Willis Navarro, MD: Good morning. Our learning objectives for today are to identify disease- and patient-related factors that predict outcomes of hematopoietic cell transplantation (HCT) and new drugs or combination therapy in older patients with myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). We also aim to help explain recent clinical trial results that have influenced the selection of patients. Objective 3 is to compare the risks and benefits of treatment options and how timing of therapy choices affects outcomes. Lastly, we will describe appropriate candidates for a prospective study on outcomes of hematopoietic cell transplantation in adults who are 65 years and older with MDS and part of the Medicare population.

I have some recent statistics from the National Marrow Donor Program®/Be The Match® registry. These are transplants that have occurred in adults with AML. As you can see, there is a trend upward for AML in general, but probably one of the most striking trends is the number of older adults who are undergoing transplantation. In 2004, it was almost nothing among patients older than age 65 and now, here in 2012, the numbers were about 364 who were transplanted for AML. Those numbers have really just vastly increased over the last 8 years.

When we look at the numbers for MDS, we see a very similar pattern. Very few transplants were conducted for those patients who were older than 64 in 2004, and then more than 200 transplants were performed in 2012 for that same population. Again, it's become clear that more patients who are older than 64 are undergoing transplantation and are now considered transplant eligible.

It is my pleasure to introduce our first speaker, Dr Gail Roboz. She is an associate professor of medicine at Weill Cornell Medical School and New York Presbyterian.

Gail J. Roboz, MD: Thank you and good morning. The question that we struggle to answer for patients with advanced MDS and AML who are over age 60 or 65 is which is right: hospice or stem cell transplant?

Often, when we're asking that question, people seem to suspect we don't know what we're doing -- and patients certainly have difficulty with that type of conversation, where everything from doing nothing to performing a highly intensive procedure might be the right answer. What we're going to do today is explain the options a little bit and then have some, hopefully, big fights or at least pleasant arguments over what to do for patients in a number of scenarios. I have been charged with talking about novel diagnostics and therapeutics for AML in older patients. I will start with these curves, which are shown often.

These are ECOG Frontline Trials, for patients under 60 and over 60.^[1] There are a couple of takeaway messages that are fairly consistent from these survival curves. The first one is that no matter where you look, whether it's in Europe or in the United States and the different cooperative groups, there have been improvements -- but no huge leaps over the last several decades -- in the survival of patients who are under 60. Many of us believe that supportive care primarily, rather than tremendous advances in therapy, have contributed to the successes. Also, no matter how you slice and dice it, overall survival is still in the 40% to 50% rate for this group of younger patients at 5 years, not anywhere near the 80% to 90% that we would like to see.

For patients over age 60, these curves remain superimposable over years and years. Unfortunately, there have been no substantial improvements in overall survival, and median survival is typically less than a year. It is more than a year for patients who achieve complete remission.

The Europeans are not doing much better than we are. These curves are pretty much the same.^[2] This paper was interesting to read because there were, in fact, relatively more patients in what we would consider the much older categories who had received intensive therapy. The take-home message is that older patients can receive intensive therapy. They can go into remission, but the overall survival remains a very elusive end point to improve.

The current treatment paradigm is to first distinguish acute promyelocytic leukemia. That's a happy lecture where patients do very well and get different treatment. The rest of those with AML, however, still undergo a remission induction program, with "7+3" still the board answer 40 years later. At consolidation, things get a little bit more nebulous, with 1 to 4 cycles of chemotherapy. Although consolidation therapy has never been shown to have a survival benefit for older patients, we tend to do a little bit anyway. Autologous or allogeneic stem cell transplant is increasingly being offered to older patients. Finally, there is the big question, which floats in and out, and that is whether or not maintenance therapy has a role in AML. Of course, it has a very important role in ALL [acute lymphoblastic leukemia], but it floats in and out of the literature in AML. I am going to show a few slides that I hope will demonstrate that it should be floating back into favor.

How do we improve current treatment?

