Kathy D. Miller, MD; David J. Kerr, MD; Maurie Markman, MD; Robert L. Comis, MD


June 20, 2011

This feature requires the newest version of Flash. You can download it here.


Kathy D. Miller, MD: Hello. I'm Kathy Miller, Associate Professor of Medicine at Indiana University School of Medicine in Indianapolis, Indiana. I'd like to welcome you to an Oncology Insights Roundtable.

Clinical Trials: What Needs to Change?

Our topic today is "Clinical Trials: What Needs to Change?" Several colleagues will be joining me in this discussion: Dr. David Kerr, Professor of Cancer Medicine from the University of Oxford, Oxford, United Kingdom, and President of the European Society of Medical Oncology; Dr. Maurie Markman, President of Patient Oncology Services and National Director for Medical Oncology at the Cancer Treatment Centers of America in Philadelphia, Pennsylvania; and Dr. Robert Comis, Professor of Medicine, Drexel University College of Medicine in Philadelphia, Group Chair of the Eastern Cooperative Oncology Group (ECOG), and President and Chair of the Coalition of Cancer Cooperative Groups. Welcome to all of you.

We have several topics that we would like to discuss today, all related to current issues in clinical trials. Let's start with a tough one, and one that some of the information at this meeting brings to the fore. Is it ethical to launch or to require a phase 3 trial when early data overwhelmingly suggest the likelihood of benefit? This question has come up in many trials, but let's start with the most recent one -- the use of a Braf inhibitor in patients who have mutated Braf with metastatic melanoma. We saw the results from a phase 3 trial at this meeting that showed a profound improvement in response and progression-free survival (PFS) and overall survival (OS)[1] -- a phase 3 trial that was done because it was mandated by the regulatory authorities. Maurie, I know you have thought a lot about this. Are those trials needed?

Unethical Phase 3 Trials

Maurie Markman, MD: Somebody says that they are needed. Your question is whether they are ethical, and I think that they are absolutely not ethical. I ascribe very strongly to the basic ethical principle of equipoise, which argues that the reason you can do a phase 3 trial is that there is a general agreement within the community that arm A and arm B are possibly equivalent, and Dr. A could just as easily pick arm A and Dr. B could just as easily pick arm B or vice versa. There may be a little disagreement, of course, and somebody might put a patient on one trial and not on another trial. But when you have data that are so overwhelming -- and I want to emphasize that that doesn't mean because an investigator says that it's overwhelming [but rather] because one has reviewed the data, looked at the CT scans, and whatever one wants to look at, and it's agreed that the data are very different -- I don't think that it's possible to say that it's ethical to do this from the perspective of the individual doctor and the individual trial. The question of what the regulators say is a different matter. I personally would not have been able to put a patient on the trial that you described for the reasons that I've given. In The New York Times Amy Harman wrote a painful story a number of months ago about that particular trial and the impact on human beings [who were not permitted to cross over to the experimental arm when their disease progressed]. I think that's difficult. There has to be a better way of doing it, and I feel very strongly about that.

Dr. Miller: We'll talk about the regulatory implications in a little while, but I have to push you on this because I can come up with examples in which many of us thought that the evidence was overwhelming. At this meeting, a phase 3 trial of iniparib, initially thought to be a PARP [poly(ADP-ribose) polymerase] inhibitor, was reported.[2] It's not clear what iniparib does, but PARP inhibition is not among its mechanisms of action. Phase 2 data at the plenary session 2 years ago led to the same angst that this is overwhelming and the drug ought to be available. Yet, when put to the test of a phase 3 trial, the results were not nearly so striking. It was not a positive trial. If you're going to make that leap, what's overwhelming?

Dr. Markman: I'm sure that you want the other panelists to voice their opinions. I'm not talking about a question of "does it improve overall survival?" I'm not talking about whether it has a statistically significant difference on PFS. I'm talking about patients with metastatic melanoma in which there is no treatment. There is no hope, no treatment. Dacarbazine is not an active drug. To say that it is, is just not true. It's a drug that has been used for 30 years because we have nothing better to offer, and I would personally argue, having treated very large numbers of patients, that you could say that the treatment is almost as bad as the disease. We have a drug (again, I have no personal experience with this drug [ipilimumab]), but the argument is that we have 50%, 60%, and 70% response rates that have genuine, major clinical impact on those individuals for 3 months, 6 months, 9 months, and 12 months. I don't see how that has anything to do with what the end result is in terms of survival.

