And what I want to do is actually then start with the introduction of the Committee Members, and what we?ll do is just go around. If everyone will introduce themselves, Mr. Ohye.

Dr. Sara Taylor is excluded from participating in today?s discussion and vote concerning Casodex. We would also like to note for the record that George Ohye is participating in this meeting as an Industry Representative acting on behalf of regulated industry.

I am an eight-year prostate cancer survivor, and a three-year throat cancer survivor. And in addition to being Chairman of the Florida Prostate Cancer Network, I am also Co-Chairman of the Florida Prostate Cancer Task Force, and I was the Founding Chairman of the National Prostate Cancer Coalition. I am on the Board of Directors of the Moffitt Cancer Center, and I served on the NCI?s Prostate Cancer Progress Review Group. However, in 1992, I retired as Vice President of what is today J.P. Morgan Chase and moved to Tampa, Florida. And in 1994, I got diagnosed with prostate cancer. That set me off on a whole new direction in life, because I will admit that I had very little understanding of the disease prior to that.

MR. CAPUTI:  Good morning everyone. My name is Anthony Caputi, and I am the Manager of Government Relations and Patient Advocacy for the American Foundation of Urologic Disease. I?m also a prostate cancer survivor, so I?m wearing two hats here. A couple of things first.

I?d like to read a statement that I prepared on behalf of my organization for Casodex 150. I?m writing to offer the American Foundation for Urologic Diseases', AFUD's, support of AstraZeneca?s application for Casodex 150 to be used as an adjuvant therapy of curative intent for patients with locally advanced, non-metastatic prostate cancer.

In the patient?s world, changes in PSA levels are the clinical indicators that guide a physician?s treatment and recommendations. As patients, any treatment that inhibits the progression of PSA, particularly for those of us who are at high risk for disease recurrence, is welcome with due consideration of potential side effects.

At this point during the prostate cancer journey, treatment options are limited, and side effects from the treatments for advancing disease can be very distressing to a man?s quality of life. The current standard of care for advanced non-metastatic disease is administration of an LHRH analog. The side effects of this drug therapy include hot flashes, loss of libido and bone loss. For many, this disruption in quality of life is very distressing.

If Casodex 150 milligrams were approved by the FDA, this would provide an additional tool for the treatment of high-risk disease. To me, the data collected during the Casodex 150 trial phase is sufficient to warrant its approval. For the high risk patient, Casodex 150 appears to be effective in inhibiting the progression of prostate cancer as defined by the standard of care in today?s clinical practice; and that is, PSA monitoring.

The side effect profile offers certain quality of life improvements over today?s standard therapy, such as reduction of hot flashes, retention of sexual interest and function, and the preservation of bone mineral density.

MR. MARFYAK:  Good morning. This is an awesome group. I didn?t expect to see so many of you here. As a former state employee of the State of Wisconsin, running a budget shop and subsequent to that, 23 years with the Department of Energy, I?ve conducted a number of hearings such as these over the years. I would point out that normally we allowed our people a good deal more time to speak than five minutes.

I?m a prostate cancer survivor, and I?m currently Co-Chairman of the Pennsylvania State Coalition. In addition to that, I am working with the NPCC, National Prostate Cancer Coalition, in setting up state coalitions all over the United States. Furthermore, I also sit as an evaluator, consumer evaluator for the Congressionally mandated program at Fort Detrick that allocates roughly $85 million a year to the study of prostate cancer.

I am here as a supporter of this request by Casodex. They have paid my way from Gettysburg to come here. I?m going to be very short and to the point. You?ll have heard all these arguments later on, so I?ll be succinct in what I have to say. I?m neither a statistician, nor a pharmacist, and so I?m not equipped to address the numbers or the science involved in AstraZeneca?s study. But I can address the proposal?s efficacy from a consumer?s point of view. After all, they are the beneficiary of whatever you decide.

My feeling has been, in examining the study that has been done here, that you?re talking basically about promise versus risk. As Tony and as my friend Bob from Florida have already stated, the consequences of prostate cancer are enormous, and until you?ve watched people die from this disease, watched the suffering that goes on, you really don?t have a full understanding of what?s involved. And anything that alleviates that, anything that creates a possibility is a welcome piece of ammunition in our arsenal to fight this disease.

On the other hand, there is a risk, always a risk. And as Tony has pointed out, there are a lot of liabilities on this. But if the patient has informed understanding of what?s involved, then this is something between the doctor and the patient to decide.

MS. NISSENBERG:  Good morning. I?m Merel Grey Nissenberg. I?m an attorney in medical litigation issues in California, and I?m here today because my pro bono work is heavily concentrated in cancer and related issues. I also represent a very large constituency. I am in my fourth term as President of the California Prostate Cancer Coalition, which is a network of individuals, healthcare providers, and every support group for prostate cancer in the state. I am also the Co-Chair for the State Coalition Advisory Board for the National Prostate Cancer Coalition. I?m a CARRA member for NCI, and I?m the legal advisor to the Cancer Task Force in San Diego, so I come here to represent a great deal, a great number of voices in asking you to recommend approval of Casodex 150 in the proposed indications.

We?re here today because AstraZeneca has conducted the largest ever randomized clinical program in men with prostate cancer. As our data demonstrates, Casodex 150 milligrams significantly reduced the risk of objective progression in these men. However, the FDA issued a not approvable letter in June, because of lingering questions about the relevance of these data to U.S. patients, so today?s presentation will answer the questions posed by the FDA in their Briefing Document.

The FDA in their Briefing Document refers to this as the sponsor?s endpoint. It?s not. It is a standard endpoint. It?s commonly used in clinical trials. It was agreed with the Agency in 1995, and it was the endpoint upon which we decided to embark upon this very large clinical trial program. The FDA's retrospective endpoint of time to bone scan progression was requested by the Agency in 1995, following the close of recruitment to this 8,000 patient study.

In establishing the Casodex 150 milligram Prostate Cancer Program, AstraZeneca consulted with prostate cancer experts throughout the world. The Casodex program is briefly outlined on this slide, and Drs. See and Soloway will be reviewing the program in much greater detail. However, it?s important to note the following points. A total of 8,113 men with localized or locally advanced prostate cancer, were recruited in less than three years -- it?s a monumental achievement -- from 353 centers, in 23 countries. This represents the largest clinical trial program ever conducted in this disease.

The next two slides cover some very important definitions for this trial program. Localized disease means that the cancer is confined to the prostate gland ? that is T1 or T2 disease, and locally advanced disease is defined as disease that has penetrated the capsule, and is designated T3 or T4 disease.

More importantly, a new treatment option now exists. Casodex has the potential to delay the onset of serious and painful disease-related complications in men with prostate cancer. And now I?ll turn the podium over to Dr. Howard Scher, who will describe the unmet need for Casodex therapy in this disease.

DR. SCHER:  Good morning. I?m Howard Scher, and I?m Chief of the Genitourinary Oncology Service at Memorial Sloan-Kettering Cancer Center in New York.

As a medical oncologist, I focus on advancing therapy for patients with prostate cancer. I?m involved in medical decision making for patients with localized disease, and the treatment of patients with more advanced and recurrent disease.

This slide shows what I?m going to discuss this morning. I?m going to demonstrate that patients with prostate cancer would benefit from additional treatment options. I will show where, in the spectrum of the disease, additional options are needed. The spectrum includes newly diagnosed patients with localized cancers who are at risk for recurrence, and patients considered for watchful waiting.

There are a number of options. A patient may elect to undergo therapy with curative intent by radical prostatectomy or radiation therapy. The choice is determined by considering characteristics of the patient?s cancer, their age, concomitant morbidities and preference. A patient may opt, or may be advised to defer treatment and undergo active surveillance, or watchful waiting, or a patient may elect to undergo castration-based therapies in the hopes of slowing the progression of their disease.

What happens when primary therapy fails? Typically, this is first manifested as a rising PSA. It is a sign that the cancer has not been cured. At this point, many men are offered castration. PSA progression is followed by objective progression on a bone scan or other imaging study. It is at this point forward that a patient?s quality of life deteriorates both from the disease and its treatment. It cannot be cured. The disease itself can cause bone pain, anemia, fatigue and/or spinal cord compromise. And the castration-based therapies that are currently available for PSA or objective progression are associated with unacceptable side effects for many patients. We clearly need better options to delay disease progression, and an option which we can discuss with our patients who are at high risk for failure. So who are these patients?

Now let?s focus on the 20 percent of patients who are treated with watchful waiting. What is the effect of no active treatment, and do these patients also need additional options? Watchful waiting is a conscious decision, or a formal recommendation to undergo no immediate therapy after the diagnosis of prostate cancer is established. These patients are felt to have competing causes of morbidity or mortality that exceed the risk of symptoms or death from prostate cancer, or wish to avoid the complications and side effects of radiation therapy or surgery.

DR. SEE:  Thank you, Dr. Scher. Good morning. My name is William See. I am Professor and Chief of the Department of Urology at the Medical College of Wisconsin. I?m speaking to you this morning as one of the Principal Investigators for Trial 23, the North American Trial.

Now I?ve been asked to address one minor housekeeping item. Dr. Kennealey alluded to some definitions that are used in this clinical trial, and we noted that you were searching your binders for those. Those are located on the outside of the back cover of your binder.

Now I?ve been an investigator in this trial program since its inception. Many of my personal patients are included in this trial, and I am intimately familiar with the details of this program, as well as the results.

I?ve organized my presentation as follows. First, I will review the objectives, the design, and the relevance of the EPC program. I will then demonstrate the efficacy results of the Casodex 150 milligram dose program, and I will place special emphasis upon those patient populations who see greatest benefit from Casodex in the EPC trial program. Specifically, patients at high risk for disease progression managed in either the adjuvant or the immediate therapy setting. And finally, I will summarize the data analysis of other clinically relevant endpoints.

As I?ve mentioned, the trial program consisted of three different clinical trial protocols. And as you heard, these were conducted around the world. By design, these trials were designed to capture patients constituting the entire spectrum, if you will, of non-metastatic prostate cancer. While the overall design of each trial program was fundamentally similar, differences in eligibility criteria and prior treatment across the three trial programs were intended to capture the spectrum of early prostate cancer patients.

The second primary endpoint of this trial program was survival. I will tell you that given we only have three years of follow-up in this trial program, survival and prostate cancer in this trial is immature. Consequently, I?m going to be focusing on the outcome data for the primary endpoint of objective disease progression.

Now there were secondary endpoints shown here. Time to PSA progression, as we?ve heard from some of the initial speakers, is a disease-relevant endpoint in this country. In U.S. practice, PSA progression is widely considered to imply primary treatment failure, and often serves as a prompt for the initiation or institution of second-line therapies. Additional endpoints included time to treatment failure, and finally, tolerability and safety, which will be addressed by the next speaker, Dr. Mark Soloway.

There?s also an interesting trend in the United States to upgrade Gleason sums. This slide is derived from the CaPSURE database of over 7,200 patients, and it illustrates what has happened over the last decade for the use of Gleason scoring in this country.

I?ve told you the EPC Trial Program represents a broad continuum of the disease process we refer to as prostate cancer. It includes patients with relatively low-risk, early disease, such as the subset seen in Trial 23, as well as patients with higher tumor burdens, such as those seen in Scandinavia and Europe.

Now patients in Trial 23 represent a subset of the overall trial program with the lowest tumor burden as evidenced by their clinical stage and pre-treatment PSA. However, it?s important to recognize that Trial 23 is not only a subset of this overall clinical trial program, but in fact, represents a subset of patients with prostate cancer which we encounter and manage in the United States.

Let?s shift for a moment from the endpoint of objective disease progression and look at the effect of Casodex on biochemical failure for the overall clinical trial program. This Kaplan-Meier plot for PSA progression, with Casodex shown by the orange line, demonstrates a 59 percent reduction in the risk of PSA progression associated with the use of Casodex.

In U.S. practice, PSA progression is regarded as treatment failure, and frequently triggers the initiation of systemic therapies, specifically hormonal deprivation. Consequently, as we?ve heard, a treatment that delays PSA progression is clinically relevant.

I previously indicated to you that patients in North America often receive hormonal therapy at the first evidence of PSA, that is, biochemical disease progression. It appears that this was true in Trial 23. In Trial 23, five times as many patients in the placebo arm, as compared to Trials 24 and 25, had medical castration therapy introduced in the absence of objective clinical progression. This truly confounds our ability to interpret the effect of adjuvant hormonal therapy for objective disease progression in Trial 23. What, in essence, we?ve ended up with is immediate adjuvant therapy, or very early androgen deprivation therapy at the first evidence of biochemical progression.

Let?s return now to a discussion of the primary data from the EPC trial program. You will recall that the FDA has also asked these data to be presented according to an analysis of bone scan confirmed progression over the first two years of the trial. These data are shown here.

First, let?s consider the potential patient treatment groups and treatment settings. This slide demonstrates a matrix of potential treatment groups based upon either adjuvant or immediate treatment, and segregated by localized versus locally advanced disease. Let?s start by looking at the adjuvant treatment setting.

Let?s first talk about patients treated primarily with radiation therapy. Based upon the literature, patients at high risk for failure of mono-modality radiation therapy include those patients with clinically staged, locally advanced disease, and an elevated pre-treatment PSA.

Multivariate analysis of the data from the EPC Trial Program confirmed the relevance of these factors in our trial, and subgroup analysis data that I?m now going to show you confirm that there was a marked benefit for Casodex in this specific subset of patients.

Let?s move on now to patients treated by radical prostatectomy. The literature suggests that those patients at greatest risk for failure of surgical mono-therapy are those with locally advanced disease, or a detectable post-operative PSA, or an elevated pre-operative PSA, or a Gleason sum greater than 7.

In reviewing the patient populations who are candidates for this treatment strategy, while again we don?t see an indication for localized disease, we do see a clear treatment effect for high-risk patients who are treated initially with curative intent with either radiation or radical prostatectomy.

Let?s now transition to this other subset of patients. Let?s talk about the use of Casodex as immediate therapy as an alternative strategy. These are patients who are not considered candidates for therapy with curative intent. These are the so-called watchful waiting patients.

Prostate cancer is the same the world over. The last time I checked, prostate cancer doesn?t need a passport. Therefore, as we review the patient matrix, the EPC trial program provides data strongly supporting the use of Casodex as immediate therapy in patients with localized disease who are not candidates for therapy of curative intent. This is in addition to the data I showed you that strongly demonstrated a treatment effect in the adjuvant setting for patients with locally advanced, that is, high-risk, disease following therapy of curative intent with either radiation or radical prostatectomy.

Now there is one thing that?s important to take away from this slide. You will note that there was no difference between the treatment arms for non-prostate cancer deaths. This observation supports the well-established safety profile of this specific agent.

By way of introduction, my name as indicated, Mark Soloway. I?m Professor and Chairman of the Department of Urology at the University of Miami, in Miami, Florida. In addition to my hat as chair, and many of you in this room have a lot of administrative responsibilities, I see about 100 patients a week as a urologic oncologist. Half of them are prostate cancer patients, and I?ve been involved with the Casodex program for a number of years, and have been participating in the Casodex 150 EPC Program, as well.

Now in the EPC program which Dr. See went through with you in detail, I think you?ll note that if we look at adverse events here as indicated on this slide, clearly the most common are breast pain and gynecomastia. But also of note, one will see that typical effects that one sees with the current forms that we have available of surgical or medical castration, if we look at asthenia, if we look at impotence, if we look at hot flashes, they?re almost the same between the Casodex 150 and placebo, so that?s a dramatic difference from what we currently have available for our patients.

Now the next slide shows withdrawal from therapy on the EPC program. And again here we?ll note that the most common reason for withdrawal, in fact, is adverse events from the Casodex 150 group. And in the placebo group, the most common reason is progression from treatment.

The incidence of adverse events and adverse events leading to death for patients on Casodex 150 are quite similar, as you see Casodex and placebo. And the events were of the type to be expected from an older population, strokes, heart attacks, COPD risk problems, et cetera. So I think in conclusion, regarding the safety profile of Casodex 150, it is a favorable profile. That?s how I would look at it.

The side effects observed are, in fact, well-characterized, they?re predictable, they are generally mild to moderate, and consistent, as I?ve mentioned, with the known pharmacology of this product. The main tolerability issues clearly are gynecomastia and breast tenderness.

Currently, many urologists, radiation oncologists and medical oncologists use medical or surgical castration in patients with non-metastatic prostate cancer. That is a quantum difference than what we saw 10 or 15 years ago, when most of our patients, in fact, had metastatic disease, and that?s where we used the androgen deprivation. In fact, despite limited clinical trial data, the most common use of LHRH analogs in the U.S. today are in patients with non-metastatic prostate cancer; that is, the rising PSA.

Surgical or medical castration, as we know, is associated with a list of adverse effects which significantly alter a patient?s quality of life. And some of these adverse events are seen here, hot flashes often requiring, certainly in my practice, additional treatment, erectile dysfunction or impotence, loss of libido. Important one for men who are vigorous, who have occupations, attorneys, physicians, accountants, et cetera, is their cognitive function. And there is now emerging data that typical forms of androgen deprivation; that is surgical or medical castration alter this not insignificantly.

And something that I think will become more important over time is that with increasing use, duration of androgen deprivation, particularly again referring to the medical or surgical castration that we use, be it orchiectomy or LHRH analog, there?s a progressive loss of bone mineral activity, and thus, osteoporosis. And I think in the future, this is going to be a major problem we?ll have to address.

Well, Casodex 150 is a different type of androgen deprivation. It is a potent anti-androgen with a once-daily pill, and I think this is important, one, for compliance having a once a day pill, but also because it frees the patient?s schedule from having to go to the physician every three or four months for that injection. It puts them a little bit more in control. Since testosterone is maintained, there?s a different side effect profile. And, in fact, sexual function and aspects of bone mineral density are, in fact, retained with Casodex.

