John P. Leonard, MD: Nathan, 1 of the things I'm struck by is how all of these things going to improve outcomes for patients. One of the things that is pretty clear to me from our discussion is that we really don't have any molecular predictors of who should get what for any of these therapies. It's really a clinical decision: you have refractory large cell lymphoma, you can get a CAR T [chimeric antigen receptor T cell], you get it. You want an oral therapy? Okay, maybe you'll get selinexor. You want a chronic therapy that's well-tolerated? Okay, maybe you'll get tafasitamab /lenalidomide. Or maybe you're not a stem cell or an aggressive chemo candidate, maybe you'll get polatuzumab. It doesn't address relapse setting. I guess we would argue we have the most data, though limited with CAR T-cells.

Beyond that, we have no molecular predictor, biomarker, profiling, any of these things that tells us what to do, other than what we've talked about today, which is your impression of the drug, your impression of your patient, and what seems to fit what they want to do. Whether or not it's let's go for the long-term remission and cure with 1 set of agents or palliation with another set of agents, how do you think we can move this forward in a reasonable way?

Nathan H. Fowler, MD: This is something we see across the cancer space, especially in lymphoma, and something that we have been talking about for more than 20 years in large cell lymphoma. It started off with ABC [activated B-cell] vs germinal center B-cell subtype, and there are more refined systems to cluster patients into different groups that came out from Margaret Shipp, MD, and Louis Staudt, MD, PhD. I think there was a recent paper in Cancer Cell looking at trying to cluster patients into different groups based upon their genomics.

To date, I have not seen any of these systems used to select treatment. That's the next step—to identify these biologic subsets of large cell lymphoma and treat them differently based upon some hypothesis that that biology can be treated by some targeted agent or some immunotherapy. These things are desperately needed. There are multiple groups that are performing these analyses, dividing groups of patients by their biological characteristics either using some marker or multiplex imaging or genomics. We really need to define the biology by some assay and then using that as a way to put patients into different treatment groups. That’s the holy grail because the one-size-fits-all approach can only move the bar so much in large cell lymphoma. CAR T-cells, which are incredibly exciting, still only have 30% or 35% of patients with long-term benefit. We don't know why that is, so we treat 100% of patients to get that 30% outcome. That's a long answer to your short question but biologic predictors are what we desperately need for this disease.

John P. Leonard, MD: Before we wrap up, I just want to ask, Matt and Kami, your thoughts about cost and how that affects your patients. We've alluded to it a little bit but, Matt, are you having trouble with getting approvals for the drugs you want to treat your patients with? Are they coming back and saying their co-pay is too high? How often does that come up?

Matthew Lunning, DO: It comes up more than it used to. I'm doing many more peer to peer discussions explaining why I'm going down a course that I am. If you think about our highest cost therapeutics right now, it probably is CAR T, unless you are on an oral therapy that is a continuous therapy that's priced out month per month. Then if you're getting to 2 years on some of those drugs, they can be expensive. That's great if you're on an oral therapy for 2 years from a relapse/refractory large cell standpoint.

One of the ways to drive down cost that isn't the cost of the product from a cellular therapy is if they're not in the hospital. If you can decrease the in-hospital cost—I guess in MD Anderson [Cancer Center],no patient was to be sitting in the hospital unnecessarily—I think that that would be one of the avenues to do some cost savings as a total bundled price. There was a lot of cost CAR T-cell data at EHA [European Hematology Association Annual Congress] just released, trying to look at modeling around this.

Different countries have different cost structures; I saw a French and a Canadian. I know that were trying to do our own certain analyses here, but I think it will factor into it. I don't know where things will be priced out in these newer IV [intravenous]and oral drugs, but I can imagine is going to be part of the discussion also.

John P. Leonard, MD:Kami, is costs affecting your patients?

Kami Maddocks, MD: Cost is becoming something that I have to talk about frequently and it can be very frustrating and uncomfortable. We have all these oral therapies, while they are available, can be very expensive. There are patients that may have a co-pay of nothing where there's patients that have a co-pay of a couple thousand dollars a month or a 1-time co-pay of $5000 and that they get it for the rest of the year free. Those are significant costs to patients and it's frustrating to have to tell somebody we have this therapy that can help them but it's just not affordable for them or its it's going to make them go broke. To Matt's point, with cellular therapy, you can get it approved by your insurance company for the therapy, but then you have to still negotiate the cost of the hospitalization and the supportive care, like tocilizumab and those things. It can become prohibitive even if you get coverage for the product. We need to find ways to make these more cost-effective. We have to do better for people in that respect.

Transcript Edited for Clarity