What I would like to do is talk about new developments in the treatment of chronic myelogenous leukemia (CML), go through some of the follow-up data on first-line imatinib, some of the newer agents, and then talk about a game plan approach to this disease.

This is the Philadelphia chromosome, which is found in all CML cases. And on a molecular level what it does is it places the bcr gene from chromosome 22 with the abl tyrosine kinase gene on chromosome 9, so you get this fusion of the 9 and the smaller 22. And this is the Philadelphia chromosome when it is seen in a cytogenetic preparation.

This is really the business end of the molecule. This is what usually normal abl, which is the tyrosine kinase, shuttles between the nucleus and the cytoplasm. But when it is hooked up with bcr, it actually becomes completely cytoplasmic, and it causes the increase in proliferation and the block of apoptosis that is associated with CML. This is also the marker that we are going to be using in cytogenetics to define complete cytogenetic remission. It is also the molecular marker that we use for polymerase chain reaction (PCR) detection, because you will have bcr-abl in CML cells, but you should never find that translocation in the juxtaposition of the bcr-abl genes in any normal cell, so it makes for a very unique condition where you can have a very sensitive assay.

One of the things we want to lead off with is how you should assess response in this disease, and there are various levels. First is complete hematological response. That is just looking at blood counts. You want to see a normalization of those blood counts, and you need to see a normalization of those blood counts by 3 months of therapy.

Then, we go to the next level with cytogenetic response that obviously has to be done with a bone marrow aspirate. You can go from having no cytogenetic response to a partial response, which means that you still have a fair amount of metaphases left, and down to a complete cytogenetic response, which is what you really want.

Regarding the criteria for response, you usually need to see a major cytogenetic response, which means two-thirds of the Philadelphia chromosomes have to be eliminated by 12 months, and we would like to see a complete cytogenetic response by 12 to 18 months.

In this molecular response, which is even a finer response, that is typically done in patients who are cytogenetically negative, and those are defined a couple of ways in the literature. One is by the major molecular response, and what this means is a 3-log reduction in the mRNA level of bcr-abl from a baseline, so a thousand-fold decrease. And then there is the complete molecular response, which means that the disease is undetectable by reverse transcriptase (RT-) PCR. This is a term that many of us do not like to use, because the undetectable rate varies from lab to lab considerably based on how good their assay is. You can make people have a complete molecular response by simply sending your assay to a bad lab, and that probably should not count.

Here is the update of the International Randomized Study of Interferon and ST1571 (IRIS) trial data, which was the 400 mg up-front study for chronic phase CML taken out now to 5 years. It shows that complete hematologic response happens very early. These are cumulative incidence plots. The other thing that we want to look at most carefully is complete cytogenetic response, so, by bone marrow, no evidence of the Philadelphia chromosome. By 12 months, most people get there, and you get kind of a dangling amount up to here.

This is a curve that is often misinterpreted. These are cumulative incidence plots, which means that these patients achieved a complete cytogenetic remission at some time. And they count on this plot if they had a complete cytogenetic remission but then popped out of it. So this just tells you that at 5 years, 90% of the people at one time had a complete cytogenetic response. It does not say that 90% of people 5 years out are in complete cytogenetic response. Those data are approximately 70%. So if you look at the IRIS data and the patients who got the imatinib, about 70% of them at this point will still be free of the Philadelphia chromosome, so about 30% either do not go into a complete response or go in and relapse.

This is the survival. I think the key survival statistic here is the event-free survival for these chronic phase CML patients who 5 years out have over an 80% event-free survival. And that is pretty spectacular when you compare it with other old-line therapies, such as interferon and the like, and it is perceived as comparable, if not better, than transplantation.

Now you can look at these data and really get a little bit more fine-tuning on prognosis based on a combination of cytogenetics and molecular data. If you look at patients and take a 12-month benchmark and you say you either achieve a complete cytogenetic remission or not, if you do not, then your risk of progression, defined by either flopping back into chronic phase or moving into accelerated phase or blast crisis, is about 30%. These are clearly patients for whom you need to think about something else.