Now really quite forward looking, remember when this trial was designed, the Scandinavian group decided for patients who had normal erectile function going onto the trial that they would use a validated questionnaire, and at regular intervals they would ask the men to fill out this questionnaire. Now we say oh, that?s a natural, but this is several years ago when they did this. And this looked at sexual function and sexual frequency.

And if we look at the data on the next slide, and again, this is one of the largest data sets asking these questions in men with prostate cancer, I think you will note that there?s very little difference between the Casodex 150 and placebo, and this is questions related to sexual function.

Now again, those of you who are familiar with various forms of androgen deprivation which lower testosterone, orchiectomy or LHRH analog, there would be a very dramatic difference here. And the next slide shows in the questions related to sexual frequency, again not much difference between Casodex. It?s a little bit lower, but dramatically different than one would expect if these patients had an LHRH analog, for example.

To address this, AstraZeneca performed a prospective randomized study of 103 men who either had Casodex 150 or an LHRH analog, and used the standard test dual emission x-ray absorption to monitor bone mineral density at indicated times. And it?s actually pretty dramatic. If you note that the LHRH analog group in the blue on the bottom, bottom because their bone mineral density diminished over time. And by week 96, there is a 5 percent loss. That?s already in week 96, two years. And the level of minus 5 percent is where you start seeing osteoporotic fractures.

In contrast, the men who took Casodex 150, they?re above the line and retained the bone mineral integrity. And as I said, when we?re talking about men who, if they elect androgen deprivation for rising PSA, they may be on these ?- they may have orchiectomy or an LHRH analog for many years. And I think in the future this is going to be a major problem.

Well, as a clinician who practices in the U.S., obviously, like many of you, there are clinical scenarios for which I would like the opportunity to discuss, and if I discuss, I would like the opportunity, of course, to prescribe Casodex 150. Now this would include patients who have received radiation therapy, who have had a radical prostatectomy, or who may be electing watchful waiting. And by knowledge, by your knowledge, by the patient?s knowledge, and patients are pretty sharp these days, and you heard some of them earlier today, they are at high risk for recurrence or progression of their disease. And I?m going to give you a couple of examples from my practice, which I hope will bring this home.

Not surprisingly as many of our patients, he was surfing the net and he pulled up one of the nomograms which are, in fact, patients know them more than I do at times, and he says, you know, Dr. Soloway, my clinical recurrence rate is about 50 percent. And he said, "Gee, I want additional therapy. I?m not happy with that. But on the other hand, I don?t want the side effect profile that I?m aware of with an LHRH analog."

Well, based upon this information we heard today from the large, and I think it?s quite large, EPC Program, we know that Casodex therapy will reduce his risk of progression by 61 percent. I think given this information, I, as a treating physician, would like the opportunity to discuss this approach with him. And then, of course, discuss it. One would want the opportunity to prescribe Casodex 150. He may or may not elect to have it, but I think that opportunity, particularly with the favorable side effect profile of this agent, should be reasonable and available to the patients.

Now from my database of over 1,100 radical prostatectomies and databases, there are some people in the room that have similar databases, this gentleman has a 50 to 70 percent chance that he will initially have PSA, and then subsequently clinical recurrence. And as we?ve heard today from the large Casodex 150 program, the EPC program, he has a 47 percent less chance that he will progress over time if he receives Casodex 150. And I think it is very important to provide the information. The information is out there, and thus, allow the opportunity, if he chooses, to have that agent.

The next patient is a little bit different scenario, and each case, of course, is somewhat different. And I think this again highlights some of the important variables that we have to deal with. This is a 77 year old, very well known. He wouldn?t probably be known to many of you, talk show host. And he deals with important personalities every day.

Well, he found out his PSA was 17. He had a biopsy and he has prostate cancer. Now cognitive function is critical to this guy?s very livelihood, and what he does every day. And parenthetically, he also has a relatively young wife. Again, unfortunately, and not without much discussion, he was said you need an LHRH analog. He was put on this, and he was devastated.

From the data that you?ve heard today, and the adverse events safety profile data, I would conclude that the benefit risk ratio for Casodex is clearly favorable in patients who are at high risk for recurrence or progression. And most of us, as oncologists, know who these people are, and increasingly the patients know who they are. Right now there?s only one systemic treatment option for them, and that is surgical or medical castration. And many men simply do not want to tolerate the side effects related to that, and that?s why they often choose, whether they have radiation or surgery, much of it is based on the side effect profile.

I think Casodex 150 does represent an important treatment option for the patients we treat on a daily basis in the United States, and it fulfills an unmet need. Thanks for your attention, and we?ll go back to Dr. Kennealey.

You?ve seen an awful lot of data today, both efficacy data and safety data, and I would like just to summarize, show the slide again, because I think this slide gets to the heart of what we?re here to talk about today.

Dr. Hoberman has recently sent to all of you a statistical addendum looking at the Early Prostate Cancer Program, and there?s some clear agreement between AstraZeneca and the FDA on a number of these issues. And I?ve put these areas of agreement on this slide. And they are, Casodex 150 milligrams reduces the risk of progression regardless of primary treatment. Dr. Hoberman has confirmed that patients in Trial 23 are at low risk for progression, and that PSA and stage are important determinants of outcome. He has also noted that objective progression may be suppressed in Trial 23 due to U.S. clinical practice. And he also noted that the central re-read of the bone scans supported the protocolled primary endpoint.

I?m going to ask my colleague, Dr. George Blackledge, who many of you remember from September. Dr. Blackledge is the Global Vice President of Oncology, and has been heavily involved in Trials 306 and 307, to give you some information about these trials.

I think it?s worthwhile putting Trials 306 and 307 into context. If we look here at the slide based on Dr. Scher?s presentation, we can see the spectrum of prostate cancer from clinically localized disease right through to clinical metastases. As is usual in oncology drug development, we began our development at the more advanced end; namely, in the presence of metastases, and that?s exactly what Trials 306 and 307 began to do.

In metastatic disease, it?s true that we fell short of our objective, and we did not demonstrate equivalence with castration with metastatic disease. We actually had a shortfall which was statistically significant of 42 days in terms of median survival, so we did not feel that we could progress with a metastatic claim at that point. However, as you can see, this trial program also covered clinically locally advanced disease.

You can see here the differences between the two populations. In Trials 306 and 307, the aim was palliative. In the EPC trial, patients chose not to undergo therapy of curative intent. Castration was considered a standard of care. It had to be for the randomization in Trials 306 and 307. There?s no standard of care for the patients entering the Early Prostate Cancer Program.

You can see dramatic differences in the median PSA, and indeed, in the T stages, with most of the patients in the Early Prostate Cancer Program having T1 or T2. And these actually not being allowed in Trials 306 and 307. So the amount of overlap that there is between Trials 306 and 307, and the Early Prostate Cancer is vanishingly small. And even in Trials 306 and 307, together with the accumulated body of data, there?s a strong suggestion that there is really no difference between Casodex and castration in terms of survival outcome. Over to you, Dr. Kennealey.

DR. KENNEALEY:  Thank you, George. And let me now go on to discuss the questions that the FDA has posed, and they?re in your Briefing Document. And I have shortened them in order to fit them on the slides.

Additional analysis of these data with now a follow-up of 4.2 years has confirmed this benefit. A substantial delay in disease progression is truly a long-term clinical benefit. We can?t lose sight of the fact that delaying progression to bony metastases is meaningful to men with prostate cancer, nor lose sight of the fact that patients with metastatic cancer face a lifetime without prospect of cure.

CHAIRPERSON PRZEPIORKA:  Thank you, Dr. Kennealey. And at this time, I?d like to ask if Dr. Schoenberg could come forward and take his seat, because I think we want to have the Committee ask the sponsor questions at this point. And I will actually open that question asking period with questions for Dr. Kennealey or Dr. See.

DR. KENNEALEY:  The question is, are the ends for localized disease in the watchful waiting patient group positive separately? And I?ll ask Dr. Charles Morris, who is the Senior Director of Oncology for AstraZeneca to respond to that question.

DR. MORRIS:  Charles Morris, AstraZeneca. Yes, the treatment effect as we?ve seen for the localized watchful waiting group overall is reproduced within Trials 24 and 25, as you can see on this slide, with the 32 percent reduction in Trial 24, and 39 percent reduction in Trial 25.

CHAIRPERSON PRZEPIORKA:  And in parallel, you?ve asked for approval for adjuvant treatment for locally advanced disease after radiotherapy or prostatectomy. And are the endpoints significant in Trials 23, 24 and 25 individually for that subgroup of patients?

For the high risk patients on radical prostatectomy, what we see on this particular slide is that the effects are seen within Trials 24 and 25, but less of an effect is seen in Trial 23. This, as has been discussed in Dr. See?s presentation, really seems to be consistent with some of the events which have been going on in terms of U.S. clinical practice, where the number of patients who received additional therapies at PSA progression was actually much higher in the United States trial.

DR. KENNEALEY:  That?s correct.

DR. MORRIS:  The length of time is important, so that?s why I wanted those clarified. The other question that I have is, am I correct that the withdrawal of patients on Casodex is 27 percent? Is that accurate?

DR. MORRIS:  Now that, to me, is a striking number. Okay? That means that nearly a quarter to a third of patients choose to come off because of a side effect. Do we know ?- so that?s a problem in and of its own, but the other question that follows that in my mind is, is that number different in the U.S. versus in the non-U.S. studies? In other words, do we have more patients in this country who came off because they perceived toxicity?

The large part of that 27 percent withdrawal rate from Casodex is due just to adverse effect, gynecomastia and breast pain, which account for 16 percent. When you take that out of the equation, there is no difference between the Casodex response. If we look across the three trials, we do see some differences in withdrawal rates, that?s true.

The withdrawal rate in Trial 23 is somewhat higher than in Trials 24 and 25, particularly with regard to gynecomastia and breast pain. With regard to other adverse events, there?s very little difference in withdrawal rates.

DR. KENNEALEY:  I?d like to ask Mr. Kevin Carroll, our Statistician, to respond to the question concerning bone scans and the bone scan re-read that we performed.

MR. CARROLL:  Thank you. Kevin Carroll, Statistician. I think it?s important to point out that the bone scan re-read exercise was designed to assess whether there was any bias in the local reading of scans. And in concordance with the review provided by Dr. Hoberman, the re-reads indeed showed no evidence of any bias in the local reading of bone scans between Casodex and placebo treated patients. And, therefore, the time progression endpoint as defined in the trial protocol at the outset, is fully supported by the re-read results. Thank you.

DR. HANNO:  My question is though, if in reality half ?- there are studies that show that in cancers where the bone scan turns positive when it?s been negative, only about 14 percent may be true positives, and people do MRIs, they do KUBs, they do other studies to see if there?s a reason for this. Even if it?s the same in both groups, if the actual number of true positive bone metastatic events is half of what you have in the study, how would that affect your results?

DR. BLAYNEY:  Secondly, a large measure of your indication hinges on what you call watchful waiting in Europe. And as a medical oncologist in this country, we do get a fair amount. I?d like to get some sense, perhaps from one of your European investigators, what actually happened during your investigation. Were there patient ?- during your clinical trial. Were there patients who were prevalent, if you will, or who were being followed in clinic, and then all of a sudden the next time they appeared their physician said, "Oh, we have this potentially new drug. Would you like to be involved?" Or was there some triggering event that said, "Now is the time for you, sir, to be involved in this clinical trial." Because those are, I think, two different iatrogenic stimuli, if you will.

DR. KENNEALEY:  I?d like to ask Dr. John Anderson, who is one of the investigators in the European trial, to respond to the treatment practices concerning watchful waiting.

DR. ANDERSON:  John Anderson. I?m a urologist from the U.K. I was involved in Trial 24. The scenario you described is exactly the case. We have a number of patients who elect to go on watchful waiting in the U.K. They are more concerned about the potential morbidity that goes with radical prostatectomy, or radical radiotherapy, and elect to sit on a watchful waiting program.

Nonetheless, these men remain concerned, and obviously when they come to clinic, the first thing we do is check their PSA, and we can?t get away from the rising PSA story. The patient?s disease may not have changed, but he starts to see something alter. Along comes a trial where we?re exploring a new agent, and we offer the patient to be randomized into this trial to see if it?s going to make a difference for him. I?m a great believer in sharing that information with patients explaining the limitations of the treatment, but also illustrating the potential benefits, and many patients with a rising PSA on a watchful waiting program were keen to be enrolled in this study.

DR. BLAYNEY:  So I?m given to understand that the stimulus to enroll in this study was some action perceived on the part of the patient, not a new trial opening in your or other centers.

DR. KELSEN:  There?s been a considerable discussion about the Gleason Score, so I have a technical question, and then a question about some of your conclusion slides. Was there a central pathology review of the European pathology slides to address the issue of Gleason stage, or was this determined by the local pathologist?

DR. MORRIS:  You remember that the primary intent of the trial was to demonstrate an effect overall in patients with early prostate cancer, irrespective of radical prostatectomy, radiation therapy, or watchful waiting. There was no specific requirement or entry criteria based on Gleason Sum, nor was there any a priori intent to analyze results in terms of ?- in relation to Gleason Sum, so we did not see a need prospectively to define a central pathology review.

DR. KELSEN:  So if I just follow that up, in CC-7, in your conclusion slide to partly address this, you made the point that Gleason was not an independent variable, and was not an important part of making a decision regarding treatment, so I?m curious how on CC-9, one of your requests for patients who derive greatest benefit is radical prostatectomy patients who have locally advanced disease and any one of several criteria. And as I read this, maybe I?m reading it wrong, it?s one or the other.

DR. KENNEALEY:  That?s correct.

DR. KENNEALEY:  Yes, that?s correct.

DR. KELSEN:  So it?s not an independent variable, but it can be chosen to choose a patient population.

DR. KENNEALEY:  You?re asking for CE ?-

DR. KENNEALEY:  37. These are the three indications ?- is that the slide?

DR. BRAWLEY:  The first question is about the LHRH agonist. Are they approved for ?- it?s my impression that there is currently not any drug approved for these indications. AstraZeneca ?-

DR. BRAWLEY:  Okay. One of the things that I frequently worry about is truth in advertising, and whenever I want to argue something, I usually use relative risk. And whenever I want to argue against something, I usually use absolute risk. Let?s look at 42 percent there overall, and please tell me if I?m wrong. I believe that refers to slide CE-13, and if we can go to CE-13.

DR. KENNEALEY:  It?s the reduction, the relative reduction in risk.

DR. KENNEALEY:  I?ll ask our statistician, Mr. Carroll, to respond to that. But I?d also preface his remarks by mentioning that this reduction in relative risk is in line with or actually exceeds the reduction in relative risk seen with Tamoxifen ?-

DR. BRAWLEY:  But that?s for mortality.

DR. KENNEALEY:  ?- in the breast cancer studies. Both for time to objective progression and for mortality.

DR. BRAWLEY:  Yeah. But see, my problem ?- I?ll already play my hand. My problem is, Tamoxifen has been shown to reduce mortality at these rates, and you?re showing that there?s a reduction in disease recurrence at these rates. There?s a difference between recurrence and death.

DR. BRAWLEY:  Okay. I?d much rather see you in your insert if you did this indication, not say there?s a 42 percent reduction, but say that 5 percent of the guys getting this drug benefit.

DR. SCOTT:  Mark Scott, AstraZeneca. You?re correct that the reduction in risk for Tamoxifen, there was a survival benefit, but the original approvals for Tamoxifen were based on a reduction risk of time to progression.

DR. BLACKLEDGE:  George Blackledge, AstraZeneca. I think we have to be quite careful about making absolute differences, because this is actually summing the data for the whole population. And actually, you could have an effect for every single patient, a smaller effect for every single patient actually making this up, so we cannot talk in terms of having to treat a hundred patients to benefit 27 or 13, or whatever it is. You can actually be getting the benefit, as we believe we do in breast cancer in the adjuvant setting, across the whole population, so I think it?s probably inadvisable to be talking about treating so many people to get so much benefit, because you can actually be benefitting the whole target population to a greater or lesser extent.

DR. BRAWLEY:  Then I?d actually prefer to see median days increase disease free survival as the way that you present the data. But again, bear with me again. CE-29, by my way of presenting the data, instead of a 47 percent decrease in risk, 6 percent of men who were treated with this drug actually end up benefitting at four years. That?s how I actually would prefer to think of it, and on CE-33, instead of a 35 percent decrease in risk, I?d prefer to say that 5 percent of men treated over four years benefit. And I think that by doing my 5 percent in the percent, I?m actually statistically averaging over the population where I?m talking to one individual man, very much in the same way you were talking, Dr. Blackledge. Thank you.

DR. CARPENTER:  I want to disagree pretty strongly with that interpretation of the data. These absolute risks are on the order of magnitude absolute which you?ve seen with Tamoxifen, you get a much more informative effect of the therapy by looking at a difference in the medians as you suggested. If you discuss it with patients, you also get very different reactions, and this has been done both ways.

It?s likely that you get a benefit across most patients, and the ?- if you analyze the curves the other way, you probably do get a sum of benefit. If you look at your progression free survival median differences, which are not presented here but which actually be quite ?- you couldn?t use medians but you have to cut it perhaps the 75^th percentile because it?s early. You?d see very dramatic differences which have spread through the population. And the vertical difference greatly under-estimates the population benefit.

DR. BRAWLEY:  Perhaps I didn?t say it clearly. I wish they had presented the data by progression free. I?m wondering why it wasn?t. And I still have a very open mind as to whether this should be favorably moved upon or not, but I wonder one, why not the progression free analysis. And two, why spend so much time talking about the relative risk and not talking about absolute?

DR. CARPENTER:  Well, the absolute risk is on the same order of magnitude as I?ve seen with treatment of node negative pre-menopausal breast cancer, which this situation has a lot of ?-

DR. BRAWLEY:  For survival, not for recurrence. Correct? Now if this were survival, if these numbers were in survival, I?d say this is a slam dunk. This is easy.

