ASCO® 2010 Wrap-up: Converting Cancer to a Chronic Disease
Maurie L. Markman, MD: Hello. I am Maurie Markman, Professor of Medicine and Vice President for Clinical Research at the University of Texas MD Anderson Cancer Center in Houston, Texas. It is my pleasure to welcome you to Medscape Oncology Insights Wrap-up for ASCO [American Society of Clinical Oncology] 2010. I am joined today by Dr. Kathy Miller, Associate Professor of Medicine at the Indiana University School of Medicine in Indianapolis, Indiana; Dr. John Marshall, Professor of Medicine at Georgetown University and Director of the Ruesch Center for the Cure of Gastrointestinal Cancers, Lombardi Comprehensive Cancer Center in Washington, DC; Dr. Mark Kris, Chief of Thoracic Oncology at Memorial Sloan-Kettering Cancer Center in New York City; and our special guest, Dr. Richard Schilsky, Professor of Medicine and Section Chief of Hematology/Oncology at the University of Chicago Medical Center here in Chicago.
Our purpose in this short discussion is to talk about our overall impressions of ASCO 2010. We were here at a very exciting meeting, a very big meeting -- it is only getting bigger -- and we have been hearing a lot of new and innovative ideas.
I will take the privilege of starting the discussion because it's actually now 2 years in a row that the first talk at the plenary session of ASCO dealt with the management of ovarian cancer, my own area of research.
Last year was very interesting and brought a very controversial discussion regarding management of patients who are asymptomatic with a rising CA-125 antigen level. That discussion suggested that perhaps that management strategy shouldn't be considered routine. This year, we took the opposite approach, and we looked at the results of a very exciting trial. The researchers in ovarian cancer have been waiting for more than a decade for a frontline ovarian cancer chemotherapy that was not intraperitoneally based and that would show an impact on progression-free and overall survival.
The Gynecologic Oncology Group [GOG] Trial, known as 218, was reported at the plenary session. The data in this trial were extremely interesting because they showed an improvement in progression-free survival, but there was no evidence that the addition of bevacizumab to carboplatin and paclitaxel improved overall survival. Many questions were raised, including the cost of this treatment, whether one could give bevacizumab only in maintenance rather than with chemotherapy, and whether to give bevacizumab at the time of recurrence. This was an 1800-patient, randomized, double-blind, placebo-controlled study. This was a very interesting study because it demonstrated that even an outstanding trial raises as many questions as it answers. This says a lot about oncology today and about the information presented at the meeting this year.
With that introduction, I want to turn to my guests and ask for impressions of the meeting in your own areas or in general.
John L. Marshall, MD: I'll come back to this maintenance approach because we saw the same sort of thing in the colon cancer world, not with a new medicine, but instead with an old medicine for us, bevacizumab. Our concept has been to keep giving chemotherapy until progression or toxicity, but we know, say with oxaliplatin, that we can't do that, we have to back off because of cumulative nerve toxicity. What do you do when you back off? This study looked at continuous chemotherapy vs chemotherapy after 6 cycles, then maintenance bevacizumab. It turned out that this approach held the curve up just as well as continuing chemotherapy. This concept is emerging, that bevacizumab perhaps is not such a good single agent in advancing adenocarcinomas but might be an agent that suppresses the regrowth of suppressed adenocarcinomas. I think both of these trials are very similar in design and concept.
Dr. Markman: Absolutely, and I would add one thing. Cancer is increasingly a very serious but nevertheless chronic disease process, certainly in the solid tumor area, and it needs to be managed. It's not a matter of just 1 treatment. It may require long-term treatment or multiple treatments over a long period of time. This raises issues of the toxicities of chronic treatments and the costs.
Richard L. Schilsky, MD: Something that we have to be really thoughtful about, though, particularly with these expensive and potentially toxic therapies, is whether we can demonstrate a survival advantage. Let's remember that the patients who are in clinical trials are mainly asymptomatic patients. The eligibility criteria for the clinical trial typically is PS [performance status] 0-1 patients, so they are asymptomatic or minimally symptomatic. To keep them on chronic therapy to slow the rate of tumor progression without ultimately demonstrating an improvement in their overall survival should cause us to question how they are actually benefitted. They were asymptomatic when they began; we put them on long-term treatment, which exposes them to toxicity and certainly increases their cost of care and, ultimately, may not improve their survival. So, where is the real benefit? I think this is a real conundrum that we face both as physicians and as drug developers. There are many questions in the regulatory environment as to whether progression-free survival is a suitable endpoint for drug approval, particularly if there is no advantage in overall survival.
Kathy D. Miller, MD: This is an area I've thought a lot about. It has been a big question in breast cancer with bevacizumab, and 1 key to answering the question that you asked -- we simply don't have. Our quality-of-life data are fairly limited compared with the clinical data. We do have quality-of-life assessments for people who remained in the trial. If they either stopped the treatment in the trial because of progression or toxicity, we don't have quality-of-life data. One of the arguments for using a therapy that substantially improves progression-free survival is if it doesn't add to the toxicity burden in a way that affects the quality of life for a group of patients who have no symptoms or minimal symptoms, even if that therapy is not free of toxicity. If I can prolong that period of time [of progression-free survival], that may be independently beneficial.
Dr. Schilsky: Time to onset of new symptoms may ultimately be that useful clinical -- and maybe even regulatory -- endpoint.
Dr. Miller: Exactly, because inherent in that assumption is that progression of cancer, at least in some patients, will bring with it symptoms of the disease. If I can prolong that [period of progression-free survival], then I prolong time to symptoms. Progression also brings with it treatment with other therapies, some of which may have more side effects on a day-to-day basis. Unless we actually have sequential quality-of-life data in people irrespective of whether they stayed in our studies, we ultimately don't have a way of sorting out whether we do more good than bad.