Now there is a different question: How do we pay for it? I understand all of those arguments, but I'm talking about making it available, figuring out how to allow the drug to be used, figuring out cost benefits -- however you want to look at it. To say that you did a randomized trial and worse, not to allow a crossover -- that was the worst part of it. I certainly could have made the argument that giving the Braf inhibitor to patients with melanoma after they progressed might have been just as active. That would have met my qualifications for equipoise.

Dr. Miller: I was thinking when you were saying that "that's a trial I couldn't enroll a patient on" --

Dr. Markman: Because they weren't permitted to get the drug when they progressed.

Dr. Miller: But if you enrolled your patients on that trial, half of them would get the drug. If you say "I don't think this trial is one that meets the requirements of equipoise, so I'm not going to participate," then none of your patients get the drug.

I would like input from our other panelists because this is an area in which reasonable people might disagree. I will assume that our other panelists are reasonable. Bob, you chair a group that does a lot of phase 2 and phase 3 trials. This must be an area you have thought about.

In Defense of Phase 3 Trials

Robert L. Comis, MD: Yes, it's a key area, and it's a key area in oncology in general. Throughout our careers we have lived through some phase 2 trials that have crossed the cure barrier -- the Einhorn regimen, for instance. After that was presented in the early 1970s, we all went home and tried it, and it worked. Treatments for Hodgkin lymphoma and non-Hodgkin lymphoma [have shown similarly positive results]. Occasionally, in oncology phase 2 studies cross the cure barrier.

Imatinib didn't cross the cure barrier in the beginning, but it was profoundly active. If, in fact, [a drug crosses] the cure barrier, then I think equipoise is out, but when you have a treatment that really looks active, it's important to prove for all sorts of reasons -- to the patients, to the practicing physicians, and to the community at large -- that in fact this treatment is an advance. Now, right now, in spite of the fact that it's a rational approach, it's based on genetics and molecular biology; it [ipilimumab] is not crossing the cure barrier, so I think to give it a rigorous test is really important. I tend to agree that in that particular study there should have been a crossover, and crossing over is a big issue that you have to deal with, particularly once you have defined a patient population. The whole issue of equipoise in the genetic era is an issue that we are going to be grappling with thankfully with these new agents that are very highly active in unique molecular settings.

Dr. Miller: David, although we do share a lot of commonalities, there are sometimes differences of thought on the other side of the pond. Is this one of those areas?

David J. Kerr, MD: No, I think we're sharing a set of common values at equipoise and support exactly what you have both said. I worked in the Radcliffe Infirmary in Oxford, first founded in 1700 something, and the first-ever trial of penicillin was done there. In the Dunn School they manufactured enough penicillin to treat 6 patients. This was the world supply of penicillin. They had a ward full of pneumonic patients who had just come back from the Second World War, and they did a randomized trial. It only took 12 patients because the answer was so overwhelmingly powerful. There is a bit of me that still thinks that biology is king. Even within modern genetics we have to contextualize it within different pathways and different ways of looking at things. I still feel a hard-line defender of the phase 3 trial, and if we have something that is wonderfully powerful and effective, that means that we can power the trial to be small, quick, fleet of foot, and to get the answer in some way that is rapid and makes sense. I like the sense of randomization because biology is king, and how can we account for that unless we balance across both arms?

Dr. Miller: So the idea of genetics and the change in our understanding of molecular biology has come up a couple of times, and our current clinical trial system does a poor job of accommodating that. With this explosion of genetic data, more rapid development of specific or sometimes less-specific inhibitors based on what we think we have learned about the biology, we have an ever-increasing number of new compounds to be tested. Is the current clinical trial system even sustainable?

