MRD Could Take Center Stage in Pediatric and Adult ALL Paradigms

Jerald P. Radich, MD

There are several ways to assess minimal residual disease (MRD) in patients with acute lymphoblastic leukemia (ALL), explained Jerald P. Radich, MD. And although there is some debate as to whether it should be termed “minimal” or “measurable,” it may become standard practice to test for MRD as a surrogate endpoint in clinical trials.

Polymerase chain reaction (PCR)-based assays and flow cytometry are two of the most commonly used tests across Europe and the United States, respectively. Though, the potential for standardization with next-generation sequencing (NGS) is unmatched and as such may become the preferred option in the future.

“The main thing that is needed is FDA regulatory buyoff,” said Radich. “Based on the data we submitted to the FDA, they thought MRD would be a potential surrogate for an endpoint and analysis, meaning you could use [MRD] at a very early timepoint—say 6 months—to declare which drug might be better and carried forward to a phase III trial.”

If MRD is accepted as a surrogate endpoint, it could enhance the efficiency of clinical trial designs and potentially bring more drugs to market.

In an interview during the 2019 OncLive^® State of the Science Summit™ on Hematologic Malignancies, Radich, member, Clinical Research Division, Fred Hutchinson Cancer Research Center, and professor, University of Washington School of Medicine, discussed the methods that are used to measure MRD and the work being done to standardize its use in ALL.

OncLive^®: Could you discuss the importance of MRD in ALL? What are the methods that are currently used to detect MRD?

Radich: Right now, there are 3 methods used to detect MRD. One looks at the immunoglobulin or the T-cell receptor changes that happen in either B or T cells. There, a patient-specific, PCR-based assay is done based on whatever specific VDJ clone rearrangement is in that leukemia. The advantage of that is you have a patient-specific marker which tends to be fairly sensitive. There, you can detect 1 leukemia cell in down to 10,000, sometimes even down to 1 in 1 million. The con [to this approach is that] it takes several days to get [the results] back, it's very expensive, and it’s very labor intensive. Also, the sensitivity varies from patient to patient. Although we say the sensitivity is generally 1 leukemic cell in 10 to 1000, we might get lucky in 1 patient where it’s 1 in 1 million because of the VDJ particulars of that rearrangement. The Europeans mostly use this approach as their way to measure MRD.

In the United States, we mostly use flow cytometry, which looks at the constellation of antigens and their distribution on leukemic cells. Here, we’re trying to look at patterns of those antigens that are not normal. The advantage of that [approach] is that it's very fast. You can get [results] turned around in 1 day. There is a fair amount of art to it. However, the sensitivity and performance vary dramatically from lab to lab, so it's very hard to get standardized.

I'm sure NGS is going to take over soon. NGS is the approach I spoke of with the VDJ rearrangements. Instead of patient-specific amplification strategies, you're basically using hundreds of primers, taking that patient's VDJ and doing thousands of PCR reactions in one solid tube. Then, we’re looking at what the gene sequence is; that's the fingerprint of the leukemia in that particular patient. That [approach] has the advantage of being standardized. In fact, it has now been approved by the FDA. The downside with NGS is that it takes longer to do than the other assays because it's approved for sequencing and is, therefore, more expensive.

Are the data with MRD the same in adult and pediatric patients?

The pediatric groups have been much more organized [with their data collection] in the United States. Basically, all pediatric cases are put on clinical trials; that’s fairly rare in adults. The amount of rich data that we have looking at any biological parameters in pediatric patients is much more than we have in adults. We have fairly robust data because we've been doing it very systematically in kids for a long time. In this setting, MRD is significantly associated with risk of relapse. With adults, our data is much spottier as a result of different methods, different groups, and much smaller studies.

We thought intuitively that the data wasn't going to be that strong or that consistent. It turns out that when we did this large meta-analysis for the FDA and the National Cancer Institute, we were able to look at enough data from different studies in adults and pediatric patients and put them all together to show that the effect of being MRD-positive is essentially identical, irrespective of age. It just happens that kids do better than adults, so those curves are shifted. However, the actual relative difference between patients who have MRD and those who don't are identical between adults and kids.

Is there a difference between “minimal” residual disease and “measurable” residual disease?

MRD is a term that came about when I started doing this about 30 years ago. It was a better way of looking at disease than standard morphology. As we have found out, the implications of “minimal” are many. Minimal suggests that patients don't have much disease, when, in fact, at diagnosis they probably have a thousand billion leukemia cells. Even [with the most sensitive assays] that can detect 1 in 1 million cells, [which denote] MRD negativity, that

still means you might have 1 million cells leftover that you can't detect. Thus, minimal is not minimal. [The term] suggests that the clinical impact of residual disease is minimal when it's not; we know it's highly associated with risk of subsequent relapse. Most of us are now referring to it as measurable residual disease, because it really reflects that this is a better measurement of disease burden in a patient.

What still needs to be addressed in this space?

The FDA has been doing some work on their own and they're coming to a conclusion where [they might accept MRD as a surrogate endpoint in clinical trials]. If that's so, it will change everything quickly because [researchers] will start looking at early phase II trials and ways to use MRD as the endpoint of the trial.

Once that happens, the amount of investment in technology from industry biotech companies to build a better “mousetrap” [will skyrocket]. They would save [a significant amount of money] by making clinical trials [more efficient] and therein send more drugs through the pipeline. Once that happens, it’s going to take flight really fast. That will drag all the diseases that are behind in technology to measure MRD upfront. This is going to be the “strong horse that pulls the big cart.”