Molecular Testing to Determine Gastric Cancer Treatment

Transcript:Johanna Bendell, MD: Now, Ian, you have a huge phase I unit at the Royal Marsden, and you have all these potentially targeted agents you could give. Do you use any of these tests, as well, more in the experimental setting than in standard?

Ian Chau, MD: Yes. We have conducted a study to try to have all-comers coming through our clinic as new patients undergo next-generation sequencing, with a panel of however many genes you want. We then try to either influence the subsequent treatment or, as you said, direct them to the relevant clinical trial. In fact, we actually presented some data at this year’s ASCO conference on our experience, but we actually faced several things in real life. Number one: when we say we want to do all these sophisticated tests, a lot of our patients are diagnosed in the referring hospital. These are gastroenterologists that work in a referring hospital. They saw a tumor. They said, according to the guideline, they’ll take six biopsies. After six biopsies, not all of them actually have tumor in them. So, by the time they bring it in, they might be down to only three of them have some tumor in them. And after that, those that have tumor, the tumor content may not be very good. And then by the time you choose the one that’s got high content and then you try to extract the DNA, you’re down to very small amounts of DNA. It does pose some challenge as we try to push this into everyday practice. It’s actually not as simple as a research study where we say we’re going to get fresh biopsies, we’re going to get them straight to the lab, process them, etc. So, I think there are many steps as we try to take this in to the large scale of the routine clinical practice to target it [with a targeted drug].

Now, of course, despite having a big phase I unit, it also depends on what kind of clinical trials are available at the time, and you’re always not able to make a match. Perhaps somewhat disappointingly, there are some data coming out to say, well, when you match it with so-called matching mutations with the clinical trials, these patients still may not necessarily benefit a lot. I think it just tells us the capacity of the cancer. And sometimes even if you can know a mutation, it does not mean that in that disease that is really the driver mutation for it. And the drug that they’re testing may not necessarily be the best of that target. So, just because you’ve got a PI3 kinase inhibitor, doesn’t mean that it will work in [all patients with] a PI3 kinase mutation. As you said, it’s a common mutation in gastric cancer. I think we certainly do want to match it. But, Jo, you run a big unit, as well. Maybe you can tell us about your experience.

Johanna Bendell, MD: I think you’re right. I think we are mostly focusing on just getting those patients tested and saying, “Okay, let’s try to find out as much about these tumors as we can.” Sometimes, we find a matching mutation or an amplification that would work for the patient, but, if not, there are new agents coming every day where we can see a pathway that might be aberrant or the mutation might make them more susceptible to, say, a PARP inhibitor that will drive us toward treating that patient in a certain way. And I think when aggregating the data, sort of like the TCGA analysis. The more we get on data in real life and real patients who had these cancers, the more we can learn as we get more agents and we get more knowledge about things. So, I think it’s important to test. But just like you’ve said, just because we find one match, doesn’t mean that everything will come true. But it also means we could find a direction to go. I think one of the most important things is to tell people not to be afraid to test. Sometimes I have clinicians saying, “I want to check the molecular profiling, but I don’t know what to do with the results.” And that’s where I say to the clinicians, “Well, use the results that would change your standard, like the HER2.” But, if you don’t know what to do with the results, that’s the perfect time to refer or call one of us or other colleagues that do the research that can say, hey, this makes sense or this doesn't make sense. But don’t feel that pressure to know all the answers when you get that result back.

Manish Shah, MD: You make a really good point about that. I think that without consolidating the different data that we have, we won’t know. It might be that diffuse gastric cancer or genomically stable gastric cancer is better treated with a taxane than irinotecan—and that may actually be important in the second-line setting—or ramucirumab might be very good for a subtype or not. I think having that data will be very important. And I just wanted to come back to one other point that was mentioned by Ian. Currently, the TCGA analysis was not only just sequencing, but also expression profiling, and methylation, and a pretty comprehensive analysis. And for most patients, with a biopsy, it’s actually challenging to get all of that information. So, with better technology, you might be able to do that. But it may be that patients need to get another biopsy to actually adequately profile their tumor, so that we can actually use it for their future care.

Johanna Bendell, MD: What Yelena was talking about is the somatic mutations, ones that happen inside the tumor. But what about the germline mutations? What about the hereditary syndromes that can lead to gastric cancer? Manish, you know a lot about these.

Manish Shah, MD: Thank you. In about 10% of gastric cancers, a patient will say that they had a family history of gastric cancer. And, of those, the most common, by far, is the hereditary diffuse gastric cancer. These are caused by a germline mutation in CDH1, which is e-cadherin. Yelena alluded to this as sort of pathognomonic for the genomically stable type. They have mutations in cell surface proteins and proteins important in gap junctions. A germline mutation or a constitutive mutation in CDH1 is associated with gastric cancer. And, in fact, this is actually very important because the risk of gastric cancer in these patients is about 1% per year starting at age 20. The standard of care is to undergo a prophylactic gastrectomy, often in their early 20s, to prevent future gastric cancers. And there have been a lot of data, at least 150 patients, published with regard to the experience with gastric cancer and prophylactic gastrectomy. And many of these patients have already had precancerous lesions or NETs in the wall of the stomach that were silent.

Other hereditary syndromes, Lynch syndrome, is associated with gastric cancer. And also FAP is associated with gastric cancer, and that’s actually really quite interesting. A mutation in the APC gene, it affects beta-catenin signaling, and not only is it important in colon cancer, which we all know about, but it’s also important in gastric cancer. It gives you a 10-fold risk. There are things to be wary about, so I always ask my patients about their family history because it comes up not too infrequently.

Transcript Edited for Clarity