First, if you can't beat them, join them. Make 7+3 better; try to figure out how to give it in the optimal manner. Then, look to the rise of the hypomethylators. These drugs are not new, but they are of increasing use in AML. Later, I will show you some of the data and some of the reasons why they might be incorporated into current treatment.

Here's what we want to do: We want to untangle everybody's individual molecular profile and personalize their treatments. We all have our glossy brochures that talk about personalized medicine and that all we have to do is figure out your mutation and then we can know what to give you. But here, the importance is to make it personalized medicine, not anecdotal medicine, and further, to try to figure out how to design useful clinical trials. Such trials would involve enriched populations and would try to treat specific molecular abnormalities with more targeted or more specifically beneficial drugs. The challenge is designing those trials to include enough patients and therefore get meaningful results.

We also have the clone wars. Everybody understands that the initial population of leukemic cells is not all the same. These cells that are mixed in at the initial disease have different biology, as do, presumably, what's left over after chemotherapy. So we're probably totally wrong to give ongoing cycles of similar drugs to patients when we know that the cells that we're treating at different points in their therapy are different. Whether or not you call them leukemia stem cells, they are biologically different.

We also want to know about brand new drugs: Will AML ever have an imatinib? Everybody hopes so.

Finally, we need to address allotransplants so that they are not so miserable and more people are candidates. Depending on the day of the week, most of us feel that allotransplant is probably still the most powerful therapy in AML. The question is: How to make it better and have more people be able to undergo transplantation?

The problem is this pie chart.^[3] AML is molecularly heterogeneous. It's a number of very different diseases, and unfortunately, every single one of these lines may benefit from a different therapy. That is the problem: One size fits none at the moment.

What are the mutations in the hematologic malignancies? I'm not going to go through the alphabet soup, but broadly speaking, there are:

• activating proliferative mutations,
• mutations affecting differentiation,
• antiapoptotic tumor suppressor gene mutations,
• mutations in the epigenetic regulatory genes, and
• mutations in the spliceosomal genes.

They are all important and relevant, and the question is: Should the identification of these mutations drive therapeutic choices?

Molecularly personalized treatment sounds better than it is, but some possibilities are emerging. For example, if you have an MPM1 mutation, it may be that ATRA is of benefit. The FLT3 inhibitors are looking better for FLT3 mutations -- they weren't looking so great in initial therapy, but they're actually looking a bit better now. Core binding factor (CBF) leukemias may have a benefit from dasatinib, with early data suggesting that dasatinib may confer some survival benefit in those patients. What about DNMT3A mutations and high-dose daunorubicin or possibly DNMT3 mutations and the hypomethylators? Finally, does TET2 really predict response to azacitidine?

These are some of the places where there is more than a glimmer of hope at the moment for guiding therapy, with respect to mutations that can more or less routinely be checked.

We go back to 1973 and earlier for cytarabine and daunorubicin. That's still where we are. Although the doses have changed, the concept has not. We have learned that 45 mg/m² of daunorubicin was better than the lower doses, and so we now know that higher is better than lower. For many years, idarubicin looked better than daunorubicin, but there is still some argument about whether that was a dose-effect issue rather than the drug itself being better.

With cytarabine, it hasn't improved overall survival, especially with respect to older patients, to double the dose, prolong the infusion, or use higher doses. These approaches have been looked at over and over again, and they still lead us back to the same model of the current 7+3 paradigm.

Intensification of daunorubicin, specifically in older patients, can be done.^[4] Regarding the initial concern, the patients in the study did not go into heart failure or have major cardiac issues. The problem was the regimen just didn't do that much. There was some benefit in patients up to age 65 to go up to the 90 mg/m² of daunorubicin × 3, but for the rest of the patients, the increase didn't confer a survival benefit. The issue of importance turned out to be not that the regimen was too toxic, but that it didn't work that well in prolonging survival. The message is that older patients can handle the therapy; it just needs to work.

The results from adding gemtuzumab suggest that the phoenix is rising. A French study^[5] at the 2011 ASH meeting showed improvement in event-free survival and overall survival in patients between 50 and 70 years old when gemtuzumab was added at a reasonably low dose. Considering this drug was withdrawn from the market in the United States, these results have generated a lot of discussion about gemtuzumab coming back to life.

