DR. CHRISTIAN: The next speaker is Dr. Michael Bookman from Fox-Chase.

DR. BOOKMAN: I realize I am the next to the last speaker for the afternoon. It is always a special treat and I am going to try to keep folks awake, if I can.

I just wanted to make one comment regarding the previous session on IP therapy. I will save that for tomorrow. That is okay.

The Gynecologic Oncology Group has been a great resource for phase I and phase II trials, and ultimately phase III trials.

Our focus has really been to look for new agents, evaluate them, and get them as rapidly as possible into the front line phase III setting, if they appear interesting and active.

We don't typically do first in human studies of brand new investigational drugs, although occasionally that has occurred. So, most of the very early phase studies are done external to the group.

We focus intensively on phase II evaluation of investigational agents and new combinations with other drugs.

Then we are in a kind of unique position to do front line phase I studies. So, we can take new combinations, like carboplatinum Taxol, when this was first being developed, and test it directly in patients with newly diagnosed disease without prior therapy, and do that in such a way that we could get data on efficacy, on feasibility, toxicity, and make recommendations about how to go forward into phase II trials.

TOP

There are a number of queues that we manage to treat our patients, not just with ovarian cancer, but also with other gynecologic tumors.

Because of the size and complexity of the group, and all the issues that have been discussed by a number of speakers over the course of the day, this is a laborious process.

It needs to be precise. We need to know exactly which studies are open when, which ones are under development, and how to close one down and open the next one, get it out to the institutions and so on and so forth.

A lot of effort is expended by the group to support this phase II effort, and the amount of effort required to support it has been increasing each year, as our complexity of studies have increased, but also as the regulatory administrative environment has increased demands upon us.

TOP

The phase II trials, as shown here -- sorry, the previous slide had the phase Is. these are the phase IIs. You can see that, in ovarian cancer with platinum resistance, platinum-sensitive disease, biologic studies, which are not distinguished between resistance or sensitive, an intraperitoneal queue that is about to be reactivated to explore some new intraperitoneal combinations, and also other tumor sites where we do a large number of phase II trials.

In addition, over the last few years, we have expanded substantially the number of studies that we can run at the same time.

So, within platinum-resistant ovary, platinum-sensitive ovary in biologics, as well as within cervix, we run two studies in each category, essentially dividing the group in two, and have an A team and a B team.

We are considering strategies to further subdivide the group to be more efficient in terms of getting studies done, amongst the institutions that are strong participants on those trials, to encourage more ownership and try to overcome of the barriers that exist.

TOP

I wasn't going to obviously review in detail the history of the group or the agents that we have studied. You will note that many of these studies were sponsored by the National Cancer Institute through the Investigational Drug Branch of CTEP, and we do try to maintain a close collaborative relationship with them.

We also do studies in collaboration with industry or with commercially available agents, and I will talk a little bit about that.

TOP

So, in platinum-resistant disease, I have just shown you a quick summary of the last few years of studies that we have been focusing on.

TOP

Platinum-sensitive disease, the studies are shown here, and I would be happy to provide a copy of these for anyone who wants to go into this in more detail.

TOP

Recently, in the platinum-sensitive queue, we have also been studying combinations with biologics, due to the interest in developing these regimens as well. We tried doing a randomized phase II trial, which is currently running with the PCIN.

TOP

Most of our attention has really been on the newer molecular targeted or biologic agents. This is the 170 queue. Again, most of these studies have been CTEP-sponsored, as you go down the list.

Right now, we are doing lapatinib, which is a dual kinase inhibitor that Janet mentioned, GW572016, as well as the histone diacetylase agent, and we just finished the first stage of accrual for the Bayer raf kinase inhibitor, which is under analysis to see if we need a second stage.

That study actually completed accrual within 35 weeks, and that shows actually the power of the group. If there is a study up and running that we are interested in, we do have the resources to accrue fairly rapidly and get the study done.

It is the timing to get to that point, to analyze the data, and then decide if we need a second stage, that sometimes gets us bogged down.

TOP

This is an example of a study, many of you have seen this information before, but again, just to illustrate the power of phase II studies in a Cooperative Group setting.

We wanted to study a receptor, trastuzimab, in patients with ovarian cancer. We needed to screen for over-expression of HER-2-NU. We had over 800 tissue specimens submitted for screening during the course of this trial.

Ultimately, approximately 95 were identified as potentially eligible. From that group, about 45 were registered on the trial, 41 were evaluable, and three patients responded.

So, there was a tremendous effort by the group to actually systematically study this particular agent in ovarian cancer.

We concluded that it wasn't active and, now that we understand more about the targeting of this particular receptor, we think this makes sense, based on the biology of HER-2 being different in ovarian cancer compared to breast cancer, which is something we didn't totally understand when the trial was being developed.

You could also say, well, there were three patients here who did benefit, and what was different about those three patients?

Actually, we don't know, but there are other studies, and you heard from Dr. Schilder and Godwin earlier about the studies with tyrosine kinase inhibitors where, in fact, with group resources, we were able to go back, look at tissue, and learn that receptor mutations were a predictor of response and a predictor of positive outcome in a very small number of patients.