Dr. Schilsky: There is also the whole issue of patient well being. Our patients live from scan to scan, and every time you can sit down with the patient and say that the tumor has not progressed, you know that they go home feeling okay. You also know that when you tell them the tumor has gotten worse, they don't feel so well. There are important psychological benefits to maintaining tumor stability, but at the end of the day the question is: at what cost to both the patient and to the healthcare system? Does that really become a viable tradeoff?
Dr. Marshall: We quibble about these issues because the magnitude of the benefit is relatively small. If the benefits were of greater magnitude, we wouldn't be struggling with all of these questions. As I sit through my 17th or 18th ASCO meeting, I long for greater magnitude of benefit. I wonder how we are going to shift our world in a way to where we get some bigger and better outcomes.
Dr. Schilsky: To some extent we have to create that world, both by developing better drugs and also by stating up front that we may not be willing to accept these small incremental benefits. That GOG trial, as I recall, had over 1800 women who participated.
It took 1800 women to demonstrate that 3.5- or 4-month improvement in median progression-free survival. That's a lot of patients to invest in a clinical trial to demonstrate a very small increment. Now you could say that it's still something. You could also say that perhaps those 1800 patients would have been better used in smaller clinical trials looking for a 6- to 12-month improvement. If you hit that goal, then you would have something that would really be a step forward.
Mark G. Kris, MD: First off, maintenance is the hottest topic. In the last year, the FDA [US Food and Drug Administration] approved 2 drugs for maintenance. I have a little different spin on maintenance therapy. To me and most of my patients, it's a no-brainer to continue maintenance therapy because you are only bringing this up in the patient for whom you have documented regression, both objectively and subjectively. You have also documented that that person finds that treatment acceptable. To continue therapy in that context is much easier. In the lung cancer arena, time off from cancer means time closer to your normal life, and that, in the absence of cure, is what we can offer our patients. I think that the real issue here is maintenance. The negative of the lung cancer trials -- the SATURN [Sequential Tarceva in Unresectable Non-Small Cell Lung Cancer] trial that led to the approval of erlotinib was just published in Lancet Oncology -- fewer than half of the patients in the induction went on to maintenance therapy.
I think everybody is saying the same thing. We need to think about what is going on here and keep our eye on the prize. If we do something for that group of people, we're affecting a lot more lives, including those patients who could benefit from maintenance, and that's where we need to shift our focus. That's a message for ASCO meetings in the future. We have to step back and think about what the real issues are and focus research.
There is the need to prioritize. I saw Dr. John Mendelsohn's presentation about the Institute of Medicine (IOM) report, and I saw Dr. James Doroshow's presentation about how we are going to improve clinical trials. One of the big issues there was prioritization, and I think that's something we have to think about. Are we going to do a trial that shows things are equivalent, or are we going to do a trial that shows that we are going to improve cure rates significantly or improve survival significantly? I think all of us would agree that the priority has to be to those trials. Cure is the number 1 goal; improve survival is number 2. Everything other than that, while important, is not as important as cure and survival improvements.
Dr. Schilsky: Mark, the other thing about the maintenance is that we really need a better understanding of the biology of disease stability. For example, if we could identify which patients were likely to have stable disease because they have a biologically more indolent tumor vs which patients were more likely to rapidly progress, that would be an advantage. John and I take care of colon cancer patients, and I'm sure we have both seen many patients to whom we decide to give a chemo holiday; then, months and months go by, and they don't progress off treatment. You know, if you had arbitrarily decided to switch to a maintenance regimen, you'd think the maintenance regimen is doing great things for this patient. However, all you have done is to identify a patient with a biologically less aggressive tumor. What we really need to figure out is which patients have a high probability of progression off treatment and therefore would benefit from maintenance and then focus our efforts in those patients.
Dr. Kris: Getting back to the molecular question, my group and the group at Vanderbilt have looked at the growth characteristics of these resistant tumors. In lung cancer, after resistance to gefitinib or erlotinib develops, a resistance mutation T790M develops. We have been able to look at the growth characteristics of those cells, and those tumors actually grow much more slowly. This has many implications. The ability to detect growth is very difficult and is even more difficult to design in clinical trials. You say what you may be seeing is slow growth and conclude that we're slowing the growth of the tumor, when you're not. The main thing is to focus on the biggest problems that patients face.
Dr. Schilsky: We had a really good example of this yesterday in the colon cancer session. There were very compelling data presented showing that patients with colon cancer at highest risk for rapid progression after first-line treatment are those who have BRAF mutations in their tumors. BRAF looks like a poor prognostic factor in general, but where it seems to have a major effect is after that initial treatment, and then these patients with BRAF-mutated tumors progress very quickly. That would be a population that we would want to focus our efforts on, at least to try to slow down that high risk for progression.
Dr. Kris: That brings us around to the molecular aspects of care, both for selection and for treatment. One of the next plenary abstracts after the ovarian abstract was the one in lung cancer. I think that study generated a lot of controversy. There aren't many phase 1 trials that are on the plenary program. It might be the first phase 1 trial that was done in a molecularly defined subset outside of a hematologic malignancy.
Dr. Marshall: PARP [poly-ADP ribose polymerase] inhibitors did a similar thing after some early findings.
Dr. Schilsky: And we did have the paper last year on the hedgehog inhibitor, which was a phase 1 trial in The New England Journal of Medicine.