We also have a new study -- the AML 16 trial -- from the Journal of Clinical Oncology from the Medical Research Council (MRC).^[6] There was a lot going on, but basically, in randomly assigning patients to daunorubicin/cytarabine or daunorubicin/clofarabine, there was also a plus/minus of gemtuzumab that was included in the randomization. There appeared to be a survival benefit for reasonably low doses of gemtuzumab in that randomization. So the question of whether or not this drug is going to be available to us is a very active one. These are the survival curves from the trial, suggesting again benefits specifically for the gemtuzumab-treated patients.

There are a number of novel anthracycline and ara-C formulations in randomized trials -- the idea being again, if you can't beat them, join them. There is CPX-351, which is a liposomal formulation of cytarabine and daunorubicin. This is a drug that is moving forward on the notion of is it just 7+3 but better: Better in terms of toxicity? Better in terms of efficacy? That remains to be seen.

Elacytarabine is an interesting fatty-acid derivative of cytarabine. The objective is to improve the ability of cytarabine to stay in the system and not get metabolized. This study actually just closed, with the results expected in the first or second quarter of the next year.

Vosaroxin is a novel topoisomerase II (topo II) inhibitor. Again, the questions include: Can we make topo II inhibition in AML better? Can we have less metabolism? Can we have less cardiotoxicity? It all remains to be seen as that trial is ongoing.

We are also still trying to figure out if there is a role for clofarabine. With that in mind, ECOG designed an ongoing, randomized trial of 7+3 vs clofarabine. We'll wait for results of that. I think clofarabine is definitely an active drug in AML, but we're not sure where to use it, in what dose to use it, and what to do with it, so the results of this trial are going to be important.

There will also be the Intergroup study from MD Anderson exploring intermediate doses of idarubicin and cytarabine vs idarubicin and cytarabine plus vorinostat vs 7+3. We will have to wait and see what happens. Investigators at MD Anderson have been very convinced that their regimen of intensified cytarabine for upfront induction therapy is better than 7+3, and we have to wait and see.

I want to talk about hypomethylators now. They are a group that is supposed to offer improved tolerability and activity in patients with unfavorable cytogenetics, which is a group that historically has not benefitted from low-dose cytarabine. They appear to offer clinically meaningful responses even in the absence of complete response in the MDS patients, but that remains to be seen in AML. But the question is: Is it possible that life is better even if you don't get a complete response? A further question is: Because they have favorable toxicity and an interesting noncytotoxic mechanism of action, at least at lower doses, might hypomethylators be able to be combined with other agents in a fruitful way?

Decitabine is a DNA demethylating agent that is fairly well known. It was shown in a monotherapy trial involving elderly AML patients taking it for 5 days to have a complete response rate of about 20% or so, with an overall survival of 7 months.^[7] It was then taken on to a randomized study in which decitabine was randomized vs a low-dose cytarabine or supportive-care type of option.

These data are important.^[8] The initial survival date did not show a survival benefit, but a later analysis at a more mature time point did. The data were convincing to the European authorities, and the drug was approved in a 5-day schedule for older patients with AML, but it did not get approved in the United States. This has generated a lot of discussion. Basically, the activity of the drug was clear in a 5-day regimen, and there appeared to be a significant benefit over supportive care or a low-dose cytarabine regimen. In terms of responses, the question was whether that did or did not translate into a survival benefit. It seemed there was a bit of playing with the statistics at the end to get the overall survival benefit to be significant.

What is even more interesting are the results from a 10-day schedule of decitabine.^[9] These are data from Bill Blum and colleagues showing a 47% complete response rate. That rivals the complete response rates that we see with 7+3 in older patients. This study included older patients with complex karyotype and high white counts. These were not smoldering AML patients -- they were real AML patients who achieved high complete response and low upfront mortality rates after receiving decitabine 10 days at a time.

