice President of Medical Research at

4 Biogen will speak about the clinical safety

5 profile. Finally, we have invited Dr. Mark

6 Lebwohl, a distinguished expert in the field of

7 psoriasis to provide a perspective from the

8 clinical view on the risk-benefit profile of

9 alefacept.

10 [Slide.]

11 In addition to Dr. Lebwohl, we are

12 fortunate to have with us a number of other

13 distinguished consultants. These include Dr.

14 Richard Cooper, a hematologist, Professor of

15 Medicine at the Medical College of Wisconsin; Dr.

16 David Margolis, Associate Professor of Dermatology

17 and Epidemiology at the University of Pennsylvania

18 and Dr. James Krueger, Associate Professor and

19 physician at the Rockefeller University. Dr.

20 Krueger heads the Laboratory of Investigative

21 Dermatology at that Institution.

22 Although they will not be making formal

23 presentations, they are here to help with the

24 discussion and answer any questions that may arise.

25 [Slide.]

1 Now, to begin my review. Chronic plaque

2 psoriasis is recognized to be a T-cell mediated

3 disease. Men and women are affected equally.

4 Although it is recognized that there is a strong

5 genetic component to this disorder, the exact genes

6 that drive the disorder have yet to be identified.

7 In appearance, the skin lesion of

8 psoriasis is a circumscribed red raised plaque.

9 These plaques are often itchy and scaly and can

10 crack and bleed. Psoriasis can also be associated

11 with a number of systemic manifestations, the most

12 common of which is psoriatic arthritis.

13 Individuals with moderate to severe psoriasis

14 typically have lesions covering 10 percent or more

15 of their body-surface area. As you will have seen

16 in the briefing document that we distributed, a

17 number of the patients in our studies actually had

18 skin involvement of up to 98 percent of their body-surface

19 area. Psoriasis is a life-long disease

20 and, as yet, there is no cure.

21 [Slide.]

22 Here is a picture of the disease that we

23 are speaking about. This is a patient from one of

24 our Phase 3 studies, a gentleman with moderate to

25 severe chronic plaque psoriasis. It is not hard to

1 understand that this disease, in addition to the

2 clinical manifestations, has a debilitating impact

3 on a patient's life.

4 John Updike, in his essay, At War with My

5 Skin, describes poignantly his own personal

6 experience with psoriasis. "They glance at me and

7 glance away pained. My hands and my face mark me.

8 The name of the disease, spiritually speaking, is

9 Humiliation."

10 [Slide.]

11 This statement powerfully captures the

12 psychosocial burden that many individuals with

13 psoriasis suffer. In fact, this has been studied

14 and, to some degree, quantified. Quality of life

15 is identified as being severely impacted in

16 patients with moderate to severe psoriasis. The

17 impact is similar to that of other serious diseases

18 such as chronic congestive heart failure and

19 advanced diabetes mellitus.

20 Understandably, these effects correlate

21 with the increased risk of substance abuse,

22 depression and suicidal ideation commonly seen in

23 the psoriasis population. Common comorbidities of

24 psoriasis include obesity, heart disease, diabetes

25 and hepatitis.

1 For all these reasons, patients and their

2 physicians are often searching for new therapies

3 and patients with advanced psoriasis often seek out

4 and are commonly treated with aggressive therapies.

5 [Slide.]

6 Current therapies to treat chronic plaque

7 psoriasis are listed here, systemic therapies.

8 There are two types. In the upper part of the

9 slide, I have indicated the disease-suppressive

10 therapies. In the lower part are the remittive

11 therapies.

12 The suppressive therapies, methotrexate,

13 retinoids and cyclosporine effectively treat the

14 disease as long as the patient takes them. When

15 therapies are withdrawn, there is usually

16 reasonably rapid return of disease, hence the label

17 suppressive. Remittive therapies such as PUVA an

18 UVB, light-based therapies, can provide disease-free

19 periods. However, to obtain these results,

20 patients must undergo frequent and repeat treatment

21 cycles.

22 Each of these important therapies is

23 associated with one or more toxicity that is

24 significant, commonly observed and often limits it

25 use.

1 [Slide.]

2 For example, methotrexate can cause

3 hepatic fibrosis and patients who receive over a

4 gram and a half of methotrexate often are required

5 to have a liver biopsy to determine whether they

6 can receive additional therapy. Cyclosporine is

7 commonly associated with nephrosis and, therefore,

8 patients cannot take cyclosporine continuously for

9 more than a year.

10 Phototherapy with PUVA has been documented

11 to increase patient risk for squamous-cell

12 carcinoma and melanoma. So, again, significant

13 limitations for therapy.

14 So, while these therapies provide

15 meaningful efficacy, their use also imposes

16 significant risk. In an effort to balance toxicity

17 and maintain reasonable disease control,

18 dermatologists have evolved a strategy of disease

19 management based on rotating the available

20 therapies. Clearly, new therapies, particularly

21 remittive agents that can induce a long duration of

22 effect will favorably impact this strategy of

23 rotational therapy.

24 It is to address this significant unmet

25 need that we have developed alefacept.

1 [Slide.]

2 To understand the rationale behind the

3 development of alefacept as a new immunomodulator,

4 I would like to briefly review the pathobiology of

5 psoriasis. As indicated a few slides ago,

6 psoriasis is clearly recognized to be a T-cell-mediated

7 disorder. In particular, memory T-cell

8 subsets play a critical role in a pathogenesis of

9 the psoriatic plaque.

10 In this section from a skin biopsy of a

11 patient with psoriasis, memory T-cells are seen

12 infiltrating the skin underlying the proliferative

13 response. These active cells are derived from CD4

14 and CD8 cells and are identified by a

15 characteristic cell-surface marker called CD45RO-positive.

16 It can be stained for and these cells

17 can, therefore, be uniquely identified.

18 Once in the skin, again as we see here,

19 these activated CD45RO-positive cells release a

20 spectrum of inflammatory mediators that stimulate

21 kerotinocyte proliferation and blood-vessel growth

22 resulting in the characteristic psoriatic plaque.

23 [Slide.]

24 The cells that I have described can be

25 identified in the blood and in the lymph organs.

1 This cartoon indicates the composition of

2 leukocytes in the blood. You can see that memory

3 CD45RO-positive cells are constituent of the T-cell

4 CD4 and CD8 population within the blood and they

5 can be distinguished from naive cells by this

6 characteristic marker.

7 Our data suggest that alefacept

8 selectively targets CD4 and CD8 memory cells and it

9 does this through its activity against the CD2

10 ligand on memory cells.

11 [Slide.]

12 Dr. Bonvini has taken you through this and

13 with somewhat more elegant slides. Perhaps he will

14 lend them to me in the future. But I will take you

15 through this mechanism again.

16 A naive T-cell that has never previously

17 seen antigen will interact with antigen-presenting

18 cells by way of the MHC and T-cell receptor. But,

19 as already mentioned, this interaction is

20 inadequate to result in T-cell activation and,

21 importantly, costimulatory pathways mediated

22 through coupling of LFA-3 and CD2 and B7 and CD28

23 are also necessary. In fact, this cartoon is,

24 itself, a simplification and there are other

25 additionally important costimulatory pathways.

1 As a result of these events, the naive T-cell

2 becomes activated. During the activation

3 process, a number of characteristic changes occur.

4 The cells proliferate so, in fact, there would be

5 more cells here than just the one and a number of

6 changes occur on the surface. In particular, there

7 is increased expression of CD2 on the surface of

8 these CD45RO-positive cells.

9 This conversion from the CD2 low state to

10 the CD2 high state is what we think imparts the

11 selectivity of alefacept to the CD45RO-positive

12 memory cell.

13 Just, also, by way of historical

14 background, LFA3 was actually cloned at Biogen and,

15 very early on, we understood the significance of

16 the LFA3-CD2 interaction and that is why we have

17 chosen this particular pathway to develop a drug

18 that interferes with this process.

19 [Slide.]

20 Here, again, is a picture of alefacept.

21 As you can see, it includes the extracellular

22 domain of human LFA3 fused to a portion of human

23 IgG1 and is, therefore, called a fusion protein.

24 It is expressed as a dimer which is held together

25 by cysteine bonds and, although it looks like an

1 immunoglobulin, it is not an immunoglobulin. It is

2 a fusion protein.

3 Now, the sequence is entirely human and

4 that is why there is very little antigenicity

5 associated with the use of this.

6 [Slide.]

7 I would like to review again alefacept

8 actions as are illustrated in this slide.

9 Alefacept can block LFA3 CD2 interactions thereby

10 inhibiting reactivation of memory T-cells. As

11 indicated here, alefacept would bind to CD2 and

12 stearically interfere with the docking to an

13 antigen-presenting cell.

14 Again, as Dr. Bonvini already indicated,

15 another effect is also mediated by alefacept.

16 Alefacept combined via the FC receptor on certain

17 cells such as natural killer cells and induce a

18 pro-apoptotic response. This is mediated through

19 the release of a protein called granzyme which

20 initiates apoptosis in the memory T-cell resulting

21 in its loss.

22 This is a generalized model. We believe

23 that this model applies at the doses that are

24 recommended for use to treat psoriasis but there

25 may be specifics about how this mechanism works in

1 the skin, in the blood and in lymph tissue and we

2 are fortunate to have Dr. Krueger here with us

3 today who has studied this extensively and,

4 perhaps, during the question period, he can comment

5 further on the specifics of this effect.

6 [Slide.]

7 This mechanism of action was tested in a

8 blinded placebo-controlled dose-ranging Phase-II

9 study in approximately 230 patients with moderate

10 to severe psoriasis. I have illustrated the

11 results here, in particular looking at the effects

12 on CD4-positive memory cells. So these would be

13 CD45RO-positive cells that are also CD4 positive

14 and CD4 positive naive T-cell, unactivated T-cells,

15 that would not express CD45RO.

16 What you can see--this was a dose-response

17 study. Here is the twelve-week dosing period and

18 this is a twelve-week follow-up period. This is

19 the placebo dose and here are increasing doses of

20 alefacept. You can see that, with increasing

21 doses, there is increased reduction in the number

22 of CD4-positive memory T-cells and the cell counts

23 start to recover after discontinuation.

24 In contrast, there is minimal effect, if

25 any, on the naive T-cells during the same dosing

1 period. It was these pharmacodynamic effects

2 coupled with the clinical effects that we observed

3 in this study that led to the development of the

4 clinical program for alefacept in chronic plaque

5 psoriasis. We are here to discuss those results

6 today.

7 [Slide.]

8 I would like to provide some additional

9 background on the overall program. We have

10 conducted an extensive toxicology program to

11 support alefacept development. In fact, we have

12 completed 35 toxicology studies in nonhuman

13 primates. We are fortunate because a nonhuman

14 primate responds somewhat similarly to humans in

15 that we can observe impacts on T-cell numbers and

16 we can look at the effect that this may have in the

17 lymph nodes and we can watch recovery.

18 For testing purposes, we have used

19 regimens up to 20 milligrams per kilogram IV weekly

20 for one year. This dosing regimen, obviously,

21 greatly exaggerates the recommended dosing regimen

22 in people, both in terms of magnitude of drug

23 delivered and length of continuous exposure.

24 [Slide.]

25 Here I have summarized the results of the

1 toxicology program for you. Alefacept was well-tolerated in

2 these animals. We observed reversible

3 decreases in lymphocyte counts, both in blood and

4 lymphoid tissues. No opportunistic infections were

5 observed in any treated animal and no reproductive

6 toxicity was observed.

7 I would like to comment on one observation

8 that was outlined for you in the briefing document.

9 In a single cyno monkey receiving 20 milligrams per

10 kilogram of alefacept weekly, we diagnosed the

11 occurrence of a B-cell lymphoma. This monkey was

12 part of a long-term treatment study and, as I have

13 mentioned, received a very high dose continuously

14 for 28 weeks.

15 In fact, this dose is the equivalent of

16 622 clinical courses. So we made this observation

17 in the setting of a highly exaggerated dosing

18 schedule. This was the only observation of

19 lymphoma in over 200 animals treated across various

20 preclinical studies.

21 [Slide.]

22 This next slide briefly outlines the

23 clinical program for alefacept which you will be

24 hearing in much more detail later this morning. We

25 have conducted 18 clinical studies and treated

1 1,357 patients with chronic plaque psoriasis and

2 240 healthy volunteers.

3 The core of our presentation focuses on

4 three randomized double-blind placebo-controlled

5 studies in patients with chronic plaque psoriasis.

6 One is a Phase 2 study and the other two Phase 3

7 studies. These studies will be discussed in detail

8 by Dr. Vaishnaw.

9 We, at Biogen, are committed to

10 understanding both the short and long-term safety

11 issues associated with the introduction of

12 alefacept as we would be with any new drug being

13 introduced into the community. We believe that

14 active monitoring of patients on therapy for

15 extended periods of time, even after a product is

16 approved, should be a key component of an

17 integrated, long-term safety and development

18 program.

19 For these reasons, most of the patients

20 coming out of our randomized clinical trials have

21 been given the opportunity to enter into a

22 comprehensive extended safety dosing study. In

23 fact, at this point in time, over 800 patients are

24 currently in extended safety dosing studies.

25 Already, some of these individuals have received as

1 many as five treatment courses over a three-year

2 period of time and it is our intention to extend

3 this program indefinitely and probably to expand

4 it.

5 [Slide.]

6 Because alefacept targets memory T-cells,

7 we have already begun to study its effects in other

8 autoimmune disorders with a T-cell-mediated

9 etiology. Currently, in addition to psoriasis, we

10 are studying psoriatic arthritis, rheumatoid

11 arthritis and sclera derma and, in fact, we

12 summarized for you, in your briefing document, the

13 results of a small study in psoriatic arthritis.

14 [Slide.]

15 Throughout the development history of this

16 program, we have had a close collaboration with our

17 colleagues at CBER. We are grateful to them for

18 their interest and guidance in all aspects of the

19 preclinical, clinical and manufacturing programs.

20 The regulatory history of alefacept is

21 outlined here. In August of 1996, we had a pre-IND

22 meeting with the agency and, shortly thereafter,

23 launched our program in the United States. In

24 1999, and end-of-Phase-II meeting was held to

25 discuss our positive findings. After agreement

1 with the agency on the design of the Phase 3

2 program, we moved forward to begin the studies that

3 we will be discussing today.

4 Now, importantly, the safety database in

5 this document is consistent with ICH guidelines.

6 In July of last year, we met again with CBER to

7 discuss our Phase 3 results and plan for filing an

8 electronic biologics license application. In

9 August of 2001, we filed the application which we

10 are happy to be here to discuss with you today.

11 Now, in March of 2002, we provided the agency with

12 an extensive safety update to this document.

13 Now it is my pleasure to introduce Dr.

14 Vaishnaw who will take you through clinical details

15 of our program. Thank you for your attention.

16 Clinical Experience

17 DR. VAISHNAW: Thank you.

18 [Slide.]

19 Madame Chairperson, members of the panel,

20 ladies and gentlemen, good morning. I am Akshay

21 Vaishnaw. I am a member of the medical team at

22 Biogen. I have been involved with the development

23 of alefacept.

24 I shall be describing two components of

25 the clinical experience to you today, namely the

1 efficacy and pharmacodynamic aspects of the

2 program.

3 [Slide.]

4 I have divided the efficacy part of the

5 presentation beginning with a brief overview of the

6 Phase 2 study and following with a detailed

7 analysis of the Phase 3 studies both the IM and IM

8 protocols. I will then move to a description of

9 the quality-of-life improvement seen after

10 alefacept therapy and close with a discussion of

11 the efficacy in some important subpopulations of

12 patients.

13 [Slide.]

14 There are three randomized placebo-controlled

15 trials that are at the core of the

16 program; a Phase 2 IV study and two Phase 3

17 studies, one by the IM route and one by the IV

18 route.

19 You can see that in the Phase 2 study, we

20 dosed patients on a body-weight basis. Here you

21 can see that is indicated as dosing in milligram

22 per kilogram. Other studies during Phase 2

23 indicated that body weight did not significantly

24 influence the pharmacokinetics of alefacept and,

25 therefore, in Phase 3, we transitioned to the more

1 convenient fixed-dose regimens as indicated here.

2 As you look to the right of this slide,

3 you can see that a substantial number of patients,

4 in fact over 1300 patients, were enrolled in these

5 three studies making them some of the largest

6 chronic-plaque-psoriasis studies ever.

7 [Slide.]

8 The findings from the Phase 2 study were

9 published by Drs. Charles Ellis and Gerry Krueger

10 in an article in The New England Journal of

11 Medicine last year and their major findings were

12 summarized as follows. They detected that

13 alefacept was associated with clinically meaningful

14 efficacy and it was superior to placebo. They

15 determined that it had a significant duration of

16 benefit.

17 Patients that had cleared their disease

18 had a median time to retreatment of ten months.

19 With respect to T-cells, given the mechanism of

20 action, they clearly illustrated that alefacept was

21 selective for reductions in memory T-cells with

22 sparing of naive T-cells. Importantly, these

23 changes correlated with efficacy outcomes. This

24 validated the therapeutic rationale in the approach

25 to Phase 3. Finally, the Ellis and Krueger study

1 allowed us to pick the optimum dose group for Phase

2 3.

3 With that, I want to turn to the Phase 3

4 studies.

5 [Slide.]

6 At baseline in the Phase 3 studies, all

7 the important background demographic and disease-severity

8 factors were well balanced. I want to

9 consider some factors related to disease status at

10 baseline.

11 Here you see data for the two Phase 3

12 studies, the IM and IV protocols. The median

13 duration of disease at baseline ranged between

14 eighteen and nineteen years. In other words, these

15 patients had established chronic plaque psoriasis.

16 If you look at the next three rows, the

17 body-surface area involvement, the PASI score and

18 the physician global, each reveals that patients

19 had moderate to severe chronic plaque psoriasis at

20 baseline.

21 [Slide.]

22 Let me illustrate that by considering the

23 BSA score. The median BSA at baseline ranged

24 between 21 and 22 percent in these studies. Now,

25 if we imagined that one palm size is about 1

1 percent of our body-surface area, then 22 percent

2 average involvement is extensive chronic plaque

3 psoriasis and a significant burden of disease to

4 these patients at baseline.

5 That conclusion is supported by the median

6 PASI score in the mid-15s and the physician global

7 assessment where over 80 percent of patients had

8 disease severity ranging between moderate to

9 severe.

10 [Slide.]

11 I have already mentioned the PASI. PASI

12 will be central to a lot of our discussions

13 regarding efficacy today. PASI is, in fact, an

14 acronym of the Psoriasis Area and Severity Index.

15 It is a widely used tool in psoriasis clinical

16 trials in order to quantify and follow disease

17 activity over time. It is a composite measure and

18 involves measurement of erythema, induration,

19 desquamation and the extent of body-surface area

20 involved.

21 Those four parameters are evaluated over

22 four parts of the anatomy; the head, the trunk, the

23 upper limbs and the lower limbs. Those data are

24 put into a formula resulting in a composite score

25 which ranges from 0 to 72. 0 is clear or healthy

1 skin. 72 is disease of maximum severity.

2 A score between the range of 10 and 30

3 typically summarizes patients with moderate to

4 severe chronic plaque psoriasis.

5 [Slide.]

6 Three endpoints will be discussed with

7 respect to the clinical trials we are reviewing

8 today. These are PASI 75--that is a 75 percent or

9 greater reduction from baseline disease severity

10 with respect to the PASI tool, a very stringent

11 endpoint. The next endpoint is PASI 50, a

12 50 percent or greater reduction from baseline

13 disease severity. Finally, the third stringent

14 endpoint is the physician global assessment of

15 almost clear or clear.

16 These two endpoints were read out both two

17 weeks after the last dose in the studies and also

18 in what we term the overall response rate. I want

19 to illustrate what I mean by that on the following

20 diagram.

21 [Slide.]

22 Here is a typical randomized placebo study

23 comparing placebo to alefacept. On the left-hand

24 part of the diagram, you can see the dosing

25 interval. Patients are receiving injections for

1 the first twelve weeks. On the right-hand side,

2 you can see they are followed for another twelve

3 weeks. That 12-plus-12 interval we term a course

4 of alefacept therapy.

5 Now, the primary efficacy endpoint was

6 conducted as a landmark analysis two weeks after

7 last dose at this single time point. Given that in

8 Phase 2 and in other studies we had determined that

9 alefacept patients often reach maximal efficacy at

10 other times often late in the follow-up interval

11 here, we also determined the overall response rate

12 for patients that achieved PASI 75 and the other

13 endpoints at any time during the course of therapy.

14 [Slide.]

15 Before we actually consider the efficacy

16 data, I want to, with the use of a few pictures,

17 consider what a PASI 50 and PASI 75 response is

18 like. It can be difficult to conceptualize them in

19 the abstract.

20 Here is a patient on the left who, at

21 baseline, has had extensive chronic plaque

22 psoriasis effect from the midline, the area above

23 the buttocks and the backs of the arms. This is a

24 patient with a score of 18.7 by the PASI 2 and

25 baseline. After treatment, the score is 5.7. This

1 patient has an almost 70 percent reduction in PASI.

2 This patient would not qualify for the

3 primary-efficacy endpoint of PASI 75 but would

4 qualify for PASI 50. She doesn't qualify for PASI

5 50. She doesn't qualify for PASI 75 because she

6 has never attained 75 or greater.

7 [Slide.]

8 Contrasting that to the PASI 75 response,

9 on the left you see a young person with extensive

10 disease again affecting the torso and the lower

11 limbs. His score is 34.3 at baseline. After

12 treatment, his score is 4.2. The percentage

13 positive reduction is 88. This gentleman would

14 qualify as a PASI 75 responder.

15 [Slide.]

16 With that background, I want to review the

17 two major studies, first the Phase 3 IM study.

18 [Slide.]

19 In the Phase 3 IM study, patients were

20 screened and randomized to one of three arms,

21 placebo or alefacept 10 milligrams or alefacept 15

22 milligrams. They received the injections once a

23 week IM for 12 weeks on the left-hand side of the

24 diagram and then there was a 12-week follow-up

25 interval. The primary efficacy endpoint was read

1 out as a landmark analysis two weeks after last

2 dose. The primary endpoint was PASI 75.

3 Note that the endpoint was read out

4 without the use of disqualifying medications; by

5 this, I mean major, high-potency topical steroids

6 or the major systemic antipsoriatic agents, and the

7 range of UV therapies that are commonly used.

8 If patients used any of those

9 disqualifying medications prior to the primary

10 efficacy endpoint, they were classified as a

11 treatment failure. If patients did not show up for

12 the primary efficacy-endpoint visit, they were,

13 again, classified as a treatment failure. This is

14 a relatively conservative approach when documenting

15 efficacy data.

16 The rules regarding disqualifying

17 medications also apply to all the other efficacy

18 data we are going to review today.

19 [Slide.]

20 In the Phase 3 IM study, PASI 75 score two

21 weeks after last dose was 21 percent in the 15

22 milligram group and 5 percent in the placebo group.

23 This difference was highly statistically

24 significant and the Phase 3 IM study, therefore,

25 met the primary efficacy endpoint.

1 In the middle you can see that, in the 10-

2 milligram group, 12 percent of patients attained

3 the endpoint contributing to this nice dose

4 response between placebo and 15 milligrams.

5 The findings from this PASI 75 tool was

6 strongly supported by an independent measurement,

7 namely the physician global of almost clear or

8 clear.

9 [Slide.]

10 Here you can see on the right that 14

11 percent of patients in the 15-milligram group

12 cleared their disease versus 5 percent in the

13 placebo group. The difference was highly

14 statistically significant.

15 [Slide.]

16 Finally, the third of the endpoints also

17 supported the conclusion that alefacept was

18 superior to placebo with 42 percent of patients in

19 the 15-milligram group achieving the endpoint, 18

20 percent in placebo. So, over a series of

21 endpoints, all stringent, we have demonstrated that

22 alefacept monotherapy was significantly superior to

23 placebo.

24 [Slide.]

25 I have just conveyed some of the landmark

1 analyses two weeks after last dose. I want to

2 contrast the findings from those to those for the

3 overall response rate where patients were achieving

4 the endpoint at times other than just two weeks

5 after last dose.

6 On the right, you can see patients who hit

7 PASI 75 at any time during a course of therapy as

8 shown with 33 percent in the yellow in the 15-milligram

9 group achieving PASI 75. This is

10 significantly greater than the 21 percent by the

11 landmark analysis.

12 You see increments for all three treatment

13 groups on the right compared to the left, but the

14 data on the right conveyed that these patients in

15 the alefacept group had more sustained responses

16 than those in the placebo group here, and we

17 therefore believe that the overall response-rate

18 data for each of the endpoints we will be

19 discussing today reflect the true clinical

20 attributes of alefacept and what patients can

21 expect to experience in terms of the course of

22 therapy.

23 [Slide.]

24 I am going to turn now to the Phase 3 IV

25 study. Patients were screened here and randomized

1 to one of three arms, Cohort 1, Cohort 2 or Cohort

2 3. All three cohorts received two courses of

3 therapy, as indicated. Each course was 24 weeks

4 long.

5 Cohort 1 received alefacept in the first

6 course followed by alefacept in the second. Cohort

7 2 received alefacept followed by placebo. Cohort 3

8 received placebo followed by alefacept. The

9 primary efficacy endpoint, as for the IM study, was

10 PASI 75 two weeks after last dose, again without

11 the use of disqualifying medications.

12 The advantage of this type of study, apart

13 from the primary efficacy endpoint for the placebo-

14 controlled component of the program here was we

15 could also ask the question, did repeated courses

16 of alefacept result in evidence for incremental

17 efficacy by examining outcomes in Course 2 for

18 alefacept with the outcomes in Course 1.

19 By examining outcomes for Cohort 2 who

20 received a single course of treatment, when they

21 are off therapy for this 36-week period, we could

22 determine how sustained was the efficacy after 12

23 injections. So, with that, let me actually turn

24 now to the data.

25 [Slide.]

1 Here we have summarized the three

2 endpoints we have spoken of, the outcomes two weeks

3 after last dose in the first course. Let's focus

4 first on the far left, PASI 75, which is the

5 primary efficacy endpoint for this study. 14

6 percent of patients in the alefacept group achieved

7 the endpoint, 4 percent in the placebo group. This

8 difference was highly statistically significant.

9 So, again, for the Phase 3 IV study, we met the

10 primary efficacy endpoint as prespecified.

11 The data from the other two endpoints

12 again support the conclusions from the primary

13 efficacy endpoint, the physician global, alefacept

14 11 percent, placebo 4 and, for PASI 50, 38 percent

15 of patients achieved the endpoint versus 10 percent

16 in the placebo.

17 Now, examining outcomes for Cohort 1 in

18 the second course, we detected evidence of

19 incremental efficacy as shown here in yellow. You

20 see that, for each of the three endpoints I have

21 just described, the response rates increased in the

22 second course. Considering the PASI 75, the

23 response rate when from 14 to 23 percent, a very

24 significant increment, similarly, for physician

25 global and PASI 50.

1 [Slide.]

2 Now, to contrast those landmark analyses

3 two weeks after last dose in each course to the

4 overall response rate where patients responded at

5 other times during the course of therapy.

6 Concentrating first on PASI 75, far left,

7 you can see in the alefacept subgroup 28 percent of

8 patients responded at some point during the course

9 of the first course of therapy. This is a doubling

10 of the primary efficacy-endpoint data, 14 percent.

11 The difference here is statistically highly

12 significantly superior to placebo.

13 The evidence of an incremental rise in

14 these overall response rates is also seen for the

15 physician global and PASI 50 with over half the

16 patients achieving PASI 50 in the first course of

17 therapy.

18 If we look at the overall response rates

19 in the second course, we see evidence of

20 incremental efficacy, 37 percent of PASI 50, 30

21 percent for patients clearing their disease and 64

22 percent--that is, almost two-thirds of patients--achieved

23 PASI 50 during the second course of

24 therapy.

25 [Slide.]

1 Now, an important area of ummet need and

2 an important attribute of potentially new therapies

3 or agents that could put the disease into

4 remission; we were interested to calculate whether

5 alefacept had disease-remittive properties and, to

6 do that, we analyzed the data from Cohort 2 who

7 received the twelve weeks of treatment and 36 weeks

8 of follow up.

9 We calculated the duration of remission

10 for those patients that had achieved the most

11 stringent endpoint, PASI 75, during Course 1. The

12 duration of remission was defined as the time spent

13 in response at PASI 50 or better.

14 The median duration of remission, as

15 defined, was seven months. This appears to be

16 significant and to suggest that alefacept is a

17 disease-remittive type of agent and the first

18 systemic immunotherapy to have this type of

19 property. The data also consolidates the findings

20 from the Ellis and Krueger paper in The New England

21 Journal of Medicine where they also demonstrated

22 efficacy duration suggestive of disease-remittive

23 properties.

24 [Slide.]

25 Here is a graphical representation of this

1 same data. We are looking at the PASI-50-or-better

2 response in those that achieved PASI 75, Cohort 2

3 in the Phase 3 IV study. The Kaplan-Meier curve

4 tracks the duration of time patients are in a

5 response of PASI 50 or better.

6 You can see 50 percent of patients are at

7 this level of response for 211 days or more. So,

8 again PASI 50 or better is maintained for a period

9 of seven months for the median number of patients.

10 [Slide.]

11 The other important area of unmet need for

12 chronic-plaque-psoriasis patients is the tremendous

13 quality-of-life deficit these patients suffer. We

14 were obliged to understand whether alefacept

15 treatment improved the quality of life.

16 [Slide.]

17 To do this, we used the tool termed the

18 DLQI, or the Dermatology Life Quality Index first

19 described by Finlay and Kahn in 1994. It has been

20 used fairly widely in dermatologic studies

21 including psoriasis studies.

22 On the left, you see the data for the

23 changes in DLQI for placebo versus 7.5 milligrams

24 IV for the Phase 3 IV study. On the right, you are

25 seeing the corresponding data for the Phase 3 IM

1 study.

2 Looking on the left at the Phase 3 IV

3 data, there is a reduction in the DLQI score for

4 those in the placebo group, 11 to 9.9. I should

5 remind you that the reduction in score is an

6 improvement in quality of life. In the alefacept,

7 7.5 milligram group, there is a significant

8 reduction from 11 to 7.6.

9 The conclusion that alefacept is

10 associated with statistically significant

11 reductions in DLQI scores was also seen in the

12 Phase 3 IM study as indicated on the right here.

13 [Slide.]

14 These types of data don't fully convey the

15 potential quality-of-life improvements patients can

16 experience. To begin to do that, the next two

17 slides address the issue of to what extent are

18 patients really improving.

19 Firstly, to what extent did patients

20 improve if they achieved PASI 75, if they achieved

21 PASI 50 or they achieved physician global. These

22 data are from the responders in the Phase 3 IV

23 study. It is a pooled analysis irrespective of

24 whether the patient was in the placebo group or in

25 the alefacept groups. Looking at the PASI 75

1 response, you can see the score transition is from

2 11 pretreatment to 2.4 if you achieve PASI 75 with

3 alefacept. That is a significant reduction.

4 Similarly, if you go to the right, you can

5 look at the physician global. The transition is

6 from 10.4 to 2.4, again a very extensive reduction.

7 Those data are not surprising because these are

8 very stringent endpoints but we were surprised to

9 see that, for PASI 50, the score went from 11.6 to

10 4.2, another very significant improvement in the

11 quality of life.

12 This data begins to give insight into the

13 importance of PASI 50 as an important endpoint for

14 these patients to achieve with this burden of

15 disease.

16 [Slide.]

17 Finally, to give the ultimate granularity

18 of what quality-of-life improvement means to

19 patients, here are data from the actual

20 subcomponents of the DLQI score for 15 milligram

21 group in the Phase 3 IM study. There are similar

22 data for the other treatment groups. What I want

23 to discuss is the extent to which patients that

24 reported being at the severe end of the scale for

25 each of these questions changed from baseline to

1 two weeks after last dose.

2 So it is a five-point scale and at

3 baseline patients are meant to fill out a

4 questionnaire saying how much embarrassment did

5 they suffer. The most extreme end of the scale is

6 very much or a lot. The proportion who answered at

7 that level at baseline was 64 percent consistent

8 with the disease burden they have.

9 After twelve weeks of treatment, 27

10 percent of patients in the 15-milligram group

11 experienced the same level of embarrassment. Their

12 impact on daily activities transitioned from 21

13 percent having very great difficulties to 7 percent

14 and as you go on down the table.

15 This is across the treatment groups. If

16 you look at the same data for patients who

17 responded to the various endpoints, you see even

18 further improvements or greater improvements in

19 these important quality-of-life domains.

20 [Slide.]

21 Finally, I would like to close the issue

22 with a discussion of outcomes in some important

23 subpopulations.

24 [Slide.]

25 First, the outcomes as a function of

1 disease severity at baseline. There appear to be a

2 lot of ways to quantify disease severity. We have

3 chosen one standard approach here. Severe disease

4 is body-surface area greater than 30 percent at

5 baseline. Less severe disease is body-surface area

6 involved in less than 30.

7 On the right, you can see the proportions

8 of patients with a BSA greater than 30 who achieve

9 the primary efficacy endpoint, 13.8 in the

10 alefacept group in Phase 3 IV study versus 5.6 in

11 the placebo group. The difference is significant.

12 The same magnitude is seen in the BSA

13 less-than-30 group, 16.2 in the alefacept group

14 versus 4.1. We have concluded that alefacept

15 efficacy is not significantly influenced by

16 baseline disease severity and patients with a broad

17 range of disease severity can be helped by the

18 drug.

19 [Slide.]

20 Now, a similar pooled analysis of all

21 Phase 3 patients so that we have very big numbers

22 here was done for patients based upon their prior

23 response status. About 80 percent of patients in

24 the Phase 3 studies reported having one of the

25 major systemic antipsoriatic agents or UV therapy

1 prior to entering into our studies.

2 Those patients were classified based upon

3 their responses as having no change or worsening on

4 the previous therapies, improving on previous

5 therapy or no prior treatment; i.e., naive to the

6 previous therapies.

7 Then, for each of those groups, we

8 assessed the primary efficacy endpoint. For those

9 that had not changed on the previous treatments or

10 worsened, 20.2 percent responded to alefacept. 3.1

11 responded in the placebo group. This difference

12 was highly statistically significant. The same

13 kind of data is seen for those that also improved

14 on previous treatments and for those that were

15 naive to previous treatments.

16 So this analysis supports the conclusion

17 that alefacept is efficacious in a broad range of

18 patients irrespective of their response to previous

19 agents.

20 [Slide.]

21 To summarize the efficacy part of the

22 presentation, we have concluded that alefacept is

23 effective in reducing psoriasis disease activity.

24 We have done this by three independent randomized

25 placebo-controlled studies. These encompass both

1 the IV and the IM routes. The data, as we have

2 discussed, are consistent and robust across all

3 endpoints and in important subpopulations of

4 patients.

5 In the Phase 3 IV study, we demonstrated a

6 greater evidence of response with the second course

7 of therapy--in other words, incremental efficacy--and we

8 demonstrated extended durations of remission

9 of seven months in patients that achieved PASI 75

10 during the Phase 3 IV study.

11 Finally, and most importantly, perhaps,

12 alefacept therapy has been shown to improve the

13 quality of life of patients in the course of both

14 our Phase 2 and Phase 3 studies.

15 [Slide.]

16 I would now like to move to the

17 pharmacodynamics. Both Dr. Bonvini and Dr. Adelman

18 have elegantly described the mechanism of action to

19 you. I now want to review the range of alefacept-mediated

20 lymphocyte effects that we documented in

21 Phases II and III.

22 To do that, I will focus specifically on

23 the Phase 3 IV study, the two-course study. We

24 have similar data from the Phase 2 study and also

25 the Phase 3 IM study. These were summarized in

1 your briefing documents.

2 [Slide.]

3 I will consider both the mean counts over

4 time to convey the range of qualitative changes

5 that we can expect to see and also convey the

6 individual patient experience because there are

7 data of clinical relevance that we should discuss.

8 Finally, I will close with a discussion of

9 the potential implications of the types of changes

10 we have seen with a specific question as to what

11 are the role of memory T-cells given that they are

12 targeted selectively by the agent. After doing

13 that, I want to consider what data do we have that

14 addresses does Biogen have evidence for integrity

15 of immune function in alefacept-treated patients.

16 [Slide.]

17 Here you see a diagram which is just a

18 variant of one that Dr. Adelman showed you earlier.

19 These are the major lymphocyte subpopulations in

20 our peripheral blood and lymphoid tissues. They

21 are dominated by two species, the CD4 and CD8 T-cell. The

22 CD4 T-cells are of two types. They are

23 either naive or they are memory.

24 CD8 T-cells, again, are of the same two

25 types, CD8 naive or CD8 memory. You will see data

1 demonstrating that alefacept selectively targets

2 CD4 and CD8 memory T-cells. From this diagram, you

3 can see that a reduction in CD4 or CD8 memory T-cells would

4 result in a reduction in the total CD4

5 T-cell count or a reduction of the total CD8 T-cell

6 count.

7 Those reductions, in turn, would summate

8 to result in a reduction of the total lymphocyte

9 count which can be easily assayed by the CBC.

10 [Slide.]

11 With that background, let me begin to

12 demonstrate the range of features. This graph

13 summarizes what is at the core of the program, the

14 selective effect of alefacept against memory T-cells. On

15 the left, you see the effect on CD4

16 memory T-cells, on the right, the effect on naive

17 T-cells. It is immediately apparent that, in the

18 memory compartment, there is no significant effect

19 in the placebo group but, in the Phase 3 IV study,

20 the dosing period was associated with a reduction

21 in counts during the dosing interval.

22 Contrasting that to the findings on the

23 right, you see no significant changes in the naive

24 CD4 T-cells in either the placebo or the alefacept-treated

25 patients. We have identical data for the

1 CD8 memory and naive T-cells.

2 [Slide.]

3 Taking the CD4 memory T-cells a step

4 further, changes in this compartment would result

5 in a change in CD4 memory T-cells as a whole.

6 Those changes are illustrated here. You can see,

7 on the dotted line, no significant changes in the

8 placebo group during dose and a significant

9 reduction in alefacept during the dosing interval

10 with an increasing count following withdrawal of

11 treatment.

12 At all timepoints, just as we saw for

13 total lymphocyte counts, the mean, and I emphasize

14 the mean, CD4 T-cell count, remains above the low

15 limit of normal.

16 [Slide.]

17 Finally, the total lymphocyte count; you

18 can see, again, in placebo, no significant changes.

19 In alefacept, significant reduction during dosing

20 and increasing counts upon withdrawal of therapy.

21 Again, the mean counts remain above the low limit

22 of normal.

23 So that is one course of therapy. Cohort

24 1 in the Phase 3 IV study had two courses of

25 therapy.

