FOOD AND DRUG ADMINISTRATION












                          Monday, May 3, 2004


                               8:10 a.m.




                           Hilton Washington

                           620 Perry Parkway

                         Gaithersburg, Maryland





         Donna Przepiorka, M.D., Ph.D., Chair

         Johanna M. Clifford, M.S., RN,

           Executive Secretary




         John T. Carpenter, Jr., M.D.

         Bruce D. Cheson, M.D.

         James H. Doroshow, M.D.

         Stephen L. George, Ph.D.

         Antonio J. Grillo-Lopez, M.D.

         Pamela J. Haylock, RN

         Silvana Martino, D.O.

         Gregory H. Reaman, M.D.

         Bruce G. Redman, D.O.

         Maria Rodriguez, M.D.

         Sarah A. Taylor, M.D.




         Michael Bishop, M.D.

         Ronald Bukowski, M.D.

         Ralph D'Agostino, Ph.D.

         Maha Hussain, M.D.

         Jan Buckner, M.D.

         Wen-Jen Hwu, M.D.

         Joanne Mortimer, M.D.

         Michael Perry, M.D.




         Kenneth McDonough (for Genasense)

         Natalie Compagni-Portis (for RSR 13 Injection)


         FDA STAFF:


         Richard Pazdur, M.D.

         Grant Williams, M.D.

         Robert Temple, M.D.



                            C O N T E N T S


      Opening Remarks, Donna Przepiorka, M.D., Ph.D.             5


      Comments by Congressman Peter Deutsch                      5


      Comments by Alex Delpizo                                  11


      Conflict of Interest Statement,

         Johanna M. Clifford, M.S., RN                          14


      Opening Remarks, Richard Pazdur, M.D.                     20


      Genta Presentation:


         Introduction, Loretta M. Itri, M.D.                    26

         Melanoma Overview, John Kirkwood, M.D.                 29

         Study GM301, Loretta M. Itri, M.D.                     36

         Clinical Benefit Summary,

           Frank Haluksa, M.D., Ph.D.                           60


      FDA Presentation:


         Medical Review, Robert Kane, M.D.                      69

         Statistical Review, Peiling Yang, Ph.D.                76

         Clinical Relevance, Robert Kane, M.D.                  86


      Questions from the Committee                              93


      Open Public Hearing                                      125


      Committee Discussion                                     152


      Allos Presentation:


         Introduction, Pablo J. Cagnoni, M.D.                  206

         Brain Metastases, John H. Suh, M.D.                   209

         The Science of RSR13, Biran D. Kavanaugh,

           M.D., MPH                                           216

         Clinical Efficacy Results,

           Pablo J. Cagnoni, M.D.                              225

         Conclusions, Paul A. Bunn, Jr., M.D.                  251


      FDA Presentation:


         Clinical Review, Kevin Ridenhour, M.D.                254

         Statistical Review, Rajeshwari Sridhara, Ph.D.        265



                            C O N T E N T S


      Questions to the FDA and the Sponsor                     278


      Open Public Hearing                                      309


      Subgroup Analysis in Clinical Trials,

         Stephen George, Ph.D.                                 314


      Committee Discussion                                     333




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


  2                         Opening Remarks


  3             DR. PRZEPIORKA:  Good morning to all and


  4   welcome to the Food and Drug Administration's


  5   Advisory Committee for Oncologic Drugs.  My name is


  6   Donna Przepiorka.  I will be chairing the


  7   committee.  I just wanted to remind everyone in the


  8   audience that the purpose of the individuals on


  9   this panel is to serve as independent consultants


 10   to the FDA.  We do not work for the FDA.  We are


 11   also not anyone who makes any decisions; we only


 12   provide advice.


 13             Our first item on the agenda--we are going


 14   to go a little bit out of order.  We want to hear


 15   first from Congressman Deutsch who has a few words


 16   to say.


 17             CONGRESSMAN DEUTSCH:  Thank you very much.


 18   I appreciate the opportunity to be here.  My name


 19   is Congressman Peter Deutsch, and I recognize that


 20   it is not at every meeting of this committee that


 21   you are addressed by a member of Congress.  Largely


 22   it is in that capacity that I speak to you today,




  1   but it is also in my capacity as an individual who


  2   has been personally affected by the specter of


  3   melanoma.


  4             On several occasions I have had basal


  5   cells removed from my body.  Thankfully, they were


  6   not malignant but their existence renders me high


  7   risk.  My dermatologist now evaluates me on a


  8   quarterly basis for melanoma and guides me on how


  9   to reduce my risk profile.  I pray that this risk


 10   never materializes but, if it does, I need to know


 11   that my physician and I have access to every


 12   therapeutic treatment available for this horrible


 13   disease.  As someone who actually hears people


 14   testify in many settings, I am trying to get your


 15   attention so actually I have pictures of my kids


 16   who both have red hair so, obviously, they are high


 17   risk for skin cancer as well especially as having a


 18   parent who has been diagnosed with basal cells.


 19   They also happen to live in Florida.


 20             Again, most of the people in this room


 21   don't live in Florida and I am not exaggerating


 22   that the school that they go to and, in fact, the




  1   schools they have gone to since pre-K, do not have


  2   hallways.  It is one of the unique things about


  3   Florida, south Florida in particular so they are


  4   literally outside all the time.  For anyone who has


  5   kids, especially in a setting like south Florida,


  6   think about the summer when you try to get your


  7   kids to wear suntan lotion.  It is not an easy


  8   thing to do.  So, this is a very real thing.  I


  9   mean, I have fights with my kids, especially as


 10   they have gotten older, about putting suntan lotion


 11   on, on a continuous basis.


 12             But it is not just for my kids; it is not


 13   for myself that I am here today.  It is for all the


 14   constituents I represent and all the citizens


 15   around the nation.  So, it is on their behalf as


 16   well that I stand before you today, not to advocate


 17   for the approval of this drug but to advocate that


 18   the mind set from which you consider this


 19   application be your own mind set--clinical


 20   physicians dedicated to the welfare of their


 21   patients.


 22             What does this mean?  That this




  1   application be a referendum on whether you would


  2   want this drug available to your patients if they


  3   were diagnosed with metastatic melanoma.  That is


  4   the standard we owe cancer patients and that is the


  5   standard government is obligated to uphold.


  6             I did not come here to preach to this


  7   committee to the extent me and Congress have had


  8   frustration with over-regulation by the FDA.  It is


  9   not of your doing; quite the opposite.  It is


 10   people like yourselves who give up your time to


 11   guide the FDA.  I cannot over-emphasize the


 12   importance of your role.  You provide the FDA a


 13   window that they otherwise do not have, a window


 14   into the real world, if you will, a world in which


 15   dying cancer patients are desperate for and must be


 16   given access to every reasonable treatment that


 17   might save their lives.


 18             As you may know, there were two relevant


 19   newspaper articles last week that got some


 20   attention in Congress.  One was an article in The


 21   New York Times about a Japanese study published in


 22   The New England Journal of Medicine proving the




  1   effectiveness of a drug called UFT in treating a


  2   form of lung cancer.  What was staggering about the


  3   article was that this same technology was rejected


  4   in this country by the FDA.  In other words,


  5   thousands of cancer patients in this country could


  6   be dying because the government failed them.


  7             What I later learned was that the FDA


  8   rejected this drug even though this very advisory


  9   committee composed of your predecessors voted


 10   unanimously to approve it and, because the FDA did


 11   not accept the recommendations of clinicians,


 12   countless Americans lack access to that drug today.


 13   That is inexcusable.


 14             In the other article, the Wall Street


 15   Journal related to this committee's hearings.  It


 16   offered no views on whether this drug should be


 17   approved but, instead, noted the absence of


 18   treatments for metastatic melanoma and a couple of


 19   vignettes about the people who took the drug.  One


 20   of those was an individual names David Bernstein


 21   who is scheduled to join us here today.  Mr.


 22   Bernstein is a fourth grade teacher from a small




  1   town in New Jersey.  The article said that Mr.


  2   Bernstein's cancer went away and he is alive today,


  3   teaching his students in his fourth grade classroom


  4   because of the drug before you today.


  5             I am not a physician nor a scientist and I


  6   have not studied the clinical data regarding this


  7   drug, but I do know this, if you find that this


  8   drug is as safe and effective as other available


  9   treatments, if it reasonably presents another


 10   possible course of treatment, by what right can


 11   government deny cancer patients an avenue to save


 12   their lives?  This is not about a passing illness


 13   for which there are other treatments.  This is


 14   about cancer, an absolutely devastating disease


 15   that has in some ways affected nearly every single


 16   American.  This is about cancer patients who are


 17   dying and desperate for a chance to live longer.


 18   It is in their interest that we must be foremost in


 19   today's hearing.


 20             I flew back to Washington last night to


 21   speak to you this morning, however, prior


 22   obligations in my district require me to actually




  1   literally turn around right now and return to


  2   Florida this morning.  I regret that I can't stay


  3   here to listen to all of the testimony but I wish


  4   to thank this committee for its time, and it has


  5   been an honor and pleasure to speak with you this


  6   morning.


  7             DR. PRZEPIORKA:  Thank you, Congressman


  8   Deutsch.  Any questions for the Congressman?


  9             [No response]


 10             Thank you, sir.


 11             CONGRESSMAN DEUTSCH:  Thank you.


 12             DR. PRZEPIORKA:  Next we will hear from a


 13   representative from Congressman Ferguson's office.


 14             MR. DELPIZO:  My name is Alex Delpizo.  I


 15   am here representing Congressman Mike Ferguson of


 16   New Jersey who, unfortunately, is in New Jersey and


 17   couldn't be here with us today.


 18             I am not a scientist or a clinician or a


 19   chemist but everyone knows a person whose life has


 20   been taken by cancer.  For me, that person was my


 21   mother.  She fought and eventually lost her


 22   six-year battle with cancer.  However, due to




  1   miracle life-extending drugs she saw two of her


  2   children get married and met her three


  3   grandchildren.  My mother was fortunate enough to


  4   experience all of the wonderful things that mothers


  5   and grandmothers experience later in life.


  6             As you know, Genasense us used to treat


  7   stage 4 metastatic melanoma.  Metastatic melanoma


  8   is currently a death sentence.  When two available


  9   therapies treat the disease and the last


 10   chemotherapy therapy treatment was approved in


 11   1975, yours is an awesome responsibility.  The FDA


 12   works every day to ensure that Americans and their


 13   food and drug supply are safe.  Your decisions on


 14   which drugs are approved are based on numbers, and


 15   numbers are very important, however, we would never


 16   want to approve a placebo.  However, an


 17   over-emphasis on statistics at the expense of


 18   patient needs does a life-threatening disservice.


 19   The failure to appreciate mean or median


 20   statistical analyses in any size sampling also


 21   fails to take into account a patient population


 22   that achieved the most dramatic overall response.




  1             Given the devastating nature of this


  2   disease and the relatively few treatments


  3   available, even marginal increases in life


  4   expectancy can clearly be the difference between


  5   rapid death and years of life extension for those


  6   patients that will see a benefit from this and


  7   other drugs.


  8             In closing, I would like to highlight the


  9   experience of one of my constituents in Montgomery


 10   Township in New Jersey.  David Bernstein was


 11   diagnosed with skin cancer and prescribed


 12   chemotherapy to remove a grape-sized tumor on his


 13   chest.  Mr. Bernstein opted to supplement the


 14   chemotherapy by joining a clinical trial of an


 15   experimental drug.  Six weeks after his first dose


 16   he received the news that his tumor had essentially


 17   disappeared.  This was two years ago.  That


 18   experimental drug was Genasense.


 19             For my mother, David Bernstein and for all


 20   of those who have been diagnosed with cancer, I


 21   respectfully request that you look favorably on


 22   Genasense and other new drug applications that can




  1   provide hope for those for whom hope is all they


  2   have.  Thank you very much.


  3             DR. PRZEPIORKA:  Thank you.  Again, I


  4   would like to ask the folks who are standing along


  5   that far wall by the doors to please step outside


  6   into the hall, or take a seat, or take a stand at


  7   the back wall only, please.  You are going to need


  8   to vacate that area immediately, please.


  9             We would like to now move on to the first


 10   item on the agenda and Johanna Clifford will read


 11   the conflict of interest statement.  Thank you.


 12                  Conflict of Interest Statement


 13             MS. CLIFFORD:  Thank you.  The following


 14   announcement addresses the issue of conflict of


 15   interest with respect to this meeting and is made a


 16   part of the record to preclude even the appearance


 17   of such at this meeting.


 18             Based on the submitted agenda and


 19   information provided by the participants, the


 20   agency has determined that all reported interests


 21   in firms regulated by the Center for Drug


 22   Evaluation and Research present no potential for a




  1   conflict of interest at this meeting, with the


  2   following exceptions:


  3             In accordance with 18 USC Section


  4   208(b)(3), Dr. Ronald Bukowski has been granted a


  5   waiver for serving on a competitor's advisory board


  6   on an unrelated matter for which he receives less


  7   than $10,000 a year; consulting with the sponsor of


  8   dacarbazine on an unrelated matter for which he


  9   receives less than $10,000 a year; and, finally,


 10   for consulting with a competitor on an unrelated


 11   matter for which he receives less than $10,000 a


 12   year.


 13             Dr. Maha Hussain has been granted waivers


 14   under 18 USC 208(b)(3) and 21 USC 505(n) for


 15   unrelated consulting for the co-developed of


 16   Genasense for which she receives less than $10,000


 17   a year; and owning stock in the co-developer of


 18   Genasense, valued from $25,001 to $50,000.


 19             Dr. Wen-Jen Hwu has been granted a limited


 20   waiver under 18 USC 208(b)(3) for her employer's


 21   contract with a competitor for an


 22   investigator-initiated study of a competing




  1   product.  The contrast is less than $100,000 a


  2   year.  Under the terms of the waiver, Dr. Hwu will


  3   be permitted to participate in the committee's


  4   discussions of Genasense.  She will not, however,


  5   be able to vote.


  6             A copy of these waiver statements may be


  7   obtained by submitting a written request to the


  8   agency's Freedom of Information Office, Room 12A-30


  9   of the Parklawn Building.


 10             We would also like to disclose that Dr.


 11   Silvana Martino has been recused from participating


 12   in all matters concerning Genta's Genasense.


 13             Lastly, we would like to note for the


 14   record that Dr. Antonio Grillo-Lopez, Chairman,


 15   Neoplastic and Autoimmune Diseases Research


 16   Institute, is participating in this meeting as in


 17   industry representative, acting on behalf of


 18   regulated industry.  He would like to disclose that


 19   he is a scientific advisor to Chiron and receives


 20   speakers fees from Roche.