If you are in a complete cytogenetic remission, then you can dichotomize into whether you had a 3-log reduction or greater, or less than 3-log reduction. And you can see here the progression-free survival is about 10%. But if you have a patient at 12 months who is in complete cytogenetic remission and has had a 3-log reduction in their bcr-abl, they are golden. Their risk of progression over the next 4 years is only 3%. So you can really start fine-tuning and telling people what they should expect their prognosis to be.

There is some evidence that even looking earlier makes some sense in determining what the prognosis is. This is from Timothy Hughes' group in Adelaide, and he is looking at an earlier molecular time point, 3 months after initiation of therapy. And what you have done here is you have taken the 3-month mark and you have said, "Well, based on your response at 3 months, what is your likelihood that in the future you will either get a major molecular response or become resistant?"

What it shows is sort of intuitive, what you would expect. For those who have had a greater than 2-log response by 3 months, their chance of going on and even being better is tremendous. On the flip side of the coin, if you had a really lousy response at 3 months, the chance that you will ever get to a major molecular response is pretty poor. So even at 3 months, you can start at least trying to hedge your bets about how a patient is going to do.

This is the acquisition of resistance. It is the flip side of what we just talked about, and it basically says that if you have really had a lousy response, your chance of going on and having resistant disease is very, very high. That is sort of intuitive.

If you look at the accumulation of imatinib resistance to find if it is either initially not responding or becoming resistant after a good response, it goes up with type of disease and phase of disease. So if you look at patients who have early chronic phase disease — that is, they start getting imatinib less than a year from the diagnosis — their chance of failure is pretty low. With later disease — they are in a chronic phase but they have had disease more than a year before they get imatinib — it is higher. If you see patients with accelerated phase or blast crisis, the chances are that they will fail sometime in the future.

The mechanisms of resistance have been talked about quite a bit, and I will not spend a lot of time on them. It is most commonly caused by a point mutation in the abl tyrosine kinase domain, which essentially inhibits the imatinib from binding, so that then the abl is released and can actually cause problems in apoptosis and proliferation. Resistance is rarely caused by bcr-abl overproduction, which is basically amplification of multiple Philadelphia chromosomes. That seems to be more a laboratory phenomenon looking at cell lines in actual patient samples.

Then there are cases where it appears that you have actually blocked bcr-abl activity, but patients still become resistant. These are probably caused by the involvement of some other kinases, and it is probably a harbinger of progressive disease.

All of these different mutations have some different responses to drugs. These are patients where cells were taken and given imatinib, and you see whether or not they were viable. This is wild-type, meaning you give imatinib in increasing concentration, and this basically falls off. And these are the various point mutations. This is the evil T315I mutation, which is completely resistant to imatinib, dasatinib, and nilotinib, and you can see cells just grow through that.

But different mutations have different responses, and as we get smarter and smarter, we will actually be able to look at resistant cases, define point mutations, and probably give people the right drug given the type of point mutation they have.

Once we have intolerance and resistance, what are the options? Well, we will talk about second-line tyrosine kinase inhibitors. Dasatinib has been FDA-approved, and nilotinib is in the FDA pipeline, and I put parentheses there because I suspect that will be approved shortly.

There is transplantation, and then there are investigational drugs, which, right now, are theoretically nilotinib and other tyrosine kinase inhibitors, and the Aurora kinase inhibitors.

This is a plot of those similar types of curves with point mutations, looking at imatinib, dasatinib, and nilotinib. And you can see whereas most of these point mutations are resistant to imatinib, most of them fall by the wayside with either dasatinib or nilotinib. Again, the bad ones are the T315I and the F317L where these drugs do not affect those very much.

If you look at the comparison between dasatinib and nilotinib, it really depends on what your philosophy of resistance is, because dasatinib blocks bcr-abl kinases but it also is a fairly broad tyrosine kinase and blocks other kinases, especially in the Src family, which may or may not be responsible for advancing disease, especially a lymphoid blast crisis.