1 [Slide.]

2 The mean CD4 T-cell changes for that

3 cohort are illustrated on this graph. On the left,

4 you see the Course 1 data. On the right, you see

5 the Course 2 data for the same patients. There are

6 three features in common that I want to go through

7 here. Number one, the rate of change during the

8 dosing interval is identical between Courses 1 and

9 2.

10 Number 2, the nadir reach for mean counts

11 is identical between Courses 1 and 2. Finally, the

12 rate of increase following withdrawal of therapy is

13 also identical between Courses 1 and 2. Note that

14 while patients are on alefacept therapy when drug

15 is withdrawn, they haven't, as yet, reached

16 baseline. At all timepoints, patients maintain

17 mean counts above the low limit of normal.

18 [Slide.]

19 Contrasting those ranges of features

20 considering the entire treatment groups and

21 starting to look at individual patients, we can see

22 the range of effects. To do the most conservative

23 analysis, what we illustrate here are the patients

24 that experienced total lymphocyte in the first row,

25 CD4 in the second row and CD8 in the third row.

1 Counts below the lower limit of normal at any time

2 point during the course of the Phase 3 IV studies

3 either in Course 1 or in Course 2. These are the

4 same patients dosed in both intervals.

5 This is a conservative approach because we

6 count patients, even if they went below normal just

7 on one occasion and came back. Given that most

8 individual's counts are very volatile, this is

9 probably an overestimate of the data. But it is

10 important we go through these carefully.

11 For total lymphocyte counts, the

12 proportions that went below normal in the first

13 course were 18 percent in the first course and 17

14 percent in the second. The CD4 T-cell count, the

15 proportions below normal, first course 44, second

16 course 44. For CD8 T-cell count, 51 percent in the

17 first course and a suggestion of incremental events

18 with 56 percent in the second course.

19 [Slide.]

20 If patients go below normal, then how did

21 they achieve counts within the normal range. I

22 have illustrated that here by looking at patients

23 who achieve counts to within the normal range after

24 twelve injections of IV therapy. These are data

25 from the Cohort 2 in the Phase 3 IV study whom, you

1 will recall, have twelve weeks of treatment and

2 then we followed for a 36-week period off drug.

3 That 36-week interval is the time course on this X-axis.

4 The Y-axis illustrates the proportions who

5 achieve counts within the normal range.

6 Immediately after the twelve injections, you can

7 see 63 percent of patients have counts within the

8 normal range. As we follow patients out, you can

9 see that, by Day 180, 90 percent of patients have

10 achieved a count within the normal range.

11 Finally, as we look at the last time

12 point, it appears that there are patients who are

13 missing while these are patients, 16 patients, who,

14 almost in all cases, were lost to follow up. Some

15 of these patients at the last point of observation

16 had counts between 300 to 400, but they disappeared

17 at any time during this interval and, for purposes

18 of summary, we just leave them missing here.

19 [Slide.]

20 The range of alefacept effects, I have

21 just described, are based upon careful monitoring.

22 In the Phase 3 studies, dosing was only initiated

23 in those with CD4 T-cell counts in the normal

24 range. Dose admission was carried out with

1 substitution of placebo for those patients that had

2 a CD4 T-cell count under 250 recalling that the low

3 limit of normal is 404 cells per microliter.

4 Finally, moving forwards, despite the fact

5 that we have not found any evidence of

6 immunodeficiency associated with the lower T-cell

7 counts, we propose a conservative approach, CD4 T-cell

8 monitoring every two weeks during therapy.

9 [Slide.]

10 Having gone through the phenomenology of

11 the pharmacodynamic effects, I now want to discuss

12 what are the potential implications for us as

13 clinicians here. That depends on a question what

14 are the actual functions of the memory T-cells that

15 are being manipulated.

16 In the physiological setting, memory T-cell are

17 important in the prevention of infections.

18 They are important in assisting B-cells for

19 antibody responses to recall antigens so when we

20 get reexposure to an antigen we have previously

21 seen, the IgG responses are critically dependent on

22 memory help.

23 Finally, they play a potential role in

24 immune surveillance in conjunction with other cell

25 types such as natural killer cells. That is in the

1 physiological setting. In the pathological

2 setting, Dr. Adelman has already discussed data

3 demonstrating that memory T-cells are important in

4 the induction of a range of autoimmune disorders

5 including psoriasis.

6 Over the next two or three minutes, I want

7 to close by addressing what sets of data do we have

8 addressing each of these points.

9 [Slide.]

10 First, the issue of infections and T-cell

11 counts. In the randomized placebo-controlled

12 studies, we divided patients into those that had

13 counts below 250 versus those that had counts above

14 250 and quantified the patients that had infections

15 after counts under 250. That number was

16 24 percent. Contrasting that to those that had

17 infections when counts were above 250, 46 percent,

18 the data suggest that lower T-cell counts do not

19 predispose to infections. Now, this is a very

20 preliminary look at this dataset. My colleague,

21 Dr. Vigliani, who will discuss the safety profile

22 with you, will go into this topic further.

23 [Slide.]

24 We have carefully studied immune-function

25 tests in patients exposed to alefacept to try and

1 determine what evidence do we have for disturbance

2 of normal immunity. To do this, we have used both

3 cell-mediated--tested responses of cell-mediated

4 immunity and responses to humoral immunity. Cell-mediated

5 responses were most robustly addressed in

6 the Phase 2 part of the program, specifically in

7 the Phase 2 IV study that we discussed earlier, the

8 Ellis and Krueger study. There, delayed-type

9 hypersensitivity skin tests were carried out to a

10 range of skin antigens using a CMI multitest.

11 Minor trends towards loss of response to

12 some of the antigens was seen but, given the high

13 false-positive and false-negative rate as well as

14 the difficulty in conducting these types of

15 studies, there are some important caveats when we

16 review these data, and I would be happy to discuss

17 those with you.

18 Contrasting that to the humoral responses,

19 these were studied in the clinical study of 46

20 chronic-plaque-psoriasis patients of the type we

21 treated during Phase 3. They were given either

22 alefacept or placebo and immunized with two T-cell-dependent

23 antigens. These are antigens that T-cells are critically

24 involved in from mounting

25 antibody responses to as documented in a range of

1 immunodeficiency studies in the literature.

2 The antigens were phi-X-174, a neoantigen

3 that patients have never been exposed to where we

4 tested both response when they were naive to the

5 antigen as well as response after reexposure where

6 we are specifically testing memory function. We

7 also tested tetanus toxoid, an antigen that we are

8 all immunized to and we have preexisting immunity

9 to. Here the tetanus toxoid is a recall antigen

10 and we are testing the memory component.

11 When we did these studies, we found that

12 alefacept treatment did not abrogate anti-phi-X-174

13 or antitetanus antibody responses.

14 [Slide.]

15 To show you those data graphically, here

16 are the phi-X-174 responses over time. The X axis

17 is the dosing interval and follow up the Y axis is

18 the mean antibody titer in log units. The primary

19 exposure is associated with a brisk rise in

20 antibody titer in both the alefacept and control

21 groups which is overlapping. This demonstrates

22 that naive T-cell function is intact in the

23 alefacept-treated patients. They can respond to

24 neoantigens.

25 The reexposure or the secondary

1 immunization is associated with a brisk rise in

2 both groups again which appears to be entirely

3 overlapping. The proportion IgG fraction in these

4 patients receiving either alefacept or placebo was

5 identical demonstrating that alefacept patients

6 undergo changes in memory-T-cell counts but that

7 these do not result in a change in their ability to

8 mount antibody responses.

9 We have similar data where we demonstrated

10 that patients had a twofold rise in antibody titer

11 against tetanus toxoid that was identical between

12 both alefacept and control groups.

13 [Slide.]

14 Finally, I want to turn to the issue of

15 what about the pathological setting, given these

16 manipulations of discrete T-cell subsets, do we

17 have data validating the therapeutic rationale as

18 originally proposed by Dr. Adelman. Here we have

19 documented the response on CD4 memory T-cells and

20 to what extent that correlated with the likelihood

21 of patients achieving PASI 75.

22 Now, on the X-axis, you see this axis

23 graded low to high where patients are divided in

24 quartiles, where the reduction in CD4 memory T-cells is

25 divided into four groups. Those in the

1 first quartile of the lowest group had the least

2 CD4 memory T-cell changes. Those in the highest

3 quartile had the greatest extent of CD4 memory T-cell

4 changes. Those intermediate had intermediate

5 changes.

6 Now, as you go from left to right, you can

7 see the stepwise increase in the likelihood of

8 response to PASI 75; 13, 23, 33 and 41 percent.

9 These are encouraging data but they are somewhat

10 indirect because we are looking for the surrogate

11 whereas the site of action is really the skin

12 lesion.

13 [Slide.]

14 To address that, Jim Krueger has conducted

15 a study over the last eighteen months asking the

16 question what do we understand about changes in T-cells in

17 the skin and outcomes after a patient is

18 treated with alefacept. Here are just some of his

19 data. What you see here is a plot of the T-cell

20 number at various time points for 21 patients pre-clin

21 versus the change in epidermal thickness at

22 those corresponding time points when the T-cell

23 number was assayed.

24 You can see the data are tightly gathered.

25 In fact, the r-value is 0.87. This suggests a very

1 tight correlation between the change in T-cell

2 number in the skin associated with alefacept

3 therapy and the therapeutic outcome.

4 The last two slides provide important data

5 validating the therapeutic rationale as originally

6 proposed.

7 [Slide.]

8 So I would like to close my presentation

9 by summarizing that, for lymphocyte

10 pharmacodynamics, both in Phases 2 and 3, we have

11 demonstrated that alefacept treatment is associated

12 with selective reductions of memory T-cells with a

13 relative sparing of naive T-cells. There is a

14 great deal of more data behind that bullet point

15 and some of those are with Dr. Krueger from his

16 studies where he has also demonstrated selectivity

17 of changes in the skin versus blood with preference

18 towards changes in the skin and also changes in

19 discrete sub-subsets of memory cells, specifically

20 those that are home to skin to mediate the disease

21 versus those that reside in the central memory

22 compartments. We can, perhaps, review some of

23 those data in the Q&A.

24 With respect to the second point, we have

25 demonstrated dose-dependent and gradual and

1 predictable changes during therapy. The findings

2 are consistent and predictable throughout. There

3 has been an increase in lymphocyte counts following

4 cessation of therapy and the reductions in T-cell

5 counts that we have seen have been correlated with

6 efficacy as I demonstrated but have not predisposed

7 to infections.

8 That is a suitable point to turn to the

9 discussion of the safety profile and I will now ask

10 my colleague, Dr. Vigliani, to come up.

11 Before she comes up, there is just one

12 point I would like to address was the

13 pharmacokinetics which I didn't discuss. The

14 pharmacokinetics are very consistent for the IM and

15 IV and there is as minor point of clarification.

16 In one of the briefing documents, there were some

17 placebo patients that were said to have alefacept

18 in their circulation during the PK assays. Those

19 patients have been revisited and we have provided

20 data to the FDA that have resolved that,

21 demonstrating that this was inference in the assay

22 at baseline. Those were false positives.

23 So, with that, Dr. Vigliani, if you could--

24 DR. DRAKE: I would like to take the

1 prerogative of the chair. I have looked at your

2 slides and the time left. So I just want us to the

3 cognizant of your allotted time. We are a little

4 bit--I don't know how you have divvied it up among

5 yourselves, but if we could try to hold--the next

6 two presenters please hold to the time schedule, we

7 would be appreciative.

8 Thank you.

9 DR. VAISHNAW: Okay.

10 Clinical Safety

11 DR. VIGLIANI: Good morning.

12 [Slide.]

13 It is my pleasure to be here today to

14 deliver the clinical-safety presentation.

15 [Slide.]

16 I will begin by defining the size and

17 scope of the clinical-safety database. I will then

18 review the most common and most serious adverse

19 events. I will review all deaths and will then

20 focus on the issue of malignancy and infection

21 since these are important areas of interest with

22 any new immunomodulatory therapy. Finally, since

23 alefacept is a protein immunobiologic, I will

24 discuss the issue of immungenicity.

25 [Slide.]

1 Let's now turn to the clinical-safety

2 database. Within clinical-safety database are the

3 876 patients from the three placebo-controlled

4 studies previously discussed. We have integrated

5 the data from these three studies and done pooled

6 analyses comparing event rates in alefacept-treated

7 patients with event rates in placebo-treated

8 patients.

9 The integrated analysis provides larger

10 numbers of patients thereby increasing sensitivity

11 for detection of trends not observed in individual

12 studies. However, important differences by study

13 occurring in the individual studies will be

14 highlighted when relevant.

15 The total clinical experience that we are

16 discussing today consists of 1157 chronic-plaque-psoriasis

17 patients from all alefacept studies in

18 which patients have received between one and five

19 courses of treatment. The comparisons presented

20 today will include the integrated placebo-controlled patient

21 experience as well as the

22 experience by course.

23 [Slide.]

24 When reviewing the placebo-controlled

25 comparisons, keep in mind that there is significant

1 disparity in terms of the number of patients

2 receiving alefacept and the number of patients

3 receiving placebo. If we compare the patient years

4 of exposure, as shown on the Y-axis, you can see

5 that alefacept exposure is more than two times that

6 of placebo exposure.

7 The person-year exposure is further

8 magnified when considering the total alefacept

9 people database. The higher person-year exposure

10 in alefacept-treated patients increases the

11 likelihood of capturing adverse events in these

12 patients. Additionally, events of low frequency

13 have an even lower likelihood of being observed in

14 the placebo group.

15 [Slide.]

16 Let us now move to a broad safety overview

17 of the placebo-controlled studies examining four

18 categories of events; incidence of any adverse

19 events, serious adverse events, discontinuations

20 due to adverse events and deaths. Here we find

21 that both alefacept and placebo groups are well

22 balanced in each of the categories. There was one

23 death in the alefacept group, a patient who

24 committed suicide related to his long-standing skin

25 disease.

1 [Slide.]

2 The safety overview by course provides a

3 similar picture. If you look across the top of

4 this table, you can see the number of patients

5 exposed during each course. Upon review of the

6 four categories, there is no broad evidence of

7 cumulative toxicity based upon this top-level view

8 of these important categories of events.

9 [Slide.]

10 If we now take a look at the most

11 frequently observed adverse events, that is those

12 seen at greater than or equal to 5 percent

13 incidence in placebo-controlled studies, we see

14 that 79 versus 83 percent experienced adverse

15 events. The range of adverse events reported is

16 typical for the population studied. There are no

17 unusual or atypical events.

18 You can see that none of the adverse

19 events occurred at a rate of 20 percent or greater.

20 This speaks to the overall tolerability of

21 alefacept and also speaks to investigators' ability

22 to maintain the integrity of the blind during these

23 studies.

24 When you compare the left-hand column to

25 the right-hand column, you can see that the groups

1 are generally well-balanced. When we look at

2 differences on the order of 5 percent or greater,

3 we find only one event, chills, occurring in 1

4 percent of the placebo group and in 6 percent of

5 the alefacept group. This is the one adverse event

6 that has consistently been associated with

7 alefacept exposure.

8 Chills were generally seen via the

9 intravenous route of administration, were generally

10 mild occurring early in the course of therapy and

11 were not associated with fever or other symptoms

12 and, importantly, did not result in discontinuation

13 of study drug.

14 One category of adverse events not listed

15 on this slide is injection-site reactions because

16 they occurred at an overall incidence of less than

17 5 percent in the integrated database. They did,

18 however, occur at a higher rate in the

19 intramuscular Phase 3 study. However, they did not

20 represent a significant tolerability issue.

21 [Slide.]

22 I would like to now consider serious

23 adverse events. These events were largely

24 considered serious based upon the regulatory

25 serious based upon the regulatory definition of

1 serious and, in most cases, this was based upon the

2 requirement for hospitalization.

3 [Slide.]

4 This table displays serious adverse events

5 seen in more than one alefacept-treated patient in

6 the placebo-controlled experience. The complete

7 table can be found in your briefing document.

8 Alefacept and placebo were well-balanced with 5

9 percent incidence of serious adverse events in each

10 group.

11 The most frequently observed event was

12 psoriasis which occurred in six patients in the

13 placebo group and in two patients in the alefacept

14 group. Serious adverse events observed both in

15 alefacept and placebo included chest pain and

16 pancreatitis. Some events show a slight imbalance

17 with higher rates in alefacept-treated patients--for

18 example, coronary-artery disorder, cellulitis

19 and myocardial infarction.

20 This apparent imbalance may be explained,

21 at least in part, by the fact that we have much

22 greater alefacept exposure than placebo exposure

23 and the number of events is small. Also note that

24 numerous single occurrences of serious adverse

25 events are not displayed in this partial table

1 accounting for the similar overall rates of serious

2 adverse events between the two groups.

3 [Slide.]

4 The rates of serious adverse events did

5 not increase with increased exposure in repeated

6 courses. So if you look along the top in yellow, 5

7 percent in the first course and going down to 2

8 percent in the fifth course experience serious

9 adverse events.

10 The range of adverse events seen were,

11 again, typical for the population studied and

12 didn't change significantly from those observed in

13 the placebo-controlled studies. Considering some

14 of the individual events noted at a higher rate in

15 the placebo-controlled experience such as coronary-artery

16 disease and cellulitis, none increased in

17 incidence with further courses of therapy.

18 Importantly, when evaluating overall

19 observed rates for events such as myocardial

20 infarction and coronary -artery disease, the rates

21 are consistent with the expected rates in the

22 general population based upon available

23 epidemiological data.

24 [Slide.]

25 I will now review the reported deaths

1 within the program. There have been a total six

2 deaths in the alefacept program to date. The first

3 four were detailed in your briefing document.

4 Three of these occurred in patients on alefacept

5 and one patient died prior to receiving study drug.

6 Two additional deaths have been reported

7 since the briefing document and are listed below

8 the line at the bottom of this table. Moving to

9 the top of this slide, we see the suicide

10 previously mentioned. This involved a 34-year-old

11 man with a lifelong history of psoriasis and,

12 unfortunately, a family history of suicide. His

13 disease was featured prominently in his suicide

14 note.

15 This case clearly illustrates the

16 psychosocial impact that psoriasis has in this

17 patient population. There were two deaths from

18 myocardial infarction. Both were middle-aged men

19 with multiple risk factors. While one occurred in

20 a patient on alefacept, the other occurred prior to

21 receipt of study drug.

22 These cases emphasize some of the

23 comorbidities in the study population. The fourth

24 patient died because of esophageal carcinoma

25 resulting from Barrett's esophagus. The two

1 remaining deaths reported after your briefing

2 document include a case of lung carcinoma in a

3 heavy smoker and a patient with a history of

4 seizures who died during a grand mal seizure in his

5 sleep ten months after receiving study drug.

6 [Slide.]

7 Let's now move to a discussion of

8 infections. In addition to collecting adverse

9 events, investigators were required to perform an

10 assessment of the patient for signs and symptoms

11 and infection at each study visit. They were

12 further required to record whether each adverse

13 event represented a new or ongoing infection.

14 Now, this prospective collection of

15 adverse events associated with infection

16 facilitated the identification and analysis of

17 these events. We have also analyzed the risk of

18 infection in relation to reductions in T-cell

19 counts.

20 [Slide.]

21 Looking first at infections that occurred

22 at an incidence of 5 percent or greater in the

23 placebo-controlled studies, 43 versus 45 percent in

24 the two groups experienced an event associated with

25 infection. There were only four events that

1 occurred at an incidence of greater than or equal

2 to 5 percent. These include pharyngitis,

3 nasopharyngitis or the common cold, flu-like

4 symptoms and nonspecific viral infection.

5 As you compare placebo to alefacept for

6 these four events, note that the groups are well-balanced

7 leading to the conclusion that alefacept

8 did not predispose to these common types of

9 infections.

10 [Slide.]

11 Now let's look at whether any of these

12 infections occurred at a higher rate in patients

13 with low CD4 counts. During the pharmacodynamic

14 part of the presentation, Dr. Vaishnaw showed you

15 the top part of this table in yellow. Note that a

16 lower proportion, or 24 percent of patients who had

17 CD4 counts less than 250 developed an infection

18 compared with 46 percent of those who maintained

19 counts above 250.

20 The rest of this table illustrates the

21 range of infections that were associated with low

22 T-cell counts. As you scan through the events,

23 note that there are no events suggestive of

24 opportunistic infections or immunodeficiency. If

25 you compare the incidences for these infections by

1 the CD4-count groupings, you see no significant

2 imbalance.

3 We have analyzed rates of infections for

4 different CD4 thresholds as well as CD8 thresholds

5 and have found no correlation between the risk of

6 infection or serious infection and reduction in

7 lymphocyte counts. The same holds true if you look

8 at data from the multiple course experience. This

9 leads to the conclusion that alefacept-mediated

10 reductions in lymphocyte counts do not predispose

11 to infection.

12 [Slide.]

13 Now let's turn our attention to serious

14 infections. Serious infections were observed at an

15 equal rate of less than 1 percent in both alefacept

16 and placebo groups. There were no atypical or

17 opportunistic infections. This is the placebo-controlled

18 experience. The data are similar across

19 the multicourse experience as described in your

20 briefing document. There were a total of 19

21 serious infections in the entire alefacept

22 database. You may notice that skin infections were

23 the most frequent category of infection in the

24 placebo-controlled experience. Therefore, we will

25 now look at this issue in greater depth focussing

1 on all serious skin infections in the entire 1300-patient

2 database.

3 [Slide.]

4 This table displays the case details of

5 all serious skin infections across the entire

6 program. These are divided into skin infections

7 and postoperative wound infections. Note that in

8 almost all of the cases, there were significant

9 risk factors which alone could account for the

10 types of infections observed.

11 For example, several patients had diabetes

12 mellitus and/or a disruption of the integrity of

13 the normal skin barrier. The first patient, a

14 diabetic, had a history of recurrent otitis

15 externa. The second had manipulated a sty with

16 resultant pre-septal cellulitis. The third had

17 multiple cardiopulmonary medical problems and was

18 treated for a presumed cellulitis, complicating

19 peripheral edema and erythema surrounding a large

20 psoriatic plaque.

21 Another patient with a history of

22 arthritis had a small finger abscess following

23 treatment of olecranon bursitis five months after

24 study drug. Another developed cellulitis

25 surrounding a Herpes simplex lesion near the eye.

1 Each of these patients had uncomplicated

2 infections and responded to conventional therapy.

3 Additionally, there was one case of cellulitis

4 resulting from a large burn and a case of toxic-shock

5 syndrome occurring two months after

6 completing alefacept. This patient experienced the

7 usual complications of toxic-shock syndrome but

8 made a full recovery.

9 In addition, three postoperative wound

10 infections were reported, one requiring

11 debridement, repeated debridement after a rotator-cuff

12 repair. This patient has since continued in

13 retreatment studies without further incident.

14 The two others included a repair of an

15 open and lacerated fracture of the tibia and a

16 surgical infection following appendiceal rupture.

17 Note also that more than 50 percent

18 underwent surgical procedures without such

19 complications. In all cases, patients were treated

20 with conventional therapies will full recovery.

21 The majority of patients continued with treatment.

22 There was no correlation between serious infection

23 and reduction in CD4 counts.

24 I would like to take a minute to discuss

25 the burn infection in greater depth as I feel that

1 it illustrates that maintenance of normal immune

2 function almost certainly contributed to a

3 favorable outcome in a high-risk patient. The

4 patient was an obese diabetic man who dropped a hot

5 radiator on his abdomen while maintaining his car

6 sustaining a large abdominal burn measuring 18 by

7 24 centimeters.

8 Despite a significant disruption in the

9 normal protective skin barrier in an area where

10 wound healing would be otherwise compromised, this

11 patient had an uncomplicated and brief admission to

12 the hospital responding to a course of conventional

13 antibiotics and topical treatments of his burn.

14 [Slide.]

15 So, with regard to infections, we can make

16 the following conclusions. The incidence and

17 nature of infections observed were similar between

18 alefacept and placebo. Low CD4 counts did not

19 appear to predispose to infections. There was no

20 evidence of increasing risk of infections by

21 course. The serious infections observed were

22 uncomplicated in nature, clinical course and

23 outcome.

24 Most importantly, we observed no

25 opportunistic infections, no tuberculosis and no

1 deaths due to infection. Finally, there was no

2 indication that the types of infections that would

3 be suggestive of a T-cell immunodeficiency were

4 observed in the association with alefacept therapy.

5 [Slide.]

6 So we have asked ourselves the question

7 why is it that we haven't seen an increase in the

8 risk of infection despite the significant T-cell

9 effects of this drug. There are a number of

10 possible reasons for this observation.

11 The first is that alefacept does not alter

12 naive T-cells allowing patients to respond normally

13 to new bacterial, viral and other antigens. The

14 second is that the effect of alefacept against

15 memory T-cells is only partial. The remaining T-cells

16 appear to be sufficient to promote antibody

17 responses as demonstrated in the immune-function

18 study previously discussed.

19 Third, there is significant redundancy

20 within the immune system with memory functions

21 divided between a number of important subsets that

22 include CD45RA-positive cells. We have also noted

23 that patients with infection are able to mount

24 increases in their lymphocyte counts. Given that

25 only 3 percent of the T-cell pool resides in the

1 circulation with the rest residing in lymph-node

2 tissue, maintenance of lymph-node integrity may

3 also explain why T-cell function appears to be

4 preserved.

5 [Slide.]

6 I will now turn to the topic of

7 malignancy.

8 [Slide.]

9 The proportion of patients with a

10 malignancy in placebo-controlled studies were less

11 than 1 percent for placebo and 1 percent for

12 alefacept. As expected in this population, the

13 most common cancer was non-melanoma skin cancer.

14 This categorization includes both squamous-cell

15 carcinoma and basal-cell carcinoma.

16 One patient in the placebo and six

17 patients in the alefacept group, less than 1

18 percent in each case, had skin cancers reported

19 during these studies. Two events of carcinoma,

20 both in the alefacept group, were cases of

21 testicular cancer and renal-cell carcinoma.

22 The patient with renal-cell cancer was

23 diagnosed with an 11-centimeter renal mass within

24 three weeks of initiation of therapy making

25 causality unlikely in that case. Prostate cancer

1 was seen in both groups. Finally, a single case of

2 skin melanoma was reported in the alefacept group.

3 This occurred in a patient with a history of PUVA

4 and UVB exposure for 60 months who had two prior

5 squamous-cell skin cancers. His lesions were

6 excised after his fourth dose of study drug. There

7 was no correlation between the development of any

8 malignancy and low lymphocyte or CD4 counts.

9 [Slide.]

10 In the multicourse experience, we have had

11 various additional malignancies reported as

12 presented in this slide with no clear trend towards

13 an increase in incidence with successful courses of

14 exposure. I will not discuss each of these cases

15 in detail today but would like to comment on a

16 single case non-Hodgkins lymphoma that was just

17 recently reported in one of our retreatment

18 studies.

19 [Slide.]

20 This involved a 68-year-old female with a

21 history of long-standing psoriasis for over 50

22 years who had previously been treated with

23 methotrexate and PUVA in the remote past. After

24 twenty doses of alefacept, she presented with an

25 isolated 2-centimeter node below her jaw. She was

1 diagnosed histologically with follicular B-cell

2 non-Hodgkins lymphoma.

3 Workup was negative for other lymphoid

4 tissue or bone-marrow involvement. The

5 histopathologic and molecular features of this

6 tumor suggest that it represents a sporadic

7 occurrence of lymphoma rather than the type of

8 lymphoma seen in association with

9 immunosuppression.

10 [Slide.]

11 To gain a perspective on the overall rate

12 of malignancy and the rate of specific

13 malignancies, we compared the observed rates in our

14 trials with those cited in published literature.

15 This slide illustrates that the overall rate of

16 malignancy, including skin cancers, of 20 per 1000

17 person years is consistent with the expected rate

18 of 29 per thousand person years in severe psoriasis

19 patients. You will note the confidence intervals

20 here.

21 [Slide.]

22 So,in summary, we have seen no evidence of

23 an increase in the risk of malignancy in alefacept-treated

24 patients. The predominant cancers we have

25 seen, as expected, as skin cancers, mainly

1 squamous-cell and basal-cell carcinoma, and the

2 observed rates in the database are within the

3 expected rates reported within the literature.

4 [Slide.]

5 Now let's turn to the issue of

6 immunogenicity. If we look at the incidence of

7 antibody development, we see that the rate of anti-alefacept

8 antibodies are 2 percent or lower both at

9 baseline and after treatment with no increase in

10 successive courses. Rates were slightly higher in

11 the IM study in the range of 4 percent.

12 The titers of anti-drug antibodies seen in

13 the patients that were positive were generally

14 below 1 to 40 and did not amplify with repeated

15 dosing. There has been no evidence of specific

16 adverse safety outcomes associated with the

17 development of anti-alefacept antibodies.

18 [Slide.]

19 Let's now summarize the safety findings.

20 Alefacept has a favorable safety profile as

21 demonstrated by evaluation of adverse events,

22 serious adverse events, infections and malignancies

23 in more than 1300 patients studied with up to five

24 courses of exposure for up to three years. The

25 incidence of adverse events and serious adverse

1 events was similar comparing alefacept to placebo.

2 There is no convincing evidence of an

3 increase in the incidence of infection or

4 malignancy or any relationship to lymphocyte

5 reductions. Alefacept has low potential for

6 immunogenicity.

7 [Slide.]

8 We are committed to understanding the

9 long-term safety of alefacept and, to this end,

10 approximately 800 patients are currently enrolled

11 in safety-extension studies the data from which

12 were summarized here today and continue to be

13 collected. At present, we have over 400 patients

14 who have received more than four courses of

15 alefacept therapy to date.

16 However, in order to best understand the

17 key long-term safety issues, we recognize that

18 large numbers of patients treated for longer

19 periods of time will need to be studied. We

20 believe that the optimal method to study these

21 issues is via an alefacept safety registry study

22 powered to specifically evaluate increases in the

23 risk of adverse events of interest specifically

24 infections and malignancies.

25 We are currently working with experts in

1 the field in order to optimally design and

2 effectively execute such a study.

3 [Slide.]

4 Today you have heard about the unmet need

5 for new therapies in the treatment of chronic

6 plaque psoriasis. I would like to conclude the

7 clinical presentation by summarizing the important

8 and unique features of alefacept.

9 Alefacept represents a novel approach to

10 the treatment of chronic plaque psoriasis by

11 selectively targeting memory T-cells which are

12 believed to be among the key pathogenic mediators

13 in psoriasis. The effects of alefacept on T-cells

14 correlate with improvement in disease activity but

15 are not associated with adverse safety outcomes.

16 A clinically meaningful benefit is

17 appreciated by the majority of patients. Response

18 is associated with significant duration of disease

19 remission. Most importantly, improvement in

20 disease activity is associated with improvement in

21 the quality of life of patients treated. Alefacept

22 is very well-tolerated.

23 These properties position alefacept as the

24 first systemic disease-remittive agent for

25 psoriasis without significant organ-system

1 toxicity. The risks and benefits of this therapy

2 have been rigorously evaluated and we believe that

3 they support the use of alefacept as a new

4 treatment option for this severely underserved

5 population.

6 I will now turn the podium over to Dr.

7 Mark Lebwohl who will discuss the risks and

8 benefits of alefacept from the treating physician's

9 perspective.

10 Alefacept Risk Benefit Profile

11 DR. LEBWOHL: Thank you very much, Dr.

12 Drake and members of the panel.

13 [Slide.]

14 I will try to catch us up.

15 DR. DRAKE: You know, Mark, thank you very

16 much. The only thing standing between these folks

17 and a bathroom break is you. Mark, I am just

18 teasing you. I just want to tell you that we are

19 glad to see you and we are glad you are here and

20 please feel free to present your information.

21 DR. LEBWOHL: Thank you very much.

22 [Slide.]

23 In addition to my role as Chairman of the

24 Department of Dermatology at Mt. Sinai, I see

25 patients about thirty hours a week and so it is a

1 pleasure to be here to tell you a little bit about

2 psoriasis and about my experience with alefacept.

3 I will spend only a couple of minutes for the

4 nondermatologist members of the panel showing you

5 some pictures of psoriasis and telling you a little

6 bit about the treatments we currently use and then

7 I will go to the risk-benefit profile with

8 alefacept.

9 [Slide.]

10 This is plaque psoriasis.

11 [Slide.]

12 You can imagine the impact that this has

13 on the quality of life of these individuals that

14 work at home, in their interpersonal relationships.

15 [Slide.]

16 Involvement of the hands and feet gets in

17 the way of day-to-day activities.

18 [Slide.]

19 Again, you can imagine what this does to

20 an individual self-image.

21 [Slide.]

22 This is just a sampling of the patients

23 whom we treated in our alefacept trials.

24 [Slide.]

25 The negative impact on quality of life

1 that psoriasis has has been compared in a number of

2 publications to congestive heart failure and

3 diabetes and found to be comparable to the impact

4 that those conditions have on patients with those

5 diseases.

6 [Slide.]

7 Fortunately, we do have some excellent

8 therapies. This is my most commonly used treatment

9 which is phototherapy with ultraviolet B. It does

10 have a number of drawbacks. First, it involves

11 treatments three times a week for at least a few

12 months a year, in many cases, for most of the year.

13 Patients need to have access to therapies so

14 someone who lives two hours from a phototherapy box

15 won't be able to get this treatment.

16 And last, but not least, it doesn't work

17 for everyone.

18 [Slide.]

19 PUVA is another superb treatment for

20 psoriasis. It is dramatically effective and, of

21 the treatments that are currently available, it is

22 the only one that provides a durable duration of

23 remission. It is associated with some of the same

24 problems; frequent treatments, access to therapy,

25 but also has been associated with the development

1 of squamous-cell carcinoma of the skin and, most

2 recently, it has been suggested that malignant

3 melanoma occurs in PUVA-treated patients as well.

4 [Slide.]

5 There are three oral medications for

6 psoriasis. The first and oldest of these is

7 methotrexate. It is associated with hepatic

8 fibrosis which has led us to guidelines which call

9 for routine liver biopsies in patients who are

10 treated with methotrexate.

11 Now, routine liver biopsies, by

12 themselves, have significant morbidity and even

13 mortality and, in this study from the Mayo Clinic,

14 a 21-year experience of over 9,000 liver biopsies,

15 1 in 300 had a significant bleed that required

16 intervention, 1 in 1,000 patients, approximately

17 died.

18 For that reason, rheumatologists, in their

19 guidelines, do not call for routine biopsies of

20 everyone who gets methotrexate but it is also clear

21 that the frequency of hepatic fibrosis is

22 substantially higher in psoriasis patients than in

23 rheumatoid-arthritis patients for a number of

24 reasons.

25 [Slide.]

1 That point is made by this patient, who is

2 a patient of mine and, as you can see, does not

3 have much psoriasis because he is now on

4 cyclosporine for his liver transplant. Incidently,

5 he used to work at Mercedes Benz and is very proud

6 of his Mercedes scar.

7 I have, in my practice, patients who have

8 either had liver transplantation because of

9 methotrexate, died while waiting for liver

10 transplantation because of methotrexate or are

11 currently on transplant lists.

12 [Slide.]

13 Probably a more acutely serious side

14 effect of methotrexate is the effect it has on bone

15 marrow. Dermatologists are pretty good at

16 prescribing this drug and we do warn our patients

17 not to take other medications. But I can't tell

18 you how often they do. Patients go to another

19 physician, are given either an antibiotic or a non-steroidal

20 antiinflammatory drug which raises the

21 methotrexate levels and results in bone-marrow

22 toxicity.

23 In this study from Ottawa, some

24 rheumatologists looked at teaching records at two

25 hospitals, teaching hospitals, and surveyed

1 physicians in the Ottawa area and came up with 15

2 cases of pancytopenia due to methotrexate. Of

3 those 15, two died, one of them directly attributed

4 to methotrexate.

5 [Slide.]

6 The second drug I would like to speak

7 about is our oral retinoids. The main side effect

8 of oral retinoids is teratogenicity. But the side

9 effects that really keeps patients from taking this

10 drug is hair loss. This woman had a full head of

11 hair. Not only did she lose her scalp hair, she

12 lost her eyebrows and eyelashes and looked like a

13 chemotherapy-treated patient. This is a very

14 unpleasant side effect.

15 In addition, there are number of

16 mucocutaneous side effect; thin nail plates, sticky

17 skin, cheilitis fissuring and chapping of the lips.

18 Here you see pyogenic granulomas which are very

19 painful. This patient had difficulty using his

20 fingers or walking because of pain from the

21 pyogenic granulomas. Hyperlipidemia is another

22 side effect of retinoids.

23 [Slide.]

24 Lastly, cyclosporine is approved for the

25 treatment of psoriasis. The main limiting side

1 effect--it has many side effects but the main

2 limiting side effect has been nephrotoxicity.

3 Essentially, if you give enough cyclosporine for a

4 long enough period of time, the vast majority of

5 patients will develop some kidney damage. As a

6 result, our guidelines call for limiting

7 cyclosporine therapy to one year.

8 [Slide.]

9 What does alefacept offer? You can

10 imagine the improvement in quality of life that

11 this patient had from the treatment he got but I

12 would like to point out that, according to the

13 protocol of this study, the bar that was set to

14 define treatment success was 75 percent improvement

15 in PASI score. This patient was a treatment

16 failure.

17 As you can see, the patient only achieved

18 66 percent reduction in PASI score. It was 75

19 percent at two weeks. So, despite this benefit,

20 this is called a treatment failure.

21 [Slide.]

22 Another patient who did not achieve 75

23 percent reduction in PASI score. Imagine the

24 difference from here to here. That is a treatment

25 failure according to the high bar that was set in

1 this study for defining treatment success.

2 [Slide.]

3 Another patient. Imagine calling this a

4 treatment failure and imagine the impact this had

5 on this patient's quality of life. This patient

6 did not achieve 75 percent reduction in PASI score.

7 [Slide.]

8 Again, here; same story. I can show you

9 photo after photo of these. This is another

10 problem with the definition of treatment success.

11 This patient achieved 75 percent reduction in PASI

12 score but not until twelve weeks after the last

13 dose. The primary endpoint was defined at two

14 weeks after the last dose.

15 So I can show you many patients who met

16 the endpoint at twelve weeks but didn't meet it at

17 two weeks after.

18 [Slide.]

19 I am only going to show you two patients

20 who did achieve PASI 75 to make two points. The

21 first point is that this patient had a remarkable

22 improvement but noticed that she improved even

23 further twelve weeks after the primary endpoint.

24 The second point that I would like to make is the

25 duration of remission.

1 [Slide.]

2 Here is the patient at baseline. Here is

3 two weeks after the last dose, dramatic

4 improvement, clear twelve weeks after the last

5 dose.

6 [Slide.]