 21             In the event that the discussions involve


 22   any other products or firms not already on the




  1   agenda for which FDA participants have a financial


  2   interest, the participants are aware of the need to


  3   exclude themselves from such involvement and their


  4   exclusion will be noted for the record.


  5             With respect to all other participants, we


  6   ask in the interest of fairness that they address


  7   any current or previous financial involvement with


  8   any firm whose product they may wish to comment


  9   upon.


 10             DR. PRZEPIORKA:  Thank you.  Once again,


 11   there are still some folks registered for the open


 12   public hearing who have not signed in.  I just want


 13   to remind you that if you do wish to speak at the


 14   open public hearing you will need to sign in at the


 15   table outside.


 16             Next, I would like the members of the


 17   committee and the other participants to introduce


 18   themselves and we will start with Dr. Pazdur.


 19             DR. PAZDUR:  Richard Pazdur, Director of


 20   the Division of Oncology Drug Products, FDA.


 21             DR. WILLIAMS:  Grant Williams, FDA,


 22   Director, Division of Oncology Drugs.




  1             DR. FARRELL:  Ann Farrell, clinical team


  2   leader for Genasense.


  3             DR. KANE:  Robert Kane, medical reviewer.


  4             DR. YANG:  Peiling Yang, statistical


  5   reviewer.


  6             DR. BUKOWSKI:  Ron Bukowski, medical


  7   oncologist, Cleveland.


  8             DR. BISHOP:  Michael Bishop, Experimental


  9   Transplantation, Immunology Branch, National Cancer


 10   Institute.


 11             DR. HWU:  Wen-Jen Hwu, medical oncologist


 12   at the Memorial Sloan-Kettering.


 13             DR. TAYLOR:  Sarah Taylor, University of


 14   Kansas.


 15             DR. REAMAN:  Gregory Reaman, George


 16   Washington University and Children's National


 17   Medical Center.


 18             DR. REDMAN:  Bruce Redman, University of


 19   Michigan.


 20             MS. CLIFFORD:  Johanna Clifford, FDA,


 21   executive secretary for this meeting.


 22             DR. PRZEPIORKA:  Donna Przepiorka,




  1   University of Tennessee, Memphis.


  2             DR. RODRIGUEZ:  Maria Rodriguez, medical


  3   oncologist, M.D. Anderson Cancer Center.


  4             DR. DOROSHOW:  Jim Doroshow, Division of


  5   Cancer Treatment and Diagnosis, NCI.


  6             DR. CHESON:  Bruce Cheson, Georgetown


  7   University Lombardi Comprehensive Cancer Center.


  8             DR. GEORGE:  Stephen George, Duke


  9   University.


 10             MS. HAYLOCK:  Pamela Haylock.  I am a


 11   nurse and I am at the University of Texas.


 12             DR. CARPENTER:  John Carpenter, University


 13   of Alabama at Birmingham.


 14             DR. D'AGOSTINO:  Ralph D'Agostino, Boston


 15   University biostatistician.


 16             DR. MORTIMER:  Joanne Mortimer, medical


 17   oncology Eastern Virginia Medical School.


 18             DR. HUSSAIN:  Maha Hussain, University of


 19   Michigan.


 20             MR. MCDONOUGH:  Ken McDonough, patient


 21   representative.


 22             DR. GRILLO-LOPEZ:  Antonio Grillo-Lopez,




  1   Neoplastic and Autoimmune Diseases Research


  2   Institute.


  3             DR. PRZEPIORKA:  Thank you to all.  I


  4   think Dr. Pazdur will open with some remarks.


  5                         Opening Remarks


  6             DR. PAZDUR:  Thank you very much, Donna.


  7   First, I would like to recognize the contributions


  8   of four ODAC members who will be leaving the


  9   committee after this meeting.  These members


 10   include our chairman, Donna Przepiorka, John


 11   Carpenter, Sarah Taylor and Bruce Redman.  We, at


 12   the FDA, recognize their efforts at providing us


 13   advice at these public meetings and, in addition,


 14   we appreciate their valuable assistance throughout


 15   the years in providing us with their insights at


 16   other FDA meetings and in reviewing and assessing


 17   protocols.  Our work and the welfare of the


 18   American public is greatly facilitated by their


 19   hours of work and their talents devoted to these


 20   tasks.  Again, Donna, John, Sarah and Bruce, we


 21   thank you for your efforts, your patience with our


 22   phone calls, and advice on some of the most




  1   perplexing issues of drug development.  Let me say


  2   this, this is not "adios" but "hasta la vista" and


  3   it is not "hasta la vista, baby."  We will be


  4   calling you; we will be in touch; this will be a


  5   continuous process that we will be dealing with you


  6   over the years, but we do appreciate your kindness


  7   and your efforts at helping us with some of the


  8   problems that we have at hand.


  9             Let's turn to the issues at hand.  This


 10   morning's meeting focuses on a drug for the


 11   treatment of patients with advanced melanoma who


 12   have not received prior chemotherapy.  I would like


 13   to spend some time addressing issues for you to


 14   consider during the presentations provided by the


 15   sponsor and the FDA staff.  These issues are


 16   important to this application but also this


 17   afternoon's application and in drug development in


 18   general, especially as we have continuing, ongoing


 19   discussions and dialogue with the committee on


 20   endpoints for drug development.


 21             The FDA has long considered the


 22   demonstration of an improved survival as the gold




  1   standard for drug approval.  An improvement in


  2   survival associated with an acceptable safety


  3   profile is of unquestionable clinical benefit.  It


  4   is assessed daily and is unambiguous.  When we, at


  5   the FDA, began our discussions with the committee


  6   on drug approval we realized that there may be some


  7   disadvantages to requiring survival improvement for


  8   drug approval.  These disadvantages include the


  9   confounding of survival analysis by crossover with


 10   frequently large patient numbers required to be


 11   enrolled on trials for survival, and the long


 12   follow-up that may be required in selected


 13   oncological diseases.


 14             This trial at hand this morning was


 15   originally discussed with the agency to be a trial


 16   with a primary endpoint of survival improvement.


 17   The trial did not demonstrate an improvement in


 18   overall survival.  We are asked to evaluate this


 19   drug for approval on the basis of secondary


 20   endpoints of claimed improvements in


 21   progression-free survival or PFS and response


 22   rates.  Please member that since this drug is added




  1   to a standard therapy we must assess the drug's


  2   contribution to that standard therapy and any


  3   claimed response rates or claims for PFS advantages


  4   represent a combination of the investigational


  5   agent and the standard therapy.  Hence, we must


  6   isolate the efficacy of the drug in assessing the


  7   drug's efficacy.


  8             Let's turn our attention to the


  9   measurement and assessment of PFS which will be


 10   discussed during this meeting on multiple


 11   occasions.  The assessment of PFS may be difficult


 12   and uncertain in unblinded trials with a small


 13   effect on this endpoint and where there is a lack


 14   of attention to clinical trial issues that are


 15   important in measuring and comparing PFS data


 16   between treatment arms.  These issues include a


 17   prospectively defined methodology for assessing,


 18   measuring and analyzing PFS.  These need to be


 19   detailed in the protocol and in the statistical


 20   plan.  Tumor progression should be carefully


 21   defined in the protocol.  The FDA and the sponsor


 22   should agree prospectively on the protocol, the




  1   case report forms and the statistical analysis plan


  2   for PFS.  There should be a prespecified analysis


  3   plan for handling missing data, especially missed


  4   assessment visits.  Censoring methods and


  5   assessment of progression in non-measurable lesions


  6   must be prospectively outlined and agreed upon.


  7   Most importantly, visits and radiological


  8   assessments should be symmetrical on the study arms


  9   to prevent systematic bias.  When possible, studies


 10   should be blinded.  This is especially important


 11   when the patient or investigator assessments are


 12   included as components of the progression endpoint.


 13   If progression is assessed by both the treating


 14   physician and an external review panel or an


 15   external radiology committee, the protocol should


 16   prospectively stipulate whose assessment will be


 17   used in defining PFS.  This cannot occur after the


 18   study data has been examined.


 19             Hence, from a practical perspective, PFS


 20   as a primary endpoint for drug approval takes


 21   meticulous, prospective planning.  The measurement


 22   of PFS progression-free survival requires rigor. 




  1   This planning is frequently lacking in clinical


  2   trials that relegate PFS to a secondary endpoint.


  3   Some practical problems outlined above in


  4   accurately characterizing the treatment of PFS will


  5   be discussed by the FDA reviewers.


  6             Provided an acceptable safety profile, one


  7   has to answer the following question, what is the


  8   magnitude of the drug's effect on PFS that would be


  9   considered clinically relevant?  A very small


 10   effect may raise questions about the very existence


 11   of this effect, especially when the study is


 12   unblinded and attention to the symmetry of


 13   assessments and handling of missing assessments is


 14   not evident.


 15             In answering whether marketing approval


 16   should be granted to an agent, two important


 17   questions need to be answered.  First, does the


 18   drug have a convincing effect that can be


 19   adequately characterized?  Secondly, and this


 20   question can only be addressed if the first


 21   question is answered in the affirmative, what is


 22   the clinical relevance of the effect?  This




  1   obviously must take into account a risk-benefit


  2   analysis.  However, benefit can only be assessed in


  3   this equation if it convincingly exists and also


  4   can be adequately characterized.


  5             I hope these comments will provide a


  6   catalyst for your considerations this morning, this


  7   afternoon and tomorrow as we discuss endpoints of


  8   drug approval.  Donna, I turn the program over to


  9   you and I will answer questions after the FDA


 10   presentations.  Thank you.


 11             DR. PRZEPIORKA:  Thank you, Dr. Pazdur.


 12   Let's go ahead and begin with the sponsor


 13   presentation, with an introduction by Dr. Itri.


 14                       Sponsor Presentation


 15                           Introduction


 16             [Slide]


 17             DR. ITRI:  Dr. Przepiorka, members of the


 18   Oncology Drug Advisory Committee, ladies and


 19   gentlemen, it is my pleasure, on behalf of Genta,


 20   to introduce the agenda and the participants for


 21   the presentation of the new drug application for


 22   Genasense in combination with dacarbazine for the




  1   treatment of patients with advanced malignant


  2   melanoma.


  3             Following my introductory remarks, Dr.


  4   John Kirkwood will give an overview of malignant


  5   melanoma and available treatments.  After Dr.


  6   Kirkwood's presentation I will return to the podium


  7   and discuss the results of GM301 in detail.  At


  8   that point, Dr. Frank Haluska will summarize the


  9   risks and benefits in the context of the disease we


 10   are treating.


 11             [Slide]


 12             By way of introducing our speakers, Dr.


 13   Frank Haluska is from Harvard University and Mass.


 14   General Hospital.  He is chairman of the CALGB


 15   melanoma committee.  Dr. John Kirkwood is professor


 16   and vice chairman of Medicine at the University of


 17   Pittsburgh and is also chairman of the ECOG


 18   melanoma committee.


 19             [Slide]


 20             In addition to our distinguished speakers,


 21   we are fortunate to have with us today a number of


 22   clinical experts in the field of melanoma,




  1   including Dr. Sanjiv Agarwala from the University


  2   of Pittsburgh Cancer Center, Dr. Agop Bedikian from


  3   M.D. Anderson Cancer Center, Dr. Paul Chapman from


  4   the Memorial Sloan-Kettering Cancer Center, Dr.


  5   Robert Conry from the University of Alabama, Dr.


  6   Peter Hersey from the University of Newcastle, all


  7   the way from Australia, and Dr. Evan Hersh from the


  8   University of Arizona Cancer Center.


  9             Drs. Bedikian, Conry, Hersey and Hersh


 10   were principal investigators in our study and


 11   together are responsible for managing approximately


 12   20 percent of patients who are on our trial.  They


 13   are available to address any issues you may have


 14   regarding patient management in the study.  Dr.


 15   Janet Wittes, formerly head of statistics at the


 16   National Heart, Lung and Blood Institute and


 17   currently president of Statistics Collaborative, is


 18   available to provide expert biostatistical


 19   consultation.  Dr. Robert Ford, chief medical


 20   officer and founder of RadPharm, is with us to


 21   address the intricacies related to the blinded


 22   independent review of radiographic studies.  I




  1   would like to now invite Dr. John Kirkwood to the


  2   podium.


  3                        Melanoma Overview


  4             DR. KIRKWOOD:  Thank you, Loretta.


  5             [Slide]


  6             Dr. Pazdur, Dr. Przepiorka, members of


  7   ODAC and the FDA, I am delighted to speak with you


  8   today about a disease that many of us here have


  9   spent all of our lives working on.


 10             [Slide]


 11             This is a disease that has risen in


 12   epidemic proportions and is 4 percent of new


 13   cancers, rising at 5 percent per year.  The


 14   mortality from this cancer is also rising and most


 15   notably for men over 50 for whom there is a 157


 16   percent increase in mortality in just the last


 17   decade.  The societal impact of this cancer is even


 18   more because of its median age of incidence in the


 19   late 40s, and it takes a toll in terms of


 20   productive life years that exceeds many more


 21   frequent cancers, even including prostate cancer.


 22             [Slide]




  1             In the past 37 years only three agents


  2   have been approved for the treatment of this


  3   disease in the advanced setting.  Not one of these


  4   agents was approved on the basis of randomized,


  5   controlled Phase 3 trials prior to their approval.


  6   None of these agents has ever shown a survival


  7   benefit.  Approval of these agents was based solely


  8   on response rate.


  9             Hydroxyurea, approved in 1967 with a 10


 10   percent response rate, has not been used in the


 11   clinical community for 20 years or more.


 12             Dacarbazine, approved in 1975 with a


 13   response rate of 23-25 percent, has more recently


 14   been summarized in an article to appear next month


 15   in the European Journal of Cancer.  The response


 16   rates that range between 7-13 percent I think are


 17   far more accurate assessments of the true response


 18   rate to this agent.  Most of these were done


 19   pre-RECIST criteria and we don't know really what


 20   the objective response rate will be in larger


 21   trials using the newer RECIST criteria that have


 22   been used for the study to be discussed today.




  1             [Slide]


  2             Turning to IL-2, the most recent agent


  3   approved for the treatment of metastatic melanoma,


  4   the IL-2 NDA pooled 8 Phase 2 small studies.  The


  5   regimen was not compared in these to any other


  6   therapy.  The approval was based upon quality of


  7   response, durable responses and, given the


  8   significant toxicity of this agent, the population


  9   that was treated was highly atypical of the general


 10   community of patients that we have to deal with in


 11   the country at large.  The median age was 42 years.


 12   The patients had in general no co-morbidity in


 13   terms of cardiac or pulmonary disease.  Most of the


 14   patients who had responses had disease confined to


 15   skin, lymph nodes and lung.  The toxicity of this


 16   regimen is so regularly, predictably severe that,


 17   in fact, specialized units are required for the


 18   administration of this agent.  Its administration


 19   is confined to specialized centers in general


 20   across the country.


