Using MRD to Monitor Therapeutic Response in  ALL 

Eunice Wang, MD: Today, with the increased surveillance by MRD [minimal residual disease] for the treatment of B-ALL [B-cell acute lymphocytic leukemia], clinicians are increasingly viewing MRD status as opposed to frank morphological disease as potentially an indicator of primary refractoriness. For example, a patient who receives up-front induction chemotherapy and consolidation therapy for B-ALL in the adult and pediatric settings would be expected to have morphological complete remission rates of over 90% with standard chemotherapy options.

However, although those patients have morphologically negative disease, the MRD status is perhaps a better indicator of whether that patient truly has achieved a therapy response. In the setting of somebody achieving a morphological response with standard therapy but remaining MRD-positive, this is increasingly being viewed as equivalent to primary refractory frank morphological disease status in terms of its implications for management and overall prognosis.

Bijal Shah, MD, MS: MRD monitoring in acute lymphoblastic leukemia is critical. We now understand that very low levels of disease can culminate in relapse. These levels of disease are well below our standard methods for detection, be they 4-color flow cytometry or morphology. In the United States, we’ve now largely moved toward MRD-based flow. This means a minimum of 6—typically 8—color flow cytometry to allow us to quantitate blast down to the 0.01% threshold.

We’re now seeing new technologies emerge, including the Adaptive clonoSEQ platform that uses next-generation sequencing [NGS], that allows us to get even lower, down to the 0.0001% threshold, or 1 x 10-6. This is extraordinarily powerful. We have data in pediatric ALL presented by Michael Pulsipher, MD, and colleagues showing that detection down to this level can predict who is more likely to relapse in the context of allogeneic stem cell transplant. We’ve known for years that using the quantitative PCR [polymerase chain reaction] and BCR-ABL going down to this level is also predictive of clinical outcome. I have little doubt that we’ll see the same thing emerge with serial studies now of this NGS platform, all the way down to 1 x 10-6.

The nice thing about the NGS platforms is just as we found with the quantitative PCR for BCR-ABL, we can now assess this on peripheral blood. The sensitivity will be lower, but the ease with which we can do this now allows us to more easily monitor for MRD in our B-ALL patients as they’re proceeding through therapy. I hope this will allow for quantitative assessments that capture relapse early.

You heard me say that blinatumomab is now approved for MRD relapse of B-ALL. It’s fascinating. Our pediatricians have beat us on this front. They’ve also found that CAR [chimeric antigen receptor] T-cell immunotherapy is similarly highly effective in the MRD setting. They had patients who had no detectable leukemia presented as part of a multicenter consortium. This is the PRWCC [Pediatric Real World CAR Consortium] for CAR T-cell immunotherapy. Liora Schultz, MD, presented this data at the 2020 ASH [American Society of Hematology] Annual Meeting. They looked at a cohort of around 200 patients. Nearly 50 had no detectable disease by flow cytometry but were simply considered high-risk. These patients benefited tremendously from CAR T-cell immunotherapy. When you look at their outcomes, my recollection is about 70% or 80% of them were still in remission after CAR T-cell immunotherapy with extended follow-up. You might argue, “They were already in remission, they didn't need it.” But as it turns out, they develop B-cell aplasia. When compared with their counterparts who had only 5% or higher blasts, the difference was profound. It was around 35% survival in that subset of patients relative to those who had lower blast burdens.

Putting this together, we now have 2 potent immunotherapy platforms. We have blinatumomab and we’re now seeing with CAR T-cell therapy evolve in the pediatric space, where we can argue that these are highly effective in the low disease burden setting. It emphasizes the need to identify these patients using these sensitive techniques and tools that are now at our disposal. This is unquestionably how we’re going to move the field forward, and how we’re going to ultimately improve that survival benchmark that I referred to earlier.

In terms of how frequently to monitor this, it’s a tough question. I don’t know that anyone has the right answer. Some of this is protocol-dependent. Whether you’re going to do it post-induction or post consolidation—which are more or less the standard—is irrelevant to me. Pick your time point and try to follow the protocol as closely as you can so the data you’re generating equates with the data that were generated on the study. Guidance for monitoring post-consolidation is still largely undefined. There’s value because picking up on the level of disease is something that is now actionable, but trying to identify those optimal time points so you are not running a test unnecessarily is still something we need to learn.

Transcript Edited for Clarity