DR. CARPENTER:  Yes. It?s too early to present this for survival. In general, the survival benefit has been in breast cancer about half that. The absolute differences seen in recurrence fairly consistently across the board in the overview. We don?t know if that?s what?s going to happen in this population because there?s not any long experience to make a judgment on that. But the ?- even the magnitude of reduction in the lapses on this order of magnitude in young women with node negative breast cancer.

DR. GEORGE:  I wasn?t going to comment on that, but since Dr. Brawley brought it up, I?ll mention it. Both are correct ways to look at things. It is a difference in emphasis. The relative risk is an average over time in these time to event things, but may translate into a very small absolute risk at the end of the day. But I have a couple of questions.

One is, I can?t help asking since I haven?t ?- we?ve heard a good explanation of how these studies differed with respect to the patient populations. I had a question as to why, not that it might affect the decision today, but it?s just ?- it?s a question left hanging for me. Why were the studies designed this way? Why didn?t you try to get more comparability among the studies, watchful waiting patients from the U.S., lower risk patients from the other countries. Is there some explanation for this?

DR. KENNEALEY:  Sure. Let me start by answering the final part of the question first. The U.S. trial was actually designed as a true adjuvant study, and therefore, watchful waiting patients were not entered into the U.S. trial. And the overall objective of the trial program was to get patients over the entire continuum of prostate cancer, and that?s why there were different entry criteria over the three trials.

DR. GEORGE:  Yeah, but am I missing something here? Wouldn?t it have been nice to have ?- at this point to have had watchful waiting patients from the U.S. on this study to sort of add to the evidence?

DR. GEORGE:  All right. I don?t want to beat that. It was just something bothering me.

Now the other question has to do with something I?m concerned about, is the follow-up information, the length of follow-up.

DR. GEORGE:  One sub-part of that is the ?- you mentioned it was a meeting of three years, I think, follow-up.

DR. GEORGE:  And then you said that everything remains the same at 4.2 years, but have we seen that data, or is that just what you?re saying?

DR. KENNEALEY:  The data ?- the three year median follow-up data was part of the original submission. The four month safety update was submitted to the FDA more recently than the submission, and that included data up to 4.2 years for safety. At the request of regulatory agencies outside the United States, we?ve looked at that data, and the efficacy data have not changed.

DR. GEORGE:  And the ?- but we?re just taking your word for it, I mean.

DR. GEORGE:  Okay. And the data presented, if I?m looking at it right, overall, if you just take all the studies together, about 11 percent of the patients have progressed, and the primary endpoint is time to progression. I?m just adding up all the studies. And it?s strikingly different, of course, by study. In the U.S. only about 5 percent. And this is over a time frame that I would just guess, although it wasn?t presented, what the median time to progression, again over all the studies, would probably be something on the order of 7 years, sort of a median time. So all those things put together, I mean, this is pretty early. Even though you?ve got some strong differences in the 24 and 25 study, this is early in this ?-

DR. KENNEALEY:  This slide shows the progression of events over the entire clinical program at the 4.2 year median follow-up, and we?re now up to about 17 percent progression rate. Again, the hazard ratio and the statistical significance is really substantial. At the three year median follow-up, there were 14 zeroes after the decimal point. There?s a few less now, but I mean, it?s still very statistically significant. And despite the fact that there are less than 20 percent progression events, it?s our belief that these data are unlikely to change.

DR. GEORGE:  And you will ?- you are still following the patients.

DR. KENNEALEY:  All patients continue to be followed for progression and survival in this trial, but again, the data are so compelling at this point from a statistical and clinical standpoint, that we believe they?re unlikely to change.

DR. SHAMES:  Yes. WE have now reviewed the 4.2 year data that was not in the supplement we?re talking about today at all. It was the earlier data. We actually, just to explain what happened, we offered the sponsor the option to resubmit this with the later data, and they elected to go this route instead, so we have not reviewed that data. And actually, it?s other information I see this morning that we have not had a chance to review.

DR. KENNEALEY:  Okay. I?m sorry. I didn?t hear the final portion of your question. What criteria were used for ?-

DR. KENNEALEY:  So the question is criteria for withdrawal. I have to apologize. I am directly under the blower so it?s sometimes difficult for me to hear, so if I ask to repeat the question, that?s why.

DR. KENNEALEY:  Okay. I?ll ask Dr. Charles Morris to respond to that.

DR. MORRIS:  The number of patients who had additional therapies introduced within Trial 23 was 10 percent on the ?- or 9.5 percent on the placebo arm, and 7 percent on the Casodex arm. The information that we have suggests that that was in response to a rise in their PSA. The intent of the trial, as you know, was to follow until clinical objective progression.

DR. ALBERTSEN:  So that?s 9.5 percent of 1,000, are you telling me, of the 1,645, or ?-

DR. ALBERTSEN:  So you?re telling me roughly 164 patients were withdrawn from the placebo arm, and slightly less from the Casodex arm because of a potential rise in the PSA, or some other explanation that would justify additional therapy.

DR. MORRIS:  That?s correct.

DR. KRIST:  One of the things that I?m interested in is the generalizability of studies 24 and 25 to the U.S. And I?m interested somewhat from the other end. I?m a family physician, and we?re part of a practice-based research network. And within our group, most of our 50 to 70 year old patients, about 70 percent of them opt to get screened for prostate cancer.

DR. KRIST:  And when looking at the characteristics of the patients with prostate cancer in Trial 24 and 25 versus 23, 24 and 25 look more advanced. I?m interested in what some of the screening and diagnosis practices are in the countries for 24 and 25, and how that differs from the U.S. And then what component of how ?- what?s been presented here is that Trial 23 was designed to look at earlier prostate cancer, but I?m interested in what component of Trial 23 showing earlier prostate cancer is more a reflection of different practices in different countries.

DR. ANDERSON:  Thank you very much. There are clearly Transatlantic differences in terms of early detection for prostate cancer. Whilst there are well-established recommendations for screening in the U.S.A. from the American Cancer Society, I understand, the American Urological Association, in the U.K. that is not the case. It was only very, very recently that the National Cancer Plan was instituted. It?s now advised that patients over the age of 50 can ask their family practitioner to have their PSA checked. The family practitioner is instructed not to raise the subject with the patient if he doesn?t.

DR. IVERSEN:  Yeah. With regard to the part of your question addressing whether there?s a difference in the way the disease appears across the Atlantic, I would say that there?s absolutely no evidence of a Transatlantic difference in tumor biology. An aggressive PSA-based detection strategy in the U.S., combined with a long natural history of the disease has introduced a lead time with more small tumors being detected in this country. However, the more than 30,000 American patients dying from the disease, and the more patients suffering from the morbidity, pass through exactly the same stages and phases of the disease as European patients do. And it is my belief that these patients, as their European counterparts needs and deserves all the best treatment options possible.

DR. KRIST:  Well, it does. And certainly there is a component of a lead time bias. There?s also probably a component though of a prognostic bias if you?re having higher screening than that.

DR. KRIST:  You?re going to find more clinically insignificant cancers. Once again, I?m interested some if the difference in the tumor characteristics between 24 and 25 versus 23, was more relation to the inclusion/exclusion criteria.

DR. KENNEALEY:  Okay. Sure. I?m going to ask Dr. See to come up to the podium in just a moment. I think what we need to explain more fully is the difference in the patient population who are candidates for Trial 23, and the totality of patients with prostate cancer who are treated surgically or with radiation therapy in the United States. It?s actually a bigger group, and that matches 24 and 25 quite well. Let me ask Dr. See to expand on his practice.

DR. SEE:  I think that the differences that we?ve observed in the risk profile, if you will, across the overall clinical trial program are driven by enrollment criteria. But in fact, those enrollment criteria were intended to capitalize upon differences, if you will, that existed in 1995 across the different nations participating in this clinical trial program.

DR. SEE:  Okay. The PSA was measured in all three trials every three months, and we looked at the differences among ?- we looked at each of the three trials to see if there was a difference in frequency that might create some bias, and there was neither a difference in frequency across the trials, or between the arms in the trial. And there was no recommendation from the protocol as to what action to take upon the finding ?- upon the results of the PSA measurement. A PSA rise was not considered to be ?- for the sake of the protocol was not considered to be evidence of objective progression, because the FDA did not and does not recognize that as a valid endpoint.

DR. REDMAN:  A follow-up to some of the points. Specifically, Dr. See?s Slide 18, where there was a difference in the median going to another line of therapy. And the question I have with that is that since survival isn?t a question, what was the difference between the Casodex arms across the trials and on each trial, the Casodex arm going onto an alternative form of therapy, and the patients who were not on the placebo going onto an alternative form of therapy. What is the median difference in that time in months?

DR. KENNEALEY:  So the question you?re asking is the median ?- again, I apologize.

DR. REDMAN:  But you have no ?- forget the median then. You have no difference in months? Any idea?

DR. MORRIS:  Well, at this point in time, obviously, we haven?t reached a median time to event. No.

DR. REDMAN:  Okay. One other follow-up question. For U.S. in watchful waiting, of the 20 percent or so of patients who go on watchful waiting, is there a sense of how many of those is a physician?s decision based on the fact that definitive therapy would not affect survival, and the patient?s request to do that?

DR. PAULSON:  Watchful waiting as it?s practiced clinically among urologists in the United States is largely a patient-driven initiative. It?s a patient-driven initiative based upon their own assessment of the risk of their disease, and also driven by their assessment of a need for lifestyle maintenance.

DR. KENNEALEY:  I think that?s one of the important reasons why we are looking to seek an indication in watchful waiting, because there?s a clear difference in the tolerance of Casodex, versus as Dr. Soloway mentioned, what is sometimes used in this setting, which is an LHRH analog.

DR. SCHOENBERG:  Yeah, I think this is a follow-up, actually, to a number of previously asked questions, but I?m curious about the definition of the population in the studies performed of patients who would be candidates for watchful waiting, because I think in contrast to some of the remarks that have been made today, my understanding in U.S. practice is that watchful waiting is offered to a very specific group of individuals. It is, although clearly impacted upon by patient preference, not solely driven by patient decision-making. I wonder if you could illuminate that for me? I have another smaller question after that.

DR. PAULSON:  Clearly, there are patients who have significant competing risks of death, who would be dissuaded from choosing some form of active therapy because it would be felt that their lifetime ?- their life expectancy from competing risks would not be affected. And we usually use, as you know, the ten year interval before we choose a therapy of curative intent. Physicians, I think, may very well sway patients one way or the other, but at least in our practice when we discuss watchful waiting, it?s largely a patient-driven decision.

DR. KENNEALEY:  I guess I?d have to say we would like to be able to do that, but the practice of medicine has not advanced far enough to say with certainty what the biology of a prostate cancer, what the biological progression of the prostate cancer is going to be based on a single point in time, so it?s not quite as easy to address it that way. We?re simply not quite smart enough. Let me ask Dr. Paulson to elaborate on that.

DR. PAULSON:  There certainly have been a series of risk factors identified which would indicate that the patient is at minimal risk for dying of their disease within a ten year frame. And Dr. Albertsen?s group published some very nice data on that showing that if your PSA ?- I mean, if your Gleason Sum is 6 or less, that your probability of dying of prostate cancer within a ten year span is relatively small.

Unfortunately, I?m not quite sure how many of those patients subsequently went on to have some form of castration-based therapy to extend their life span during that interval.

DR. PAULSON:  As you?ve correctly brought up, Dr. Schoenberg, there is an intermediate grade or an intermediate survival expectation for patients that have Gleason Sum 7 disease. There?s a fair amount of controversy as to whether if it?s Gleason 4-3 or 3-4, depending upon the predominant volume of disease, the relative risk. However, the data in radical prostatectomy series would state that if you have margin positive disease with Gleason Sum 7 as your pathology, you have, I believe it?s about a 50 percent probability of having a PSA failure within five to seven years. And the survival data subsequently with secondary therapies I believe has pushed that to somewhere around 14 to 15 years. But with margin positive disease, that?s a different risk group than just Gleason 7 that is organ confined, as you know very well from the data at Hopkins.

DR. BRAWLEY:  Yeah. I?m somewhat motivated by the old data via early versus late prostate cancer treatment papers in metastatic disease. Do you plan on continuing to follow these groups to determine the overall survival and mortality rates of people who were treated early with Casodex versus those who had delay therapy because they were randomized to get the placebo?

DR. KENNEALEY:  We have information on all patients on first treatment following ?- first treatment for prostate cancer following withdrawal from therapy. We don?t have information on subsequent therapies beyond that, and we will be following all patients for initial progression, initial objective progression and survival.

DR. BLAYNEY:  The analogy has been made to breast cancer, and also your again stimulus for watch and wait treatment was a rising PSA. In breast cancer, it seems to me the ?-

DR. KENNEALEY:  I?m sorry. Go ahead.

DR. BLAYNEY:  We heard in Scandinavia that when patients had an event, and often it sounded like rising PSA was what triggered the enrollment in this trial. The analogy would be the rising tumor marker in breast cancer. And to my understanding, rising tumor markers in CA15-3 or 27-29 in breast cancer have not been useful in prolonging survival, when treatment is inaugurated based on a rising tumor marker, so I think your ?- I hope that you?re correct in that there is going to be some clinical benefit and some survival benefit to inaugurating in the watch and wait population treatment with your drug based on a rising PSA. But I think the analogy in breast cancer ?- the analogy to breast cancer remains to be proven.

Secondly, the issue of gynecomastia, which was ?- and breast pain was a big issue for treatment withdrawal. When I was trained many years ago before I had all this gray hair, there was breast radiation to men who were going on castration as a preventative measure for gynecomastia. And I don?t know if any of your experts have any experience on low dose breast radiation to prevent that.

DR. ANDERSON:  Thank you very much. I?ve been using Casodex both in the trial setting, and also in my clinical practice in the U.K. for some years. You?re right to raise the issue of gynecomastia, and it?s something we address with the patients early-on. We?ve seen the figures. We?ve seen that it occurs.

In my experience, it?s not a major problem, and we do not see patients withdraw once they know the benefits of treatment. I think the high withdrawal rate we see, the EPC data, reflects unknown efficacy of the drug, but recognized toxicity.

What I see in my clinical practice is a different thing. I see patients who we address up front with gynecomastia is an expected event in someone who?s on Casodex, but it?s manageable, and it should not cause a problem. If the patient has an issue, then irradiation of the breast tissue before we start treatment is something we would discuss with them. We re-address the issue once they?re stabilized on treatment, and withdrawal of treatment is always an option. But what I?m absolutely persuaded by is the benefits that the patient perceives in terms of delaying disease progression far outweighs his concerns about the potential toxicity of the treatment. And I feel, therefore, that in my practice it?s very important to have that option to offer the patient, and involve him in the discussion of the relative benefits of delaying disease progression against possible side effects. He should have the option.

DR. BLAYNEY:  I agree with the involvement of discussion. Finally, Dr. Kennealey, in your slide CI-10, which was the chronology of development, there?s a compound and it may come up later in the morning or earlier afternoon about your dealings with the FDA. Were you meaning to imply that there was a change in the rules or the change in the parameters which you used, or you were required to use to develop this compound?

DR. KENNEALEY:  You want to see the results of the FDA ?- the endpoint of time to bone scan progression? Yes, let me just bring that up for you. This slide shows the bone scan progression, the endpoint requested by the FDA. As with the primary analysis, there is an overall benefit in favor of Casodex with a reduction of 37 percent. That was seen primarily in Trials 25 and 24, as was ?- you know, this really parallels the primary analysis, and actually confirms that analysis.

DR. HANNO:  I have a question with regard to indication number two. I know you pick your words very carefully when you go for an indication, and here the indication is localized non-metastatic prostate cancer in patients for whom therapy of curative intent is not indicated. To me, and I think to a lot of urologists, that would mean patients who don?t have a ten year life expectancy, because therapy of curative intent would be indicated in the other people, whether or not they chose to accept it.

My question is, would it be better to change that to not planned, and are you looking for therapy in patients who have locally advanced disease, and that?s why they don?t have therapy of curative intent planned, or they don?t have a life expectancy to warrant it, because those are two very different groups.

DR. KENNEALEY:  Yes. We?d love to ?- we submitted the original indication last year, and as we reviewed our data and reviewed our plans towards coming here, it became very clear that indicated was not the right word. And that intended, either by the patient or physician, would be a much more appropriate word in our indication. And that would be something that we would want to change.

DR. MARTINO:  I want to add something to this comparison of breast cancer to prostate cancer, and it?s in reference to the use of tumor markers. There actually is a trial that was reported and done in Europe. It was presented in poster form at it was either ASCO or San Antonio about five years ago. It was a patient population that was being followed by tumor marker, and the tumor marker used was a 15.3.

DR. BRAWLEY:  I mean, granted, and I accept that it?s very likely that knowing that your PSA is down improves your quality of well-being. I accept that. The other question is, and again, I?m heavily influenced by Dave Byers? studies that showed that early treatment had no greater effect on survival than later treatment granted in a different group of individuals or different patient population. And it was 30 years ago with different drugs, but I?m wondering do you have any data on response to LHRH agonists in men who progressed after being on Casodex?

DR. ANDERSON:  I have been using Casodex long enough to see men progress, and standard practice for me would be to institute them on an LHRH analog. Response rate is about a third as measured by PSA response, and I think that?s in keeping with any other first line hormone treatment where the patient escapes hormone control.

I don?t know of any sequential studies to say that one sequence is better than another, but where I draw reassurance is when I look back at the 306-307 data, where patients were either treated with Casodex, or with castration. Now in those studies, both patients progressed in either arm. They were treated with alternative second line treatments, and we know that there are no major survival differences for the two groups, so I?m reassured that there?s no difference in the biology of the tumor after the patient has been treated with Casodex.