Dr. Kerr: I think we have only scraped the surface of it, and for me it's all about context. To make an assumption that there is a single dominant oncogenic pathway that if we take out all the others will fall aside has proved to be simplistic and naive. There's crosstalk. There's pathway redundancy, etc. There are some examples in some of the leukemias in which there are pathways that seem to dominate in that way, and although we're terribly attracted to the concept of personalized medicine, of linking a drug to a companion diagnostic or a biomarker of some sort and segmenting the patient population -- right patient, right drug, right time, and so on -- I think we're standing in the foothills of it. I think as Francis Collins from the National Institutes of Health (NIH) has said, we are the standard leaders for personalized medicine -- more than cardiology, diabetes, and so on -- but I think we're in the foothills of it. The more we get into it, the more we are going to have to understand about the relative complexity that there are many types of melanoma, not just a single one. But the future is bright.

Dr. Miller: In some ways they are almost competing needs. The understanding of the biology and the greater subsetting of patients means that in reality, every disease that we take care of may be a rare, orphan disease, which means that we need greater collaboration to do clinical trials because no one center -- no matter how large -- will be able to do that. I have actually been struck in some ways that our colleagues in Europe do that better than perhaps we do. The neuroblastoma study by the pediatric group at the plenary session[3] -- a group that did not exist until they recognized the need to do this study of the most common solid childhood tumor, but still overall a fairly rare tumor -- brought together many centers across 20 countries with all of the financial and regulatory barriers for the sole purpose of doing this trial, purposes that we'll now extend. How have you been able to do that?

Cooperative Groups and Global Trials

Dr. Kerr: I'm a great fan of the cooperative groups, and I think we are following where you have led. I would be very interested in your experience. However, still in the United States I am guessing that 2%-3% of eligible patients are entered into trials. In the United Kingdom, now with concerted effort and by building research networks with investment from funders, from government, from our National Cancer Institute (NCI) equivalent, we now have 15% of eligible patients represented in clinical trials. I guess that depth of experience means that the trial results will be more generally applicable to the population of patients who we're looking after. I like the notion of celebrating our ethnic and genetic diversity. I like global trials. Look at all that we have learned from doing trials with our colleagues in Japan, China, and the Middle East and how we have delivered molecular cues from there. I would like to see us doing more global trials, how our cooperative groups could interact holding hands across the pond and stuff like that. It makes sense.

Dr. Miller: Bob, the cooperative groups are undergoing transformation right now driven by many factors including the financial reality of NCI's budget. What changes have happened, and what changes are coming in our cooperative group system?

Dr. Comis: First, I'll address the issue of biomarker-driven studies. In the Eastern Cooperative Oncology Group (ECOG) we have, as Kathy well knows, just coordinated 2 studies, which are biomarker driven: one using the genetic markers of the Oncotype DX® [Genomic Health; Redwood City, California] in which 10,000 women were screened, and another in colon cancer in which we looked at 18Q deletions and microsatellite instability. It wasn't on fresh tissue. It was on archival tissue, but we were able to establish a system that turned around that information for doctors and patients within 5 days. So it can be done, and I think we have to do it and we will do it.

The system in the United States is being restructured and reorganized. We are going to go from 9 adult groups down to 4 adult groups. I'm most intimately involved with what we are doing in ECOG, and I'm very excited about it because we are combining our forces with the American College of Radiology Imaging Network, which is the most sophisticated imaging network that exists in the country. It will give us tremendous capabilities in biomarkers and adaptive designs. Each of the groups in the consolidation has the vision of the future in mind, so I don't think it's a bad thing. I think it will help prepare us for the future.

The difficulty is that the government is strapped. Worldwide there is an economic crisis, but we are in the midst of this biologic explosion, and I think we have to work together with the government. We have to work together with industry, and we have to work together with our partners throughout the rest of the world. There has to be an IT system that really allows us to communicate and work together. I agree with the diversity issue. The more that we do together the better off we are, and we'll need it. No single institution, perhaps not even a single country, could do a study as quickly as it needs to be done.

Why Community Oncologists Shy Away

Dr. Miller: Maurie, although the academic institutions frequently design and lead large studies, most of the patients come from community centers.

Dr. Markman: Right.