7 Here is the patient twenty-three weeks and

8 only a little over nine months after, you see the

9 psoriasis coming back, nine months, but still,

10 compared to her baseline, a dramatic benefit.

11 [Slide.]

12 Who should receive alefacept? First of

13 all, I believe that it should be limited to

14 patients who have substantial psoriasis. Patients

15 who will have limited disease that would respond to

16 topical therapy certainly would not be the patient

17 I would put on alefacept.

18 In my practice, I will continue to use UVB

19 before alefacept. I think that this is an old and

20 safe and effective treatment. But, for some

21 patients for whom it is impractical, or for

22 patients who simply don't respond to UVB, I think

23 that alefacept is a valuable addition.

24 As far as PUVA, I believe it should be

25 used in rotation with PUVA. The toxicity, the

1 carcinogenicity of PUVA has clearly been related to

2 the cumulative dose. If you can rotate patients

3 from PUVA to other therapies, you can minimize that

4 cumulative dose and, thus, minimize the risk of

5 skin cancer.

6 As far as methotrexate and cyclosporine,

7 given their known toxicities in my hands, I would

8 prefer to use alefacept before methotrexate and

9 cyclosporine.

10 [Slide.]

11 A couple of points about managing

12 alefacept patients. First, it has been studied

13 both IM and IV and I believe that both should be

14 available, there are some patients who don't like

15 needle sticks. If you use it IV, you can draw your

16 blood through the same injection site that you give

17 the intravenous infusion. But, more important, in

18 patients who are covered head to toe, it is

19 sometimes painful to go through a thick plaque and

20 it may be practical, in some patients, to give it

21 IV.

22 As far as monitoring, you have already

23 heard the suggestion that CD4 counts be obtained

24 every two weeks. Examining our patients is going

25 to be very important because we are not going to

1 give this drug to patients who didn't respond in

2 the past. So, for patients who do respond are the

3 ones who are going to get future courses.

4 I would also say that if you look at the

5 way this trial was designed, it was designed to

6 maximize exposure. In real life, it will probably

7 be given less often. If you look at the

8 statistics, there was a twelve-week rest period.

9 The large majority, in fact, I believe over 90

10 percent of responders, maintained their response at

11 twelve weeks. We are not going to treat patients

12 who are still clear. We are going to wait until

13 their psoriasis starts to come back.

14 So I think that, in real life, it is not

15 going to be given with just a twelve-week break.

16 It is going to depend on the patient.

17 Last, but not least, as with any new drug,

18 we are going to have to observe patients for as yet

19 to be determined side effects that we have not seen

20 in these initial studies.

21 [Slide.]

22 As far as overall benefit-risk ratio is

23 concerned, long-term exposure will weigh heavily in

24 the benefit side of this because, as I mentioned,

25 it will not be given to patients who do not

100

1 respond. It will only be given in the future to

2 patients who have responded in the past.

3 I would point out that, in the study, the

4 majority of patients did respond. If you look at

5 the PASI 50 scores, 64 percent of patients after

6 two courses achieved PASI 50. After one twelve-week course,

7 56 percent achieved PASI 50. So the

8 majority do respond.

9 This will reduce the risk part of this

10 ratio because we can monitor lymphocyte counts. If

11 they fall, we simply withhold the drug. The

12 duration of remission weighs heavily on the benefit

13 side because there are very few treatments we have

14 that will give us this duration of remission.

15 Last but not least, I believe it is

16 important that this drug be approved so that we do

17 have an alternative to the hepatotoxicity of

18 methotrexate and the nephrotoxicity of

19 cyclosporine.

20 With that, thank you.

21 DR. DRAKE: You are amazing, Dr. Lebwohl.

22 Thank you.

23 What I would like to do now--I am going to

24 take the prerogative of the chair and I am going to

25 shift the schedule just a tiny bit because I have

101

1 gotten lots of signals from around the table that a

2 bathroom break is in order, instead of waiting

3 until 11:15.

4 But, right before we do that, those of you

5 on the committee who have worked with me before

6 know that I distinguish between points of

7 clarification and discussion. What I would like to

8 do now is just take a few minutes to see if there

9 are any specific points of clarification that you

10 would like to ask any presenters from the sponsor

11 before we move on to the FDA presentation

12 afterwards.

13 Then the discussion will begin after we

14 finish everything. So, are there points of

15 clarification that you would like to ask any

16 presenter from the sponsor right now? You also

17 will have another chance, but I just thought there

18 might be something burning.

19 Yes, Lloyd? Or, Dr. King, I guess I

20 should say. He helped train me so it is very easy

21 for me to bounce back into the familiar role there.

22 DR. KING: Thank you. My point of

23 clarification is, in reading the background, it

24 seemed to be that the response to the fixed dose

25 did not matter about the weight of the patient;

102

1 that is, you gave it and the response to the T-cells and all

2 that was the same. Seeing the

3 complications were in diabetics, and being

4 diabetic, I wonder if the sponsor had looked at the

5 role of diabetes, weight and response that they

6 saw.

7 DR. VAISHNAW: We have not specifically

8 addressed the issue of diabetes, weight and

9 outcome. If you were interested in understanding

10 the issue of diabetes and the potential issue of

11 infections, we have some data to speak to that.

12 Was that the--

13 DR. KING: One of the clinical

14 observations is that diabetics are more predisposed

15 to serious infections and other things. I just

16 wondered if that was not something you could tease

17 out because it may have something to do with

18 diabetes and infections.

19 DR. VAISHNAW: In the database of over

20 1500 individuals exposed, the number of serious

21 infections that would see were low. In the

22 placebo-controlled studies, it was under 1 percent

23 both in the alefacept and the placebo group.

24 So, whilst that is an important topic,

25 there really weren't sufficient number of

103

1 infections to study within the diabetic subgroup to

2 definitively determine a relationship or not.

3 DR. DRAKE: Other points?

4 Because we are little bit over, although I

5 must say that Dr. Lebwohl did a great job in

6 catching us up, what I would like to do is call for

7 a ten-minute recess. We will reconvene in ten

8 minutes. I hope we can make that goal. We will

9 aim for it; all right? Thank you.

10 [Break.]

11 DR. DRAKE: I would like to invite the FDA

12 to begin their presentations. I would really like

13 the audience--would the audience please be seated

14 or step outside the room.

15 I believe the first presentation by the

16 FDA is Dr. Marzella. You are the gentleman leading

17 off. Please proceed.

18 FDA Presentation

19 DR. MARZELLA: Madame Chairman,

20 distinguished members of the advisory committee,

21 ladies and gentlemen, good morning. In the next

22 hour, we will consider the FDA perspective on the

23 efficacy and safety of alefacept.

24 [Slide.]

25 The FDA presentation has two main

104

1 objectives. The first objective is to confirm the

2 analysis and the interpretations of the key

3 clinical data that you have already heard this

4 morning from the sponsor. The second objective is

5 to point out, and hopefully explain, areas where

6 there are different points of view about the

7 interpretation of the data.

8 These areas are primarily in things such

9 as safety where the clinical data are too few or

10 inconclusive to provide definitive answers. We

11 will be asking the committee to discuss these

12 issues and provide guidance.

13 [Slide.]

14 Biogen is seeking to market alefacept for

15 the treatment of adults with chronic plaque

16 psoriasis. As you have heard, the clinical trials

17 evaluated patients with moderate to severe disease

18 which was defined as involvement of greater than 10

19 percent body-surface area. Patients had

20 previously received or were judged to be candidates

21 for systemic therapy or phototherapy.

22 [Slide.]

23 You have heard this morning already about

24 the significant impact that this disease has on a

25 lot of Americans. It is seen in about 2 percent of

105

1 the U.S. population. There is a genetic component

2 in the disease. Caucasians are affected primarily,

3 other ethnic groups less commonly. There are two

4 peaks of onset, one which is at around twenty years

5 of age and one which is in later years, around

6 sixty.

7 [Slide.]

8 Psoriasis in children tends to have a more

9 severe disease expression. There is also a family

10 history associated. Biogen has requested and

11 received from the agency a deferral of the

12 requirement to conduct pediatric studies. The

13 agency will ask the committee to provide advice on

14 the need and timing of pediatric studies of

15 alefacept in children.

16 [Slide.]

17 As you have heard again this morning,

18 psoriasis is a hyperproliferative disease. It is

19 associated with significant morbidity particularly

20 in the 30 percent or more patients who have

21 moderate to severe disease. We have heard about

22 the impact that this disease has on quality of life

23 and it is well known that it is associated with an

24 increased risk of suicide.

25 [Slide.]

106

1 Let's move on to the analysis of the

2 clinical trials. In my presentation, we will go

3 trial by trial to sort of highlight the key points.

4 The clinical study of alefacept began with single-dose dose-

5 escalation studies of IV and IM dosing in

6 110 healthy subjects and continued with multiple-dose dose-

7 escalation studies in patients with

8 psoriasis.

9 [Slide.]

10 The healthy volunteer study showed rapid

11 maximal reduction in CD4 cells and CD8, primarily.

12 They decreased up to 40 and 70 percent of baseline

13 respectively. Time to recovery was generally hours

14 to days but occasionally lasted several weeks.

15 There was a suggestion of dose relationship of the

16 effect on lymphocytes.

17 The effects of alefacept on lymphocytes

18 will be discussed in more detail when we talk about

19 the Phase 2 and Phase 3 studies. Let me mention

20 another finding of the early studies which was a

21 rise in neutrophil counts which rose to about

22 sometimes as high as four times normal. This rise

23 usually peaked at around 4 hours and it was not

24 associated with changes in body temperature.

25 No other hematologic abnormalities were

107

1 seen. Consistent with this protein configuration,

2 alefacept has a long elimination half-life, about

3 250 hours. The initial study showed that the IM

4 route of administration was approximately 50

5 percent less bioavailable than the IV route.

6 [Slide.]

7 Let's move on to the main Phase 1

8 multiple-dose dose-escalation study which was done

9 in patients with psoriasis. As you can see from

10 the slide, the doses bracketed ranged from 0.005 to

11 0.075 milligrams per kilogram IV, and a regimen of

12 intramuscular dosing was also tested. The

13 treatment schedule consisted of once weekly

14 administration for eight weeks.

15 The main safety observation from this

16 Phase 1 study was the relationship between dose and

17 reduction in lymphocyte counts. The number of

18 subjects with low lymphocyte counts and the

19 duration of low counts increased with dose. At the

20 highest dose level, some subjects experienced

21 prolonged decrease in CD4 and CD8 counts, up to 53

22 days and 117 days, respectively. Again, we will

23 have more to say about these drops when we talk

24 about the Phase 2 and Phase 3 data.

25 This was the first study to give

108

1 information on the time course of drops in

2 lymphocyte counts. Various patterns of change were

3 observed. An important general observation was

4 that lymphocyte counts following an initial drop

5 did not continue to decline as dosing continued.

6 The study also examined delayed type

7 hypersensitivity to intradermal challenge with

8 various antigens. Antigens were applied before the

9 treatment and after the end of the treatment

10 intradermally to non-lesional skin. A number of

11 patients tested positive at baseline and negative

12 post-treatment to specific antigens. In the

13 example shown here, which is the most dramatic, for

14 example for tetanus, there were eight shifts from

15 positive to negative out of a total of nine

16 patients who were positive at baseline and no

17 patients shifted in the opposite direction.

18 [Slide.]

19 Let's discuss next the Phase 2 and Phase 3

20 studies.

21 [Slide.]

22 Let's consider first the general design

23 issues. The studies were randomized, double-blinded and

24 placebo-controlled. An important

25 provision for maintaining the study blind was

109

1 Biogen's use of a laboratory physician who

2 evaluated the laboratory data. The physician

3 ordered placebo substitutions if T-cell counts were

4 below specified thresholds for age and laboratory

5 range.

6 Now, in brief, let me characterize what

7 the three main Phase 2 and Phase 3 studies were.

8 Study 708 was a Phase 2 dose-ranging study that

9 used weigh-based IV dosing. 711 was a fixed-dose

10 IV administration study that evaluated two courses

11 of treatment. Finally, 712 was a dose-comparison

12 study that used fixed-dose intramuscular

13 administration.

14 For all these courses, the drug was

15 administered once weekly for a total of twelve

16 weeks.

17 [Slide.]

18 Let's discuss the primary efficacy

19 outcomes. The primary outcome in Study 708 was a

20 static PGA of mild or better. In Study 711 and

21 712, the main efficacy outcome was a 75 percent

22 reduction in PASI score from baseline.

23 Now, the handling of patients who used

24 disallowed therapies during study was as follows.

25 In Study 708, any topical antipsoriatic drug was

110

1 allowed on specific areas of the body such as

2 groin, scalp, palms and soles. Low potency topical

3 corticosteroids were allowed on any skin lesion

4 other than target lesion.

5 Systemic therapy and phototherapy,

6 however, were not allowed. However, in the primary

7 efficacy analysis, patients who used disallowed

8 treatments were not considered treatment failures.

9 On the other hand, in the Phase 3 studies, namely

10 711 and 712, patients who received systemic therapy

11 or phototherapy were considered treatment failures

12 for the primary efficacy analysis and for most

13 secondary analysis.

14 It is important to note that the

15 prespecified time to assess treatment outcome was

16 two weeks after the end of treatment.

17 [Slide.]

18 There is a suggestion in a number of

19 studies that patients continued to respond to the

20 study treatment beyond the prespecified time point.

21 This is a plausible suggestion given, as you have

22 heard, the long half-life of the drug and also the

23 long duration of its pharmacodynamic effect.

24 However, as we will discuss in detail,

25 there are some caveats to take into consideration

111

1 in interpreting treatment responses in the follow-up period.

2 For this reason, we think that this

3 hypothesis about response needs further

4 corroboration.

5 [Slide.]

6 Let's go, then, to recap, in the next

7 slide, what 708 was, again a dose-ranging study.

8 The dose groups were placebo, 0.025, 0.075 and

9 0.15 milligrams per kilogram IV. Certain

10 concomitant antipsoriatic medications were allowed

11 and dose--and this is an important provision of all

12 of the trials from now on--was withheld if CD4

13 count was less than 300 in this particular study.

14 [Slide.]

15 The next slide indicates, as a sponsor has

16 already shown, that 708 provided evidence of

17 treatment effect. Based on the primary efficacy

18 outcome, there was a 20 percent absolute increase

19 in the proportion of responders.

20 The primary outcome did not provide

21 sufficient information about the relative clinical

22 activity of alefacept doses. However, secondary

23 efficacy analysis such as PASI and pharmacodynamic

24 analysis did allow further delineation of a dose

25 response and, ultimately, this was the dose that

112

1 was selected for the Phase 3 study, intravenous

2 study.

3 [Slide.]

4 Evidence of treatment effect can be seen

5 starting at about 60 days after the beginning of

6 treatment. This is the placebo plot. These plots

7 are for the alefacept groups. This line indicates

8 the time for assessment of endpoint which was two

9 weeks after the end of the treatment period. So,

10 again, there is a suggestion that both in the

11 placebo group and in the alefacept arms, patients

12 continued to respond. The issue is going to be to

13 see--for instance, if one looks at the alefacept

14 group, what is the contribution of placebo in

15 addition to other issues that we will talk about in

16 a moment.

17 [Slide.]

18 This figure is taken--a very elegant

19 figure--from the sponsor's study report. What this

20 shows is the response of lymphocyte counts in Study

21 708 to dosing. The bar here shows the duration of

22 the dosing period. These are the various groups.

23 As you can see, there is a nice dose response in

24 terms of decrease in lymphocyte counts.

25 The pattern of drop is also informative.

113

1 It tends to be greatest within four weeks and,

2 after that, it sort of stabilizes. Following the

3 end of the treatment period, you will notice that,

4 for the groups, there is a tendency for the counts

5 to recover. However, by the last observation in

6 the study, the counts have not returned to

7 baseline.

8 [Slide.]

9 As Biogen indicated, obviously, these are

10 mean data. To look at specific clinically

11 meaningful effects in patients, we have to go to

12 another type of analysis which essentially looks at

13 the proportion of patients that fall under specific

14 thresholds at any time during the treatment course.

15 In this particular case, we are looking at

16 CD4 T-cell counts but the same phenomena can be

17 seen with other T-lymphocyte subsets. Namely, what

18 is occurring is that there is a definite dose

19 response in the proportion of patients who, at any

20 time, have decrease in CD4 cell counts below

21 normal.

22 The other interesting thing is that the

23 magnitude of the drop is also dose dependent. You

24 will notice that, as we go from low dose to high

25 dose, the proportion of patients falling below a

114

1 clinically significant threshold, potentially

2 clinically significant threshold, of 200 also

3 increases.

4 [Slide.]

5 The next slide also shows the correlation

6 of this finding, namely that the laboratory

7 assessing physician ordered substitution of blinded

8 study drug with placebo whenever he observed

9 abnormal CD4 counts. So what this slide shows also

10 is a dose relationship in the proportion of

11 patients who had to receive placebo substitutions

12 because of a drop in CD4 counts. Again, the

13 percentage is dose related and I will remind you,

14 this is the dose that was tested further in the

15 Phase 3 study.

16 A caveat here is that, for this analysis,

17 only patients who completed treatment and received

18 all twelve injections were used.

19 [Slide.]

20 There was some suggestion, in the previous

21 study, that there might have been some shift in TDH

22 testing. Again, to remind you, this was done using

23 a commercial test kit and the antigens, about a

24 dozen of them, were applied intradermally before

25 treatment and then after the end of treatment.

115

1 Again, there is noise in this data but there is a

2 suggestion that the alefacept groups had, perhaps,

3 a higher number of shifts than placebo. This is

4 not consistent for all antigens.

5 [Slide.]

6 If we go to the next group, we can see

7 that, perhaps, there is a trend with Proteus but

8 not with Trichophyton. So we think that this is

9 suggestive data and one should be mindful of it

10 particularly because it has a lot of plausibility

11 due to the mechanism of action of the drug. We

12 will be asking the committee to provide advice on

13 this issue.

14 [Slide.]

15 So, in conclusion, then, 708 provided

16 evidence of treatment effect. The sponsor used

17 pharmacodynamic and secondary efficacy outcomes to

18 identify a dose that appeared to have a suitable

19 risk-benefit profile and, in particular, the high

20 dose was not chosen because, as you saw, about 50

21 percent of patients had to have reductions for

22 lymphocyte counts.

23 The onset of response tended to occur

24 towards the latter part of the dosing period--it

25 began after 60 days in this study and the median

116

1 time in response plus treatment, I didn't actually

2 show the data but it was estimated to be around 70

3 days. I will show that in more detail in further

4 studies and I also indicate how that was analyzed

5 because you have heard different estimates and I

6 want to try to reconcile them and explain how they

7 were arrived at.

8 [Slide.]

9 The study also confirmed that alefacept

10 induces dose-dependent reduction in total

11 lymphocyte counts and lymphocyte subsets primarily

12 CD4 and CD8. Lymphocyte counts did not return to

13 pretreatment baseline by the time of the last

14 hematology assessment which was twelve weeks post-treatment

15 in all subjects.

16 There were also safety observations

17 related to infections and malignancy but we will

18 discuss those as the sponsor has done in the

19 integrated safety analysis.

20 [Slide.]

21 Let's move on to Study 711 which was the

22 Phase 3 intravenous dosing study.

23 [Slide.]

24 This study compared alefacept given IV as

25 a 7.5 milligram fixed dose to placebo. The study

117

1 was also designed to evaluate two treatment courses

2 of alefacept. A minimum interval of twelve weeks

3 was specified between treatment courses to allow

4 for recovery of lymphocyte counts before a second

5 treatment course.

6 Note that in the first treatment course,

7 Cohort 1 and Cohort 2 received alefacept so, for a

8 lot of the analysis, these two cohorts are pooled

9 and are referred to as the combined alefacept arm.

10 The comparator group for that analysis will be

11 Cohort 3 which received placebo in the first

12 treatment course.

13 [Slide.]

14 The primary efficacy outcome was the

15 proportion of patients again who experienced PASI

16 75 percent improvement. As you can see, after

17 placebo adjustment, the proportion of responders is

18 10 percent. These are the confidence intervals of

19 the difference. As you can see, they exclude zero.

20 Using a criterion of PASI 50 percent improvement

21 from baseline, the placebo-adjusted rate is 28

22 percent. These are the confidence intervals around

23 that difference.

24 Using a criterion of PGA almost clear or

25 clear, the absolute difference, after adjustment

118

1 for placebo, is 7 percent. So we are in basic

2 agreement with the finding of the sponsors that

3 there is evidence of a treatment effect--it is 10

4 percent--that the evidence of efficacy is

5 corroborated by secondary efficacy outcomes. And

6 we agree with the sponsor that all of these

7 outcomes, and there are several others, in general,

8 track very well with each other, perhaps not

9 surprisingly because they essentially assess the

10 very same manifestations of disease.

11 [Slide.]

12 Let's spend a little bit of time looking

13 in detail at this slide which tries to examine the

14 changes in median PASI score over time over two

15 treatment courses. Let me, again, explain that

16 there are two treatment courses here and that the

17 sponsor defines a treatment course as an initial

18 dosing interval which, as you see here, is twelve

19 weeks followed by a follow-up period, which is

20 another two weeks, followed by an interval which

21 can be more than twelve weeks to allow for patients

22 who were clear before--and, of course, did not

23 qualify for redosing, as well as to allow for

24 patients who had variable intervals of times during

25 which their CD4 counts were too low for

119

1 retreatment.

2 The reason that this plot is truncated

3 here is that that interval is nonlinear and it is

4 variable.

5 Let's look at the various groups again.

6 This is the placebo group, the brown line. This is

7 the alefacept-placebo group and this is the

8 alefacept-alefacept group. It is important to note

9 that the median scores for all three groups were

10 similar at the beginning of the first treatment

11 course.

12 So, when one compares the combined

13 alefacept group at the end of the treatment period

14 at endpoint to the placebo group, one sees that the

15 median score in the alefacept arm is lower than the

16 placebo group. This is, of course, consistent with

17 the primary efficacy outcome using a responder

18 analysis.

19 It is informative to ask what happens

20 after the second treatment course. First of all,

21 one notices that, in the follow-up period, there is

22 a tendency for the median PASI scores to rise in

23 the treatment group. Following a second treatment

24 course, you can see that there is a further decline

25 in median PASI score.

120

1 There are two ways to look at the

2 magnitude of the second treatment response. One

3 can use as baseline the first treatment course, as

4 the sponsor has done, and that results in a greater

5 estimate of proportion of responders. If one looks

6 as baseline the first treatment course, the

7 magnitude of the second treatment course is lower.

8 In any case, I think it is reasonable to

9 conclude that this plot shows that that two

10 treatments are active, the two courses of treatment

11 are active. A little bit inconsistent with this

12 observation, however, is the fact that in the

13 placebo arm, you can see that an initial placebo

14 response following a course of alefacept, this

15 group ultimately ends up where the other group ends

16 up who received two courses of treatment.

17 Now, of course, for the purpose of this

18 comparison, we are doing a landmark analysis and we

19 are purposefully disregarding the area under the

20 curve which shows that this group did, in fact,

21 benefit. But I am pointing this fact to sort of

22 point to some of the potential complications in

23 comparing these effects.

24 Another comparison that is informative is

25 to look at the alefacept-placebo group. One can

121

1 see that, over the course of about nine months,

2 essentially all of the treatment response is lost

3 and one goes back, then, to the placebo-placebo

4 level. So, again, if you are now thinking back on

5 what the sponsor talked about in terms of median

6 responses of nine months, you sort of have to

7 wonder about that interpretation.

8 The final point that I wanted to make is

9 that, interestingly, there is a maintenance of

10 response following the end of the treatment. The

11 maintenance of response occurs in both the active

12 and the placebo group. So the comparison of these

13 two is not straightforward.

14 I have throw a lot of sort of analysis at

15 you and, of course, I want to sort of make it clear

16 that these are all post hoc analyses, but I think

17 that it is informative to carefully look at these

18 values and try to interpret the various effects of

19 this treatment regimen.

20 [Slide.]

21 Let me go next quickly to the observed

22 mean changes in patient-reported outcomes. I think

23 that the FDA and the sponsor are in complete

24 agreement on what the data show. Actually, as you

25 saw in the meeting package, we--meaning I--misinterpreted

122

1 some of the values and we corrected

2 that in the agenda. But there is no disagreement

3 on the figures.

4 The only thing that I want to point out,

5 as the sponsor did, I guess, is that there is some

6 response in the placebo group and that if one looks

7 at the absolute difference, it is in favor of

8 alefacept. But the question is how meaningful this

9 is.

10 [Slide.]

11 This is for the DLQI which was considered

12 the primary score. Looking at another scale, the

13 DQOLS, there is also, again, a response in placebo.

14 Again, negative scores mean improvement. If you

15 compare the difference between arms, there is a

16 difference in favor of alefacept. But, again, the

17 question is how clinically significant that

18 magnitude is.

19 [Slide.]

20 Moving on to the next slide, we want to

21 look at an estimate of the duration of a 75 percent

22 reduction from baseline in PASI in those patients

23 who achieved a response at the end of the

24 treatment.

25 As you can see from this Kaplan-Meier

123

1 plot, a rough estimate of the median duration of

2 treatment response is, perhaps, about 100 days or

3 so in the alefacept arm and it is about--I think it

4 is about 30 days in the placebo arm. Again, this

5 is looking at--it is, admittedly, a somewhat

6 conservative analysis looking at patients who

7 achieve and maintain a 75 percent response.

8 [Slide.]

9 There was a question earlier about effects

10 of weight on treatment response. This post hoc

11 analysis did suggest that if you look at treatment

12 responses in placebo and alefacept and you divide

13 them weight quartiles that, if you look at the

14 patients in the heavier weight quartiles, that the

15 proportion of responders corrected for placebo is

16 very low. We have a 4 percent, 5 percent and this

17 contrasts with about 18 percent treatment effect

18 adjusted for placebo in patients with lower body

19 weight.

20 Then, if you look overall to try to

21 increase the power, if you make a cut point which

22 is roughly close to the median, and we used for

23 this greater than 85 and less than 85, again, you

24 can see that there is about a four-fold difference

25 in response in favor of patients with lower body

124

1 weight.

2 Now, of course, it is not clear what this

3 association is due to. There are multiple factors

4 but it certainly raises the question of whether

5 patients with greater body weight are being

6 appropriately dosed.

7 [Slide.]

8 The next slide shows the relationship

9 between efficacy and CD4. The sponsor also showed

10 this correlation. I think that the main point that

11 we would like to make here is that is, indeed, a

12 correlation but that the correlation is very weak.

13 This is taking total CD4 counts. The sponsor

14 showed data focusing only on memory cells.

15 There are two ways of looking at these

16 data. You can look at the--this data, let me

17 explain what this shows. This is categorizing

18 patients in terms of magnitude of response. Here

19 we have patients that respond 75 percent or more,

20 50 to 75, less than 50 percent. The question that

21 we ask, then, within each of these groups, what

22 proportions of patients have low CD4 counts.

23 There are two ways of looking at the data.

24 If you look this way, we just calculated the

25 numbers. I don't happen to have them in front of

125

1 me, but another way, perhaps, intuitively to look

2 at the data is to look at the proportion of

3 patients who had 75 percent improvement who were

4 below 300. There is 33 percent of these as opposed

5 to 11 percent who were below 50.

6 So you have to look at these two numbers,

7 11 percent less than 300, 68 percent greater than

8 400. So there seems to be a correlation. If you

9 look at nonresponders, more tend to be over on this

10 side whereas if you look at patients who responded

11 more, more tend to be on the opposite side.

12 However, if you look at--oh; thank you.

13 My office director actually calculated these

14 numbers so I have to give him credit. The

15 percentages are 53 percent for 75 percent

16 improvement, 36 percent and 31 percent. So there

17 is a general correlation.

18 However, if one tries to estimate what

19 proportion of the drop in CD4 accounts for the

20 response, you can see that the correlation is very

21 weak. So, by this estimate, and I have to

22 acknowledge Dr. Chao's analysis for this, only 4

23 percent of the treatment effect can be accounted

24 for by dropping CD4s. So it is a modest

25 correlation at best.

126

1 [Slide.]

2 The next slide also is a busy slide but I

3 think it is very informative. So I will try to

4 spend a few minutes to try to go over that. This

5 is essentially a correlate of the slide that you

6 showed before except that, now, this one asks what

7 happens to median CD4 counts over time in patients

8 who receive two treatment courses.

9 There is a lot of, I think, informative

10 points to be made here. One is that if one looks

11 at the alefacept-alefacept that, following an

12 initial alefacept treatment, there is a tendency

13 for the counts to recover. But, by the time that

14 you get a second treatment, you still haven't

15 recovered to baseline and, in fact, these data

16 suggest that you get a cumulative drop in counts.

17 You go from basically a median of 600 to

18 400. I want to emphasize that these are

19 essentially median counts. These are not in the

20 individual patients.

21 The other point to make is that--actually,

22 this is a very important point to make. This

23 particular study has the best controlled data on

24 long-term safety of a single alefacept treatment

25 because, as you will remember, this group got

127

1 alefacept only during this three-month treatment

2 interval. Then they had a three-month follow up

3 and then they went into a placebo phase where they

4 got three months of placebo followed by another

5 three months of placebo follow up.

6 So, the interesting point here to note is

7 that nine months after the end of the treatment,

8 the median CD4 counts are still low so there is

9 substantial duration of time that it takes for CD4

10 counts to recover.

11 Of course the clinical significance of

12 this is unknown but we would argue that, in view of

13 the suggestion that these effects may be

14 cumulative, that they are long-lasting, that

15 caution and conservatism is called for interpreting

16 the data.

17 [Slide.]

18 Let's look at the same analysis that we

19 talked about earlier. This one now looks at drops

20 below normal in individual patients. These are the

21 proportions of patients that fall below specific

22 thresholds. As you can see, at any time, there is

23 a proportion of patients that drop below threshold.

24 [Slide.]

25 In comparing Course 1 and Course 2 as well

128

1 as comparing multiple treatment courses, the

2 problem is that there is a potential enrichment in

3 patients who are resistant to the potential toxic

4 effects of the product. So these analyses are

5 essentially potential underestimates of what the

6 potential for cumulative toxicity would be for this

7 product.

8 If you carefully noted the numbers in the

9 treatment cycles that the sponsor showed, I think

10 that there was a substantial drop, at least 50

11 percent or more, with each treatment cycle. So the

12 conclusion that there is no cumulative safety risk

13 of adverse events with cumulative cycles has to be

14 tempered by the realization that there is a

15 substantial drop in the number of patients with

16 subsequent cycles.

17 [Slide.]

18 We agree with the sponsor's interpretation

19 that most of the effects are seen in CD4 and CD8

20 counts, particularly in memory cells. However, we

21 would like to point out, and I am not showing the

22 data here, that if you look at individual patients,

23 there are patients who also experience drops in

24 naive cells. NK cells also do show a drop. It is

25 not that dramatic. If you look at mean percent

129

1 changes at nadir, there are drops both in placebo

2 and in the alefacept groups so there is a small

3 differential, but it is reproducible and the counts

4 return to normal.

5 So the point we are making here is that

6 potentially there is a range of CD2-positive cells

7 that can be affected by the drug. Again, the

8 clinical consequences of that may be benign but are

9 certainly unknown at this point.

10 [Slide.]

11 The next slide, again, shows the same

12 issue which is important for clinical use of this

13 product which is the proportion of patients that

14 require placebo substitutions because of CD4

15 counts. Of course, the proportion is--the total

16 numbers of patients is as you see here.

17 This is in the first course, second

18 course, and this is in the drug course of this

19 particular group.

20 [Slide.]

21 So, in conclusions for 711, the trial

22 demonstrated convincingly that alefacept was

23 superior to placebo. The placebo-adjusted response

24 rate was 11 percent absolute. Alefacept was also

25 active for a second treatment course and, depending

130

1 on where one pegs the baseline, the response was

2 either 15 percent or 6 percent.

3 There was a suggestion that body weight

4 was associated with a differential effect on

5 response. There is insufficient data in subjects

6 weighing less than 50 kilos. In the clinical

7 trial, these patients were dosed at about one-third

8 less but there is no enough experience to indicate

9 whether there is sufficient rationale for making

10 that recommendation for these patients. The

11 patient-reported outcomes also showed trends in

12 favor of alefacept.

13 [Slide.]

14 In terms of immunologic parameters, it is

15 clear that alefacept lowers lymphocyte counts.

16 CD4s and CD8s are affected most, NK cells to a

17 lesser degree. Consideration should be given to

18 the potential that lymphocyte reductions may be

19 cumulative and the decrease in CD4 counts are only

20 weakly associated with treatment response.

21 [Slide.]

22 Now, lymphocyte counts may not return to

23 baseline for up to nine months treatment, certainly

24 on average, and certainly they were identical in

25 specific patients, individual patients. The

131

1 pharmacologic effect was potentially greater

2 without appropriate monitoring because one rule

3 that was strictly adhered to in the clinical trial

4 is that weekly monitoring and that the dose was

5 held if counts were less than 250.

6 [Slide.]

7 Let's move on to the intramuscular dosing

8 study. This was this design.

9 [Slide.]

10 This was a study that compared two

11 intramuscular doses of alefacept, 10 and 15

12 milligrams, weekly for twelve weeks to placebo.

13 The stratification was by the two variables of PASI

14 score and prior systemic therapy.

15 [Slide.]

16 These are the efficacy outcomes for the

17 study. We agree completely with the sponsor's

18 interpretation. The placebo-adjusted difference

19 for the 15 milligram dose group is about 17

20 percent. The confidence intervals around that

21 difference between the two groups excludes zero.

22 Interestingly, as the sponsor indicated, the 10

23 milligram dose is also active. In fact, there is a

24 suggestion--I shouldn't say there is a suggestion

25 of a dose-dependent effect, but let me leave it

132

1 that it is intermediate.

2 The p-value that was calculated was about,

3 I think, 0.04. The reason that it did not make it

4 into significance was because of the multiplicity

5 of comparisons, the prespecified p-value was 0.025.

6 So there is a definite suggestion that this is also

7 active. Again, if you use secondary outcomes,

8 let's say 50 percent improvement or a PGA of almost

9 clear to clear, that this is supported by the

10 secondary efficacy outcomes.

11 [Slide.]

12 As in 711, there was a suggestion, at

13 least in the 10 milligram dose group, that

14 retreatment response was associated with weight.

15 As you can see here, these are the proportion of

16 responders in patients in the highest quartiles.

17 This is the next highest above the mean and these

18 are the two lowest. There is certainly a

19 suggestion that patients, again, with higher body

20 weights do not respond as well as patients with

21 lower body weights.

22 This effect was not seen, however, in the

23 15 milligram dose which is what the sponsor is

24 seeking for a label.

25 [Slide.]

133

1 This slide, again, shows the relationship

2 between efficacy and CD4 counts. If anything, in

3 this particular slide, the correlation is a little

4 bit even weaker than in the previous study. I

5 think roughly 2 percent of the response can be

6 accounted for by CD4 counts. I don't think I will

7 go into the details there.

8 [Slide.]

9 The time to treatment response is shown in

10 this slide. Consistent with what was seen in

11 earlier studies, the onset of response is fairly

12 late in the treatment period. This was the time to

13 endpoint. This is the period of dosing. As you

14 can see, time to response, this is the placebo arm.

15 These are the two active arms.

16 There is a difference between the two but,

17 as you can see, separation occurs fairly late,

18 around after Week 9 or so of the treatment period.

19 Again, there is this suggestion that there are

20 additional responders in the post-treatment period.

21 [Slide.]

22 The sponsor--I should have given Biogen

23 credit for the previous plot as well as this plot--this

24 shows the median duration of treatment

25 response. As you can see, this is the placebo

134

1 curve--I cannot read this number from here. It is

2 probably 43--right; it is 43. Actually, let me

3 make sure that I don't misrepresent that. Anyway,

4 it is roughly maybe around 30 or so. It is very

5 hard to see the slides from here. For the active

6 arms, it is around 60. I will stand corrected if I

7 don't read this. Is that reasonable? Okay.

8 [Slide.]

9 Again, we entirely agree with the sponsor,

10 with their analysis of the mean changes in patient

11 reported outcomes. Again, the placebo group tended

12 to respond as well as the active arm but the mean

13 difference between groups favored alefacept.

14 Again, the question that we would like to ask the

15 committee is does this provide additional

16 clinically meaningful information for the label,

17 for a potential label.

18 [Slide.]

19 This analysis looks at the proportion of

20 patients who have abnormal CD4 counts at any time

21 during the treatment period. As you can see, the

22 proportion of patients with abnormal counts and the

23 thresholds that they reach are certainly higher in

24 the active arms confirming previous results.

25 [Slide.]

135

1 The subjects with abnormal cell counts at

2 the last visit is shown here. There is about 8

3 percent of patients at the last visit whenever that

4 happened to occur have abnormal CD4 counts.

5 [Slide.]

6 So the efficacy conclusion for this study

7 is that, compared to placebo, the 15 milligram

8 group is superior and the placebo-adjusted response

9 is 15 percent. The 10 milligram group has

10 intermediate activity. Response for body weight is

11 different in the 10 milligram dose group depending

12 on which cohort you are in and the association

13 between efficacy and reduction in CD4 counts is

14 weak.

15 [Slide.]

16 For patients who responded at any time,

17 the median time to response is certainly near the

18 end of the 84-day dosing period, approximately 90

19 days for both alefacept and placebo groups. The

20 median duration of response in this particular

21 study was 40 days for placebo and 64 days for

22 alefacept. Again, this is a 75 percent criterion.

23 [Slide.]

24 Alefacept, then, induced decreases in CD4

25 and CD8 cell counts. They persist until the end of

136

1 the study in some patients. I didn't show the data

2 but there was a proportion of patients who

3 developed alefacept antibodies, 4 percent, as the

4 sponsor indicated.

5 Let's look at the summary of safety. Here

6 we have, I think it is fair to say, some

7 differences in interpretation with the sponsor.

8 Before going into the integrated safety, I want to

9 comment on the toxicology data. As my colleague,

10 David Green, who made this slide, would like to

11 point out, that similar toxicities were observed at

12 the 1 and 20 milligrams per kilogram dose.

13 So, given the fact that no nontoxic doses

14 were identified, we are not sure what the linearity

15 is between the toxicity of 1 and 20. Potentially,

16 there might be some saturation effect. So we have

17 a word of caution about that.

18 Perhaps another fair caution is that if

19 you look at the animal that, as Dr. Seigel pointed

20 out, developed a lymphoma, the pharmacodynamic

21 correlate of that was some drop in CD4 counts which

22 was that dramatically different, if I remember. I

23 shouldn't, perhaps, be so glib, but it was

24 dramatically different from what one sees in

25 humans.

137

1 [Slide.]

