John P. Leonard, MD: I want to ask Andy about the issue of MRD [minimal residual disease]. That seems to be a very big deal with mantle cell lymphoma. We’ve had a number of studies correlating MRD to outcome. There was a bit of an update at ASH [American Society of Hematology Annual Meeting] about this issue. How do you see MRD being used in the future, depending on how things play out?

Andrew D. Zelenetz, MD, PhD: The fundamental observation, which has pretty much held true no matter what assay you use, you have a somewhat worse outcome if you’re MRD detectable vs MRD undetectable. But it all depends on what test you use and when you look. We saw the MRD results from the LyMa101 trial, and they used a relatively insensitive assay. It has a sensitivity of 1 of 104, which is actually what we use in CLL [chronic lymphocytic leukemia], and we know that it’s a validated threshold. In the LyMa101 trial—remember, it was chemotherapy, then high-dose therapy, and then maintenance—they looked at before the high-dose therapy and after the high-dose therapy. Before the high-dose therapy, if you were detectable, you had an inferior outcome. It was statistically significant barely, P value of .04. The curves did not dramatically separate, but it’s partly because both curves went down. Clearly, there were a bunch of people who had MRD in the MRD undetectable group because they were using insensitive tests.

Interestingly, we looked at our own trial, which I mentioned before, the R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone]–lenalidomide followed by rituximab–HiDAC cytarabine, followed by rituximab-lenalidomide maintenance. We actually used next-generation sequencing with immuno-sequencing and looked at different threshold levels. When we looked at a threshold of 10-5, it really did not predict for outcome. We used a threshold level of 10-6, and all of a sudden the curves are like you could drive a truck between them. The people who had truly undetectable MRD had a very flat curve, and the people who had anything detectable—10-6 or above—failed.

We need to learn more. There’s a big interaction between the treatment, the testing, and the depth of sensitivity, so we have to figure out how to best apply MRD tools to a particular treatment setting. And if it works in 1 treatment setting, it may not absolutely work in the next. What’s important is you prove that MRD is valuable with treatment X, so then when treatment X becomes standard, you can use MRD to help predict what’s going to happen. But because it works with treatment X doesn’t mean it’s going to work with treatment Y. That’s 1 of the lessons we learned from the data that were presented back to back at ASH.

John P. Leonard, MD: Tycel, you may have referenced this earlier: There’s an ECOG study looking at MRD negativity with 1 in up-front therapy looking at MRD negativity as a primary end point, and the other using MRD negativity as a point of randomization to decide whether to do a transplant. That that is accruing very well. I don’t know if you’ve enrolled patients on that trial, but what’s your impression of that approach.

Tycel Jovelle Phillips, MD: We are participating in the transplant portion of that study, and we have enrolled some patients. Oddly enough, obviously they come back for the maintenance whether they get to transplant or not. We are trying to open the other trial as well, the up-front study. The trial in and of itself has worked out fairly well. It has been almost seamlessly at our institution, so we haven’t run into any issues. Thus far, all the patients we’ve sent that have actually had a clone that has been able to be followed. That portion of the study has worked out before because historically there have been some issues, depending on how old the tissue is, of not being able to find the clone to follow. At least in that sense we haven’t had any troubles.

I’ll be looking forward to seeing what the results are. And if you are MRD undetectable with the Adaptive Biotechnologies assay, whether transplant is really needed in this patient population. If not, obviously that also moves us potentially further from giving cytotoxic agents, especially if you can get to MRD undetectability with some of these other drugs and maybe have more of a finite treatment plan. You would have pretty much a benchmark to transition to maintenance, which I think we all agree is very important in mantle cell lymphoma, irrespective of whatever you give up front.

Transcript Edited for Clarity