The other side of the angle is that nilotinib, which has actually been specifically designed to inhibit bcr-abl, has a much narrower spectrum of tyrosine kinases, which may be good if you really think that bcr-abl is driving it. If you think that some of these other enzymes are driving progression, then it may not be so good, but these are the kind of different philosophical approaches to those two drugs.

Let's talk about dasatinib in chronic phase patients who have failed imatinib. There are three studies I want to go over very briefly: the evidence that 70 mg bid works, dasatinib versus high-dose imatinib, and a study looking at the optimal dosing of dasatinib.

This is looking at patients who were in chronic phase and either failed imatinib or they were intolerant to imatinib. And just look at the red here. This looks at complete cytogenetic remissions in these patients. This is the total, about 50% of patients got complete cytogenetic remissions on dasatinib. If you break it up depending on why people got in the trial, as you might expect, resistant cases have a little bit less, and intolerant patients have a little bit more, so about 75% of patients who were put on this study because they were intolerant to imatinib achieved a complete cytogenetic response. That is the gold standard we are looking at here.

If you look at the different types of mutations, different types of mutations had different responses. These are complete cytogenetic responses, and you can see some of these mutations, there are a fair number of those. If you go to the other side of the spectrum, the T315I patients, no one responded. And in-between is in-between.

Cytopenias are common, as they are with imatinib and the other tyrosine kinase inhibitors. These are probably a fact since you get clonal changes as you give the drug. The leukemia clone goes away, and there is usually time that you have to buy until the normal hematopoiesis takes on. If you look at Grade 3 to 4 toxicities, this is fairly common in this drug, and it bears watching. You cannot give these drugs and just not watch white counts and the like. This is an important complication that you at least have to follow.

Types of toxicities, all grades, and Grade 3 and 4 are across the spectrum, as with most of the tyrosine kinases, with fatigue, dyspnea, rash; but none of these blow out as really, really striking as far as the Grade 3 and 4 are concerned. These are all higher than with imatinib 400 mg, and, again, they are not anything really to write home about.

If you look at progression-free survival in this group, it is quite good. These are patients who are either intolerant or resistant. The total is shown in white, and you can see for the total at almost 2 years, there is about a 90% progression-free survival, which is pretty darn encouraging. And for the intolerant patients, it is really good. So these people are on drug, they get a cytogenetic response, and they maintain it quite well. I think these are pretty encouraging data with this drug.

We know 70 mg bid works. The question is, does it work better than imatinib? If you have someone who has become resistant to imatinib, can you just jack up the dose from 400 mg to 800 mg, and is that different than putting them on a new drug?

There was a randomized trial with a crossover arm looking at patients who are resistant. Obviously, none of the intolerant patients.

This is a busy slide, so I just want to direct you to one part here that I have highlighted in yellow. If you look at the initial treatment, either with dasatinib or high-dose imatinib, and look at complete cytogenetic response, 40% of them in the dasatinib arm versus 16% in the high-dose imatinib arm. And if you look at major molecular response, 16% versus 4%. I think that shows a fair amount of increasing activity in this setting for dasatinib.

If you look at time to treatment failure or progression-free survival, both show the same thing. The dasatinib arm wins over the high-dose imatinib arm.

What is the best dose? Well, 70 mg bid is what the marketed dose is, but there is some question about whether that should be the best dose. And so patients were randomized to either just getting 100 mg or 140 mg and having two different dose schedules — 100 mg once a day, 50 mg bid, versus 140 mg once a day, 70 mg bid.

If you look at progression-free survival, they all fall together pretty darn closely here. There is some evidence that actually 100 mg once a day may in fact be better than 70 mg bid. There is obviously not a huge difference in efficacy.

The other question is, is there a change in toxicity? Here is the progression-free survival again, 100 mg once a day, 70 mg bid. But if you look at some of these criteria like dose interruptions, there are fewer dose interruptions on the 100 mg. There is less congestive heart failure (CHF). There are fewer pleural effusions. So I think all of this points in the direction that, in this case, a smaller dose delivered once a day may be as good or may be better.