2 So I think the concept to emphasize here

3 is that if one looks at pharmacodynamic effects in

4 addition to dose toxicity, one, perhaps, would

5 adjust downward the safety factor that one is

6 dealing with in the toxicology data and apply that

7 to the human.

8 Let's look at the issue of serious adverse

9 events. The sponsor indicated that the incidence

10 of serious adverse events was the same, 5 percent

11 in both placebo and alefacept arms. But what the

12 sponsor also pointed out was that there was a

13 disproportionate amount of patients in the placebo

14 arm who had serious adverse events which were

15 called psoriasis.

16 We didn't have a chance to go back and

17 analyze those closely, but it is a reasonable

18 assumption to make that these are essentially--the

19 disproportion is because this is essentially a

20 manifestation of efficacy.

21 So another way, then, to consider the

22 safety experience is, perhaps, to exclude patients

23 that have serious adverse events due to psoriasis

24 because one would expect a disproportion in the

25 placebo patients. If one recalculates the data

138

1 this way, then the proportion is 3 percent in

2 placebo and 5 percent in the alefacept arms.

3 The other point to make is that the

4 sponsor indicated that the incidence--these are,

5 admittedly, very low numbers but it is the best

6 controlled experience that we have. It is in

7 Course 1. So the intervals of exposures are

8 comparable. We have a controlled experience.

9 I think that it is not reasonable to sort

10 of discount both of these as being less than 1

11 percent. Again, the numbers are low but another

12 way to look at this is that there is a signal, that

13 the relative proportions are higher in the

14 alefacept arm.

15 This is further supported when one goes

16 and looks clinically at the description of the

17 serious adverse events. The numbers are a little

18 bit different. We excluded one patient from the

19 placebo group because that patient had pancreatitis

20 due to alcohol intoxication and he was classified

21 as an infectious event. So, excluding that event,

22 we have one patient who was a patient with chronic

23 COPD who developed decreased O2 saturation, was

24 admitted, was treated with oral antimicrobials and

25 improved.

139

1 One would contrast that with patients who

2 had peritonsillar abscess, serious cellulitis. For

3 instance, the diabetes mellitus patients, we

4 confirmed the sponsor's observation that this was

5 in a patient with a preexisting risk factor, but

6 this maybe makes it more likely that, perhaps, a

7 signal might be seen in this population.

8 So the fact that the patient had several

9 episodes of external otitis and that, in this

10 particular instance, developed necrotizing facial

11 cellulitis requiring debridement and intravenous

12 antimicrobials is certainly, we would argue, a

13 complicated situation.

14 There are examples, also, from the

15 noncontrolled data. For instance, we would argue

16 that the patient who developed cellulitis is not

17 atypical in patients with psoriasis. But this

18 particular patient developed septic shock and

19 developed renal failure, respiratory failure. With

20 good medical care, he did survive. But, again, we

21 would argue that that is a complicated event.

22 There was another patient, again this one

23 with diabetes mellitus, who had a very complicated

24 course following repair of a rotator cuff. He had

25 multiple abscesses, had to have multiple operating-room

140

1 debridement and wound up, finally, with having

2 to be reoperated and having some residual loss of

3 range of motion.

4 So, again, the numbers are few but we

5 would argue that caution is called for in the

6 interpretation of these numbers.

7 [Slide.]

8 Let's look at malignancies. Again, the

9 sponsor sort of chose to interpret this as less

10 than 1 percent. But, again, there is potentially--the

11 numbers are few but there is a suggestion of a

12 signal, we would argue, potentially. The

13 interesting fact is that the skin cancer seen in

14 the placebo group was a basal-cell carcinoma.

15 There were two basal cells in the

16 alefacept arm and four squamous-cell carcinomas,

17 and the percentages you have to have those. So,

18 again, we would argue that clearly the observation

19 period is short. There are questions about whether

20 we are dealing with development of cancer,

21 promotion of cancer, a clinical diagnosis of

22 cancer, but we think that this cannot be ignored.

23 [Slide.]

24 Let's look at the incidence during

25 treatment of anti-alefacept antibodies. We agree

141

1 with the sponsor's analysis. In the IV group, the

2 incidence was less than 1 percent. The highest

3 titer was 1 to 160. The proportion of patients,

4 not surprisingly, who developed antibodies was 4

5 percent which is notable. The highest titer was 1

6 to 40 and there was no evidence--we agree with the

7 sponsor that these titers resulted in adverse

8 events or loss of efficacy.

9 [Slide.]

10 Let's look, then at the overall

11 conclusions.

12 [Slide.]

13 Alefacept efficacy; the responders

14 compared to placebo, by a criterion of PASI 75--75

15 percent in PASI from baseline, the placebo-adjusted

16 response is 10 to 15 percent higher--it is 10 to 15

17 percent in the alefacept-treated groups. Using

18 PASI 50, the response is 25 percent.

19 Now the median time to response is

20 approximately 90 day both by the IV and IM route.

21 Again, this may seem plausible given the lag time

22 following the pharmacodynamic effects. Then the

23 median duration of response is approximately 105

24 days or 64 days. As we have cautioned, the

25 interpretation of this response is fraught with

142

1 dangers and it is something that needs to be

2 confirmed with additional studies.

3 [Slide.]

4 With regard to reduced lymphocyte numbers,

5 it is clear, as the sponsor has indicated, that

6 phenotypes with higher levels of CD2 counts, with

7 CD2 expression, are affected most. This means T-cells with

8 memory phenotypes. But, again, we would

9 point out that in individual patient-data listings,

10 there were examples of patients who had also naive

11 cells affected. This did not show in the mean

12 counts.

13 NK cells were also affected to a minor

14 extent. There is a suggestion that needs to be

15 considered that the reduction may be cumulative

16 with additional therapy cycles. Again, the comment

17 that we would make, looking at cumulative cycles,

18 is that, given the considerable dropoffs in numbers

19 with subsequent cycles, it is very difficult to

20 interpret that data.

21 Recovery to normal levels or to baseline

22 is slow and/or incomplete in some patients. That

23 data, again, beyond the second cycle is incomplete.

24 [Slide.]

25 We would like to focus the key issue of

143

1 what is the significance of the reduction in CD4

2 and CD8 cells in terms of clinical events. I think

3 that the sheer magnitude of the drop, as I have

4 shown in different studies in as much as 50

5 percent, suggests that the impact is likely, very

6 likely, to go beyond psoriasis or immunity or any

7 specific--recall to any specific antigen and it is

8 likely to impact on immune defenses in general.

9 Again, this is an interpretation of the

10 magnitude of the drops. We would argue, also, that

11 there are some signals. There is some suggestion

12 of decreased DTH responses. This is something that

13 was also observed in the animal data. There is a

14 high plausibility for this effect being seen, so

15 the fact that we would, perhaps, admit to

16 overinterpreting this. But it seems to us to be a

17 signal.

18 There are trends in increased incidence of

19 infections and malignancies that also cannot be

20 disregarded. We acknowledge that the database is

21 small to assess risk but, perhaps, this is the

22 best-controlled way to certainly look at the

23 incidence of infections.

24 [Slide.]

25 W also would like to caution that

144

1 reduction in CD4 counts may be potentially greater

2 without strict monitoring. The sponsor should be

3 complimented for their strict monitor and adherence

4 to safety in the studies. The drug was withheld if

5 CD4 counts were less than 250 and we would think

6 that this would be the regimen that ought to be

7 continued until this additional data that longer

8 periods of observation are just as safe.

9 The other question is the issue of we

10 don't know what happens to noncirculating T-cell

11 pools. We are looking at, basically, a pool that

12 is in the circulation. We know from animal data

13 that lymphoid tissues are all affected. But,

14 obviously, this is not easy to evaluate in humans.

15 [Slide.]

16 So we conclude with this slide indicating

17 that there would appear to be need for long-term

18 monitoring of immune function using clinical and

19 laboratory assessment. More data are needed.

20 Large-scale long-term studies are needed to assess

21 the risk of infections in neoplasms and we are

22 encouraged to see that the sponsor is giving strong

23 consideration to how to design these studies.

24 We have a question for the committee about

25 what is the appropriate timing of the safety and

145

1 efficacy studies in children.

2 DR. DRAKE: Thank you very much. Gosh;

3 you know, this is just a ton of material and I want

4 to compliment both the sponsor and the FDA for

5 concise, thorough presentations. It is a

6 tremendous amount of information to cover, as those

7 of us who spent hours on our briefing books know.

8 I want to do just a second of housekeeping

9 because the notion of this being a holiday weekend

10 and people have already come up to me, would you

11 believe this early in the morning, being concerned

12 about missing flights because the flights are all

13 booked full because of the holiday weekend.

14 So I want to make sure we get our work

15 done on time. That is one reason I have been kind

16 of tight with the time this morning, not to be

17 punitive but to make sure I keep my committee

18 intact until we get to the vote. So I think that

19 is real important.

20 What I would like to do is we have a

21 little bit of time before lunch, so, at this

22 moment, I would like to allow some Q&A to occur. I

23 would like some questions to be directed toward the

24 FDA or the sponsor. Dr. Swerlick, I know you are a

25 nonvoting member but you are here because of your

146

1 expertise, and so I want to absolutely encourage

2 you to participate in the question and in the

3 discussions. You just can't raise your hand when I

4 get to that point. I am not sure why. That just

5 has to do with the process of the FDA.

6 Questions from the Committee

7 DR. DRAKE: Questions for anyone from the

8 committee? Seth? By the way, for those of you

9 don't know, if you will just raise your hand and

10 signal me, I make a little note of who has got

11 their hands raised and I will call on you in the

12 order that I spot you.

13 I have now seen Elizabeth and Seth.

14 DR. STEVENS: The question is for Dr.

15 Marzella and it relates to your observations about

16 possible differential benefit based on patient

17 weight. Did you do analysis on risk for adverse

18 events based on weight and did you see any

19 difference between the heavier and the lighter

20 patients in that regard?

21 DR. MARZELLA: We did look at that and we

22 did not see a correlation. We looked, for

23 instance, at effect of weight on CD4 counts and the

24 correlation was not that strong. I wonder if the

25 sponsor has any comments on that?

147

1 DR. VAISHNAW: I can clarify with just a

2 few brief comments. We, in fact, did divide the

3 Phase 3 patients both from the IV and IM into

4 weight quartiles and examined the adverse-event

5 rate by weight quartile and we saw no trend that

6 was at variance between the various weight

7 quartiles.

8 DR. STEVENS: Thank you.

9 DR. DRAKE: Elizabeth, and then Dr. Katz.

10 DR. ABEL: This was also in regard to the

11 weight, Dr. Vaishnaw. If there is a dose-response

12 curve in terms of effect on lymphocyte counts and

13 the patients of low body weight would be more

14 affected, do we have any data on decreased

15 lymphocyte counts in the patients with low weight

16 compared to high weight and why was this milligram

17 per kilogram dosage schedule abandoned?

18 DR. VAISHNAW: Let me take the issue of

19 lymphocyte changes in the lower weight segments.

20 If I could have Slide 1051, please.

21 [Slide.]

22 This slide illustrates the CD4 memory

23 cells which are the key targets which we defined in

24 our presentation of the drug and the extent of

25 change in the CD4 memory T-cells by the four weight

148

1 quartiles indicated. You can see that there are no

2 significant changes between the four weight

3 quartiles. I already made a comment as to the

4 safety which parallels with this.

5 The second part of your question is

6 important to us in terms of why did we transition

7 from milligram per kilogram to fixed-dose regimens.

8 Essentially, that relates to several factors. One

9 is, in order to insure that in Phase 3 and beyond

10 we could have an accurate calculation of dosing and

11 so that people didn't have to kind of fiddle around

12 with vials and calculate the dose that was

13 required, it is a safety issue and we thought it

14 would be preferable to have a fixed dose. It is

15 more convenient and more accurate. That is the

16 reason why we transition.

17 We had pharmacokinetic data in Phase 2

18 that demonstrated that body mass between lean

19 individuals and heavier individuals was not a

20 significant influence on the major pharmacokinetic

21 parameters. So we took the 0.075 milligram per

22 kilogram dose which was optimum risk-benefit in

23 Phase 2 and converted that to the fixed-dose

24 equivalents in Phase 3.

25 DR. DRAKE: Dr. Katz?

149

1 DR. KATZ: Dr. Vaishnaw, I just want to

2 have two points of clarification. In the cohort

3 that got the two--the drug-drug cohort, you said

4 there was evidence then that they got further

5 improvement. But in the second part of that drug-drug

6 cohort, there was no continual placebo

7 control; is that not correct? In other words, it

8 was placebo-drug. There is no placebo-placebo so

9 there is no control over that continued improvement

10 with placebo. Is that correct?

11 DR. VAISHNAW: I need to, indeed, clarify

12 that point. So, to do that, let me have the Phase

13 3 IV study design slide, just to begin with that to

14 refresh myself.

15 [Slide.]

16 What you see here is, as you say, we were

17 analyzing the response rates in Cohort 1 during

18 Course 1 and comparing them to Course 2 coming to

19 the conclusion that there was evidence of

20 incremental efficacy. You are inquiring as to

21 whether a formal placebo control comparison was

22 conducted.

23 One of the things I want to point out

24 whilst we are on this diagram is Cohort 2, who

25 became placebo in the second course, had that

150

1 prolonged duration of benefit that was the

2 carryover. So this tends to confound the

3 comparisons versus placebo in the second course.

4 If we go to Slide 123, now--

5 [Slide.]

6 On the left, you see the outcomes for

7 Cohorts 1 and 2 in terms of PASI response rates

8 over time. These are data we have already

9 discussed. At the bottom, you see the placebo

10 group. In the second course, Cohorts 1 and 2 which

11 represent the yellow line here were broken out into

12 those that received alefacept again, and that is

13 the yellow line there, and those that received

14 placebo.

15 You can see that there is a substantial

16 carryover effect because the proportions of

17 patients who are responding at PASI 75 are clearly

18 significant. So the placebo-controlled comparisons

19 were carried out and I will go on to discuss them

20 now. But there is significant underestimate

21 because of this carryover effect and the persistent

22 benefit in the population group.

23 Finally, if I could have Display 414 from

24 the briefing document which is where these data

25 were summarized for you.

151

1 [Slide.]

2 This is a complicated table but let's just

3 focus on the second part here. So this is Study

4 711. It is IV study, Course 2 outcomes. Here is

5 placebo response rate and here is the alefacept

6 response rate. Two weeks after last dose, the

7 response rate in the placebo group was 7 percent.

8 Note that it is higher than the response rate in

9 the first course of the placebo group. This is the

10 late carryover effect.

11 When we compare the 7 percent response

12 rate here in the placebo group for Cohort 2, in the

13 second course, versus Cohort 1 who received drug,

14 it is 23 percent in the alefacept group and the

15 difference was highly statistically significant.

16 DR. KATZ: But that group that got drug-placebo

17 weren't really--they were decreasing

18 because they came off the drug in the first--so we

19 are really not getting a true placebo response in

20 the second course. So it is not a true comparison.

21 DR. VAISHNAW: It is not a true comparison

22 and it tends to weight against alefacept so to

23 speak because of this carryover effect of the

24 alefacept effect from the first course into the

25 second placebo course. It was a formal

152

1 prespecified placebo-controlled comparison, but the

2 response rate in the second course, in the placebo

3 group, is still influenced by the alefacept they

4 were exposed to in the first course.

5 DR. DRAKE: Dr. Seigel, I think, has a

6 comment on that question.

7 DR. SEIGEL: There is no question, I think

8 as was pointed out, that the data indicate that

9 patients who get the second course do better, which

10 is to say compared to where they start the second

11 course and, at the end of the second course, they

12 are somewhat better.

13 If the question is whether there is a

14 cumulative effect, they reach a better status on

15 the second course then they did on the first

16 course, aside from the carryover issues, there is

17 another complicating factor here which is that

18 there is some amount of dropout in between the two

19 course. I think, in the controlled study, it may

20 only have been 20 percent of patients, or

21 something.

22 In larger and uncontrolled studies, the

23 dropouts are for any of a variety of reasons. Some

24 or nonresponses. Some are toxicities. Probably

25 some are that they are still in response and not

153

1 interested in getting it again, whatever they are.

2 So you are not necessarily comparing the

3 same patients when you look at the percent

4 response. You are looking at percent responses of

5 a somewhat smaller denominator on the second

6 course. So we have had, for that reason as well,

7 trouble making any definitive determination as to

8 whether there is any evidence of cumulative

9 benefit.

10 DR. KATZ: Thank you. One more question.

11 May I?

12 DR. DRAKE: Yes; please.

13 DR. KATZ: On the diagram that you have on

14 primary efficacy endpoint in Phase 3 based on prior

15 therapy, the point also should be made that only 9

16 percent in the people who improved on previous

17 treatment, which you are taking 100 percent of

18 people who improved on previous treatment because

19 that is in that group, in this study, only

20 9 percent over placebo improved with the drug.

21 So, in human terms, taking 100 percent of

22 people who respond, the drug is only having 9

23 percent--unless I am missing something--9 percent

24 improvement in those people. In people who had no

25 change with previous systemic treatment, there is a

154

1 17 percent response over placebo. Is that correct?

2 DR. VAISHNAW: Right. We illustrated

3 these data terms as one point but the treatment

4 effect is consistent over placebo irrespective of

5 the high response status to the other therapies. I

6 think you have paraphrased the data with respect to

7 this group that reported improving to previous

8 agents.

9 The other data set that I would like to

10 point out here is the differential between placebo

11 response rates for those that reported no change or

12 worsening on the previous therapies and the 20.2.

13 So that is an approximate 17 percent differential

14 to those that responded to alefacept.

15 So this is just a spectrum of analysis to

16 see whether patients are likely to respond to

17 alefacept based on their previous response status.

18 DR. KATZ: Thank you.

19 DR. DRAKE: You may have commented on

20 this, but I have a quick question on that last

21 slide. The previous therapies, were those all

22 systemic or were those both topical and systemic.

23 DR. VAISHNAW: No; those were all the

24 major systemic and--

25 DR. DRAKE: That's what I thought it was.

155

1 Okay; thank you.

2 DR. LEBWOHL: May I also comment that that

3 is PASI 75 and it is at the primary endpoint two

4 weeks after. So anyone who would have achieved

5 PASI 75 six weeks after or twelve weeks after would

6 not be counted there and also anyone who would have

7 achieved PASI 50 wouldn't have been counted there.

8 DR. DRAKE: Dr. Morison.

9 DR. MORISON: I had a couple of questions.

10 The first one, I guess I am getting back to this

11 weight business because one of the things that

12 strikes you with that data no matter which way you

13 look at it is that the actual response rate in

14 comparison to some other systemic therapies is

15 really very low. You come away with the idea, what

16 is the chance that people who are not responding,

17 not reaching 75 or not reaching 50, are actually

18 being underdosed.

19 Is that an issue you have thought about

20 addressing?

21 DR. VAISHNAW: As Dr. Marzella summarized,

22 in the Phase 3 IV study, there was a trend towards

23 lower response rates as you went significantly

24 above 100 kilograms. In the Phase 3 IM study, we

25 didn't see the same type of variation. and those

156

1 are the data summarized here for the PASI 75

2 response rate two weeks after last day. So, again,

3 this is the kind of primary efficacy-endpoint

4 analysis.

5 You can see, in the upper weight segments,

6 you don't see the tail-off in the response. So

7 certainly IM is an option for patients who are in

8 the higher weight category.

9 The other point that you made that I would

10 like to address is the issue of efficacy. If we go

11 to Slide 1059.

12 [Slide.]

13 On the left you see the stringent two

14 weeks after last dose landmark analysis of the

15 right, the overall response rate. What these

16 overall response rates are informing is of,

17 perhaps, very significant clinical efficacy with

18 the majority of patients responding at the level of

19 PASI 50. We provided several lines of evidence

20 demonstrating the kind of quality-of-life benefit

21 patients are attaining with PASI 50.

22 Certainly, in a population like this with

23 this burden of disease with the types of other

24 factors at play in terms of baseline severity,

25 potentially previous response, poor response to

157

1 previous agents. We think these kinds of profiles

2 are very significant and helpful.

3 Mark, do you want to comment on the

4 clinical relevance of the--

5 DR. LEBWOHL: I hope that some of the

6 photos that I showed you express the importance of

7 PASI 50. The PASI score is one that is a high

8 hurdle to climb if you ask for 75 percent

9 improvement because if someone starts out with

10 severe disease over a large body-surface area and

11 has a dramatic reduction in the severity of

12 disease, say from a 3 to 1 in all parameters but

13 has the same area involved, you won't necessarily

14 achieve a PASI 75 in that patient even though the

15 quality-of-life benefit is dramatic.

16 DR. DRAKE: I would like to comment just

17 quickly from a historical perspective. This

18 committee has had, in March of 1988 and October of

19 1988, there were meetings that were just to discuss

20 how to evaluate patients with psoriasis, and what

21 was the utility of the PASI score and what was the

22 physician's global assessment and how did those all

23 weigh together.

24 I can just tell you that we had experts

25 around the table who couldn't come to closure on

158

1 it. We decided the PASI score was certainly far

2 from perfect. We decided the physician's global

3 assessment was probably better. But we also

4 acknowledged that it is almost impossible to put

5 all patients with psoriasis into one bucket because

6 they have different types of psoriasis, different

7 locations, different everywhere.

8 So I would encourage the committee to

9 think more globally and not get hung up on a

10 specific number but more what your gestalt is

11 because everyone around this table understands

12 psoriasis. I don't know how to tell you how to

13 think about it except that I wouldn't get too hung

14 up on a number because the PASI number is not a

15 great number. We just don't have a great

16 substitute for it.

17 If anybody comes up with one, I am certain

18 the FDA and all of us would be very interested in

19 that. So, if that is of any help on this scoring

20 business.

21 DR. VAISHNAW: Could I just also, just

22 interject there, Dr. Drake.

23 DR. DRAKE: Yes.

24 DR. VAISHNAW: Dr. Krueger has also been

25 studying the issue of what is efficacy and he has a

159

1 different approach, and perhaps, Dr. Krueger, do

2 you want to discuss some of your findings with

3 respect to efficacy at a more kind of skin--

4 DR. DRAKE: If it is efficacy related to

5 this, Dr. Krueger, but not a whole new scheme for

6 efficacy. When I was asking for additional

7 comments on PASI, I don't mean to develop a new

8 scheme right now.

9 DR. VAISHNAW: No, no, no. It is not with

10 respect to--

11 DR. DRAKE: Okay; good. My Executive

12 Officer will kill me if I get us off schedule that

13 much.

14 DR. KRUEGER: I have generated some

15 alternate analysis of patients treated with

16 alefacept in a small study that I conducted.

17 DR. DRAKE: Excuse me. Dr. Krueger, would

18 you mind identifying yourself and where you are

19 from.

20 DR. KRUEGER: I am Dr. Jim Krueger. I am

21 from the Rockefeller University. I am a

22 dermatologist.

23 DR. DRAKE: I knew that. I was just

24 checking. Actually, we need it for the record.

25 DR. KRUEGER: I want to say that I have,

160

1 under an investigator IND, conducted an independent

2 study of the effects of alefacept and have used

3 what I view as hard endpoints in a histological

4 assessment of plaques to look at both the response

5 and to look at T-cell effects of skin because T-cell are

6 clearly differentiated home to different

7 compartments and this gives us some direct idea of

8 the disease-relevant T-cell population.

9 DR. DRAKE: Dr. Marzella, have you had a

10 chance to review this information he is about to

11 share with us?

12 DR. KRUEGER: He has not because my data

13 are independent of the Biogen submission under an

14 investigator IND.

15 DR. DRAKE: I would like an opinion. I

16 don't know if we can discuss it at this time. I

17 would like an opinion from the FDA because we

18 really kind of have to have it on schedule.

19 DR. SEIGEL: An opinion as to procedure

20 regarding the data?

21 DR. DRAKE: Yes; procedure.

22 DR. SEIGEL: We don't ban the presentation

23 of new data. We would caution that no data look

24 quite the same after we have analyzed them as they

25 do when they first come to us. I don't mean to

161

1 cast aspersions. So that is something you want to

2 bear in mind but it is certainly up to the chair to

3 see whatever data you choose.

4 DR. DRAKE: Jim, because of time

5 constraints, not that we would disregard your data,

6 but please go ahead. Can you keep it brief.

7 DR. KRUEGER: I will actually limit it to

8 this one slide.

9 DR. DRAKE: Oh; that is really brief.

10 [Slide.]

11 DR. KRUEGER: This is an assessment of

12 what happens to epidermal hyperplasia in patients

13 that either fail to respond or respond to alefacept

14 based upon an endpoint where keratin 16 is either

15 eliminated from lesions or continues to be

16 expressed.

17 So, in the nonresponding patients here, we

18 have very little change happening on the average in

19 this epidermal hyperplasia. This is a group of

20 eight responders out of thirteen in a study that I

21 set up. They are unselected in that these are all

22 sequential enrollees. What we have here is, over

23 the thirteen weeks of treatment, sequential

24 measures of thickness showing a progressive

25 reduction.

162

1 What you can see here at the end is an

2 endpoint that is not so terribly different from the

3 thickness of normal skin. In each of these

4 instances, keratin 16 is turned off. You can see,

5 associated with this in the responding patients,

6 are really dramatic reductions and progressive

7 reductions in the number of T-cells that are

8 infiltrating the epidermis whereas, in the

9 nonresponding patients, the corollary data are that

10 there are not progressive and much lower magnitude

11 changes in T-cell in tissue.

12 So I think, based upon this objective

13 endpoint, it says that this drug is capable of

14 turning off hyperplasia. I have gene expression

15 measures that say all inflammation that is

16 associated and driven by T-cells is also turned off

17 in skin lesions.

18 The problem with the PASI, I believe, is

19 that it is a stochastic measure. I just need to

20 say this, that a 75 percent improvement in the PASI

21 doesn't translate to a 75 percent improvement in

22 disease. In fact, it may be a 95 percent

23 improvement in disease reflected by the PASI of 75.

24 DR. DRAKE: Thank you, Jim.

25 DR. SEIGEL: Just one additional

163

1 perspective. I think we certainly agree with the

2 sponsor that PASI 75 is a relatively high bar. I

3 would also agree that there is not a linear

4 relationship between PASI and amount of clinical

5 benefit. Also, any cut point is an insensitive

6 measure of benefit. Some people probably had a 20

7 percent and would have, on placebo, had a 0 percent

8 or something like that and there is potentially

9 some benefit there.

10 Two things to speak to just to understand

11 and counterbalance against that is that, by any

12 standard, there is a "response rate" in the placebo

13 arm. We wouldn't call that necessarily a placebo

14 response in the sense that it may not have been

15 induced by placebo. It may simply be regression to

16 the mean. People tend to enroll in studies and see

17 their doctors when they are doing poorly because of

18 the cyclic nature--not cyclic nature, but variable

19 nature over time of the disease, when people enroll

20 in studies at times when they are doing poorly,

21 they are often likely to get better on the placebo

22 arm.

23 Some of that was observed here. So when

24 one looks at the placebo rates, as we did, when one

25 looks at the different cutoffs, one needs to also

164

1 look at the placebo rates. So, when you look at

2 the PASI 50, I guess as was pointed out, the

3 response rates go up on both placebo and

4 nonplacebo. They go up differentially. So,

5 instead of seeing a 10 or 15 percent difference,

6 you see I think it was a 23 and 28 percent

7 difference between groups, something larger but

8 still, again, in the 25 percent range.

9 The other thing I would note is another

10 way of looking at this, because of the problem with

11 cut point, are the data on the median score of

12 patients or mean or other aggregate data which Dr.

13 Marzella presented, and just to summarize briefly

14 in one or two sentences, the placebo patients on

15 the first cycle of the study went from a median

16 score of 15 to 12 at their primary endpoint and

17 went from 15 to 8 on treatment.

18 So their status was 8. The treated

19 patients were at 8 whereas the nontreated patients

20 on median was at 12. Again, there is not

21 necessarily a linearity in terms of what the

22 implications of disease are. So one-third lower

23 PASI may or may not mean being one-third or two-thirds as

24 ill. Those comparisons are judgmental

25 and hard to come by.

165

1 DR. VAISNAW: We do have some data that

2 addresses that if there is inflation in the placebo

3 rate and the alefacept rate, how can we

4 differentiate between the extent of benefit in

5 alefacept versus placebo. When we examine the

6 number of times patients hit the endpoint in the

7 placebo group, they hit it many fewer times than

8 those in the placebo groups

9 Although the rates of proportion

10 responding are as we have discussed, the responses

11 you see with the alefacept group tend to be more

12 sustained and so, therefore, of clinical relevance.

13 DR. DRAKE: Dr. Morison, you had a quick

14 follow up?

15 DR. MORISON: Just a quick question for

16 clarification, really. You gave huge doses to the

17 nonhuman primates.

18 DR. VAISHNAW: Yes.

19 DR. MORISON: I presume that the

20 conclusion you would draw is that these animals are

21 much much much less sensitive than humans because,

22 otherwise, they would be dead, wouldn't they?

23 DR. VAISHNAW: Did you say more or less

24 sensitive.

25 DR. MORISON: Much less sensitive. In

166

1 other words, have you got any information on if you

2 give the same sort of dose as you have given in

3 humans, 10 to 15 milligrams, does that produce any

4 change in the primate?

5 DR. VAISHNAW: The object of the nonhuman

6 primate studies, as always, was to really push the

7 test system, as they say in the jargon, that is to

8 give as high a dose as possible for as long as

9 possible to induce changes, to look at the

10 potential range of events that can occur.

11 Under those circumstances, I think, as you

12 are intimating, we would also urge some caution

13 because you start seeing changes which may not be

14 necessarily representative. So, for example, in

15 the 20 milligram per kilogram dose group in the

16 nonhuman primate, we saw over 80 percent reductions

17 in lymphocytes which are far in excess of what we

18 see in man at the therapeutic regimens requested.

19 The other point to note there is that, at

20 those levels of reductions in the nonhuman primate,

21 you lose that selectivity which we spoke about

22 during that main presentation where, with the

23 therapeutic regimen, you see an effect on memory

24 not on naive. In these nonhuman primates with

25 these reductions in excess of 80 percent, you are

167

1 hitting everything.

2 So you start getting into a setting where

3 the toxicologic findings that may or may not occur

4 always are relevant but you can't be sure that they

5 are the result of the kind of mechanism that is

6 operative in man.

7 We have got studies at lower doses. Those

8 more closely resembling the clinical regimen are

9 associated with T-cell reductions of a lower

10 degree. In those settings, we did not see any

11 significant changes of clinical note.

12 DR. DRAKE: I am going to ask Dr. Weiss--

13 DR. WEISS: I just was going to ask if Dr.

14 Green who is at the FDA, who is a toxicologist who

15 reviewed the animal data, if he would just make a

16 comment regarding the data.

17 DR. GREEN (FDA): Thank you. I think that

18 our interpretation of the an toxicology data is at

19 variance with the sponsor and that although a very

20 high dose of 20 milligram per kilogram was used in

21 many of their studies, pharmacodynamically, in

22 terms of immunological endpoints, there was,

23 oftentimes no difference between 20 and 1 milligram

24 per kilogram as Dr. Marzella pointed out.

25 I think that we would find that, for very

168

1 many of the important characteristics such as CD4

2 depressions, we would find a great similarity

3 between the response of the cynomolgus monkeys and

4 other studies including baboons and that which was

5 seen clinically. So I think that some of the

6 factors that have been suggesting that there is a

7 very high difference in terms of safety factors

8 will not translate out.

9 As was pointed out, there is no no-effect

10 dose or no nontoxic dose that we know of. I would

11 say there is probably a grade equivalence between

12 the nonhuman-primate studies and the clinical

13 situation.

14 DR. DRAKE: So the agency is at variance

15 with the sponsor on this issue of dosing. And you

16 are concerned--clarify just a bit more for me.

17 DR. GREEN (FDA): I think we are at

18 variance in terms of the safety factors that were

19 reported. Although there is a difference in the

20 time that the animals were exposed, they gave a

21 factor of, as I recall, about 600. In other

22 documents, they have said there is about a 200

23 safety factor. But that is based on a dose, 20

24 milligram per kilogram, which is functionally

25 equivalent to a much lower dose, and the 1

169

1 milligram per kilogram is approximately, even by

2 the sponsor's statements, about thirteen-fold

3 different than the clinic which puts it exactly in

4 the ball park.

5 DR. DRAKE: Right. Good. Thank you very

6 much.

7 I have a whole list of questions. You

8 guys are getting into this. This is great. Dr.

9 Abel is next, then Dr. Tan and Dr. Swerlick, Dr.

10 Taylor, Dr. Morison, Stevens, Epps and Katz. That

11 is the order in which I seen your hands.

12 DR. ABEL: I have two questions. One is

13 this drug seems to have--it does have a selective

14 action on the memory T-cells. Point of

15 information; do we know what the proportion is of

16 memory T-cells to naive T-cells and could this

17 somehow have to do with responders versus

18 nonresponders, those people who have a lot of

19 memory T-cells and the drug selectively inhibiting

20 them? What are the ranges in normal subjects?

21 DR. VAISHNAW: There is a very wide range

22 of CD4 and CD8 memory T-cell counts in normals. We

23 have generated the largest pharmacodynamic database

24 of this type of lymphocytes in humans to our

25 knowledge.

170

1 [Slide.]

2 For example, here, you can see, at the

3 top, for CD4 memory T-cells, the point I am making

4 about this very wide range.

5 With response to the specific point that

6 did baseline counts for these memory cells predict

7 outcome. The answer to that is no. The most

8 important predictor of outcome, looking at the

9 memory cells that are targeted, was the extent of

10 reduction seen on a percentage basis.

11 That goes back to that slide I showed in

12 the core presentation where, for those that had the

13 greatest reductions in the so-called fourth

14 quartile, 40 percent of them achieved PASI 75.

15 DR. ABEL: Thank you. My second question

16 has to do with therapies that were disallowed. In

17 some of the Phase 1 I believe dose-ranging studies--or that

18 they allowed. There were exceptions to

19 the rule. They allowed them to use treatments,

20 antipsoriatic treatments on the scalp, topicals,

21 palms and soles.

22 Was this the same in the Phase 3 studies

23 that they were allowed to use topical steroids or

24 other antipsoriatic treatments to the palms, soles,

25 groin area, scalp?

171

1 DR. VAISHNAW: I am happy to address that.

2 The Phase 3 setup is described on this slide.

3 [Slide.]

4 These are the therapies that disqualified

5 patients and classified them as treatment failures.

6 So, if you took any of this range of agents from

7 the top down, and they include the phototherapies

8 and the major systemic agents. At the bottom, you

9 see if patients indiscriminantly used moderate-potency

10 topical corticosteroids, D analogues, et

11 cetera, as in beyond the palms and soles and the

12 scalps, then they were treatment failures from that

13 point on.

14 So if we look at the data by taking into

15 account all of these, then the primary efficacy

16 data which we report and the agency reported are

17 what you get. So you are looking at the effect of

18 alefacept as a monotherapy.

19 So the entire efficacy dataset you see

20 today is devoid of the use of these agents

21 respective to all the endpoints.

22 DR. ABEL: But certain sites, they were

23 allowed to use these topical agents in certain

24 sites, and that does have an impact on the PASI. I

25 think if I recall the scalp and the face are 6

172

1 percent of the total body-surface area, and each

2 palm and sole is another 1, 2, 3, 4 percent if you

3 are counting palms and soles. So was that taken

4 into account and subtracted from the PASI response?

5 DR. VAISHNAW: Right. So let's deal with

6 that with Slide 1211.

7 [Slide.]

8 In order to address the issue of how

9 robust are the conclusion from the primary efficacy

10 endpoints, we did what is termed a sensitivity

11 analysis in the jargon. What you see here are the

12 response rates under three sets of conditions;

13 first PASI 75 responders irrespective of the

14 disqualifying medications. We went through that

15 list just now.

16 The response rates you see here are 4

17 versus 15 for placebo versus 7.5 and 7 versus 22

18 for the IM study. In the middle, you see what is

19 termed the prespecified primary efficacy endpoint

20 and those are the data we discussed in the main

21 presentation and the data exactly as we spoke

22 before, and the agency also commented on those.

23 Finally, at the bottom, we looked at the

24 range of medications of the type you are

25 suggesting. I think the agency was also interested

173

1 to explore this further. In their briefing

2 document, they had two tables, Table 29 and Table

3 53, that brought up the issue of these medications

4 that have been used.

5 Then, when we disqualified those patients

6 from the analysis, again we found that the response

7 rates were stable and very comparable to the

8 primary efficacy analysis. So, by these analyses,

9 we have concluded that the data are devoid of the

10 use of the effect of the list of disqualifying

11 medications that we had and also the medications

12 pointed out by the--

13 DR. ABEL: I wasn't talking about patients

14 who were disqualified because they were

15 indiscriminantly using. I was talking about

16 patients who were using in the allowed sites and

17 how that affected the PASI.

18 DR. VAISHNAW: The last analysis just

19 takes them out of the analysis. I can't

20 specifically comment for those patients that were

21 using it on the scale, to what extent it had any

22 effect on their PASI.

23 DR. DRAKE: I think that is the answer.

24 By the way, for the folks from the FDA, when the

25 questions are asked the sponsor is answering, but

174

1 if you guys have an answer or a counter answer,

2 please speak up.

3 DR. VAISHNAW: I think Dr. Lebwohl is

4 indicating to me that he just wanted to make a

5 point.

6 DR. DRAKE: But, before that, Dr. Bonvini

7 had his hand up.

8 DR. BONVINI: I had a comment on your

9 previous question pertaining to the selectivity of

10 action. Again, we have no contention on the

11 evidence that memory cells are substantially more

12 affected than the T-cells in this context. That

13 may be due because these are selectively targeted

14 or perhaps because memory cells tend to die much

15 more rapidly, more quickly, be more susceptible to

16 an action by alefacept or some other agent who

17 might target them.

18 There is evidence that memory cells may be

19 prone to apoptosis. The fact is that we don't know

20 what the exact mechanism of action is. This may be

21 semantic to some extent, but it may not necessarily

22 be in the terms of the selectivity of targeting in

23 one case versus targeting of the whole population.

24 As a matter of fact with higher doses in the animal

25 studies, more than just memory cells were affected.

175

1 DR. DRAKE: Dr. Lebwohl.

2 DR. LEBWOHL: Just to address Dr. Abel's

3 comment. It was first double-blind placebo-controlled so

4 that the impact on PASI score would

5 be seen both in the active treatment group and in

6 the placebo group. At the investigator's meeting,

7 many investigators were unhappy with the prospect

8 that patients would be treated with twelve weeks of

9 placebo and twelve weeks off therapy, almost six

10 months, with no therapy at all on visible areas,

11 scalp and hands.