 21             [Slide]


 22             IL-2 responses were noted in 16 percent of




  1   patients treated, about one-third of whom had


  2   surgery to maintain this complete response, and 10


  3   percent partial responses, defined using pre-RECIST


  4   criteria.  The most salient aspect of the IL-2


  5   benefit in these patients has been the long


  6   duration of response observed in some patients.


  7   While the median duration of patients treated at


  8   large was 9 months, the median duration for


  9   patients who achieved complete responses was


 10   greater than 5 years.  Unfortunately, the number of


 11   those complete responses alive is rather small.


 12   The drug-related mortality with this treatment in


 13   this series was 2 percent, further compromising


 14   this relative benefit.


 15             [Slide]


 16             Over the years there have been many


 17   attempts to improve upon the therapeutic benefit of


 18   dacarbazine.  The largest of the trials conducted


 19   in the last five years are summarized in this


 20   slide, beginning with the IL-2 experience which was


 21   Phase 2 and, therefore, for which no comparator


 22   exists.




  1             These include the Dartmouth regimen,


  2   adding tamoxafin to BCNU, cisplatin and


  3   dacarbazine; two regimens of biochemotherapy


  4   including one that the Eastern Cooperative Oncology


  5   Group and the Intergroup presented to the ASCO


  6   meetings just a year ago, now enrolling 416


  7   patients; and a similarly large study from the


  8   EORTC that has not yet been published; as well as a


  9   publication just recently in JCO from the French


 10   group with a total number of more than 1000


 11   patients in which overall there has been no


 12   combination that has shown a statistically


 13   significant difference in overall response rate, in


 14   complete response rate, in durable response rate or


 15   in progression-free survival.


 16             [Slide]


 17             I appeared last before this committee in


 18   1999 in relationship to metastatic melanoma.  In


 19   that setting, it was to introduce the application


 20   for temozolomide.  This is an oral equivalent of


 21   dacarbazine that I think no one questions was


 22   equivalent to dacarbazine.  The committee did not




  1   vote to approve that agent which achieved


  2   equivalency in a trial that had been targeted upon


  3   superiority.  But since that time I think it has to


  4   be admitted that temozolomide has been the most


  5   widely used drug in the community across the


  6   country.  The FDA briefing that you have before you


  7   suggests that Genasense is, in fact, comparable to


  8   temozolomide.  I would argue that it is not.


  9             The overall response rate for the


 10   temozolomide application was not significantly


 11   different.  The complete responses, identical; the


 12   durable responses, not detailed; and the


 13   differences in progression-free survival with an


 14   asymmetrical interval of assessment for the two


 15   arms, as Dr. Pazdur has just spoken about,


 16   significant but 11 days.


 17             The other major difference about


 18   temozolomide is that this agent was already going


 19   to be available to the community at large for trial


 20   exploration, and the agent that we are going to


 21   discuss today will not be available if it is not


 22   approved today.




  1             [Slide]


  2             In summary, despite more than 25 years of


  3   work and low response rates with the single agent


  4   dacarbazine, this agent remains the reference


  5   standard for the field.  No single cytotoxic drug


  6   nor any biological agent or combination has been


  7   shown to be superior to single agent dacarbazine in


  8   relation to survival.


  9             Relative to dacarbazine, no large


 10   randomized, multicenter comparative study has ever


 11   shown a statistically significant benefit in


 12   overall response rate, in complete response rate or


 13   in progression-free.


 14             High-dose IL-2 is a useful agent that many


 15   of us use for selected patients who lack


 16   significant co-morbidity and who are willing to


 17   accept its side effects.  This drug is not suitable


 18   for the majority of patients who present to us with


 19   metastatic melanoma and is particularly unsuited


 20   for patients who are elderly.


 21             [Slide]


 22             I would conclude that metastatic melanoma,




  1   upon which I have focused the last 33 years of my


  2   work, is a drug-refractory neoplasm.  We need new


  3   agents desperately.  Thank you.


  4                           Study GM301


  5             DR. ITRI:  Thank you, Dr. Kirkwood.


  6             [Slide]


  7             Genasense is an example of a new class of


  8   drugs called antisense.  Antisense is fundamentally


  9   a protein knockout strategy.  Genasense inhibits


 10   Bcl-2 production.  Bcl-2 is a protein and is


 11   believed to be an important mediator of cancer cell


 12   resistance to chemotherapy.  Genasense is


 13   administered for 5 days before chemotherapy,


 14   reduces Bcl-2 production and renders the cancer


 15   cell more susceptible to chemotherapy.  In this


 16   way, Genasense is postulated to enhance the


 17   efficacy of chemotherapy.


 18             [Slide]


 19             Bcl-2 is ubiquitously expressed by


 20   melanoma cells.  Five days of continuous IV therapy


 21   with Genasense prior to the administration of DTIC


 22   resulted in approximately 70 percent reduction in




  1   Bcl-2 levels in melanoma cells taken from patients


  2   before and after Genasense treatment.  These


  3   results provided the rationale for a Phase 3 study


  4   in patients with advanced malignant melanoma.


  5             [Slide]


  6             This study is the largest randomized trial


  7   ever conducted in patients with advanced malignant


  8   melanoma.  It was an open-label, multicenter trial


  9   involving 139 investigational sites in 9 countries


 10   around the world.


 11             The primary endpoint was overall survival


 12   and the secondary endpoints included


 13   progression-free survival, antitumor responses


 14   using computer calculated RECIST based on


 15   evaluations of site tumor measurements; durable


 16   responses which were defined as responses lasting


 17   longer than 6 months; and, of course, safety in all


 18   patients.


 19             [Slide]


 20             Patients received either DTIC at the


 21   standard dose of 1000 mg/m                                                

                          2 or the same dose of


 22   DTIC preceded by a 5-day continuous infusion of




  1   Genasense at a dose of 7 mg/kg/day.  Patients were


  2   stratified according to the three major prognostic


  3   factors for melanoma, ECOG performance status 0 or


  4   1-2; the presence or absence of liver metastases;


  5   and normal or elevated LDH levels.  Patients could


  6   receive up to 8 cycles during a treatment phase


  7   which were administered every 21 days.  Restarting


  8   evaluations were performed at the end of every two


  9   cycles.


 10             It is important to note that the timing of


 11   interval measurements were fixed and similar in


 12   both arms, and they were prospectively defined with


 13   FDA agreement, with the temozolomide review issues


 14   clearly in mind.  Crossover was not permitted from


 15   the DTIC arm into the Genasense arm, and follow-up


 16   was continued for 2 years in both arms of the


 17   study.  Patients on the Genasense arm only could


 18   receive up to an additional 8 cycles of the


 19   combination therapy in extension protocol GM214 if


 20   they achieved at least stable disease by the end of


 21   the treatment phase and it was considered to be in


 22   the best interest of the patient, in consultation




  1   with the treating physician.


  2             [Slide]


  3             The statistical assumptions for this study


  4   were based on an overall median survival for DTIC


  5   of 6 months which was derived from published


  6   reviews.  Genasense was postulated to add an


  7   additional 2 months, for total a median survival of


  8   8 months; 750 patients would provide 90 percent


  9   power to see a difference between groups, with an


 10   alpha level of 0.05.  It was assumed that accrual


 11   would be constant at 30 patients per month.  In


 12   agreement with FDA, an analysis was planned when at


 13   least 508 deaths had occurred on the study.


 14             [Slide]


 15             The two groups were balanced for age and


 16   gender.  The median age of patients in this study


 17   was 60 years but patients ranged in age from 16 to


 18   93.  Approximately 40 percent of our patients in


 19   this study were greater than 65 years of age and,


 20   remarkably, more than 10 percent were more than 75


 21   years of age.


 22             [Slide]




  1             The two groups were equally balanced with


  2   regard to baseline performance status and


  3   approximately half of all patients were symptomatic


  4   at baseline.


  5             [Slide]


  6             Similarly, the two groups were balanced


  7   with respect to the major prognostic indicators


  8   including time from initial diagnosis, LDH/disease


  9   site distribution and prior immunotherapy which


 10   consisted primarily of alpha interferon


 11   administered as an adjuvant therapy in both groups.


 12             [Slide]


 13             Forty patients who were randomized into


 14   the study did not receive treatment.  The primary


 15   reason for this is that in the DTIC arm some


 16   patients, later being randomized to the standard of


 17   care, were unwilling to travel or withdrew consent


 18   once they learned they would not be receiving


 19   experimental therapy.  The amount of DTIC delivered


 20   to both groups was equivalent.  Overall, the


 21   addition of Genasense did not require dose


 22   reduction of DTIC.




  1             [Slide]


  2             This is a summary of the efficacy


  3   parameters which, taken together, provide evidence


  4   for the benefit of combining Genasense with DTIC.


  5   I will discuss each of these in more detail in


  6   following slides.


  7             Although not statistically significant,


  8   improvement in overall survival was noted for the


  9   Genasense group.  Statistically significant


 10   improvement was noted in both progression-free


 11   survival and response rates, and I will shortly be


 12   showing you some interesting updated results


 13   regarding complete responses in this study.  We


 14   also saw a positive trend in patients with durable


 15   responses.


 16             [Slide]


 17             The FDA has raised a number of


 18   considerations for the committee's review.  These


 19   include response rate concordance; the impact of


 20   interval assessments on progression-free survival;


 21   the impact of missing data on progression-free


 22   survival; baseline differences in prognostic




  1   factors; and the influence of non-U.S. sites on


  2   response rate.  I will address each of these issues


  3   separately in the appropriate sections of my


  4   presentation.


  5             [Slide]


  6             This Kaplan-Meier plot of overall survival


  7   shows that both arms outperformed expectations.


  8   DTIC was associated with a 7.9 month median


  9   survival as opposed to the expected 6 months, and


 10   Genasense treatment resulted in a 9.1 month median


 11   survival.  These differences were not statistically


 12   significant.  Please note that the overall survival


 13   curves begin to separate at 6 months and the median


 14   follow-up at the time of database lock was 7


 15   months.


 16             [Slide]


 17             The addition of Genasense was associated


 18   with an overall response rate of 11.7 percent as


 19   compared to 6.8 percent for DTIC alone.  This


 20   difference is significant, with a p value of 0.019.


 21   Use of the stringent RECIST measurement system has


 22   historically reduced response rates in other




  1   studies by 25-50 percent when compared to


  2   investigator determinations.


  3             [Slide]


  4             It is appropriate at this point to discuss


  5   how responses were calculated in this study.  The


  6   investigators did not determine response.


  7   Investigators measured lesions and entered these


  8   data onto an electronic case report form.  The


  9   computer then calculated whether the response met


 10   criteria for RECIST.  RadPharm was only contracted


 11   to review responding patients.  The sponsor was


 12   provided with measurements of target lesions and


 13   evaluations of non-target lesions by RadPharm.


 14   These measurements were also assessed by the same


 15   computer algorithm using RECIST criteria.  RadPharm


 16   reviewers were blinded as to the treatment arm and


 17   all clinical information in which tumors had been


 18   selected by the sites as target lesions.  All marks


 19   made by the sites on x-rays were removed.


 20             There are three major reasons why RadPharm


 21   readings might not have been strictly concordant


 22   with the site measurements.  These include the




  1   evaluation of different target lesions with


  2   different measurements, the absence of important


  3   clinical information regarding preexisting lesions


  4   and controversy regarding the reporting of normal


  5   or residual lymph node tissue.


  6             [Slide]


  7             The patient on this slide had extensive


  8   liver metastasis at baseline which resolved


  9   completely during treatment.  This patient has


 10   remained in complete  clinical remission for


 11   approximately three years.


 12             [Slide]


 13             Due to the presence of a persisting liver


 14   lesion in the same patient, RadPharm was unable to


 15   confirm a complete response.  By procedure,


 16   RadPharm was unaware that this was a documented


 17   preexisting cystic lesion that was benign.  This


 18   patient is being cared for by Dr. Hersey who is


 19   here with us today and can answer any questions you


 20   might have regarding her treatment course.


 21             [Slide]


 22             In the next case, which demonstrates how




  1   the absence of medical history can confound


  2   concordance, a biopsy-proven metastatic lesion of


  3   the frontal sinus was read by RadPharm as


  4   incidental sinusitis.  Because this patient had


  5   undergone a Caldwell Luck enterotomy with removal


  6   of the inferior turbinate due to metastatic


  7   melanoma, RadPharm reasonably assumed that this was


  8   an infectious process and did not confirm the


  9   response.


 10             [Slide]


 11             Because RECIST criteria do not provide


 12   guidance for the interpretation of normal lymph


 13   nodal architecture at the site of previous disease,


 14   RadPharm could not confirm complete response in the


 15   next case and several others like it.  Despite


 16   complete regression of the tumor next to the blood


 17   vessel, here, RadPharm could only assign partial


 18   response due to the presence of small residua.


 19             The PET scan results for this same patient


 20   confirmed complete clinical response and shows no


 21   residual evidence of a viable signal post


 22   treatment.  The FDA did not review any of these




  1   x-rays and based their concordance judgments solely


  2   on raw measurements in percent reductions provided


  3   by the sponsor at their request.  I urge the


  4   committee to address questions regarding


  5   radiographic reviews to Dr. Robert Ford, who is


  6   here with us today as an expert consultant in


  7   radiology and who personally reviewed all of these


  8   films.


  9             [Slide]


 10             Seventy-one responding patients were


 11   evaluated by RadPharm and 60 of these were


 12   considered to be evaluable; 11 patients were not


 13   evaluable due to the poor quality of photographs or


 14   films or the absence of lesions which could be


 15   considered measurable by RadPharm.  Five of these


 16   cases occurred in the Genasense arm and 6 occurred


 17   in the DTIC arm.


 18             Point-to-point concordance for two time


 19   point evaluations were available for 38 patients


 20   and give the concordant rate of 63 percent which is


 21   consistent with literature citations for


 22   evaluations of this nature.  Two additional




  1   responding patients were confirmed to be responses


  2   but were assessed differently by the site and by


  3   RadPharm.  Eight cases were consistent at a single


  4   evaluation and were within 10 percent of response


  5   at the second evaluation.  Four patients, such as


  6   the ones I have previously described to you, were


  7   easily explained by the absence of appropriate


  8   medical history.  If we include only the 40


  9   responders confirmed by RadPharm and agreed to by


 10   the FDA on treatment comparison, Genasense is


 11   completely consistent to DTIC as demonstrated by


 12   odds ratios.  If only those 40 responses considered


 13   to be confirmed by both RadPharm and the FDA are


 14   included, odds ratios reveal a 91 percent


 15   improvement in response rate by RadPharm compared


 16   to an 82 Percent improvement in response for


 17   Genasense as reported in the NDA.


 18             [Slide]


 19             These cases were randomly selected by FDA


 20   and included 40 cases in each arm of the study.