This is what we can say with dasatinib in this setting. For patients who fail chronic phase, you will have a complete cytogenetic response in at least 50%. It does not work in T315Is. Dasatinib is probably better than high-dose imatinib, and probably the dose that is going to stick is 100 mg once a day.

Let's talk about nilotinib. This is a drug that has more activity than imatinib and is fairly specific just for abl.

A similar study was done where you take patients who are intolerant or resistant to imatinib and put them on nilotinib. Here we have the results. About 40% of patients will get a complete cytogenetic remission. Here the resistant and intolerant patients are fairly similar, but this is the number I think that you should stick in your mind, about 40% of the patients get a complete cytogenetic remission.

The usual suspects as far as tyrosine kinase inhibitor toxicities. Nothing really sticks out as being especially unusual.

Grade 3 and 4 hematologic toxicity is probably higher than with imatinib but along the same range as dasatinib.

Here is the time to progression for these patients. So you start here, you get a good response and you either come out of it or stay in it. By about a year and a half, about 70% of patients have stayed without progression. So you have lost about 30% here, so something to think about. These drugs are probably not going to be curative in these kinds of patients. It certainly buys you some time.

In the chronic phase setting, about 40% of patients achieve complete cytogenetic response. It is well tolerated. It is probably an efficacious drug, but it is probably not curative.

What do we do with advanced phase disease? This is the problem that all of us worry about because, right now most patients in chronic phase will do well, but if you do not monitor them carefully, there is an intrinsic relapse rate. All drugs — imatinib, nilotinib, dasatinib — and transplantation work a lot better in chronic phase than accelerated phase or blast crisis. There is biological evidence to suggest now that this really is not a 3-step disease, that accelerated phase and blast crisis are really the same thing. Once you have gone to accelerated phase, the horse is out of the barn, so it behooves you to really watch these patients closely.

But if one slips through and you get accelerated phase or blast crisis, what do you do? There are a variety of agents. There is dasatinib or nilotinib. You can go to an investigational drug, or you can try transplant.

Let's look at dasatinib in advanced phase CML and Philadelphia-positive acute lymphoblastic leukemia (ALL). If you look at all of these cases together, you get complete cytogenetic remission in about 32% of cases. Myeloid blast crisis in 26%. A little bit better, 46%, in lymphoid blast crisis. Pretty good initial response, 50% complete cytogenetic responses in Philadelphia-positive ALL.

The problem is they do not last. This is progression-free survival for accelerated phase disease. Not bad. Blast crisis, lymphoid, myeloid, and Philadelphia-positive, terrible. So you can get people into complete cytogenetic remissions with these drugs, with this advanced phase, but they will not stay, so you cannot count on these things curing anyone.

Nilotinib is basically the same story. If you look at patients in accelerated phase and look at major cytogenetic response, it is about 36% with complete cytogenetic response at 22%.

Here is the total time to progression for the accelerated phase patients. The disquieting thing is that there is really no plateau on all of these curves, that you are continually drifting down, so the clock is really ticking in these patients.

Since we have mechanisms to monitor these patients closely, it behooves us to do it. Here are the current National Comprehensive Cancer Network (NCCN) guidelines for patients who are on imatinib. At diagnosis, you get baseline cytogenetics and a PCR. You can use fluorescence in situ hybridization (FISH) if for some reason you cannot pull a marrow, and that will confirm the diagnosis. It does not tell you anything about additional cytogenetic abnormalities, though, so that is a problem if patients are in accelerated phase. It certainly is not sensitive enough for molecular monitoring.

When the patient is responding, we do bone marrow cytogenetics at 6 and 12 months and a peripheral blood for PCR bcr-abl every 3 months. When a patient achieves a complete cytogenetic remission, they get a PCR every 3 months. You can do marrow cytogenetics every 12 to 18 months. We can talk about the rationale for that later.

When bcr-abl transcripts are rising, you can increase the bcr-abl testing to monthly and look for abl mutations.