12 So they bore down on the sponsor to add

13 that possibility with weak topical steroids in

14 those areas.

15 DR. DRAKE: I have just a quick request.

16 I have to ask everybody in the room who has a cell

17 phone to please turn it off. I am embarrassed to

18 ask that because the very first cell phone that

19 rang was mine. So I have now turned mine off. If

20 I have to turn mine off, so do all you guys. I

21 appreciate your cooperation on that issue.

22 Dr. Tan.

23 DR. TAN: The incidence of adverse events

24 in the alefacept group is consistently higher. The

25 incidence in the alefacept group is consistently

176

1 higher than those in the placebo group. I wonder

2 if this trend is statistically significant where it

3 is stabilized. Is there any statistical analysis

4 about this adverse event--

5 DR. VAISHNAW: Right. So the issue did we

6 power the studies or do we have a statistical

7 insight into the rates of adverse events that we

8 have seen. So, in keeping with the usual approach,

9 the studies were powered for efficacy rather than

10 safety.

11 DR. TAN: No; I understand that.

12 DR. VAISHNAW: To take the question of

13 have we had a statistical approach to some of the

14 rarer events, for I think my colleague, Dr.

15 Vigliani, addressed that with just one of our

16 sites. We have others of that type. But, for

17 example, if you take the total malignancy rate, the

18 rate expected is within the rate expected for this

19 type of moderate to severe psoriasis population

20 when you look at the rates reported in the

21 literature. The means and confidence intervals are

22 almost overlapping.

23 We have similar data for other types of

24 rare adverse events. The other point, of course,

25 is that in the alefacept group, there were far

177

1 greater numbers of patients. So the period

2 observation of patient years observed is greater

3 for alefacept in the placebo-controlled studies and

4 so you are more likely to pick up rare events

5 DR. TAN: But in terms of it, you look at

6 infection, you look at neoplasm, but they are all

7 like relative instances, like at least doubled,

8 more of these.

9 DR. DRAKE: Dr. Seigel?

10 DR. SEIGEL: Certainly, I think in the

11 areas that we highlighted concern about, which were

12 serious infections, and this is corrected; these

13 are in the controlled trials and patients in both

14 groups were followed approximately six months in

15 the course, 0.9 versus 0.2 percent. For a

16 malignancy, 1.1 versus 0.5 for the subset of skin

17 malignancies, I think it also around 0.9 versus

18 0.2. None of those comparisons are statistically

19 significant. We are talking about a handful of

20 cases.

21 I think, as Dr. Marzella correctly said,

22 they have raised concerns. They hardly stand as

23 definitive evidence of treatment-associated adverse

24 effect. But, if there are adverse effects at the

25 levels suggested, at a half percent per half year

178

1 increase, or about a 1 percent year increase, if

2 those do exist, then these trials--the controlled

3 part of the data here are well under-powered to

4 look at that.

5 DR. VAISHNAW: The other way we have

6 addressed the issue given the low incidences of

7 numbers in both the placebo and the alefacept group

8 is to ask ourselves the questions are the rates

9 increased over time with multiple course of

10 exposure because one might expect to see a rise in

11 the rates of serious infections if that is one of

12 the points of debate.

13 We have consistently failed to see a lack

14 of rise in the infection rate with multiple course

15 of exposure. Under the issue of low numbers, these

16 are other ways to look at it. The last point I

17 would make on the topic is that naturally we, like

18 the agency, are very diligently addressing the

19 issue of what is the risk of infection in this

20 population and does the agent predispose to that.

21 The central question there to ask has been

22 that, given that this is an agent that targets T-cells, is

23 there a pattern of events in terms of

24 infections or malignancies that are representative

25 of T-cell immunodeficiency. Most of us are very

179

1 familiar with the pattern of infections you would

2 expect to see in T-cell immunodeficiency and we

3 have failed to consistently see that and both we

4 and the agency included in our briefing documents

5 that we have not seen a relationship between

6 alefacept treatment and the occurrence of

7 opportunistic infections or atypical infections.

8 DR. TAN: Of the 2 million patients with

9 psoriasis, how many of them would be as severe a

10 psoriasis as you defined?

11 DR. VAISHNAW: Of the 2 million patients,

12 how many would be classified as moderate to severe

13 DR. TAN: Yes.

14 DR. VAISHNAW: I am not an expert on this.

15 Dr. Lebwohl will correct me, but I believe of the 2

16 million or so in the U.S., probably 20 percent are

17 moderate to severe.

18 DR. LEBWOHL: The number from the survey

19 of the Psoriasis Foundation was 7 million psoriasis

20 patients and someone had a number of 30 percent.

21 Certainly, there are a minimum of half a million

22 and probably about a million severe psoriasis

23 patients.

24 DR. DRAKE: Dr. Swerlick, finally.

25 DR. SWERLICK: Thank you. A comment about

180

1 some confusion in definitions. It is easy to get

2 confused as returning to baseline as opposed to

3 returning to normal.

4 DR. VAISHNAW: Yes.

5 DR. SWERLICK: In terms of looking at T-cell

6 counts, I think we should try to be really

7 explicit about sort of defining that. The reason I

8 raise that has to do with the next series of

9 questions I have. Do we really know if there is

10 any difference between normal CD4 counts and normal

11 memory-cell counts in psoriatics versus normal

12 individuals or individuals with other inflammatory

13 skin diseases?

14 DR. VAISHNAW: Shall I take that question?

15 DR. SWERLICK: Yes. Anybody.

16 DR. VAISHNAW: We are privileged to have

17 the largest database on this topic so I guess I

18 have to answer this. What we have found is that if

19 we look at the entire cohort of alefacept-treated

20 chronic plaque psoriasis patients at our disposal

21 for analysis, there is a minor elevation in the CD4

22 and CD8 memory counts versus the healthy volunteer

23 database that we have.

24 There are lots of caveats to that kind of

25 comparison, clearly. It is not an order of

181

1 magnitude. It is maybe a 5 to 10 percent

2 elevation. It reaches statistical significance but

3 we do detect that. The issue hasn't been addressed

4 in the literature as yet.

5 Dr. Krueger?

6 DR. KRUEGER: I would like to comment also

7 because I think you raise a very important point,

8 that return to normal and return to baseline may be

9 different kinds of considerations. From study of

10 psoriasis patients outside of this study, there

11 have been two kinds of expansions of T-cells that

12 have been found in the peripheral blood of

13 psoriasis patients.

14 One is that there is a higher proportion

15 of CD25-positive T-cells. Those are proliferative

16 T-cells. One might conclude, therefore, that if

17 those were reduced, there was some reduction, they

18 are about 10 percent elevated over normal, that you

19 could say that a 10 percent reduction might, in

20 fact, bring these people back down to normal.

21 The second thing is there is an expansion

22 of Type 1 T-cells, so psoriasis is a disease of

23 immune deviation. Again, there is about a twofold

24 elevation of Type 1 T-cells in psoriasis patients

25 compared to normals.

182

1 So, in my view, if you take both of these

2 sets out, you might, in fact, derive sort of a

3 normal set for these patients that might have a

4 reduced number from their baseline.

5 DR. VAISHNAW: Thank you, Dr. Krueger.

6 DR. BONVINI: Can I ask a question to Dr.

7 Krueger?

8 DR. DRAKE: We are not done with you

9 DR. BONVINI: Sorry, Dr. Krueger. Your

10 CD25-positive T-cells were affected to CD25

11 negative by alefacept? In other words, binding

12 appears to be identical as far as I understood. I

13 was wondering if actually the susceptibility to the

14 two subsets is identical.

15 DR. KRUEGER: CD25-positive T-cells tend

16 to be CD2 high. Therefore, they are affected

17 selectively by this drug, if that answers the

18 question

19 DR. BONVINI: Can you repeat that?

20 DR. KRUEGER: I said CD25-positive T-cells, the

21 activated T-cell group which tends to be

22 memory T-cells, have high levels of expression of

23 CD2 and therefore they are selectively reduced by

24 alefacept

25 DR. BONVINI: Comparing CD25, the high

183

1 level of expression of CD25 and the low level of

2 expression in memory cells.

3 DR. KRUEGER: Yes. In fact, in peripheral

4 blood, there are about 20 percent of circulating T-cells

5 that are CD25-positive. The other 80 percent

6 of CD25-negative.

7 DR. VAISHNAW: Just to finish that point,

8 I think neither Dr. Marzella nor myself included

9 these data. This was addressed in one of the

10 earlier Biogen studies, the issue of CD25-positive

11 cells. Indeed, the findings that Dr. Krueger is

12 reporting from his study were corroborated by the

13 findings in ours that, as expected, CD25 are

14 preferentially targeted.

15 DR. DRAKE: Dr. Swerlick?

16 DR. SWERLICK: Is there any data looking

17 at conventional therapies such as methotrexate or

18 even systemic corticosteroids and their effect on

19 lymphocyte CD4 counts? Are they equivalent to what

20 is seen? Are they larger? Are they smaller? Is

21 it known?

22 DR. VAISHNAW: I am not familiar with the

23 investigations of methotrexate and its effects on

24 CD4 T-cells in psoriasis. Again, I appeal to

25 someone from one of our consultants because they

184

1 are dermatologists. They might be familiar. As I

2 think Dr. Krueger mentioned, there was an

3 investigation of methotrexate and its effect on

4 memory T-cells, I believe.

5 DR. KRUEGER: I have to say, for the most

6 part, these are not points that were taken up in

7 the prior studies of older drugs simply because, at

8 that time, we weren't thinking about T-cells in

9 this disease. Subsequent studies haven't really

10 looked at that.

11 DR. SWERLICK: One last question, and that

12 is getting back to the studies with DTH, again, we

13 are studying patients undergoing this therapy. Do

14 we know what we are comparing it to? For example,

15 if you put a series of DTH reactions on normals,

16 what is the reproducibility? How many of those

17 individuals change from negative to positive or

18 positive to negative?

19 DR. VAISHNAW: To address that, I would

20 like to bring Slide 1110 up, please.

21 [Slide.]

22 These are, I think, the data that Dr.

23 Marzella was drawing your attention to during part

24 of his presentation. So this is the DTH response

25 converting from positive to negative in the Phase 2

185

1 IV study.

2 At the bottom, I point out an important

3 caveat and this begins to address the issue you

4 have raised. Less than 30 percent of patients were

5 reactive at baseline. So this is one of the

6 caveats when you are interpreting the data. The

7 next point is the issue of how many people just

8 convert from positive to negative without the

9 influence of alefacept. Do we have any insight?

10 The response to that is yes. If we look

11 at the placebo group here, you can see significant

12 conversion rates to negativity. These are

13 patients, of course, that didn't receive placebo.

14 So I would argue that yes, you are raising some

15 important caveats. The performance of these tests

16 is difficult. Their clinical implications are not

17 well understood.

18 Whilst, as Dr. Marzella said, and we

19 acknowledge there are some trends for one or two of

20 these, the fact that so many patients are not

21 reactive at baseline, the fact that many normals

22 convert to negative and the fact that for many of

23 these antigens that are on this table, the antibody

24 response is much more dominant than the T-cell

25 response for protection.

186

1 We would have our own set of caveats for

2 interpretation of these data but these are

3 precisely the data that Dr. Marzella showed.

4 Slide 1111, if we could go to that.

5 [Slide.]

6 This is the same type of analysis. This

7 is a less-conservative analysis that we also did

8 just to see how things spun out because, for the

9 last analysis, if you converted from positive to

10 positive and then negative, because there were two

11 time points at which they were reevaluated, if you

12 were positive on one and negative on the other one,

13 you were counted as a negative.

14 Here, this is an analysis of the data

15 where, if you were positive at baseline and you

16 were positive in one of the two post-treatment

17 visits, you were counted as positive and you start

18 seeing loss of the trend.

19 So we acknowledge what Dr. Marzella is

20 saying, but we have had interpretation difficulties

21 with this assay.

22 DR. DRAKE: Dr. Morison had a comment on

23 this.

24 DR. MORISON: I would comment, anybody who

25 has used this particular system, there is so much

187

1 noise in the system, I don't think the results mean

2 anything. I am amazed you actually picked that as

3 a means of looking. Looking at DNCB sensitization

4 would have been much more attractive an approach

5 than this.

6 DR. VAISHNAW: To that point, that is why

7 I drew your attention, also, in fair balance, to

8 the phi-X-174 study which is pioneered by Hans Ochs

9 who is a leader in the investigation of

10 immunodeficiency. Both Ochs' literature and many

11 others have demonstrated that failure of response

12 to phi-X-174 is clearly correlated with

13 immunodeficiency.

14 DR. SEIGEL: I had a question about that,

15 though. It looked like, from your slide, that the

16 primary immunization to phi-X-174 was given at the

17 time of the onset of treatment, not at the time

18 when the patient had become lymphopenic but prior

19 to where the lymphopenic effects of the drug had

20 kicked in.

21 DR. VAISHNAW: I would be happy to address

22 that, Dr. Seigel. Can we have the slide from the

23 main presentation because this does require a

24 clarification for Dr. Seigel.

25 [Slide.]

188

1 This slide was corrected within the last

2 48 hours just to try and make it simpler. This is

3 in error so you are quite right to point that out.

4 Let's go and clarify for the audience the actual

5 data.

6 If would could have the CD4 and CD8

7 changes and their relative timing to the point of

8 immunization, please.

9 [Slide.]

10 Here we have the conversion. So you can

11 see, in orange, is the reduction in CD4 memory T-cell count.

12 In blue, you see the naive T-cells

13 which are relatively constant. You can see here is

14 the primary challenge. It is back in the middle of

15 the period of exposure to the drug. And here is

16 the rechallange.

17 So this study which was designed in

18 conjunction with the agency, was a kind of maximal

19 test of the hypothesis that if you push the T-cell

20 experience, will these patients mount antibody

21 responses. Our conclusions were yes.

22 DR. DRAKE: Dr. Swerlick, are you done?

23 DR. SWERLICK: Yes.

24 DR. DRAKE: Dr. Taylor

25 DR. TAYLOR: I had two points I wanted to

189

1 make. One of them has already been taken care of

2 and that has to do with the PASI score. I think it

3 has been adequately pointed out that PASI 75 is a

4 very, very high bar to reach and probably doesn't

5 reflect how much clearing that occurs in patients

6 with a PASI 75 response.

7 The other point had to do with dosing by

8 weight. It seems to me that the company seems to

9 be resistant to dose by weight but yet there has

10 been some evidence here that dosing by weight may

11 have been better in some respects. For example,

12 some of the heavier people were underdosed and some

13 of the lighter people had to have their dose

14 withheld because their CD4 counts dropped too low.

15 So is it too late to dose by weight?

16 DR. VAISHNAW: Just to go to that issue.

17 We found an evidence, just as Dr. Marzella pointed

18 out, of diminishing response at the higher weight

19 ranges in the IV study but not in the IM. So the

20 IM route provides an option for patients across all

21 weight ranges.

22 Now, in the 10 milligram group in the IM

23 study, yes; there was also a slight loss of

24 response at the higher weight brackets, but the 50

25 milligram--you know, our conclusion of the data has

190

1 been that we don't conclusively show that kind of

2 trend.

3 So there is a validated dose option and

4 route for the full spectrum of patients. With

5 respect to the IV, we acknowledge the point that

6 has been brought up by the agency and we look

7 forward to working with them whether we need to do

8 further studies to determine the optimum approach

9 in the heavier patients via the IV route.

10 DR. SEIGEL: I would just like to comment

11 that the lack of a dose response observed in the 15

12 milligram IM population was based on the heaviest

13 quartile--well, not exactly quartile, but the

14 heaviest subpopulation you saw there had a 22

15 percent response. That was six responders out of

16 27 patients. A confidence interval around that

17 range could include that that true response range

18 was well under 10 percent, not 22 percent.

19 So we do not know that there isn't a dose

20 response on the 15, or a weight-related response on

21 the 15. It may well be we simply don't know.

22 I would also add that in terms of is it

23 too late, I am not sure that the agency would be

24 comfortable recommending a higher dose than tested

25 in heavier people because there are suggestions

191

1 that it may work better, but not all toxicities or

2 efficacies vary with weight. But what certainly

3 wouldn't be too late to do would be to look at

4 whether the tested dose versus a somewhat higher

5 dose, for example, in heavier people--whether a

6 higher dose had a better efficacy-safety profile if

7 we were interested in that. So, further study

8 could be done.

9 DR. DRAKE: I think, Lloyd, you had a

10 comment on this?

11 DR. KING: Just a follow up. Body weight

12 can reflect large people who are not obese if you

13 are thinking football players, et cetera. It also

14 can reflect adult-onset diabetes. That is often

15 used as the marker. Since people with diabetes are

16 less likely to respond well to treatments for

17 psoriasis and are likely to have increased

18 susceptibility to infections, it seems to me that

19 there is a surrogate marker that you may want to

20 look at rather than just say big people.

21 To distinguish this body weight over 100

22 kilograms predisposes to IV decreased

23 responsiveness, I suggest that the sponsor consider

24 using serum hemoglobin A1C as a surrogate marker

25 for decreased responsiveness to treatment and

192

1 predisposition potential to infections.

2 DR. DRAKE: Thank you, Lloyd.

3 DR. VAISHNAW: Thank you for your comment.

4 DR. KING: Then I have a second comment.

5 DR. DRAKE: I have already taken you out

6 of order. Go ahead and finish it up.

7 DR. KING: According to where you are,

8 similar observations that all politics are local, a

9 general assumption is that immune reaction and

10 psoriasis are ultimately localized to the affected

11 skin. In essence, the alefacept is targeting the

12 entire population T-cells to deplete the terrorist

13 T-cells that are going to target the psoriatic

14 skin. Surrogate markers, other than just measuring

15 just cell population, being the ultimate product

16 would be quite helpful.

17 It seems to me that, since the sponsor has

18 already done a preliminary study, studying

19 psoriatic arthritis using serum C-reactive protein

20 as a marker for inflammation, it would seem

21 appropriate to use that signature for psoriasis not

22 affecting the joints.

23 So C-reactive protein would be a great

24 marker for that since it is also a marker for

25 things like atherosclerosis and inflammation in

193

1 general.

2 DR. DRAKE: I am going to move to Dr. Epps

3 in just a minute but I saw Dr. Wilkins in here

4 earlier. This PASI thing keeps coming up. Is he

5 still in here? There he is. John, do you have

6 anything to add? Dr. Wilkins was kind of the FDA

7 honcho on those October meetings on the PASI. I

8 thought you might have something to add to what has

9 been said.

10 DR. WILKINS: No. This is a CBER meeting.

11 DR. DRAKE: I know it is a CBER meeting.

12 I read all these transcripts last night. I thought

13 I had it in my head but I thought, well, I will

14 just double-check with you, Dr. Wilkins to see if

15 we have missed anything. All right.

16 Now that we have digressed. Dr. Epps. I

17 am going to ask you because you haven't had a

18 question yet and then I want to go to the people

19 who have second rounds of questions.

20 DR. EPPS: I just have a couple of quick

21 questions, hopefully. The drug we are referring to

22 right now is the human fusion protein. Without

23 revealing secrets, what does that mean?

24 DR. VAISHNAW: No secrets. The

25 extracellular domain of LF3--

194

1 DR. EPPS: No; I mean is it pooled

2 products? Is it recombinant?

3 DR. VAISHNAW: Oh; it is recombinant. It

4 is a recombinant fusion protein produced by a

5 mammalian cell line.

6 DR. EPPS: Okay; great. Is there any idea

7 what the etiology to the transient neutrophilia

8 might be?

9 DR. VAISHNAW: Dr. Marzella pointed out

10 some findings from some of those smaller, earlier

11 studies. In the Phase 3 studies and Phase 2

12 studies where we have very large analyses of over

13 1300 individuals, we failed to confirm any evidence

14 for alefacept changing neutrophil levels. So we

15 don't know how to consider the significance of

16 that.

17 DR. SEIGEL: So you had measured, like, 4

18 hour and 24 hour--I mean, that when it was seen in

19 the first study. You measured that in the 1300

20 patients?

21 DR. VAISHNAW: Oh, right. No; that is a

22 point of clarification. We didn't. Those were

23 measured at weekly intervals. But if there had

24 been a sustained effect on neutrophils, then I

25 would say we would probably have detected it given

195

1 the approach to the studies and we failed to see

2 that.

3 DR. EPPS: In regards to the delayed type

4 hypersensitivity and tetanus and diphtheria, have

5 any of those patients been retested or would they

6 respond to a booster?

7 DR. VAISHNAW: The best way to answer that

8 is to go back to that graph that was in error, but

9 it would make the point for us to answer your

10 question.

11 If we could have the phi-X.

12 [Slide.]

13 What we have here is that, at the index

14 point here, when patients are in the middle of

15 dosing, they had had challenge with phi-X-174.

16 Then, six weeks later, they are being rechallenged.

17 So it is the surrogate for a booster that we would

18 do with a conventional immunization. You can see

19 that there is a brisk rise which parallels the

20 changes in the control group.

21 The other thing to point out is that the

22 IgG content in both groups is identical which is

23 reassuring regarding the integrity of the memory

24 cells to help the B-cells despite the action of

25 alefacept

196

1 What you are looking at here on the left

2 is the percentage of patients that had IgG greater

3 than 30 percent in their phi-X-174 response. You

4 can see control and alefacept are identical. Then

5 these patients went on to have further challenges

6 in the follow-up period and that is the third and

7 fourth. We didn't do that in the control group.

8 When they had the third and fourth challenges, they

9 did boost their responses further and the responses

10 were in a logarithmic scale on the last.

11 The ultimate responses at the fourth

12 challenge were exactly what is reported in the

13 literature for this antigen for which there is a

14 lot of existing information.

15 With respect to the booster with tetanus,

16 we also identified that tetanus immunization in

17 this same study was associated with a twofold rise

18 in both control and alefacept groups as predefined

19 in the study.

20 DR. EPPS: Lastly, according to your

21 protocol, you had a four-week washout period for

22 systemic immunosuppressants. Do you think that

23 that may be too brief and, perhaps, the prolonged

24 depression in the CD4 counts may be due to a

25 confounding factor or some kind of a synergy there?

197

1 DR. VAISHNAW: That is an issue we

2 analyzed by looking at patients that had or had not

3 had systemic agents or UV prior to the onset of the

4 immunotherapy with alefacept. You don't find any

5 significant changes in the pharmacodynamic profile

6 in those that are coming off those agents and then

7 going on to alefacept versus those that are not

8 coming off those agents.

9 For the same reasons that I think you are

10 intimating, we also looked at the safety profile by

11 that type of analysis and we found no difference if

12 patients had previously been exposed to

13 immunotherapies versus if they had.

14 DR. EPPS: So there may be suppression

15 regardless of whether or not they had been on it.

16 DR. VAISHNAW: In other words, the changes

17 that we are witnessing and discussing today are the

18 effects of alefacept rather than a combination of

19 effects from previous agents and alefacept.

20 DR. DRAKE: Dr. Marzella, you had a

21 comment?

22 DR. MARZELLA: I wanted to follow up on

23 the question of neutrophilia because potentially it

24 is a signal that alefacept may be inducing some

25 activation of inflammatory or chemotactic factors.

198

1 One reason that I think that it was striking how

2 elevated it was in the Phase 1 studies.

3 The other point that is relevant, as has

4 been pointed out, a lot of the patients in the

5 studies have a great deal of cardiovascular risk

6 factors. So there is a high proportion of

7 cardiovascular events--well, I shouldn't say a high

8 proportion, but I was struck looking at the

9 listing, by how many patients had cardiovascular

10 events.

11 So I think it is reasonable to ask whether

12 there is some potential relationship and to look

13 further into this issue of what is the potential

14 significance of the neutrophilia.

15 I know that it is not associated with--I

16 didn't notice any drops in platelet counts. There

17 was no fever. But I think it is potentially

18 something that might be followed up.

19 DR. VAISHNAW: I take your comments--

20 DR. DRAKE: Dr. Stevens.

21 DR. STEVENS: I have a number of

22 questions. Just a follow-up to that last one. Do

23 neutrophils express the appropriate FC receptor to

24 bind this molecule?

25 DR. VAISHNAW: You know, I am not an

199

1 expert on that. The answer is yes. I am getting a

2 nod from my scientific colleague here. I don't

3 know about the expression levels and whether they

4 can support the kind of mechanism that we are

5 describing.

6 DR. SEIGEL: I was just going to

7 interject. That also speaks to part of our concern

8 about safety. I think we agree with the company

9 that, in this experience, we haven't seen any

10 signal of the types of opportunistic infections you

11 would find with T-cell depletion. But the immune

12 system is complex. CD2 exists on CD8 cells, CD4

13 cells. It exists on some B-cell precursors and

14 some other cells in the immune system.

15 LFA exists on some of those cells. FC

16 receptors exist on a broad variety of cells. All

17 of those cells interact with each other and the

18 cytokines that the CD4 cells make interact and

19 activate all of those cells.

20 So there exists at least as theoretical

21 possibilities that any aspect of immune--or

22 inflammation can be influence. If the finding of a

23 neutrophilia, somewhat transient, but highlights

24 that, I think, as an issue.

25 DR. STEVENS: That brings me to another

200

1 one of my questions which is can you educate me on

2 the role of CD2 in T-cell ontogeny. We are going

3 to be asked to consider the use of this in

4 children, perhaps young people. Can you tell us

5 whether CD2 is important in the development of T-cell

6 responses during young childhood and

7 childhood, role in thymic development, et cetera?

8 DR. VAISHNAW: Now you really have me at a

9 weakness. Either Dr. Krueger or--Jim, do you want

10 to come up?

11 DR. STEVENS: I won't ask you to do math.

12 DR. KRUEGER: There aren't good human data

13 on that but there have been knockout mice made with

14 the CD2 deficiency. Those mice develop T-cells

15 normally. The immune abnormality that exists, if

16 you will, in these animals is that they appear to

17 be about tenfold less susceptible to a given

18 concentration of antigen, and that is we think CD2

19 dials up, or dials down, the threshold at which T-cells

20 become antigen-activated.

21 So I think, from that, and I will admit

22 that that is not completely reassuring data for

23 humans since there may be some differences in

24 development. But, to the first step, it says that

25 there should be a developmental problem. What

201

1 there might be is some differences, then, in let's

2 say immunization responses or some other

3 acquisition of acquired immunity in some early

4 childhood period when immunologic memory is being

5 acquired.

6 So I wouldn't want to go back too early in

7 terms of kids that are exposed.

8 DR. VAISHNAW: Thank you, Dr. Krueger.

9 DR. STEVENS: Thank you. To follow up on

10 Dr. Morison's question, you have shown data that

11 does not appear to affect primary immunization or

12 transition from naive to memory in a T-dependent

13 humoral immune system as well as minimal effect,

14 possibly, in the recall cell-mediated immunity

15 system. Do you have any data about the transition

16 of naive to memory in cell-mediated immune process

17 such as contact hypersensitivity or in DTH, itself?

18 DR. VAISHNAW: We don't have that. We

19 have been working with the agency throughout the

20 program to try and conduct immune test systems that

21 are reliable, reproducible across multiple centers

22 and where we can interpret the data. You have seen

23 two aspects to that. You have seen the DTH and we

24 have discussed the pros and cons of that data

25 there. You have seen the other approach which has

202

1 been more robust across multiple centers, and that

2 is the phi-X approach.

3 But we don't have data to that point. The

4 only point I would make is given that some of these

5 things are difficult to assess in a controlled

6 fashion because of the types of assays involved, we

7 have repeatedly asked ourselves the question what

8 is happening in the safety database.

9 The corollary to a defect in the kind of

10 conversion you are talking about is evidence of

11 opportunistic infections or a pattern of infections

12 that are suggestive of problems in terms of T-cell

13 immunodeficiency and we have failed to detect that.

14 I guess my concern also didn't come only

15 from infection but also the hint that, perhaps,

16 there may be an increase of malignant risk in

17 treated patients. So it was more that rather than

18 infection that was bringing that concern

19 DR. DRAKE: Dr. Morison has a comment.

20 DR. MORISON: I would agree with that.

21 That is the reason I raised the DNCB assay, an

22 assay which is reproducible across multiple

23 centers. It is an easy assay to do. There is

24 correlation, at least in the mouse and, to some

25 extent in the human, that if I had to develop a DTH

203

1 response to a contact sensitizer like that, it is

2 correlated with the development of skin cancer.

3 So there is good reason for doing that,

4 not just looking at the immune system and it is

5 quite separate and distinct from the infector in

6 infectious diseases.

7 DR. VAISHNAW: With respect to the point

8 of the potential for a signal in the malignancy

9 situation, maybe I could just review the squamous-cell

10 carcinoma rates that we observed because

11 squamous-cell carcinoma in many other settings

12 where there is high intensity of duration or

13 immunodeficiency is a good signal for occurrences

14 of--it is a good sentinel event indicating

15 significant immunodeficiency.

16 [Slide.]

17 In the placebo-controlled comparisons, I

18 think both Dr. Marzella and my colleague pointed

19 out that there was a numerical excess of squamous-cell

20 carcinomas in the alefacept-related patients.

21 Because of the excess numbers of patients in the

22 alefacept group versus placebo, in those

23 comparisons, we have been concerned whether it is a

24 kind of false-positive signal.

25 The only way we have found to try and

204

1 contextualize the rates we have observed is this

2 type of comparison where you look at the rate in

3 the alefacept placebo-controlled studies at 12.5

4 squamous-cell carcinoma per 1,000 patients years,

5 in the entire alefacept database, where we have

6 1,056 patient-year experience, you can see the rate

7 is stable. It is 13.3. These are patients that

8 are going over multiple courses.

9 So, if there was significant ongoing

10 immunosuppression, one might detect an elevation in

11 this rate here. Finally, at the bottom, you see

12 the expected rates that Drs. Stern and Margolis and

13 others who have been trying to address this issue

14 in the literature have documented.

15 So, at least from these comparisons, at

16 present we have concluded that the rates that we

17 have documented are within those expected. In the

18 sense of what is in store for the future, clearly,

19 as we indicated and as Dr. Marzella indicated, this

20 is a topic that is going to give continued study

21 for us because we are obliged to do that. It is

22 new therapy and a registry should help us address

23 that.

24 DR. DRAKE: Dr. Stevens, are you done?

25 DR. STEVENS: I had another question on

205

1 the topic, if somebody had a follow-up question--

2 DR. DRAKE: You have another question.

3 Dr. Abel, was your comment on this?

4 DR. ABEL: It relates, in a way, to side

5 effects and skin potential carcinogenicity and skin

6 cancer.

7 DR. DRAKE: Is it a question or a comment?

8 DR. ABEL: It is a question as to whether

9 we have data, and you may have mentioned this

10 already, in the patients who did develop cutaneous

11 malignancies, what their prior treatments were that

12 made them at risk; in other words, the PUVA-treated

13 patients would be, perhaps, at greater risk.

14 DR. VAISHNAW: We can go through that.

15 DR. ABEL: Cyclosporine.

16 DR. VAISHNAW: I haven't shown you the

17 data but we have those data for you if you wish to

18 review them. Would you like to do that?

19 DR. ABEL: I don't know if we need to do

20 that now.

21 DR. DRAKE: That is sort of borderline

22 between question and discussion.

23 DR. ABEL: It brings up issues as far as

24 recommendations and contraindications with regard

25 to prior--

206

1 DR. DRAKE: It brings up all kinds of

2 issues. If you would just address the facts and

3 then we will do the discussion this afternoon. If

4 you have a factual slide you want to show us.

5 DR. VAISHNAW: There is a factual slide.

6 DR. DRAKE: I figured you had one. You

7 are very good. I am impressed.

8 DR. VAISHNAW: I will ask my colleague,

9 Dr. Vigliani, to step up and walk you through this.

10 It is a little bit busy.

11 [Slide.]

12 DR. VIGLIANI: These represent each of the

13 individual patients who experienced squamous-cell

14 carcinomas within the study population. We have

15 indicated here the patients by course as to when

16 they developed these squamous cells. What you see

17 is that the majority actually were observed within

18 the first course and then there were additional

19 squamous cells reported in subsequent courses,

20 although the subsequent course diagnoses of skin

21 cancers actually were restricted to a couple of

22 patients who seemed to be experiencing multiple--if

23 we take the first patient, for example, in looking

24 at the baseline history, we see that that patient

25 who accounts for, actually, a total of six

207

1 squamous-cell cancers had a prior history of

2 squamous-cell cancers, had a prior history of PUVA

3 as well as UVB, methotrexate and cyclosporine.

4 So you see that there are a number of

5 preexisting risk factors based on prior therapies

6 as well as, in some patients, prior history of

7 squamous cell.

8 We actually have a slide that looks at

9 baseline characteristics that just defines this

10 across the entire database.

11 [Slide.]

12 In this slide, what you see are some

13 baseline characteristics of the patients indicated

14 on the left. On the top of the slide, you see the

15 proportion of alefacept-treated patients who

16 developed squamous cells and/or basal cells and how

17 these risk factors compared to patients in the

18 entire alefacept population.

19 So, looking at a prior history of

20 squamous-cell or basal-cell, what you see is that,

21 for squamous cells, 25 percent versus 1 percent

22 developed squamous cells had a prior history of

23 squamous cell. You can see similar imbalances for

24 prior treatment.

25 So I think what we can conclude from this

208

1 is that patients who developed these cancers were

2 patients that were at high risk.

3 DR. VAISHNAW: I think the other point

4 that, perhaps, we should make here is that, at

5 baseline, we noted that, given that squamous-cell

6 carcinoma, itself, is a predictor of subsequent

7 risk of squamous-cell carcinoma, there was an

8 imbalance between alefacept and placebo groups.

9 The placebo group was one individual that had had a

10 previous SCC. In the alefacept group, there were

11 eleven individuals. So that, perhaps, also plays

12 into the debate.

13 DR. DRAKE: We are running into lunch time

14 and I want to make sure people have time to grab a

15 bite to eat because people get cranky when they

16 don't eat. We don't want to fool around with that.

17 I have Dr. Katz left on my list and Dr.

18 Swerlick left on my list. You are okay? No more

19 questions? Anybody else with questions?

20 DR. STEVENS: I still have one more

21 question. I yielded for the follow up.

22 DR. DRAKE: You yielded for the follow up.

23 I understand. So you are next and then Dr. Katz.

24 Dr. Raimer, do you have any questions?

25 DR. RAIMER: No.

209

1 DR. DRAKE: Ms. Knudson, do you have any

2 questions?

3 MS. KNUDSON: My questions have to do with

4 adding children and that can come later.

5 DR. DRAKE: Okay. So we will do Dr.

6 Stevens' last question and then Dr. Katz' question

7 and then we will move to lunch and then reconvene.

8 Dr. Stevens?

9 DR. STEVENS: Thanks. I am trying to

10 integrate all the information that you gave us with

11 respect to the CD4 counts effects on--or T-cell

12 counts and the effect as well as potential safety

13 issues. You showed us that it took about six weeks

14 to really knock out the T-cell population, yet you

15 were dosing for twelve weeks.

16 I wonder about the variability between

17 patients in their attainment of that lymphopenic

18 state or relative lymphopenic state. I want to get

19 an understanding of why the monitoring is at 250

20 cells per microliter, why that, maybe, is a magic

21 number. Could we increase the potential safety or

22 further ameliorate the safety questions by raising

23 that threshold to a higher point.

24 There were a number of patients in whom

25 you withheld doses because of the lymphopenia. So

210

1 the question is was this repeated lymphopenia in

2 the same patients or one episode spread out evenly

3 among a number of patients. I guess, ultimately,

4 what I am getting at is trying to understand the

5 cutoff for holding the dose and also the rationale

6 behind the twelve weeks of dosing rather than some

7 other number.

8 I guess the other factor that plays into

9 that is the amount of time after you have finished

10 dosing patients in which they maintain this

11 relative lymphopenic state.

12 DR. VAISHNAW: So there were several

13 questions there. Let's go one by one. I think the

14 first one was the issue of the rates of dose

15 omission because of a CD4 count under 250. If we

16 looked in the Phase 3 studies, obviously the most

17 controlled setting, 10 percent of patients in the

18 IV study had that kind of transient dip and needed

19 a substitution. It was 5 percent in the IM.

20 Then you mentioned the issue of, well, are

21 there patients that get a more kind of multiple

22 count below 250 and would require multiple

23 substitutions. Indeed, there were 2 percent of

24 patients in the IV study had that type of event in

25 the first course and when the same patients were

211

1 retreated in the second course, there were none.

2 For the Phase 3 IM study, no studies had multiple

3 counts under 250 of the type you describe.

4 Now, the question of the choice of 250 has

5 been important to us. We have thought very hard

6 about it. The low limit of normal is 404 for CD4

7 T-cells. A CD4 count of 300 was elected in the

8 Phase 3 studies. We saw very encouraging safety

9 profile with that.

10 For Phase 3, the agency worked with us on

11 the designs on those studies and they were aware of

12 the threshold that we picked which was 250. You

13 have seen the safety, efficacy and other data in

14 relation to regulating dosing around that

15 threshold.

16 A couple of things, looking back at this

17 whole experience maybe that are important to

18 acknowledge is that we have been intrinsicly

19 conservative and we should have been and we are

20 because we don't understand everything there is to

21 understand about alefacept lymphocyte safety and

22 efficacy although I might act as if I might.

23 We have a lot to understand and we want to

24 be conservative. We have a count of 250 because we

25 understand the safety profile around that now. We

212

1 propose moving forward with that. As multiple-course

2 experience increases and our safety profile

3 is defined over multiple courses, I think we can

4 revisit the issue of whether 250 is or isn't. At

5 the moment, we have data that supports 250 as a

6 rationale choice.

7 The final thing I would say about the

8 choice of 250 is that it is very much--it is all to

9 do with what is happening in the blood. It does

10 not necessarily mean that this is what is going on

11 in the extravascular compartment. If you look at

12 the individual patient profiles over time, and for

13 those patients that got infections, you very often

14 see a brisk rise in lymphocyte count far above

15 normal, in fact.

16 What that teaches us is that we are

17 looking in the blood. There is massive repository

18 outside the blood and the function, there, of those

19 lymphocytes is described by the safety profile and

20 in the lymphoid tissues by the phi-X-174

21 experience.

22 So I have given a long-winded answer, but

23 I think I have addressed most of your points.

24 DR. DRAKE: Dr. Katz.

25 DR. KATZ: Getting back to the clinical

213

1 study, and maybe I missed it in the briefing book,

2 but the people who recorded these rather minor side

3 effects like chills, were they the same people

4 evaluating the patient for improvement?

5 DR. VAISHNAW: Whether people getting the

6 chills were the ones that achieved significant

7 improvement?

8 DR. KATZ: No.

9 DR. VAISHNAW: I'm sorry.

10 DR. KATZ: Was the same investigator the

11 same physician evaluating chills, IM reaction, as

12 was evaluating improvement in the PASI?

