Panelists: Charu Aggarwal, MD, MPH, Hospital of the University of Pennsylvania; Hossein Borghaei, DO, Fox Chase Cancer Center; H. Jack West, MD, Swedish Cancer Institute
H. Jack West, MD: Let’s turn to tumor mutational burden [TMB]. On our tweet chat someone asked about that, and I think lots of people are wondering, should they be ordering it? And if so, how should they be using it in practice? Charu, I’ll start with you here.
Charu Aggarwal, MD, MPH: So I don’t routinely get TMB at this point in time. We do our gene sequencing in house, and TMB is not something that we are able to clinically make a decision on yet. I think it’s cumbersome. You have to know which test to order, how to interpret the test that you order, and how to apply it. I don’t think there are any clear guidelines. I think regarding PD-L1, we still can say, yes, there were trials driven based on PD-L1 [programmed death-ligand 1] expression. I just don’t feel the same way about TMB. Now, I do think that it’s a relevant biomarker for future clinical trials and currently what we are doing, in terms of patient selection, in conjunction with other biomarkers as part of a clinical trial. So I think TMB is very relevant. I just don’t clinically use it routinely.
H. Jack West, MD: Now Hoss, I know you have a lot more experience and spend a lot of time thinking about this. What do you do, and what would you recommend for the oncologist in practice?
Hossein Borghaei, DO: Right. I mean, the second question is easier to answer. As long as we don’t have a clear approval associating a specific treatment with survival advantage to TMB, I don’t think you absolutely need to get TMB. So I think in the absence of a clear clinical indication, there’s no need to do it. The new in-house panel promises to give us TMB because we’re switching to a bigger sample or a bigger platform. So maybe we get that. But for now, if you send your molecular testing to companies that are out there, you get TMB basically for free as part of the whole sequencing that you get back in your results. So if you use 1 of these other entities, you get your TMB data back and there is no extra charge for it, right? So it can be part of it. And right now we’re getting that.
So as part of my EGFR [epidermal growth factor receptor] and everything else, I do get the TMB results back. But do I use it clinically? As I said, I’ve had a couple of examples of patients with negative PD-L1 and low TMB in which I’ve had to have a discussion with the patient; or in the second-line or third-line setting, if I’m looking for a clinical trial for somebody who’s been through I-O [immuno-oncology], I do look at the TMB to say, do I think this patient has a higher likelihood of responding to another immunotherapy or something to that effect? So I don’t use it for everyday clinical practice, and I’m not saying that I use it to make a decision, whether I should give chemotherapy and I-O or I-O to somebody. But I do look at that when I’m trying to decide subsequent lines of therapy and what I might be offering to the patients. I think TMB does have a lot of issues to overcome, kind of like where we were with the PD-L1 testing, and everybody had their own antibody and platform and we needed a blueprint to sort of tell us what these tests are. I think something like that is in progress for TMB. We know that work is being done. So I think before long, we’ll have some way of sort of standardizing the TMB testing, which is absolutely needed.
But as long as there is just PFS [progression-free survival] advantage or maybe a little bit of a response advantage and not OS [overall survival], I will quote somebody who we like, who said, “Don’t bring PFS to an OS fight,” if you remember that.
H. Jack West, MD: That’s mine.
Hossein Borghaei, DO: I know it’s yours. I think it’s hard to argue to use the biomarker in everyday clinical practice. So I think those obstacles have to be overcome for us to say TMB is going to be part of our everyday use. But I fully believe we’re going to get to a multiplexing. I think we’re going to be looking at TMB, PD-L1, and a couple of other markers within the next 2 to 5 years to decide who should really get I-O and who should not get I-O.
H. Jack West, MD: I think that the potential for negatively selecting…
Hossein Borghaei, DO: Yes.