Dr. Kris: The designers of the study said, "Here is a targeted therapy that is going to work only where you have the target. We have available technology to find that target, so let's develop the drug and give the drug to those patients with that target." That's what they did, and they had an unbelievable waterfall result; 90% of the patients had some kind of benefit. They had a very high objective response rate, and those responses seem fairly durable in patients who were accustomed to receiving cytotoxic chemotherapy weekly or every 3 weeks. Suddenly, they were taking a tablet twice a day, almost like patients who take tamoxifen in breast cancer.
Dr. Miller: I wonder if the investigators saw the same thing, as we got more HER2-targeted therapies and bevacizumab therapies that didn't have chemotherapy toxicity. This came immediately after the big days of bone marrow transplants. I remember enrolling women in their first bevacizumab phase 2 trial who had undergone transplant for metastatic disease. A couple of the responders actually complained about the lack of side effects.
Dr. Kris: They thought that that was a sign the treatment wasn't working?
Dr. Miller: Yes, but put yourself in their position. They had had the disease for years. They had undergone a bone marrow transplant, and they got through a lot of miserable therapy by telling themselves that this is what I need to do to stay alive, to get this tumor under control, and to get better. When treatment was so easy, they essentially said, "I'm not in the fight anymore. It can't be this easy." We were all saying that was the goal, but for patients the shift [from a miserable therapy to an easier form of treatment] took some time to accept.
Dr. Kris: Some patients have a difficult time with that, and they are almost happy to have a side effect because it's a sign that the treatment is aggressive.
Dr. Miller: It's doing something.
Dr. Kris: Also, sometimes it leads them to question us -- if we are giving them a treatment that is not causing side effects, is it really that good? But you're exactly right. Giving them more effective, less toxic treatment is what we want to do.
Dr. Schilsky: However, I think one of the things about which we need to be careful is the issue of adherence to treatment. Kathy, I'm sure you see this in your breast cancer patients, particularly with the development of more and more oral therapies. The issue of adherence to treatment is huge. There are developing data showing about 50% of patients actually stop taking their oral medicine within about 6 weeks of beginning treatment, and that's a terrible adherence rate for a therapy that is meant to be given on a long-term basis. Obviously, some oral therapies are more or less toxic than others, and you'd have to wonder about whether the patient who gets rash and fatigue is more likely to stop sooner than the patient who gets a less uncomfortable side effect. In general, we know from other areas of medicine that people stop taking their antihypertensives, people stop taking their anti-ulcer medicines. However, if we are going to try to convert cancer to a chronic disease, we're going to have to figure out ways of getting patients to stay on their treatments.
Dr. Marshall: Do we have to do a phase 3 study in the ALK [anaplastic lymphoma kinase] inhibitor? How can you?
Dr. Kris: There is one underway today.
Dr. Marshall: But how long is it going to take?
Dr. Kris: Your patient who has an EML4-ALK-positive tumor is going to be asked to participate in a phase 3 trial that randomizes them to get a standard chemotherapy, which, by the way, has been shown to improve survival, vs the EML4-ALK inhibitor. At progression, the patient can then cross over to the EML4-ALK inhibitor.
Dr. Marshall: With no overall survival; we'll lose that in the crossover.
Dr. Kris: We'll likely lose that, but that's okay. I think we'll have a very robust examination of benefit. The one nice thing about these therapies is that usually the benefit is obvious no matter what time you get it. So, even in that third- or fourth-line setting, I think you're going to be able to see pretty substantial benefit.
Dr. Miller: Are you worried the trial won't get done?
Dr. Kris: I'm worried the trial won't get done.
Dr. Miller: Why?
Dr. Kris: Show me the trial that tested this cancer for cisplatin. Why didn't we do that trial? We didn't do that trial because it was obvious.
Dr. Schilsky: Well, this trial will get done because there's no other way to get the drug.
Dr. Miller: There's not another way to get the drug. The drug is not likely to get approved in the current climate without the trial.
Dr. Marshall: But did oxaliplatin or cetuximab get approved without phase 3 trials?
Dr. Miller: If I were in charge of regulations, I'm not sure that I would mandate a phase 3 trial to approve this drug, but the FDA probably will.
Dr. Markman: We had that same issue with PARP inhibitors in ovarian cancer. PARP inhibitors occur in 5%-10% of an uncommon disease, and randomized trials are being done [in PARP inhibitors], but people have questioned that requirement.
Dr. Miller: It doesn't strike me that it would be hard to do this study.
Dr. Kris: If I could design the system, I'd probably have a 2-tiered approach. I would get the drug approved in a very strict way and only in the mutated patients. Then, of course, it's a totally different question to ask whether it is better than chemotherapy, because, ultimately, people are going to be asking should I give chemotherapy and, if so, in what order? We have seen that the patients with EGFR [epidermal growth factor receptor] mutations with non-small-cell lung cancer have a 50% response rate to chemotherapy. It's just amazing. Clearly, chemotherapy helps many of those patients. The question is knowing which is the best treatment and in what order [to give the targeted therapy and the chemotherapy]. However, even in those groups you get more benefit from the targeted therapy.
Dr. Schilsky: I think you're raising a really important question, Mark, about the need for randomized trials with some of these targeted therapies. For example, do we know whether the prognosis of lung cancer patients with the ALK mutation is different from other patients? We know that the prognosis of patients with the EGFR mutation tends to be better. If you do a nonrandomized trial and select patients based on the biomarker, give them a drug -- and you are selecting a group that has a better prognosis -- then, it looks like they do better. Is that because of the drug or is that because of the patient selection?
Dr. Kris: We started routinely doing ALK testing about a year ago, and none of those patients have died. Even for non-small-cell lung cancer, that is a pretty amazing result, and it speaks to what you are saying.
