Advancing Precision Medicine Through Comprehensive Genomic Profiling (CGP) - Episode 7
The rationale for migrating away from single, spot testing to comprehensive genomic profiling assays in oncology.
Bruce Feinberg, DO: I keep stepping back, taking the one step back because I don't want to lose our audience. But for most of the audience, they probably are more familiar with NGS [next-generation sequencing] than they are with CGP [comprehensive genomic profiling]. But NGS is the method and CGP is the outcome of that method, if I might think of it that way. But if you could go into it, into a little bit more depth and distinguish between what we were doing in terms of a single test, and looking for potentially a protein vs the gene. Then, this migration from single and spot test into these small profiles into more comprehensive profiles, and using lung cancer as an example.
Mark G. Kris, MD: Well, I'm going to challenge Eugean a little bit. If I think a patient is anemic, I want a hemoglobin. I don't have any right to get a platelet count or white blood cell count? Is that where we're going now?
Bruce Feinberg, DO: No. That's…
Mark G. Kris, MD: Wait a minute. I mean who decided a zillion years ago for doctors that needed hemoglobin, they got the white blood cell count and the platelet? I didn't care about that. But for technological reasons, it made sense to bundle those. That's the same thing here. How did those 500 gene panels happen? They took experts in all the common cancers. They said, "What genes are relevant to your cancer", and they added them all up.
These aren't 500 genes to interrogate lung cancer or sarcoma. These are maybe the 50 genes that are important sarcoma, the 20 genes important in lung cancer. They develop this multiplex testing, as an efficient way of doing it. It wasn't like people were doing research: "Let's just see what this gene and sarcoma means in lung cancer.” No. That gene was being done for the sarcoma people to establish risk or choose treatments. It was a convenience, I'll call it and a technological necessity, I'll say to do this.
There's also the practical thing too about the spot testing. When you need tissue, there's only so much tissue people have. If you just keep going back and taking another piece of tissue to do this and that test, the tissue is exhausted. Talk about the cost of an NGS test. The cost of a biopsy is a lot more, let alone the danger to the patient, but I don't want to go out. Forgive me for saying that. But these tests were done as a convenience, just like the CBC. I mean somebody decided that you guys were going to pay for platelets and white [blood cell] count whether we needed it or not.
Bruce Feinberg, DO: Eugean, sorry you have the target on you, you're their target, but before you respond, so Mark, if you were going to do more of a restricted lung panel only, would that be 30 genes that you'd be looking at?
Mark G. Kris, MD: Well, probably not.
Kenna R. Mills Shaw, PhD: Not, if you want TMB [tumor mutation burden] and lung cancer in particular, you need to have enough landscape. That's really what's driven, I think, a lot of the expansion of the size of some of these platforms is the need to get enough landscape to also be able to call tumor mutation burden—which has become a diagnostic—that's approved in a tumor agnostic way for many patients.
Mark G. Kris, MD: I have to say Kenna, in the lung cancer community the TMB is just not accepted. You should say that the genes are being added to this panel, though. I don't think that's the reason, I mean that's the consequence. But it is being added because different diseases feel we want to see this gene. At least, that's the way it's done at our place.
Kenna R. Mills Shaw, PhD: I'll have a chat with your scientists, but there's a lot of genes on all of these panels that really are not either truly significant from a biological, from a clinical, or prognostic perspective. I have the gene list and about half of the genes on these lists are not actually targetable or actionable from any perspective. There are mutations….
Bruce Feinberg, DO: Eugean, just let them argue, they're going make your case. They're going to make your case.
Eugean Jiwanmall, MPH, MBA: I was actually taking a breath because I was like "I was the target, no pun intended. “I was the target; however, somehow, there was divine intervention and Kenna and Mark disagreed on tumors—for folks for listening, TMB is tumor mutation burden. This would be really getting technical here. I'm going to try to balance between being going down that line, but I'm glad we're bringing it to tumor mutation burden. But there's a number of shots that have been fired. I'm going to try to tackle as many as I can in simplest terms. Of course, as the conversation goes on, we'll touch upon more.
Let me go back to actually what Bruce said. Because I think that's going to be and what he was asking. He was asking about NGS and CGP. Folks, I can understand NGS. We are not talking about that long ago. I grew up with this. When I was in a lab, when I was in an academic lab, I was in my academic setting, it was called Sanger sequencing. I don't even know if folks actually talk about that anymore. But that's just a little fun, I'm having this Sanger sequencing. But that's Sanger sequencing because you were looking at 1 marker for say, let's make it easy, so it was the easiest world out there, even for the care or somebody who's evaluating. Let me stop calling it just a care. We're talking about institutions that actually evaluate the evidence that is out there. We have to make evidence-based decisions. That's the bottom line I'm going to keep coming back to whenever I try to answer questions, because this is where folks are like "Why are you doing this'? We [need] multiple resources that we can actually use to evaluate, to establish medical necessity, that is safety and effectiveness of a particular intervention. In this case, we're talking about testing. Going back to my nostalgic world.
We're talking about Sanger sequencing. There, you have a marker, you're working hard, you talk about your step thing. You talk about how to do back in the day. You have multiple ways. You can even do sequencing on that, but it gradually built up. You were dealing with a small volume. You were dealing with one marker or maybe a couple of markers that are out there then things started exploding, rightfully. Technology actually developed, so you got into something called next-generation sequencing. Nice, advanced platform that can look at multiple markers at the same time started coming up. That was called panel testing. I've spoken about this at multiple events over the years. When panel testing first came, everybody was like "No, absolutely not. You can't do these."
Then, there was different sizes of panels that were there. What started happening was that technology advanced. You got this testing that was accessible to you. We had to differentiate between the research aspect of it and the actual clinical aspect of it. It took time. It took time for the data to aggregate; for some of the terms that we have talked about, Kenna's talked about, Mark has talked about. It took time for the clinical validation to happen for the technical and accuracy, and precision to be developed for these platforms. We are not even questioning that anymore for NGS, if folks are actually familiar with it. Can you move on to the fact that you are going from small panels to larger panels and then larger?
At the same time, something else happens, now bringing cost into this because first we make medical decisions and then we get into cost. The cost started to come down too. The ones who were responsible for paying for it, their strategies changed too. Yes, as a firm believer, a believer in looking at evidence, looking at their literature and guidelines, that's my role, that's why I get paid. I'm going to always stick to that, but I also want to bring in that there was a cost-effectiveness portion, that this is going back to what Mark was bringing up, who decided? There are multiple factors here but the bottom line is evidence, and then you build upon any cost-effectiveness that is out there as well. When the panel testing was happening, people were grappling with it. There were some panels out there that try to get the clinical utility data. Overall, they understood that they're never going to be able to establish the genotype to phenotype correlation or the biological relevance and simple terms for all the markers that were part of it.
Transcript edited for clarity.