Transcript: Benjamin P. Levy, MD: In closing, Phil, do you just want to give us a little bit of a forecast? Put out your crystal ball here. Where are we heading in the next 5 to 10 years with precision oncology for solid tumors?

Philip Agop Philip, MD, PhD, FRCP: I have to remind people that they’re rare cancers, but they’re also very common cancers. What do we do with them? It’s great to treat 1%, but what happens to the 99%? And that’s what we should be starting to think about. At this moment in time, most people think NGS [next-generation sequencing], NGS, NGS. But is NGS enough? In my opinion, obviously not because that leaves out a lot of people. If you think of KRAS as NGS detected, that’s not a driver. It’s just an ER [estrogen receptor] negative, not to give a patient tamoxifen.

So I think the future will be moving into us better molecularly profiling patients. That has to be the case, otherwise we’re stuck with NGS in the 1%, 5%, or whatever percentage we’re talking about. And that’s really very important. We really have to be working with people who do these tests, so that they develop them in ways that are more clinically meaningful and serve our purposes beyond just giving us a percentage of patients who have 1% or 2% of it. So NGS is very helpful, but it’s not the answer for precision medicine for the majority of the patients.

Benjamin P. Levy, MD: Do we think—and I think we can all agree, for now at least—that the sequencing technologies have outpaced what’s in our therapeutic armamentarium? We understand a lot about tumors right now genetically, but we a haven’t yet developed the drugs that we need, and we’re finding these rarer fusions and mutation—in our G fusions, FGFR, RET in the lung realm, and other solid tumors. Are we moving more toward the basket studies as we start to look forward?

John L. Marshall, MD: Actually, I was going to go a little more into fundamental biology. I learned that cancer was clonal. It’s clearly not.

Benjamin P. Levy, MD: No.

John L. Marshall, MD: And the emergence of resistance and resistant clones, and whether they have emerged or they were there all along, where you biopsy, tumor heterogeneity. So as we drill down on these tumors, our fundamental understanding of cancer is changing. The NGS lessons and our targeted therapies have taught us that over and over again. Where I think we’re also heading is trying to measure the polyclonal nature of cancer. And what is that relationship between the host and the tumor? Immune therapy is teaching us a lot about that. So I’m hoping that this polyclonal understanding and our better understanding of hosts will actually give us the kind of information that will be the next breakthrough.

Marcia S. Brose, MD, PhD: What I’d also add is that in our molecular tumor board, for instance, 1 of the big areas that we just jump over is the fact that many of the mutations we’re finding now are passenger mutations and mean nothing. I think that if we’re going to be moving forward with NGS, even to start with, we have to make it really clear, especially to patients. We get a number of calls for, “Oh, here’s my profile. Does it do anything?” You see a list of passenger mutations, and probably none of them are actually disease causing. And so the problem is now we have a test that is very good and sensitive for picking up these wonderful things when we do have a target, but it’s also very good at picking up a lot of junk. So the future is going to have to be more sophisticated in gaining support.

Philip Agop Philip, MD, PhD, FRCP: I mean, some of it can be junk, but some of it is reported as unknown significance.

Marcia S. Brose, MD, PhD: Sure.

Philip Agop Philip, MD, PhD, FRCP: And I think it’s our responsibility. Something came up in the discussion a little earlier: We have to share the data.

Marcia S. Brose, MD, PhD: Right.

Philip Agop Philip, MD, PhD, FRCP: I think that’s very important because you have so many people doing these things, and the clinical data are out there. I don’t know how much of it is really being filtered or is trying to be matched with the NGS or other data. It may be that we will discover more and more, and we have those powerful informatics. But then again, you have differences.

Marcia S. Brose, MD, PhD: I remember, back in the day, when we did BRCA1 and BRCA2. There was a whole database, and every time you got 1 you put it into the database. Every single time there was a hit, it was disease. It gave us all that information regarding which were disease-associated and which weren’t. But now it’s like we’ve multiplied that by several hundred. We don’t have databases for each gene, but I think we need to do that—still collect that data.

Mark Agulnik, MD: What’s also fascinating is that this is the first time for which every single time I get back another report, it’s almost as if there’s an opportunity to learn something. But every time I get back another report, it’s almost as if this is the first report I’ve ever received. I’ve never received 2 reports that were the same. So the learning curve is exponentially growing. Our understanding about the diseases we treat is ever changing. The biggest challenge—and we just did a questionnaire study with patients—is to temper expectations. The expectations are enormous from patients. My expectations are very minimal. I expect to see what I know I’m going to see, which is very little regarding actionable markers. Again, most things being passengers very little or unknown significance. Patients’ expectations are enormous. Despite the fact that I say, “Less than 1%,” regarding the chance that I’m going to find something, they hear 1% as 80%. It’s just impossible to translate that at that stage.

Philip Agop Philip, MD, PhD, FRCP: But what are the chances of winning the lottery?

Mark Agulnik, MD: I tell them that as well.

Benjamin P. Levy, MD: What a fascinating discussion. We will deep dive into NGS. This whole idea—of drivers versus passengers, subclonal mutations, tumor heterogeneity, host versus tumor factors—is a very complex story making the treatment fascinating and, at the same time, very, very challenging. So we will table some of this more in-depth conversation to the last section, which I think is important when talking about how we interpret results and how we use them within our disease states.

Transcript Edited for Clarity