Dr. Miller: That has certainly been true in the US cooperative groups, but there are unique pressures and unique issues that come from doing clinical research in the community setting and ones that become more complex as we talk about getting fresh tissue and biomarkers and the turnaround time that is needed. It feels a bit like that drives more people from the community out of participating in trials at a time when, for all of these issues, we most need to bring them in.

Dr. Markman: This is truly a tragedy because I think, and I'm sure Bob would agree, that we have an enormous resource in the community oncologists who want to participate. They want to figure out how to participate because they believe that this is good for their patients. This is good patient care. It gives patients the opportunity to receive agents early. They want to participate, but it's very hard. The financial pressures are tremendous. The funds that are available certainly from the federal government to support this work are very difficult, and for all of these reasons it becomes more and more difficult. The more they [physicians] have to do, the more information that they have to provide, the more tests that they need to do -- which, quite frankly, they may think are only important for regulatory purposes, not for patient care -- [the more] it drives them away. People say it's a revolution. I have to agree -- a revolution in our understanding, and I don't use that word lightly because we have heard "revolutions" for our entire careers. But we are there. This meeting showed that we are there, not in every tumor --

Dr. Miller: Is a revolution above or below a paradigm because there are lots of those, too?

Dr. Markman: Years ago, I talked to my renal cell cancer colleagues who used to have no treatment, and they were smiling. At this meeting, doctors who have made a career of taking care of patients with melanoma, and now, we see their excitement at being able to offer their patients something. We're going to see more of these [advances]. Clearly, trials are what allow us to get to where we're going. We have more products; we have more drugs; we have more strategies driven by spectacular science; and we get to the point at which [we're wondering]: How do you do the trials? How do you get the patients on the trials? What kinds of trials? What are the requirements? How much does it cost to do the trials? Who is going to help do the work? We have to figure this out, and I don't have the answers. It's not getting easier. It's getting harder.

The Endpoint Debate: OS or PFS?

Dr. Miller: I'm going to come back to you for an answer in a minute because the endgame of all of this is if we want these drugs to be available to our patients, they have to be approved by the regulatory authorities. David, in Europe what does it take to get a drug approved? What's a valid endpoint that will get you from a trial to approval?

Dr. Kerr: That's a very interesting question. I was at a meeting yesterday, a fantastic meeting of some wonderful patient advocates for a variety of tumor types, and we had the same debate. These were very sophisticated patients, and we had a very interesting and detailed debate about OS vs PFS as an endpoint. I think as an academic scientist trying to work out [whether] a drug has some effect on the tumor biology, things like PFS are valid endpoints in terms of hard science and understanding the validity of the drug and the target and so on. Stepping back from that, looking at value in cancer care, I'm attracted to OS. It's difficult because of multiple lines of treatment. How can we ever truly define what the contribution of an individual drug or treatment has actually been? It's like our earlier discussion about crossover confusing OS but giving us a cleaner endpoint and being more balanced to the patients involved in it, so it's really tricky. If I were a professor of public health, if I were a health minister, if I were a funder, then I'd be thinking that for the population of patients OS is what I want to invest in. That's what I want to understand. However, as physicians, as academic physicians doing the best for our patients, we need to understand if drugs work, and that pulls me back towards things like PFS, that sort of thing, as a valid end point endpoint.

Dr. Miller: If we take Maurie's position that overwhelming evidence of benefit -- you [Maurie] would certainly include Bob's examples of [instances] when clearly you have cured a disease -- but I think you were willing to go for [evidence of] overwhelming benefit short of cure.

Dr. Markman: Oh, absolutely.

Dr. Miller: So if we suddenly made you "king of the world" or "head of the FDA," what should it take to get a drug approved?

Dr. Markman: I think the problem here is that we are talking about different things. If we could eliminate the discussion of cost, we would probably be having a different discussion. Perhaps even the FDA would have a different discussion, I don't know, and perhaps even in Europe they have different discussions, but cost can't be eliminated.

Dr. Miller: Drugs do cost money.

Dr. Markman They cost a lot of money, and they are costing a lot more money.

Dr. Miller: Someone has to pay for them.