My group also achieved similar data, about a 40% complete response rate, with 10-day decitabine in patients who were older with poor prognostic features.^[10] These results raise the question of whether decitabine is a better approach than 7+3. Is this a "kinder, gentler" way to get into remission? That remains to be seen.

There's an ongoing ALLIANCE trial right now, 11002, that is accruing very well. I anticipate that it might be fully accrued in the next 6 months or so. It is looking at decitabine in a 10-day schedule vs decitabine combined with bortezomib. Whether or not bortezomib is additive is obviously an important question, but what it will also allow us to do is see what the 10-day decitabine schedule offers in a cooperative group setting, with specific emphasis on these being real-world patients. One of the main complaints about older AML studies is that they didn't include the types of patients clinicians actually see. They didn't include the frail older person who walks into clinic. The ALLIANCE trial is really a study of, "give us your huddled masses, your bad creatinines, your poor performance status patients." All of those patients have been included, and complete response rates will be examined.

Also in this study, we're trying very hard to examine quality of life and other analyses that should make the results a little bit richer in terms of understanding whether the patients actually benefit from the regimen. We are including quality-of-life assessments and comprehensive geriatric assessments -- including questions like: Can you get up? Can you walk around? Can you have a life? This is to try to determine whether the responses achieved and the ways they are achieved are really worth it. We're also doing extensive correlative studies, which I hope will add to the data.

I'm finishing a trial at the moment of decitabine plus plerixafor. The concept is based on the idea that plerixafor could mobilize at least other progenitor cells, and possibly also leukemia stem cell populations, and that might make decitabine work better. This study will be closing soon.

There is also a new DNA methyltransferase (DNMT) inhibitor that is trying to capitalize on decitabine and make it better by increasing protection from deamination. This is the compound that is currently in clinical trials. And again, the idea is to take what is an old drug and make it better.

On this slide, the different colors of the different dots are meant to underscore the fact that the leukemia cells at diagnosis are not all the same.^[11] They may look the same under the microscope, but they're not. What is continually shown is that populations are left over with conventional chemotherapies. This is obvious clinically, because we know that complete remission doesn't lead to cure in all or even most of these patients.

We also know that patients who have minimal residual disease (MRD) do much worse.^[12] One of our major frustrations is that if we could just keep the people who got into remission in remission, we would already have a substantial improvement over where we are right now in AML. So, the question is: How do we standardize monitoring for MRD (the Europeans are ahead of the United States in terms of standardizing), and then, of course, how do we target it? How do we design trials to eradicate those residual populations?

Patients who have high stem cell frequencies at diagnosis relapse more rapidly.^[13] Whether or not you call them a stem cell isn't the point: If you have more of these populations, you do worse.

Our major drug for the last 40 years, the one that we say is the best, doesn't eliminate this population.^[14] So it's clearly a huge problem.

But, what do we do about it? Especially for older patients for whom additional cycles of intensive consolidation may not be feasible, you need new options. A randomized trial of oral azacitidine vs placebo in AML patients post-remission is opening. Azacitidine is used post-transplant for these residual populations. There is also an interesting, post-remission antibody-based trial of an anti-IL-3 agent. The concept here is for an immunotherapy-based therapy to try to eradicate minimal-residual disease.

Is there a magic bullet? We don't have one yet in AML, but all of these different pathways -- FLT3, PI3 kinase, and NFκB -- appear to be important for the residual populations that are left over after the blasts go away initially. The hope is that there might be a way to use 1 or more of these drugs to target residual populations without necessarily taking the patient through transplant or perhaps even after transplant.

There is more of a trend this year toward intensive chemotherapy. That is an interesting pendulum switch because in the last couple of years, everybody has been floating toward the less-intensive regimens for older patients. One thing that we know and that we now know again is that if you have failed hypomethylating treatment, if it worked for you and then it stopped working either for AML or for MDS, that is very bad; those patients have a very dire prognosis. Finally, there was a depressing result from the MRC trial of clofarabine vs low-dose cytarabine.^[15] Clofarabine actually doubled the response rate, but there was no survival benefit overall.

Dr Navarro: Thank you, Dr Roboz.