Part 3:
MRD assessment in clinical trials and clinical practice

Kwee Yong, PhD, FRCP, FRCPath
University College London, UK

I want to end by talking a little bit about showing some data, which you may well have seen, on the early results of the CAR-T studies, and of course the BCMA targeted products are in the lead. This is bb2121, it's fairly standard CAR-T study in which patients are leukapheresed and then conditioned and infused with the CAR T-cells. This is a second generation CAR construct with a 41bb with a CD3 zeta. The key thing to note here is that assessment is done early, so there is an early opportunity to assess MRD.

Here, you can see the swimmer plot with the cell dose here on the Y-axis, and also the response of the patients. You will see in the different colors are that responses, so light blue is complete responses, darker blue are the VGPRs and the PRs and MRs are in light green and dark green. But the stars tell you the MRD status. You can see that many patients, particularly above the cell dose, are achieving MRD negative response very early. Some of them maintain it and do well, but you can see that there are relapses here.

What's also important is the MRD negative response in the bone marrow seems to precede biochemical response, possibly because of the half-life of the paraprotein. But this might tell you that for some types of treatment where the response is achieved very rapidly with tumor reduction, perhaps we actually need a dynamic assessment of MRD, and certainly with a full diseased response. This is published by Noopur Raje in the New England Journal not long ago. Here you see that the progression-free survival of patients who received more than 150 million CAR T-cells here is really only 11.8 months. More importantly for patients achieving MRD negative response, that median progression-free survival is only 17.7 months.

Now, it's unfair to compare this data with the data on the upfront studies that I showed you earlier. First of all, these are multiply relapsed patients. They are much further on in their treatment, and of course this is a bone marrow-based MRD assessment, and we have no information on extramedullary disease. Important considerations when interpreting MRD results, we need to think of the timing of assessment, but also there are times when MRD negativity may not be enough. It might depend on the biology of the disease. There is a question of sensitivity, and I'm not entirely sure that it is treatment agnostic. It might depend on whether your treatment is tumor-directed or immune-directed. If your treatment is directed at the microenvironment, then maybe you have to have some sustained treatment. You need your CAR-T product to persist in order to maintain MRD negative response.

So, when would MRD be used for in the clinical practice? I promised to talk about clinical practice. We might use it to make patient-level decisions. My patient was MRD positive. I could only counsel him that he should have a transplant, but he would not. Three or four years down the line, now he remains in stringent CR, even though he's MRD positive. I have not talked about treatment de-escalation, but I will do in a minute. You may also, outside the context of clinical trials, evaluate prospectively the efficacy of change in algorithm and in service approach.

We've started using tandem ASCT in my institution, mainly because in the UK we still don't have access to maintenance therapy. But we will assess these patients for MRD so that we can get a prospective evaluation of the efficacy of our treatment change. You can evaluate a real-world cohort, and of course stopping maintenance. To finish, I thought I'd show you how in our next frontline UK national study for newly diagnosed patients eligible for ASCT, we are trying to use MRD status to stratify according to intensity of treatment. Patients will receive KCRD inductions, and that is the winning arm for myeloma 11 plus, and standard and high-risk patients will go down different pathway, but essentially, the induction is the same. After transplantation, patients who are MRD positive will be stratified to four different cassettes, one of which will be LEN maintenance, and some of these will have combinations with CD38 antibodies.

For patients who are MRD negative, they will receive maintenance, single-agent maintenance, which will be a CD38 antibody. Then if they maintain MRD negativity after 12 months, they will be randomized to continue or to stop. The high-risk patients will be stratified between escalation cassettes post-transplant, but we will of course track MRD, and we will follow MRD six-monthly for 18 months and then monthly.

So, to conclude, I do think MRD is a new goal of therapy in myeloma, but we need to interpret this in the context of disease, genetics and biology. I do think that we need deep sensitive methods, particularly in high-risk disease. But in patients without high-risk disease, perhaps MRD negative responses may be less important and you need to see that goal in the context of the patient priorities, frailty and performance status. I do think that we will be able to use MRD information to help direct treatment, and I use it in my personal practice for individual patients on a case-by-case basis. But I do think that now we're beginning to use it to stratify patients in clinical trials, we will soon get the evidence base to be able to use it in clinical practice on a much wider level. Thank you.