CHAIRPERSON PRZEPIORKA:  A follow-up question while you?re standing, sir. Did I understand you to say that when you had patients on Casodex, you would start them on second line salvage therapy on the basis of a PSA, and follow the PSA only?

DR. ANDERSON:  I?m sorry. Could you repeat the question? It is noisy here.

DR. ANDERSON:  No, you did not hear me say that. When a patient fails hormone treatment, it?s a full clinical picture. The patient either becomes symptomatic, he has a rapidly rising PSA or has changes perhaps on his bone scan that would indicate treatment. It would just ?- it would not just be on a PSA progression.

DR. ALBERTSEN:  Just a quick follow-up on Dr. Brawley?s question, again to help me understand some of the British data. In England, on Trial 24 when you started patients on Casodex, and when they failed Casodex, they moved on to anti-androgen therapy, was the length of time of response on anti-androgen therapy equivalent to what might have been perceived had they started initially on anti-androgen therapy and run the whole way? Basically, what I?m trying to get at, was there additional response time by sequencing the drug, rather than starting initially on anti-androgen therapy and just waiting to see what happened?

DR. KENNEALEY:  Let me just clarify. You meant after Casodex, you mean LHRH. You didn?t mean anti-androgen therapy?

DR. ANDERSON:  I don?t have any personal data to support that, neither do I know of any studies that would support it either. It?s relatively early to be able to answer that question confidently, but I do reflect back to the 306-307 data, where those sequences occurred in each arm where there were no differences.

DR. ALBERTSEN:  Where I?m coming from is, it would strike me from a clinical perspective that Casodex in this country would probably substitute in many instances for LHRH agonist therapy very early on in the practice. And so what I?m trying to grasp is, are we likely to see an extension of survival which we haven?t seen certainly in Trial 23. And I?m not saying you have data here. What I?m getting at as you?re talking about your clinical practice. Is this ?- because what you?re quoting is the survival on Casodex versus standard castration therapy. There was no survival difference, so what?s striking me is that Casodex becomes a substitute, but doesn?t increase survival. Is that a wrong conclusion?

DR. ANDERSON:  Well, I think that?s probably accurate. There?s probably ?- there?s no evidence to suggest one way or the other to my mind.

CHAIRPERSON PRZEPIORKA:  Other questions? Hearing none, what I?d like to do is take a break for 15 minutes, be back here at 10 minutes after 11. Thank you.

CHAIRPERSON PRZEPIORKA:  Okay. We?ll start with the FDA Presentation. Dr. Daniel Shames.

DR. SHAMES:  Good morning. I?m Dan Shames. I?m the Director of the Division of Reproductive and Urologic Drug Products at the FDA. Before I start my remarks, Dr. Hoberman, our statistician, has asked to make a few comments.

DR. HOBERMAN:  Thanks, Dr. Shames. In retirement, I thought I could be quiet as a mouse, but unfortunately not. The Sponsor made comments about statements I made in my review, and I just wanted to clarify a couple of them because they were made out of context, and I want to provide the full context for the record. Also, I do want to make it clear that I?m a statistician, I?m not a clinician.

The first statement they made was they concluded from my review that Casodex reduces the risk of progression regardless of primary treatment. That was true in Europe, but I wish they had made it clear that I don?t think that there?s a shred of evidence data that supports efficacy in the United States.

The second point is that patients in Trial 23 are at low risk for progression. I think that was a consensus that we all had that what they failed to mention was that I did do an analysis which tried to account for the lack of treatment difference in Europe, in the United States, based on the different distributions of prognostic variables. And for reasons unknown to me, was unable to account for that difference. So even though that I understand that in the clinical community, I guess it?s sort of common wisdom that one of the reasons that the response rates were so low in the United States in comparison to Europe was that the course ?- where the patients were in the course of their disease, statistical methods surprisingly couldn?t confirm that, and I was disappointed.

The other thing that they quoted was, ?Objective progression may be surpressed in Trial 23 due to U.S. clinical practice.? I have no business saying that as a statistician, except as in the context of the review, it was one of a litany of different things that might have accounted for the differences between Europe and the United States. And one of the reasons I mention that is because I was involved in the application of Rilutek, the only treatment on the market for ALS, in which there was evidence of efficacy in Europe, but there wasn?t a shred of evidence of efficacy in combined Canadian and North American trials, so I want to make that clear.

There are a couple of very quick things that I would like to point out also for the record, that I want to emphasize, although I know that Dr. Monroe is going to refer to these. One is that when the sponsor showed the slide about the results of 306 and 307, they combined the results and showed that they didn?t make a confidence interval for equivalence or non-inferiority. But the whole point of that was that the FDA decided that those trials could not be combined because of positive results in one, negative results in the other. And when you put them together, there was a wash, and that?s important to point out.

The only other thing I wanted to point out is that this business about the modeling in order to find a subgroup in which there would be efficacy in a high risk subgroup, that modeling exercise is, in a sense, a reasonable thing to do, but I think you ought to keep in mind that when labeling the drug, it?s a very risky business when you?re talking about what are going to be the cut-offs in the prognostic factors, which have been identified in the model. So it may be that they can find a difference in a subgroup when restricted to certain risk variables like PSA end stage, but how the Gleason Score came up, how the PSA score came up as the boundary for who should be eligible for Casodex 150 is certainly not clear, and those cut points must be due to a degree of data dredging that comes from the model that was used to find the prognostic factors.

DR. SHAMES:  Hello. I?d like to first thank the Division of Oncologic Drug Products for the close cooperation and advice we?ve had from them during the course of our reviews of the various applications for Casodex 150. And I also appreciate ODAC for taking the time to advise us regarding this challenging issue being presented before us today.

I would say that whatever differences any of us have we, I?m sure, all agree that these issues ?- the general issue of prostate cancer and the specific issue before us today is extremely challenging.

The FDA?s presentation will consist of the following. I?m going to discuss background and review issues. Dr. Monroe, who is the medical team leader for this product, will review the clinical trial data, and then I?ll come back and summarize the review issues, and introduce the questions.

As you all know, prostate cancer is a disease that can exist along a very wide continuum, from microscopic clinically inapparent, to advanced hormonally insensitive disease. For many elderly men, the disease exists in our bodies for many years until they die of something else. Therefore, the treatment for prostate cancer must take into account the dictum ?primum non nocere?, or first do no harm.

We must be careful not to expose patients with early prostate cancer to unnecessary toxicity without proven benefit. Recent evidence indicates that the most precise method to predict a patient?s disease stage and, therefore, decide on appropriate therapy, such as surgery, radiotherapy and others are outlined in the trial, which I will mention further.

It should be noted that Gleason Scores, which are based on glandular patterns of tumor under low power magnification is essentially the most accepted method of pathologic grading in the United States, were not used in the non-U.S. trials. Traditional pathologic grading was used in the studies outside the U.S., and the results were extrapolated into Gleason Scores. It was more than ?- really there was no central laboratory. The pathology was translated into Gleason Scores, and someone did make the comment about the translation of 7 versus 7, 8, 9, 10, and we also had that problem.

From my reading of the literature, it appears that the best way to ?- at the moment to define cancer sub-populations are the clinicals, the Gleason Score and the PSA. It is true that they all are independent predictors, but the papers that we read tell us that the three together are the best predictors, and add the most precision to the staging and prognosis paradigm.

The results from the sponsor?s combined analysis revealed that Casodex failed to meet the pre-specified parameter to declare non-inferiority castration. In other words, demonstrate non-inferiority to castration in terms of survival. Casodex was to be no more than 25 percent worse than castration with respect to survival. However, the combined analysis as previously shown the confidence interval was 36 percent.

In recently published large meta analysis of single therapy androgen suppression in men with advanced prostate cancer which was published in April of 2000, the author stated, ?The evidence from eight trials involving 2,700 patients suggests that non-steroidal anti-androgens were associated with a lower overall survival compared to castration. The data from the Casodex trials in the meta analysis, especially

Those who treat patients with prostate cancer are familiar with the phenomenon of anti-androgen withdrawal syndrome, which is a paradoxical anti-androgen stimulation of prostate cancer, perhaps resulting from prostate cancer receptive gene mutation. I don?t think we?re absolutely sure. A similar mechanism could be in operation to explain the survival disadvantage of Casodex compared to castration in patients with advanced prostate cancer.

Now, it is true that the Trial 306 and 307 were on more advanced patients, and there was only some overlap. But, quite frankly, we don?t know what long-term treatment of Casodex is going to have ?- the effect is going to have on survival because we?re very early in the process here. And considering the data we have, this is an issue of great concern to us.

I?m going to now tell you what the review issues that Dr. Monroe is going to speak about in just a minute or two. Our efficacy concerns have to do with the fact that the trials are really not long enough to demonstrate enduring efficacy as we have defined them. The Gleason Scores are ?- we consider invalid in Trials 24 and 25, not only for the technical reasons I told you, but also -- as Dr. Monroe will point out -- there were inconsistencies between the clinical stage and outcomes and pathology between the U.S. and the non-U.S. trials.

Also, the data proposed to support efficacy in the U.S. is based on a retrospective subgroup analysis. As far as safety, there is a very high discontinuation rate from adverse events, a high instance of gynecomastia and breast pain. And as you will learn, for many of these people the gynecomastia is irreversible. There?s also some concern about liver toxicity.

Additional review issues were the issue I just discussed, which is questionable. We?re not sure about long-term survival, even in this patient group. We do not believe it?s been demonstrated that there?s a quality-of-life or sexual advantage clearly demonstrated, especially a quality-of-life advantage regarding Casodex. And the three trials that we are presented are heterogeneous populations with different treatments. And the non-U.S. trials reflect different practice patterns.

DR. MONROE:  Hi. I?m Dr. Scott Monroe from the Division of Reproductive and Urologic Drug Products. I was originally going to start by saying "good morning," but we?ve almost reached noontime because of the lengthy -- but I think very important -- discussions that we?ve had prior to this time. Earlier this morning, the sponsor presented their data and their interpretation of these data that were submitted in support of the two indications for Casodex that you see showing on the screen in front of you. In support of these indications, as you?ve heard, the sponsor conducted three multi-center randomized placebo-controlled clinical trials.

One of the most important characteristics of all patients across all trials was that they were all supposed to be negative as far as bone scans. That?s a very important unifying characteristic that crossed all of the trials. An important difference was that treatment in Trial 23 was limited to two years, while treatment in both Trials 24 and 25 continues. And this also applies to the adjuvant patients in Trials 24 and 25, so you can see there is a difference between these trials. It?s not just whether they?re watchful waiting or not, but a decision was made in one adjuvant group treatment would be discontinued, while in the other two trials which also have adjuvant therapies, treatment is ongoing.

Since the sponsor has reviewed the overall design of these clinical trials, their similarities and differences, and the overall efficacy and safety findings, I will avoid re-reviewing these topics. Rather, I?ll try to limit my presentation to significant clinical review issues. And these review issues include differences between the sponsor and the division regarding study endpoints and data analyses, and interpretation of clinical findings.

This discordance is also apparent when we look at tumor differentiation and pre-treatment PSA values. And once again, we can see -- looking at median PSA values at the bottom of that slide -- they?re lowest in the U.S., highest in the non-U.S. studies, and this is going exactly the opposite than what was determined to be the histopathology of these tumors.

There is also a discordance between the baseline histopathology or Gleason Scores and disease progression. On this slide, we?ve listed for you, for each of the trials, the percentages of patients in the prostatectomy group who had Gleason or tumors reported as poorly differentiated, or Gleason Scores of 7 or greater. And whether we?re talking about 7s or 8s, I don?t want to belabor that point. We have to go with the data as it was presented to us by the sponsor. But I think the important point I?m trying to make here is that there is less tumor dedifferentiation, at least as assessed by the pathologists in the various trials in the non-U.S. studies; namely, those were better differentiated tumors, yet we can see that the incidence of disease progression as assessed by a positive bone scan was much higher in the non-U.S. trials. And this certainly was problematic for us in terms of interpretation of the baseline disease characteristics of these patients.

We discussed this somewhat earlier this morning. I?d like to expand on that just a little bit here. The sponsor preferred a time-to-disease progression endpoint, where progressive events were based either on local or distant events of disease progression confirmed by bone scan, x-ray, CT, MRI, ultrasonography, or biopsy, or death due to any cause in absence of progression.

The rationale for the FDA?s preferred endpoints and analysis was based on concern acknowledged by the sponsor that blinding could not be maintained because of the anticipated high incidence of gynecomastia and decreases in serum PSA in the Casodex-treated patients. We felt that the inability to maintain blinding could result in significant assessment by us. In addition, specific criteria for local disease progression did not appear to be provided in the study protocols, and there was no central blinded review of events classified as progression, whether they be bone-scan-documented or otherwise. Because all the protocols mandated a bone scan at two years post- randomization, it was believed that this endpoint, along with death, would be least subject to possible assessment bias.

In each of Trials 24 and 25, there was statistical evidence that treatment with Casodex delayed disease progression. In Trial 23, however -- the only trial that enrolled patients in the U.S. -- there was no significant difference between the two treatment groups. And as you saw earlier this morning, the same conclusions regarding the effect of Casodex treatment on disease progression were obtained using the sponsor?s preferred endpoints and analyses.

The sponsor?s original proposed indication for Casodex 150 milligrams was immediate hormonal therapy, or adjuvant therapy to treatment of curative intent, patients with non-metastatic prostate cancer. Such an indication would encompass virtually all patients with non-metastatic prostate cancer, and was not, in our opinion, supported by the submitted data.

The data supporting this change for the adjuvant treated patients, based on the FDA preferred endpoints, are shown on this slide. And what we can see here -- and let?s focus primarily on Study 23 -- that by this criteria, there were very few patients in the U.S. who had disease progression, as assessed either by a positive bone scan or death within the two-year period after randomization. As you can see, there are only four in the Casodex group, six in the placebo group, and clearly these were too small to make any conclusions regarding the potential benefit of Casodex in this group.

We noted during our review of the sponsor?s background document that the definition of a patient of high risk for disease recurrence appears to have expanded somewhat, and this new expanded definition is listed in the lower portion of the slide. These patients remain those with locally advanced stage T3-4 disease and detectible post-surgical PSA values, but also include pre-surgical PSAs of greater than 10, or a Gleason of 7 or greater. And the criteria for a patient treated by radiation has been loosened somewhat, so that a pre-radiation value of 4 would qualify an individual for being at high risk for recurrence or disease progression.

Earlier today, the sponsor showed you a number of Kaplan-Meier curves based on these definitions. And I want to first bring to your attention that the prostatectomy and radiotherapy labels are reversed on this slide. The data in the upper half of the slide are from the radiotherapy group, those in the lower portion from the prostatectomy group. And we agree with the Kaplan-Meier curves that you saw earlier, if you accept the sponsor?s endpoints, that in the high risk group of patients as defined -- as I showed you just a moment ago -- there were statistically significant reductions in the proportion of patients who had disease progression.

Now the earlier analyses were based on very small numbers, and their significance is questionable. But here, presumably we have 712 patients that are at high risk for disease progression, yet we don?t see any benefit at this stage of Casodex therapy. And this, again, underlies the concern of the division that logic would say you could just transfer these data and information from the European studies to the U.S. studies, but it just hasn?t worked out when we apply them, or the sponsor in this case has applied them to the actual data.

I do not have a slide for the radiotherapy patients. The sponsor actually showed one, I believe, earlier where they showed that in Trial 23, using the definition of high risk for recurrence, there were only four events -- I believe -- in the U.S. trial, if I remember. And if I also remember correctly, of those four events, two occurred in the Casodex group, two in the placebo group. So once again, I?d say there is a problem as to what is happening with the U.S. patients.

Now subsequent to the division?s not approving the NDA for Casodex 150 milligrams, the sponsor revised the proposed indication for the "watchful waiting" or monotherapy group. And in the revised indication, as you?ve heard this morning, the sponsor now recommends that immediate treatment or monotherapy be limited to patients with localized Stage T1/T2 non-metastatic prostate cancer. It?s our understanding that this modification was made because of the concerns of the division that Stage 3 and 4 patients were very similar to those in the previous studies, where there were concerns about survival in the Casodex-treated patients.

Sponsor has also shown or at least told you that the effects of Casodex treatment in these earlier stage patients was statistically significant, and that we do agree with that based on the sponsor?s endpoints and time-to-event analysis. But if you apply to these patients the endpoints and analyses that the FDA feels are more appropriate because of the concerns about assessment bias, we can see that there are certainly strong trends in support of Casodex, but that the upper bound of the 95 percent confidence limits for the odds ratios for both of the studies extend above one. Whether these will drop down below one or whether they have in your more mature data which you?ve not seen, we just don?t know. But at least on the data that we?ve had a chance to review, neither of these studies would have crossed the bound that would have met the criteria for statistical significance.

These are the baseline disease characteristics for the patients in the T1/T2 "watchful waiting" groups in Trials 24 and 25. And at first glance, one would say these Gleason scores look like those that are seen for U.S. patients that frequently are managed by surveillance. But our concern is that, based on the data where we could compare tumor histopathology across the European or the non-U.S. studies and the U.S. studies, we felt that these patients had under-reporting for poorly differentiated tumors. Clearly, there was a difference in reporting, so that even though these Gleason scores would be very comparable, I believe, to what type of patient might be a candidate for "watchful waiting" presently in the U.S., we believe that these scores are unreliable and, therefore, we can?t conclude that these patients had the same type of tumors as those patients that are frequently given the option, or at least advised that watchful waiting would be a reasonable option for them.

The last efficacy topic that I?d like to review with you this morning relates to survival. And what I?ve summarized on this slide are the percentage of deaths related either to prostate cancer or other in each of the three trials. The data in the upper half of the slide are those data that were submitted at the time of the initial submission as part of the efficacy component of the application. And they had a data cut-off date of June in 2000. With the safety update, we received additional survival data, which had a cut-off date of September, 2001, represented in the lower portion of the slide, and I think we ought to focus on those numbers since they are more current.