H. Jack West, MD: If someone is negative for PD-L1 and negative for PD-L1, that may be the patient who is especially unlikely to benefit. But do we want to know that? Will patients take that and say, “OK, now I happily accept that I’m not going to get it,” or is this something that’s going to be a tool used by payers as a barrier or a requirement? Not that it would be wrong to actually do some kind of means testing to see that you could benefit. So how practically do you see it being implemented to support that concept of negatively selecting the patients least likely to benefit?
Hossein Borghaei, DO: I think that’s a major issue. If you can avoid the toxicities of these drugs, even though it’s only 15% of patients who get really severe toxicities, look at all of the randomized head-to-head to chemotherapy kind of things in the second-line setting. But still, these toxicities are real. These patients end up in the hospital sometimes because of these toxicities. So if you can select a patient population that’s absolutely not going to benefit from I-O, forget about just the finances. Not that you can, but let’s just put that aside for a minute. Just from a toxicity point of view, I think it’s very important. Moreover, it gives an opportunity to think about that patient by population and say, “OK, what is it, biologically, that these patients are telling us, and what would be the better way of treating them?” So it’s a negative selection for the currently available I-O proper drugs that we have, but who knows down the road. Maybe a new IDO [idoleamine 2, 3-dioxygenase] would be perfect for this patient population, or a LAG3 [lymphocyte activation gene 3] would be perfect for this patient population. So it opens up those sort of avenues that we haven’t explored so far. But I think a negative selection would be extremely helpful to me, not giving a drug that could cause a lot of adverse effects to a group of patients.
Charu Aggarwal, MD, MPH: It would help you personalize therapy.
Hossein Borghaei, DO: Absolutely.
H. Jack West, MD: Do you see TMB as being specific for identifying patients more likely to benefit from NIVO [nivolumab] or NIVO/IPI [nivolumab plus ipilimumab], or do you see this as being a marker of benefit from immunotherapy regardless of the specific agent?
Hossein Borghaei, DO: I would say from Dr Naiyer Rizvi’s work and from other people at Memorial [Sloan Kettering Cancer Center], we know it works for pembrolizumab also. So I think the currently available checkpoint inhibitors, the PD-1 [programmed cell death protein 1], PD-L1, the CTLA4s that we have in the clinic now, we have a lot of retrospective data to suggest that PFS and response can be better in the TMB group. So I think it’s not just something for IPI and NIVO, I think it is something that’s applicable to what we’re using right now. Now again, is it going to be applicable to the OX40s or the CD73s, or the LAGs? That information, I don’t have, and I’m sure we can generate that information. But it just does seem to be a little bit of an intrinsic tumor property that’s telling you that I have a lot of mutations, and therefore that might be more meaningful when it comes to choosing an immunotherapy-based approach.
H. Jack West, MD: Let’s look just a couple to 3 years ahead. Do you see TMB being a routinely used tool? And if so, do you see it replacing PD-L1, being 1 component alongside it, or even maybe being a bigger composite that may include gene expression profiling? Where do you see patient selection for immunotherapy and TMB in just a few years?
Charu Aggarwal, MD, MPH: I think TMB is here to stay. There have been multiple studies that have shown that TMB can predict response to immunotherapy, not necessarily immune combination versus chemotherapy, but just immunotherapy in general. It’s an inherent property, as Hoss said, off the tumor’s heterogeneity or mutational burden to basically predict how many epitopes they have, how many antigens they have, and whether they will respond to immunotherapy. Do I think that I will be using it in the future as the sole marker? No. I don’t think it will replace PD-L1 completely. I do think that we’ll be looking at, sort of like a panel. I think it will be an immune panel, just as we order a gene sequencing panel to look for molecular alterations. I think we’ll be ordering an immune panel, and wouldn’t it be wonderful if we had 5, or 10, or 15 things for which we could come up with an immune score and say to the patient, “I think you should get chemotherapy.” Or, “I think you should get LAG plus chemotherapy.” I think that’s where we’re headed, and it would be such a wonderful but complicated world.
H. Jack West, MD: Hoss?
Hossein Borghaei, DO: I agree. I think multiplexing is the way to go, and I think that’s where we’re going to get to.