13 DR. VAISHNAW: Yes. So the clinical

14 examination of patients was by a blinded

15 investigator who was evaluating both the PASI and

16 the physical status of the patient from the safety

17 viewpoint; yes.

18 DR. KATZ: I may have missed in the

19 briefing book, what percentage had IM reactions the

20 first time?

21 DR. VAISHNAW: We can address that--I'm

22 sorry?

23 DR. KATZ: What percentage of the patients

24 getting the drug had that?

25 DR. VAISHNAW: I will ask my colleague,

214

1 Dr. Vigliani, to walk you through the data that we

2 have addressing that.

3 DR. VIGLIANI: As I mentioned in my

4 presentation, if you look at the overall integrated

5 database, you would actually find that less than 5

6 percent of patients had injection-site reactions.

7 However, we did see a higher frequency in the IM

8 study.

9 I will just present to you here the data

10 on injection-site reactions from that study.

11 [Slide.]

12 What you see was that there were 8 percent

13 of patients with an injection-site reaction in

14 placebo, 13 percent in the 10 milligram and 19

15 percent in the 15 milligram. These are any

16 injection-site reaction.

17 If you look at the number of injections

18 that were associated with an injection-site

19 reaction, counting the total number of injections,

20 you see that the majority of injection-site

21 reactions were reported on one occasion, some on

22 two and infrequently with multiple injections.

23 [Slide.]

24 Just to further characterize the

25 injection-site reactions by severity, on this next

215

1 slide, what we see is that the majority of

2 injection-site reactions or 84 percent in the 15

3 milligram group were mild, 16 percent moderate and

4 no severe injection-site reactions.

5 In the IM Phase 3 studies, we had no

6 patients discontinuing due to injection-site

7 reactions.

8 DR. KATZ: I would like a comment,

9 perhaps, from the group statisticians, as far as

10 blind goes, I was concerned about the severity of

11 the injection-site reactions. Do you think this,

12 in part, negates the blind of the study because

13 there is 11 percent more injection-site reactions

14 seen by the physicians evaluating that, number one,

15 and, number two, the 6 percent chills versus 1

16 percent.

17 Considering the margin of efficacy, we are

18 talking about 10 percent, 25 percent. Are we

19 talking about something relevant? Can we have the

20 statistician comment on that?

21 DR. VIGLIANI: Can I just put back up the

22 injection-site reaction slide again, that first

23 one, just to look at what types of injection-site

24 reactions these were, or maybe I don't need the

25 slide. But the most frequent injection-site

216

1 reaction actually was just injection-site pain.

2 No; I guess I don't have a slide of that. Sorry.

3 So the most frequent injection-site

4 reaction was pain.

5 DR. KATZ: It was 19 percent versus 8

6 percent. The other thing was on the chills. I

7 have another question for Dr. Lebwohl and then I am

8 finished, Lynn.

9 DR. DRAKE: That's fine.

10 DR. KATZ: Mark, first of all, thank you--

11 DR. DRAKE: Mark, how come you keep

12 standing between us and break? Have you noticed

13 that this morning?

14 DR. KATZ: Mark, thank you for your

15 clinical slides which had answered questions of

16 mine, not being used to these studies, what is 50

17 percent, what is 75 percent. I certainly would

18 agree with you that 50 percent is, in a clinical

19 basis, very much appreciated by the patient.

20 I would revise my thought that 50 percent

21 isn't so great and would agree with you that is

22 quite impressive. However, you used the figure of

23 60 percent of people comparing to methotrexate. I

24 am sure, clinically, that is going to be a clinical

25 judgement for everybody and I appreciate your

217

1 experience because you have more than anybody else.

2 But you say 60 percent respond yet, even

3 with a PASI of 50 over the placebo, there is only

4 24 percent response. That is in the IM study.

5 There is a 9 percent clear or almost clear over

6 placebo. So when you consider the experience we

7 have with methotrexate of whatever--Figure 1 in the

8 briefing book, it said 60, but I think usually

9 85 percent is quoted and they get equal response.

10 I wondered why you would say you would pick this

11 over methotrexate as a drug.

12 DR. LEBWOHL: First of all, largely

13 because of toxicity. I think first the

14 hepatotoxicity, which is long-term, which I think

15 we can monitor for, but secondly those occasional

16 instances of pancytopenia that happen because of

17 accidents that happen out there. I view

18 methotrexate, at least with what we know about it

19 and, admittedly, we don't have long-term data on

20 alefacept, but short-term, I do believe that this

21 is a safer drug.

22 That is why I would put this ahead of

23 methotrexate. As far as efficacy, no question

24 methotrexate is a highly effective therapy. I

25 think that before we started using PASI 75 or clear

218

1 or almost clear as endpoints, if you ask me how

2 often does it work for methotrexate, I would say 80

3 percent of the time.

4 You said 85 percent of the time. I think

5 if you applied the same bars, you would find the

6 numbers probably a little bit higher than alefacept

7 but not as much as you think. Someone told me that

8 there was a poster at the SID that did that and, in

9 fact, found the two comparable.

10 Lynn mentioned the October meeting of the

11 FDA in which this high bar was discussed. Part of

12 discussion was even if only 5 percent of patients

13 achieved the endpoint because they knew they were

14 advocating very high endpoints, as long as it was

15 statistically significant, it would pass.

16 I think that what we are looking at here

17 is precisely that scenario. You know, we are

18 looking at the drug that the patients were very

19 happy getting, the patients who responded were

20 ecstatic getting. But a lot of the patients who

21 were ecstatic didn't achieve PASI 75 exactly two

22 weeks after they finished dosing.

23 The other issue that you mentioned with

24 Dr. Vigliani I want to say that the chills were in

25 the IV study, I believe. Is that right? In the IM

219

1 study, I don't think the chills occurred. I don't

2 recall. I don't think that, to the investigators,

3 that pain at the site of injection certainly didn't

4 lead us to believe that that was active or placebo.

5 That was only the first one or two injections.

6 So I don't think that we could have

7 distinguished the patients on the basis of pain at

8 the site of injection and the chills were in the IV

9 study, not the IM.

10 DR. KATZ: Thank you.

11 DR. VAISHNAW: Could I just add a brief

12 comment to that. The database that we have is

13 interesting to probe from a variety of viewpoints

14 and it gives interesting insights into the unmet

15 need in this population.

16 About 10 to 20 percent of patients at

17 baseline had abnormal liver-function tests. I

18 think it kind of underscores the point that Dr.

19 Lebwohl has just been making about the potential

20 for the current agents and where the scope of new

21 agents is to help patients like that. 10 percent

22 of patients had a hypertension at baseline and they

23 would be concerned about cyclosporine.

24 DR. DRAKE: What I would like to do now is

25 two things. First of all, I want to thank the FDA

220

1 and sponsor for wonderful presentations. I have no

2 doubt that the sponsor will hang around for this

3 afternoon for the discussion. That is sort of a

4 given.

5 But I would also hope that Dr. Lebwohl and

6 Dr. Krueger, your comments and your expertise have

7 been most appreciated and I hope you will be

8 available to the committee this afternoon if we

9 have specific questions. We would very much

10 appreciate it.

11 Let's aim for--I this is a short lunch.

12 I'm sorry. But still we need to try to aim for

13 1:30 because of the public comment. We are in

14 recess until 1:30.

15 [Whereupon, at 1 o'clock p.m., the

16 proceedings were recessed to be resumed at 1:30

17 p.m.]

221

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

2 [1:40 p.m.]

3 DR. DRAKE: With respect to this

4 afternoon, we have a very ambitious agenda to say

5 the least. I must compliment the FDA. These

6 questions are terrific but there are a lot of them.

7 The only critique I can make is this should have

8 been a day-and-a-half meeting, I swear, because

9 this biologic is a new one for dermatology.

10 We are asking lots of questions and the

11 committee is involved. It is fun to see this kind

12 of intellectual dialogue with everybody just trying

13 to do the right thing here. So I am tickled.

14 I had a question or two that I wanted to

15 ask. This is going to be directed towards the

16 sponsors. I know it is all time-and-done, for the

17 sponsor to be done, but I saved my question. Dr.

18 Marzella had a slide that was on animal toxicity.

19 I was interested because it was kind of before all

20 the data was in.

21 What I was quite interested in is could

22 the FDA or the sponsor--and, by the way, I gave

23 both the FDA and the sponsor notice ahead of time

24 that I was going to ask this question so everybody

25 could kind of have their act together here, but I

222

1 want to know what the recent status of the animal

2 studies are. I want an update because I think one

3 of the most serious things that this committee will

4 have to consider is the safety issue.

5 That is clearly foremost on everybody's

6 mind and I want to know if there is an update, any

7 more recent information, on studies with respect to

8 animals and primates. Who has the information on

9 that because there is always last-minute

10 information but it doesn't make it in our book.

11 DR. VAISHNAW: I will invite my colleague

12 from Biogen to comment on that.

13 DR. GREEN (BIOGEN): Good afternoon.

14 DR. DRAKE: You are?

15 DR. GREEN (BIOGEN): My name is James

16 Green and I am referred to as the chief

17 toxicologist at Biogen at times like this.

18 DR. DRAKE: Welcome.

19 DR. GREEN (BIOGEN): I am currently Vice

20 President of a group called Preclinical and

21 Clinical Development Sciences and I am intimately

22 involved in this study as well as well as worked

23 with the FDA on a number of these issues over the

24 past.

25 To update briefly, I think what I will do

223

1 is just give you a general sound bite of what the

2 overall profile of the safety program looks like

3 for alefacept in animals. You heard the incidence

4 of lymphoma, single incidence. That was one

5 incidence of B-cell lymphoma that was observed out

6 of 228 animals, primates that had been treated with

7 alefacept, one out of 228 animals that have been

8 treated with various courses of alefacept from

9 periods ranging from three months to one year.

10 With the exception of the lymphoma that

11 Dr. Marzella described and Dr. Green reported, the

12 profile in primates is one that is relatively

13 uneventful, no opportunistic infections for animals

14 treated at high doses for periods ranging from one

15 month to 52 weeks, for doses that are

16 pharmacologically active and superpharmacologically

17 active.

18 The hallmark tissue change that would have

19 been observed consistently in studies of one-month

20 duration up to 52 weeks would be a subtle decrease

21 in the T-cell-dependent regions of the spleen or

22 the lymph nodes. This is a truly expected effect.

23 It is one that we have seen consistently between

24 studies and, in fact, it is one that is very, very

25 subtle in nature.

224

1 One of the comments that I will make about

2 the 52-week study which is in contrast to some of

3 the shorter-term studies which went from one month

4 to three months is that 52 weeks of treatment is

5 high-dose intensity exposure, that is consecutive

6 weekly dosing.

7 It is very different than the clinical

8 regimen and the intent of that study is essentially

9 to identify possible alerts or possible flags. We

10 view, and I don't think we have any disagreement

11 with the agency on their interpretation, is that

12 the observation of this single lymphoma in heavily

13 treated long-term immunosuppressed animals is not

14 unexpected and, in fact, could be viewed relative

15 to other immunosuppressive agents and put in that

16 context.

17 DR. VAISHNAW: Just if I would close that

18 comment with some clinical commentary. As Dr.

19 Green just discussed, indeed cyclosporine-associated

20 lymphoma is also well-recognized in the

21 nonhuman primate starting at therapeutic regimens.

22 The prevalence of those in the nonhuman primate

23 setting is about 25 to 30 percent in the similar

24 species when parallel types of studies have been

25 conducted.

225

1 You have heard about the prevalence for

2 us. The clinical implications are clear to us.

3 [Slide.]

4 I can probably just close that last point

5 with this.

6 DR. DRAKE: I knew you would have a slide.

7 I just knew it.

8 DR. VAISHNAW: In the cynomolgus monkey

9 setting, if you look here on the far right, post-transplant

10 lymphoproliferative disorder which are

11 B-cell tumors occur at a prevalence of 25 to 30

12 percent in association with cyclosporine. So we

13 have a similar situation here that, with alefacept,

14 we have observed the one B-cell lymphoma. The

15 prevalence is nowhere near this, of course, but it

16 is a finding of note.

17 We are taking that data seriously. In the

18 clinical setting, we have observed no B-cell

19 lymphomas related to immunosuppression and we have

20 clearly made this a subject of long-term study and

21 we know we will have to study this in the post-approval

22 setting as appropriate.

23 DR. DRAKE: Dr. Seigel?

24 DR. SEIGEL: Just to be clear, then, you

25 said this is not unexpected in heavily treated

226

1 animals and you pointed that out. But you wouldn't

2 have expected this to occur spontaneously without

3 treatment, this sort of lymphoma; is that right?

4 DR. GREEN (BIOGEN): I think the

5 experience in nonhuman primates is that this is a

6 rare observation. These is relatively healthy

7 animals and, in fact, the conditions that have been

8 described long-term, high-dose, heavy pretreatment

9 are associated essentially with this kind of

10 observation that has been viewed in other contexts.

11 I think the important point with that

12 cyclosporine is that cyclosporine dose is the

13 therapeutic dose. In fact, that data was reported

14 several years ago at an advisory committee meeting,

15 a subcommittee of the xenotransplantation group

16 that was held with CBER.

17 DR. DRAKE: I saw Dr. Green step up to the

18 table from the FDA. I would like your comment on

19 my same question, please.

20 DR. GREEN (FDA): The most recent report

21 we have had from the company was last week,

22 approximately. At that time, they reported to us

23 the end-line portion of the 52-week weekly dosing

24 study in cynomolgus monkeys. In the original form

25 of this study, which was a nine-month study, there

227

1 was the incidence of the lymphoma that was observed

2 and then that was converted to a twelve-month study

3 which has just ended and now a one-year observation

4 period has followed for the surviving monkeys.

5 But I think of the findings which was

6 somewhat surprising, at least to me, was a

7 treatment-related localized hyperplasia of B-cell

8 lineage which occurred in three of six low-dose

9 animals, 1 milligram per kilogram, and five of five

10 of the high-dose animals which was the 20 milligram

11 per kilogram.

12 The importance of this finding is that it

13 is unclear as to what its origin is. It might

14 reflect a reactive or adaptive response but it

15 cannot be distinguished even by the committee we

16 have had from reviewing pathologist from those

17 cases which might represent an immune-suppressed

18 related hyperproliferative response.

19 So you have basically the situation of T-cell

20 suppression against a background of B-cell

21 proliferation in which there is, in the animal who

22 had the B-cell lymphoma, was also noted to have an

23 Epstein-Barr-like virus infection which is common

24 among these animals.

25 So the one-year observation period will be

228

1 an important aspect of determining the safety

2 profile of this particular biologic.

3 DR. DRAKE: This is very important for

4 those of you who might have wandered in late. I

5 apologize. We should have box lunches for the

6 committee members prepared and we will try to do

7 that in the future.

8 But I asked the question, for those of you

9 who walked in late, what was the most--I was

10 concerned about one of Dr. Marzella's comments

11 about toxicity in animals. I know so many of you

12 have been skirting around that issue and so I asked

13 what the most recent update was because there is

14 always stuff that they have that doesn't make it

15 into our briefing book.

16 You have just heard the company and the

17 FDA's perspective on it. So, if I understand this

18 right, there has just been one case of lymphoma but

19 there is also this B-cell proliferation that you

20 are seeing, or hyperplasia, rather, that you are

21 seeing in this group.

22 We are not quite certain what that means.

23 It could be a precursor or it could be. Dr. Green

24 from the FDA, would you clarify that just a little

25 bit more for me?

229

1 DR. GREEN (FDA): I think you are exactly

2 right. It is not known. I think it was surprising

3 that there was a hyperproliferative research. The

4 consequences of that hyperproliferative response

5 are basically unknown. They could possibly be the

6 harbinger of something adverse or they could be a

7 normal response which, over the course, the

8 recovery period, will diminish and not present any

9 issues.

10 But, at this point, that is an unresolved

11 point.

12 DR. GREEN (BIOGEN): I think the other

13 perspective that I could add to what Dr. Green has

14 added, again, viewing the B-cell hyperplastic

15 responses within the context of the single

16 incidence of lymphoma. We have had these

17 observations extensively peer-reviewed by

18 veterinary pathologists and human medical

19 pathologists. The conclusion that they reach is

20 they say, well, this is not an unusual kind of

21 hyperplastic finding that we see in heavily

22 immunosuppressed patients, patients that would be

23 in the transplant setting.

24 In fact, those animals that would have

25 been in the transplant dataset that Dr. Vaishnaw

230

1 showed, if looked at histologically, it would not

2 be unusual to see those similar kinds of changes.

3 They are categorized and recognized as uniformly

4 being reversible, nonneoplastic and it is not with

5 any probability that they progressed to anything

6 more serious when treatment is stopped.

7 We have other nonhuman primate data in the

8 registration submission that hasn't been discussed

9 here. But these studies have incorporated long-term

10 recovery periods and, as part of our peer-review process, we

11 have gone back and looked--these

12 are very, very subtle changes. It is only with

13 hindsight and foreknowledge of the single incidence

14 of lymphoma that these tissues have been looked at

15 very, very carefully.

16 What we have found is that we had seen

17 focal evidence in previously conducted studies of

18 the same kinds of findings, but when these animals

19 essentially were put on long-term recovery periods,

20 upwards of seven months, they completely reverse.

21 So that pattern is consistent with what I think the

22 human experience has been in patients that have

23 been heavily immunosuppressed.

24 DR. DRAKE: Dr. Green?

25 DR. GREEN (FDA): Just to provide a little

231

1 bit more information, as best I recall, there were

2 two longer repeat-dose studies in nonhuman

3 primates. One was a seven-month baboon study and

4 the other one was a 44-week cynomolgus monkey. The

5 study that was recently reported to us in unique in

6 the length of time that the animals were dosed.

7 As I recall, the 44-week cyno study didn't

8 have similar findings. So it may be that some

9 place between 44 weeks and 52 weeks, where just

10 running this study again produced these results. I

11 would also point out that, although there can be

12 honest disagreements about how to evaluate this

13 material, the lower dose, the 1 milligram per

14 kilogram dose is, in our opinion, clinically

15 relevant.

16 DR. DRAKE: But you said three out of

17 five.

18 DR. GREEN (FDA): Yes; with the low dose

19 DR. DRAKE: At the low dose, and five out

20 of five of the higher dose.

21 DR. GREEN (FDA): Yes. It is clearly a

22 pharmacologically active dose.

23 DR. VAISHNAW: I would agree with Dr.

24 Green that there are no findings that we have here

25 that are not of clinical relevance in terms of

232

1 trying to understand their implications for us in

2 the clinic. What we would say is that there is an

3 opportunity here to identify a subset of events

4 that we should focus on in the clinical setting.

5 In dosing 1500 individuals at the clinical regimen,

6 which contrasts very significantly with the regimen

7 that has been explored here in this nonhuman

8 primate setting, both in terms of dose, in terms of

9 duration and in terms of the intensity of exposure,

10 that we have not had any immunosuppression-related

11 lymphomas or lymph adenopathy in the human setting.

12 But we cannot disagree and acknowledge

13 that this is data of clinical relevance and

14 something that has to be the subject of studies as

15 the database expands in the postapproval setting.

16 We propose a registry type approach to understand

17 the incidence, if any, of immunosuppression-related

18 events like that.

19 DR. DRAKE: Thank you very much. I am

20 going to move to the public comment.

21 Open Public Hearing

22 I am very delighted to see public comment.

23 Sometimes, we don't have it at these meetings and

24 so it is delightful.

25 Gail Zimmerman from the National Psoriasis

233

1 Foundation. Welcome, Gail. We are delighted to

2 have you here.

3 MS. ZIMMERMAN: Thank you for that

4 introduction, Lynn, and I am glad to be here in

5 behalf of the National Psoriasis Foundation. I am

6 President and CEO. The Foundation was founded in

7 1968 by patients and physicians interested in

8 helping people with psoriasis and psoriatic

9 arthritis.

10 We spend our time providing information to

11 the public on psoriasis and also serve as an

12 advocate, we hope, effectively on behalf of

13 patients.

14 Our funding comes principally from

15 patients and their families. 70 percent of our

16 budget is from the public. 20 percent comes from

17 the pharmaceutical and biotech industry. 10

18 percent of our budget, of that money, goes to our

19 operating budget and the other 10 percent goes to

20 special projects, principally medical education for

21 physicians.

22 I am here today on behalf of the

23 foundation to communicate our support for the

24 approval of, if I may say, Amevive. The other word

25 I stumble over sometimes, alefacept. We support

234

1 that approval because we believe very strongly that

2 there is a need for more treatments. There are too

3 few treatments out there for people with moderate

4 to severe psoriasis.

5 I wanted to communicate the reasons we

6 believe that and also I have brought three members

7 of the Foundation who have psoriasis to let them

8 share briefly their story with you on coping with

9 the disease.

10 In the twenty years I have been at the

11 Foundation, I have discovered it is difficult for

12 many people to quickly appreciate the impact of

13 this disease. It is physical but it has a

14 tremendous emotional component that is often hard

15 to grasp if you are not intimately involved in

16 treating it or in working with patients.

17 I wanted to tell you briefly about a

18 survey we did this last couple of months. We did a

19 national survey funded by Biogen and Immunex-Wyeth-Ayerst.

20 We went to them. We saw an opportunity to

21 obtain funding to do a national survey, a public

22 survey, to measure the incidence of psoriasis and

23 psoriatic arthritis and to establish some

24 benchmarks about treatment. We were trying to find

25 out is it only our members that are in need of more

235

1 treatments or is everyone feeling the same way; is

2 it a representative population.

3 So we conducted this study and we finished

4 it in January. We defined moderate to severe

5 psoriasis as anything over 3 percent BSA. Based on

6 that, we concluded or estimated there are 1.5

7 million moderate to severe psoriasis patients in

8 the country.

9 In surveying them, in taking a small

10 random sample of that group, 78 percent said they

11 were not currently on any systemic therapy

12 primarily due to side effects of lack of efficacy.

13 That is a big number. Frankly, that reflects what

14 our membership has told us in our small member

15 surveys. There is a great reliance on topical

16 steroids, still.

17 So we feel very strongly that we want to

18 encourage new treatments. We feel that Amevive

19 offers a potential safety profile that makes it a

20 tool, a desirable tool, to add to the physician's

21 treatment kit. We think there are many patients

22 out there that would like this therapy because of

23 that potential safety profile and its ease of

24 administration.

25 So, with that, I want to just conclude to

236

1 say that I brought three members. These members,

2 two of whom have used Amevive, we have asked them

3 here because we wanted to hear--this is their story

4 to tell you how they felt after this treatment.

5 The third is a member who is not on treatment

6 currently, or has just started treatment, and who

7 has been on every treatment out there for psoriasis

8 just to give you a brief overview of how it feels

9 to make choices today about treatment and to live

10 with the disease.

11 Thank you.

12 DR. DRAKE: Thank you, Gail.

13 I guess the first one is Ms. Diane Lewis.

14 There is nothing like hearing from patients who

15 actually have to deal with this disease to

16 understand how important it is that we have good

17 therapies for them. You are really a hero to come

18 tell us about your experience, sharing your life

19 with us and we thank you.

20 MS. LEWIS: Thank you very much. Good

21 afternoon. First, I would like to say that myself

22 and the next two speakers are lay people. This is

23 our personal testimony and we are nervous and I ask

24 you please turn off your cell phones because that

25 ring could really throw us off. So, person-to-person,

237

1 please turn them off. Thank you.

2 My name is Diane Lewis. My age of onset

3 was nine after a strep-throat infection. I have

4 had this disease for twenty-four years. My family

5 has been members of the National Psoriasis

6 Foundation since 1986. I am currently in treatment

7 at the Psoriasis Daycare Center at the University

8 of California, San Francisco, under Dr. Ku. I am

9 using a combination of bath PUVA and topical

10 steroids.

11 My list of treatments include natural

12 sunlight, LCD 20 percent, topical steroids,

13 Dovonex, anthralin, gacrimin outpatient, which is a

14 combination of UVB and topical tars, systemic

15 steroids, Accutane, methotrexate three times. I

16 have had a liver biopsy and climatotherapy at the

17 Dead Sea three times.

18 That is just about everything that you can

19 possibly name. I have not been on cyclosporine.

20 For the last twelve years, I have had a total time

21 of either totally clear of less than 15 percent for

22 only four months. That is not very much. I am

23 generally totally covered. The highest I have ever

24 been is 95 percent.

25 The time factor of treatments is

238

1 extensive. It is hard to balance friendships,

2 career and a life with having to go to a

3 dermatologist or a day-treatment center all the

4 time. I have lost jobs over the fact that I had to

5 go into gracrimin. They would not hold my job for

6 me.

7 It has been also difficult for my

8 education as stress is a factor and finals is

9 always difficult and I have actually had professors

10 and universities say to me, "But it is just a

11 little skin thing." When I can't move and I can't

12 walk, it is not just a little skin thing.

13 In the last twenty-four years, I have

14 dealt with the shame that comes with psoriasis, of

15 wanting to cover yourself, of feeling like you have

16 no control over your body. It is very difficult.

17 The bonus of that is yesterday, when I was riding

18 the local metro, nobody would sit next to me so I

19 got to sit all by myself and I wasn't crowded. You

20 always have to find the silver lining.

21 There is intense isolation with this

22 disease. It is very difficult to communicate what

23 it feels like to constantly be in pain, itching,

24 not sleeping at night, waking up stuck to your

25 sheets because you are bloody, having blood stains

239

1 on your clothing and constantly having to dust

2 yourself.

3 There is also a fear of rejection. This

4 has affected my intimate relationships. It is very

5 difficult for somebody you are involved with for

6 you to say, "I'm sorry, but I don't want to be

7 touched right now and, not only that, I don't want

8 to be touched for the next three months." It

9 destroys intimacy.

10 It is also hard in friendships because you

11 don't want to burden your family and friends with

12 constant complaining but sometimes it is how we

13 feel. Growing up with psoriasis, it has been

14 difficult, as I become an individuated person, to

15 create an identity that is separate from psoriasis.

16 As such, in my early twenties, I went into a severe

17 depression for five years. For three of those

18 years, I was afraid to leave my home. I would

19 leave my house once a week to do my grocery

20 shopping and to see a therapist.

21 I was a total victim to this disease and I

22 have slowly climbed out of it to the point where,

23 in 1998, I was able to backpack by myself around

24 the world.

25 There is also intense desperation

240

1 associated with this disease, desperation to find a

2 treatment that works, desperation to find a doctor

3 who can deal with it. Not many dermatologists can

4 deal with the severity of my disease as they don't

5 have the instruments. There are actually

6 dermatologists who don't have phototherapy in their

7 offices and they will put you right onto

8 methotrexate or they will just keep giving you

9 topical steroids because they are not comfortable

10 giving you systemics.

11 It is very difficult finding a

12 dermatologist who can deal with this and I am very

13 lucky that I live in San Francisco and that I have

14 the Psoriasis Daycare Center where they are able to

15 give me a variety of options. Nonetheless, I have

16 to accommodate this disease. I have had to find a

17 profession that will allow me to have total

18 flexibility where I can take off three months at a

19 time to deal with my disease and be able to not

20 work 9:00 to 5:00 as, in the mornings, I have to

21 take two-and-a-half hours to go and have my bath

22 treatments.

23 I live three blocks from the Psoriasis

24 Daycare Center so that it is easy for me to go in

25 the morning and get my treatments and not blow it

241

1 off.

2 It is also hard to find piece of mind. I

3 want to tell you that, at one point, when I was

4 depressed, the level of desperation and my desire

5 to have relief would be that I would actually slice

6 some of my plaques off with an exacto knife for

7 that 10 seconds of relief so that the tightness

8 wasn't there, so that the itching wasn't there, and

9 it was the only way I could get it to go away

10 knowing full well that, within 10 seconds, intense

11 bleeding would start and I am sure immediate

12 keratinization. That is desperation.

13 There are not a lot of treatments out

14 there for severe psoriasis. I am a young woman. I

15 want to keep my liver and I want to keep my

16 kidneys. So I ask you to really consider this

17 treatment. I am very honored to represent all the

18 patients with severe psoriasis here in the United

19 States.

20 Thank you very much.

21 [Applause.]

22 DR. DRAKE: Thank you very much, Ms.

23 Lewis. Bless you for coming forward. It is very

24 helpful.

25 Is it Ms. Maryellen Crawford is next?

242

1 MS. CRAWFORD: I am here today. I came

2 with the National Psoriasis Foundation from

3 Portland. I am Maryellen Crawford. I am a

4 psoriasis sufferer. At the age of thirty-three, I

5 was in a car accident and my elbows became very

6 inflamed. The doctor said, oh, when you go home,

7 they will clear up. They didn't and I was

8 diagnosed with psoriasis.

9 Over the years, I have had as much as 75

10 percent. Now I am down to 1 percent, which is a

11 joy. Living with the consequences of the lesions

12 is difficult, both emotionally and practically.

13 People staring at me, moving on buses and in

14 movies, in plays, so that they don't have to

15 possibly touch or come in contact.

16 Not swimming with my children in the local

17 pool. I have never been told exactly that I can't

18 go in, but you know they would rather I didn't. In

19 the neighborhood, the children would ask my kids,

20 "What is the matter with your mother? Has she been

21 burned," or "Is she contagious?" and then maybe not

22 coming to the house to play. Or, at school

23 functions, they would ask me to volunteer. With

24 the kids I knew once they would get a look at the

25 legs or the arms that they would shy away, so I

243

1 didn't do it. I stayed home.

2 My husband also had to live through this.

3 He lived through the bleeding, the itching at

4 night. When I was near tears, he would comfort me.

5 I wished, lots of times, that it would just go

6 away.

7 Only wearing the long sleeves, summer and

8 winter, not only for yourself the embarrassment,

9 but the people around you would become very aware

10 of how they felt and you didn't want them to feel

11 uneasy. So, lots of times, you would stay home.

12 You wouldn't go where you wanted to or with your

13 children.

14 The bedsheets and the clothing would

15 always be stained either with the blood or with tar

16 treatments that you were on. The skin would become

17 very, very tight and then crack and bleed and it

18 made sleeping almost an impossibility. The

19 scarring that you will live with the rest of your

20 life.

21 Seeking medical help often was a

22 nightmare. You would go from doctor to doctor

23 getting tar treatments, different ones maybe, but

24 the results were always the same. They didn't

25 help.

244

1 I gave up going to the physicians because

2 I was discouraged and just medicated myself with

3 what I had learned through the years. Then, one

4 day, I read a little article and it said that there

5 was going to be a study and it had very little side

6 effects. I jumped to the phone. I couldn't wait.

7 That is when I read about Amevive. I was so

8 excited that it had been tested in Europe with

9 success and that it had supposedly very little side

10 effect.

11 The drug Amevive, in the study that I was

12 on, was an incredible experience for me. The side

13 effects are minimal, just a little nausea after my

14 shot and usually I go home and rest and I am just

15 good as new. For the first time in all these

16 years, I feel whole. There are days when I get up

17 and I have forgotten that I have had psoriasis and

18 the memories of the anguish and the embarrassment.

19 I would seek out Amevive in a second, even

20 though it hasn't been approved. I was that

21 thrilled. That is why I am so honored today to

22 have been asked to talk about it. I just want to

23 shout it from the rooftops. Everyone I know with

24 psoriasis I have tried to tell them about it, that

25 there is hope, don't give up.

245

1 Even though I am considered to have mild

2 psoriasis, the hurt and the mental anguish has been

3 no less difficult than someone with severe

4 psoriasis. It is my hope that the committee would

5 approve Amevive very quickly.

6 Thank you for the honor of being here

7 today.

8 [Applause.]

9 DR. DRAKE: Thank you very much, Ms.

10 Crawford. We really appreciate you coming.

11 Mr. Morton, welcome.

12 MR. MORTON: Thanks for having me. I am

13 almost in tears. I have only had this disease for

14 about three years so I am really an infant in the

15 world of I guess wisdom, I should say. I really

16 don't know where to start. I had something all

17 written down so I guess I am just going to read it

18 for you guys.

19 Imagine slightly bumping your elbow on a

20 cupboard or a door and needing a band aid. Imagine

21 combing your hair and ripping out the chunk of your

22 scalp on accident. Imagine wanting to get a

23 haircut but being too embarrassed to go to the

24 barber. Let me ask you a question. Have you ever

25 been in an accident where you have broken a limb or

246

1 maybe had a bandage and had people ask you, "What

2 happened?" and, after while, maybe it gets a little

3 bit annoying. If you have had psoriasis, you have

4 experienced it and it is annoying.

5 I want to ask you also to picture yourself

6 as a young man or woman, mid-twenties, maybe early

7 twenties, and you have grown up so far normally,

8 maybe played sports, had girlfriends, had

9 boyfriends depending on your gender, I guess. Keep

10 in mind, that you are in your prime, the time when

11 you are supposed to be having fun and possibly

12 finding your soul mate.

13 You wake up with this lesion on you. It

14 is small at first and the next day, it is a little

15 bit bigger. Then, over time, maybe it multiples.

16 So you go to the doctor and he tells you try this

17 and that and writes you a few prescriptions and you

18 leave his office feeling absolutely no resolution.

19 A month or two goes by and you have been

20 using the treatments, topical probably. They are

21 not helping you. You go clothes shopping now no

22 longer for what it is in style or what looks good

23 on you but what will cover your hideous lesions.

24 Let's say once you were a happy person,

25 maybe even good-looking. The good-looking person

247

1 you once were had degraded. You once played in the

2 sun and now you just stay inside. Everything you

3 once took for granted, like taking a shower or a

4 walk or playing basketball with friends or maybe

5 even asking out a pretty girl all seems awkward and

6 uncomfortable.

7 Let's say you had good self-esteem which

8 you thought was unbreakable. It wasn't.

9 Unfortunately, that was me. I was on an

10 experimental drug which had no noticeable side

11 effects to me. It helped me be again the person I

12 once was and, from my understanding, I have been on

13 it for the last two years, it is not an absolute

14 cure. However, it is a step in the right

15 direction.

16 It is a little different from most or all

17 treatments. Like I said, I haven't been as

18 experienced as Ms. Lewis over there. But if you

19 live the way I have for the last few years, believe

20 me when I tell you that you would this drug also.

21 Thank you.

22 [Applause.]

23 DR. DRAKE: Thank you very much, Mr.

24 Morton. We really appreciate your sharing with us.

25 Ms. Zimmerman?

248

1 MS. ZIMMERMAN: Excuse me, Dr. Drake. I

2 just needed to clarify that our expenses for this

3 trip out here, the patients and myself and the

4 staff, were paid for by the Foundation.

5 DR. DRAKE: Thank you very much.

6 Dr. Menter? Welcome, Dr. Menter.

7 DR. MENTER: Dr. Drake, thank you. I

8 appreciate the opportunity to come to speak to you

9 today in this public forum portion. Basically, I

10 would like to address three points. Number one,

11 who am I. Number two, why am I here. And, number

12 three, why do I believe new therapy is needed for

13 the treatment of psoriasis.

14 From a personal point of view, why am I

15 here? I have, just from a conflict of interest

16 point of view--just as Gail said, I have paid my

17 own way here. I am a consultant for Biogen. I

18 have participated in clinical-research studies both

19 for Amevive as well as for almost all the

20 "biologic" drugs that are currently under

21 development.

22 Basically, I have two brothers with

23 psoriasis. I have lived with them for twenty-five

24 years. They all live with us in Dallas. I have

25 tended to their psoriasis and just like we have

249

1 very eloquently heard, I have gone through the

2 struggles that they have had dealing with

3 psoriasis.

4 I also had the fortunate experience of

5 chairing the National Gene Bank for Psoriasis these

6 last ten years under the auspices of the National

7 Psoriasis Foundation and was able to travel around

8 the country looking at families with psoriasis,

9 large families with psoriasis, fortunately one of

10 which was able to produce a gene for psoriasis for

11 our gene bank.

12 I was amazed, just like you have heard

13 today, how often fathers, grandfathers, kids,

14 cousins, nephews when we got these families

15 together, never knew that their loved ones has

16 psoriasis. It is a hidden disease. You can just

17 have to read John Updike's personal experiences in

18 his book on how a psoriasis patient has to suffer.

19 Basically, it is a hidden disease and I

20 think the time has come, just as we have heard

21 today, for this psoriasis disease to come out and

22 for people to recognize that this is as disease on

23 a par with other chronic inflammatory disease,

24 asthma, diabetes, arthritis, Crohn's disease,

25 diseases of the autoimmune system, of the immune

250

1 system, that have a similar long-term chronic

2 course.

3 So that is why I am here today. I also

4 treat a number of psoriasis patients and have done

5 for the last twenty-seven years in Dallas. We have

6 a large psoriasis treatment center, just like you

7 heard from Diane, similar to what Dr. Ku has in San

8 Francisco. Currently, we have, at last count last

9 week, 565 patients taking systemic therapy for

10 psoriasis, the three main therapies you have all

11 heard about earlier this morning.

12 So why am I here? What is the reason for

13 me to come here and try to have ten minutes of time

14 to speak to you about psoriasis. You have heard

15 the quality-of-life issues from the patients. You

16 have heard the presentations this morning about the

17 drug, the efficacy, the safety data.

18 Basically, I believe there is a

19 significant reason to have new drugs for psoriasis

20 for one main reason. We have good drugs currently.

21 The three systemic drugs currently, methotrexate,

22 cyclosporine, Soriatane and PUVA, the light

23 treatment, give us good results in I would say 60

24 to 70 percent of patients.

25 On the other hand, and I think this is

251

1 critical, we cannot look at psoriasis any more as

2 short-term-treatment disease. Patients currently

3 with all the treatments that we have, systemic

4 treatments, relapse within six to eight weeks when

5 getting off the drug.

6 We cannot keep patients long-term on some

7 of these drugs because of the side effects you have

8 heard about. So, from a quality-of-life point of

9 view, it is critical that we look for drugs that

10 will improve quality of life by improving

11 remissions, either on treatment if it is safe or

12 off treatment for longer periods than six to eight

13 weeks.

14 A psoriatic hates one thing. They had

15 being cleared and then allowed to relapse six to

16 eight weeks later. They will tell you this. We

17 need to look at psoriasis as a long-term, chronic

18 inflammatory disease that needs long-term control

19 like a diabetic takes an insulin shot every day,

20 when an arthritis patient has to stay on long-term

21 treatment. We need to find drugs that will allow

22 us to maintain a stable course for these psoriatic

23 patients out there.

24 From a perspective point of view, I have

25 lived through Soriatane coming to the market. I

252

1 use Soriatane. I have lived through methotrexate.

2 With methotrexate, we have a 30-year track record.

3 I think Mark Lebwohl may have mentioned that three

4 patients underwent liver transplantation for

5 methotrexate. These are patients at our

6 institution who have been overdosed with

7 methotrexate.