If we look just at prostate cancer-related deaths, we can see that there are small numeric decreases in both of Trials 24 and 25, but that in Trial 23, we see just the opposite. There?s a small numeric increase. I think the conclusion is that there?s probably no impact on survival at this time in patients with prostate cancer.

And similarly, if we look at other events we again see a little increase or a decrease, depending on which study we?re referring to. And if we go to the bottom, the data represented in yellow, we can see that the differences within any study tend to be very small. And we would concur with the sponsor that at this time there isn?t any evidence that treatment with Casodex is having any impact on survival either way.

So, to summarize what we?ll call unresolved efficacy issues, we can lump these into perhaps three broad categories. The first category concerns the maturity of the studies, and since only 15.6 percent of the patients using the sponsor?s endpoints and analysis, or 9.3 percent using the FDA?s preferred endpoints, have had an event of disease progression, we think that these are early studies. The results from these studies are quite early. And that the long-term benefit of treatment at this time is unclear in the absence of survival data, or a survival difference, or meaningful quality of life data.

The third area is what is the risk benefit ratio for immediate therapy or monotherapy in patients with localized disease. This time I?d like to spend just a little bit of my presentation to go over some of the safety observations. This slide shows the disposition of patients in each of the trials. And as you can see, the percentage of patients who terminated prematurely exceeds probably 30 percent or more in all of the trials. And in some trials, the percentage is higher in the Casodex, in others it?s higher in the placebo group.

If we look at Trial 23, the U.S. trial, we can see that the patient terminations due to adverse events in the Casodex group far exceeded those in the placebo group. What the reason for this imbalance is, we don?t know, except for the development of those adverse events, which we?ll discuss in a moment. And actually, adverse events exceeded in the Casodex group was a greater cause for premature withdrawals in all the studies. However, as you can see in the European studies, disease progression was a more common cause for withdrawal in the placebo patients than in the Casodex-treated patients.

The most common adverse events, as you heard earlier, are those related to the pharmacology of the drug, its anti-androgenic or its estrogenic activity. And as a result of the drug?s pharmacology, 73 percent of the patients across all the trials reported breast pain, and 67 percent reported gynecomastia. You can compare these to the much lower incidence in the placebo groups of 7 and 8 percent respectively.

Because of the high incidence of breast pain and gynecomastia, I?d like to focus on this adverse event in somewhat greater detail. The figure that I?ve taken from the sponsor?s integrated summary of safety in which the proportion of patients without event, in this case gynecomastia, are represented as a function of time. And, as you can see, that by approximately one year or so after the onset of treatment, about two-thirds of the patients have developed gynecomastia. This percentage increases slightly, but most of this occurs certainly within the first year of treatment.

A very high proportion, as I?ve mentioned to you earlier, of patients do experience gynecomastia or breast pain. We can see that across the studies, this was a complaint reported by 86 percent of the patients. And patients withdrew from the study because of gynecomastia in what we believe is a significant number. Across all the studies, 16 percent of patients withdrew because of gynecomastia or breast pain in the Casodex groups, compared to less than 1 percent in the placebo groups. And in the U.S. trial, I believe this number was approximately 20 percent, even slightly greater.

We?ve heard earlier this morning about quality-of-life issues, and in these particular trials there wasn?t any effort to really assess quality-of-life in any systematic manner. There was very limited quality-of-life data. There was data on maintenance of sexual function in the study that was conducted in Sweden. And I have to confess, I?m personally a little bit befuddled by these data, because the data that I show before you here, I thought I had taken directly out of the sponsor?s integrated summary of safety again, and this would suggest that in these patients, both in the placebo and Casodex patients, there?s a very significant diminution of sexual function over time. And in the bar graphs that we saw earlier, they didn?t give this impression at all, so perhaps we have misinterpreted these data. Perhaps you could explain the difference.

But if we interpret these data correctly, it appears that in both groups in this particular population as assessed by this instrument, there was significant decrease in sexual function over time. Whether one can put any credence on these data we don?t know, because of the rapid fall within 12 weeks of treatment onset in both treatment groups.

In these studies, the incidence of life-threatening or fatal hepatotoxicity was similar in both the Casodex and placebo treatment groups. However, as shown on this slide, there was an increase in what was defined as clinically relevant changes in ALT or AST or bilirubin levels in the Casodex patients relative to the placebo patients. And roughly anywhere from two ?- this difference was two- to four-fold higher in the Casodex patients relative to the placebo patients. There was also a greater percentage of Casodex patients who withdrew due to liver-related adverse events, perhaps two- to three-fold greater in the Casodex group, as well.

Life-threatening or fatal hepatotoxicity was rare and similar in both treatment groups. However, clinically significant or clinically relevant -- to use the sponsor?s terminology -- arises in ALT or AST values, and withdrawals due to hepatic adverse events were two- to three-fold greater in Casodex-treated patients.

At this time, I?d like to return the presentation to Dr. Shames, who will summarize our concerns about this particular application.

DR. SHAMES:  Thanks, Scott. First, I want to tell you where we are in a regulatory sense. We issued a non-approvable letter for this supplement involving Trials 23, 24, 25, and in that letter, we stated that we wanted to see more mature trial data to find out ?- to answer some of the questions that we have before us today. We also asked that, if it were possible for the sponsor to get the slides from the foreign studies and really do Gleason scores, but we believed that the essential issue was that these were sort of post hoc subgroup analyses, and that there was hypothesis testing, and they should choose well-defined successful subgroups, and perform well-controlled trials after the results that they?ve seen currently.

I just would like to take time to discuss ?- clear up an issue regarding the interaction of the FDA with the sponsor regarding these endpoints. In the notes that I can see as far back as it goes, there was disagreement regarding the ?- what we consider an objective endpoint, the protocol-driven bone scan as opposed to the more investigator-driven endpoints, which the sponsor used. That disagreement appears right from the very beginning, because we were concerned about the possible unblinding, probable unblinding, perhaps, of gynecomastia, and the fact that Casodex in some variable way in itself will reduce PSA.

Now let?s go on to the review issues, which are the core of the concerns that we?ll discuss in the questions. We are concerned about the small number of progression events, and even fewer survival events after three years, to draw conclusions about long-term use of Casodex 150. As mentioned several times, one of the key parameters used in the U.S. for disease staging -- the Gleason score -- was improperly used in the non-U.S. trials. In addition, there was disturbing inconsistencies between the pathology and clinical outcomes between the three trials.

Certainly on the face of it, you would, you know, consider that sexual function will be decreased in people who are ?- men that are castrated. But on the other hand, we have other issues when we?re using Casodex. And in fact, it?s very unclear whether -- when we?re dealing with quality-of-life -- we improve it with Casodex versus castration, or placebo. As a matter of fact, in the paper that I mentioned before in The Annals of Internal Medicine, April 2000, which looked at a large amount of analysis of androgen monotherapy in advanced prostate cancer, the authors concluded that treatment withdrawal, the most reliable indicator of adverse effects are less with LHRH agonists versus non-steroidal anti-androgens.

The review regarding safety is particularly concerned because this drug has the potential for being used in a very wide population. We had some other review issues. There is concern over the potential, as we?ve talked about, survival detriment is too early to tell, and we feel there may be evidence for us to be concerned about that, and there is some biological plausibility.

That?s the end of our presentation, and thank you.

CHAIRPERSON PRZEPIORKA:  Thank you, Dr. Shames. And the floor is now open for the Committee to ask questions to the FDA. And I?ll start by asking, could you explain briefly, please, why you would not accept x-ray results or biopsies as an endpoint for this study?

DR. SHAMES:  Well, I think you mean ?- well, because I think they were ?- some of those were investigator-driven, and we?re concerned that there may been some selection bias. You mean, x-ray ?- you mean a precipitation of the x-ray, or the biopsy occurred because of some change in PSA or knowledge of what the control ?- what the arm was.

DR. MONROE:  I guess I?ll expand on that just a little. The documentation for these other events was very inconsistent in the application. They weren?t assembled by the sponsor in any way that you could actually determine, in most cases, exactly what was going on. They were reported just as having occurred, and it was not possible to really decipher what these were. Some of these were local events, and some were distant events. And without having had more documentation as to the nature of these other events -- forgetting about whether they were driven by factors or not -- we just couldn?t place any reliability on them. And I did address the fact that there was no central review of these. There was no clear criteria in the protocols defining what some of these events needed to be. I believe some of these events may well have been enlargement of the prostate, perhaps, for those patients in the "watchful waiting" group. Yet, we couldn?t find anything that said it had to be an enlargement by a specific size or not. It seemed to be driven only by the assessment of the investigator that an event of progression had occurred.

DR. CHESON:  Well, I can see the issue that you?re raising about the size of the prostate, but a positive biopsy is a positive biopsy. No matter what drives it, it?s evidence of progression. And even a positive x-ray is a positive x-ray, but that raises the other point, which troubles me no end, that there wasn?t any central review of the x-rays. And I really find that hard to believe.

If I could ask one question which sort of came to me, because often we miss therapeutic leads because they?re hidden in toxicities and whatever. Did either the agency or the company look to see if there was any correlation between the development of gynecomastia and outcome, because this may tip you off as to whether this is really a biological effect, and may be a plus instead of its being a minus. But I?m agreeing with Donna, that it?s sort of bothersome that you?d sort of write off things which are not subjective, and are purely objective, like positive biopsies and things, which I think, you know, in all fairness probably are evidence of progression.

DR. KRIST:  Going on the theme about the sponsor's versus the FDA?s endpoints, to a certain extent I disagree. I mean, I do think an objective result is an objective result. There?s a higher risk of missing those objective results in the placebo patients because you might not look for it. But that?s something that I think is a difficult thing to think about.

I?m curious, though. You presented data showing that if you looked at just Trial 25 in the "watchful waiting" group, and looked at low-risk people by FDA endpoints, that there was no benefit. And then you also showed in 24 looking at the prostatectomy group who are high risk, and by sponsor?s endpoints, that there was a benefit. I?m curious for Trial 24, if you were to believe in the FDA endpoint, if high-risk FDA endpoint and prostatectomy, if there was a relative difference, I didn?t see that number.

DR. MONROE:  I think I?m going to have to ask you to go through that step by step, please, and then I can address each piece of it.

DR. KRIST:  The big question I had was that you showed a slide doing the subgroup analyses, and you showed that on Trial 24, in the patients who had locally advanced or the high-risk disease, who had prostatectomy, and you went by sponsor?s endpoints, that there was a difference in outcome between Casodex and placebo. And I?m interested in Trial 24 for that same group, instead of going by sponsor?s endpoint, but by FDA endpoint, for Trial 24 who are a high risk, who had prostatectomy ?-

DR. MONROE:  We?re talking about the adjuvant treatment prostatectomy patients.

DR. MONROE:  Okay. And on the slide that I showed, and we?re not set up to immediately go back, but I believe on that slide, which is ?- let me just find that for you so you can all look at it here. Were you talking about Slide 26, perhaps? Could you refer me to the slide on the handouts? That was the high risk prostatectomy FDA analysis. Is that the one you?re referring to?

DR. MONROE:  Well, there?s clearly nothing there. We could show you what those data look like if you use the sponsor?s analysis, if you wished. In other words, you would find events due to this category of other objective events, which would have an impact on the absolute numbers. Is that what you?re asking?

If you want to go to the backup slides, Randy. Okay. That?s the same data, I believe, but using the looser definitions where you would include all events. And you would see, we have more events, and we see that there are some numeric differences, only one of which by this analysis has a confidence interval that is less than one. But if you go to the next slide and look at what?s driving these, I think this may answer your question. That?s Study 23 at the top, so we can see there?s actually one more positive bone scan in the Casodex group than in the placebo group. There were two more other objective events, and I can?t tell you precisely what those were at this moment. And there appeared to be four deaths, but those deaths, three of the four had nothing to do with prostate cancer. So it seems that as you go into these subset analyses deeper and deeper with small numbers, you can come out with almost any kind of an outcome. And that is what?s being driven by these.

I think perhaps the most compelling data were those of the sponsors using the criteria that they are now using for high risk, where we have a lot of events occurring, but yet there was no difference between the two groups. And if we could go back to slide -- I guess it?s 30. And here we don?t have to bicker over whether we should or should not include these other objective events. This includes all of those other non-bone scan driven events, and there is just no difference in the ratios between the Casodex and placebo treated patients in Trial 23.

DR. CHESON:  I was ?- this is just probably not even a very smart question, but I just want a point of clarification. There were those patients in the ?watchful waiting group?, who were treated on the basis of the PSA that went up, which have been alluded to as, perhaps, protocol violations. How did you handle them in the analysis? Did you include them? Were they censored at some point, and does that make any difference?

DR. MONROE:  Neither we nor the sponsor handled them any differently. This question of whether or not they should have been included was not addressed at all in the sponsor?s original submission. And because of the fact that our basic analysis had not yet shown -- or maybe never would show -- that there was a statistical difference in at least the "watchful waiting" local patients, it wasn?t pursued further. If he were to look at all of the watchful waiting patients, Stages 1, 2, 3 and 4, as a group, they do show a statistical effect as a consequence of Casodex treatment, but those ?- but one has to remember that there are many more events occurring in the T3 and T4 patients, and once those are removed, you have many less events to assess whether the change is statistically significant or not. And that is what we saw with our analysis based on the low risk component of the "watchful waiting" population.

DR. ALBERTSEN:  My question drives to the choice of endpoints for these trials. In some respect, it?s almost an artificial construct in contemporary urologic and oncologic practice, in the sense that a bone scan and a survival are really downstream endpoints, to the point of 10 years and almost 15 years later when you look at some of the data in the literature. And my concern is: by excluding PSA progression as a potential endpoint, don?t we potentially run the risk of missing a benefit, in the sense that if patients -- specifically in Trial 23 -- move on to LHRH agonist therapy when they have PSA progression, this happens before they even have a chance to achieve the endpoint the FDA is looking for. And, therefore, asking for more mature data, while very valid, probably puts that off until about 10 years from now. And I?d like your comment on that.

DR. SHAMES:  Well, as you know, that?s sort of a controversial issue, and we have yet to have the hard data that we felt we needed to use PSA as a surrogate endpoint essentially. So we have not yet accepted that. I mean, that?s still being debated internally and externally, as you know. It?s a somewhat difficult issue. I think the ?- okay. Go ahead.

DR. ALBERTSEN:  But the whole decision before us rests on that very decision, because the trial is clearly too short to demonstrate an effect from anti ?- at least in my opinion -- to direct an effect from a hormone therapy given over two years, given the lead time of five years. At a minimum lead time of five years that PSA testing has achieved, we are identifying patients in the U.S. considerably earlier than our European and Scandinavian colleagues. We?re operating on them, so an adjuvant trial based on those patients given for only two years, I think at virtually any endpoint you select, it would be difficult to demonstrate the difference.

So the trial, as constructed, is stacked heavily in favor of not demonstrating efficacy, unless you ran that trial for a period of ten years. And I?m not sure if the FDA or the sponsor were prepared to do that up front, and we?re stuck now deciding what do you do with the information at hand.

DR. SHAMES:  Well, I mean, we grapple with this issue daily because prostate cancer has this particular problem. We do trials regarding prevention of prostate cancer. And so, you know, perhaps if we had had ?- the problem that we have ?- the essential problem we have here is how to communicate who these people are that we can ?- that physicians will prescribe medication. And basically, although there are certain aspects, there are certain data here that shows a difference, there?s no question about it. We could?- we?re not able with the data before us to, we feel, adequately communicate to prescribers who exactly is supposed to get this medication, and make sure that we weren?t giving it to patients too early, and they were taking this for years and years, you know, without ?- we don?t know, perhaps possible adverse events -- or giving them later where they might overlap with the patients that might have some survival disadvantage.

So that, you know, as you?ve seen in the questions, and we?ll discuss the questions. That?s part of our basic difficulty here.

DR. ALBERTSEN:  If I can follow up, I think realistically were this drug, if it were to take ?- if FDA were to give its approval, I doubt it would be used as the indications are listed. Realistically, I think what would happen is most urologists and oncologists would give this drug to patients who demonstrate aggressive PSA doubling time. Specifically, if you refer to the Pound data from JAMA, anyone with a doubling time greater than ten months is at great risk of disease progression.

I suspect this drug is going to be used as a substitute for another drug that is not indicated for use in early prostate cancer, but is uniformly used in this country, specifically an LHRH agonist. The risk factors for them is the potential for osteoporosis, and all the other complications we all grapple with. So, therefore, the average clinician is going to try to weigh the complications of one drug that?s not indicated with another drug that?s not indicated. Realistically, that?s what?s going to happen. And I kind of scratch my head looking at the data presented, and feel I?m kind of an artificial construct, or an Alice in Wonderland scenario. In fact, what we?re discussing bears little resemblance to what will happen the minute this drug gets approved. Your comments on that.

DR. SHAMES:  All right. Well, that happens all the time. That?s part of the problem we have here, actually. Part of my concern is that the drug will not be used the way it should be used. And we will have ?- and I did say we might have hundreds of thousands of men using this drug, in either late stage, early stage, in-between, when we have some indication that there might be in some people a survival disadvantage, you know. And so that?s right, absolutely. That is absolutely one of our concerns. It?s always one of our concerns, but in this case, given the huge potential population that could be using this drug, we?re concerned about it not being used in the right population. And that?s why -- since we had a lot of difficulty determining who exactly this should be used in -- that?s why we?re not ?- one of the reasons we didn?t approve it. We couldn?t rely ?- conceptually, of course, we realize people in Europe and Sweden are not biologically different than the U.S. I mean, obviously that wasn?t the problem. The problem was, the data were difficult for us. You know, we didn?t understand the problem with the pathology, you know, the internal inconsistencies.