Dr. Markman: If for the moment, we can take that off the table in terms of the discussion -- I don't want to take it off the table completely because it is important -- but my definition of clinical benefit is not just if the patient lives 5 or 10 years. The melanoma example -- I'll come back to that because it's so uncommon.

[Until recently], not a thing you can do -- pain and suffering. I'm reporting what others say because I don't treat melanoma, but [with ipilimumab] patients' pain goes away. They are able to carry on with their lives. No one would suggest that it's a cure, but it might last for 3 months, 6 months or 12 months when there are no other options. That, to me, is clinical benefit. Now, sure it's not curative therapy, but I consider that to be clinical benefit. You are helping patients. That is what we should be doing, and I believe that is a valid endpoint.

Recognizing that [a drug provides clinical benefit], you then have to talk about cost and our society's ability to pay for it. But I do believe that these are separate discussions, and we shouldn't mix them when we are trying to talk about clinical benefit. If we are talking about cost, let's talk about cost. If we're talking about clinical benefit, let's talk about clinical benefit because I'm concerned in many of our discussions that we are throwing in cost, and that's really what we're talking about, but we don't have evidence of clinical benefit.

Dr. Kerr: A surrogate discussion, isn't it?

Dr. Miller: We'll leave cost off the table for a minute. Bob, Maurie has retired and you're the new head of the FDA. What's your definition of clinical benefit?

Clinical Benefit as an Endpoint

Dr. Comis: It's better to be in remission than not, so I think that treatments that have a profound effect on PFS and a profound effect on the momentum or trajectory of the disease have to be considered as having an impact on the overall disease process and the overall patient's experience in that process. So I think that PFS, being in remission and having your symptoms alleviated are extremely important, and I do agree that the survival endpoint, which the FDA has been obsessed with until recently, is one that really needs to be re-examined. We're in the midst now as you very well know, Dr. Miller, of a situation in which your study has shown that we were able to double PFS.[4] It wasn't associated with a clear survival benefit, but back to the design of studies, instead of having 700 patients where you showed that you could double PFS, you needed 3000 patients to show that you improved survival by a few months. So there is a balance here, but being in remission associated with quality-of-life issues -- not just measurements but the kinds of things that Maurie's talking about -- is very important. I don't think any regulatory agency should turn its back on that, and thank God that the European agency hasn't with respect to bevacizumab and breast cancer, so we're hoping that people in the United States will be able to benefit from that as well. But who knows?

Dr. Miller: We should be clear here. We all want our patients to live longer.

Dr. Comis: Exactly.

Dr. Markman: Yes, there's no objection.

Dr. Miller: It is simply that we would all agree that there are times when there is real benefit to patients to be had that stops short of an improvement in survival.

Dr. Markman: Or an improvement in survival that you could document to be statistically significant in a randomized, phase 3 trial.

Dr. Miller: In a trial that was feasible to do on the basis of the size and the statistical considerations.

Quality of Life: Does It Measure Up?

Dr. Kerr: Clinical benefit, you described it beautifully, and I want you to capture that into a scale that we can quantitate and establish because quality of life just now doesn't do it for me.

Dr. Comis: In fact the regulators in the United States have not agreed on what measures should even be done, but when you present something that doesn't show an improvement in quality of life, they say, "oh, no quality-of-life improvement," and they haven't really declared what the measurements should even be.

Dr. Miller: The quality-of-life scales that are used have several problems. They were designed a couple of decades ago, so they were designed to capture symptoms of fairly advanced cancer and toxicities primarily of cytotoxic chemotherapy, so if you have a patient population that is not particularly symptomatic and a therapy that doesn't behave in its toxicity profile like a cytotoxic chemotherapy, they [the quality-of-life assessments] are simply not a match. Then, there are the practical trial issues. Although we might design a study that says we are going to get the quality-of-life assessments at these weeks on everybody, in reality what happens is once they progress, they say "sayonara." They have moved on to other therapies, and you get no more data. So you get both the problem of fewer data over time, and you get the quality-of-life assessments on people who from both a disease and toxicity front were doing well enough that they were still on the trial.