I think we have to think about this in a different way. We need a hybrid clinical trial approach for these targeted agents. I think that if you have a strictly defined molecular target and a very clear demonstration of benefit, targeted agents could be made available with a proviso that more testing is to be done in larger populations of patients who don't exactly fit the enrollment specifications of the first trial. Ultimately, there can be a comparison of the targeted therapy to available therapies. Doctors can weigh how good that targeted therapy is and begin to ask questions about the right treatment order or whether it should be given in conjunction with chemotherapy. Actually, while we are talking today, there is a trial being presented at the meeting comparing erlotinib alone with erlotinib and chemotherapy, the purpose of which is to figure out what protocol works best in each arm of our mutated patients.
Dr. Marshall: You reminded me of one of my peeves about the meeting itself: you're not going to see his poster, and I'm not going to go see yours. We don't get to learn the lessons from the other diseases.
Dr. Schilsky: That's why we have the highlights-of-the-day session in the morning.
Dr. Marshall: But that's really a CliffsNotes version of a lot of clinical stuff.
Dr. Schilsky: And, remember, there is the virtual meeting. I can put in a plug for that for ASCO.
Dr. Marshall: Those are all good, but I do wonder about shifts in the format over time that might allow a place where those of us who are going to subspecialty meetings can come and actually share what we have instead of just redoing it. I have thought of shifting some of the poster discussions to our subspecialty meetings, bringing only the best of the day to the big meeting, where we can actually spend time thinking about each other's lessons.
Dr. Kris: To your point, what I am talking about tomorrow is the development of a process where we would genotype specimens when patients come at diagnosis. When setting this up, we were asked by our hospital to develop a process that could be used in colorectal cancer and breast cancer. So, it speaks to your point. It's something that was meant to be generalizable. How that gets discussed at ASCO, I'm not sure.
Dr. Schilsky: This is a very important issue.
Dr. Markman: ASCO is a fantastic meeting. It has only gotten bigger. It has gotten even more difficult to get to the various sessions, and there is so much interaction between the various tumor types and molecular markers. Do you have any suggestions how ASCO can try to bring us all together?
Dr. Marshall: I have been stewing on it, and we do have ASCO-sponsored subspecialty meetings where the GI gang, all 2500 of us, huddle up in Orlando for a few days. Frankly, if we had new fresh data, the next tier of abstracts that might have normally come to the big meeting, if we had those to work on and think about and stew on and then bring what we think are the best of those to share with everybody at the big meeting, it would increase the value of both meetings. The big meeting becomes really an amalgamation of the data that are brought forward in the subspecialty meetings, and the subspecialty meetings would give us the intimate one-on-one time to study a lot of that data. It's just a thought, but it would seem to me a nice model to use what we have and grow the structure.
Dr. Kris: I went to a session this morning on PI-3 [phosphatidylinositol 3] kinase, and it was really nicely structured. They had a basic lecture. This is a target that cuts across a lot of different diseases, so they had the breast cancer, melanoma, gynecologic, and lung communities there, and they all came together in that room with phase 1 people. If we find more natural synergies like that, you'll get people from different specialties and different aspects of medical oncology to share a lot of stuff. I thought it was terrific.
Dr. Miller: Since I was outed as the ASCO Program Chair for next year, I'll tell you a couple of the things we have been talking about. We have all been on the program committee where there is an abstract that clearly crosses boundaries, such as a new drug with a phase 1 trial in lung and breast cancer that has PI-3 kinase activating mutations. The breast group wants that, the lung group wants that, and I have seen people nearly come to blows wrestling over the abstract. There is no question you get very different audiences. This probably comes up most with tumor biology, patient and survivor care, or the health services research. If there is something that 2 program committees feel strongly is high profile and that the unique audiences that come to their oral sessions would want to see, there ought to be a way to co-sponsor for those 2 audiences and have it presented twice. It gets different audiences, and that may help some of the cross-pollination you have been talking about.
Dr. Kris: No matter how much you want such a topic discussed on the plenary level, there could be another level for the discussion.
Dr. Miller: And this may be that other level -- for presentations that are thought to be important enough by more than 1 group. We are also working with the education committee to think about the clinical science symposia and how to make those more broadly target- and theme-based, rather than narrowly focused on disease phase. I have also asked the committees to think about other ways to use those discussion slots, because if I have to sit through one more discussion where somebody just shows me the same slides and repeats what was just said, I may start throwing things at the screen. Some topics really need a discussion that is actually a discussion, and some molecular topics may actually do better if presented as a preview by a basic scientist who tells us about the topic.
Dr. Schilsky: This is similar to last year's topic of PARP inhibitors, where we had a presentation of the science behind the topic, followed by the clinical discussion. That's a great idea.
Dr. Miller: It could present the science behind the study; it could cover the state of clinical care and questions in the field. Which of those formats will work best is going to depend on what the abstracts are. I'm asking the committees to consider how we can best fill a specific number of slots so that the program is useful. If they think they've only got enough for 2 hours, then end the session at 2 hours and have people walking out saying that was a great breast cancer session and I've got an hour free to go to a colon cancer session, or wherever. That's better than having presenters feel like they have to fill the time slot.
Dr. Kris: That's a great idea. Also, it would be good to keep the number of discussants open for a number of abstracts. I have another slightly radical thought. People are more a slave to the ranking than to the program, and an abstract that gets a ranking number that is not as programmatic to the day may ruin the flow of the program.
Dr. Markman: What a terrific discussion! I want to thank our wonderful faculty and I want to thank the audience for listening. I hope you enjoyed this discussion. I certainly did. Thank you for your attention.