The thing with the Gleason score, you know, was difficult. That?s one of the reasons we asked the sponsor to perhaps, if it?s at all possible, to get the slides, and better define the population. That?s what we?re trying to ask there.

DR. GEORGE:  I had a question about the further follow-up, and I guess it?s related to the definition of the primary endpoint, as well. How would you ?- if you do further follow-up, which is, I think, certainly needed -- if you do that, what are you going to gain unless you expand the endpoint some to ?- if you?re going to include bone scan only, you?re going to have some issues there with people who have clearly progressed, but just didn?t have a bone scan yet. And you?re going to have also issues of requiring a bone scan at future times. Have you thought about that, I mean, beyond the two years? I mean, if you just said two years, then there?s not much point in following up beyond two years, if that?s what your major endpoint would be. DR. SHAMES:  Well, quite frankly, I was ?- I wonder if I?ll often see what happened to survival ultimately. I mean, that was one of my main reasons for asking for follow-up. I?m not sure there?s going to be a tremendous difference.

DR. GEORGE:  Well, one of the things that was presented in the FDA presentation was, further confirm that the durations of the time-to-progression information is maintained, and that?s more than just survival. I mean, I agree certainly with the survival, but the ?- you?re going to have to think hard about ?- that endpoint thing is not going to go away just with further follow-up.

DR. GEORGE:  That?s good. Would you do a bone scan in someone who clearly has progression, I mean, that you already know has progression? Is at advanced stage and, you know, you just didn?t do the bone scan. You did other things.

DR. MONROE:  You?d have to ask the sponsor exactly what they?ve told their investigators. It was our impression that these were supposed to be confirmed by bone scan to address the concerns about not being able to maintain blind, but they would need to address that.

DR. GEORGE:  I was just concerned, I guess if bone ?- it says that bone scans are supposed to be done every two years. If you didn?t do a bone scan because a patient had had clearly advanced disease because of other markers, and maybe died even of ?- I guess the death would be the endpoint, but you still would have had ?- presume if you had done a bone scan earlier you would have spotted it, but you didn?t. And I just wonder if there are going to be patients that you?re going to end up not counting as progression that were clearly progressions?

DR. CARROLL:  Thank you for the clarification. What the protocol required was that in patients who had not previously progressed, then a bone scan was scheduled at every two years. There was no intention to ?- for a patient who progressed, say, at three years, to then do a bone scan at four years if that would not be clinically indicated.

DR. GEORGE:  Right. But that then gets to be a problem in the FDA. Okay. If it were ?- but it wouldn?t have been picked up as via a bone scan. It would have been a rising PSA or something, I think, what they?re talking about.

DR. REDMAN:  Just to reiterate on the endpoint, and really just ask the FDA directly, if I was coming to them with a trial of an intervention in this setting, you?ve agreed that PSA is not valid. You?ve shown data that you don?t accept bone scan because of the inherent error in bone scan that all of us who practice are well aware of, that it?s very difficult to read a bone scan that?s not attached to a patient and interpret it. So I guess other than survival, overall survival, what endpoint is acceptable if you ask the sponsor to redo the trial, narrowly define the patient population. What is the end point going to be?

DR. SHAMES:  We showed some of the problems with bone scan, but I don?t think we said we wouldn?t accept that as a protocol defined bone scan in everybody. A situation where we ?- in this particular case, and perhaps some of the advisors can comment on this -- we were concerned about, particularly with Casodex because of this unblinding issue. So there?s some issues that are particular here that may not be relevant to, you know, other trials with other drugs we?re not as concerned about unblinding.

DR. KELSEN:  This is a follow-up to the comment that was made about patients who have a rapidly rising PSA, have the option right now. They get an LHRH antagonist, but Casodex is commercially available in the United States at a 50 milligram tablet, and I?m pointing at the open form. At least one patient who is taking 150 milligrams a day. This is a question to the urologists, either here or from the sponsor. How frequently is this therapy being used now in this area? Do we know anything about that at all?

DR. ALBERTSEN:  I?ll just make an anecdotal comment. I know it?s beginning to make its way into the medical community. How pervasive it is, I think, depends on how close you are to a center that?s promoting it, or to an individual person who might be promoting it. But I think the very real issue is, as the long term concerns over LHRH agonists, which have become the de facto treatment for men with rising PSAs, could Casodex be the substitute? And that?s the reality that?s playing out on the street. But I have no idea how prevalent it is yet.

MR. OHYE:  Thank you. I have a rather fundamental question about the change in endpoints. Some people say I?m older than dirt because I?ve been attending ?- before I retired I?ve been attending End-of-Phase-2 meetings, dozens upon dozens of them. And at the conclusion of the End-of-Phase-2 meeting, the sponsor generally goes away with a good idea of what would be needed to register the drug.

If I look at the history of this product, they had that End-of-Phase-2 meeting in 1995, and -- was this division in existence in 1995? It was not. It was the Metabolic and Endocrine Division then. I believe the practice then was to allow the sponsor to carry away key elements of agreements made, and then they would come back with a protocol based on the agreements made, sort of a quasi contract with the division. And it appears to me from the record, that they did that. And over the course of the next three years, they completed enrollment. And then a year later, the target was moved, and I think it would be very instructive for me as a representative of industry to find out how the heck that happened. Because, you know, these studies ?- you?re talking about a huge study here, costs millions of dollars. I?m sure you all are operating on good faith, but this moving target is quite troublesome to me personally, and I?m quite sure to others. I have other questions, if I may, but this is just a fundamental procedural question.

DR. SHAMES:  You?re absolutely right, in that when people come for an End-of-Phase-2 meeting we do everything we can to make sure that everything is appropriate so they can do their trials, and we?re all in agreement. You know, we ?- there was disagreement about a lot of these issues since we?ve been here maybe the last three or four years, so I cannot say if we would have ?- those of us here would have agreed to all of this, you know, the three different trials with the totally heterogeneous groups and that kind of thing. And certain other ?- we would have advised perhaps about Gleason scores, or central readings, things like that. So it was pretty much after the fact that we got here and looked at the trials and found the problems.

Now it brings us ?- and I agree that it ?- the overall procedure should be that we stick to the ?- what we say at the end of Phase 2, and then we let them go and then evaluate it.

Now sometimes people don?t ?- and I don?t know if that?s the case here. Sometimes that doesn?t happen anyway. You know, we can?t stop trials for sign problems. However, we are faced with this situation at it occurs now. I agree with your point and, you know, the only thing I can say is perhaps the people that reviewed it had a different view than we did, you know. It was before our division even was assembled.

But the issue before us is what we have to deal with right now, so that?s, you know ?-

MR. OHYE:  But nevertheless, we deal ?- as sponsors we deal with the agency as an entity, and as not just individuals? opinions, but we deal, you know, with the FDA as a whole.

DR. SHAMES:  Your point is very well taken, and I absolutely agree that the procedure should be that we come to agreements at the end of Phase 2, and hold to those agreements unless there?s some scientific reason that comes up later that causes us to have a problem.

MR. OHYE:  I think we all agree if there?s an overwhelming scientific reason to change prior commitments, but if we deal as sponsors deal with the agency as an entity, then we should be able to rely on prior commitments.

I have a question for Dr. Monroe. I?m sure you weren?t implying -- and I just want to make this clear -- that there is anything wrong with the sponsor proposing a change in indication, because that goes on, you know, frequently during the development of a product. I just want to make sure that I understood that right.

DR. MONROE:  Well, the need to change an indication does come up not infrequently, where an indication has to be modified to best reflect what the data are. What we were, I think, trying to convey to you is that, in an effort to identify who would best be served by the use of this drug as we reviewed it and brought up these issues, the indication was changed on multiple occasions, because the data clearly didn?t support certain things. Adjuvant issues in early disease, and we were surprised that the application even came in with such a broad indication, because there was just no data to support that. When those issues were brought forth, the sponsor acknowledged that and made these changes which perhaps could have come in initially.

And that I think it?s also a reflection that I don?t believe that we or the sponsor -- I shouldn?t really speak for the sponsor -- has adequately characterized who would benefit by adjuvant therapy. On that last slide I showed, we pretty much took aside all of the issues about what the endpoint should be. And those data I showed, they showed absolutely no effect of Casodex in U.S. Trial 23, based on the sponsor?s present definition of high risk. And it?s our concern that -- let? say the drug were to be labeled as that, and you say high risk. Well, are you going to define what high risk is, or are you going to leave it to each practitioner? I don?t know.

But if you use criteria that I think are generally accepted, as we use our high Gleason score, post-operative PSA, and a high pre-operative PSA, we didn?t see any benefit of the drug. And that?s our dilemma. We just don?t know who would be served well by taking this drug.

MR. OHYE:  I beg the Committee?s indulgence. I have two short questions, if I may continue. I believe I heard the conclusion that some of the data didn?t support a clinical benefit when you were making reference to some of your slides. This may be useful for deliberation when we discussed the questions. Are there data likely to predict clinical benefit?

DR. SHAMES:  Let me go ?- first, I would like to go back to the other issue, because Dr. Hirsch reminded me that in 1995, because he looked this up, we did not ?- we discussed the issue of the endpoints and did not totally agree with the endpoints actually, and the endpoints were ?- and as I said, you know, we can?t stop trials for design problems. We can only stop them for safety problems. Can you repeat the question, because I ?-

MR. OHYE:  I was making reference ?-

CHAIRPERSON PRZEPIORKA:  I?m sorry. Before you go on, can I just address that, please. Just switch hats as a former member of an IRB, and to encourage you to re-look at that issue specifically, because from the IRB point of view, to put a patient on a trial which will not give you an answer is a safety concern.

DR. BRAWLEY:  Can I ask, was there an ?- pardon me for interrupting. Was there an agreement between the FDA and AstraZeneca about what relevant endpoints would be for these studies that would lead to approval?

DR. SHAMES:  Since my ?- I have been there in various capacities -- we could not totally agree. We knew what the endpoints were, but we did not agree about these other endpoints, the objective progression endpoints.

MR. OHYE:  I?m afraid I?m taking too much time, but one last question, if I may. I was trying to ask -- I remember hearing the conclusion that the data didn?t support the finding of a clinical benefit. My question was, and I thought this might be useful when we go into our own deliberation, were there any data likely to predict a clinical benefit?

DR. SHAMES:  I think an objective ?- however we defined it, an objective progression would be information that we would consider clinically important, or delaying objective progression in a way that was not biased.

DR. MONROE:  Well, if I could add a little bit to that. The question of accepting the bone scan data is really not a question. I raised the concern that there is an inherent lack of accuracy in these methods, and we are concerned that a degree of inaccuracy needs to be considered when you?re looking at small absolute differences between treatment groups in trials that show very few events. We?re not saying that the bone scans are not acceptable as an assessment. We think that was a very valid endpoint, and would be done in a way that would be subject to minimum bias.

MR. OHYE:  I?d like to reserve my comments for the general discussion. Thank you.

DR. BLAYNEY:  I have two things. One gets to the issue of bias, as you put it, or this unblinding effect by Casodex, or unblinding effect. When you?re treating a patient, was your concern that gee, Mr. Smith, you have gynecomastia. You must be getting the active drug, and we?re going to ignore this urinary retention, or this new bone pain, or this rising PSA. Whereas, in somebody who is not having gynecomastia, they?re treating investigator would jump on a similar symptom in some differential manner.

DR. SHAMES:  I think it?s the general issues that we?re concerned about, regarding bias. You know, whatever ?- if you know that 80 percent of the patients ?- if the trial is essentially unblinded, then it?s not the kind of trial that gives us the same kind of data as a blinder trial, and we were trying to get the best data possible.

DR. BLAYNEY:  I mean, the endpoints you showed seemed to be very unsubjective. I mean, death, bone scan progression or some other PSA progression seemed to be uninfluenced by observer interpretation. So I think the suspicion or the concern that you had in setting an endpoint that gynecomastia or some clue that a patient was getting active treatment turns out to be ?-

DR. SHAMES:  Well, it?s true that death and bone scan are ?- we accepted death and bone scan. And PSA is a separate issue which, you know, needs other discussion. We don?t have the data right now to use it as a surrogate endpoint, and many of the other events were driven by the investigators.

DR. BLAYNEY:  But I?m saying that I don?t see that an investigator would have that ?- would be biased by the ?- he?d do the same ?- an investigator is likely to do the same thing, regardless of whether he thinks the patient is getting active treatment or not.

DR. BLAYNEY:  Fair enough. The last thing, I think, you?re asking us to make a judgment here basically on subset analysis since 1995 and this inauguration of the trial. You said well, wait a minute. Let?s now retrospectively or encouraged the sponsor to retrospectively define a group of men whom they thought might benefit based on a retrospective subset analysis. That?s something that has been looked at askance at this committee level, and I?d like to hear why you would want to do that.

DR. BLAYNEY:  No. Why you would encourage the sponsor to bring forward an application for an indication based on a ?-

DR. SHAMES:  No, it was not a sub ?- what they brought ?- their original indication was essentially everybody in this trial. Their original indication was everybody. Everybody who doesn?t have metastatic disease was the original indication. We didn?t think that was reasonable. However, the data supported it. And we?d said that and they went and tried to find the appropriate subgroups. And, you know, we are ?- you know that?s what happened.

DR. BLAYNEY:  And based on that, you know, we might as well just go home, because they haven?t demonstrated that. But I think there?s efficacy there that we need to ?- I would encourage you all to find a way to take care of ?-

DR. SHAMES:  Actually, the advice we?re looking for here, I mean, quite frankly, we are ?- you know, I said this to the sponsor. We could not figure out how to communicate who these people are to be treated with this medication. That was our basic problem, because there were various trial design problems, aside from the pathologic problems, et cetera, et cetera. And that?s reflected in the questions, and we certainly didn?t want to go approving this for everybody since this is not a totally benign drug, at least what we know about it.

DR. BLAYNEY:  I think Dr. Albertsen?s comments are right on point. It?s likely to be a drug that is used as a substitute for another non-approved drug. And this is going to be an expensive drug that patients will have to put that into their and their physician?s calculus.

DR. HANNO:  Just two very quick points. One is, since the bone scan data is so critical in this, I really think that some confirmation of which of the bone scans are truly positive would be helpful in at least calculating the absolute risk, because it may be much lower than it appears in these data. And there are plenty of studies that show how bone scan data is kind of unreliable. Even though it may be the same unreliability in both sides, it doesn?t mean that the risk is significant.

Second, aren?t we ultimately talking about if we?re going to use delay in objective progression as the endpoint and agree to that, don?t you really need an answer on whether immediate hormonal therapy versus delay, versus intermittent hormonal therapy? Which of those ?- is there a problem between them? I mean, that?s really the underlying issue here that we don?t have an answer to, and that we?re sort of skating around, I think. And in the absence of that, you?re really looking at survival and quality of life data.

DR. BRAWLEY:  Well, I was asking what was the FDA and AstraZeneca?s agreement back in 1995. And it just seems somebody had gotten up ?- I just saw him out of the corner of my eye, and they weren?t allowed to speak.

DR. SCOTT:  Mark Scott, AstraZeneca. If you ?- the interpretation of the minutes is there were a couple of different endpoints that were under debate. We designed the Casodex clinical trial program as one large program, where the analysis we proposed was based on objective progression as we defined it in each trial. The difference really was whether you would need to have two clinical trials to support that indication, if in fact it was time to progression.

The overall analysis could be done if survival was the endpoint of interest, but we focused on clinical progression as the primary endpoint. And we agree that we did have the conversations about the potential for unblinding, but I believe that we?ve demonstrated in the application that blinding was not ?- or the unblinding was not present due to the frequency of the assessments being similar between treatment groups, and across studies.

Oh, yes. I was just struck by lightning. One of the ?- and I?m very sympathetic to what Dr. Blayney said, because once we ask the sponsor to go back and find a subgroup in which there would efficacy, it was practically doomed to failure from the beginning, and the reason is that the U.S. trial was a null trial. There was no room for anybody to get true benefit. And so what happened was that the sponsor went back and did what they could, and had a very, very small subgroup, which showed a leaning towards results that were similar in Europe. The problem is that it was based on so few results, so few patients that it simply wasn?t reliable, and it was hard to take seriously.

The slide that Dr. Monroe showed, which happened when you took into account more patients to try to increase the size of this high risk population, the whole thing blows up in your face. You?re including more people in a null trial, and the hazard ratio goes right back to one, so it?s sort of like a rubber band. You pull it out, and it?s going to snap back. So I think that in this data is futile to try to go back and try to find a believable subgroup of patients in the United States who actually benefitted from the drug. And we?re in this problem because we have something in Europe that is trying to be extrapolated to the U.S.

DR. MARTINO:  This is terribly reminiscent of breast cancer. The real problem that I see here has to do with the fact that the gentlemen who were enrolled in America actually did too well to show you much of a difference if there could be one. It really comes down to that simple problem, which is in no way unheard of in adjuvant therapies. And I really think that?s at the gut of all of this, that the patients did so well that there?s no way, at least with this length of follow-up and this volume of relapse, which we?re really in the range of what, 5 percent or so for that one trial. In all fairness, how could you expect that there would be much of a difference unless you had a true miracle. And you can turn that data inside out and upside down, and it?s not going to change unless either you add more patients or more time passes.

CHAIRPERSON PRZEPIORKA:  Other comments. Okay. Then before we break for lunch, we have two more individuals who want to respond at the open public hearing after hearing all the data, so I?d like to call to the podium first, Mr. John Page from Us Too! International.

MR. PAGE:  My name is John Page. I?m President and CEO of Us Too! International. We are a 501(c)(3) prostate cancer education and support group. As a matter of disclosure we do, in fact, get funding from a number of sources. One of them is AstraZeneca, but do not interpret that to mean that someone has paid me to come here and make the comments that I am about to make, because any people that know me, know that that is probably the farthest thing that could be from the truth here. And unfortunately, I have not been in this kind of a situation before, so if some of my comments come out as too aggressive, please forgive me.