 21   X-rays were collected from around the world and


 22   included assessments which occurred in the




  1   follow-up period after NDA cutoff.  As a


  2   consequence of this unplanned review of cases,


  3   RadPharm was able to identify additional responses


  4   which occurred in the follow-up period after NDA


  5   cutoff.  These important clinical findings prompted


  6   Genta to evaluate all patients in follow-up who met


  7   RECIST criteria for response during at least one


  8   time point during the treatment phase and all


  9   patients who ended the treatment phase without


 10   disease progression and who had received no


 11   intervening therapy.


 12             [Slide]


 13             As with response, we observed good


 14   concordance regarding the conclusions about time to


 15   progression between the investigational site


 16   assessments and RadPharm determinations.  When the


 17   site assessments and RadPharm determinations for


 18   time to progression are compared, both showed a


 19   benefit for the Genasense group.  RadPharm


 20   assessments of time to progression in the Genasense


 21   group were generally longer than the site


 22   assessments.




  1             [Slide]


  2             Six additional responses have been


  3   identified which occurred in the follow-up period


  4   after the NDA submission and all were in the


  5   Genasense group.  Only complete responses are


  6   reported since they are the ones most unequivocally


  7   associated with clinical benefit and constitute a


  8   result not commonly observed with single-agent


  9   DTIC.  Three of these complete responses were


 10   upgraded from the partial response category and 3


 11   were patients with long-standing stable disease.


 12   Information regarding these additional responding


 13   patients was submitted to the FDA on April 9th of


 14   this year.


 15             It is important to note that the submitted


 16   database has not been updated or altered in any


 17   way, nor are we attempting to change the data


 18   provided in our NDA.  We wish simply to inform you


 19   of important and frankly unanticipated clinical


 20   findings.  These responses all occurred in the


 21   absence of other intervening therapies and have


 22   been documented by duplicate CT scans using the




  1   same RECIST criteria as specified in the protocol.


  2   The physicians caring for several of these patients


  3   are here with us today and are able to answer any


  4   questions you may have directly.


  5             [Slide]


  6             Complete responses were evenly distributed


  7   by gender and generally exhibited the same


  8   demographic pattern as the overall population.


  9   Importantly, one-third of the responses occurred in


 10   patients with elevated LDH and half were observed


 11   in the worst AJCC prognostic categories, M1b and


 12   M1c.


 13             [Slide]


 14             Survival for the complete responders


 15   ranges from 15 months to more than 3 years on the


 16   Genasense arm, and 19 to 21 months on the DTIC arm.


 17   The plus signs denote ongoing responses.  Two


 18   patients have died, one on each arm of the study.


 19             [Slide]


 20             The evolution of the complete responders


 21   on this study is shown in this slide.  The two


 22   responding DTIC patients are shown in yellow for




  1   comparison.  The solid bar denotes the database


  2   cutoff of August 1, 2003 and is the information


  3   contained in the NDA.  The dotted line denotes the


  4   date of the FDA inquiry that precipitated review in


  5   the follow-up period after database cutoff.


  6             As you can see, partial responses tend to


  7   occur later in the Genasense arm and evolved over


  8   time into complete responses.  Three of the


  9   Genasense responses, similar to what has been


 10   described for IL-2, have been surgically


 11   maintained.  Once again, all responses were based


 12   on strict RECIST criteria with duplicate


 13   measurements and no patient received intervening


 14   therapy.


 15             [Slide]


 16             Returning now to the data previously


 17   reported in the NDA database, the duration of


 18   response is presented using a box-and-whisker plot


 19   on this slide.  The red line denotes the median.


 20   The top of the box is the boundary of the third


 21   quartile and the bottom is the boundary of the


 22   first quartile.  As you can see, the medians are




  1   similar but an important difference is observed in


  2   the third quartile, resulting in a longer mean


  3   duration of response in patients who received


  4   Genasense.


  5             [Slide]


  6             Durable responses, defined as responses


  7   lasting at least 6 months, were more than doubled


  8   in the Genasense group, as shown in this slide.


  9             [Slide]


 10             Median progression-free survival for the


 11   Genasense group was 74 days as compared to 49 days


 12   for the DTIC group.  The relative risk of having


 13   progressive disease or death was reduced by


 14   approximately 27 percent in the Genasense arm.


 15   These differences are highly significant, with a p


 16   value of 0.0003.


 17             Time to progression was performed as a


 18   sensitivity analysis for progression-free survival.


 19   The results were very similar and showed


 20   approximately a 27 percent reduction in the risk of


 21   progressive disease.  In this analysis, 11 patients


 22   who died without documented disease progression




  1   were censored to the day of last lesion


  2   measurement.  These 11 patients constitute the only


  3   difference between progression-free survival and


  4   time to progression in this study, and explain why


  5   the two curves are so similar.


  6             [Slide]


  7             Genta conducted multiple sensitivity


  8   analyses to address possible biases in the


  9   calculation of progression-free survival.  In all


 10   instances the hazard ratios remained stable and all


 11   were statistically significant, attesting to the


 12   robustness of the observation.  The most common


 13   concerns regarding progression-free survival


 14   analyses include the impact of scheduled assessment


 15   and missing data which can potentially be a source


 16   of bias.  Several of the methods used by Genta


 17   address these issues and all confirm the conclusion


 18   derived from the original planned analysis.


 19             [Slide]


 20             FDA has performed four analyses using


 21   interval censoring techniques.  Hazard ratios are


 22   not reported for this method.  Approach number one




  1   specifically addresses the issue of assessment


  2   schedule bias and remains statistically significant


  3   in favor of Genasense.  Approaches two, three and


  4   four address both assessment schedule and missing


  5   data biases taken together.  Approaches two and


  6   three remain statistically significant in favor of


  7   Genasense.  Only approach four, which represents a


  8   rather extreme case assumption, and I will show you


  9   an example of this on the next slide, resulted in


 10   an insignificant p value and would have resulted in


 11   the deletion of almost half of the data.


 12             [Slide]


 13             Using this example of patient data by


 14   interval censoring technique number four all of the


 15   data in yellow would have been thrown out because


 16   the investigator failed to repeatedly record the


 17   absence of brain metastases.  I would encourage


 18   committee members to address any questions you


 19   might have for the sponsor regarding this analysis


 20   technique to Dr. Janet Wittes.


 21             [Slide]


 22             In order to address FDA concerns about




  1   potential differences for baseline variables to


  2   affect efficacy endpoints, progression-free


  3   survival results and response rates were adjusted


  4   for the variables of age, gender and AJCC LDH


  5   disease site criteria.  Results show that both


  6   hazard ratios and odds ratios remain stable and all


  7   results remain statistically significant.  Thus,


  8   there was no apparent impact of potential baseline


  9   imbalances on results.


 10             [Slide]


 11             An additional concern has been raised


 12   regarding benefit for patients in the United States


 13   when response rates are examined by country.  This


 14   tree plot shows that confidence limits overlap and


 15   point estimates are similar for the United States


 16   and non-United States.  There is, of course,


 17   expected variability in some countries with small


 18   sample sizes but no evidence exists that the


 19   beneficial effect of the Genasense combination is


 20   different in the United States than it is outside


 21   the United States.


 22             [Slide]




  1             In summary, we have demonstrated


  2   radiographic concordance and superiority of


  3   Genasense regardless of who reviews the x-rays.


  4   Progression-free survival was not biased by missing


  5   data or interval assessment irregularities.  No


  6   effect on endpoints was observed related to


  7   baseline demographic variables and similar benefit


  8   was observed for both U.S. and non-U.S. patients on


  9   the study.


 10             [Slide]


 11             Turning now to safety, adverse events were


 12   generally increased in the Genasense arm, as can be


 13   expected with add-on therapy.  The committee is


 14   referred to the briefing document provided by the


 15   sponsor for details of adverse events.


 16   Importantly, no new or unexpected adverse events


 17   were observed in the study which have not been seen


 18   with DTIC alone.  We did see an increase in the


 19   incidence of fever, which is a well-known effect


 20   related to Genasense as a single agent, as well as


 21   an increase in neutropenia, thrombocytopenia and


 22   catheter-related complications.  Safety data were




  1   regularly and carefully monitored by an independent


  2   drug safety monitoring board who at no point


  3   identified any safety concerns in the study.


  4             [Slide]


  5             There is an increased incidence of grade


  6   3-4, as well as serious events of thrombocytopenia


  7   in the Genasense arm.  The word "serious" in this


  8   context is defined in its regulatory context and


  9   generally means the need for hospitalization or the


 10   prolongation of hospitalization.  However,


 11   bleeding, which is the major clinical consequence


 12   of this laboratory abnormality with grade 3-4


 13   bleeding, serious bleeding--serious bleeding


 14   related to thrombocytopenia, shows no difference


 15   between the arms.  Similarly, the number of


 16   patients who required platelet transfusions with


 17   the absolute number of units transfused were no


 18   different between the two treatment arms.


 19             [Slide]


 20             Neutropenia exhibited a similar pattern as


 21   thrombocytopenia.  The incidence of grade 3-4 and


 22   serious events was increased in the Genasense arm. 




  1   Although higher in the Genasense arm and largely


  2   related to the presence of a central line, the


  3   incidence of grade 3-4 and serious neutropenic


  4   infections was generally low in both groups.


  5             [Slide]


  6             Not surprisingly, catheter-related


  7   complications occurred almost solely in the


  8   Genasense arm and the incidence was consistent to


  9   that reported in the literature for central venous


 10   catheters.  Injection site infections occurred in


 11   approximately 4 percent of patients and thrombotic


 12   events occurred in approximately 2 percent of


 13   patients receiving Genasense, whereas injection


 14   site reactions occurred only in the DTIC group


 15   where peripheral lines are generally used for DTIC


 16   administration.  Two patients in the Genasense arm


 17   received their 5-day Genasense dose in 5 hours due


 18   to a mis-programming of the pump.  Both of these


 19   patients experienced nausea, fever and


 20   thrombocytopenia.  Both patients recovered


 21   completely within 48 hours and had no sequelae


 22   related to the overdose.  Both patients went on to




  1   receive the additional cycles of therapy and one of


  2   these patients has achieved a PR after 7 additional


  3   cycles of treatment.  We are hopeful that


  4   subcutaneous and other alternative dosing methods


  5   in development will mitigate the need for a central


  6   line and its attendant complications.


  7             [Slide]


  8             Adverse events leading to discontinuation


  9   were increased in the Genasense arm.  However, the


 10   majority of events in both arms were related to


 11   disease progression.  In this study disease


 12   progression could be reported as an adverse event.


 13   Importantly, adverse events resulting in death and


 14   deaths which occurred within 30 days of the last


 15   dose of study drug were no different between the


 16   two treatment arms.


 17             [Slide]


 18             In summary, this study was the largest


 19   randomized trial ever completed in patients with


 20   advanced malignant melanoma.  The study was


 21   carefully conducted; showed internally consistent


 22   results; and demonstrated compelling clinical




  1   benefit.


  2             We believe that we have addressed all of


  3   the study questions given to ODAC for


  4   consideration.  Finally, we believe that the study


  5   shows consistent clinical benefit, which will be


  6   summarized by Dr. Frank Haluska in his closing


  7   remarks.


  8             In closing, I would like to thank the


  9   patients and their families, the physicians, the


 10   nurses and the site coordinators who made the study


 11   possible.  I would also like to thank the dedicated


 12   and professional employees of Genta who worked


 13   tirelessly to contribute to the treatment of cancer


 14   patients.  Thank you for your attention.  Dr.


 15   Haluska?


 16                     Clinical Benefit Summary


 17             DR. HALUSKA:  Thank you, Dr. Itri.


 18             [Slide]


 19             My task today is to provide you with a


 20   summary of the data that you have just seen, that I


 21   think have been so clearly presented, as well as an


 22   overview and some context for the clinical trial.




  1             [Slide]


  2             I think the best way to do this is to in


  3   our minds assume the role of ODAC and if I were a


  4   member of ODAC right now I would have two major


  5   questions.  The first of these is that the sponsor


  6   here has failed to meet the primary endpoint of the


  7   study, which is survival--can I still approve this


  8   drug?  I think the answer to that question is an


  9   emphatic yes.  Dr. Pazdur has already commented


 10   that although meeting a survival endpoint is


 11   desirable and is the gold standard, the failure to


 12   do so does not preclude approval, and I think that


 13   is germane here.


 14             I addition, I think it is important to


 15   consider the recent regulatory history of the


 16   melanoma field, specifically with regard to IL-2


 17   and temozolomide.  IL-2, as you know, was approved


 18   several years ago based on the rate, the quality


 19   and the duration of the responses, data that we are


 20   presenting here, and I think these data are


 21   stronger because they are the result of a


 22   randomized, prospective trial, albeit with




  1   secondary endpoints.


  2             The other drug that I think is relevant is


  3   temozolomide and, as Dr. Kirkwood has already


  4   explained, the data are better for Genta than for


  5   the temozolomide submission as well.  So, I think


  6   that this drug is approvable despite the failure to


  7   meet the primary endpoint.


  8             The second question that must be on your


  9   mind is do the secondary endpoints confer or


 10   support the conferral of clinical benefit?  Are


 11   they strong enough to support approval of this


 12   drug?  I do think that significant clinical benefit


 13   is strongly suggested by these data.  So, let's


 14   consider that.


 15             [Slide]


 16             These are I think the most important


 17   endpoints of this study.  Again, I want to stress


 18   that they were prospectively identified as opposed


 19   to, for instance, IL-2s which were the result of


 20   Phase 2 data.


 21             The first of them is the overall response


 22   rate.  The overall response rate approaches 12




  1   percent versus 6.8 percent in the DTIC arm.  This


  2   is an improvement.  In this field, no improvement


  3   with statistical significance has ever been


  4   demonstrated in response rate for advanced


  5   melanoma.


  6             We have demonstrated improvement in


  7   complete responses, 11 versus 2.  This is


  8   significant as well and, again, this has not been


  9   demonstrated in a reaction study.  I think the IL-2


 10   experience is relevant to both of these.  As I


 11   said, IL-2 was approved on the basis of the rate,


 12   the quality and the duration of survival.  We have,


 13   in this trial, 9 patients that are alive, an


 14   increment that is not seen in the DTIC trial, and I


 15   want to point out that IL-2 was approved on the


 16   basis of 10.  So, this is certainly in keeping with


 17   previous decisions that have been made.


 18             The final issue is progression-free


 19   survival, 74 versus 49 days, nearly an additional


 20   month for patients who are presenting to their


 21   oncologist.  That is an extra visit a patient can


 22   come to their oncologist without having been told




  1   that their disease is progressing.  This, to my


  2   mind, is clinical benefit.