8 We have a huge big transplant population

9 at our institution in Dallas. Three out of the

10 first 200 patients transplanted were psoriasis

11 patients who had had too much methotrexate. So we

12 cannot treat with cyclosporine for longer than a

13 year, with PUVA for periods of time without skin-cancer

14 risk.

15 So why, to answer my third question, do I

16 believe we need a new treatment for psoriasis? I

17 have polled, out of the 500 patients we have plus,

18 between the three of us, and we do psoriasis

19 treatments on a daily basis and psoriasis clinics

20 on a daily basis, I have polled our patients, would

21 you prefer a weekly injection, a monthly injection,

22 recognizing there are other drugs coming down the

23 pipeline that may have different manners of

24 administration. This has been done. The British

25 have published a publication showing, as well, that

253

1 the vast majority of patients would prefer a weekly

2 or a monthly injection if this will keep them clear

3 for longer periods of time than is currently

4 available except for PUVA which does keep people

5 clear for longer periods of time.

6 The vast majority of patients will tell

7 you, give me a weekly injection. If it is safe,

8 and I recognize this is a major problem with a drug

9 that is new--not a major problem, but something

10 that we all have to consider--but having started

11 with cyclosporine in the 1980s where we didn't know

12 much about it, methotrexate in the '70's that we

13 didn't know much about, recognizing that those

14 drugs took a long time to be approved, they have

15 helped our patients but we need more.

16 We need more medicines available for our

17 patients currently today. Half the patients drop

18 out of treatment because of concerns about side

19 effects and almost a third of our dermatologists in

20 the country will not utilize systemic treatments

21 currently.

22 Therefore, in the last two minutes, why do

23 I believe we need a new treatment for psoriasis? I

24 have talked about the current drugs we have

25 available. They will continue to be utilized.

254

1 Dermatologists do a wonderful job in mixing and

2 matching medications probably as well as any other

3 specialty. I believe should this panel decide to

4 approve alefacept that dermatologists will find the

5 most expedient way to utilize this drug with safety

6 criteria that dermatologists being fairly

7 conservative people in the majority will recognize

8 and understand.

9 Drug holidays off treatment is important

10 to minimize side effects. I think I have already

11 mentioned that the three drugs we currently have

12 available we cannot get patients off these drugs

13 for longer than six to eight weeks without them

14 failing and sometimes failing fairly substantially.

15 So that is drugs with the safety profile

16 that we understand, affording long-term remissions,

17 are very critical. Too many patients have

18 withdrawn from treatment, as I have said. I do

19 believe that the problems that you have heard about

20 so eloquently from the patients and the NPF are

21 real and afford us the opportunity to take 6

22 million lives in the United States, improve the

23 quality of their lives and improve the treatment

24 that we currently have available.

25 I would urge the panel to take into

255

1 consideration all that has been said and consider

2 not only safety profiles, not only improvement, but

3 the tremendous need in the marketplace for patients

4 to have better treatment.

5 The final point I would like to make is

6 that psoriasis, as you have heard today, is a

7 disease of young people. The vast majority of

8 patients with psoriasis present before the age of

9 35 when body image is important. They are

10 developing their body image. Those of us who are

11 older recognize that our paunches are getting a

12 little bit bigger and our hair is getting thin, but

13 the bottom line is when a person is fifteen,

14 twenty, twenty-five and their body image has not

15 yet been established, looking at themselves in the

16 mirror every day and recognizing their psoriasis is

17 an important factor in their own self esteem.

18 Females have equal representation with

19 psoriasis. Currently, a twenty-five to thirty-year-old

20 female or a thirty-five-year-old female

21 contemplating pregnancy cannot take any of the

22 drugs we currently have available. So we need to

23 have drugs available that have a safety profile

24 that we can understand, we can follow, we can

25 watch, we can be conservative and we can improve

256

1 the quality of life for our patient population.

2 Thank you, Dr. Drake.

3 DR. DRAKE: Dr. Menter, thank you for a

4 very passionate and well-thought-out presentation.

5 We appreciate your taking time to come.

6 I also have to tell you that I want to

7 also thank Ms. Lewis for helping me make my

8 announcement about the cell phones because I forgot

9 again. So you helped me. So thank you very much.

10 There is more than one way to skin a fish, isn't

11 there. Thank you so much.

12 We do appreciate so much, Gail, you and

13 all your representatives coming. It takes time out

14 of people's days and lives but it is important for

15 people to put these things in perspective. The

16 committee has to weigh efficacy and safety, which I

17 think is our foremost issue, it is important to

18 hear from patients so we know why we are all here.

19 So thank you again.

20 Committee Discussion and Vote

21 DR. DRAKE: Now, here we go, group. We

22 are down to the real serious nitty gritty now. We

23 are now into just the committee deliberations.

24 The sponsor will be asked not to comment

25 unless called upon during this time period because--it is as

257

1 much a time issue as anything, but this

2 really is the committee's time to think about

3 things and discuss it.

4 As you can see, we have a lot of

5 questions. I have tried to have some time lines

6 that are rational about most of this. I would like

7 the committee to think about how much we have to

8 cover and keep your comments as abbreviated as

9 possible and pertinent. Maybe we can get through

10 this agenda.

11 I may change the order. I am going to

12 change the order just a little bit. I am going to

13 take the Chairman's prerogative. We are going to

14 take Roman numeral I first followed by IV because I

15 do not want us to miss the crux of the issue with

16 people, perhaps, having to leave or running out of

17 time. Frankly, each one of these questions could

18 take a day in and of themselves. They are

19 wonderful questions and they are wonderful

20 thoughtful propositions. So there was some real

21 thought that went into it.

22 Roman numeral I, I am not going to read

23 the whole thing but I would just like to highlight.

24 Let's start with Part A. It is about lymphocyte

25 reduction and risk of infection. Just to make a

258

1 few quick summary points, in Study 711,

2 approximately half the participants experienced at

3 least a single occurrence of the CD4 cell count

4 below the lower limit of normal at any time during

5 a treatment.

6 That was kind of a point. Then the next

7 point the has been made is that the total

8 experience of patients receiving more than two

9 cycles is limited. The third point--these are

10 safety concerns. You understand this doesn't rule

11 anything in or out. With every drug we have these

12 issues and so it is just kind of important to

13 highlight them and see if we think the risk-benefit

14 ratio is where it ought to be.

15 Third is a central issue, interestingly

16 enough. It is where the lymphocyte reductions

17 result in clinical sequelae. Serious infections

18 were reported in about 0.2 percent of placebo and

19 0.9 percent of active drug in the treated patients.

20 There didn't seem to be an apparent relationship

21 between lymphopenia and infections and there were

22 no opportunistic infections observed, which I think

23 is important.

24 Then I think, in the fourth paragraph, one

25 of the points I want to make is that normal

259

1 lymphocyte and CD4 cell counts were required before

2 the first treatment cycle and normal CD4 cell

3 counts were required for subsequent cycles. These

4 are kind of the major points upon which the agency

5 based their questions to us.

6 Have I given that an accurate summary?

7 Dr. Weiss, do you have anything to add to that?

8 DR. WEISS: No; that is fine. Thank you.

9 DR. DRAKE: Okay, good. Depending how

10 much the committee wants to get into, I think the

11 first thing--the only one of all these questions,

12 of all these Roman numerals, that we need to vote

13 on today, so you will know that, too, is No. IV.

14 Roman number IV is where we will have a vote.

15 Otherwise, these are questions, discussions and I

16 may ask for a sense of the committee, just a sense

17 of what you are thinking, to give the agency some

18 direction of how the committee is thinking, but

19 they are not votes.

20 So has the sponsor generated sufficient

21 data premarketing to characterize treatment-related

22 effects on lymphocyte reductions? What say you?

23 Listen to me. I have been listening to O'Reilly

24 too much using his same quote.

25 Dr. Raimer?

260

1 DR. RAIMER: I think we do need to follow

2 patients if the drug gets approved to watch whether

3 we have a registry or exactly how it is done, I

4 think the numbers of infections need to be

5 monitored.

6 But I am very encouraged by the fact that

7 we don't see opportunistic infections. These were

8 over a fairly large number of months so I think if

9 it were really going to be a very significant

10 problem that probably would have shown up in the

11 studies that have been done so I feel reasonably

12 comfortable and not totally comfortable. I think

13 it is definitely going to need to be monitored

14 because it definitely is a potential problem. But

15 I feel reasonable comfortable at this point in

16 time.

17 DR. DRAKE: Dr. Swerlick?

18 DR. SWERLICK: I have a question regarding

19 what level of safety we are talking about. We are

20 able to identify, or potentially identify,

21 significant infections in a patient population,

22 about 1,300 patients extending over a few years.

23 If we are looking for adverse events that are going

24 to occur 1 in 10,000 or 1 in 100,000 or more, how

25 many patients are we going to have to follow for

261

1 how long? Perhaps the people from the FDA can

2 address that issue.

3 DR. SEIGEL: Following patients for rare

4 events that have a significant background you could

5 follow forever and not determine if you don't have

6 a controlled population. If you are talking about

7 rare events that are very uncommon in the

8 population, certain specific types of tumors, liver

9 failure or whatever, those will stand out in a

10 postmarketing.

11 If you are talking about an increase in

12 the incidence such as these data might suggest of

13 something like cellulitis. That is certainly going

14 to happen to patients without the treatment, I

15 think the answer is, especially given that these

16 patients will be on and off this therapy and

17 several other therapies, that you will not know,

18 outside of controlled studies.

19 DR. DRAKE: Bob, you just hit on the crux

20 of the question, how do we know when safety is

21 enough safety. I don't think this committee ever

22 knows. Sometimes, you just have to keep tracking

23 and see what happens. But I think the important

24 thing is we don't turn something loose that we

25 think might cause imminent harm would be the way I

262

1 would approach it.

2 DR. SWERLICK: I would like to know the

3 standards so we don't set the standard in such a

4 way that it could never be approved.

5 DR. DRAKE: I see.

6 DR. SWERLICK: If we set a standard that

7 is so difficult--and I am trying to get a feel for

8 where the standard is.

9 DR. SEIGEL: The laws and regulations

10 speak to safe and effective and for biologics say

11 pure and potent. I can tell you that the long

12 tradition with the FDA and its advisory committees

13 is that safety is certainly considered in the

14 context of benefits. Many of the drugs that are

15 used to treat cancer wouldn't be considered safe if

16 used to treat a common cold or a simple headache.

17 So it is a judgmental risk-benefit but

18 there is not a lot of formal guidance I can give as

19 to what a standard is in that regard.

20 DR. DRAKE: Dr. Katz?

21 DR. KATZ: To go along with what Bob just

22 said, isn't it difficult for us to discuss this in

23 an isolated manner without integrating it with

24 efficacy. I know, Lynn, that we have to discuss

25 one thing at a time, but you are probably willing

263

1 to have certain risk if you are clearing up 90

2 percent of people. If you are clearing up 15

3 percent of people, maybe you are willing to accept

4 lesser risk even in a disorder such as this.

5 As Bob said, we need a little more

6 guidance before we make an agreement whether this

7 is acceptable or not, an acceptable risk for this

8 condition.

9 DR. SEIGEL: Excuse me, and let me clear

10 up and in answer to Dr. Swerlick's question because

11 I wasn't sure if you were asking what is the

12 standard for how safe is safe enough, or how much

13 data is data enough.

14 DR. SWERLICK: Both.

15 DR. SEIGEL: Because I answered the first

16 one, but there is a guidance for how much data and

17 it was alluded to in the sponsor's presentation.

18 It is one developed in the international

19 harmonization process which speaks about drugs for

20 chronic disease and suggests that there should be--the

21 numbers that come to my mind are in the 1,000

22 to 1,5000 range of exposures, 300 to 600 at least

23 for six months of therapy, 100 for a year of

24 therapy.

25 But that guidance is also full of provisos

264

1 where certain signals arise. Where there are

2 concerns about serious rare events, you may need

3 more or whatever. So it is to be taken in the

4 context of the science. But that is the guidance

5 given to provide an approach to identifying rare

6 events that may occur in chronic therapy that are

7 not anticipated.

8 There has been some discussion since those

9 went into effect some probably seven or eight years

10 ago, and given some the concerns about adverse

11 events being discovered with drugs after their

12 approval as to whether those guidances are

13 adequate. For many drugs, we have larger numbers

14 than that.

15 DR. SWERLICK: Basically, the first

16 question points to use of surrogate markers to try

17 to predict whether or not something untoward will

18 happen in the low-frequency event. The difficulty

19 with that is that we really don't know--even if we

20 see drops in lymphocyte counts, how do we interpret

21 all that?

22 I guess the crux of my question is that it

23 is not really if something untoward will ultimately

24 happen in one patient who is receiving this drug.

25 If you give it to enough people, something bad is

265

1 going to happen whether it is related or unrelated.

2 Ultimately, what is the frequency that we will find

3 acceptable? Will that be 1 in 10,000, 1 in 100, 1

4 in 1,000? That is where I am uncomfortable

5 because, ultimately, that is where we are called

6 upon. And I don't know what the standard is.

7 DR. SEIGEL: Right. That is why I was

8 answering the first part. That is determined in

9 the context of anticipated benefits. There isn't a

10 standard. What is acceptable in one disease and

11 for a highly effective drug versus a less effective

12 drug or for a more serious versus a less serious

13 disease is going to vary and it is usually a matter

14 of common--by saying it is common sense, I don't

15 mean to say it is easy. It is not easy, but it is

16 not a hard number.

17 DR. DRAKE: Dr. Morison.

18 DR. MORISON: I think one of the issues is

19 how are you going to follow the patients, not just

20 how many patients have you got but how are you

21 going to follow them. The example immediately

22 comes to mind is the multicenter study on PUVA

23 therapy here in the United States. They followed

24 1,500 patients and, after about ten years, had

25 about a 98 percent follow-up rate on those 1,500

266

1 patients and found an increased risk of squamous-cell

2 carcinoma within two and a half years of the

3 approval of the treatment whereas, by comparison,

4 the European study has 3,500 patients that, after

5 about five years, was only following 1,500 of those

6 patients and it took ten years to find an increased

7 risk of squamous-cell carcinoma.

8 So, when you are talking about a registry

9 or following patients, I think it has to be clearly

10 defined what you mean by following patients. Are

11 you taking a population of patients and making sure

12 someone is keeping tabs on those patients and

13 looking at them at regular intervals because,

14 otherwise, you could have a lot of ex-PUVA patients

15 or UVB patients or sun patients out there with

16 squamous-cell carcinoma and you won't detect them

17 unless someone is very carefully following those

18 patients.

19 So the use of the word "registry," I think

20 should be defined rather than just drug registry.

21 DR. DRAKE: Let's discuss both parts of

22 the questions then, of the first question and the

23 second question, since we have kind of wandered

24 into that.

25 Dr. Epps?

267

1 DR. EPPS: I guess I would like see more

2 data although two cycles is more than one, I don't

3 necessarily think it is multiple. Certainly, if,

4 according to the testimony of people who have

5 experienced this medication, if they really like it

6 and they think it helps them, then certainly more

7 cycles could be performed for longer studies and

8 more data.

9 I think it would also be important to

10 interview the people who dropped out, find out why

11 they dropped out, who didn't have side effects,

12 necessarily. Is it because they couldn't wait? Is

13 it because they had an untoward effect or whatever.

14 But I think that is important to know, too,

15 collecting the pro and the con for any medication

16 because, although we hear the testimony of people

17 who benefit from it and, of course, we all want

18 medications for psoriasis and more options.

19 I am in a pediatric group and my options

20 are much more limited. I hear the stories of

21 people won't hold their hand and won't play with

22 them. So I am very aware of the other side

23 effects, but I am also very aware of the long-term

24 safety effects and we will get to the pediatric

25 questions later, but I think if there are adults

268

1 who are willing to move forward and have multiple

2 cycles, I think it would be important to collect

3 that data.

4 DR. DRAKE: I think we can mix some of the

5 kiddie stuff in with this right now. Everybody

6 commented to me about kiddie stuff during break, so

7 make your comments, if you will, just kind of right

8 along with that. If we look at children right now,

9 what do you think about this? Should pediatric

10 patients be included in this now? That is one of

11 the agency's questions.

12 Do we need specific studies in pediatric

13 patients? You are a pediatrician.

14 DR. WEISS: Just let me clarify, too, that

15 I think what is on the table, a question that we

16 will hopefully get to, is Roman numeral IV, an

17 indication for use in adults. The question, then,

18 would be for pediatrics because the sponsor is not

19 actually asking right now.

20 DR. DRAKE: I know that.

21 DR. WEISS: The question would be if and

22 when to study children.

23 DR. DRAKE: I have a suggestion, then. In

24 the interest of time and streamlining the process,

25 this is an important clarification. The sponsor is

269

1 not asking for children. The children is sort of a

2 second phase in the process. Let's focus our

3 discussion now on adults and get through the

4 primary adult stuff because this is not a request

5 by the sponsor to do children.

6 So we could put that off and address that

7 later, time permitting. Is that fair enough, Dr.

8 Weiss?

9 DR. WEISS: That is correct.

10 DR. DRAKE: Good. We solved that. Boy,

11 you saved me some time there. Good job, Dr. Weiss.

12 I want to ask a question. I want a sense

13 of the committee. That second part, given that the

14 sponsor is proposing the product be indicated for

15 multiple cycles, please comment on the adequacy of

16 the data to support multiple-cycle use. We have

17 had data on two cycles.

18 I want a sense of the committee. This is

19 not a vote. This is just a sense. Do you think

20 that this data is sufficient at this time for us to

21 go ahead and think about--do we need more data--I'm

22 with you a little bit. The efficacy almost comes

23 before the safety but do you think--let's for the

24 moment assume that the efficacy was okay and we are

25 thinking about recommending approval of this.

270

1 Do you think that we have enough data in

2 terms of cycles or should, perhaps, the number of

3 cycles given be limited initially until further

4 data is collected? What is your sense of the

5 committee? Dr. Abel, do you have a comment on

6 that?

7 DR. ABEL: My sense is that there should

8 be some limitation. If, indeed, the responses last

9 up to nine months, then, hopefully, the responders

10 are going to be the ones that will be treated. But

11 the ones who don't show response won't have

12 multiple cycles to try to push them to be

13 responders and maybe increase the possibility of

14 toxicity side effects.

15 There are some who aren't responders. I

16 have to maybe get a better feel for the percentage

17 but there are excellent responders, there are

18 moderate responders and there are some that clearly

19 may be nonresponders. But I would not like to see

20 those nonresponders being pushed with multiple

21 cycles to try to get them to be responders and just

22 treat them every twelve weeks, I mean after only a

23 twelve-week interim.

24 DR. DRAKE: Dr. Seigel and then Dr. Tan.

25 Dr. Tan, did you have a comment on the--

271

1 DR. TAN: Right on this.

2 DR. DRAKE: Okay; excuse me, Dr. Seigel,

3 he kind of had his hand up first.

4 DR. SEIGEL: That's fine.

5 DR. TAN: I think we discussed in the

6 morning that we don't have--there really isn't

7 enough data to differentiate the benefit of the

8 second course is due to the carryover effect of the

9 first course. So there wasn't enough data as we

10 discussed in the morning, I think.

11 DR. DRAKE: Dr. Seigel?

12 DR. SEIGEL: I just wanted to make sure

13 that the committee understood, as they discussed

14 this and particularly since you asked the sponsor

15 and they have been very compliant--they are

16 remaining quiet--to note that there is two-cycle

17 data in the controlled clinical trial. There is a

18 limited number of experience with patients on

19 third, fourth and fifth, I think 150-some odd on

20 third and another 120 who have had four or five

21 cycles.

22 They are subselect groups. They are not

23 studied on the same controlled protocol but there

24 is some experience available with additional

25 cycles. Then probably in comparing, like,

272

1 lymphopenia issues, if you look at the 80 people

2 who had four cycles or the forty-some odd who had

3 five cycles, they are a subgroup, people who might

4 have had certain types of either durable responses

5 or unfavorable responses in early cycles aren't

6 getting later cycles. It is a little hard to

7 understand, but there is, indeed, some data

8 available on longer cycles.

9 We are not comfortable, I think, with the

10 amount.

11 DR. DRAKE: You are not comfortable with

12 the amount? Okay. So the agency has got a level

13 of discomfort. Solves that.

14 Any comments on how to discuss the optimal

15 ways to generate additional data on infectious

16 risks? Lloyd. It is 2 under A under Roman numeral

17 I, please discuss optimal ways to generate

18 additional data on infectious risks.

19 DR. KING: I had suggested one of the

20 surrogate markers would be the C-reactive protein.

21 There is a whole body of information, such diverse

22 things as atherosclerosis, et cetera. The best

23 predictor is not the lipid profile but the C-reactive

24 protein as studied in the Framingham study

25 of nurses.

273

1 So it seems to me that, if you are going

2 to have cells, the question is whether they are

3 potent or not; that is, the product being released

4 could be an acute-phase reactant. So it seems to

5 me that one of the populations that keeps coming

6 up, diabetes, atherosclerosis, psoriasis and so

7 forth, I, for one, believe that psoriatics are much

8 higher risk as a subpopulation for atherosclerosis

9 and heart disease than one would imagine.

10 Part of that may be the C-reactive

11 proteins. So I would suggest that it is oftentimes

12 difficult to culture things. We all have a lot of

13 things--you can't culture strep from cellulitis.

14 It is like 10 percent. So I would suggest

15 measuring C-reactive protein and other parameters

16 would tell you whether or not the up or down pool

17 of T-cells did or did not produce the biological

18 assassins.

19 DR. DRAKE: Bob and then Dick.

20 DR. SWERLICK: I would just inject a word

21 of caution again using surrogate markers. The

22 difficulty is that, unless you study that within a

23 population of psoriatics who have not been treated

24 with this drug, you don't know how to interpret it

25 because the gold standard becomes whether you can

274

1 actually diagnose an infection or not.

2 Therefore, in order to generate sufficient

3 data to know whether or not the drug sets people up

4 for increased numbers of infections, you just have

5 to follow a lot of people for a long period of time

6 and compare them to controls that were followed for

7 a long period of time. Otherwise, I am not sure

8 how to interpret the surrogate data.

9 DR. KING: They already have data on

10 psoriatic arthritis. So one of the ways,

11 potentially, to get into the issue of children and

12 psoriasis is look at C-reactive protein. They are

13 already doing biopsies. I am not sure they are

14 biopsying joints of children. So maybe our

15 rheumatology colleague could help us more this kind

16 of phenomenon, but I agree, you can't always

17 diagnose infection. But if you have psoriatic

18 arthritis and you are already getting response and

19 you are measuring C-reactive protein as your

20 surrogate marker, I am talking about that specific

21 population.

22 DR. DRAKE: Dick?

23 DR. TAYLOR: I may have some confusion

24 with regard to the registry. I am not sure what

25 that is going to include. But it appears to me

275

1 that if the registry was inclusive enough, it could

2 tell you about lymphocyte counts after four, five

3 or ten cycles and it could tell you about the

4 malignancies and it could tell you about some of

5 these things that we are concerned about and maybe

6 make it easier for us to worry about the efficacy

7 and not so much about the toxicity.

8 So maybe somebody could explain what is

9 going to be in the registry or maybe it could be

10 expanded to include some of these things. Who is

11 going to control the registry? Who is going to do

12 it? Is it on all patients?

13 DR. DRAKE: With all due respect, I would

14 like to ask Dr. Seigel have you guys thought about

15 a registry? Where is the FDA on this?

16 DR. SEIGEL: I think the company has

17 proposed one. Whether or not we would be

18 discussing with them whether a registry is the

19 right way to proceed will depend, in significant

20 part, on the determination as to whether to approve

21 the drug now. I think some of the issues can be

22 addressed well in a registry. Other issues are

23 better addressed with randomization and controls.

24 So, obviously, we are looking for some

25 guidance and to make some guidance and to make some

276

1 decisions as to where to move forward. So I don't

2 know that we have had substantial input yet as to

3 registry design. We have not.

4 DR. WEISS: Oftentimes, registry

5 discussion comes when we are talking about

6 approving a product and then these would oftentimes

7 required postmarketing commitments and we would

8 discuss in much more detail at the time of an

9 approval about the size of the registry and the

10 amounts of data to be collected and the types of

11 periodic follow up to the agency that would be

12 coming in.

13 There are lots of details. There is a lot

14 that can be done right now. There hasn't been much

15 discussion in that regard.

16 DR. DRAKE: So we are not quite there yet.

17 Since you stood up, and I don't, by any means mean

18 to be rude, would country give us a quick sentence

19 from the sponsor? But I really want to keep this

20 committee-focused right now.

21 DR. VAISHNAW: The first half of the

22 sentence is that there are over 800 patients in

23 safety-extension studies and the current snapshot

24 of the database reveals several hundreds in the

25 fourth and fifth course is different from the

277

1 different you are reviewing right now. The safety

2 profile remains the same. If that is helpful to

3 that panel to know that.

4 Secondly, the registry study, we are in

5 active dialogue with experts and we feel there are

6 a number of good ways to move forward and

7 definitively answer the question is the risk of

8 something like squamous-cell carcinoma elevated

9 and, as a sentinel event, our hypothesis would be

10 that a discrete elevation in the rate of that would

11 be telling in terms of potential for other types of

12 risks, and this is a tractable problem.

13 DR. DRAKE: Thank you very much.

14 Seth?

15 DR. STEVENS: I would just like to

16 comment, with all due respect to Dr. King, about

17 the use of surrogates. I would agree that the way

18 to follow infection is clinically to look for

19 infection. I think that we associate things based

20 on our clinical experience in the past. I think an

21 example of that this morning was, for example,

22 chills which we normally associate with infection.

23 There were chills. There wasn't strong

24 evidence for infection. I think when using

25 biological-response modifiers and things like that,

278

1 some of our old associations don't carry over. I

2 think when the thing that you really are interested

3 in is something that we are trained to do, that

4 doesn't involve expense or risky tests, I think

5 that is the best way to monitor for those events.

6 DR. DRAKE: Other comments on this first

7 question, on this first section, on the safety, the

8 lymphocyte reduction. Lloyd?

9 DR. KING: I am still concerned about this

10 line that says who is going to follow up and

11 monitor the lymphocytes if you turn it loose? It

12 has been my experience there is a whole lot of off-label use

13 and, once you open the door, it is the

14 Harvard law that, under defined conditions, the

15 organism will do as it dadgum well pleases.

16 The idea of the registry actually is

17 intriguing to me because, having been involved in

18 the fuss about Accutane back and forth, it seems to

19 me that the study will get the results you plan for

20 but it is the unexpected things that, if you turn

21 it loose, people are going to be so--as you heard,

22 "I want something, even if it is going to be

23 dangerous for me."

24 Then, after the fact, after you have taken

25 three courses of, say, arsenic for asthma you find

279

1 out fifteen years later it causes cancer. So I

2 think the idea of registry really has to be

3 hammered out and actually who is going to follow

4 these people because if you just turn it loose and

5 say all you have got to do is take a skin injection

6 once a week, I can imagine that there will be whole

7 lots of nondermatologists and other people doing

8 this because it happened to me with Accutane. So I

9 am concerned about the registry.

10 DR. DRAKE: Dr. Weiss and Dr. Seigel, what

11 I am hearing, to kind of summarize what I have

12 heard, is that the sense of the panel is that there

13 probably needs to be a registry or some semblance

14 of a registry, perhaps some follow-up studies,

15 either before or after, preapproval or

16 postapproval, but clearly some follow-up studies.

17 Probably two cycles is very limited

18 information upon which to base long-term

19 conclusions. So, as you get into multiple cycles,

20 I think you are clearly going to need more

21 information about what happens to lymphocytes, what

22 happens to infections, what happens to the whole

23 malignancy notion.

24 I think there are all kinds of things that

25 would need to be followed out either before or

280

1 after approval. Is that a fair assessment from the

2 committee's perspective? Lloyd?

3 DR. KING: Yes.

4 DR. DRAKE: Does anybody have additions or

5 corrections to what I have just said? Dr. Weiss

6 and Dr. Seigel, is that adequate for you guys? Do

7 you need more information before I move on to the

8 next one?

9 DR. WEISS: I think that is adequate.

10 Thank you.

11 DR. DRAKE: Okay. You notice I didn't say

12 is that exceptional because I don't think we have

13 given you any exceptional help there. But I think

14 we are a little baffled ourselves exactly how to

15 proceed. So at least we can try to help you.

16 Let's talk about B, the changes in antigen

17 response. In Study 708, the number of DTH shifts

18 from plus to minus was higher in the treatment

19 group compared to placebo. So let's look at the

20 questions. Should all individuals be evaluated for

21 latent t.b. infection with a tuberculin skin test

22 prior to therapy? If latent infection is

23 uncovered, discuss how such individuals should be

24 managed with respect to use of this drug.

25 Comments on that question? Bob?

281

1 DR. SWERLICK: I don't think it should be

2 any different than using any other

3 immunosuppressive. Essentially, if you put

4 somebody on prednisone or you put somebody on

5 cyclosporine or Immuran, you are going to end up

6 managing it the same way. So at least they have to

7 be held to the same standard.

8 DR. DRAKE: I think that is a very simple

9 answer to this question, just make it the same

10 standard as other immunosuppressives. Any

11 additions or comments to that?

12 DR. SWERLICK: The only other question

13 about the PPD, it may be meaningless because these

14 patients may have been put on other

15 immunosuppressives which may modify it. So I think

16 it has to be sort of determined, an algorithm

17 depending on whether or not they have been on

18 immunosuppressives before.

19 DR. DRAKE: Other comments on that

20 question? Should subject monitoring include

21 periodic assessment of DTH?

22 DR. SWERLICK: My comment on that it is

23 such a miserable test. I am not sure to interpret

24 it so it would be hard for me to require them to do

25 that.

282

1 DR. DRAKE: I saw almost everybody at the

2 table shaking their head no. So you got an answer

3 there. Number 3, should the sponsor perform

4 studies to evaluate the ability to respond to

5 immunization such as pneumococcal or influenza

6 vaccines? Lloyd?

7 DR. KING: If you are going to address the

8 pediatric population or older people where you do

9 that for--where they COPD, et cetera, I think the

10 answer would be yes. I think you really have to

11 talk about if you are going to vaccinate against

12 Asian flu which may knock people out.

13 The same reason you knocked out the age

14 population not getting this drug early on, I think

15 you have to say that a recommendation would be

16 high-risk populations, children and older people

17 with disabilities, the answer would be yes.

18 DR. DRAKE: Help me, Lloyd. Are you

19 saying we should not give it to these patients or

20 do it with due consideration?

21 DR. KING: No, no. I'm sorry. I'm saying

22 if you are going to give it to these populations,

23 addressing the issue of children, then you are

24 going to talk about is the immunization going to be

25 effective.

283

1 DR. DRAKE: Let's talk about adults

2 because we are not on kids yet.

3 DR. KING: Adults in high-risk

4 populations, I think it should be periodically

5 tested to see if they are going to respond to the

6 flu shots or whatever in the same way you want to

7 know if they are going to resist Asian flu or

8 whatever. I think you are going to have to have

9 populations you recommend testing.

10 DR. DRAKE: Other comments? Elizabeth?

11 DR. ABEL: I think this might apply to all

12 of the potential side effects, change in antigen

13 response, malignancies. We have talked about who

14 are candidates for this treatment but I think we

15 also have to think what population groups may not

16 be candidates or what population groups there might

17 have to be special cautions written up in the

18 package inserts. These might be not just children

19 but--well, we are not talking about children but

20 previous treatment in regards to, say, PUVA or

21 cyclosporine, geriatric patients, et cetera.

22 DR. DRAKE: I think what I am hearing, the

23 sense of the committee is saying one needs to use

24 reasonable and rational precautions in high-risk

25 populations.

284

1 DR. KING: Yes.

2 DR. DRAKE: Is that a fair assessment?

3 Dr. Weiss? I see that is not enough; right.

4 DR. WEISS: Now, that is helpful. When we

5 get beyond the letter questions, if there is a

6 recommendation for market approval from this

7 committee, we have several questions about what

8 populations it should be indicated and studies in

9 other populations.

10 But one of the questions, and we have had

11 experience with these kinds of studies in other

12 therapies such as anti-TNF strategies where the

13 question specifically is if you have an adult who

14 is being treated on a chronic basis, and they are

15 coming in for their yearly flu shot, is it

16 important to have a study, and these studies can be

17 done in a controlled fashion, to determine whether

18 or not these individuals actually can mount or have

19 a blunted response to the standard vaccinations

20 that they might be getting while they are on

21 treatment.

22 DR. DRAKE: Thoughts on that question?

23 DR. SWERLICK: I think it might be helpful

24 to interject any previous experience you have with

25 the anti-TNF biologics if those answers are

285

1 appropriate to questions that are being posed here.

2 In particular, actually, I was thinking about the

3 previous question about repeated courses. How has

4 this been handled before and what was the

5 justification for those criteria?

6 I think that is really useful information.

7 DR. SWERLICK: I think for both anti-IL2

8 receptors, anti-CD25 products and anti TNF-receptor

9 products, we have rather routinely had, I think

10 almost invariably had, postmarketing commitments to

11 study the impact of those on vaccination of

12 recipients. I am not sure I could generalize what

13 the results of those studies are. There is some

14 controversy in some cases.

15 DR. DRAKE: Seth?

16 DR. STEVENS: I think that some of my

17 hesitancy is when we talk about moving the use of

18 this drug to different populations and the task

19 before us today. So in terms of not an increased

20 risk of influenza in the patients that were treated

21 with this drug to date, those sorts of things give

22 me a certain perspective. Then when you start

23 saying, well, what about elderly people who should

24 be getting these vaccines that were not

25 specifically studied, that is where I start to lose

286

1 my solid footing.

2 So I guess I just have that as a comment,

3 not to sort of derail things but I think that I

4 have agreed essentially with what we just heard

5 from the FDA and from the other committee members.

6 DR. DRAKE: Dr. Taylor

7 DR. TAYLOR: Do a small study. Figure out

8 what is going on.

9 DR. DRAKE: You want to do a small study,

10 figure out what is going on. Premarketing?

11 Postmarketing? Or either?

12 DR. TAYLOR: Either.

13 DR. DRAKE: So that gives you some

14 flexibility. I have a question about lymphocytes.

15 Somehow, I still haven't got it about the potential

16 nonrecovery. It seemed like there was a small

17 percentage of patients who never recovered. This

18 is one time I am going to ask Dr. Seigel, perhaps

19 you can help. If not, then I am going to go to the

20 company because I am still confused about how

21 important an issue is that and what must we do

22 about this recovery, and is it important.

23 DR. SEIGEL: I will defer, actually, to

24 Dr. Marzella but, except to briefly summarize, as I

25 understand the data, a lot has to do with how you

287

1 define recovery. If you talk about recovery to the

2 lower limit of normal as opposed to recovery to

3 baseline as has been pointed out, that will differ.

4 Over a period of nine months, there is

5 not, in aggregate, a recovery to the pretreatment

6 levels, whether those depressions are clinically

7 significant and what level of recovery is

8 important. Lou, do you want to add to that?

9 DR. DRAKE: Maybe we are knocking out the

10 bad guys that need to be knocked out anyway and

11 hopefully they will recover with more normal

12 lymphocytes. How is that for doing a short cut?

13 DR. SEIGEL: Not bad.

14 DR. DRAKE: You know what I am trying to

15 say.

16 DR. MARZELLA: I guess you are either an

17 optimist or a pessimist or you want to see the data

18 before you make a decision.

19 DR. DRAKE: Thank you, Dr. Marzella. That

20 is just terrific. We have really clarified this

21 issue.

22 DR. MARZELLA: I think that, obviously, it

23 is a profound biologic change. To be honest, the

24 clinical significance is not known, but that

25 doesn't mean that we don't need to follow these

288

1 patients and document when, in fact, a recovery

2 occurs.

3 There is similar experience in other

4 indications. For instance, we have seen other

5 products that cause lysis of T-lymphocytes that

6 cause profound depressions. It takes sometimes

7 years for these counts to recover. We still don't

8 have the full picture of what it means but I don't

9 think we can afford to ignore it. I think that we

10 need to understand what happens.

11 There is a suggestion, at least with two

12 cycles, that these decreases can be cumulative. It

13 will be important to clearly understand whether

14 they are or not. So my sense is that they need to

15 be followed.

16 DR. DRAKE: I would ask the committee--I

17 agree with you on that, actually. That is my

18 sense. The question is is this important enough to

19 be done preapproval or postapproval. Does the

20 committee have a sense on that? Is this something

21 that can be done after approval to follow it out or

22 does it need to be done ahead of time?

23 DR. ABEL: I think it depends on the

24 number of cycles these patients are going to be

25 receiving.

289

1 DR. DRAKE: No, no. That is not the

2 question. If we decide to approve it, they will be

3 receiving cycles.

4 DR. ABEL: Well, that's true.

5 DR. DRAKE: So that is not the issue.

6 DR. EPPS: I think it should be done

7 before. Most of these people have only had two and

8 they still haven't recovered. That is just my

9 feeling. I think we need more data.

10 DR. DRAKE: We just heard Dr. Marzella

11 say, and I am not being argumentative. I am trying

12 to be a little bit of a devil's advocate. We just

13 heard him say that sometimes it takes years for us

14 to figure this out. In terms of risk-benefit, do

15 we want to deprive--if we decide this is

16 efficacious, do we want to deprive patients of this

17 drug?

18 DR. EPPS: At what risk?

19 DR. DRAKE: At what risk? I don't know.

20 That is the question I am posing to you guys.

21 DR. MARZELLA: If I can make another

22 comment. Another option would be to reconsider the

23 thresholds that one allows patients to decrease to.

24 That could be also tailored to specific

25 populations, some that are more susceptible,

290

1 obviously. So there are different ways of

2 approaching this.

3 DR. DRAKE: That is actually a very good

4 suggestion is modify the level that you allow them

5 to decrease to so that it is not particularly

6 dangerous so if it continues to go on, you have got

7 a little give room in there until you collect

8 further data. Is that what you are trying to say?

9 DR. MARZELLA: That is one option, I

10 think.

11 DR. DRAKE: That is one option. Good

12 idea. Seth?

13 DR. STEVENS: I would just like to say

14 that that was part of where I was coming from with

15 my question this morning about the relationship of

16 these picking 250 versus 300 cells. I guess, just

17 to balance Dr. Epps, I would be inclined to say

18 that those studies could be done after rather than

19 before because--for a long list of reasons.

20 DR. DRAKE: A sense of the committee. How

21 many think it could be done before? This is just a

22 sense of the committee. I am just going to have

23 them hold their hand up so I can kind of get a

24 sense. I am not getting by name at all. I am not

25 voting. I just want a sense.