Introduction
Maurie L. Markman, MD: Hello. I am Maurie Markman, Professor of Medicine and Vice President for Clinical Research at the University of Texas MD Anderson Cancer Center in Houston, Texas. It is my pleasure to welcome you to Medscape Oncology Insights Wrap-up for ASCO [American Society of Clinical Oncology] 2010. I am joined today by Dr. Kathy Miller, Associate Professor of Medicine at the Indiana University School of Medicine in Indianapolis, Indiana; Dr. John Marshall, Professor of Medicine at Georgetown University and Director of the Ruesch Center for the Cure of Gastrointestinal Cancers, Lombardi Comprehensive Cancer Center in Washington, DC; Dr. Mark Kris, Chief of Thoracic Oncology at Memorial Sloan-Kettering Cancer Center in New York City; and our special guest, Dr. Richard Schilsky, Professor of Medicine and Section Chief of Hematology/Oncology at the University of Chicago Medical Center here in Chicago.
Our purpose in this short discussion is to talk about our overall impressions of ASCO 2010. We were here at a very exciting meeting, a very big meeting -- it is only getting bigger -- and we have been hearing a lot of new and innovative ideas.
Maintenance Therapy: Managing Cancer as a Chronic Disease
I will take the privilege of starting the discussion because it's actually now 2 years in a row that the first talk at the plenary session of ASCO dealt with the management of ovarian cancer, my own area of research.
Last year was very interesting and brought a very controversial discussion regarding management of patients who are asymptomatic with a rising CA-125 antigen level. That discussion suggested that perhaps that management strategy shouldn't be considered routine. This year, we took the opposite approach, and we looked at the results of a very exciting trial. The researchers in ovarian cancer have been waiting for more than a decade for a frontline ovarian cancer chemotherapy that was not intraperitoneally based and that would show an impact on progression-free and overall survival.
The Gynecologic Oncology Group [GOG] Trial, known as 218, was reported at the plenary session. The data in this trial were extremely interesting because they showed an improvement in progression-free survival, but there was no evidence that the addition of bevacizumab to carboplatin and paclitaxel improved overall survival. Many questions were raised, including the cost of this treatment, whether one could give bevacizumab only in maintenance rather than with chemotherapy, and whether to give bevacizumab at the time of recurrence. This was an 1800-patient, randomized, double-blind, placebo-controlled study. This was a very interesting study because it demonstrated that even an outstanding trial raises as many questions as it answers. This says a lot about oncology today and about the information presented at the meeting this year.
With that introduction, I want to turn to my guests and ask for impressions of the meeting in your own areas or in general.
John L. Marshall, MD: I'll come back to this maintenance approach because we saw the same sort of thing in the colon cancer world, not with a new medicine, but instead with an old medicine for us, bevacizumab. Our concept has been to keep giving chemotherapy until progression or toxicity, but we know, say with oxaliplatin, that we can't do that, we have to back off because of cumulative nerve toxicity. What do you do when you back off? This study looked at continuous chemotherapy vs chemotherapy after 6 cycles, then maintenance bevacizumab. It turned out that this approach held the curve up just as well as continuing chemotherapy. This concept is emerging, that bevacizumab perhaps is not such a good single agent in advancing adenocarcinomas but might be an agent that suppresses the regrowth of suppressed adenocarcinomas. I think both of these trials are very similar in design and concept.
Dr. Markman: Absolutely, and I would add one thing. Cancer is increasingly a very serious but nevertheless chronic disease process, certainly in the solid tumor area, and it needs to be managed. It's not a matter of just 1 treatment. It may require long-term treatment or multiple treatments over a long period of time. This raises issues of the toxicities of chronic treatments and the costs.
The Real Cost of Trials That Show Only Incremental Benefit
Richard L. Schilsky, MD: Something that we have to be really thoughtful about, though, particularly with these expensive and potentially toxic therapies, is whether we can demonstrate a survival advantage. Let's remember that the patients who are in clinical trials are mainly asymptomatic patients. The eligibility criteria for the clinical trial typically is PS [performance status] 0-1 patients, so they are asymptomatic or minimally symptomatic. To keep them on chronic therapy to slow the rate of tumor progression without ultimately demonstrating an improvement in their overall survival should cause us to question how they are actually benefitted. They were asymptomatic when they began; we put them on long-term treatment, which exposes them to toxicity and certainly increases their cost of care and, ultimately, may not improve their survival. So, where is the real benefit? I think this is a real conundrum that we face both as physicians and as drug developers. There are many questions in the regulatory environment as to whether progression-free survival is a suitable endpoint for drug approval, particularly if there is no advantage in overall survival.
Kathy D. Miller, MD: This is an area I've thought a lot about. It has been a big question in breast cancer with bevacizumab, and 1 key to answering the question that you asked -- we simply don't have. Our quality-of-life data are fairly limited compared with the clinical data. We do have quality-of-life assessments for people who remained in the trial. If they either stopped the treatment in the trial because of progression or toxicity, we don't have quality-of-life data. One of the arguments for using a therapy that substantially improves progression-free survival is if it doesn't add to the toxicity burden in a way that affects the quality of life for a group of patients who have no symptoms or minimal symptoms, even if that therapy is not free of toxicity. If I can prolong that period of time [of progression-free survival], that may be independently beneficial.
Dr. Schilsky: Time to onset of new symptoms may ultimately be that useful clinical -- and maybe even regulatory -- endpoint.
Dr. Miller: Exactly, because inherent in that assumption is that progression of cancer, at least in some patients, will bring with it symptoms of the disease. If I can prolong that [period of progression-free survival], then I prolong time to symptoms. Progression also brings with it treatment with other therapies, some of which may have more side effects on a day-to-day basis. Unless we actually have sequential quality-of-life data in people irrespective of whether they stayed in our studies, we ultimately don't have a way of sorting out whether we do more good than bad.
Dr. Schilsky: There is also the whole issue of patient well being. Our patients live from scan to scan, and every time you can sit down with the patient and say that the tumor has not progressed, you know that they go home feeling okay. You also know that when you tell them the tumor has gotten worse, they don't feel so well. There are important psychological benefits to maintaining tumor stability, but at the end of the day the question is: at what cost to both the patient and to the healthcare system? Does that really become a viable tradeoff?
Dr. Marshall: We quibble about these issues because the magnitude of the benefit is relatively small. If the benefits were of greater magnitude, we wouldn't be struggling with all of these questions. As I sit through my 17th or 18th ASCO meeting, I long for greater magnitude of benefit. I wonder how we are going to shift our world in a way to where we get some bigger and better outcomes.
Dr. Schilsky: To some extent we have to create that world, both by developing better drugs and also by stating up front that we may not be willing to accept these small incremental benefits. That GOG trial, as I recall, had over 1800 women who participated.
It took 1800 women to demonstrate that 3.5- or 4-month improvement in median progression-free survival. That's a lot of patients to invest in a clinical trial to demonstrate a very small increment. Now you could say that it's still something. You could also say that perhaps those 1800 patients would have been better used in smaller clinical trials looking for a 6- to 12-month improvement. If you hit that goal, then you would have something that would really be a step forward.
Mark G. Kris, MD: First off, maintenance is the hottest topic. In the last year, the FDA [US Food and Drug Administration] approved 2 drugs for maintenance. I have a little different spin on maintenance therapy. To me and most of my patients, it's a no-brainer to continue maintenance therapy because you are only bringing this up in the patient for whom you have documented regression, both objectively and subjectively. You have also documented that that person finds that treatment acceptable. To continue therapy in that context is much easier. In the lung cancer arena, time off from cancer means time closer to your normal life, and that, in the absence of cure, is what we can offer our patients. I think that the real issue here is maintenance. The negative of the lung cancer trials -- the SATURN [Sequential Tarceva in Unresectable Non-Small Cell Lung Cancer] trial that led to the approval of erlotinib was just published in Lancet Oncology -- fewer than half of the patients in the induction went on to maintenance therapy.
The Need to Focus Research and Prioritize Trials
I think everybody is saying the same thing. We need to think about what is going on here and keep our eye on the prize. If we do something for that group of people, we're affecting a lot more lives, including those patients who could benefit from maintenance, and that's where we need to shift our focus. That's a message for ASCO meetings in the future. We have to step back and think about what the real issues are and focus research.
There is the need to prioritize. I saw Dr. John Mendelsohn's presentation about the Institute of Medicine (IOM) report, and I saw Dr. James Doroshow's presentation about how we are going to improve clinical trials. One of the big issues there was prioritization, and I think that's something we have to think about. Are we going to do a trial that shows things are equivalent, or are we going to do a trial that shows that we are going to improve cure rates significantly or improve survival significantly? I think all of us would agree that the priority has to be to those trials. Cure is the number 1 goal; improve survival is number 2. Everything other than that, while important, is not as important as cure and survival improvements.
Molecularly Identifying Patients at Highest Risk for Progression
Dr. Schilsky: Mark, the other thing about the maintenance is that we really need a better understanding of the biology of disease stability. For example, if we could identify which patients were likely to have stable disease because they have a biologically more indolent tumor vs which patients were more likely to rapidly progress, that would be an advantage. John and I take care of colon cancer patients, and I'm sure we have both seen many patients to whom we decide to give a chemo holiday; then, months and months go by, and they don't progress off treatment. You know, if you had arbitrarily decided to switch to a maintenance regimen, you'd think the maintenance regimen is doing great things for this patient. However, all you have done is to identify a patient with a biologically less aggressive tumor. What we really need to figure out is which patients have a high probability of progression off treatment and therefore would benefit from maintenance and then focus our efforts in those patients.
Dr. Kris: Getting back to the molecular question, my group and the group at Vanderbilt have looked at the growth characteristics of these resistant tumors. In lung cancer, after resistance to gefitinib or erlotinib develops, a resistance mutation T790M develops. We have been able to look at the growth characteristics of those cells, and those tumors actually grow much more slowly. This has many implications. The ability to detect growth is very difficult and is even more difficult to design in clinical trials. You say what you may be seeing is slow growth and conclude that we're slowing the growth of the tumor, when you're not. The main thing is to focus on the biggest problems that patients face.
Dr. Schilsky: We had a really good example of this yesterday in the colon cancer session. There were very compelling data presented showing that patients with colon cancer at highest risk for rapid progression after first-line treatment are those who have BRAF mutations in their tumors. BRAF looks like a poor prognostic factor in general, but where it seems to have a major effect is after that initial treatment, and then these patients with BRAF-mutated tumors progress very quickly. That would be a population that we would want to focus our efforts on, at least to try to slow down that high risk for progression.
Dr. Kris: That brings us around to the molecular aspects of care, both for selection and for treatment. One of the next plenary abstracts after the ovarian abstract was the one in lung cancer. I think that study generated a lot of controversy. There aren't many phase 1 trials that are on the plenary program. It might be the first phase 1 trial that was done in a molecularly defined subset outside of a hematologic malignancy.
Dr. Marshall: PARP [poly-ADP ribose polymerase] inhibitors did a similar thing after some early findings.
Dr. Schilsky: And we did have the paper last year on the hedgehog inhibitor, which was a phase 1 trial in The New England Journal of Medicine.
