MRD Testing in ALL: Sensitivity and Availability

Transcript:

Aaron C. Logan, MD, PhD: It’s clearly very important to understand what your MRD assessment is actually telling you. There was a very important study that the American College of Pathology did where they surveyed about 120 labs around the country and asked, “What is your actual threshold of detection using multiparameter flow?” And fewer than half actually stated that they can report an MRD result in ALL to the level of 10^-4. So, there are many centers where you cannot get a bona fide MRD result, because they’re not running enough events on the flow cytometer. You need to know that when hematopathologist says, “I don’t detect any disease,” how much did they look? Did you go to 1 in 1,000? Did you go to one 1 in 10,000? Did you push more cells to the cytometer and maybe get 10^-5? You have to know that.

Now, the molecular techniques have the advantage of being more sensitive, and historically ASO-PCR (allele-specific oligonucleotide polymerase chain reaction) was a very useful technology. However, it’s a single patient test because each of those tests has a patient-specific primer and probe that are used during the quantitative PCR. That has been accepted by the Europeans as an acceptable way to measure MRD, and they’ve been very successful at setting up centralized labs that do ASO-PCR.

In the United States, it’s unlikely the FDA would ever approve a single-patient test. Thankfully, the next-generation sequencing approach has entered with multiplex PCR so that every patient can undergo the same assay with the same reagents by using this multiplex PCR and sequencing all the amplimers that come from that amplification, as long as you know what the original clonotype or immunoreceptor sequence that denoted the patient’s cancer was. You can look for that at a deep level, down to 10^-6 in subsequent samples. And so, there have been many studies that have compared the specificity and sensitivity of flow versus ASO-PCR flow versus next-generation sequencing. Above 10^-4, they actually have great correlations.

If you have access to a bona fide validated flow-based MRD assay, that’s great. Use it. But if you don’t, consider using the next-generation approach because it’s now just something you send out to a reference lab. Above 10^-4, you can make the same decisions: Should the patient go to transplant or not; whether you should use flow cytometry as your assay, as long as it’s a validated MRD assay; or if you’re using the next-generation sequencing approach.

What we do need to learn is—and I totally agree with Bijal—we need some studies to tell us, “If I can look deeper, if I can make that dynamic range go 2 logs deeper and tell me the patient has 3 clones out of a million still left in their bone marrow at week 16, does that convey a higher risk of relapse? Is that a patient who I still need to transplant?” We don’t know that right now, but the technology will enable us to do that. Hopefully, we can get those studies done.

Bijal D. Shah, MD: Anecdotally, we do know the context of Philadelphia-positive leukemias. We’ve seen it. If someone fails to develop a complete molecular remission, we know that patient is at high risk of relapse. And so, we’re already getting a sense of what we’re going to see and extrapolating from Philadelphia-positive leukemias. We’re starting to get a sense of what we’re going to see in the Philadelphia-negative with that increased sensitivity. The real question—and I’m going to turn to Ryan in a second here—is, are we ready to embrace it in the absence of clinical trials that support it?

Ryan D. Cassaday, MD: I think in the right context. Aaron’s point is very true, and that’s my take on this. Again, I have to offer this with the caveat that for my clinical flow cytometry, I have access to a reference lab with COG-quality flow cytometry. But I think if you have access to one of the more established technologies that have been used in clinical trials—multiparameter flow that’s reliable or ASO-PCR if it’s BCR-ABL—those would be my usual recommendations. But if you don’t have access to one of those good assays that have been validated and have been studied, then I think next-generation sequencing-based MRD detection would be useful. While there are relatively limited data available, to my knowledge, that tell us how to precisely use next-generation sequencing to make clinical decisions, I would say that MRD testing is definitely better than no MRD testing.

Aaron C. Logan, MD, PhD: I don’t think we need a prospective study to validate a technology above the level of 10^-4. We have lots of retrospective studies. Thousands of samples show great correlation between the technologies above 10^-4. So, I don’t think we need a prospective study to decide whether you can use one or the other to risk stratify for transplant eligibility or things like that. Again, if you want to go deeper than 10^-4, that’s where we probably need some additional work.

But let’s face it: The reason that COG and the Europeans have incorporated MRD testing is they centralized the care of the patients. Pediatric patients always get treated in an academic center. In a European center, or in European countries, they mostly get treated in academic centers and they’re all doing the same MRD through the European MRD and the European flow consortia. In the United States, it’s the Wild, Wild West. Adult patients are largely treated in community centers, and there’s no MRD testing being done because there is no way to do it.

One of the advantages of the next-generation approach is the availability of a central lab that will just take a tube of bone marrow with EDTA (ethylenediaminetetraacetic acid) in it and tell you what the clonotype is. And then subsequent to therapy, you can send another bone marrow aspirate in EDTA and they’ll tell you how much disease is there. The availability of that test really democratizes the access to MRD. Any physician in the country can use that assay. Some MRD assessment is better than no MRD assessment. I think the guidelines are right: Every patient needs MRD assessment, because it is the most important prognosticator for the outcome of that patient. You get 1 chance. If they’re not MRD-negative at some milestone—and we can quibble about end of induction or week 16, I think they’re both useful—you’ve got to get that patient to transplant and probably do something to try and achieve MRD negativity prior to the transplant, so that the outcome of the transplant has a great opportunity to be successful.

Transcript Edited for Clarity