291

1 Who thinks they can be done afterwards?

2 Okay; we are getting somewhere, then. That's good.

3 I hope you guys recorded that the committee split

4 but it seemed to me the sense was that--I am going

5 to restate it. The sense is that there are some

6 members of the committee who feel it should be done

7 premarketing but there is a greater number of the

8 committee that thinks it could be done

9 postmarketing.

10 But I think you are getting a sense that

11 there is a high level of caution that should be

12 exercised in this arena and certainly very careful

13 follow up and perhaps periodic reviews, maybe even

14 back before this committee sometime in the future

15 or back before the FDA, certainly, within a

16 rational period of time because I think the risk is

17 nobody wants it to get away from us because we are

18 uncertain about what we are going to see with

19 repeated cycles.

20 Is that a fair expression? Is that a nice

21 summary of where the committee is? Dr. Epps, you

22 don't agree. Feel free to speak up.

23 DR. EPPS: I am just listening.

24 DR. DRAKE: Okay.

25 DR. SWERLICK: I have a question. Is

292

1 there any data that would suggest that the average

2 T-cell count, CD4 count, seen after the infusion

3 which is within the normal range confers a risk of

4 infection to any population?

5 DR. DRAKE: There is no evidence of that

6 that we have been presented.

7 DR. SEIGEL: That CD4 counts such as were

8 observed here confer risk of infection to other

9 populations in other settings?

10 DR. SWERLICK: Yes.

11 DR. DRAKE: That statement was made that

12 both the sponsor and the FDA were in agreement on

13 that during the presentations.

14 DR. STEVENS: I guess I would just raise

15 the issue that entity that was popular several

16 years back of idiopathic CD4 lymphocytopenia in

17 which there were opportunistic infections and

18 malignancies that were associated with low CD4

19 counts that persisted in the absence of HIV and so

20 on.

21 That would be the only other instance that

22 I could consider.

23 DR. SEIGEL: I think not all CD4

24 lymphocytopenia is the same. In most cases, you

25 are going to have functional disturbances.

293

1 Sometimes, you have clonal deletions. Sometimes

2 you have selective memory or naive, depending on

3 the drug and the disease. So I am not exactly sure

4 how to approach that question.

5 DR. DRAKE: The Chair has recognized Dr.

6 Krueger.

7 DR. KRUEGER: I would like to make two

8 very brief comments. The first is I have, in a

9 study of effects on memory cells, subsetted the

10 memory-cell effects into long-term memory which are

11 called central-memory cells and then other cells

12 that are called peripheral memory cells which are

13 the bad guys, if you will. They are the short-term

14 effectors that end up at the skin and produce

15 psoriasis.

16 There is a relatively small effect of this

17 drug on decreasing the number of the long-term

18 memory cells. Instead, the effect is mainly in

19 this short-term expanded population. That, to me,

20 gives some comfort in the idea that long-term

21 memory is not being abrogated. But my studies are

22 limited to a single course and don't address the

23 multiple-course issue.

24 Secondly, I want to say that there were

25 studies done in England with an antibody called the

294

1 CAMPATH antibody many years ago which was

2 profoundly T-cell-depleting and produced T-cell

3 counts that were regularly below 100.

4 There were, in that setting, some

5 immediate concerns with infection seen but there

6 has actually now been many, many years of follow up

7 of patients that have stayed regularly with T-cell

8 counts below 100. In that setting, while there is

9 some risk, it is clear that it is a very different

10 risk setting from the AIDS population where the T-cell risk

11 below, let's say, 250 or 200 cells is

12 quite high.

13 So I think the risk of immunosuppressive

14 for an individual T-cell count really depends on

15 the circumstance.

16 DR. BONVINI: Dr. Krueger, could you

17 please state--I haven't seen the result of this

18 study that you have referred to now. Is this

19 derived from in vitro experience, in vivo, and if

20 these were patients, how many patients are involved

21 in the calculation?

22 DR. KRUEGER: May I have the Chair's

23 permission to show a slide?

24 DR. DRAKE: Yes.

25 You notice how he just happened to have

295

1 that at his fingertips?

2 [Slide.]

3 DR. KRUEGER: This is a measure in twenty-one

4 patients that are treated with alefacept with

5 the intravenous administration at the standard

6 dose. So this is the effect on these two groups of

7 cells that are called central memory and infector

8 memory. The overall effect on memory CD4s is about

9 a 30 percent reduction. What you can see is that

10 this long-term memory group is affected much less

11 than this and the p-value for this difference is

12 incredibly--

13 DR. BONVINI: Based on CCR7?

14 DR. KRUEGER: Based on CCR7 and CD4 who

15 have RA negativity as well as a lineage marker. It

16 was a four-color flow experiment. There is a

17 fourth antigen in this. So these are actual in

18 vivo data for psoriasis patients treated with the

19 drug.

20 DR. BONVINI: Were these responders,

21 patients--

22 DR. KRUEGER: This is a mixed group. I

23 will tell you that the responding patients tend to

24 have more depression of this group of cells

25 compared to nonresponders but that, in the

296

1 nonresponders--I'm sorry; this differential is

2 extremely well preserved.

3 DR. DRAKE: Dr. Katz?

4 DR. KATZ: I have a sense in our

5 discussions on the last two points that there is

6 some anxiety about the safety. If that is the

7 case, why need this be rushed without gathering

8 more patients? We are talking about 1,000

9 patients. We are talking about multiple courses of

10 how many patients, 300 patients.

11 It is a definitely effective drug but I

12 don't see the urgency before they gather--if there

13 is a little uncertainty with many more patients,

14 then that would be more valid to take the risk.

15 But, otherwise, we are dealing with small numbers

16 and anxiety around the table. The question is

17 everybody is talking about labeling and follow up

18 and so forth. Don't you think that that should be

19 done before it is released?

20 DR. DRAKE: Dr. King?

21 DR. KING: I guess if you take it in

22 context, I tend to think biologics and chemicals

23 like methotrexate are two different things.

24 Insulin has been around a long time. It is a

25 biologic. Growth factors for the hematopoietic

297

1 disorders, and so forth, are biologics. So there

2 is a great deal more information than you would

3 think out there.

4 This is building on that, not starting de

5 novo. So when you think about this product, you

6 are really talking about there is not any known

7 effect on the liver or the kidney. So now you are

8 talking about what is the effect on the immune

9 system which is what it is targeting. It is not

10 going to target the central nervous system or the

11 liver or the kidney. What you are really talking

12 about is what is your long-term risk for an

13 infection or cancer or whatever.

14 I have the bias that, basically, skin

15 cancer starts for most people in childhood. So you

16 are not literally going to survey cancer effects

17 for a long time except in a registry-type study.

18 So if those of us who are diabetic waited

19 for a long time until we got total risk issues on

20 insulin, most of us would be dead. So I am

21 comfortable with a registry as long as we define

22 what we are measuring and I haven't heard anything

23 here to tell me that infection was up or cancer is

24 up. All we really had a potential bogeyman of what

25 it may or may not do.

298

1 DR. KATZ: There is a little, not

2 statistically significant data, but there is a

3 little direction on most cancers and infections.

4 This is really not analogous to hormone-replacement

5 therapy. You are interfering with immune response.

6 Hopefully, this is going to be completely safe and

7 it will afford the 10 to 25 percent of patients

8 over placebo with effective treatment, but I am

9 just saying that, perhaps, more patients should be

10 treated.

11 DR. KING: Actually, I beg to differ with

12 you because I don't think of any difference between

13 a cytokine and a hormone. The immune system

14 releases peptides and peptides hit receptors and

15 that is how hormones work, at least the peptide

16 hormones work.

17 DR. DRAKE: Seth?

18 DR. STEVENS: I think we are back to the

19 question that Dr. Swerlick asked to start us off

20 which is how safe is safe enough. I think if we

21 repeated all the studies and we doubled the length

22 that they were followed and doubled the number of

23 cycles, maybe the statistics would shake out and

24 maybe they wouldn't.

25 But I think we are looking at shades of

299

1 gray rather than eventually reaching black or

2 white.

3 DR. ADELMAN: Madame Chairman?

4 DR. DRAKE: Yes.

5 DR. ADELMAN: Would it be possible for me

6 to put up one slide that just might help focus on

7 this conversation?

8 DR. DRAKE: Yes.

9 DR. ADELMAN: We recognize the challenge

10 and the concern about how much data are necessary

11 to approve a fundamentally novel drug in an

12 indication that has significant need. As some have

13 said, how much data is enough? You never really

14 have enough. That is why, in the context of our

15 conversation, we have discussed our commitment to

16 going forward with a very structured organized

17 registry or trial after approval that we would

18 envision would collect thousands of patients and

19 carefully monitor their long-term outcome from

20 safety and focussing on some of the key issues that

21 have been raised today which are absolutely correct

22 and relevant for concern.

23 But what I want to do is just point out

24 that the process continues even today as we speak

25 because there are 800 patients who are in various

300

1 stages of retreatment. The serious adverse events

2 we hear about immediately when they occur. So I

3 think that this slide, as of May 20th, so this is

4 current--you can see that right now, up to Course

5 5, we actually have 116 patients currently

6 receiving their fifth course of therapy.

7 The number of serious adverse events is

8 listed here. You can see that there are serious

9 adverse events that occur at all courses, but we

10 haven't seen anything new or unusual that we

11 haven't discussed today, and the trend is not

12 toward increasing incidence of serious adverse

13 events.

14 So we feel that this process is ongoing.

15 The agency is being made aware of this information.

16 They will be made aware of the information up to

17 and through an approval date and we will probably

18 expand the size of this group that we are

19 following.

20 But this is the core group to address the

21 question that has been raised which is how safe is

22 multiple treatment. These patients are undergoing

23 multiple treatment and we are carefully monitoring

24 their lymphocyte counts, incidence of infection,

25 incidence of malignancy and any other untoward

301

1 event that occurs.

2 DR. DRAKE: Dr. Adelman, we actually--I

3 think the whole committee appreciates that you are

4 doing that. We also appreciate that you have done

5 a valiant effort to give us the information we

6 need.

7 What I am hearing around the table though,

8 and I must really restrict this in the future to

9 the committee, please, I want to ask the sponsor

10 not to come to the microphone. If a committee

11 member wants to address a specific question to the

12 sponsor, you have my absolute permission to do so,

13 but we must allow the committee time for their

14 deliberation without a point-counterpoint at every

15 turn because much of this will fall out in the

16 discussion.

17 I have chaired many of these committees.

18 You would be amazed at how much falls out during

19 the discussion of intelligent people sitting around

20 the table thinking about it.

21 So I would like to continue, please, with

22 the committee deliberations. Dr. Tan?

23 DR. TAN: I was going to point out the

24 data just presented, I think, the follow-up data is

25 biased. I think the patients who don't respond,

302

1 you are not going to give him alefacept again;

2 right?

3 DR. DRAKE: A little slower.

4 DR. STEVENS: He said the study is biased

5 because the people who don't respond are not given

6 further rounds of alefacept.

7 DR. TAN: If the patients don't respond,

8 they won't get this drug again. So, therefore, if

9 you follow up those patients, you are always

10 studying those patients who respond. But, when you

11 first give the drug, the biologics, to the

12 patients, you don't know whether that patient is

13 going to respond or not.

14 DR. DRAKE: That's right. There are no

15 predictors. Absolutely. That is a very good

16 point. Thank you.

17 DR. EPPS: One-tenth of people were at the

18 fifth course than started out, 1,300 in the

19 beginning, 116 were at the end. So there was quite

20 a bit of drop off for whatever reasons. We don't

21 really know.

22 DR. DRAKE: Other comments? Bob?

23 DR. SWERLICK: I don't think you can

24 interpret that data necessarily that way. Those

25 people were staggered in how long they had been on

303

1 it. So some may not have been on the drug long

2 enough to be through the fifth course. Some may

3 have responded and stayed clear.

4 DR. EPPS: But that's what we don't know,

5 how many cleared.

6 DR. SWERLICK: But I think that is a

7 separate issue. You are talking about efficacy

8 versus safety issue.

9 DR. EPPS: That is an important issue.

10 DR. SWERLICK: In terms of the number of

11 patients who have undergone the fifth course, it

12 comes back to the same question I asked earlier.

13 If we are going to set a standard, a higher

14 standard, is it going to be an eternally moving

15 one. What I am trying to figure out is how many

16 patients would we have to study in order to detect

17 a certain frequency of adverse events and how many

18 patients would need to be studied.

19 So, if this isn't enough, how many

20 patients would be enough? I don't have the

21 statistical background to answer that, but are we

22 talking about another 1,500 patients? Are we

23 talking about 15,000 patients? How many is that?

24 DR. DRAKE: Dr. Seigel?

25 DR. SEIGEL: It depends on what you are

304

1 looking for, but if something doesn't occur in the

2 background and then you study 150 people and you

3 don't see it, you can be pretty sure that the rate

4 is 2 percent or less from a statistical

5 perspective.

6 If you increase that to a thousand people,

7 you can be pretty sure that it is a quarter of a

8 percent. So it is going to change. If it has a

9 background occurrence, as serious adverse events

10 go, as I said, it may be hard to tell no matter how

11 many you study whether it is real.

12 DR. SWERLICK: Because I have the same

13 anxiety regarding this whole new class of

14 medications, but if our response to that is simply

15 to say, well, we need to study more, we need to

16 study more, again, it comes back to how much is

17 enough. It has to be reasonably defined.

18 DR. DRAKE: Dr. Abel and then Dr. Tan.

19 DR. ABEL: Why couldn't we vote to approve

20 it with some limitations and not feel that it may

21 be--

22 DR. DRAKE: It is certainly one of the

23 committee's prerogatives.

24 DR. ABEL: Because cyclosporine was

25 approved for one year. Maybe there are some

305

1 thoughts about multiple cycles within a certain

2 time period and it could be approved with

3 qualifications.

4 DR. DRAKE: The committee can make any

5 recommendation they want to to the agency. We are

6 free to make a recommendation of--here are your

7 options. You can turn the whole thing down and

8 recommend that it not be approved. We are not the

9 final deciding authority, you should know. We are

10 just an advisory body to the FDA. They will make

11 the decision.

12 But we can recommend based upon our

13 deliberations that it shouldn't be approved at all.

14 You can recommend that it be approved but with some

15 caveats; here is what we think you ought to

16 continue to look at. Or you can say, boy, we think

17 it is great. Let's go. You have a range and that

18 is what we are here for.

19 We are to give the agency advice. They

20 will make the final determination based upon what

21 they have heard from the sponsor, from our experts

22 and from you guys. So your role here is to help

23 advise the FDA staff on what they might want to

24 look for irrespective of what our recommendation is

25 because they do not have to abide by our

306

1 recommendation.

2 But we certainly can make lots of them.

3 We have a lot of fun.

4 DR. SEIGEL: We will appreciate all of

5 them. Thank you.

6 DR. TAN: I had one more. I think this

7 has been brought up several times. I think, in

8 terms of the incidence rate, probably you want to

9 consider this in terms of the adverse events for

10 the alternative therapy as well.

11 DR. DRAKE: Okay. I am going to make kind

12 of a summary statement here. Would you all agree

13 that if we look at Part C under malignancies, it

14 says they went from 0.5 in placebo to 1.1 for

15 treated patients. I think the very same set of

16 questions could be asked about malignancies that we

17 have just asked about the rest of this section.

18 Is it fair for me to say that we want to

19 translate almost all of our comments from A and B

20 to C? The very same questions about malignancy are

21 going to apply. Yes?

22 DR. MORISON: With one proviso, that

23 infections will crop up probably early.

24 Malignancies may crop up late. So you could be two

25 years into a course of therapy and then start

307

1 seeing malignancies.

2 DR. DRAKE: I agree with you totally. We

3 need to have a longer time line for monitoring for

4 malignancies.

5 DR. MORISON: To go to the extreme, you

6 might say, well, you have got to look at these

7 people for fifteen years before you start finding

8 melanomas.

9 DR. DRAKE: Look at PUVA. Two years was

10 the earliest.

11 DR. MORISON: Two-and-a-half years.

12 DR. DRAKE: Two-and-a-half years was the

13 earliest; yes. So you will need a time line on

14 malignancies because they just are slower. No

15 matter what we do with it, you need a longer

16 monitoring period for that.

17 I must admit, I still am a little

18 concerned. The safety data that we just heard on

19 the animals bothers me just a little bit. I really

20 think that hyperplasia of the B-cells really must

21 be monitored to see what--it could just be

22 reactive, but it also needs to be in the monitoring

23 portfolio to make sure that that doesn't signal

24 anything important.

25 Now, then, Dr. Weiss and Dr. Seigel, do

308

1 you have enough information on Roman numeral I or

2 what other questions would you like to pose to us

3 or ask the committee?

4 DR. WEISS: I think you have addressed

5 those as well as anybody could.

6 DR. DRAKE: Yes; it is a little hard. But

7 we are getting there. At least we are pulling out

8 some information. As far as I am concerned, III

9 and IV sort of go together because the first

10 question on IV is how safe and effective is it.

11 So I want to devote just a couple of

12 minutes to efficacy. I want to talk about efficacy

13 for just a moment and then we will do IV because I

14 want to make sure we get that out of the way

15 before--the question on III, on efficacy outcomes,

16 because I think this is a quick for us, on the

17 outcomes part, the question is--we are back to

18 PASI. Is it okay to suggest that perhaps we have

19 discussed PASI already? Can we dispose of that

20 first question? Don't you have enough information

21 on opinion on PASI?

22 DR. WEISS: Yes; that's fine. It is

23 really more the issue about have they shown it to

24 be effective and then the overall risk-benefit

25 integration.

309

1 DR. SEIGEL: I would simply add, however,

2 that the question, although asked about in the

3 evaluation of this product and I think we have

4 heard well about the use of this in the evaluation

5 of this product probably has implications for what

6 sponsors seek to show for a variety of other

7 products that come along in psoriasis.

8 So, to the extent that there might be

9 suggestions, as some have said, that the PASI 75 is

10 insensitive or too high a response rate for trials,

11 I think, in the interest of time and getting

12 today's job done, it would be okay to skip over

13 that.

14 But, if we don't come back to it, we

15 might, at some future point, want to discuss with

16 this committee what are the optimal endpoints given

17 what we know now for new psoriasis trials.

18 DR. DRAKE: I couldn't agree with you

19 more. I think that we have grappled with this

20 issue on two separate committee meetings already

21 and I think it wouldn't hurt to have a third one

22 because we have got all kinds of stuff in the

23 pipeline that this committee and the agency are

24 going to consider.

25 So the more well-defined we can get is

310

1 going to help the sponsors. It is going to help

2 the committee. It is going to help you. So, Dr.

3 Seigel, I totally agree with you. I think that is

4 an extraordinarily important comment. The bar of

5 75 percent I think is reasonably high. On the

6 other hand, I bet you if we had some other slides,

7 we could show some other folks who didn't improve

8 as much.

9 I think you need to see the whole spectrum

10 as you are making these decisions is what I am

11 trying to say. You need to see some of the placebo

12 patients to get a sense. You need to see the whole

13 spectrum if you are going to be making

14 determinations about the PASI score, I think.

15 Let's talk about efficacy. Let's have

16 just a little bit of open discussion about efficacy

17 before we actually go to the vote because we

18 haven't discussed that. I want comment from the

19 members of this committee about efficacy of this,

20 whether you use the PASI score, the physician's

21 global assessment, whatever you use. What are your

22 reactions regarding the data and the information we

23 have received.

24 Dr. Abel?

25 DR. ABEL: I think you have to look at all

311

1 three assessments, PASI, physician's global

2 assessment and the quality of life. I think that

3 the efficacy seems very impressive especially in

4 terms of the fact that you think of this as a

5 remittive therapy and that there are going to be

6 long remissions and we don't have any treatments

7 for psoriasis that are like that except for PUVA,

8 and maybe UVB.

9 So I think it definitely has plus.

10 DR. DRAKE: Dick?

11 DR. TAYLOR: I agree. I am impressed with

12 the efficacy of this product. I think, in looking

13 at the patients that we have seen and hearing from

14 patients that have received it, I agree that

15 looking at all three of the parameters for

16 evaluating efficacy, that they are all good. As I

17 said before, I think the PASI 75 is much too high

18 and PASI 50 would probably be more reasonable.

19 If you did that, then the efficacy is very

20 impressive.

21 DR. DRAKE: As a custom I have, I like to

22 go around the room and make sure everybody talks

23 when we get to this point because I want to hear

24 what everybody has to say about both efficacy and

25 additional comments on safety.

312

1 Bob, would you start. Dick, you can

2 repeat what you have or not, but everybody in the

3 room be thinking about what you want to say because

4 I am going to call on everybody.

5 DR. SWERLICK: My impression is that when

6 compared to what I use now, this drug seems like it

7 will be as effective or more effective and

8 potentially even safer than some of the other

9 poisons that I have to resort to using.

10 I had one other issue to raise.

11 DR. DRAKE: Please.

12 DR. SWERLICK: That has to do with the

13 safety. This is likely to be combined with other

14 biologics. That actually hasn't come up yet.

15 Should we wait until--that has to do with the

16 product labeling or--

17 DR. DRAKE: Yes; let's wait on that. But

18 using it in conjunction with something else is a

19 problem no matter what we approve, or don't

20 approve. It is just absolutely an issue. But, for

21 now, I would like to keep it sterile. Let's assume

22 this is a sterile process.

23 I don't know if I am going to invite you

24 again or not because you ask too hard questions.

25 I'm just teasing, you understand. They are very

313

1 important. Do you want to comment on safety while

2 you are at it?

3 DR. SWERLICK: Again, I think that it is

4 not if something happens to somebody on this drug.

5 It is when. But I think, compared to the risks

6 associated with everyday life that this compares

7 well with other therapies given the information we

8 have on hand now, and that the amount of additional

9 study that would be required to identify the low-frequency

10 catastrophic events, the 747 going down

11 in New York City sort of business.

12 The numbers involved in that sort of study

13 would be huge you will pick it up in postmarketing.

14 That is my bias.

15 DR. DRAKE: Dr. Taylor, please give us

16 your total range of thoughts.

17 DR. TAYLOR: As I said before, I think

18 this is a sufficiently efficacious agent to

19 consider approval. I would agree that, compared to

20 other treatments that I presently use all the time,

21 this is at least equal if not better than most.

22 I think the other issue is that, as far as

23 the risk is concerned, I think many of the problems

24 that we have all identified will be identified in a

25 registry if the registry is set up well enough and

314

1 it will be identified much more rapidly

2 postmarketing than it will be premarketing. So I

3 would think that we are not going to get the

4 numbers premarketing that we need to make the

5 decisions. So I would think that we know enough

6 about the risk right now to go ahead.

7 DR. DRAKE: Dr. Abel.

8 DR. ABEL: I would agree with that. I

9 think it should be approved now. I think it

10 compares favorably, more than favorably, with other

11 systemic therapies for psoriasis. I, too, am

12 concerned about the risks and the repeated courses,

13 the number of cycles, the time interval. I think

14 that we have to develop guidelines to decrease the

15 risk of potential side effects and monitor these

16 patients very closely long-term for both short-term

17 infections and long-term for infections and

18 malignancies. And there may be some caveats

19 written into the approval.

20 DR. DRAKE: I am going to derail my own

21 process here. I wanted to ask everybody a quick

22 opinion about dose. Without it being a total

23 discussion, I forgot we didn't address that. I am

24 going to go back to you three and ask you to give

25 me your opinion on dose and then would the rest of

315

1 you include that as we go around the table.

2 Bob, tell me what you think about dose.

3 DR. SWERLICK: I am confused. The

4 pharmacokinetics would suggest that the dose is not

5 going to be critical, but there is enough data that

6 would suggest that there may need to be dose

7 adjustment for certain subgroups of individuals

8 based upon size, not necessarily just weight but

9 other factors.

10 I think, again, it is one of those things

11 that it can be hashed out post-approval.

12 DR. DRAKE: Do you recommend further

13 studies on that?

14 DR. SWERLICK: Yes.

15 DR. DRAKE: Dr. Taylor

16 DR. TAYLOR: I have already given you my

17 impression of dose earlier on. I really think it

18 ought to be weight-adjusted rather than a given

19 dose.

20 DR. DRAKE: Dr. Abel?

21 DR. ABEL: That makes sense to me. We

22 talked about that early on and I would favor the

23 weight-based. But that doesn't seem to apply with

24 IM, so if it were just IM, it seems to be okay to

25 use the fixed dose. I am wondering about the

316

1 options for IM versus IV. How are we to choose?

2 Why are both of these routes being offered?

3 If it is just the fixed dose, then maybe

4 IM is the ideal way for it to be given.

5 DR. DRAKE: Thank you. Now, Ms. Knudson,

6 we certainly haven't heard much from you today. As

7 an IRB person, you probably have quite a few

8 comments on safety and everything else. So please

9 share them.

10 MS. KNUDSON: My concern, of course, is

11 that this is a highly vulnerable population. I

12 suspect that as soon as it is approved, there will

13 be many, many, many patients who will want to take

14 the drug and could be followed. So long-term

15 effects I think could be found with some ease as

16 long as that registry is set up appropriately.

17 I think the safety is certainly better

18 than toxins that are used currently. This is

19 infinitely better. It seems to be at least as

20 efficacious. I don't think I can comment on the

21 dose except I am concerned about children and size

22 and if children are going to be included.

23 DR. DRAKE: Thank you. Dr. Stevens?

24 DR. STEVENS: With respect to dose, I

25 think we have heard the issues with respect to

317

1 weight and all of that. The other side of that

2 observation, of course, is that if you think that

3 it is less effective in heavier people, then the

4 data would shake out that it would be more

5 effective then we are thinking globally for the

6 lighter people when we look at the entire cohort

7 that was studied. So I think that is a

8 postmarketing issue.

9 My remaining question with respect to

10 dosing goes back to what I mentioned earlier about

11 the reduction of lymphocytes at six weeks. I think

12 you can always redesign experiments and studies.

13 There are infinite variations that you can do on

14 these. My question, with respect to dosing, is the

15 twelve-week dosing regimen as opposed to a shorter

16 one. But, again, I think that is one for

17 postmarketing.

18 I am also impressed with efficacy, as

19 everyone else has mentioned and I agree with the

20 comment that was made explicitly by Dr. Tan but

21 reiterated by the others that the question before

22 us with respect to safety goes towards if we do not

23 allow this therapy to be available, what will these

24 patients be doing otherwise.

25 They will be using these other therapies

318

1 that have been demonstrated to have safety issues.

2 DR. DRAKE: Dr. Katz.

3 DR. KATZ: As far as using other therapies

4 with safety issues, no doubt this therapy will have

5 safety issues also but that would have to be

6 acceptable. As far as safety, thus far, probably

7 the safety profile is fairly good. There are some

8 indications, though, that there may be problems. I

9 feel that there is not enough people who have been

10 treated with these indications, with infection and

11 malignancy, that we have to be much more cautious.

12 Also, as far as efficacy, there is no

13 question is it very impressively efficacious in a

14 small number of patients. Now, there are people

15 here who treat more psoriatics than I do, although

16 I have my average patient share. But some people

17 have psoriasis clinics and so they know more than I

18 do.

19 So when they say it is as efficacious as

20 anything, then I respect that. However, with a

21 PASI even of 50 which we will say is good, 24

22 percent over placebo--24 percent. Now I ask those

23 who said it is as efficacious as the others, do you

24 not get more than 24 percent improvement with PUVA,

25 with methotrexate, 80 percent, 80 percent with

319

1 PUVA. Of course, I am talking with Dr. Morison

2 here so he can address that.

3 Those have their risks over decades. As

4 physicians, we have to make that judgment with our

5 patients whether they are willing to subject

6 themselves to those risks. But I think that it

7 would be a useful alternative after more studies

8 are done but, certainly, clear or almost clear 9

9 percent over placebo, and 16 percent PASI 75

10 certainly shows that it is efficacious, but I

11 wouldn't agree with its being impressive.

12 The other thing that bothers me a little

13 and I would admit that this may be irrelevant,

14 especially with respect to what Mark said and he

15 couldn't differentiate it. But I wonder about the

16 blind being negated in part so that, really, the

17 efficacy is even really less than we are told here

18 because the same physicians are--I mean, there was

19 a difference in I think it was 11 percent it the IM

20 reaction. So I have my reservations.

21 DR. DRAKE: Thank you, Dr. Katz. Dr.

22 Morison?

23 DR. MORISON: In addressing the three

24 issues, I think as far as weight is concerned,

25 everything I have heard today sounds confusing to

320

1 me. It makes sense to me that a milligram per

2 kilogram approach would be the best approach but

3 hearing all the data, I am confused as to whether

4 that is going to be possible to sort out with

5 further studies. Certainly, to me, it would be an

6 ideal approach.

7 So far as the PASI 75 is concerned, I sort

8 of take exception to the comments that have been

9 made to some extent. Let's say I am in a different

10 camp. I am used to dealing with narrow-band UVB

11 and Hoover's main treatments and they certainly do

12 exceed PASI 75. Hoover, you can clear people to 95

13 percent in a very consistent way.

14 I think you can clear 90 percent of

15 patients with PUVA and UVB to 95 percent clear.

16 So, certainly, those treatments have a higher

17 standard.

18 Having said that, I would 100 percent

19 agree with everybody's comments that we need more

20 agents because certainly I have patients who are in

21 trouble, end stage, can't get in for treatment and

22 I would love some more agents to use to treat

23 psoriasis because certainly the ones we have now,

24 methotrexate and Soriatane, and cyclosporine have

25 lots of drawbacks.

321

1 The final point is my only real concern is

2 safety. I think we are sort of launching into a

3 biologic experiment where I am not quite sure we

4 are headed. When I say that is the one concern I

5 have is malignancy because the psoriasis population

6 is a unique population, quite different from

7 rheumatoid arthritis patients and such like.

8 This is a group of patients who spend a

9 maximal amount of time down at Ocean City. They

10 have had a maximal exposure to UVB and many of them

11 had a lot of exposure to PUVA. They are all primed

12 for the development of skin cancer. Almost the

13 whole severe group of patients with psoriasis are

14 primed to develop skin cancer. It is something

15 that is going to take a few years to develop.

16 We have already seen it with cyclosporine.

17 I hope we don't see it with this particular agent.

18 That is why I think that we need a very solid

19 follow up to detect it as early as possible.

20 DR. DRAKE: Thank you, Dr. Morison. Dr.

21 Epps?

22 DR. EPPS: Thank you. I think I have made

23 some of my impressions known. Of course, we all

24 wish we had more agents to use. I would have hoped

25 that statistically and otherwise it would be

322

1 stronger in support of this medication even though

2 I do tend to think beyond just the nine months of

3 improvement. Even a twenty-year-old could have a

4 life-expectancy of fifty more years. And we just

5 don't know.

6 Of course, we are not going to wait fifty

7 years, but my point is that even if, in this brief

8 period, there was malignancy potential, I think we

9 need to think very seriously about it even as Dr.

10 Morison has already alluded to, PUVA exposure, UVB

11 exposure and also natural-light exposure.

12 The other signal is infection. Sometimes,

13 it is not the opportunists that we see. It is the

14 severe common infection. It is the ones that we

15 see all the time which are more severe or act

16 differently that we need to watch for.

17 Should we get to the dosing, perhaps a

18 body-mass index may be a better way to look at it

19 rather than just kilos. There have certainly been

20 a lot of things in the media recently about

21 overweight of Americans and other ways to look at

22 that, but BMI may be one way of dose as opposed to

23 just straight kilograms.

24 DR. DRAKE: Dr. Epps, thank you. Dr.

25 King?

323

1 DR. KING: I am struck by the three

2 different ways of measuring effectiveness but my

3 mother was a business woman and she always said

4 that, "You may have it, but the customer may not

5 buy it." So PASI always reminded me that the

6 physician and the patient were looking at the same

7 thing. You could agree on how much you have. The

8 physician global was what the doctor thought was

9 there, but the quality of life is what the patient

10 perceives.

11 So I have always put more emphasis on how

12 much did the person perceive that I had done for

13 them, how much did their psoriasis improve.

14 Sometimes, people go away happy with, say, 50

15 percent or even a small patch that was on her face

16 and yet they could cover up the rest of it.

17 So I am struck that this is efficacious.

18 It may not be the total body cure, but there are

19 lots of folks who have not only no access to a

20 psoriasis daycare center, they have no access to a

21 dermatologist.

22 So I come down on the side of a unit dose

23 and access where people can inject themselves under

24 the supervision of the dermatologist, et cetera, so

25 they don't have to figure it out. They are not

324

1 going to give themselves IV this drug or any other

2 drug. Having taken insulin shots, myself, I will

3 tell you I would much rather have a fixed dose than

4 trying to calculate what I was supposed to take.

5 So I come down on the side this is

6 efficacious as a nice alternative. It doesn't

7 interfere with the liver or kidney and you have a

8 certain population of patients that just can't take

9 these. So, for a home-therapy unit dose,

10 efficacious may be not the barn burner, then I come

11 down on the side of approval of this drug with

12 appropriate monitoring. I would worry lots about,

13 as I counted in this recent review on biological

14 therapy for psoriasis, there are already twelve

15 agents in the pipeline so you we have to be careful

16 what we say for the first agent like this in this

17 category that we don't give either the FDA or the

18 manufacturers unreasonable expectations and too

19 high a bar so that it won't become available to

20 patients.

21 DR. DRAKE: Thank you, Dr. King. Dr. Tan?

22 DR. TAN: I do consider that the agent is

23 efficacious with impressive duration of remission.

24 But I don't think there is sufficient data to

25 suggest whether it should adjust for the weight

325

1 level, whether or not it needs to be further

2 studied.

3 DR. DRAKE: Thank you. Dr. Raimer.

4 DR. RAIMER: As has been brought out by

5 several individuals, we certainly do need more

6 treatment options for psoriasis. Fortunately, most

7 of the ones we have, their side effects don't occur

8 until we have given them several months of

9 treatment. So I would sort of really like having

10 this as another option to rotate people onto as

11 another treatment.

12 Obviously, all of us have patients who are

13 sort of out of options. They can no longer take

14 methotrexate. They don't respond to other drugs

15 and we do need another drug to be able to treat

16 these severe patients who are out of options.

17 My main concern also is with the potential

18 of malignancy eventually developing. I am not as

19 worried about skin cancers even though that is not

20 insignificant because we can watch the skin. If we

21 follow these patients closely, we can remove these

22 lesions when they are small before they are a

23 problem.

24 I think internal malignancy is more of a

25 worry, but these are probably not going to show up

326

1 for years, maybe. So I would be in favor of doing

2 postmarketing studies to watch for malignancies

3 rather than holding the drug up at this point in

4 time.

5 Finally, I would be for a standardized

6 test also with more studies looking at patients on

7 the heavy and light end, maybe looking to see if

8 doses need to be adjusted for those patients. Some

9 more studies for heavy and light folks, but I would

10 be in favor of a standardized for the majority of

11 folks.

12 DR. DRAKE: Terrific. I am ready to call

13 for a vote on Question Roman numeral IV if Dr.

14 Seigel and Dr. Weiss have no objection. Is there

15 anything else you want me to get on the table

16 before I call for a vote? It is okay?

17 Dr. Swerlick, we are sorry. You have been

18 so helpful but you can't vote. What I would like

19 is to vote--I think I will put them together

20 because, if we recommend approval, the safety and

21 effectiveness go together. That is the FDA's

22 primary mission, is it safe and effective. So we

23 are going to put them together.

24 I would like a show of hands from voting

25 members on--oh; we have to do each one? Okay,

327

1 fine. We are going to go around the table with a

2 vote. This question that you are voting on is has

3 the sponsor shown that this biologic is safe and

4 effective for use in adults for chronic plaque

5 psoriasis.

6 DR. KING: Wait, wait, wait. You didn't

7 address the issue of candidates for or there is

8 something--they failed out of methotrexate,

9 whatever. You are just saying naive patients who

10 have never been treated with anything else.

11 DR. WEISS: I guess the first question is

12 do people believe it should be recommended for an

13 approval and then we can get to potentially what

14 population.

15 DR. DRAKE: Lloyd, what I thought we were

16 going to is--

17 DR. KING: I was just bringing that

18 question up.

19 DR. DRAKE: Once we get to that, then we

20 are going to--actually, I am going to have you go

21 to that and to children and to other populations

22 and to labeling; all right.

23 DR. KING: Right.

24 DR. DRAKE: But is everybody clear on the

25 vote? Please identify your name and your vote

328

1 DR. TAYLOR: Richard Taylor. I vote

2 positive for approval.

3 DR. ABEL: Elizabeth Abel. I vote yes,

4 for approval.

5 MS. KNUDSON: Paula Knudson. I vote yes,

6 for approval.

7 DR. STEVENS: Seth Stevens. I vote for

8 approval.

9 DR. KATZ: Robert Katz. I vote for

10 nonapproval at this time.

11 DR. MORISON: Warwick Morison. I vote for

12 approval.

13 DR. EPPS: Roselyn Epps. I vote against

14 approval at this time.

15 DR. KING: Lloyd King. I vote for

16 approval at this time with the appropriate registry

17 and directed by the FDA.

18 DR. TAN: Ming Tan. Vote for approval

19 with caution on the second course.

20 DR. RAIMER: Sharon Raimer. I vote for

21 approval.

22 DR. DRAKE: The Chair records a vote of

23 eight for and two opposed. Is that correct? Does

24 everybody agree?

25 DR. SEIGEL: I would just like to point

329

1 out--because we have a lot of confusion on and

2 during and after these advisory committees. What

3 we ask for is a vote as to whether this is safe and

4 effective in terms of meeting the clinical

5 standards for approval.

6 DR. DRAKE: I stand corrected.

7 DR. SEIGEL: I assume that is the vote we

8 received and that's fine. The only reason I

9 highlight that is because, as was mentioned and is

10 not a subject for discussion, there are issues

11 regarding the manufacturing this product and making

12 sure it meets other standards that are not on the

13 table now that we are not putting forward right now

14 to this committee.

15 So I take those votes for approval as

16 indicating that, with regard to safety and

17 efficacy, it meets appropriate standards for

18 approval.

19 DR. DRAKE: I totally--I misstated that

20 although I thought I had covered--I did cover it

21 earlier but I should have restated it. We are not

22 approving or disapproving. We are giving our

23 recommendation to further the approval process to

24 the FDA, that we think this would be a nice drug to

25 get on the market with certain follow up,

330

1 registries, et cetera. That is the vote of the

2 committee.

3 DR. SEIGEL: Right.

4 DR. DRAKE: And that is reflected eight to

5 two. Fair enough? As the Chair, I didn't vote. I

6 tend to vote when it is a tie. And one abstention.

7 I try to remain neutral so that I facilitate and

8 don't bias. So I try very hard not to bias the

9 committee.

10 I want to tell you that I apologize. I

11 have got to leave. I have a mom that is ill and I

12 just ca