So, conducting these studies in the setting of a group with tissue, with support, with archival information, really allows us to, over time, look at our data in different ways, and try to learn from it as we move forward.

TOP

Now, these challenges have been mentioned by the introductory comments of our panel here today and by other speakers as well.

We really value our collaborative opportunities with CTEP, but we also value our ability to work with the pharmaceutical industry, because there are so many interesting new agents that are coming forward at this time.

We also, on occasion, have run studies with commercial agents. We are trying to wrestle with prioritization in terms of how much should we focus on new agents versus combinations, and how many single arm trials should we be doing just to establish a bar for a particular drug or regimen, versus randomized studies that ultimately might provide more information than just did a regimen cross a threshold or establish a level of activity.

How do we do this prioritization?

TOP

Well, the most important thing is not the money, but clearly, in the current environment that we are in -- and you can tell this slide is dated by the price of regular gasoline -- but this is a challenging environment to be working in.

We are juggling a lot of complexity and resources with very little financial support to help carry that effort forward and the demands are becoming more complex all the time.

However, the way we really do it is by the science or, at least at the time of our semi-annual group meetings, the focus is on the science.

There are financial issues that obviously have to be integrated with the process, but we do focus on the science.

Here is where I show you the molecular pathway of the cells and signal transduction.

TOP

This is an aerial view of Wildwood, New Jersey .

There are a lot of different pathways that are obvious from this slide. Those of you who have observed this at different levels are familiar with what I am going to talk about here today.

There is a pathway of individuals that march from the boardwalk to the beach. Some do this on their own free energy, some have passive or active mediated transport of various means, because it is a long walk, about a half a mile to the water there.

Other individuals seem to have some mutation that gets them into a loop, where they go round and round and round and just never stop until their time runs out.

Then there are people who have really unusual mutations, that go on these rides here, that really leave them getting sick and frustrated, and it is questionable why they do that, but I don't know.

There is a whole level of activity at this magnification you are not really seeing, and it is the boardwalk itself.

TOP

This is kind of a low-res view, but a close up of the boardwalk shot through a cell phone camera, and then colorized a little bit to look like a poster.

I like it because this is what we see when we look at a cell, and when we look at the market place and when we look at drug development, because it is noisy, there are a lot of bright lights and colors.

You can't really magnify it to look at it any more closely. There is a lot of noise and confusion. There are people here yelling, just put your money on the line. All you have to do is get one ball in and you can win, take anything you want.

Everyone is trying to get you to participate and be a part of the action, but it is hard to make decisions. If you have children and you give them $10, it is gone in 30 seconds. That is not a problem. If you are an adult and you are more thoughtful you say, can I really get that basketball in the hoop, is that really the thing I should be choosing today, and what are the odds of having an agent be successful as we move through this developmental pathway. Those are a lot of really good questions for which we haven't defined perfect answers.

TOP

Now, the challenges that we face have been mentioned. Timing is everything. You have heard that it is too long to generate protocols, to activate studies and complete studies.

I don't deny that. I wish we were more efficient in what we do, and my goal, in being part of the group activities and being part of this meeting today, is to try to learn and think and work with you to develop ways to be more efficient in this process.

Financial support is a huge issue. The groups are under-funded. We have inadequate per capita reimbursement to our institutions to support data management and care of patients, and inadequate support, in my assessment, for the infrastructure of the group that is needed to run these trials.

I am not saying this to be critical of the funding agencies, but just really to state a point that has an impact on how we do our studies.

The administrative regulatory overhead is incredible. The protocol development and review process is very complex and time consuming, involving multiple conference calls, joint meetings, particularly when you are working with investigational drugs that actually look good, because then you have to pull the FDA into the process, because what if it needs to be a registration trial, what if the data might be used to establish an indication for a drug, which is stuff we want to do. We want to get these things out there and get them approved, but it makes for a very prolonged protocol development process. GOG has suffered in that regard recently, trying to get their next front line phase III trial up and running.

The local institutions also have it challenging and NCI, in what have been described to us as simplifications, have created useful tools, that nonetheless are more complicated than what we used to do.

I mean standard old response criteria were much easier to fill out on forms than RECIST, but RECIST, we are told, is a better way to do things.

The new CTC, with 254 categories, is better and more precise than what we used to do before with 45 categories but, again, it is more complicated and more involved in getting accurate data.

Safety reports are a particular concern, because our institutions are inundated with a number of reports that have to go to IRB and be logged and recorded, that never affect the design of a study or an amendment or the consent form because, if they did, then they would be an amendment, and that is a huge challenge as well.

In spite of efforts at informatics within the National Cancer Institute thus far, we still have groups all using different forms and different procedures and different electronic data interchange modes, not to mention that every sponsor is different, if we do studies outside of the national group programs. So, these are the big challenges.

TOP

How it affects me at one institution? This is what we have had to do for regulatory activity at Fox Chase over the last few years.

