Role of MRD Testing in Acute Lymphoblastic Leukemia

Richard M. Stone, MD: Let’s move on to consolidation. Specifically, we routinely use MRD [minimal residual disease]. Ryan, in general, what technology do you use to measure MRD, and how do you use it?

Ryan D. Cassaday, MD: I have the good fortune of working at the University of Washington Fred Hutchinson Cancer Research Center. Our in-house flow cytometry is 1 of 2 reference labs used by the Children’s Oncology Group, from whom much of our knowledge about risk-adapted therapy for ALL [acute lymphoblastic leukemia]using MRD is based in the United States. I am primarily using that assay. It is available as a reference laboratory for other sites looking for a good reputable place. I will also BCR-ABL, RT-PCR [reverse transcription polymerase chain reaction] for patients who are Ph-positive Philadelphia chromosome–positive as an adjunct. I have not routinely incorporated the next-generation sequencing [NGS] assay, the clonoSEQ assay from Adaptive Biotechnologies, mostly because I’m not quite sure what to do with that result. We saw some nice data from Nick Short, where they have retrospectively looked at how it can risk-stratify patients getting hyper-CVAD [cyclophosphamide, vincristine, doxorubicin, dexamethasone]. But I’m still a little unsure how to use that result to make treatment decisions. I have a hard enough time using MRD flow and hyper-CVAD [cyclophosphamide, vincristine, doxorubicin, dexamethasone] and other agents in adults to know what to do. I’m not sure that NGS adds to it.

Richard M. Stone, MD: When is the key time to look after the initial therapies?

Ryan D. Cassaday, MD: For me, it’s within a month. If I’m using the C10403, it’s at the end of induction. For hyper-CVAD [cyclophosphamide, vincristine, doxorubicin, dexamethasone], I do it after cycle 1A. We have data from our center and also from The University of Texas MD Anderson Cancer Center. Basically, the earlier you’re MRD negative, the better. Then I’ll check it again about 3 months, the end of remission consolidation for 103, and then after cycle 2B of hyper-CVAD [cyclophosphamide, vincristine, doxorubicin, dexamethasone]. I may do it sooner for patients who have a high burden of disease after induction if I’m a little nervous. But those are my standards.

Richard M. Stone, MD: Bijal?

Bijal Shah, MD: We at Moffitt Cancer Center have adapted the Adaptive clonoSEQ assay for most of the patients. Unless they are transferred to us from another center, and they’ve already had some large burden of their therapy there—meaning hyper-CVAD [cyclophosphamide, vincristine, doxorubicin, dexamethasone] or whatever they receive locally. Provided we can get a good diagnostic marrow, and this is the case for most of our patients, we use the clonoSEQ assay.

Richard M. Stone, MD: At the end of induction?

Bijal Shah, MD: This occurs at the end of induction, cycle 1A and 1B. There’s about a 10% MRD conversion. Nobody knows what that means yet, but there are some patients who convert after cytarabine, methotrexate based. This may be true among patients with T-ALL [T-cell ALL], so we tend to use post-1A and post-1B for our MRD monitoring.

If I have a very high-risk patient—a Philadelphia chromosome–like for example, or a TP53, or someone who has a very low level of MRD postconsolidation, I may repeat that at later phases. But it’s more about what is going to be my approach to get them to the transplant.

Richard M. Stone, MD: I want to ask Ryan about the MRD-based flow. Do you have a cutoff for what you consider positive?

Ryan D. Cassaday, MD: The assay that we have at the University of Washington conservatively estimates the level of sensitivity at 0.01%. That is the threshold I will use. I’m of the opinion that any residual disease is positive. Whether it’s at the very threshold of detection or 0.05 or 0.2%, I’m not sure there are substantial differences in those patients.

Richard M. Stone, MD: If you detect it, you’re saying it’s positive.

Ryan D. Cassaday, MD: I wouldn’t go so far as to say I’d only consider it as a categorical dichotomist result. There’s some nuance in the numbers. But I’m not reassured by somebody who’s 3 months into therapy and their level of disease is 0.002%. That’s still a problem.

Richard M. Stone, MD: Bijal, the clonoSEQ. How do you look at that?

Bijal Shah, MD: Same.Anything that is positive is positive. We found that even at levels of 10-6, those patients are at high risk of relapse.

Richard M. Stone, MD: Because it’s obviously more sensitive than the flow-based detection. We won’t worry that it’s too sensitive in any way, shape, or form.

Bijal Shah, MD: The positive predictive value has been exceptional. If it’s positive, it’s real.

Richard M. Stone, MD: We’ll come back to that in a second when we talk about therapy. Jae, what do you do at Memorial Sloan Kettering Cancer Center?

Jae Park, MD: We predominantly use flow cytometry as our MRD assay method, and then we are certified at the lab now. We’ve been fortunate to use them in conjunction. We are also using LymphoTrack, which is different from NGS the Adaptive Biotechnologies is using. But similarly, the sensitivity of 10-6 with an NGS. We struggle between the 10-4 and 10-6 or sometimes below that with a better sensitivity. But I agree with Ryan and Bijal: Any detectable levels certainly makes me nervous. What’s a good therapeutic option is the question that we’re struggling with, but it certainly raises concerns for these patients.

I typically do assess them at the end of the month 1 induction. Month 3 is particularly important because that certainly changes the therapy at that timing with a continuation of their current regimen. The 1 point is that MRD assessments should be done through the consolidation. Just because you’re MRD negative at the end of 3 months doesn’t mean you’re going to stay there. Continued monitoring for that is also important.

Transcript Edited for Clarity