Looking at the Big Picture in DLBCL 

John Leonard, MD: I now want to move to the big picture of large-cell lymphoma and even the relapse setting but I think a lot of these questions apply to the upfront setting and ask Loretta, your thoughts on kind of- we've got the concept of the chemotherapy free regimen and you were part of a study that looked at a chemotherapy free regimen upfront lead by Jason Westin and Anderson. Do you think we're getting to a point where there's some patients that won't get chemotherapy, whether it's in the upfront setting or could be cured in the relapse setting without- with some combination on novel agents without chemotherapy? What's your big picture take on where that's going?

Loretta Nastoupil, MD: I think that the data is probably the most compelling I've seen thus far to really question, do we have to give chemotherapy to every patient. I think the biggest challenge we've faced so far is how do you define high risk patients? I don't think it's cell of origin or at least not by immunohistochemistry because I think that's probably derailed a lot of the studies in the last 10 years. I think the other issue is that we probably over enrich our studies, particularly the randomized trials, with very good risk patients. So once again, I think we need to revisit how we define high risk or if we could identify biomarkers that may select patients that are going do really well with a non-chemotherapy approach. I'm optimistic that we should do it and we could do it. I think I'm still pessimistic that we'll be able to do it at least if we keep following the recipe thus far, where you give R CHOP+drug X and you make it so hard to enroll the patient that that lead time, again selects out for such good risk patients, we're never going to move the needle.

John Leonard, MD: Greg, what are your thoughts about where things are going and in particular, you've been doing biomarker driven studies at least around cell of origin, but we're now moving beyond that. We have- we haven't talked yet about, but I think it's an important and exciting area the ctDNA and the concept of profiling patients that way perhaps or tracking them and trying to identify risk. Where do you think that's going in the scheme of things?

Greg Nowakowski, MD: I think pendulum has swung a little but against biomarker driven trials, at least on the traditional biopsy and the reason for this is exactly what was mentioned, in the past to get those biomarkers to center. This was delaying initiation of therapy to the patients in the trial, which resulted in patient selection and possible failure of some of those trials. With some of the currently designed front line studies are being just launched, you can see this effort in capturing those patients early on without a real time biomarker qualifying patients for the study. I mentioned earlier this tafasitamab R square CHOP combination is moving forward as a Phase 3 study against R-CHOP. In this study, this was presented in Trials in Progress, there's actually limitation on the- how long after biopsy you can enter the study and limiting that crew only to high-risk patients. There's this effort, conscious effort made to capture those patients and pretreatment is allowed as well. We've learned a lot, how to deal with those trials. In terms of the MRD, I- that's a very hot topic and very exciting. I think changes early on in MRD maybe will let us identify those patients who are more likely to relapse. So that's another area of development where we could actually develop adaptive trials based on MRD. This will actually allow you the time, because everybody would start on the first drug of R-CHOP maybe and then if you're MRD positive or you didn't have certain redaction, you would then go for additional therapy or be randomized in the study. But the tradition development of the biomarkers, going back to initial question has been a little more difficult because of the difficulty capturing those patients. We now have also the smaller classification of different patients with different plasters coming from various group including Blue Stout and others. The challenge with those complex is going to take quite a long time before they could actually produce the results on the trial. I'm not optimistic we'll be able to apply those soon to assign patients a specific therapy in multi-arm study.

John Leonard, MD: I think it's fascinating because we- the pendulum goes back and forth. We've had cell of origin, we had drugs that we thought could target it, we had trials, they really didn't pan out. Although there are still some studies are going on I should say. So, then you have the further sub-classification, the latest group of clusters is more subgroups. Some targets associated with the subgroups. We had from Ari Melnik, my colleague at the ASH plenary he and Menzia presented data looking at BCL10 mutations suggesting that there were subsets of patients within that group that depending on what mutation they had, could be sensitive to a MALT1 inhibitor and not sensitive to a BTK inhibitor. The point being, the details of that aren't important but the point being the slicing and dicing is just going into these different branches and different subgroups that are really going to be hard to even identify in a clinical trial, let alone develop a study where you could randomize to a control group and see if you're making a difference. It's really going to be tough. Then we have these agnostic treatments like CAR T cells that it doesn't matter what you have, it seems to be useful.

Greg Nowakowski, MD: Exactly and each of those biomarkers can affect the patient selection as you pointed out, even the biopsy types. To perform some of those complex analyses you need to have a bigger size of the biopsy, we all know that. It's sometimes very difficult to do just with the fine needle aspirates. One of our fellows actually presented the results at ASCO as well, that even if you look at the outcome of patients per biopsy, the needle biopsy versus larger excisional biopsy. There's a difference in outcome of those patients. Presumptively because some of the sicker patients, we just got this needle biopsy and this is what it was. Now in the patients who could wait we could get more excisional material, which is then sent for sequencing. A lot of this genetic information is actually coming from those ideal patients, if you would, who actually are relatively less rapidly progressing then others.

TRANSCRIPT EDITED FOR CLARITY