I am not, by the way, a prostate cancer patient, but I do work with them, have over the last three years on a daily basis, and so I speak really more as a patient advocate, as opposed to some of the gentlemen this morning we who have spoken really from the heart, and having to deal with this full on. I have had access to the data over the course of the last year, and so I?ve had a good opportunity to look at it.

I am also someone who has worked in healthcare for the last 30 years almost, and have a science and engineering background, so I do have an appreciation for statistics and dealing with research, but I do not consider myself, thankfully, simply a researcher or statistician, so I?m really addressing you really from the human perspective. And in dealing with that from the human perspective, I think I look at this as trying to define do patients deserve the information that is contained in these studies.

I hear the FDA talking about their determination of who should get this drug, and perhaps it?s just a generation that I have, but I find that not what I would hope the FDA would be looking at. I enjoy the FDA protecting me from bad things, and I enjoy the FDA giving me information that I can rely on, but I would hope that people would recognize that as a patient, it?s really up to the patient and his or her physician to determine what course of action is best for them. I don?t know that anyone in this room who?s not having to make that decision on a personal level should be excluding information given to the patient. And I think that?s really the crux of what I see us talking about here, is will you allow information about this trial to be provided to physicians and patients, and have them make a determination about whether or not this treatment option should even be considered.

I enjoyed the comments about the real world perspective, and whether or not you are setting up an Alice in Wonderland reality, but from all intents and purposes from a patient perspective, PSA rise is, in fact, a de facto standard that a patient uses to determine whether or not his disease is progressing. Whether we want to argue about that in theory and in research terms, the reality of the practice is a patient goes to his physician when he sees a PSA rise and says I?m afraid my disease is progressing, and I want to have another treatment. Good or bad, that?s the reality of the situation.

And I think the risk benefit equation can only be, and can best be determined by that patient and his family and his caregivers, because when it comes down to it, and you look at the NIH mandate that a patient be responsible for his or her care and the decisions thereof, the patient can?t make that determination unless they are given information. And there is no information that can be given unless this drug is approved. And so I find that by cutting off, prospectively cutting off even the discussion of potential benefit, and I think what I?m hearing after reviewing the data for a good number of months, and in listening to the FDA and AstraZeneca today, what I?m hearing is that there is definitely benefit coming ?- we?re trying to determine what group of patients that is, but there seems to me benefit, substantial benefit. Depends on how you define that, but at relatively small risk, and I think as long as the patient is informed up front about what those risks are, there seems to me very clear, and one of the gentleman from the U.K. who talks to his patient about expecting gynecomastia or breast pain. I think that that?s really what should happen. The patient should be fully informed.

Whether or not that happens is really a clinical decision, but when you?re looking at 40 to 60 percent reduction in progress of this disease, to me from a patient?s perspective, that would be very significant. And when the risks are identifiable, potentially manageable, and this is not an irreversible ?- I mean, the patient can go off this if he chooses, or if he decides that it?s not something he wants to do, I think that that?s really a patient decision. And again, as a patient advocate, my point is empower the patient with the information to make that decision.

I will use one statistic, I guess, because I think that it?s important. There were at one time more than 100 men in this room. If we use statistics, my guess is 20 men in this room will come down with prostate cancer. If I use further statistics about recurrence, 5 to 10 of those men will have a recurrence and potentially be seeking treatment options for which they have, at this point, limited or no treatment option availability.

Would you, if you were one of those five or ten men, want to prospectively eliminate a potential option with known risks. I guess that?s the question I leave you with, because as a patient advocate, a patient is really looking at options that are out there. When there are no options out there, and currently for a subset of patients, there are no treatment options out there, this represents a viable treatment option if they and their caregiver chooses to do it. And again, I think that?s the most empowering thing you can do today, is allow the patient to have the information upon which to make a decision that affects their life. Thank you very much.

MR. PAGE:  Right now, Casodex 150 is not an approved drug, and it is not available, widely available. As a result, if it?s not approved, it really doesn?t even come up in conversation except as an off-label indication. And I think that if you are assuming that it is okay for us to continue to treat patients routinely in off-label activities, then I say by all means. I mean, you cannot approve this, but I think if you?re going to be open and honest, and the reality of the situation is, patients are looking for something that may give, as the data indicated, perhaps a two year disease progression free life, that?s a quality of life indicator that a lot of men are going to accept.

DR. BRAWLEY:  Have we seen data that show that there?s a two year progression free interval with this drug?

MR. PAGE:  I looked at the data that is presented, and if you look at the Casodex versus the placebo, there is a ?-

MR. PAGE:  It may be one year, it may be two years, it may be three months. There is certainly what appears to be ?- and again, I?m going based on what the data is. The data does not appear to be contradicted by the FDA. There does appear even in the FDA analysis to be a benefit in time to progression. Their time to progression does not include PSA, but I can tell you, and Dr. Brawley, you probably know from your own practice, that a patient PSA rise is, in fact, a standard that a patient uses, whether the FDA or researchers choose to use that or not. It is, in fact, the de facto patient standard.

MR. FAY:  I?m sorry to say good afternoon. I have no conflict of interest, but when I offered to come and say a few words here, AstraZeneca did agree to reimburse me for my out-of-pocket expenses.

I?m a retired chemical engineer from Wilmington, Delaware. In February, I?ll be a six year survivor of T4 node positive prostate cancer and I watched my father die a horrible and degrading death from prostate cancer. I?m the community coordinator for the Wellness Community Delaware. I?m a Director and Secretary of the First State Prostate Cancer Support Group, and I also volunteer at the American Cancer Society and Christiana Cares, the principal health provider in northern Delaware. And as part of Christiana Care?s Cancer Outreach Program, I volunteer with a group of African American men to promote prostate cancer awareness and screening.

In these roles I know and speak to dozens and dozens of men who have or are at risk of having prostate cancer. I think I can speak realistically about how men at the grass roots level feel about prostate cancer. And I will tell you, and I can speak very confidently of this, there are three concerns that men have related to prostate cancer, other than survival. The first concern is number one on everybody?s list, loss of sexual activity. Number two on just about everybody?s list is incontinence, and number three, is hot flashes. Rarely hear any mention or concern of gynecomastia, and that?s a fact.

I?m simply going to build on what John said, and skip some of the things. I agree with everything he said virtually word for word. I?d like to talk about the African American community where I really am familiar. There are many men there who refuse to be screened or who avoid, or delay treatment after diagnosis of prostate cancer, because they fear loss of sexual ability. These men represent a self-selected and I think largely unidentified de facto group of watch and waiters. If these men had the 150 milligram dose of Casodex available with its very low level of adverse effects, adverse sexual side effects, they would elect earlier intervention, and thereby distinctly improving their likelihood of living longer and better.

These fears extend across the whole male spectrum. I?d like to give you three quick examples. Dr. Soloway talked about some patients from the doctor?s perspective. I?m going to talk to you about them from the patient?s perspective, and from my perspective. And I?m going to talk to you about three men that I know personally, that are friends, that I talk to. At the Wellness Community we deal with the emotional aspects of cancer. We talk, like you never heard men talk. There is no secret unbarred in our discussions, but anyhow, one is an African American man. He delayed additional treatment when he had a rising PSA following radiation out of fear of hot flashes, just that, hot flashes. He allowed his prostate cancer to progress until it was untreatable, and he died, I think prematurely.

George, a Caucasian has a Gleason Score of 6 and a PSA varying between 15 and 20. He?s gambling on watchful waiting because he does not want the side effects of any currently used treatment. Lou is a 76 year old Caucasian who has radiation. On Monday night at our support group meeting at the Wellness Community, he was in tears as he described the pressure he gets from his 78 year old wife, new bride, when he loses his ability to have an erection following injection of an LHRH agonist, so he stops taking the injection. His erectile dysfunction, his erectile function returns. I did stumble, and his PSA rapidly climbs to 80 before he panics and resumes the injections and the devastating cycle restarts.

These men need the option of taking the 150 milligram dose of Casodex, and they need it now, not years from now when every I is dotted, and every T is crossed. The risk of taking it, as I read the data, is very small, and the potential benefit is great. Give us, the patients, the opportunity to make the choice. I beg you. I beg you to approve today?s application. Thanks.

CHAIRPERSON PRZEPIORKA:  Thank you very much, Mr. Fay. Is there anyone else who has a comment to make? In that case, I want to actually thank both Mr. Page and Mr. Fay for their courage to do this after the presentations, and really address the data from a patient?s perspective, and come here to do that. Thank you.

We?ll break now and return at a quarter to 2. Actually, I?m sorry, 2:00, but Dr. Templeton-Somers wants to make an announcement first.

DR. TEMPLETON-SOMERS:  One of the big advantages of holding an open advisory committee meeting is that discussions like these take place in an open forum. It?s unusual for this particular committee to have a lunch break in the middle of an application, and I?d like to put forth a gentle reminder to everyone in the room that discussions of this application with the Committee should wait until this afternoon when our open meeting resumes and everyone can hear and participate. Thank you.

CHAIRPERSON PRZEPIORKA:  Dr. Shames, you gave a very nice introduction to the questions earlier today, so unless there?s anything burning that you wish to add, I will dive right into it. Fine, let us dive. So we?re going to discuss Casodex for adjuvant therapy to radical prostatectomy and radiotherapy of curative intent in patients with locally advanced non-metastatic prostate cancer who have a high risk for disease recurrence or immediate treatment of localized non-metastatic prostate cancer in patients for whom therapy of curative intent is not indicated.

We?ll go through the questions one by one, ask if anybody has any comments, and then call the question and take the vote, except for the essay questions, which I don?t think we have today.

DR. KRIST:  Well, I ?- looking overall, it?s kind of tricky for me hearing both perspectives. I have some reservations with the subgroup analyses. I have reservations with combining the data on the studies, and my inclination is to say it seems like there?s some form of a benefit, but I can?t put my finger on as to who it is. And I also think that whether that conveys into a survival benefit, there is some room for question with that, particularly with U.S. patients, because I think that the populations treated and diagnosed in the U.S. probably are very different than those treated in other countries.

DR. KELSEN:  I think this is a question again to the urology, the advisers at the table. My impression is that although the 23 Trial in the U.S. focused on a different population, people undergoing curative therapy, that many patients in the United States are seen with the same stage of disease as patients seen in Trials 24 and 25. And it?s not that we don?t have patients in the United States who present with these states of disease or rapidly develop them, but we just didn?t ?- they just weren?t studied in this particular cohort. Am I correct?

DR. HANNO:  I think you are correct. I think some of these patients get treated with the LHRH analogs, and I think that it would be more appropriate to compare it to that than to placebo, because a lot of these patients don?t get treated with placebo when they?re in that stage, or at least have that in the mix.

DR. SCHOENBERG:  Well, I think we should just have the caveat that although that?s true, we see a very significantly declining number of the more advanced stage disease, precisely because of the very proactive approach that the American medical community and patients have taken toward early diagnosis.

DR. ALBERTSEN:  My concern in the indication is the term adjuvant, and then moving on to define high risk. I don?t think the data I?ve seen convinced me that true adjuvant use of this drug is sufficient since Trial 23 basically showed no difference. Where I begin to hedge a little bit is the question of high risk patients. And indeed, those are more easily identified post surgery, when you begin to see a rising serum PSA. Unfortunately, they?re not the subjects of any of the trials, and they?re the ones of most interest to the urologists. As a consequence, I?m struggling on how best to interpret the trials of the more advanced disease, because I believe they do show efficacy. But then you have the question, are the European population sufficiently generalizable to the American population? I think they are, but I think the American population has just been identified a good five to seven years earlier, implying that you probably need to have people on this drug for at least five to seven years before you see the true survival differences, or at least the efficacy differences that you see in the European trials.

DR. GEORGE:  I think it?s ?- the problem here is a follow-up issue, particularly on 23. And I think we do need further follow-up to be sure of what we?re getting. And one of the things in saying that, we?re in a situation where we would like to have the answer faster. I mean, we would like to have fast answers, but we?re in a setting where it?s very difficult to do. It?s a long time before recurrence, and certainly before we have much information about survival. And that?s just a tough situation.

We?d like to have very good surrogate markers that would spot all this, but we don?t have them. And so my take on this is, these are very interesting results, but follow-up is a big issue.

DR. REDMAN:  I guess the point here is significant long-term benefit. I think, you know, following patients in 23 for the next 20 years isn?t going to really answer because of the patient population. But if you look at the two European trials, I don?t think 2.6 or three years is adequate follow-up to say that there?s a long-term benefit versus a potential short-term benefit. I mean, it?s a quandary in oncology, do we treat you now with toxicity, and there is toxicity to this drug, more so than no treatment. Or do we wait until you develop symptomatic progression and treat you at that time for ?- you know, overall survival is a long-term benefit.

DR. BRAWLEY:  I would agree that the issue of long-term benefit is the real problem here, that just two to three years doesn?t do it. If there was a trial that showed that there was a survival benefit, that would be, in my mind, a slam dunk and very easy to recommend approval. But we haven?t even really proven that we make the patients feel better. We?ve shown some indications that show that maybe some of the patients feel better. And we?ve also shown that, at best, 85 to 90 percent of the patients who would be treated wouldn?t even need therapy to begin with, and 30 percent of folks are going to drop off, so I guess I have some real problems and reservations here.

DR. MARTINO:  I think I have a different reaction to this than what I?m hearing around the table. These are patients where the word adjuvant, I think, does apply. They are patients without obvious distant metastases, and we?re looking at their first suggestion that they have metastatic distant disease. Okay? That is the adjuvant setting. And what you normally see when you do adjuvant trials is you don?t see survival advantage until years later. You tend to see that there?s a difference in terms of when patients have their first evidence of recurrence.

I think that?s what you?re seeing in Studies 24 and 25. You?re not seeing that in Study 23. Many of us could have almost predicted that you wouldn?t see it this quickly in that particular U.S. trial. So for me, there really is nothing here that disagrees with what I recognize is a basic principle, that this is a hormonal disease where hormonal therapy to a small degree, which is the problem with all of our adjuvant trials. It is that 2, 3, 4 percent if you?re lucky, that you see a difference between a treated and untreated group. So I think for me, there is value to this therapy in the patients that were treated with it, which is what the question states. The issue of whether that can be translated to the American population is a different issue for me.

DR. BRAWLEY:  Yeah. That gets back to an earlier point. The reason why we don?t see median time to progression in the two arms is neither arm has actually lived to median time to progression yet. That really means that we?re not treating very much at this juncture. I mean, if there is a benefit, and the advocates really need to understand this. You know, if you?re talking to that black guy up in Connecticut, you need to tell him there?s a one out of 25 chance that this pill may help you, and a one in three chance that you?re going to drop off the pill because of side effects. You really need to tell him that.

CHAIRPERSON PRZEPIORKA:  Other discussion? Then I?ll call the question once again. Across Trials 24 and 25, only 15.6 percent of patients or 9.3 depending on endpoints had objective progression of cancer or died. At the time of cut-off median follow-up was 2.6 and 3.0 years. In the absence of meaningful survival data or quality of life benefits, are these studies sufficiently mature to conclude with a reasonable level of confidence that patients treated with Casodex in these trials will derive clinically significant long-term benefits? Dr. Redman.

CHAIRPERSON PRZEPIORKA:  That?s three yes and 13 no. Question two, do the data that is clinical stage PSA level and lack of valid Gleason Score from Trials 24 and 25 allow for the adequate definition of a patient population that can extrapolated from the non-U.S. studies to a defined group of U.S. patients who will derive significant benefit from Casodex therapy?

DR. CHESON:  Point of order, there?s a second part to the first question that wasn?t addressed. If not.

DR. GEORGE:  I?m the one who noticed that. He?s the aggressive one who spoke up, but I had ?- I thought we were going to address that because that was an important part if we did say no.

DR. GEORGE:  And to me, again getting back to the follow-up issue, and it?s still the ?- what are these types of information we can maybe talk about later I guess, number two, and so forth, but the key with respect to follow-up is to follow-up until the number of events is higher, the percentage of events. I won?t go through all that again, but it?s still, even in the ones like 025, it?s still a low percentage of overall events. And events I?m talking about here are either death, bone scan progression, or even this other progression. You just add them all up, especially in the death category, of course, but in the others as well. The overall percentage is still low enough to be disturbing - not disturbing, but to be unreliable with respect to the long-term issues. Even though I agree that unless you change definitions of endpoints, the early results aren?t going to change. But we do need to know the long-term, even granted that everybody either dies or progresses eventually. You still want to see what happens later, and have more reliable answers. So the kind of information I would say we need first and foremost, is a higher percentage of events in all these categories.

DR. CHESON:  Yeah. And just all kidding aside here, I want to agree my friend Otis over there, in that you have to look at long-term in the context of the natural history of the disease. So whereas, you?re saying it?s going to be two years, three years, four years, five years, you have to recognize what the median survival is, and look at these events that Steve was talking about in relationship to that sort of a time point.

DR. SCHOENBERG:  This is ?- you?re calling for comments. Correct? I?m concerned because I think the groups of patients are very significantly not comparable. And that doesn?t mean, as I think people have noted previously, that there aren't interesting and compelling data to be derived from the European trials, but the U.S. population is very different, the one that was studied. And because of the confounding problem of identifying exactly which pathologic entities were being studied in 24 and 25, I?m particularly concerned that we?re going to have difficulty identifying who?s going to benefit. And I am very concerned as this particularly touches upon the issue of watchful waiting, so I am very concerned about those two particular problems.

CHAIRPERSON PRZEPIORKA:  Can I ask if you can give your opinion on the subgroup separately? I hear that for the localized disease group with watchful waiting you don?t believe that the subgroups, that the two groups are comparable, but what about the high risk group? Would they be more comparable, or not comparable at all?

