Peer Discussion on Maintenance Therapy in NSCLC

Moderator:Corey Langer, MD

Panel:Mark G. Kris, MDBenjamin P. Levy, MDMark Socinski, MDHeather Wakelee, MD

Langer:One major evolving area of discussion in non-small cell lung cancer is the use of maintenance therapy. I think first we should begin by discussing why it should be used and with which drugs.

Two potential scenarios: first, a patient starts on a standard ECOG 4599 regimen, paclitaxel/carboplatin/bevacizumab. Are there any data to support continuing bevacizumab in such a patient? Should there be switch maintenance in that setting?

An equally popular regimen, perhaps even a more popular regimen in the US these days, is pemetrexed/carboplatin. What are the data for that regimen?

We’ll start off with Mark. Please highlight some of the pertinent studies—the original JMEN, the PARAMOUNT trial, the AVAPERL, and finally the PointBreak trial.

Socinski:Initially, we had ECOG 4599, where we didn’t call it maintenance, we called it treatment until progression, and that was the paradigm that was established. We don’t have a true maintenance trial. We never isolated the event, but that’s what most of us do when we use bevacizumab in the first-line setting.

And then after we established that second-line therapies are effective, we took the 3 second-line approved drugs and used them earlier and we called it maintenance. We limited it to patients who had 4 cycles with either responding or stable disease. We also saw a survival benefit in JMEN and the SATURN trial, leading to FDA approvals for maintenance strategies for pemetrexed and erlotinib. And then, obviously, all of those maintenance trials were done without the presence of bevacizumab. But then the question became, in a bevacizumab-eligible population, what’s the appropriate adjuvant maintenance approach? ATLAS, although a positive trial, did not show an overall survival advantage because the primary endpoint was progression-free survival.

Langer:Unlike SATURN.

Socinski:Unlike SATURN. So the next wave became the use of pemetrexed. Pemetrexed was initially studied as a switch maintenance drug. Then, based upon the PARAMOUNT, a positive trial for the effect of continuation maintenance, both in progression-free survival and overall survival, we had the next generation of bevacizumab-related trials, both PointBreak and AVAPERL, that really asked the question of bevacizumab plus or minus a pemetrexed maintenance therapy. The trials are of different design. PointBreak randomized patients before the initial 4 cycles and then patients received carboplatin/pemetrexed/bevacizumab followed by with bevacizumab alone or bevacizumab plus pemetrexed as maintenance.

Langer:So really 2 separate variables.

Socinski:Two variables. That trial was a completely negative trial. The endpoint was overall survival. It did not show a difference. They did not show a difference in overall response rates. In my opinion, it showed a modest statistically significant benefit. I’m not overwhelmed by the clinical significance of the progression-free survival benefit that was seen.

AVAPERL had a different design. Patients got 4 cycles of cisplatinum/pemetrexed/bevacizumab, and then those patients that didn’t progress and made it through 4 cycles were randomized to bevacizumab versus bevacizumab plus pemetrexed. AVAPERL actually showed a more striking difference in progression-free survival than PointBreak showed, but we don’t have the final word on overall survival on this.

Langer:But it’s not positive. Either way, it was not good.

Socinski:It’s not positive, right.

Langer:One could also argue that it was an underpowered trial.

Socinski:Right . Again, that brings up the ECOG trial, which is the next logical extension in this question of what’s the appropriate maintenance strategy in this population.

Langer:Heather, can you discuss the ongoing ECOG effort?

Wakelee:Sure. The principal investigator for that trial is Suresh Ramalingam. What that study is really trying to do is answer some of these maintenance questions. The backbone of the study is the carboplatin/paclitaxel/bevacizumab ECOG 4599 regimen, and patients who have been on for the 4 cycles without progressing are, at that time, randomized to either stay on bevacizumab alone (a la ECOG 4599), which is the standard arm, or to stop bevacizumab and switch to pemetrexed maintenance.

Langer:So a variation on the JMEN approach.

Wakelee:A variation on the JMEN approach and, really, for the first time, a randomized trial of whether or not to use bevacizumab as maintenance. Now, it’s not bevacizumab or nothing. It’s bevacizumab versus a single chemotherapy without the bevacizumab. So it’s not going to fully answer the question, but there were a lot of concerns that we would never be able to enroll to a trial where the bevacizumab was stopped and nothing else was started, and so that’s why it’s only a 3-arm trial.

Langer:The third arm is the combination.

Wakelee:The third arm is the combination. So you’ve got continuation bevacizumab, the 4599 control; stop bevacizumab and go to pemetrexed; or go on both. That trial is accruing quite well. It’s going to be a little while before we have the data out, but hopefully we’ll be able to finish enrollment in 2014.

Langer:It will be one of the biggest maintenance trials. I think it’s slated to have nearly 1300 patients.

Wakelee:We just increased it a little bit, so it’s very exciting.

Langer:Outside of a clinical trial, Ben, do you use maintenance?