  3             [Slide]


  4             What is the context of these findings?


  5   These are the data from the five largest randomized


  6   trials that have been conducted in melanoma and the


  7   trial in front of you today is the largest.  There


  8   are 2019 patients that have been treated on these


  9   trials and until today there has never been a


 10   significant clinical improvement for any of the


 11   measures that we are discussing today.  Response


 12   rate has not been shown to be improved and it is


 13   shown to be improved here.  Complete responses have


 14   never been documented in a randomized study to be


 15   improved and they are improved here.  And,


 16   progression-free survival has never been shown to


 17   be improved and it is improved here.  I think this


 18   trial sets itself apart from the progress in the


 19   field in the last few years and I think that is why


 20   it requires your careful consideration today.


 21             [Slide]


 22             To summarize that, patients value




  1   responses and value complete responses.  The FDA in


  2   the past has made it clear that these are important


  3   criteria to consider and, in fact, there are no


  4   melanoma drugs approved that have been approved on


  5   any other criteria.


  6             You might ask is a 10 percent response


  7   rate, or the order of magnitude of 10 percent,


  8   important to patients and I think it is with, I


  9   think, the recent approval history and data on


 10   responses in other malignancies, particularly in


 11   lung cancer.  The IRESSA experience that has


 12   recently been clarified with data published last


 13   week suggests that a 10 percent response rate is


 14   clinically important.  We understand the biological


 15   basis of some of these responses and a 10 percent


 16   response rate can certainly change the field; it


 17   can certainly change a patient's life.  So, I do


 18   not think that a 10 percent response rate in and of


 19   itself argues against approval.


 20             What about the magnitude of time to


 21   progression?  A month, I think, is important.  Data


 22   that Carey Kilbridge and my colleagues have




  1   examined with regard to how melanoma patients view


  2   their experience strongly suggest that any


  3   additional time without being told their disease is


  4   progressing or without the presence of disease is


  5   important to them.  In my opinion, what the


  6   sponsors have shown today constitutes clinical


  7   benefit for the melanoma patient.


  8             [Slide]


  9             What about safety?  When we research a


 10   treatment for our patients we do it based on an


 11   evaluation of risk versus benefit.  What are the


 12   risks of this therapy?  The sponsor has shown that


 13   there are no new or unexpected adverse events


 14   concomitant to treatment with DTIC and Genasense.


 15   There is no difference in the treatment-related


 16   deaths between the two arms.  There is an increase


 17   in fever, neutropenia and thrombocytopenia.  Some


 18   of this is likely due to catheter-related


 19   complications and this is certainly not the only


 20   agent on the market or potentially on the market


 21   that would be administered with a pump.


 22             Finally, Genasense is still better




  1   tolerated than other alternatives for melanoma


  2   patients and, again, I think a review of the


  3   literature is germane here.


  4             [Slide]


  5             These are three of the trials for which we


  6   have good safety data in comparison to the trial in


  7   front of you today.  They demonstrate that the rate


  8   of complications for the DTIC arm is certainly


  9   similar to what was seen in other studies with


 10   regard to grade 3 or 4 neutropenia and grade 3 and


 11   4 thrombocytopenia, and certainly the rates of


 12   complications that can be attributed to the


 13   combination of Genasense and DTIC are less than


 14   what we see with other alternatives for melanoma


 15   patients.  I think that argues that this is a safe


 16   combination and the risk-benefit analysis is


 17   completely reasonable to be attributed to therapy.


 18             [Slide]


 19             Conclusions--I think this is a novel drug.


 20   It is the first of a class of agents that has been


 21   shown to be efficacious by several measures.  It


 22   takes into account our genetic understanding of




  1   this disease.  It is in keeping with the movement


  2   in the field broadly for targeted therapy and I


  3   think that should be taken into consideration.


  4             It confers a clinical benefit with DTIC by


  5   multiple measures that I think have been reliably


  6   demonstrated in this large clinical trial that


  7   include response rate, complete responses and


  8   progression-free survival.  And, it has a


  9   predictable and manageable safety profile.


 10             [Slide]


 11             Melanoma is refractory to current


 12   front-line therapy.  You have heard and I think you


 13   will hear further today that we need new agents.


 14   This product is safe; it is effective when combined


 15   with DTIC to treat stage 4 melanoma.  In other


 16   words, this drug works.  I think it is up to you to


 17   define today what "works" means but I don't think


 18   we can discard the randomized trial demonstrated


 19   improvement in response rate, in progression-free


 20   survival and in complete response rate.


 21             A final comment--I am supposed to be here


 22   as a dispassionate expert, scientifically objective




  1   and clinically removed but I don't think I can


  2   completely play that role because I do take care of


  3   melanoma patients.  The melanoma field has been


  4   criticized for trying to consistently hit the


  5   clinical home run.  But this represents progress.


  6   It is incremental progress.  It is not a clinical


  7   home run but it is incremental progress, and if we


  8   are ultimately going to make real progress in this


  9   disease to cure it, it will require the


 10   accumulation of incremental progress.  Allow us to


 11   make incremental progress; make this drug available


 12   to our patients.  Thank you.


 13             DR. PRZEPIORKA:  We are going to hold


 14   questions for the first presentation until the FDA


 15   presentation has been completed.  Dr. Kane, if you


 16   could begin?  Thank you.


 17                         FDA Presentation


 18                          Medical Review


 19             DR. KANE:  Thank you.


 20             [Slide]


 21             Good morning.  My name is Robert Kane.  I


 22   am the medical reviewer for this NDA and I will be




  1   presenting the FDA review along with Dr. Peiling


  2   Yang, our statistical reviewer.


  3             [Slide]


  4             I would like to recognize our primary


  5   review team members for this NDA.


  6             [Slide]


  7             Randomized, controlled trials


  8   prospectively designed with clear, quantitative


  9   endpoints statistically analyzed provide the basis


 10   to assess the merits of new drugs.  Clinical


 11   judgment translates these findings for best patient


 12   care.  Our presentation today will include


 13   requirements for new drug approval based on federal


 14   law and regulations; aspects of ODAC review of


 15   temozolomide which are relevant to today; the FDA


 16   examination of the Genasense, oblimersen, NDA; and


 17   concluding remarks.


 18             [Slide]


 19             In the FD&C Act of 1962 substantial


 20   evidence of effectiveness was required by Congress.


 21   This was defined as evidence from adequate and


 22   well-controlled investigations, generally




  1   understood to mean at least two such studies for


  2   new drug approval.


  3             [Slide]


  4             The FDAMA legislation in 1997 indicated


  5   that one trial may suffice for approval with


  6   confirmatory evidence.  The guidance document on


  7   effectiveness in 1998 indicated that for a single


  8   trial to suffice it should be of excellent design,


  9   internally consistent with highly reliable and


 10   statistically strong evidence of an important


 11   clinical benefit, such as an effect on survival,


 12   and a confirmatory study might be difficult to do


 13   for ethical reasons.


 14             [Slide]


 15             New drug approval can take two forms.  For


 16   regular approval a sponsor needs to show clinical


 17   benefit.  Accelerated approval uses a surrogate


 18   endpoint reasonably likely to predict clinical


 19   benefit and requires subsequent confirmation of the


 20   benefit.


 21             [Slide]


 22             Here are the currently approved drugs for




  1   metastatic melanoma.  In the past response rate was


  2   the primary basis, as you have seen and as you have


  3   already heard, for hydroxyurea and for dacarbazine.


  4   Survival times were, and continue to remain, in the


  5   range of 5 to 9 months.  More recently,


  6   improvements in the quantity or the quality of


  7   survival have served as the basis for approval.


  8   Also as you have heard, the aldesleukin,


  9   interleukin-2, approval was heavily related to the


 10   very long complete responders, some in excess of 5


 11   years.  Complete responses will be abbreviated as


 12   CRs on this slide.


 13             [Slide]


 14             I would like to remind the committee that


 15   the evidence for interferon supported approval for


 16   its adjuvant use although it is often used in the


 17   treatment for metastatic disease.  The temozolomide


 18   evaluation by ODAC in 1999 is relevant and


 19   instructive for today's review.


 20             [Slide]


 21             This NDA contained one main open-label


 22   study, the primary endpoint of which was survival




  1   time.  It was designed to show a 3-month survival


  2   benefit for temozolomide alone over DTIC alone.


  3   Secondary endpoints were progression-free survival,


  4   abbreviated here as PFS, and response rate, RR.


  5             [Slide]


  6             The results of this study showed no


  7   survival benefit for temozolomide over DTIC.


  8   Median survivals were 7.7 versus 6.4 months.  For


  9   progression-free survival the difference was found


 10   to be highly statistically significant with a


 11   log-rank p value of 0.002.  However, the median


 12   progression-free survival difference was only 11


 13   days.  When an ample size is chosen for a survival


 14   endpoint the statistical significance of small


 15   differences in early endpoints can appear


 16   magnified.  Response rates were not significantly


 17   different.


 18             [Slide]


 19             Temozolomide was not approved.  The study


 20   failed to demonstrate the primary endpoint of


 21   survival benefit.  Progression-free survival, a


 22   secondary endpoint, was of small magnitude at best.




  1   No symptomatic benefit was observed and a proposed


  2   post hoc 6-month survival analysis was not


  3   convincing.


  4             [Slide]


  5             For Genta's NDA, here are the important


  6   study dates.  The Phase 3 protocol began in July,


  7   2000.  The data cutoff date was August 1, 2003, and


  8   this represents excellent accrual to the study.  On


  9   December 8, 2003 the NDA was submitted for FDA


 10   review.


 11             [Slide]


 12             Genta has just presented their trial


 13   design.  I would like to emphasize a couple of


 14   points.  This was a very large, multicenter,


 15   multinational, unblinded study.  This was an add-on


 16   of Genasense to DTIC.  Prolonged central venous


 17   access is required for the 5-day infusions of


 18   Genasense.  Genasense may be abbreviated as G or


 19   G3139 on our slides.  The protocol specified an


 20   independent review, a blinded group, to assess


 21   responders.  Also, the ability to deal with an


 22   ambulatory infusion pump was required.




  1             [Slide]


  2             The primary endpoint was survival.  The


  3   design was to detect a superiority in survival.


  4   The protocol included seven secondary endpoints,


  5   listed here.


  6             [Slide]


  7             The trial design was to identify a 2-month


  8   median improvement in survival time from 6 months


  9   with DTIC alone to 8 months for the addition of


 10   Genasense to DTIC.  The primary analysis for the


 11   trial was to be the unadjusted log-rank analysis


 12   for the intent-to-treat population.


 13             [Slide]


 14             The study disposition of patients showed


 15   that less than half the patients were still on


 16   therapy after the first assessment about day 42.


 17   Most patients went off study because of progressive


 18   disease; 44 percent remained on study after the


 19   first assessment.  As I mentioned, the data cutoff


 20   date was August 1 and analysis occurred at 535


 21   deaths.


 22             [Slide]




  1             In the primary endpoint analysis, using


  2   the protocol-specified analysis with the


  3   intent-to-treat population, no survival benefit was


  4   demonstrated by adding Genasense to DTIC treatment


  5   versus DTIC alone.  These are the actual survival


  6   results.  As you have already seen, the hazard


  7   ratio was 0.89 and the log rang p value for the


  8   survival difference was 0.18.


  9             Dr. Peiling Yang will now provide a more


 10   detailed examination of the progression-free


 11   survival.


 12                        Statistical Review


 13             DR. YANG:  Thank you, Dr. Kane.


 14             [Slide]


 15             As seen in Dr. Kane's presentation, the


 16   study failed to demonstrate efficacy in the primary


 17   endpoint of overall survival at a two-sided alpha


 18   level of 0.05.  From a statistical perspective, an


 19   efficacy demonstration based on any other endpoint,


 20   such as progression-free survival, would only infer


 21   a false-positive error rate.  Despite this concern,


 22   the secondary endpoint, progression-free survival,




  1   was evaluated and the important question is


  2   regarding progression-free survival.


  3             [Slide]


  4             We have doubt regarding the applicant's


  5   findings and, second, as Dr. Kane will be


  6   discussing, there are questions regarding its


  7   clinical significance.  This will be summarized in


  8   this presentation.


  9             [Slide]


 10             My review of the progression-free survival


 11   is as follows, review of applicant's analyses and


 12   results; then the major FDA concern about


 13   assessment times; then additional FDA concerns.


 14             Let's first review the applicant's


 15   analysis and results.  Progression-free survival


 16   was defined as time from the data of randomization


 17   to the date of disease progression or death.  The


 18   data of disease progression was recorded as the


 19   assessment date when disease progression was


 20   documented.  If the assessment was on different


 21   days, then the latest date among all assessments


 22   was used by this applicant to represent the




  1   assessment date in that cycle.


  2             [Slide]


  3             This slide summarizes the applicant's


  4   results.  The protocol specified as secondary


  5   efficacy analysis or progression-free survival was


  6   the log-rank test with the missing data imputed by


  7   the last observation carried forward method.  The p


  8   value based on this approach was very small.


  9   However, in a large trial a small p value can be


 10   observed even if the treatment effect is small.


 11   During the review process FDA requested the


 12   applicant to analyze the data using a different


 13   approach by censoring patients at the last


 14   assessment date when at least 50 percent of target


 15   lesions were measured if the disease had not


 16   progressed yet.  The p value based on this approach


 17   was also very small.  However, when analyzed by


 18   this approach the observed median progression-free


 19   survival in the combination therapy dropped by 13


 20   days and in the control arm dropped by only 1 day,


 21   as presented in this table.


 22             [Slide]




  1             An important question is raised while


  2   interpreting the results of the analysis of


  3   progression-free survival.  Is the applicant's


  4   finding a true finding?


  5             [Slide]


  6             FDA has a major concern in evaluation of


  7   progression-free survival, that is, imbalance in


  8   observed lesion assessment times between treatment


  9   arms.  The next few slides address this concern.


 10             [Slide]


 11             Lesions were to be measured every 6 weeks


 12   during the treatment phase.  In practice, this did


 13   not always occur.  Even when they were assessing


 14   the planned cycles there were still differences in


 15   timing between the two arms.  Because this is a


 16   very large open-label trial involving two different


 17   regimens, one administered on 6 days and the other


 18   only 1 day and because the claimed difference was


 19   very small, FDA was concerned that the observed


 20   differences in progression-free survival might be


 21   affected by systematic bias.  One potential bias


 22   could be caused by differences in the time of




  1   lesion assessments.


  2             [Slide]


  3             We must remember a critical difference


  4   between the analysis of survival and of lesion


  5   progression.  The date of death, represented by the


  6   star, will not change regardless of the evaluation


  7   schedule.  With progression measurement, however,


  8   the date we assign for progression is usually the


  9   date of a scheduled visit occurring sometime after


 10   the actual progression date.  It should not be


 11   surprising that assessing progression at longer


 12   intervals leads to a longer time to progression.