You see that, although our approvals, amendments and adverse event reports has been fairly constant, the number of safety reports is now close to 3,000 per year that we file, and have to be logged and received by our institutional review board. Many of you probably have similar experience at your institution. It is just hard to keep up with it

TOP

It has necessitated a huge expansion in staff, required at an institutional level, to run trials.

I mean, a lot of what we do is manage Cooperative Group studies, but we need over 40 people in an independent cancer center environment, to be able to do that, and to follow all the regs and everything else.

TOP

Now, we can solve these problems, and there are shared responsibilities, and this is Washington, D.C. So, I thought this would be particularly appropriate.

Some of this is pretty simple, and we just need to talk to each other, communicate better, and try to share ideas and information.

So, what are the solutions? One, we have got to work better with CTEP and the Investigational Drug Branch.

TOP

We had, actually, what I thought was a thoughtful and useful meeting yesterday that Ted Trimble kindly pulled together with some representatives from the IDB to just sort of re-energize that relationship and make sure that we had a good representation on our developmental therapeutics committee from CTEP, and move forward in terms of conference calls, meetings, and joint prioritization.

Translational objectives, this has been discussed earlier, and I won't belabor it. I think the most important thing is selective rather than universal incorporation of translational objectives, in view of the cost and complexity of those analyses.

Use core labs whenever possible, collaborate with committees in the SPORE to facilitate these studies, utilize the tissue bank as much as possible for handling specimens, and the statistical and data center for tracking and data.

Those are very concrete improvements that are really enabling us to do better translational research with more reliability and with more verification of data and outcomes.

TOP

More solutions. Some of this has already been mentioned, and incorporating something like CA 125 as a criteria for response.

Now, you might not use that as the primary endpoint of a trial, but it could be a decision point for whether or not you go to second stage, or whether you go from a phase II to a phase III or something like that.

Biochemical response could also be an endpoint, but we have to study it. We can't just assume that it is going to be okay, because we don't really understand, in particular, with some of the newer biologics, how CA 125 may be affected by them.

We would very much like to enroll patients with non-measurable disease. Ovarian cancer patients, as our advocates will attest, tend to be monitored fairly closely with exams, history, CA 125, CAT scans, and so on and so forth.

Frequently, we are diagnosing recurrence with very small volume, asymptomatic disease that is not measurable. Therefore, they are not eligible to go into most phase II studies as currently written, but they generally get treated anyway, and then they are not eligible to go into phase IIs, because they have already been treated.

So, it is a self-defeating process, and we need to find a way to include those patients within our phase II trials, and I think this is a very important point.

We need to look at new approaches to design trials to address these points. With Mike Sill and Mark Brady and other folks in the statistical and data center at GOG, we have been starting that process.

First of all, Mike has developed a statistical model with dual primary endpoints that looks at both response rate and progression free survival.

In some cases, we like to use single stage accrual designs to avoid the delay between the first and second stage, particularly when we are working with active agents, and we would like to look at more randomized phase II trials for a number of reasons.

TOP

So, you know that bevacizumab has activity in this disease, and Bob Burger touched on this earlier, and he presented this abstract at ASCO.

Overall, there was a response rate which was unanticipated. The trial was originally written with progression-free survival as the primary endpoint, because we know that cytostatic agents don't shrink tumors.

To our surprise, and much to our pleasure, there was a substantial response rate in this setting, and then what appears to be an improvement in time to progress, or progression free survival.

So, the group would now like to do more with this agent, which I think makes a lot of sense. So, there was a proposal that Brad Monk brought forward in collaboration with Genentech and the National Cancer Institute to look at combinations of bevacizumab and erlotinib, which has been discussed in a number of different tumor sites.

I think how the group responded to that opportunity illustrates, again, what might be part of the solution.

TOP

So, here is what we have now approved as a concept, that is under rapid development, which is bevacizumab in both arms, as in the prior trial, and then one arm with an oral placebo and one arm with erlotinib.

Then, limited it to patients with ovarian cancer less than or equal to three prior therapies, measurable or evaluable, but we are not requiring RECIST measurable disease, but we are controlling the performance status and vital organ function, as you might expect.

This is going to use a single stage design, because it already includes a known active agent in this disease. As I mentioned, we will have RECIST and non-measurable disease as well.

It is what I would call a decision based randomized phase II design. It is not strictly non-comparative, it is not really comparative, but it has a decision rule attached to it, where the statistics are powered as such to determine if further study is indicated, meaning a fully powered phase III, which Genentech has agreed to support, or no further study, or inconclusive.

With that, we have a sample size of about 140 patients, which is very reasonable, for a two arm randomized phase II trial.

This is not written in stone. This could change. Mark Brady is shaking his head. I must have said something wrong. It is an example of how we have tried to think creatively about the problems we face and try to work to try to solve them.

Russ Schilder earlier used the expression between a rock and a hard place.

TOP

I think we are there, but I am still there. I am still having a good time, and I will be doing this for a little while longer. Thank you.

TOP