Levy:I do. I think we have to keep in mind that despite the excitement for all these driver mutations, the large majority of our patients don’t fit into the population that can receive targeted therapies and we have to default to chemotherapy, at least for now. I do offer my patients maintenance. My go-to regimen is generally 4 cycles of carboplatin/pemetrexed. If they’ve achieved at least stable disease and they’re tolerating therapy, I will offer them single agent maintenance pemetrexed.

Langer:The PARAMOUNT approach.

Levy:Based on the PARAMOUNT and based on the survival advantage that was demonstrated. Pemetrexed has now had 2 studies, the JMEN and the PARAMOUNT, with switch and continuation maintenance, respectively, that showed both a progression-free survival and an overall survival advantage. Based on those studies, I do offer maintenance pemetrexed.

Langer:But one of the major arguments has been the absence of the mandatory crossover in the control arms of those trials to the study drug. Unlike the Fidias trial, which looked at what some called immediate second-line therapy but was really maintenance therapy with docetaxel, and then true second-line therapy with docetaxel. These trials did not automatically feature true second-line therapy with pemetrexed.

Socinski:I think it should be pointed out that on the Fidias trial, those patients who either got immediate docetaxel or delayed docetaxel had no difference in survival.Langer:But there is an attrition, 40%.

Socinski:Absolutely. So it’s been suggested, it’s not when you get the drug, it’s if you get the drug.

Levy:I think that’s an important point. Unfortunately, we can’t predict when the functional decline and disease progression are going to take place with our patients.

Langer:It’s precipitous.

Levy:Yes. Because of that, I try to offer maintenance therapy up front. The studies have their flaws. We do not have an early versus delayed pemetrexed study, but the studies we have are fairly convincing, at least in my mind. This drug is well tolerated, with a low emetogenic potential.

Langer:Probably one of the best tolerated drugs we have in our cytotoxic armamentarium.

Levy:Absolutely. So, based on these studies, I will offer my patients a maintenance regimen.

Langer:Mark, what are your viewpoints on this: maintenance or no maintenance? And if it is maintenance, do you just continue the original regimen?

Kris:To me, it’s the biggest no-brainer in oncology. To put it in perspective, remember you’re only talking about a small number of people who have had objective benefit from this drug. You have determined that this drug is safe. You have also turned to the patient and said, “Mrs. Jones, has this drug been good for you? Does it inhibit your lifestyle or not?” If the answer is yes, the drug has been good and, no, it hasn’t inhibited my lifestyle, why would you ever stop it? To prove that continuing the drug works, we have PARAMOUNT, AVAPERL, PointBreak, and ECOG all saying continue therapy in that patient. Please, remember, we’re talking about only 15% of patients at this point.

Langer:Probably more.

Kris:Okay, 25%, but we’re not talking about everybody. We’re talking about a very specific patient.

Wakelee:In the trials that have been done, it’s usually about 60% of patients who start with first-line therapy that go on to get the maintenance.

Kris:In SATURN it was under 50%, actually. In ATLAS, it was about 60%.

Langer:And AVAPERL was two-thirds.

Kris:Yes, but again, it’s a small number of patients. All things have to be true, all the stars align. You say it works. You say it’s safe. The patient says, “Doc, this makes sense to me. My life is good because of this regimen.” I would never stop it.

Socinski:I think you’re right. This is a duration of treatment question: how many drugs are you going to continue and for how long. We know that there are some drugs, like pemetrexed, that can be administered for a long time and people tolerate them. There are other drugs, like the platinums, which have pretty tough cumulative toxicity, so it’s difficult to maintain patients on those drugs.

Kris:And you ask that question every visit and if the answer is, “Doc, I can’t use my keyboard,” you’re done.

Wakelee:We’d like to catch it before that though.

Kris:Yes. The art of oncology: talking to the patient.

Langer:Mark, out of curiosity, do you continue the single agent or do you continue the combination?

Kris:I continue whatever I can. As you pointed out, if you give cisplatin, eventually you’re going to get hearing problems or you’re going to get renal problems. If you give a taxane, you’re going to get neuropathy. It’s stopping it at the right time; Heather’s point is absolutely correct.

Langer:And you’ve gone beyond 4 to 6 cycles of the platinum?

Kris:Absolutely. Actually, I turned to your and Stinchcomb’s paper, Mark. You converted me when you put all these trials together about duration of therapy, number of cycles, and in every one of them, people have better survival getting more cycles of therapy.

Langer:Mark, you stop at 4 for the platinum compounds.

Socinski:Right, at least for the platinum compounds or a taxane. We do have a couple of switch drug maintenance approaches. I agree with Ben; I think maintenance is an option. One of the issues with maintenance therapy is that, for the majority of our patients, the median age is 71. Because people have comorbidities and lots of issues, it’s hard for me to think about continuous treatment. We all have had the experience where we say, “Well, we could do maintenance, but we could also take a break.” Maybe patients want a break, some time off therapy. I think, as Mark said, you’ve got to be a doctor; you’ve got to follow them closely. I always say to my patients, “My job is to know when your tumor is getting worse before you know it’s getting worse.” We don’t have a lot of studies in terms of surveillance protocols and the like, but I think time off therapy is very therapeutic for many patients. Therefore, the whole continuous treatment in an incurable disease…let’s face it, in the non-oncogenic driver population, we’re still dealing with a median survival of 10 to 12 months. And it’s tough to keep people on the treatment with the cost of toxicity.