 13             [Slide]


 14             To address this concern FDA summarized the


 15   time from the date of randomization to each of the


 16   first 3 observed assessments in this pivotal trial.


 17   Included in this summary are those assessments


 18   which occurred by the time of disease progression


 19   or death and where there was at least one target


 20   lesion measurement.  The observed median times from


 21   randomization to each of these assessments were


 22   obtained for each treatment arm.  They were 48




  1   versus 43 days to the first assessment; 94 versus


  2   87 days to the second assessment; and 137 versus


  3   129 days to the third assessment.  The p values for


  4   the log-rank test comparing the entire curves were


  5   also obtained for each assessment.  Note that the


  6   difference in timing of lesion assessments shows


  7   striking statistical significance, with p values of


  8   the same order of magnitude as the claimed


  9   difference in progression-free survival.  This


 10   finding raises a concern that all or some of the


 11   observed progression-free survival difference were


 12   caused by this systematic bias in lesion assessment


 13   times.


 14             [Slide]


 15             These are the times to the first


 16   assessment curves.  Please note that these are not


 17   time to disease progression curves.  The blue curve


 18   represents the combination therapy and the red one


 19   represents DTIC alone.  On the horizontal axis we


 20   have the time from randomization to the first


 21   assessment in days.  On the vertical axis we have


 22   the proportion of patients who had the first




  1   assessment later at a given time.  As seen here,


  2   the blue curve stayed above the red curve all


  3   along, suggesting a systematic delay in the first


  4   assessment time in the combination treatment arm.


  5             [Slide]


  6             Similar patterns were observed in the time


  7   to the second assessment curves.


  8             [Slide]


  9             And to the third assessment curves.


 10             [Slide]


 11             Imbalance in assessment times may have


 12   impact in several ways on the analysis of


 13   progression-free survival.  The first impact is


 14   that bias may be introduced in estimating


 15   progression-free survival.  Second, with a large


 16   trial even a small imbalance between treatment arms


 17   may lead to incorrect conclusions.


 18             [Slide]


 19             This slide illustrates the first impact.


 20   A hypothetical example is given here to illustrate


 21   how imbalance may be introduced in estimating


 22   progression-free survival.  In this example,




  1   suppose that the actual day of disease progression


  2   was day 35 post randomization for both patients,


  3   one in the control arm and the other in the


  4   experimental arm.  However, the first assessment


  5   for the patient in the control arm was on day 42


  6   and for the patient in the experimental arm it was


  7   on day 48.  The recorded days of disease-free


  8   progression will be on days 42 and 48 respectively.


  9   These recorded days, not day 35, will be the


 10   observations used in the analysis.


 11             [Slide]


 12             This slide illustrates the impact of


 13   systematic bias by a simulation study.  In the


 14   simulation study progression-free survival was


 15   generated from identical distribution in both arms


 16   with a median of 50 days and 300 subjects in each


 17   arm.  However, a systematic increase by 2 days in


 18   assessment times in one arm was introduced.  In 98


 19   percent of the 5000 simulations p values were less


 20   than 0.05.  This illustrates that even with a small


 21   imbalance in assessment times between two arms the


 22   chance of falsely concluding treatment effect can




  1   be very high when, in fact, there is no treatment


  2   effect at all, also the chance of incorrectly


  3   concluding increases as the sample size increases.


  4             [Slide]


  5             An additional FDA concern is about missing


  6   data.  Missing data was observed in both treatment


  7   arms, especially for non-target lesions which also


  8   had an influence on the determination of disease


  9   progression.  In this study lesion assessments were


 10   not always performed in planned cycles.  Also,


 11   lesions were assessed at baseline or assessed post


 12   baseline.  In the presence of missing data bias


 13   could be introduced in estimating treatment


 14   effects, especially in an open-label study as this


 15   is.  This is a common problem in assessing


 16   progression in most of the studies.


 17             [Slide]


 18             This slide summarizes the progression-free


 19   survival findings.  The claimed progression-free


 20   survival benefit in the combination therapy over


 21   DTIC alone may not be a true finding because of


 22   imbalance in assessment times between treatment




  1   arms.  The true progression-free survival benefit


  2   of the combination therapy over DTIC therapy alone


  3   was confounded by imbalance in assessment times


  4   between treatment arms.  Thus, true treatment


  5   effect with respect to progression-free survival


  6   cannot be isolated.  The chance of falsely


  7   inferring progression-free survival benefit could


  8   be high.  Even if there was, indeed, no benefit, it


  9   will be magnified by increasing the sample size.


 10   Missing data is always a concern in oncology


 11   studies evaluating progression as an endpoint.  The


 12   confidence in the amount of difference in


 13   progression-free survival is diminished in the


 14   presence of missing data and may allow introduction


 15   of bias, especially in an open-label study.


 16             [Slide]


 17             Finally from a statistical perspective,


 18   this large randomized, open-label study failed to


 19   demonstrate the protocol specified primary efficacy


 20   based on the overall survival benefit with respect


 21   to the secondary efficacy analysis of


 22   progression-free survival because of systematic




  1   bias in ascertainment.  It is not clear whether the


  2   benefit of progression-free survival in the


  3   combination therapy over DTIC alone exists.  If it


  4   exists, the magnitude is uncertain.  Also, there


  5   are multiplicity issues with analyses conducted to


  6   support the efficacy.  Dr. Kane will address the


  7   clinical relevance.


  8                        Clinical Relevance


  9             DR. KANE:  Dr. Yang has provided a


 10   detailed assessment of some of the concerns related


 11   to progression-free survival.


 12             [Slide]


 13             To summarize these concerns, assessments


 14   in this study were done at 6-week intervals.  The


 15   progression-free survival difference, however, was


 16   only in the range of 2-3 weeks.  The


 17   progression-free survival difference is highly


 18   statistically significant but may be fully


 19   accounted for by asymmetry in the timing of


 20   assessments between the two arms.  The magnitude of


 21   the effect size is uncertain.  The real problem is


 22   what is the clinical relevance.




  1             [Slide]


  2             The Division examined all of the secondary


  3   endpoints of the protocol for the possibility of


  4   patient benefit, given the fact that the overall


  5   survival analysis failed.


  6             [Slide]


  7             We will next look at the response rates


  8   among the secondary endpoints.  The data submitted


  9   at the time of the original NDA submission and


 10   analysis, as has been presented here, indicated


 11   that the Genta investigator-determined responses


 12   were derived from an algorithm using tumor


 13   measurements from the case report forms.  In that


 14   examination, 11.7 percent of patients were reported


 15   as responders to the combination versus 6.8 percent


 16   with DTIC alone.  The p value for this difference


 17   was 0.018 and the actual difference was just under


 18   5 percent.


 19             The study protocol also called for a


 20   blinded independent review and confirmation for all


 21   responders.  The protocol stated that all


 22   radiographs, as well as photographs of cutaneous




  1   lesions, were to be provided to this review group.


  2   The blinded independent reviewers, as you have


  3   heard, reported different response rates, 6.7


  4   percent response for the combination versus 3.6


  5   percent for DTIC alone, a difference of 3.1 percent


  6   and of borderline significance.  Ordinarily,


  7   adjudication by an independent review is considered


  8   to be the definitive response rate.


  9             [Slide]


 10             Some of this discordance may be due to


 11   technical difficulties, such as providing the


 12   independent review group with the appropriate


 13   images.  However, we must point out that 5 complete


 14   responses, which constituted all of the responses


 15   in the initial NDA submission identified by the


 16   Genta site investigators--there were 3 in the


 17   combination arm and 2 in the DTIC alone arm.  None


 18   was adjudicated as complete responses by the


 19   independent review.  Forty-four percent of the


 20   responders by the Genta site investigators were


 21   determined as not assessable or unconfirmed by the


 22   independent review.  For 49 percent there was full




  1   concordance for the response category between Genta


  2   and the independent review.


  3             [Slide]


  4             You have also heard that on April 9th--a


  5   couple of weeks ago--Genta provided new data on


  6   responders.  This new data is being examined.


  7   There are problems with data that is developed


  8   outside of the study protocol.  There can be


  9   ascertainment bias between arms when an analysis is


 10   not prospectively planned.  Subsequent therapies,


 11   such as surgery not being part of the protocol


 12   treatment, may not be applied symmetrically.


 13             [Slide]


 14             Turning to duration of response, another


 15   secondary endpoint, this is Genta's analysis.  This


 16   data is skewed data and, therefore, we refer to the


 17   median to describe it and the medians are quite


 18   similar.


 19             [Slide]


 20             For durable response rate Genta has


 21   provided this analysis.  This was a prespecified


 22   secondary endpoint.  The difference was not




  1   significant.


  2             [Slide]


  3             Performance status is a measure of


  4   functional capacity.  There were no differences in


  5   performance status observed between study arms to


  6   suggest a benefit for adding Genasense to the DTIC.


  7             [Slide]


  8             For tumor-related symptoms, there were no


  9   differences in symptoms observed between study arms


 10   during the treatment.


 11             [Slide]


 12             This slide introduces the adverse events


 13   which represent the toxicity safety endpoint for


 14   the study.  You have heard from Dr. Itri that the


 15   grade 3-4 adverse events, the serious adverse


 16   events, and the adverse events leading to


 17   discontinuation all were increased with the


 18   addition of Genasense to DTIC.  Since the DTIC


 19   doses were the same, the increased toxicity is


 20   likely due to the Genasense.


 21             [Slide]


 22             This represents the hematologic toxicity




  1   which you have already heard.  There was more grade


  2   3-4 neutropenia and thrombocytopenia on the


  3   combination arm.


  4             [Slide]


  5             For non-hematologic toxicity, all adverse


  6   events were more frequent on the combination arm


  7   with the addition of Genasense.


  8             [Slide]


  9             In total, there were 18 patients with


 10   upper extremity thrombosis on the combination arm


 11   compared to 3 on the DTIC alone arm.


 12             [Slide]


 13             In summary, the Genasense trial failed to


 14   achieve its primary protocol-specified endpoint.


 15   No survival benefit was demonstrated with the


 16   addition of Genasense to DTIC compared to DTIC


 17   alone.  The efficacy of the control arm, DTIC


 18   alone, is consistent with that of other studies.


 19             [Slide]


 20             Looking again at the secondary endpoints,


 21   these are usually considered to be exploratory and


 22   for progression-free survival there is no precedent




  1   for progression-free survival as evidence of


  2   clinical benefit for metastatic melanoma.  This may


  3   not be a true finding.  The progression-free


  4   survival difference between the two arms may be 13


  5   or 25 days depending on which censoring technique


  6   is chosen for missing data.  The clinical relevance


  7   is uncertain.


  8             [Slide]


  9             For response rate, the difference from


 10   DTIC alone may be in the range of 3-5 percent.  No


 11   complete responses in the original NDA submission


 12   were confirmed by the independent blinded review


 13   committee.  The clinical relevance of this result


 14   is uncertain.  Thus far, response rates in these


 15   ranges have not conferred survival benefits for


 16   metastatic melanoma.  For the durable response


 17   rate, no significant difference.  Response


 18   durations were practically identical.


 19             [Slide]


 20             For performance status no benefit was


 21   observed from the addition of Genasense to DTIC


 22   over DTIC alone.  Symptomatic benefit was no




  1   different.  There is greater toxicity with the


  2   Genasense combination than for DTIC alone.  Thank


  3   you.


  4                   Questions from the Committee


  5             DR. PRZEPIORKA:  Thank you for the review.


  6   We are now going to open the session for questions


  7   to either the sponsor or to the FDA.  Dr. Cheson?


  8             DR. CHESON:  I am sure the 11 or so


  9   patients out there still in remission will be


 10   disturbed to know that modeling suggests that they


 11   shouldn't be there.  We have heard some difficult,


 12   complicated analyses of modeling suggesting that


 13   what we heard from the elegant presentation from


 14   Dr. Itri and her co-workers might not be as


 15   clinically relevant.  So, we have one side


 16   suggesting one set of outcomes showing clinical


 17   benefit, then the computer modeling and the FDA


 18   suggesting perhaps that these are not reliable.  I


 19   would like to hear from the company, from Dr.


 20   Wittes, their side of this spin.


 21             DR. WITTES:  The issue about the potential


 22   for bias that can come from interval censoring and




  1   from missing data we knew about and, in fact,


  2   looked at--I need the slide, yes, that is the one.


  3             [Slide]


  4             In fact, that is why we did some of the


  5   sensitivity analyses.  These sensitivity analyses


  6   look at three different kinds of things, the


  7   missing data and the interval censoring, and the


  8   last three are the ones that look at interval


  9   censoring, the by-cycle analysis, the assumed


 10   progressive disease, back to the scheduled


 11   visit--these are three different ways of trying to


 12   adjust for the interval censoring.  What you see is


 13   some changes in hazard ratio but quite similar to


 14   what they were before and then statistically


 15   significant p values.


 16             [Slide]


 17             Next slide, CC49--the FDA's approach for


 18   interval censoring, which is a method due to


 19   Michael Fay, is a non-parametric approach.  It is a


 20   score statistic and, again, the p value remains


 21   statistically significant.  So, yes, there


 22   certainly is a differential time to measurement in




  1   the two groups but analyses that adjust for that


  2   time still show a statistically significant


  3   benefit.


  4             DR. PRZEPIORKA:  Dr. D'Agostino?


  5             DR. D'AGOSTINO:  Janet, the procedure the


  6   FDA used is not unreasonable.  I am asking a


  7   question but it is a set of assumptions that could,


  8   in fact, underlie some of the differences we see,


  9   and I guess the point that the FDA was making, I


 10   thought, was that you could chip away at these


 11   differences not only in statistical significance


 12   but magnitude of difference, clinical difference,


 13   and that I think should be taken into account with


 14   the interpretation of these techniques.


 15             DR. WITTES:  I agree, Ralph, but can we go


 16   back to that 49?


 17             [Slide]


 18             Here is the chipping away.  I mean, the


 19   chipping away is to look at both the interval


 20   censoring and the missing data.  I think if you


 21   approach four, which is the one that is most


 22   chipped, if you look at what that does, it is the




  1   Michael Fay approach to interval censoring plus a


  2   very conservative method for missing data, and let


  3   me describe that a little bit because I think it is


  4   important to know what happens here.


  5             There are basically three kinds of missing


  6   data.  There are those that Dr. Itri showed where


  7   there is an assessment, it is clear and then you


  8   don't keep on looking at that--the no lesion.  That


  9   is one source.  There is another kind of missing


 10   data where you have an assessment.  At the next


 11   assessment you don't measure that lesion and then


 12   subsequent to that you do measure it and there is


 13   no progression.  So, to me, that isn't really


 14   missing.  If you take away those two and leave the


 15   missing data where you really can't know whether


 16   there is an assessment or not, this method becomes


 17   an 0-3 again.  So, I think if you chip it away you


 18   still get evidence of benefit in progression-free


 19   survival.


 20             The other thing to remember is that from


 21   the point of view of complete responses there is no


 22   issue at all about either interval censoring or




  1   missing data.


  2             DR. PRZEPIORKA:  Dr. D'Agostino?


  3             DR. D'AGOSTINO:  But just again though, we


  4   are left in the dilemma of how do you respond to


  5   the data as collected, as the assessments were made


  6   and so forth, and there is uncertainty in terms of


  7   how comfortable some of us are with the p values.


  8   I think also with a large study you can generate


  9   very large p values with small differences and


 10   maybe some of that is here also.  Again, p values


 11   are important but there is clinical significance


 12   the way these numbers draw closer together by, I


 13   think, relatively comfortable assumptions that is


 14   of concern I think.


 15             DR. WITTES:  I think someone else should


 16   address the clinical significance.


 17             DR. PRZEPIORKA:  Dr. Temple?


 18             DR. TEMPLE:  Janet, one of the things


 19   about 0.003 is that you don't worry about


 20   adjustment for multiplicity and stuff like that.


 21   It kind of blows you away.  But with the smaller p


 22   values that you get from some of the other things




  1   you did that might become an issue.  Do you have a


  2   view as to how one should take into account the


  3   fact that this is not the primary endpoint?  It is


  4   one of at least several things one could have done.


  5   What would you say the right kind of adjustment


  6   would be in a case like that, assuming that some of


  7   the closer to 0.05 p values were the ones that


  8   might count?


  9             DR. WITTES:  Yes, I don't know the answer


 10   to that.  I mean, if the question is what is the


 11   type-1 error of this study, I think one can't


 12   really answer that question.  Of course, one looks


 13   at consistency.  One worries about the potential


 14   for bias and, again, I feel that those complete


 15   responses kind of avoid--they become a different


 16   kind of criterion.  But if you ask me what is the


 17   type-1 error rate, I don't know.


 18             DR. PRZEPIORKA:  Dr. D'Agostino?


 19             DR. D'AGOSTINO:  Just again, when you look


 20   at the secondary endpoints after you have a failure


 21   in the primary endpoint, the whole


 22   interpretation--just to reinforce what you just




  1   said, no one around this table is going to be able


  2   to put a real p value on any of these things that


  3   we have given that the primary didn't turn out to


  4   be statistically significant.


  5             DR. PRZEPIORKA:  Any other questions from


  6   the committee?  Dr. Hwu?


  7             DR. HWU:  I have a question for Dr. Itri


  8   regarding the design of this trial, especially the


  9   regimen used in this large trial for the


 10   experimental arm.  The initial scientific


 11   indication of this incremental improvement in the


 12   treatment of melanoma was based on the Phase 1 and


 13   2 trial, which was published in Lancet by Jansen


 14   and colleagues in 2000.  The Phase 1 and 2 trial


 15   design was extremely careful.  They screened the


 16   patients who had shown in tissue increased


 17   expression of Bcl-2.  Also, the pharmacokinetic


 18   study was done very carefully and was a clinical


 19   correlate of the tissues at the level of decrease


 20   of Bcl-2 expression.  Also, there is correlation


 21   with responses.


 22             The regimen used in that trial was very,




  1   very reasonable in design.  They were giving


  2   infusion on day 1 to day 14, continuous infusion.


  3   Clearly by day 5 the Bcl-2 expression was maximally


  4   down-regulated.  DTIC was given from day 5 to 9 in


  5   divided doses of 200 mg/m                                                

                       2 every day for 5 days.


  6   In other words, when DTIC is infused in patients,


  7   the G31 and 39 Genasense treatment also continues.


  8             Now, the response was clearly shown in the


  9   M1a group, the patient with skin metastases or


 10   lymph node metastases.  No response was noted in


 11   the lung or visceral organs.  However, the


 12   responses were impressive.  Even one patient who


 13   had prior DTIC had a partial response.


 14             My question to Dr. Itri is why we changed


 15   the protocol which has clearly demonstrated


 16   scientifically that it worked as a target therapy


 17   and now we have changed to 5-day infusion of


 18   Genasense followed by 1 infusion of DTIC and even


 19   forgot that DTIC is not an active chemotherapy


 20   agent by itself; it requires hepatic activation to


 21   its active metabolite MTIC?  We do know that the


 22   company provided a pharmacokinetic study that, yes,




  1   the continuous infusion of Genasense that achieved


  2   the maximal plateau level within 10 hours if you


  3   were giving it at the 7 mg/kg/hour rate--I am


  4   sorry, per kilogram--however, once the infusion


  5   stopped, less than 10 hours later the level for the


  6   Genasense clearly dropped to what we call the


  7   biological active level of I think 1 mcg/L.


  8             So, I would like to know before we launch


  9   this large Phase 3 trial are there any other Phase


 10   2 studies, other than the safety, well-tolerated


 11   5-day infusion by 1 day of DTIC, that have shown


 12   that there is tissue correlation and also efficacy


 13   as shown by the Phase 1 and 2 trial.  Thank you.


 14             DR. WALL:  I am Dr. Ray Wall, from Genta.


 15   Dr. Hwu, I think I will take a whack at those


 16   questions since I was around at the time the study


 17   was done and took it with Dr. Haluska down to FDA,


 18   and Dr. Itri was not.


 19             The Genasense study was informative.  I


 20   would point out to the committee it was a Phase 1


 21   studies that looked at a couple of different doses


 22   of Genasense at that time and also looked at a




  1   couple of different routes of administration, both


  2   subcutaneous administration as well as continuous


  3   IV infusion.  So, it was Phase 1 and it was a total


  4   of 12 patients.  It was published in Lancet in year


  5   2000.


  6             What we had found both in that study and


  7   also in a variety of other studies, some of which


  8   are presented in your briefing book, are a couple


  9   of things with respect to the biological activity


 10   of the drug.  The pharmacokinetics are very well


 11   described and I will skip them for the time being.


 12             What we see in human tumor cells


 13   subsequent to administration of Genasense is that


 14   the onset of the down-regulation of Bcl-2 at the


 15   protein level, not the RNA level but of the protein


 16   level seems to occur at least as early as day 3 and


 17   is maximal at day 5.  The one other thing that had


 18   been a very, very important driver of our clinical


 19   schedule is that the continued administration of


 20   Genasense beyond day 5, if the dose is not changed


 21   you do not seem to get any further down-regulation


 22   of Bcl-2 at the protein level.




  1             I didn't bring a lot of blots in my back


  2   pocket here but I think I can show you one from a


  3   melanoma patient, if I can have MA-25, please?


  4             [Slide]


  5             This is a Phase 1 study looking at a very,


  6   very low dose.  This is a dose that is about 20


  7   percent of our Phase 3 doses, and this is from the


  8   Jansen study looking at continuous infusion over a


  9   14-day period.  Again, you see maximal


 10   down-regulation by about day 5 and, despite the


 11   fact that the infusion is continued, you don't see


 12   any further decrease in the down-regulation of


 13   Bcl-2 protein effect.  These are human tumor cells,


 14   serial biopsies of patients with malignant


 15   melanoma.


 16             So, from these data and from other data


 17   that have been obtained from a variety of other


 18   patients and other cells, both malignant cells as


 19   well as normal cells, that molecular information


 20   has been used to drive the clinical studies,


 21   including the one that you have seen today.


 22             So a couple of things, one is we use




  1   rather short infusions to maximize the


  2   down-regulation of Bcl-2 so that that effect is


  3   maximal at the time that chemotherapy is


  4   administered and we don't continue beyond.  Dr.


  5   Tony Tolcher, who actually is in the audience, has


  6   done some of the best scheduling work but, again,


  7   modeling preclinically, suggesting that when you


  8   administer Genasense with chemotherapy the effect


  9   is maximized when you administer Genasense in


 10   advance of chemotherapy.  The second thing that he


 11   has shown is that there seems to be no advantage to


 12   overlapping Genasense with chemotherapy.  The final


 13   observation from the Tolcher lab is that if you


 14   reverse the sequence, if you give Genasense after


 15   chemotherapy is administered, then you basically


 16   eliminate the synergistic effect.  So, the


 17   constellation of these kinds of pharmacodynamic


 18   events have driven the schedules that you have seen


 19   here today in Phase 3.


 20             DR. PRZEPIORKA:  Before you leave the


 21   podium, just one more question to follow-up, how


 22   long is the effect once the infusion is




  1   discontinued?


  2             DR. WALL:  As was pointed out, the


  3   half-life of this drug is around 3-4 hours and


  4   fundamentally disappears probably by about 10-12


  5   hours.  The data are a little fragmentary and


  6   mostly derived from in vitro cell culture studies,


  7   but it does look like the half-life of Bcl-2


  8   protein is in the order of 16 to about 22 hours.


  9   So, you would expect that if you get complete


 10   shut-down of Bcl-2 production by knocking out the


 11   messenger RNA, then pharmacokinetically within 5


 12   half-lives or so you should have no protein within


 13   the cell, and recovery would be equally as rapid as


 14   soon as it is shut back on.


 15             DR. PRZEPIORKA:  Dr. Temple?


 16             DR. TEMPLE:  Dr. Itri or others, there was


 17   a lot of discussion about the responses.  You


 18   clearly had two different ways of calculating


 19   responses, one based on investigators and the other


 20   based on RadPharm.  My presumption was that the


 21   RadPharm analysis existed because the study was


 22   open and that is a common thing to do, to have a




  1   blinded analysis of the response rates.  In your


  2   presentation though I gather you were disappointed


  3   with what RadPharm produced and you considered it


  4   inaccurate.  Could you clarify the intended role,


  5   what happened and whether you think there ought to


  6   be a further blinded analysis, or what?  This is a


  7   somewhat unusual situation and it wasn't clear what


  8   the original intent was.  As Dr. Kane said, usually


  9   when you have a group like that, they are the


 10   primary analysis.  Was that not true?  Just what


 11   was the arrangement?


 12             DR. ITRI:  That was not true here.


 13             DR. TEMPLE:  Then why did you do it?


 14             DR. ITRI:  The response per statistical


 15   analysis plan was RECIST measurements based on


 16   investigational site measurements that were then


 17   calculated by computer to see whether or not they


 18   met criteria for a partial response or a complete


 19   response.  That is primary and that is what is


 20   reported.


 21             The use of RadPharm--and I think it is


 22   important to note that it was only responding




  1   patients that they looked at so if we were going to


  2   rely on RadPharm to actually give us a response


  3   rate for the study they would have had to review


  4   everyone.  They were really used by us for quality


  5   control purposes.  We wanted to make sure that the


  6   relative numbers we were seeing were consistent


  7   with what has been reported in the literature; that


  8   the concordance rates weren't really out of whack.


  9   I think that the best person to speak about this is


 10   Dr. Ford because he can put this into real context


 11   and explain what the literature shows, and really


 12   how we stack up in terms of other studies that have


 13   utilized a similar review.  Is that okay?


 14             DR. TEMPLE:  Anything is okay, but you


 15   have two somewhat separate, somewhat different


 16   calculations based on the ones that went to them.


 17   Usually that is distressing and I guess the further


 18   question I have is do you have some way of


 19   resolving this?  Should this be subjected to


 20   another blinded review where people get the whole


 21   files, or something?  I mean, as it is, you can see


 22   why it is sort of troublesome.  For example, all of




  1   the complete responses they didn't think were


  2   complete responses although you feel that complete


  3   responses are very important for the reasons Dr.


  4   Cheson mentioned earlier.  That is troublesome, and


  5   now you have found more which we haven't had a


  6   chance to review yet, but the same problem could


  7   arise there too.  So, it does seem important to


  8   figure out what it all means.


  9             DR. ITRI:  I really think you need to talk


 10   to Dr. Ford about this.


 11             DR. TEMPLE:  Whatever you like.


 12             DR. ITRI:  But the other issue is that,


 13   you know, if the agency would like us to submit


 14   these x-rays for review and if that would make you


 15   more comfortable, we would be totally willing to do


 16   that.  We believe that what is being called lack of


 17   concordance really relates to the fact that Dr.


 18   Ford is going to elucidate now.  And, it would not


 19   be a problem; we would be so happy to sit with


 20   anyone and give you the clinical data that supports


 21   this because these are real and the patients are


 22   alive, most importantly.  So, we would welcome a




  1   chance to sit down and review these x-rays.


  2             DR. TEMPLE:  While you are at that, that


  3   is the second question I was going to ask you and


  4   maybe you want to answer them both.  The survival


  5   curves don't seem to have different tails on them.


  6   So, I am a little confused about where the


  7   long-term survivors you are referring to come from


  8   if they are not in the survival curve, or maybe the


  9   curve has been extended.


 10             DR. ITRI:  We provided update survival


 11   information to the agency--


 12             DR. TEMPLE:  I just need the one you


 13   showed though.


 14             DR. ITRI:  Well, that was an early cutoff


 15   so we don't really know what the tail is doing.


 16   That was the 7-month median.


 17             DR. TEMPLE:  It is really Dr. Cheson's


 18   question I am following up on, if there were a


 19   small subset of people that got really important


 20   responses, wouldn't you see a difference in where


 21   the tails end up?


 22             DR. ITRI:  It might be too early to see it




  1   on that curve.


  2             DR. TEMPLE:  Well, that means they are in


  3   both groups then.  There are long-term survivors in


  4   both groups.  Is that right?


  5             DR. ITRI:  There are some long-term


  6   survivors.


  7             DR. WITTES:  It depends on the nature of


  8   the censoring, where the censoring is.  So, some of


  9   that could be showing up before the edge of the


 10   tail occurs because they haven't been followed long


 11   enough.  I mean, the fact that they come together


 12   doesn't eviscerate the point.  You have to look at


 13   where the specific events occurred relative to


 14   censoring.


 15             DR. TEMPLE:  That is fair enough.  There


 16   was reference to at least some people who were


 17   getting really spectacular benefits and I would


 18   have thought that would show up as curves where the


 19   flat part is here on one and the flat part is below


 20   on the other.


 21             DR. WITTES:  They are censored.


 22             DR. TEMPLE:  They are censored because




  1   they haven't been on long enough--


  2             DR. WITTES:  It is like three years.


  3             DR. FORD:  Well, thank you very much for


  4   the opportunity to address the committee on this


  5   topic, the topic at hand being how does an


  6   investigator who sees the patient on a daily basis


  7   or a regular basis assess response compared to how


  8   an independent review facility would assess


  9   response in the same patient in a remote location,


 10   not having access to the clinical information.