DR. SCHOENBERG:  Well, as I think we?ve discussed previously briefly, the problem, and this may be reflective of an issue, a Transatlantic practice difference, is that I think the groups that were studied in 24 and 25 are not ?- if not vanishing, substantially diminished in U.S. practice. And it?s not that those people don?t exist, but they are substantially less common. So I think yes, there are compelling data in the higher risk groups, but again, we do have problems with definition. And I believe one of the issues we've discussed previously was the lumping of 7 with 8, 9, and 10. That is, to my mind, very problematic, so yes, I?m intrigued by that population, but I think more information will be necessary to extrapolate this to the U.S. population.

DR. BLAYNEY:  I would take a contrary-wise point of view. I think for men who don?t want radiation, but who might have localized disease discovered at surgery, the data from 24 and 25 might be useful in helping them and their physicians make some decisions. For men who may have a very high Gleason Score of 8 or 9, who would otherwise fit ?- be much like the 24 and 25, I think in my practice in my community, that would be ?- that data would help in decision making, so I would say ?- my answer to this question would be yes.

DR. SCHOENBERG:  Could you be very specific about how it will aid in decision making? I?m just curious.

DR. BLAYNEY:  Some men don?t want radiation after a positive surgical margin is discovered. Traditionally, those people are recommended, and there may be some benefit to survival to salvage radiation. If they don?t want radiation, this is an ?- I think there?s data that this treatment might be an option for them.

DR. ALBERTSEN:  In my mind, the data from Trials 24 and 25 do provide evidence of efficacy for men with more advanced disease. The question then becomes how do you interpret advanced disease in the context of a U.S. population. Even though this wasn?t test tried, I think the physician community is treating patients with rapidly rising or high doubling time PSAs as men at high risk of failing. That probably can?t be incorporated in the labeling of this drug, but in terms of how do you do the walk across the Atlantic. In my mind, that?s how it would be done. So, therefore, I do believe the data support efficacy in men with more advanced, or as they say, high risk disease, and I would define these, as I think people do in clinical practice, as people with rapid doubling times.

DR. BRAWLEY:  Earlier there was discussion of subset analysis and how subset analysis should be avoided. I just want to weigh in and reiterate that subset analysis really should be avoided. It is my feeling that the groups defined in 24 and 25, especially in Trial 25, can be extrapolatable, and you can find people like that in the United States. Although they are few and far between, I think you can find folks, so I ?- it might seem contrary to my first vote in the previous statements, but I do believe that you can extrapolate from the foreign trials to find similar patients in the United States.

CHAIRPERSON PRZEPIORKA:  And I think like Dr. Brawley, I also believe that we?ve seen positive results here, and that there are probably patients in the U.S. who would fit the same criteria as in the U.S. trial. Where I would disagree also is that the statistical analysis that was presented was not the most eloquent I?ve ever seen for looking for subsets of populations and risk factors for regression, and whether or not the treatment arm gives benefit, so I would disagree and say that the data as presented are probably not adequate enough to define the population very well. Other comments? Dr. George.

DR. GEORGE:  There was the ?- Dr. Hoberman, if he?s still here. He is here, that he had done a number of analyses with respect to trying to adjust for factors to see if the great discrepancies between all these studies could be reconciled in certain ways; that is, to use the 24 and 25 data and see if you could predict some of the things that were observed on the 23 study. And if I?m interpreting this right, I think the answer was you couldn?t really do it. Maybe you?d like to elaborate on that a little bit.

DR. HOBERMAN:  Yeah, I did it, and the numbers didn?t come out right. I was not able to close the gap that should have been closed in the U.S. by doing that progression projection from Europe. I?m not sure exactly how this fits into the question that was being asked, but ?-

DR. GEORGE:  I?m just using that as an example, and why I?m kind of reluctant in this area, because what ?- you know, if you assume ?- in general, it?s not a problem that you have studies with markedly different distributions of variables, as long as you have enough patients treated in groups that you can use statistical procedures to kind of adjust for that. In this case, it didn?t seem to work out right; that is, something wasn?t right, either the variables weren?t measured properly, the models that were being applied didn?t fit, and the European studies didn?t seem to apply to the U.S. for some reason. It would take a lot more looking at this issue to convince me that the results from 24 and 25 could be used.

DR. REDMAN:  I got a little confused by Dr. Albertsen?s comments. You state that you wanted to somehow define a high risk population in a population that wasn?t in these studies. Somebody with a doubling PSA, so nine months after they?ve had the surgery, the PSA ?- or radiation therapy, the PSA is doubling. These studies, if I?m correct, looked at you had your radiation therapy. We?re not going to wait for PSA doubling. You?re either going to go on Casodex, or you?re going to go on observation, so I don?t know how that high risk population fits into what these studies showed.

DR. ALBERTSEN:  That?s why I made the comment as the way I did. When I looked at the data, the only patients that I can determine have any benefit are the ones who are high risk, i.e., the European patients, who in general have a disease that?s more advanced, precisely because they don?t do or have not been doing in 1995 aggressive PSA testing as we do in this country. So in 1995, you had a lead time introduced for most American patients which you didn?t have in Europe. And that?s why, in my mind, these two populations aren?t comparable.

When you use this drug very early on in the course of disease, it plays out over 10 or 15 years. You?re not expecting, and I would be astounded to see any benefit in the first five years. Hence, if this company were to run the trial for 15 years, I think we might see a difference, so that remains to be tested. But if I had to guess one population that might be most likely to achieve, and again, this is a bit of a leap of faith, it?s the very patients who we see failing radical prostatectomy or radiation therapy. And we know from the Pound data published in JAMA about two years ago, that men with PSA doubling times less than 10 months will generally progress to metastatic disease within 8 years, and will die from their disease within 13 years. But that?s the time frames we play out here. And again, it?s a bit of a leap of faith, but when asked the way the question was structured, if I could identify a population, considering the tools I have in 2003, that?s probably the best way I could estimate such a population.

I haven?t got a clue as to what those numbers are. I was wondering if our urologist did, and I gather it?s hard to extrapolate those numbers. But I think there is a population in this country that doesn?t get screened, that does present late, and that might well mimic the European population. It would be exactly what you?d be looking for. And they don?t go on clinical trials, so there?s no data for it. I actually don?t think 23 applies to this question at all. Twenty-three is a different issue, totally a different study. Just because the trial was done in Europe does not mean it doesn?t apply to American patients.

DR. BRAWLEY:  Yeah. David, you can perhaps correct me if I?m wrong. Grace Lu-yao showed that 30 to 40 percent of men who get radical prostatectomy in this country relapse by PSA within five years, many of them within two years. And it is now very common that those individuals get off-label LHRH agonist. I think a growing population will probably get Casodex off-label, despite what one of the advocates said a little earlier. But AstraZeneca is actually to be congratulated for doing studies looking at these populations right after initial therapy. I wish we had similar data with the other drugs that are being used in them, but one possible place for Casodex and one study that still does need to be done is in that 30 to 40 percent of Americans who after radical prostatectomy, or after ?- I don?t know the percentage after radiation therapy, who have a rising PSA. And unfortunately, that trial if it is powered for survival, is going to be a 15 year trial. But I must point out that AstraZeneca has presented data, very elegant data looking at Tamoxifen at 10 and 15 years of data, so we just need to do the equivalent in prostate cancer. And in Tamoxifen, they show the survival benefit which may or may not be available if we use Casodex as an adjuvant.

DR. ALBERTSEN:  Yeah. Otis, I agree with you. If I had to just pick the trial that I?d want to see today to make the decision, it?s basically men who have rising PSAs following radiation or surgery, who now face a choice, or face the problem of what do you do. What?s happening is most of them are getting LHRH agonist with the associated risks of libido, hot flashes and osteoporosis. They?d like an alternative, something that avoids those risks. However, ultimately I?m not sure which the better therapy is. What troubles me about the data is that Casodex 150 was not as good as castration in the M1 trials. Therefore, with an alternative that clearly works better in advanced stages, but it comes at a price of quality of life much earlier on. And since we?re dealing with a chronic disease, the question is what?s the appropriate choice.

The trial that needs to be done is in men with progression following surgery or radiation. Is it best to give LHRH agonist, Casodex or placebo? I suspect no one is going to fund such a trial. It will take too long to sort out, so instead we?re grappling with the data that?s being handed to us.

DR. SCHOENBERG:  Yeah. I think this is sort of a more general comment echoing some of the things that Peter has just said. AstraZeneca, as is everyone who works on prostate cancer and works on it seriously, is to be congratulated for doing difficult trials, but the disease is what it is. And if it takes a long time to get an answer, that?s the ball game. And at least from a practical clinical perspective, it?s very hard to accept intermediate endpoints that have not been validated, that are of questionable value within the context of a given study. And I think one thing we need to keep in the back of our minds, probably everyone here is thinking about it, is that while it is great to offer patients choices, and all of us want to do that, I think that at some level it is unethical to represent that a choice, any choice is a good choice. And part of our job, obviously, is to make sure that drugs that we recommend or don?t recommend get that recommendation for a reason. And I think the strength of the data here make it problematic for, at least for an American urologist like myself, to weigh in strongly and say sure, there?s clear evidence that this will be beneficial.

I think what we can say is that it may be in a very small and defined segment of a population, which really requires a much longer study, unfortunately. That?s the ball game we?re playing here.

MR. OHYE:  I?d like to reserve a comment after the vote between this question and the next, if I may.

CHAIRPERSON PRZEPIORKA:  I?m sorry to tell you, eight yes, eight no. Mr. Ohye, which may actually give us the deciding vote.

MR. OHYE:  As you know, I?m non-voting. However, I think the FDA in their wisdom under Subpart H in the regulations, have provided a way out for us here, because if we were to change this question to read and I?ll only refer to the last line of the question. Change it as follows, ?Define groups of U.S. patients who are likely to derive a clinical benefit from Casodex therapy?, we?re talking about using the provisions of Subpart H or accelerated approval as we did yesterday, where you have data that is likely to show, likely to provide a clinical benefit. And then we have the burden of the sponsor to show at a subsequent time, later data, could be more mature data from this study, because this is certainly not a failed study. We?re not trying to save a failed study. This was a good study. It was carried out pursuant to the agreements reached by FDA at the End of Phase 2 Meeting.

As you heard from Dr. Albertsen, we know that this drug is going to be used, available. We know the drug is going to be imported from Canada. I can go on the web site as soon as it?s available in he Canadian market. I can go on my computer and get it, and FDA has made an announcement that they?re not going enforce importation of use of drugs for individual patients, so this cries out for, I think, the sponsor and the agency to provide adequate directions for use, and to move forward with this drug under the provisions of Subpart H. Thank you.

DR. HANNO:  I don?t think it?s clear basically if any would. I don?t think we know, and we would just be guessing. And I think the key missing element in this whole discussion is, is there data that immediate hormone therapy improves survival? I mean, if it does, then a lot of this ?- then Casodex may turn out to be a great drug. In the absence of that data from any trials on any type of hormonal therapy, we are really in the dark about this, and we?re just sort of guessing, so I would just say it?s not clear if any would.

DR. HANNO:  Part C of Question Three. I?d like to see either survival data, or quality of life data suggesting that early treatment in preventing the onset of metastatic lesions improves the quality of life, regardless of effects on survival, compared to treating when PSA rises or bone metastases appear. And my concern is that this indication in the U.S. for Casodex would imply that Casodex improves survival of all patients after definitive therapy. And it would become ?- there would be widespread use of adjuvant Casodex in virtually everyone who gets a radical prostatectomy or radiation therapy. And I?m not sure that would be warranted, but I think that might well happen.

DR. MARTINO:   I don?t think it?s fair to say that this trial, number 23, does not show a benefit. See, that implies that we know the future, and we don?t. The reality is that this is a trial where the relapse rate is quite small at this point in time in a patient population where that could have easily been anticipated to be the case, so the question implies that we understand that even in this population, there can never be a benefit. I think that?s an assumption on our part.

CHAIRPERSON PRZEPIORKA:  And actually, if I recall the data, and I wrote it down from 23 specifically, the patients who had radiotherapy had a significant improvement or a significant lack of progression being reduced from 40 percent to 28 percent. It was the prostatectomy patients who had no benefit and no progression, as you had pointed out earlier, so I think you?re correct. And I would echo what you said about we?re not there long enough, although the radiotherapy patients certainly look like they derive benefit. Other comments regarding 3(a)? Dr. Albertsen.

DR. ALBERTSEN:  I agree with Dr. Hanno. Ultimately, the use of hormonal therapy has not yet been demonstrated to increase longevity. And the sad part is, I see any window of opportunity to prove this in a clinical trial as probably beginning to draw to a close, so I?ll address number C. Given the group of patients we know that are high risk for disease progression and death from prostate cancer, are those men with rapid PSA doubling times following a definitive therapy. That?s the population that?s key, and whether ?- and how much Casodex or any hormonal therapy alters that natural history is debatable. But that is a group that requires further study, and that?s the group that if we are extrapolating from Trial 24, you?d be extrapolating to.

CHAIRPERSON PRZEPIORKA:  Other comments? So, Dr. Shames, I think the answer to ?- this was an essay question for A is, we can?t define a population based on what we have now. We would like longer follow-up data.

DR. ALBERTSEN:  I?d answer no to that.

DR. ALBERTSEN:  That?s defining who is getting watchful waiting in this country, because usually the persons getting watchful waiting in this country are people who are felt to be at low risk for disease progression. It?s precisely those patients that you?re going to have the least amount of efficacy from Casodex. The ones who have a high probability of progression are the ones who might benefit, and that?s why I voted no.

DR. ALBERTSEN:  I think a potential group are those patients who elect not to have surgery or radiation, yet who have clear progression of PSA. I?m dealing with a gentleman in my community I just saw yesterday who?s been watching himself for five years. He?s 65 years old, and he just does not want to have surgery or radiation. His PSA is now up to 16. I know eventually he?s going to die from this disease. He?d be the perfect candidate to put on something like this, so I think there?s a group out there, but I think it?s small. But to a blanket statement of all patients who choose watchful waiting, that I think is inappropriate because most of the patients who choose watchful waiting are usually men in their late 70s and early 80s whose natural history is such that they are going to most likely die of a competing hazard, rather than their prostate cancer. So, therefore, treating them with Casodex just gives them the morbidity from the treatment. It is unlikely to achieve any benefit for them.

DR. MARTINO:  I guess I?m unsure now as to ?- particularly in Trial 25, the Scandinavian trial, where I thought I had seen that about 80 percent of the patients in that trial were of this population. I guess I?m not sure why it is that those patients are so different from the American population. I remember the gentleman who addressed the nature of the patients, making the statement that the patients that went into Trial 25 were men who had a time from diagnosis to entry in trial of three months, so how is that so different from what might happen in this country? I?m very confused on this patient selection issue.

DR. ALBERTSEN:  Yeah. I think the issue has to do with PSA screening in this country. The U.S. populations were probably comparable in 1990, but are no longer in 2003. Maybe in 2005, as the Europeans begin to more aggressively screen PSA, you?ll see a congruence. But right when this study was done in 1995, you probably had the most disparity from what was being done in this country compared to what was being done in Europe, so therefore, the pool of patients who are slated to be enrolled in Trials 24 and 25, by and large had more advanced disease than their American counterparts. So again, it goes down to the lead time associated with this disease. And if we?re talking about a lead time of at least five years, and possibly longer as a result of PSA screening, when you make a new diagnosis of someone who?s in their mid to late 70s, who has at best a 10 year survival, you could see where the benefits begin to get very marginal.

DR. MARTINO:  I understand the basic concept, and I agree with the basic concept. I think what I?d like is, I?d like to hear from the PI from the Scandinavian trial that in fact, when they entered patients on this, that they really were not selected from the point of view of being diagnosed, screened based on PSA that ?- the impression is that these were gentlemen in Europe who actually were symptomatic, and therefore, diagnosed. I?d like to know if that?s correct or not.

DR. SCHOENBERG:  Yeah. I mean, just as a follow-up to Dr. Albertsen?s comments, and to clarify this. There are guidelines in the U.S. for recommending watchful waiting to patients with prostate cancer. Not everybody follows them. Clinicians are accorded a relatively broad degree of latitude in advising patients about the problem, but this population is, I believe, by definition substantially different than the U.S. watchful waiting population. Using the criteria of PSA density, the amount of biopsy material that is actually positive for carcinoma, and the Gleason Score, which by definition should be less than 6, or equivalent to 6 but not greater, so I suspect they?re really very different populations.

DR. KRIST:  I agree with all the statements made. I mean, the other thing that concerns me too with the watchful waiting group is it seems like there?s a narrow window of potential benefit, where there?s a group of patients who really aren?t going to derive any benefit from treatment with Casodex, just due to the nature of their disease, and would be better off with just watchful waiting, and no therapy. And then even on the other end of the spectrum with us seeing in Trial 306 and 307, there?s the other end of the spectrum where there?s a potential group with more advanced disease who might opt for watchful waiting, who would be better off with LRHR analog, or medical castration instead.

DR. HOBERMAN:  I simply have a point of information. You made a statement about in 23, radiotherapy patients having an efficacy advantage. And I?m curious where that is coming from.

DR. HOBERMAN:  That doesn?t square.

DR. MONROE:  I believe you were shown percentages, but I think the number of events were only like four total, and so you have maybe some very ?- the data by looking at percentages I don?t think are appropriately represented. I just want to bring that up, that if one looks at actual number of events, I think by looking at percentages, you get a wrong impression as to what really occurred.

CHAIRPERSON PRZEPIORKA:  Thank you for clarifying that. Other comments? Dr. Shames, do you want us to take a vote on 4(a), or what I?m hearing is that essentially we don?t believe that the patient population in 23 and 24 is something that we frequently see in the U.S., but there is a population in the U.S. who really does need to be studied, whom we would predict would have benefit from this, to answer Part C, which means another study.