Langer:So, at least in your population, it’s continuing negotiation.

Socinski:Yes, absolutely.

Langer:Every time you get a scan. Have you considered alternative strategies; perhaps 4 cycles and then a break for 6 to 8 weeks?

Socinski:I say to my patients, “You’re kind of on autopilot for 4 cycles unless a disaster happens, and then we’re going to look at your scans and discuss what to do then.” I tell them right up front, sometimes the right answer is to take a break, and sometimes the right answer is to continue.

Langer:Introduce the concept of maintenance at the get-go, not when it comes up.

Socinski:Yes. I agree with that.

Levy:I agree with everything that Mark was saying. I think that the discussion has to start at the beginning of treatment, and then you have to integrate the art of medicine with the science of medicine. Patients routinely come to me and, when I tell them we’re going to do maintenance, they ask, “Is this going to be for the rest of my life?” I think the answer is we don’t know. Probably not; we’ll take a treatment break. I think you really have to tailor your decision on patient preference, motivation, performance status, and if they’re deriving a benefit from the drug. Each patient is different; we shouldn’t be wedded to just one paradigm when it comes to maintenance.

Langer:So individualization is key. What about insurance issues? Have any of you run into scenarios where you’ve gotten pushback from the patient’s insurer saying either no maintenance or, if you’re considering a combination, which granted is not standard, one or the other?

Socinski:We have not, but if we have insurance issues, it’s really up front when deciding what to do. Once you’re on therapy, in my experience, they don’t come back and question your medical judgment in terms of what you’re doing for the 6 cycles.

Langer:That has not been an issue with the ECOG trial.

Wakelee:No, it hasn’t been, and in the continuation, maintenance tends to never be an issue. The switch maintenance can sometimes come up, but I haven’t personally had a problem with that.

Langer:What about cumulative toxicities? I mean, pemetrexed is a generally safe drug, but you do see some unusual problems. Mark, any comments?

Kris:Yes, several unusual toxicities. However, I give a lot of pemetrexed and bevacizumab together, so it’s hard for me to pull out which one caused the problem.

Langer:As maintenance or as second-line?

Kris:As continuing—make the best choice up front and stick with the winner. What you see is, I’ll call it an edema syndrome. And, to be honest, there’s nothing you can do when that syndrome starts but to stop both drugs. I think that’s the answer. People say they notice generally lower extremity edema, and it just progresses. Cutting one drug out, lowering the doses doesn’t seem to help.

Langer:I’ve seen a fair amount of epiphora, which is hyperlacrimation, and also some degree of fatigue, which is not really captured by the data in the PARAMOUNT.

Kris:The eye problems in my patients happen up front. We aggressively try to deal with that with various eye drops. It’s counterintuitive, that the problem causing tearing is that their eyes are too dry. It’s really tough to get that across. And also the plugging of the tear ducts, which can be done.

Socinski:Yes, I’ve seen that.

Kris:A lot of ophthalmologists are involved here. They try to aggressively treat these patients. Again, it’s the whole package. If the patient says it’s worth it, then it’s worth it. If not, you’re done.

Wakelee:I haven’t had too many problems with the hyperlacrimation. I have seen that more with the docetaxel than with pemetrexed or with bevacizumab. Where we get into issues, the counts can sometimes start to have a little bit more of a lag in coming back.

Langer:And you’ve seen some renal toxicity.

Wakelee:Yes, I’ve seen renal toxicity in some of my patients. It’s not a high percentage. For some patients, it’s something that’s happened around cycles 3 and 4; for others, it’s been 4 months or longer. The fatigue can be a bit of a problem, and I also have seen the edema. What I have done with a number of patients, when they get out 4, 5, 6 months, we spread treatment out to a 4-week schedule as opposed to a 3-week schedule. For patients in that setting, who have gotten out that far, there’s something about their disease where it is perhaps slow growing, or at least sensitive to the treatment, and I feel comfortable spreading that interval. Is there data to support that strategy? No, but that’s what I’ve done.

Langer:And as both Marks have pointed out, this is an incurable disease and we really have to balance patients’ quality of life and their life events down the pike and schedule treatment.

Socinski:I actually do 4 to 6 weeks with bevacizumab. It has a very long half-life. If you’re at steady state, does the 6 versus 4 weeks really make a difference? I give patients that choice because they have a lot going on.

Kris:If the patient tells you, “By the way, I have 2 great weeks and a bad week, can we make it another week?”, the answer is yes. Talk to the patient.

Langer:It’s quite clear 10 years ago this was barely on our radar. Now, at least with this panel, it’s part of the standard. I don’t think we were routinely giving it before 2000 and now it’s become part of the standard therapeutic paradigm. As Ben pointed out, this is not the majority of patients. Most patients, unfortunately, do not have a driver mutation or a fusion protein, and there are still some skeptics out there. I suppose, in a perfect world, the 5508 trial would have a control arm with immediate crossover, perhaps to the combination at the time of progression. But that trial would probably be undoable in the US.

Wakelee: