The Biology of Extending Survival in Advanced CRC - Episode 7

Exploring Global Variations in Molecular Profiling


John L. Marshall, MD: We need to know for most of our patients, pretty early on, what molecular profiles they have. So, Paul, at Indiana University, when does a colon cancer patient get profiled and what tests are being done?

Paul R. Helft, MD: Historically, we have done most of our testing, at least the basic testing, in-house. We have a large enough pathology laboratory to do that. More and more, though, we’ve moved to doing next-generation sequencing on all of the patients up front. The trouble is, there’s always a lag time of around 4 to 6 weeks, depending on which test you order. That lag time is frustrating. Even though we, for the most part, know that a 4- to 6-week lag time is not biologically or medically important, it is certainly psychologically important for patients to wait.

John L. Marshall, MD: What markers do you want to know for every colon patient?

Paul R. Helft, MD: Well, at a minimum, we would get microsatellite instability testing.

John L. Marshall, MD: On all colon patients? We’ll come back to that.

Paul R. Helft, MD: Yes, on all colon patients. We would get all-RAS testing, BRAF testing, and HER2 testing, now. Those would be the 4 sine qua nons.

John L. Marshall, MD: Is everybody nodding? Is there anything that we’re missing? So MSI, of some way, RAS, all-RAS, in the United States, I see lots of patients who have a KRAS wild-type report but not the other RAS genes, and the answer there is you’ve got to send it again.

Tanios Bekaii-Saab, MD: Yes, you have to have all-RAS.

John L. Marshall, MD: And the next-generation sequencing? What’s the approach in Europe for doing this? Is this done in your own lab? Are you using companies? Is it getting reimbursed?

Fortunato Ciardiello, MD, PhD: This depends on the country. In Italy, where it is a public health system, there are some qualified pathology labs that usually work in a network with an external quality assessment scheme to validate them. This works pretty well.

There are regions of the country where the system is much better organized, and so you have the answer in a reasonable time, within a couple of weeks of working time. There are some regions where this is still a problem. Most of Europe is like this. To my knowledge, in France, it’s the best organized system because, for a long time, they established those 28 centers that are designated for all molecular characterization with tumors. They’ve been much better in pursuing to get 100%, ideally, of patients tested, not only in colorectal cancer but also in other diseases.

John L. Marshall, MD: In most of the patients that we see with metastatic disease, particularly if they present with a synchronous metastasis, the biopsy we get is from colonoscopy, and that’s the only sample you have. And nowadays, we don’t often re-biopsy the metastases and things. Does that primary tumor adequately reflect the molecular profile of the metastasis?

Paul R. Helft, MD: What we know so far is that, especially in a patient who walks in with a synchronous primary metastasis, most of the time—really, the vast majority of the time—according to studies so far, there’s alignment in the molecular features.

John L. Marshall, MD: What we see there matches what we see?

Paul R. Helft, MD: Right. It’s a whole different story, of course, once patients have been treated—once they’ve had time with their cancer.

John L. Marshall, MD: Dirk, do you have a difference of opinion at all?

Dirk Arnold, MD, PhD: No, not at all. It perfectly represents the synchronous metastasis. It also represents the metachronous metastasis if there is not onset very late. We do know this.

John L. Marshall, MD: So, you don’t necessarily need a re-biopsy after adjuvant therapy?

Dirk Arnold, MD, PhD: No, I wouldn’t do it. If this is within a reasonable timeframe, not 5 years later, I would say there’s no necessity.

John L. Marshall, MD: I’m going to actually differ on this position. The more we have been doing repetitive molecular testing, the more we’re seeing these things change. I’ve even seen 2 tests with the same sample, going to 2 different companies. One’s RAS wild-type and one has a mutation.

Dirk Arnold, MD, PhD: This is bad quality assurance.

John L. Marshall, MD: There is the idea that “the test is the test” and in every other disease, like lung cancer or breast cancer, the standard is re-biopsy.

Paul R. Helft, MD: But just to push back on you, John, what do you think the rate of differential testing is?

John L. Marshall, MD: Very low.

Paul R. Helft, MD: About 5%?

John L. Marshall, MD: If that. If you had a RAS-mutated patient and now, it’s 2 years later and you’re thinking, “Well, I’m out of things to do,” is this a patient that should be re-biopsied?

Dirk Arnold, MD, PhD: You’re critically right. We say that if you treat patients with a mutation with anti-EGFR antibody, we may harm them. Therefore, it’s critical to see this and not only that the treatment is not adding any benefit. We’re only exposing the patient to unnecessary toxicity, and we may actually harm the patient. So, therefore, we really have to ensure this. I’m personally not so much interested in this kind of synchronous versus metachronous different time point biopsy. The question is clearly, what do we now learn with the new tools from the cell-free plasma DNA of these tumors?

John L. Marshall, MD: Something that’s being embraced in lung cancer, where mutations can emerge and you change your therapy, is that I’m also seeing the circulating tumor marker people back off their position to say, “We’re not really good at diagnosing those mutations, but we’ll help you follow them.” This is an evolving field.

Tanios Bekaii-Saab, MD: There is definitely a little lack of concordance, even in that lung study. These are patients. The recommendation would be to consider this for patients where you can’t access, essentially, with a biopsy. But your point is well taken. The question that remains is, of course, what do you do with discordant results?

John L. Marshall, MD: You pick your favorite one.

Tanios Bekaii-Saab, MD: Exactly. But if you have RAS wild-type liver metastasis or vice versa, RAS-mutated primary, and the colon, primary, is still intact, what do you do?

Paul R. Helft, MD: You treat the metastasis.

Dirk Arnold, MD, PhD: Right.

Fortunato Ciardiello, MD, PhD: In any case, if the tumor was mutant at the beginning, 10 years later it will still be mutant because RAS mutations are very early in the development. They are founding mutations. So, it could be very complex that you lose this mutation. The opposite can happen.

Tanios Bekaii-Saab, MD: Sure.

John L. Marshall, MD: The MSI test. Paul is saying we’ve got to do it in everybody. Does everybody agree? Does every colon cancer need MSI testing?

Tanios Bekaii-Saab, MD: Yes.

John L. Marshall, MD: How about every solid tumor?

Fortunato Ciardiello, MD, PhD: In the ideal world, yes.

Paul R. Helft, MD: Every gastrointestinal solid tumor?

John L. Marshall, MD: Every GI solid tumor?

Tanios Bekaii-Saab, MD: Well, because this is what we do. The answer is complex, but I do think that we should test MSI in all patients.

John L. Marshall, MD: I would want to know. If I had metastatic cancer, I want to know.

Tanios Bekaii-Saab, MD: If the resources allow, and this is the problem.

John L. Marshall, MD: Who’s not going to allow it?

Tanios Bekaii-Saab, MD: Let’s say, in pancreas cancer….

John L. Marshall, MD: Oh, if you have enough tissue?

Tanios Bekaii-Saab, MD: Yes, but it’s less than a 1% chance.

John L. Marshall, MD: Would you biopsy somebody just for that 0.6% expression?

Tanios Bekaii-Saab, MD: No, but I have a patient, for example, who had an MSI-positive tumor in the pancreas, but we identified it in the setting of Lynch

syndrome, and so that was easy. The patient had a complete response to a PD-1 inhibitor. We see those responses, and we see some dramatic responses, in some patients who have MSI-high status who traditionally don’t respond well to chemotherapy and have really bad cancers. That also is a resource issue. In the ideal world, you should test everyone for MSI, if you can.

John L. Marshall, MD: Fortunato, how is it tested in Italy right now?

Fortunato Ciardiello, MD, PhD: The problem in Italy, as in Europe, is that we do not have a regimen to reimburse the PD-1 antibodies for the treatment. And so, it’s still within clinical trials or in centers where you can have access to any kind of clinical trials.

John L. Marshall, MD: You can’t just get it? It’s not paid for by the health system?

Fortunato Ciardiello, MD, PhD: Because it is not approved yet.

John L. Marshall, MD: Will this new approval for pembrolizumab in MSI-high tumors change things for you

Fortunato Ciardiello, MD, PhD: You have to consider that usually, on average, the European Medicines Agency approval follows about 3 to 6 months on the average FDA approval and reimbursement policies are within different countries in different ways.

John L. Marshall, MD: Do you use IHC? Or is it more of a gene test?

Fortunato Ciardiello, MD, PhD: This is another issue, because, obviously, by protein expression IHC, you can have genetic analysis. Ideally, you should use a genome analysis, like next-generation sequencing platforms, because this will help you to have multiple assessments at the same time. But again, we are in an ideal world and if you have to limit your resources to the material you have, you have to prioritize. This could be very complex.

John L. Marshall, MD: We’ll talk more about that.

Paul R. Helft, MD: I just want to make a quick comment, and it sort of follows on what Tony was saying. If you look at MSI-high colon cancer, the overall response rate of the PD-1 inhibitors is not actually that astounding. It’s just that some patients have these amazing and very durable responses. And so, everybody deserves an opportunity to have that.

John L. Marshall, MD: I call it sort of swinging for the fences, although it’s very solid. We’ll talk a little bit more about the incorporation.

Paul R. Helft, MD: Even though the numbers are low.

Transcript Edited for Clarity