That brings us to who should get abl mutation testing: patients in chronic phase who have an inadequate response — defined as no hematological response at 3 months, minor cytogenetic response at 6 months, or no major at 12 months — or any time when patients lose response. For accelerated phase and blast crisis, the incidence is high enough at diagnosis that it is a good thing to do from the get-go.

Here are the treatment options. For chronic phase, imatinib 400 mg, although some people like to go higher; transplant; and the best option is a clinical trial. The Intergroup trial right now is 400 mg versus 800 mg imatinib versus dasatinib. For resistant disease, more imatinib, but I think people are being less encouraged to do that now; dasatinib, or nilotinib when it gets approved; transplant; or trial drugs. There are at least two tyrosine kinases out there in phase I trials right now. This is an Aurora kinase inhibitor, which has the added benefit of actually being active on T315I mutations. And for accelerated phase and blast crisis, again, tyrosine kinases, transplant, or trial.

What about transplant? Well, 2 days ago at Fred Hutchinson Cancer Research Center (FHCRC), we had a celebration because it was the 50th anniversary of the first documentation of human transplantation reported by E. Donnall Thomas in the New England Journal of Medicine. You have probably never read this paper. It is actually amazing what they did at that time and how transplant went from being something that was just completely radical —they could not even do human leukocyte antigen (HLA) testing then — to a place now where it is a procedure that has literally cured tens of thousands of people.

One of my favorite quotes is this. This is Don's West Texas humor in the discussion: "In an atomic age with reactor accidents, not to mention the stupidities with bombs, someone is going to get more radiation than is good for him." And it is interesting because the pitch in this whole paper was not for using this procedure as a cure for leukemia but as a cure to salvage people who had life-threatening radiation exposure, and how it morphed into transplantation is phenomenal.

We have to remember that transplantation is a good option, because it has been shown to be curative, and the results set a pretty high bar. These are looking at unrelated donors and related donors from FHCRC from 1992 on. For related donors, disease-free survival is 80% at 8 years. It is more like 10 years now. This goes up to age 65. The unrelated donor only goes up to age 50, so there is an age bias in these comparisons. But in chronic phase patients less than 50, it does very well. You can see the problem is that there is a big drop from chronic phase to accelerated phase, and a bigger drop than we see with blast crisis.

This just emphasizes that monitoring is to be done carefully because if a patient does have a donor and does need a transplant, you want to do it in chronic phase rather than in accelerated phase or blast crisis.

One of the big questions has been, if we give imatinib for long periods of time, does that screw up transplant results, because for every other drug we ever looked at, it does, right? If you give hydroxyurea, if you give busulfan, if you give interferon, if those patients get those drugs for ages, transplant results suffer because of toxicity and because of relapse. So is that effect true in transplantation? It probably is not. There was a study by Michael Deininger that was published looking at toxicity and relapse, and it is a difficult paper to interpret because there are a lot of Philadelphia-positives in there. There are various types of CML patients in there, but it showed no major effect of previous imatinib on transplant results.

There was a bigger study that came out in Blood last year by Vivian Oehler that looked at approximately 200 patients and matched them against age controls from FHCRC, City of Hope, and Stanford Comprehensive Cancer Center that had not been exposed to imatinib.

Here is the "no imatinib" chronic phase arm. These are the patients who had imatinib prior to transplant. This is the accelerated phase, "no imatinib," "yes imatinib." Here is blast crisis. The interesting thing here is that blast crisis patients in the age before imatinib did badly, but blast crisis patients who were given imatinib for some period of time to debulk them actually did surprisingly well. So I think that it is very optimistic that we might be able to salvage some people who pop into blast crisis with these drugs.

As for future questions, we need to discover predictors of response. When people walk in the door, we can tell if they are going to be responding to the drugs or not, so we can guide them more carefully. We have to determine whether or not the more potent kinase inhibitors should be used as first-line therapy rather than salvage. Does it make more sense to bang on this disease harder and accept the toxicity? We need to evaluate if combining agents is safe and efficacious, develop strategies to eradicate residual disease, and develop strategies to override resistance mediated by the T315I mutation.