Dr. Kerr: Exactly.

Dr. Miller: So not surprisingly they are never different.

Dr. Kerr: We were constrained. When we were putting funding in for trials, our NIH equivalent [required] quality of life as an endpoint or we wouldn't get funding so we put it in, but it was uninterpretable. It was just terrible masses of data that we couldn't interpret that didn't really add any value to how we closed the trial or how we supported and looked after our patients. It just became a burden on our backs rather than a joyous exploration as to how we could actually demonstrate benefit. It just became a donkey on our backs if there is such a thing as a donkey on one's back.

Dr. Miller: To quote a famous US Senator, it's a "bit like obscenity." We have a hard time defining it; we have a hard time measuring it, but we recognize it. We all have very little difficulty when we walk in a patient's room knowing overall whether the patient is better, or not better. It's usually quite obvious. I frequently tell fellows that I look at the x-rays a bit as an afterthought because they typically tell us what we already know.

We do have to -- for at least a minute or two -- talk about cost. I have to go back to one of the ethical questions with which we started. If we design trials completely outside of any cost considerations, and find a therapy that everybody including our friends at the FDA and EMEA [European Medicines Agency] would agree improves survival and has clear benefit yet most if not all of our patients simply cannot afford, have we done them good? Was this an ethical thing to do? Sorry, Maurie, this was the question that I was going to come back to ask you.

The Question of Cost

Dr. Markman: This is a serious problem, and this is the reason I personally think that as a society, we have to have a discussion about cost. We have to have a discussion about cost with our third-party paying world, which can include insurers and Medicare. We have to have a discussion about cost with the biotech/pharma industry as a group and [gain] an understanding of "what is your investment vs your profit vs the benefit that you [patients] obtain?" For example (and this is entirely theoretical), if you had a drug that in fact improved survival by 26 months, that could give you a pretty good premium. Compare that with another drug that could give you significant improvement in the quality of life (by a valid measure, such as no pain or symptoms for 3 months) that could give you a different premium. This is a big discussion, but the key here is that before they ever start, the companies [need to] know what the ground rules would be. If it wouldn't be that [the pharmaceutical companies say] we're spending all of this money, and at the end of the day we can only charge this much. It makes it easier to do the trials. This is a big discussion, but the key here is that before they ever start, the companies need to know what the ground rules would be. It makes it easier to do the trials. I know with the ECOG study what had to be done as a company was to go back afterward to get all the x-rays and have them looked at. That's a cost, and they passed that on. You could argue wouldn't it have been better to say [to the company]: "Okay, you're going to make a little less profit over the next couple of years?"

Dr. Miller: You made less profit because your drug had benefit, but less benefit.

Dr. Markman: It's a different discussion from "should we have the drugs." It's related, but it's different. How we pay for the drugs is a very important but separate discussion.

Dr. Miller: I personally think -- not being a European but sometimes wishing I was -- that the Europeans have gotten that part right, that they have done a much better job separating cost from the discussion. The EMEA approves drugs and makes regulatory decisions, and then individual countries and their own health systems make their own independent decisions. Can we afford it? Can we pay for it? Is the benefit enough to justify the cost, and really where do we want to put our money? It seems to me that those 2 related but different issues are less jumbled [in Europe].

The NICE Way of Doing Things

Dr. Kerr: There is a societal issue here, and for me it's about transparency and openness. We don't suffer from a lot of that, and I think the more of that the better. We have a thing in the United Kingdom called NICE, the National Institute for Clinical Excellence ("cost-effectiveness" if we're being honest about it). We take drugs that are approved, and with some very transparent and open pharmacoeconomic modeling, we say the quality-of-life years added for drug X and Y are Z -- however many months or not it might be -- and the cost of that is Y. Then it's up to society to decide where we set the gain in that. If we do it deliberately, if we apply the same methodology to each drug, and are open and transparent [about it], citizens can debate, and industry can come in and we say "this is what we think we can afford." People know and understand; it's not secret and hidden, or just driven by politicians or an attractive young mother with breast cancer making a plea for a certain drug -- at least [that way] we have a system for doing it.