Dr. Kris: The designers of the study said, "Here is a targeted therapy that is going to work only where you have the target. We have available technology to find that target, so let's develop the drug and give the drug to those patients with that target." That's what they did, and they had an unbelievable waterfall result; 90% of the patients had some kind of benefit. They had a very high objective response rate, and those responses seem fairly durable in patients who were accustomed to receiving cytotoxic chemotherapy weekly or every 3 weeks. Suddenly, they were taking a tablet twice a day, almost like patients who take tamoxifen in breast cancer.
The Psychological Value of Side Effects
Dr. Miller: I wonder if the investigators saw the same thing, as we got more HER2-targeted therapies and bevacizumab therapies that didn't have chemotherapy toxicity. This came immediately after the big days of bone marrow transplants. I remember enrolling women in their first bevacizumab phase 2 trial who had undergone transplant for metastatic disease. A couple of the responders actually complained about the lack of side effects.
Dr. Kris: They thought that that was a sign the treatment wasn't working?
Dr. Miller: Yes, but put yourself in their position. They had had the disease for years. They had undergone a bone marrow transplant, and they got through a lot of miserable therapy by telling themselves that this is what I need to do to stay alive, to get this tumor under control, and to get better. When treatment was so easy, they essentially said, "I'm not in the fight anymore. It can't be this easy." We were all saying that was the goal, but for patients the shift [from a miserable therapy to an easier form of treatment] took some time to accept.
Dr. Kris: Some patients have a difficult time with that, and they are almost happy to have a side effect because it's a sign that the treatment is aggressive.
Dr. Miller: It's doing something.
Dr. Kris: Also, sometimes it leads them to question us -- if we are giving them a treatment that is not causing side effects, is it really that good? But you're exactly right. Giving them more effective, less toxic treatment is what we want to do.
Adherence to Treatment Is an Issue With Oral Therapy
Dr. Schilsky: However, I think one of the things about which we need to be careful is the issue of adherence to treatment. Kathy, I'm sure you see this in your breast cancer patients, particularly with the development of more and more oral therapies. The issue of adherence to treatment is huge. There are developing data showing about 50% of patients actually stop taking their oral medicine within about 6 weeks of beginning treatment, and that's a terrible adherence rate for a therapy that is meant to be given on a long-term basis. Obviously, some oral therapies are more or less toxic than others, and you'd have to wonder about whether the patient who gets rash and fatigue is more likely to stop sooner than the patient who gets a less uncomfortable side effect. In general, we know from other areas of medicine that people stop taking their antihypertensives, people stop taking their anti-ulcer medicines. However, if we are going to try to convert cancer to a chronic disease, we're going to have to figure out ways of getting patients to stay on their treatments.
Are Phase 3 Trials Needed (or Feasible) for Targeted Therapies?
Dr. Marshall: Do we have to do a phase 3 study in the ALK [anaplastic lymphoma kinase] inhibitor? How can you?
Dr. Kris: There is one underway today.
Dr. Marshall: But how long is it going to take?
Dr. Kris: Your patient who has an EML4-ALK-positive tumor is going to be asked to participate in a phase 3 trial that randomizes them to get a standard chemotherapy, which, by the way, has been shown to improve survival, vs the EML4-ALK inhibitor. At progression, the patient can then cross over to the EML4-ALK inhibitor.
Dr. Marshall: With no overall survival; we'll lose that in the crossover.
Dr. Kris: We'll likely lose that, but that's okay. I think we'll have a very robust examination of benefit. The one nice thing about these therapies is that usually the benefit is obvious no matter what time you get it. So, even in that third- or fourth-line setting, I think you're going to be able to see pretty substantial benefit.
Dr. Miller: Are you worried the trial won't get done?
Dr. Kris: I'm worried the trial won't get done.
Dr. Miller: Why?
Dr. Kris: Show me the trial that tested this cancer for cisplatin. Why didn't we do that trial? We didn't do that trial because it was obvious.
Dr. Schilsky: Well, this trial will get done because there's no other way to get the drug.
Dr. Miller: There's not another way to get the drug. The drug is not likely to get approved in the current climate without the trial.
Dr. Marshall: But did oxaliplatin or cetuximab get approved without phase 3 trials?
Dr. Miller: If I were in charge of regulations, I'm not sure that I would mandate a phase 3 trial to approve this drug, but the FDA probably will.
Dr. Markman: We had that same issue with PARP inhibitors in ovarian cancer. PARP inhibitors occur in 5%-10% of an uncommon disease, and randomized trials are being done [in PARP inhibitors], but people have questioned that requirement.
Dr. Miller: It doesn't strike me that it would be hard to do this study.
Dr. Kris: If I could design the system, I'd probably have a 2-tiered approach. I would get the drug approved in a very strict way and only in the mutated patients. Then, of course, it's a totally different question to ask whether it is better than chemotherapy, because, ultimately, people are going to be asking should I give chemotherapy and, if so, in what order? We have seen that the patients with EGFR [epidermal growth factor receptor] mutations with non-small-cell lung cancer have a 50% response rate to chemotherapy. It's just amazing. Clearly, chemotherapy helps many of those patients. The question is knowing which is the best treatment and in what order [to give the targeted therapy and the chemotherapy]. However, even in those groups you get more benefit from the targeted therapy.
Dr. Schilsky: I think you're raising a really important question, Mark, about the need for randomized trials with some of these targeted therapies. For example, do we know whether the prognosis of lung cancer patients with the ALK mutation is different from other patients? We know that the prognosis of patients with the EGFR mutation tends to be better. If you do a nonrandomized trial and select patients based on the biomarker, give them a drug -- and you are selecting a group that has a better prognosis -- then, it looks like they do better. Is that because of the drug or is that because of the patient selection?
Dr. Kris: We started routinely doing ALK testing about a year ago, and none of those patients have died. Even for non-small-cell lung cancer, that is a pretty amazing result, and it speaks to what you are saying.
