Relapse After CML Treatment Discontinuation

Transcript:Naval Daver, MD: It’s very important to define what a relapse in chronic-phase CML is. In most cases—especially if the patient is being closely monitored, either by his local community physician or by a leukemia expert and an academic physician—the frequent way we will find relapse is a molecular relapse. So, we will see a patient who has been on a tyrosine kinase inhibitor (TKI), many times for years, who has been on routine surveillance every 6 months or every year who now starts having a rise in his transcript that is confirmed on 2 subsequent readings. That usually will be the initiating factor that makes us shift our treatment to start considering another TKI, and we will also do molecular testing to look for secondary-resistant mutations at that time. In that case, usually when we find that the patient is progressing and he does not have any resistant mutations, if he were on one of the second-generation TKIs, we would choose the other one. For example, if he were on nilotinib, we would consider switching to dasatinib, or bosutinib if he were on dasatinib, or vice versa.

The other form of relapse that we see—and this happens quite frequently, especially in patients who have become noncompliant or poorly compliant—is that they may have been taking a tyrosine kinase inhibitor for a few years and then they had to stop it, they didn’t come to follow-ups, or they were taking it very erratically. They will come in with a blast phase, or an accelerated phase, of chronic myeloid leukemia. In those cases, the outcome is much poorer and the overall survival for blast phase CML or for AML from an underlying CML is less than about 20% to 25% at 1 year. In general, for those patients, we usually will use high-dose chemotherapy or hypomethylator-based combinations with a second-generation tyrosine kinase inhibitor.

We do have a study of decitabine in combination with dasatinib for patients who have a blast phase or an AML arising from a chronic-phase CML, and it has shown encouraging response rates of about 60% to 70%. However, the survival data at this time are not completely matured. So, depending on the type of relapse, we would choose either another second-generation tyrosine kinase inhibitor—in some cases, maybe a third-generation, especially if that patient happens to have an T315I mutation—or we may go to a combination approach with a tyrosine kinase inhibitor, plus either a hypomethylator therapy or chemotherapy if that patient has full blown relapse with an accelerated or blast-phase CML.

In general, we have seen that in patients who were on a first-generation tyrosine kinase inhibitor, let’s say on imatinib, and achieved a major molecular remission—and they lost that major molecular remission at some point, either due to compliance or due to acquisition of resistance—when we challenge those patients with a second-generation tyrosine kinase inhibitor, whether it’s dasatinib or nilotinib or bosutinib, the chances of achieving response are still very high, usually above 70% to 80% to get them into major molecular response.

Now, when we have had the patient already on a second-generation tyrosine kinase inhibitor, they start acquiring resistance, and we’re switching to another second-generation tyrosine kinase inhibitor, the achievement in major molecular response is slightly lower, somewhere in the ranges of 50% to 70%. And once the patients have had 2 second-generation tyrosine kinase inhibitors—so, let’s say they were on dasatinib for a while, then started having increase in their transcript and we switched them to nilotinib—then they had a response maybe for a while, and then they started acquiring resistance and they did not have a T315i. If I put them on ponatinib, for example, at that time, the response rates with ponatinib are about 40%, which is very favorable to what we would have got by switching them to another second-generation TKI, such as bosutinib.

So, the bottom line is that even in patients who have failed 2 or 3 tyrosine kinase inhibitors, we can get them back into major molecular response in anywhere between 35% to 50% or 55%. But in general, those are not very durable and that’s when we start considering other modalities, such as adding on therapies that may enhance the activity of tyrosine kinase inhibitors. And some of these drugs include agents such as interferon alpha, JAK inhibitors, maybe immune checkpoint-based therapies or even consideration of transplant if they have indeed failed 3 or more tyrosine kinase inhibitors.

Kendra Sweet, MD: The DESTINY trial looked at patients who were being treated with either imatinib, nilotinib, or dasatinib and had achieved an MMR and maintained that for a minimum of a year, and then dose-reduced their TKI by 50% and kept them on that 50% dose reduction for a total of 12 months. What they found with this study is that in those patients who had MR4.0, only about 2.5% of patients had what they defined or what they called a “molecular relapse” with the dose reduction. So, very few patients had a molecular relapse with dose reduction of their TKI if they went into it with an MR4.0 or better. In the patients who had a response somewhere between MMR and MR4.0, it was still only 18% of patients who had a molecular relapse, meaning, again, loss of MMR.

What this is telling us is that once patients have a sustained, relatively deep molecular response, it is possible to lower the dose of their TKI. And they found that TKI-related side effects improved significantly with dose reductions. In those patients who had a molecular relapse after the dose reduction, all of those patients regained their prior response when the dose was increased back to their initial starting dose. So, again, there doesn’t seem to be a huge negative side to attempting to lower the dose as we may improve toxicity and patients may maintain their same level of response and continue to do quite well.

I suppose this data could suggest that, in some patients, we are overtreating them by continuing to treat them with the standard doses of these TKIs, given that these data suggest that even in those patients who have a response between MMR and MR4.0, less than 20% of them are going to relapse if you dose-reduce. So, certainly, there are patients who will do just fine with a lower dose and don’t need to be treated at the standard doses for long-term therapy. The hard part is that I don’t know that we know which patients are which when we go into this. If we’re going to start thinking about dose reducing in these patients who have sustained deep molecular responses, we need to be sure that we’re monitoring them very closely so we can catch it early if they do start to relapse.

Javier Pinilla-Ibarz, MD, PhD: The question of if treatment discontinuation can be done in a second attempt has been recently addressed by an abstract reported at the last ASH meeting. In this abstract, these authors tried to see what the rate of discontinuation was after first failure of discontinuation. Very interestingly, those authors showed a significantly high percent of the patients, around 40%, where it was a successful discontinuation. However, there is a very important caveat to really discuss in this trial, and the caveat is the following: initial French trials that start to do discontinuation, they really fixed the threshold to restart the drug once a patient has any detectable disease. To say the patient has complete molecular response, when the patient has any detectable disease, patients will start again on therapy.

In this initial kind of threshold, it was changed in the subsequent trials. And we know these days that even early detection or very small detection of this transcript is not really a sign of relapse if the patient doesn’t really lose the MMR. If you really look at this trial, it’s very interesting that this 40% category of patients may be the patients who really belong, who were placed back on the drug the first time because they had some levels. However, many of those patients, because of the sensitivity of the pCR technique, may be successfully discontinued because they will never really increase the pCR above 0.1%. So, this is an important point that tells us that, again, MMR is an important threshold time point that we should consider at the time to re-introduce a TKI.

Transcript Edited for Clarity