Harry P. Erba, MD: So, we started talking about RT PCR for making the diagnosis, and we started talking about the sensitivity. Jorge, can you tell us a little bit more about this assay and the level of sensitivity that is needed for monitoring patients today?

Jorge E. Cortes, MD: Yes. So, there are a few important things to remember about the PCR. No. 1, both you and Mike have mentioned that this is something that you can do and it correlates very well between blood and bone marrow. That, therefore, becomes very useful, particularly for follow-up because we can do it in the peripheral blood and it works well, so that’s important.

No. 2 is that we need to remember that this is a test that is a ratio. It’s a ratio of the abnormal gene of the BCR-ABL to a control gene. So, part of a difficulty sometimes with a proper sample is that the laboratory has to be able to measure a certain number of copies of the control gene to determine that that sample is adequate. If they cannot detect that number of copies, the sample may not be adequate at all or it may not be able to detect a deep enough response. So, for example, if you only detect, let’s say, 10,000 copies of the control gene, that PCR is only good to 4 logs and it’s not able to go any deeper. Unfortunately, that’s information that’s not always available from the lab and makes the interpretations a little bit difficult. But we need to understand there is variability between the sensitivity. And this is also from sample to sample in the same laboratory from the same patient; one sample may be able to detect different levels. So, that’s something that’s important.

Another important thing is we talked about the levels at baseline. The different control genes may be useful in different ways at base. And in the United States, I don’t know what you use in your institutions, but most laboratories use ABL as a control gene. And that one gives a baseline level that is not very appropriate for calculations of decline and things like that because the ABL control gene also detects the BCR-ABL. So, you can get values that are not accurate. Actually, you can get values that are above 100% because of that misamplification. Now, it’s important to do it, as you mentioned earlier, because you at least know that you can detect it and it gives you that ability versus one of the atypical transcripts, but the value itself doesn’t tell you much.

When you use other control genes, then you were able to have a more accurate representation. Then you can calculate slopes from the beginning to 3 months and things like that that can be predictive of long-term outcome, etc. But it is important that because we are more and more interested in getting to these deeper molecular responses, we make sure that it’s a laboratory that has that good sensitivity, those good-quality controls. And let’s not forget that international standardization is not the only thing that matters for a lab. You can standardize the tests but still have a poor sample quality. And that sometimes is not on the control of the laboratory even. Sometimes it’s degrades in transit and things like that. So, it’s not necessarily a technical issue; it’s just part of the test. So, a lot of the variability sometimes is difficult to assess for the clinician, for any of us. But always keep these in mind.

The other thing that I think is important to remember is that there’s a coefficient of variability, and even in the best labs, there is some variability in the test. Therefore, we—I think I can speak for all of us—never act on 1 sample that has a different value because there is variability. There is a coefficient to variability that you always want to confirm, look at trends, and look at changes rather than just 1 individual sample and react one way or another, understanding that these can just be variations along a certain number.

Michael Mauro, MD: The only other thing I would add to this may be to try to have patients go consistently to 1 laboratory because that then removes the variability between laboratories because they’re all aiming, as Jorge just said, to the international standard. They’re correcting; they’re trying to aim for an international standard value, but there’s variability there, too.

Harry P. Erba, MD: So, let’s come back to that point. Because as you know, as we go through this segment today, we’re going to be talking about some of the new data regarding treatment-free remissions. And with that is a potential goal for patients. I think it becomes even more critical when you’re treating a patient, making sure that you’re getting good-quality data right from the start. And you’re right about the ABL control. A laboratory that you can trust will tell you how many copies of the ABL control were in the sample because that gives them some idea of whether the sample was degraded. If you don’t see enough of the ABL control, then there might be a problem with just the sample integrity.

So, all these things are very important, and because we are busy clinicians and not scientists, we need to put that in the hands of a laboratory that has this test well validated there and to measure down to a level that is clinically important for us. I think that’s another thing we’re going to talk about as we go along talking about initial therapy: What are the clinically important levels of response to ensure long-term survival?

But now we have a new thing that we’re going to be talking about in terms of treatment-free remission, and you clearly need deeper levels to do that. So, you need, as you said, from the start, a laboratory that a patient is going to. And unfortunately for patients, and for clinicians, we don’t always have control over that. But I think we don’t always have control over that. I think we need to start taking control and speaking for our patients. I think our hospital administrators need to understand from us the importance of not switching between laboratories based on what reimbursement or what cost deal they could get this year versus next year.

OK. So, with all of that, how do we know that the outcomes with ABL tyrosine kinase inhibitors actually improve? Which randomized trial showed us that?

Michael Mauro, MD: Well, we’ve had a steady stream. The CML story has changed basically entirely in our careers in the past 10 to 15 years. We just had a publication of long-term data from the IRIS trial showing us TKIs, where the prototype drug imatinib was vastly superior to interferon with the cytarabine if patients could receive it, that the survival of patients was dramatically different; the natural history of CML has changed. As you mentioned, it’s a whole new world, the treatment-free remission, and it’s just built from there. We obviously saw the gaps with imatinib that not all patients responded and they’re resistant. Or in some cases, intolerance was an issue. And fortunately, our industry partners came up with second-generation inhibitors. Nilotinib has shown tremendous activity both in the second-line and now in the first-line setting and has really taken us to the end zone with treatment-free remission dasatinib and so on. So, I think the outlook for a patient with CML now is just dramatically different. They should have a normal life span expected no matter what their age, that things can go well if we handle it properly. So, it’s a great story.

Transcript Edited for Clarity