I think we have to think about this in a different way. We need a hybrid clinical trial approach for these targeted agents. I think that if you have a strictly defined molecular target and a very clear demonstration of benefit, targeted agents could be made available with a proviso that more testing is to be done in larger populations of patients who don't exactly fit the enrollment specifications of the first trial. Ultimately, there can be a comparison of the targeted therapy to available therapies. Doctors can weigh how good that targeted therapy is and begin to ask questions about the right treatment order or whether it should be given in conjunction with chemotherapy. Actually, while we are talking today, there is a trial being presented at the meeting comparing erlotinib alone with erlotinib and chemotherapy, the purpose of which is to figure out what protocol works best in each arm of our mutated patients.
ASCO 2011: How to Increase the Cross-Pollination of Ideas
Dr. Marshall: You reminded me of one of my peeves about the meeting itself: you're not going to see his poster, and I'm not going to go see yours. We don't get to learn the lessons from the other diseases.
Dr. Schilsky: That's why we have the highlights-of-the-day session in the morning.
Dr. Marshall: But that's really a CliffsNotes version of a lot of clinical stuff.
Dr. Schilsky: And, remember, there is the virtual meeting. I can put in a plug for that for ASCO.
Dr. Marshall: Those are all good, but I do wonder about shifts in the format over time that might allow a place where those of us who are going to subspecialty meetings can come and actually share what we have instead of just redoing it. I have thought of shifting some of the poster discussions to our subspecialty meetings, bringing only the best of the day to the big meeting, where we can actually spend time thinking about each other's lessons.
Dr. Kris: To your point, what I am talking about tomorrow is the development of a process where we would genotype specimens when patients come at diagnosis. When setting this up, we were asked by our hospital to develop a process that could be used in colorectal cancer and breast cancer. So, it speaks to your point. It's something that was meant to be generalizable. How that gets discussed at ASCO, I'm not sure.
Dr. Schilsky: This is a very important issue.
Dr. Markman: ASCO is a fantastic meeting. It has only gotten bigger. It has gotten even more difficult to get to the various sessions, and there is so much interaction between the various tumor types and molecular markers. Do you have any suggestions how ASCO can try to bring us all together?
Dr. Marshall: I have been stewing on it, and we do have ASCO-sponsored subspecialty meetings where the GI gang, all 2500 of us, huddle up in Orlando for a few days. Frankly, if we had new fresh data, the next tier of abstracts that might have normally come to the big meeting, if we had those to work on and think about and stew on and then bring what we think are the best of those to share with everybody at the big meeting, it would increase the value of both meetings. The big meeting becomes really an amalgamation of the data that are brought forward in the subspecialty meetings, and the subspecialty meetings would give us the intimate one-on-one time to study a lot of that data. It's just a thought, but it would seem to me a nice model to use what we have and grow the structure.
Dr. Kris: I went to a session this morning on PI-3 [phosphatidylinositol 3] kinase, and it was really nicely structured. They had a basic lecture. This is a target that cuts across a lot of different diseases, so they had the breast cancer, melanoma, gynecologic, and lung communities there, and they all came together in that room with phase 1 people. If we find more natural synergies like that, you'll get people from different specialties and different aspects of medical oncology to share a lot of stuff. I thought it was terrific.
Dr. Miller: Since I was outed as the ASCO Program Chair for next year, I'll tell you a couple of the things we have been talking about. We have all been on the program committee where there is an abstract that clearly crosses boundaries, such as a new drug with a phase 1 trial in lung and breast cancer that has PI-3 kinase activating mutations. The breast group wants that, the lung group wants that, and I have seen people nearly come to blows wrestling over the abstract. There is no question you get very different audiences. This probably comes up most with tumor biology, patient and survivor care, or the health services research. If there is something that 2 program committees feel strongly is high profile and that the unique audiences that come to their oral sessions would want to see, there ought to be a way to co-sponsor for those 2 audiences and have it presented twice. It gets different audiences, and that may help some of the cross-pollination you have been talking about.
Dr. Kris: No matter how much you want such a topic discussed on the plenary level, there could be another level for the discussion.
Dr. Miller: And this may be that other level -- for presentations that are thought to be important enough by more than 1 group. We are also working with the education committee to think about the clinical science symposia and how to make those more broadly target- and theme-based, rather than narrowly focused on disease phase. I have also asked the committees to think about other ways to use those discussion slots, because if I have to sit through one more discussion where somebody just shows me the same slides and repeats what was just said, I may start throwing things at the screen. Some topics really need a discussion that is actually a discussion, and some molecular topics may actually do better if presented as a preview by a basic scientist who tells us about the topic.
Dr. Schilsky: This is similar to last year's topic of PARP inhibitors, where we had a presentation of the science behind the topic, followed by the clinical discussion. That's a great idea.
Dr. Miller: It could present the science behind the study; it could cover the state of clinical care and questions in the field. Which of those formats will work best is going to depend on what the abstracts are. I'm asking the committees to consider how we can best fill a specific number of slots so that the program is useful. If they think they've only got enough for 2 hours, then end the session at 2 hours and have people walking out saying that was a great breast cancer session and I've got an hour free to go to a colon cancer session, or wherever. That's better than having presenters feel like they have to fill the time slot.
Dr. Kris: That's a great idea. Also, it would be good to keep the number of discussants open for a number of abstracts. I have another slightly radical thought. People are more a slave to the ranking than to the program, and an abstract that gets a ranking number that is not as programmatic to the day may ruin the flow of the program.
Dr. Markman: What a terrific discussion! I want to thank our wonderful faculty and I want to thank the audience for listening. I hope you enjoyed this discussion. I certainly did. Thank you for your attention.
Source...