Mutational Landscape of Squamous NSCLC

Transcript:Benjamin Levy, MD: Let’s move on and talk now about the relevant genetic alterations in squamous cell lung cancer. For better or for worse, squamous cell hasn’t shared the same advances that adenocarcinomas have had related to targeted therapies. I think things may be changing. I think that we’re now learning that there may be relevant genetic alterations in squamous cell that are potentially actionable. But the story is a little murkier in terms of these genetic alterations, a little more complex.

Every time I try to learn about this I get a little more confused. It’s not the nice aligned pie graph that we see in adenocarcinomas with these mutually exclusive genetic alterations that we can wed to targeted therapies. It’s such a clean story and it’s been such a success in the adenocarcinoma population. Given its complexity and Paul, you’ve been a thought leader on this nationally and have been running trials and looking at targeted therapies in squamous cell.

Perhaps you can distill this down, maybe talk initially about the TCGA data and then talk about what you’ve been doing in terms of what’s relevant and what potentially is actionable, some of the drugs out there. I’m curious to have you give input on your experience.

Paul K. Paik, MD: Sure. I think that there are a couple of points you had mentioned which are important. One I think is the back history to genomic discoveries in squamous lung cancer. I think part of the problem is with the discovery of EGFR mutations back in 2004 and the realization soon after that these don’t happen in squamous lung cancers with any appreciable frequency meant— that squamous lung cancer got left behind because all these subsequent discoveries also that we’ve made, by and large, were not present in squamous lung cancer.

So, you’ve had this big gap when there was an explosion in therapeutics and investigations in lung adenocarcinomas that just did not happen in squamous lung cancers. This changed really with the cancer genome atlases (TCGA) effort, this multi-institutional network of very dedicated investigators who were tasked with trying to really figure out, on a very comprehensive basis, the genomic underpinnings of lots of different cancers.

The TCGA analysis for squamous lung cancer was published in 2012, but even before then at national conferences, we were getting glimpses for these alterations. The bottom line is that they don’t fit well into a pie. That’s exactly right. So, whereas we have the comfort of a pie for lung adenocarcinomas, we just don’t have that for squamous lung. And, there are a couple of reasons for that. The first is that there’s a lot of overlap. For the drivers that we’ve detected that I’ll mention, they sort of mix and match together for any given patient and segregate out in a not very clean fashion.

On a superficial level, we can detect something that we think is a driver oncogene in the majority of patients, in excess of 60%. And the most common ones, the ones that have been the lead targets for us have been fibroblast growth factor receptor 1 (FGFR1) amplification, alterations in the upstream PI3-kinase pathway which is very common in lots of cancers (it just happens to be more common in squamous lung cancer), changes in THE redox pathway (NFE2L2 and keap-1), some small frequency changes in a gene called DDR2 which is a little bit not WELL characterized but involved with collagen signaling with the extracellular matrix, and then a whole host of other alterations in cell cycling and dysfunctions on the tumor suppressor genes. We’ve tried to target these, particularly the ones that happen relatively commonly—FGFR1 amplification upwards of 20%. This may be dwindling down—and particularly PI3-kinase alterations. We’ve presented these data at national conferences. One has been published already for upstream PI3-kinase targeting, and they’re by and large negative studies, unfortunately.

Benjamin Levy, MD: And these are single agent therapies or in combination with chemotherapy?

Paul K. Paik, MD: That’s right. So they’re all single-agent therapeutics, which is an important thing to call out because we had gone under the assumption, which is a fair assumption based on the preclinical data and our experience with lung adenocarcinomas, that if something is a driver, it should be targetable by single pathway inhibition. And that’s just not been the case. We’ve seen modest responses for patients who have FGFR1 amplification, which is to say that this is not the entire story. There’s a lot of work going on to try to figure out why that’s the case. Many of us believe that amplification alone does not correlate with activation of the pathway. There’s a discrepancy of this in protein overexpression. And, the other part of this is that there’s just a lot of overlap with THE FGFR1 and PI3 kinase pathways, with G1S checkpoints.