 11             I think that there is little written in


 12   the medical literature about this topic, but there


 13   are two particular studies that I would like to


 14   review kind of as a background for this discussion.


 15   The first was a study that was published in the


 16   Annals of Oncology in 1997.  The author was a


 17   radiologist and that was a review of a 100-patient


 18   ovarian cancer trial.  In that review there were 24


 19   claimed responders who were reviewed by an


 20   independent review facility and in that instance


 21   there were 14 patients who were concordant, that


 22   is, deemed to be responders by the independent




  1   review facility and deemed to be concordant with


  2   the investigator.


  3             There was a second study that was done,


  4   also published in 1997 in the Journal of Clinical


  5   Oncology.  It was a review of a renal cell trial


  6   where there were 133 subjects who were reviewed.


  7   In that review an independent review facility


  8   reviewed those studies and the responses were


  9   concordant in 62 out of those reviews.  In that


 10   article you can see the concordance, that is, site


 11   same PR to independent review facility saying PR


 12   was approximately 60 percent, and in the second


 13   study it was lower, on the order of 48 percent.


 14             Now, with that as a background, there is a


 15   significant difference in the methodologies in


 16   which those reviews were performed.  That is, in


 17   those examples the investigators who enrolled the


 18   patients in the trial were actually part of the


 19   review process.  A radiologist sat down with the


 20   films, made the measurements and reviewed the


 21   images in concert with the physicians who knew much


 22   more about that patient, that is, had the




  1   additional clinical history that the radiologists


  2   would have at the time of the review.


  3             Now, that as a background, discussing the


  4   current study, the current study was a radiology


  5   only review.  When it was performed there was no


  6   clinical information provided.  In that instance,


  7   even in that particular setting the concordance was


  8   63 percent.  So, 63 percent of the time that the


  9   investigators assessed the response on this trial,


 10   the independent review facility assessed the same


 11   response.


 12             DR. TEMPLE:  When they are different how


 13   do you know which one is right?  When they are


 14   different, non-concordant, how do you decide which


 15   one is right?  I am sure I understand that


 16   different groups will reach different conclusions.


 17   Sometimes these special committees have a


 18   tie-breaker when they don't agree.  But what is one


 19   supposed to do that when they are non-concordant?


 20   How do you decide which is true?


 21             DR. FORD:  Well, in this particular


 22   setting the investigator-determined response was




  1   chosen.


  2             DR. TEMPLE:  When?  I mean, was this


  3   prospectively defined in the protocol how any


  4   discrepancies were going to be handed?


  5             DR. ITRI:  Yes, it was.


  6             DR. TEMPLE:  So, the protocol was clear


  7   that the investigator-determined conclusion, or the


  8   analysis based on the investigator--


  9             DR. ITRI:  The investigator measurements


 10   were fed into the computer and that is what was to


 11   be used for determination of response.


 12             DR. PRZEPIORKA:  Dr. Rodriguez?


 13             DR. RODRIGUEZ:  Yes, this is a follow-up


 14   to the question by Dr. Hwu because I didn't hear


 15   the response to part of her question, that is, you


 16   know, this is a biologically targeted agent and one


 17   assumes that one is going to look for the


 18   appropriate target or that one would select


 19   patients who are appropriate to be treated with


 20   this drug.  I didn't hear whether all patients


 21   entering on the study were screened, if their


 22   tumors were screened for expression of Bcl-2 or if




  1   there had been an attempt to quantitate category of


  2   patients because, obviously, some patients are


  3   going to be appropriate for trial and others are


  4   not.  Was that done?


  5             DR. WALL:  That is a very good question.


  6   Can I have slide MA-18, please?


  7             [Slide]


  8             The challenge with Bcl-2 is the ubiquity


  9   of Bcl-2 expression in melanoma.  So, this is not


 10   comparable, for instance, with HER2 expression in


 11   breast cancer in which the incidence of expression


 12   in advanced cases is on the order of 20, 25 percent


 13   so that you would not want to treat 100 percent of


 14   women.  You could theoretically benefit 25 percent


 15   so the absolute response rate would be 5 percent of


 16   your total.  In general, we chose melanoma because


 17   of the very, very high prevalence of expression


 18   which in these studies, whether you look at


 19   immunohistochemistry, which is the blue bars, or


 20   RT-PCR of excised specimen, you are talking about


 21   something in the range of 90, 95 percent expression


 22   of tumors.




  1             So, the kinds of correlations that you are


  2   going to be able to make with respect to


  3   over-expression we thought, going into this study,


  4   were going to be extremely limited due to the very


  5   high prevalence of baseline expression.  Again, it


  6   certainly influenced our choice of melanoma as one


  7   of the early targets for this particular disease.


  8   After that it is not clear where you could go if


  9   you were going to look at percentage


 10   down-regulation.  That meant serial biopsies of


 11   fresh tissues from multiple sites, handled very,


 12   very carefully, centrally managed, exponential


 13   increases in cost and ability to manage--that


 14   simply overwhelmed us as a small company.  So, we


 15   figured we would pick a big tumor in which would be


 16   an unquestioned level of very, very high expression


 17   at baseline but it did preclude the ability to make


 18   subset selections based on--at least at the stage


 19   we were dealing with this in 2000--Bcl-2 expression


 20   per se.


 21             DR. PRZEPIORKA:  Dr. Hwu?


 22             DR. HWU:  I agree that choosing melanoma




  1   as this malignancy is very important based on what


  2   we know of Bcl-2 over-expression.  My question to


  3   you that you didn't answer is based on your current


  4   regimen with some 300 patients.  Have you any data


  5   to show that it clearly reproduced your finding in


  6   the previous Phase 1 and 2 using completely


  7   different regimens?


  8             DR. WALL:  Well, the Phase 1 study, as you


  9   know, did not show correlations.  It really was not


 10   appropriately powered to look for correlations


 11   between baseline Bcl-2 expression and percentage of


 12   down-regulation.  That is very difficult to model


 13   even preclinically.  I am not sure I am answering


 14   your question.


 15             DR. HWU:  I don't agree that that is not


 16   the conclusion from the publication.  Clearly the


 17   CR person that has the highest incremental decrease


 18   of Bcl-2 is the percentage of decrease; it is not


 19   the total amount of expression.  That is what I


 20   learned from the paper.


 21             DR. WALL:  I think you need to keep in


 22   mind that it is a Phase 1 study.  That patient got




  1   a rather low dose.  The majority of patients were


  2   actually not serially sampled.  And, the ability to


  3   make inferences with respect to those kinds of


  4   correlations with a total N of 12 is I think very


  5   problematic.


  6             DR. HWU:  To make a correction, the


  7   patient got the highest dose level of 6.5 and she


  8   had 70 percent--


  9             DR. WALL:  And that blot was shown to you,


 10   by the way.


 11             DR. HWU:  --and the patient had never


 12   received any chemotherapy prior either.


 13             [Slide]


 14             DR. WALL:  Right, and here is the blot


 15   from that patient that Dr. Itri showed.  I think


 16   the major point, however, is with an N of 1 in a


 17   sample size of 12 in a Phase 1 study we didn't feel


 18   like we could make inferences.  I would say that


 19   one of the advantages of being an oncologist is


 20   that you can fall back on issues related to


 21   maximally tolerable dose and we felt that the dose


 22   used in this study for the Phase 3 study was




  1   comfortably above the threshold that we needed to


  2   achieve down-regulation of Bcl-2, which is a dose


  3   just above what this particular patient got.  Did


  4   that happen in 300 patient?  We don't have that


  5   information.  The willingness of patients to be


  6   serially sectioned for us to obtain this


  7   information on a fresh basis is rather limited and


  8   it was simply not part of the study.  It


  9   overwhelmed our capabilities in year 2000 and was


 10   not done.


 11             DR. PRZEPIORKA:  If Dr. Tolcher is here, I


 12   have a question.  In the in vitro studies is there


 13   a threshold amount of Bcl-2 that needs to be


 14   down-regulated to in order for the chemotherapy to


 15   show synergy?


 16             DR. TOLCHER:  That is a very good question


 17   and it is not well addressed.  Most of the models


 18   are, you know, somewhat artificial and in vitro


 19   versus in vivo really has no strict correlation.


 20   We functioned for a period of time with the


 21   assumption that 1 mcg/mL is probably the minimum


 22   effective concentration.  In almost all of the




  1   studies published to date we have a steady state


  2   concentration of 5 mcg/mL as an average.  So, based


  3   on the work that was done preclinically, published


  4   by Martin Gleave and others, we are well above what


  5   we would need in the in vitro setting but, again,


  6   the major caution always is that it is hard to


  7   relate what are the necessary concentrations in


  8   vitro to what are the necessary plasma


  9   concentrations for maximal effect.  Does that


 10   answer your question?


 11             DR. PRZEPIORKA:  I guess I was asking what


 12   is the amount of Bcl-2 intracellularly that we need


 13   to get the level down to in order to see the


 14   synergy with chemotherapy.


 15             DR. TOLCHER:   An excellent question.  You


 16   know, the issue is that it is dynamic so one


 17   doesn't know necessarily.  You are lowering it so


 18   that you essentially are shifting the equilibrium


 19   in favor of apoptosis.  You clearly do not need to


 20   extinguish all the Bcl-2 to have a pronounced


 21   effect in vivo.  In fact, you probably only have to


 22   drop it below some threshold and that threshold is




  1   unknown.  It gets more complex as well in that


  2   there is a diversity of Bcl-2 expression in


  3   different tumors.


  4             So, what I would say is that it is not


  5   necessarily a simple equation where you have to


  6   drop it below X amount.  It may be very dependent


  7   on the chemotherapy that is given with it.  So, it


  8   is not clear.  The certainty is that we do know


  9   that you do not have to extinguish all the Bcl-2 to


 10   have a synergistic effect preclinically.


 11             DR. PRZEPIORKA:  Thank you.  Dr. Bishop?


 12             DR. BISHOP:  I am relatively new to all


 13   this so I don't know if this question is


 14   appropriate or not but I am going to turn it to Dr.


 15   Kirkwood and Dr. Haluska.  You made passionate


 16   pleas for the treatment of metastatic melanoma in


 17   this randomized study.  So, would this treatment,


 18   Genasense plus DTIC, become the standard of care in


 19   the control arm for future CALGB and ECOG studies


 20   respectively?


 21             DR. HALUSKA:  I think that is a reasonable


 22   proposition.  I think that the context of this




  1   trial's conduct is that we have never shown any of


  2   these improvements and I think we shouldn't lose


  3   site of the fact that we are chipping away, as has


  4   been articulated, at numbers that have not been


  5   able to be chipped at away before because they


  6   haven't existed.  So, I think that that is a


  7   decision to be made by the community, but an


  8   improvement clinically like we have seen should be


  9   the standard against which other stage 4 therapies


 10   will be compared.  I think that is reasonable.


 11             DR. BISHOP:  Let me make it more specific


 12   then.  In your future randomized trials will this


 13   become the control arm?  The data with DTIC we know


 14   is not very impressive yet that is the community


 15   standard outside of immunotherapy.  So, as you plan


 16   your future trials, and you believe these results


 17   are impressive enough, will that become the control


 18   with which new therapies will be developed and


 19   compared to?


 20             DR. HALUSKA:  I wish we had new therapies


 21   to compare to now.  I would have to say that it is


 22   hard to view the future when those new therapies




  1   become available.  The landscape for drug


  2   development for melanoma right now includes other


  3   targeted therapies.  None of them is at the stage


  4   where we would choose a comparison arm like this


  5   but the short answer to your question is yes.


  6             DR. PRZEPIORKA:  Dr. Kirkwood?


  7             DR. KIRKWOOD:  I agree with Frank's


  8   conclusion so I think this is an incremental


  9   advance.  I think this is something that we have


 10   been trying to do in the studies that I reviewed


 11   and have not succeeded to do.  Obviously, if one


 12   were going to take survival as an endpoint in a


 13   future study it could still be dacarbazine but I


 14   think that we are talking here about response rate


 15   and we don't have anything that has reliably before


 16   shown response rates and complete response rates


 17   incrementally advanced as this has, with the single


 18   exception of high dose IL-2, which we have spoken


 19   about previously.


 20             DR. HALUSKA:  Something else occurs to me.


 21   I don't think it is the agency's job to support our


 22   research endeavors strictly.  I mean, their job is,




  1   as I understand it, to make agents available for


  2   public consumption.  But, clearly, these decisions


  3   do affect our research and we have, for reasons


  4   that are not clear to any of us who work in


  5   melanoma, been very unsuccessful in improving


  6   overall survival.  I don't believe that as long as


  7   we hold that out as the only endpoint that we can


  8   meet that we are going to meet it because it has


  9   been such an impediment.  But there is nothing in


 10   my mind that prevents small improvements in these


 11   sorts of endpoints from accumulating with addition


 12   of different agents and you can envision a variety


 13   of other things that you could add Genasense to


 14   that might also prove additive to the responses and


 15   progression-free survival we have seen today.


 16   Ultimately, that is how I think we are going to


 17   make real progress with the survival endpoint in


 18   this field.


 19             DR. PRZEPIORKA:  Dr. Redman?


 20             DR. REDMAN:  Thank you but Dr. Kirkwood


 21   answered my question.


 22             DR. PRZEPIORKA:  Other questions from the




  1   committee?  Dr. Tolcher, could you please come back


  2   to the microphone?  We need to have you identify


  3   your affiliation, please, for the record.


  4             DR. TOLCHER:  Sure.  I came actually today


  5   without personal compensation by Genta or any of


  6   the pharmaceutical sponsors, although my travel


  7   arrangements have been paid for Genta.  I have been


  8   the principal investigator on three clinical


  9   studies and have acted as an occasional advisor to


 10   Genta and Aventis and have been compensated with


 11   honoraria for those less than $10,000.


 12             DR. PRZEPIORKA:  Thank you.  Hearing no


 13   other questions, we will break for ten minutes and


 14   return at 10:40 to begin the open public hearing.


 15   We will need to begin the afternoon session on time


 16   so please be on time for the next part.


 17             [Brief recess]


 18                       Open Public Hearing


 19             DR. PRZEPIORKA:  If we could have the


 20   doors closed, please, we will begin the second half


 21   of this session.  This is the open public hearing


 22   and we actually had many individuals who wanted to




  1   speak this morning and, in order to give everyone


  2   who is registered a chance to participate and to be


  3   fair to all, we will be following some fairly


  4   strict procedures.  We have a timer.  Each speaker


  5   has been allotted two minutes and at the end of the


  6   two minutes we will ask that speaker to return to


  7   their seat and the next speaker to immediately


  8   begin.  Due to considerations of fairness and these


  9   restrictions of time, only speakers who have


 10   registered will be allowed to come to the podium.


 11             Both the FDA and the public believe in a


 12   transparent process for information gathering and