NICE is becoming increasingly influential in Northern Europe. People are looking to it. Whether in the United States you would have something that's similar, you see, it's an adaptable model. If you say as a society that we want to set the dollar equivalent for life here rather than here, or here as it may be in India, or here as it may be in sub-Saharan Africa, I'm interested in seeing in the global cancer village, we're talking about cost at the very top end of the market. If I'm a cancer doctor working in Ghana, I have a very different view: The health minister of Kenya has $8 per capita of the population for his entire health spending, $8 per person per year for everything. So we're arguing at the top end, and I think there's a scalable model; it's transparency, openness, and honesty, and people buy into that.

Dr. Miller: Bob, the patient advocates have played a really big and growing role in the cooperative groups, and these are issues that they are very involved in and very concerned about. Do you see this as a topic that will start to be investigated in cooperative group trials? There's a lot of cost that we don't know about. It's easy to figure out the drug costs, but there are a lot of other things that we do as a component of care or things that are ineffective. Maurie mentioned dacarbazine earlier. It doesn't work, but it still costs money.

Does Money Drive Clinical Trial Design?

Dr. Comis: I certainly think that we are going to be getting more into health economics in association with our trials. In some of the adjustments in the group system there are experts in comparative effectiveness research being integrated into heath economics research. This is a critical component of the clinical trials process, but I would like to emphasize that the clinical trials process is part of the scientific continuum. We design trials to try to prove that we have improved patient care, standards of care -- whatever the measure may be. I don't think that cost should drive the design of trials. The design of trials is driven by opportunity. I accept the fact that we all have an obligation once we show that something is active, but I would be very concerned about having a system that somehow impugned the development of scientific progress just because it might be expensive. The most cost-effective way to treat cancer is to cure it, period. I don't need an economist to show that, but I do think that the hard part here is in macroeconomics. You have to understand the macroeconomics of this in an individual country, and so I think it is a complicated situation, but I do think that it's imperative for us to start to address these questions early. I don't want to see us shying away from a potential advance because it might be expensive.

Dr. Miller: Some have argued that in some ways the finances already drive trials because so many of our trials are done by industry rather than with public funding and in the cooperative groups. In industry the decisions about trials that are done are certainly not devoid of science. I wouldn't suggest that, but they are very much informed and influenced by what it is going to take to get my drug registered, which gets Maurie not participating because the trial is not ethical. Is there a market? What might the results of this trial do for my market share and my ability to sell more of the drug? There are times when those trials include very important scientific questions and times when from a scientific standpoint they might be a bit ho hum.

Dr. Kerr: I think you've touched on something important, as did Bob. In addition to the element of cost is the cost of trials in that when we have an increasing number of new drugs and new chemicals rushing bottlenecked into the clinic, patient availability improving, the quality of our science improving, but the cost of the trials wherever is $50-$100 million to do a large, randomized, phase 3 trial -- $50-$100 million. How did that happen? I just wonder if when actually running the trials, the regulatory bodies need to look at just where the costs lie. Could we do less monitoring? Could we take samples? Could we actually reduce the cost? If we reduce the cost, does that mean that we can more quickly gain entry for more new interesting, your important point, and intelligently designed drugs rather than --

Dr. Miller: And perhaps get more enthusiasm from our colleagues in the community who have been a bit put off by the layers of regulations and monitoring?

Dr. Comis: In a study we did a few years ago, we estimated that 35 cents of every research dollar that is spent in the clinical research process goes toward regulatory issues, 35 cents on every dollar.

Dr. Miller: I would bet that it's more now.

Dr. Comis: That's right. That's 2000-2005. This is a huge issue particularly in the United States, and I think that it does impair the ability of interested clinicians, clinical investigators from both the community, and academic settings in the United States to participate in trials. The regulatory morass here is strangling.

Closing Remarks

Dr. Miller: So much has changed, and there is much still that needs to change in the way we conduct clinical trials both in the United States and in Europe. But this discussion has at least helped bring to the forefront and focus some of those major issues. I want to thank my colleagues and our audience for joining us today. This is Kathy Miller for Medscape Insights: Oncology at ASCO® 2011.