Benjamin Levy, MD: So they’re not mutually exclusive.

Paul K. Paik, MD: They’re not. Yeah, that mutual exclusivity which allows us to construct a pie for lung adenocarcinoma does not exist really for squamous cell carcinoma. So, the future directions really are combination therapies, but that, as you know, as we all know, is very challenging to perform on a number of different levels. But, in a nutshell, that’s essentially where we stand. Where we have a lot of information, more than we’ve ever had, it’s much harder to take action on.

Benjamin Levy, MD: Again, looking at what’s considered to be potentially actionable, it’s been confusing for me. I’ve recently looked at what we’re testing in Ion Torrent panel, our 50-gene panel, and many of the things you mentioned we aren’t looking at. FGFR, we are, but fusions and amplification we certainly are. So do we need another diagnostic platform to test for those relevant alterations in squamous or can we use a comprehensive platform that includes those?

Paul K. Paik, MD: I think from a research perspective, comprehensive platforms are great. It allows us to capture that complexity to then tease apart exactly what’s going on in a rational basis. I think for now, because we can’t take action really on any of these things, from a practical standpoint, it doesn’t really matter. Once we’re able to find something that works, then we can talk about expanding these platforms. But, for now, again, there’s nothing really that’s important to test clinically outside of some of these more atypical alterations for EGFR mutations and ALK rearrangements that happen in lung adenocarcinomas that we potentially can take action on.

Benjamin Levy, MD: Ed, any thoughts on the targeted paradigm and squamous cell carcinoma?

Edward S. Kim, MD, FACP: Yeah, I agree with both you and Paul’s comments on this area. To go back in a couple of sort of philosophical directions, one is the science is great, then we have to do it. Serendipity is also great, as well. And let’s face it, and it’s very rare that I’m the oldest person on a panel, but I’m now feeling that right now.

Benjamin Levy, MD: Congratulations.

Edward S. Kim, MD, FACP: I’ll accept that as a compliment. The fact is that even in 2004 when the two seminal papers came out in New England Journal and in Science, that it was not an immediate acceptance that EGFR mutations were in fact this important. And it was very difficult to permeate that through the medical oncology community and even among lung experts. We were still discussing the benefits of a quantitative assays versus a qualitative assay, and it wasn’t until 2008, 2009, where we first got our first indication with one of these drugs, erlotinib, in an adenocarcinoma population. And so the same story could be said of ALK in the early studies where that was not something that came out of the lab as an ALK-targeted drug, and it was totally a clinical experience in both cases where patients were treated, amazing results were identified, and it got taken back to the lab. And, it has subsequently led to great science which has included resistance mutations, such as T790 and drugs that target that.

They’re doing the same with ALK right now with these second-generation and next-generation inhibitors. So, I think it’s awesome the way it’s evolved, but I think the importance of it is what’s in front of you testing the patient. There’s no better substitute for trying to treat a patient who needs therapy with a drug in a clinical trial to do that. And I think that’s reflected upon. And if I go back to your pie analogies, both of you, and maybe I’m just getting hungry for pie right now, but adenocarcinoma I view as the best pie, maybe it’s your mother’s pie or someone’s pie that smells good, tastes good, but you only get a taste of it and the rest of it you can’t have because we don’t have that whole pie filled with mutations that are druggable. We have a portion of it, a small percent.

Squamous is like you really want pie, but it’s the worst pie you’ve ever had before, and it’s right in front of you and you don’t even want to take a bite because it’s just a mixed mess of things. You don’t know what you’re going to test. So I agree, larger panels in the context of a research program are very important. We’re doing that. We just joined as a care center of excellence site, and so we have a 600-gene panel with over 40 expression markers that are included with every sample that we send. So we’re going to try and gather as much information as we can in that type of manner.

Transcript Edited for Clarity