Somatic Mutation Testing for Ovarian Cancer

Transcript:Bradley J. Monk, MD: Germline testing is important—not just for BRCA, but a panel. You said that’s for the family. Now, let’s talk about the tumor. You talked about somatic. Should there be other genes in the tumor that we test? With other genes, we just agreed, but are there other somatic genes that we should test for other than somatic BRCA?

Gottfried E. Konecny, MD: To summarize, most studies have just looked at BRCA1 and BRCA2 mutations, be they germline or somatic. And there are fewer studies that have looked at all-comers—basically platinum-sensitive disease—and that are retrospectively analyzing what are the markers that reflect response to PARP inhibition. Truly, there are other genes that come up, like RAD51C, methylation of the same, and PALB2. But they’re small numbers. In the end, if the genes are biologically relevant for breast and ovarian cancer, I believe they’re likely going to be predictors.

Bradley J. Monk, MD: But we have a molecular signature that includes BRCA-like genes, which are predictive of PARP sensitivity. That’s called homologous recombination repair deficiency, or HRD. Do you test for HRD? You told us you do a panel of germline. Do you do a panel in the tumor for somatic, which would essentially be HRD?

Katie Moore, MD: I think, like everyone else, I’m still evolving. What we’re moving to is a decision tree where you do a germline test, and patients who are positive are going to get a PARP inhibitor, and you’re going to counsel their family on whatever is the appropriate targeted therapy. Those who are negative in that population of patients—BRCA wild-type patients—we are increasingly thinking that’s a group to look at testing the tissue in and doing one of these assays. We don’t agree on the assay yet, but there’s an assay that looks only at loss of heterozygosity as a predictor of response to PARP inhibitors and probably other DNA response agents. Then, there’s another assay that looks at a combination of 3 tests, including loss of heterozygosity, to try and pick out those patients where, at some point in their tumor’s journey, that tumor could not fix its DNA well.

Robert L. Coleman, MD: Can I make comments about that?

Bradley J. Monk, MD: Yes, but she had a presentation basically showing that by doing next-generation sequencing, you can get an HRD signature, and there are other actionable mutations. So, should we do a panel of germline and then next-generation sequencing of the tumor?

Gottfried E. Konecny, MD: I think a possible solution would be that, at the initial diagnosis, you would do germline testing. And then there are some data that suggest it’s more important to test at the time of a recurrence, so that would be maybe 2 or 3 years later. When patients suffer recurrence, you would then go and do a tissue biopsy because it would add information on somatic alterations and it would also add information on potential changes over time that occurred in the tumor. So, that’s what I like to do.

Robert L. Coleman, MD: I agree. I think we tend to do that anyway. I think the data we have available are that there’s relatively similar findings as long as the patients haven’t been exposed to a PARP inhibitor. If you’re just looking at chemotherapy, pre and post, you know in the ARIEL trials—where we actually looked at the archival specimens—the ones that were done right before the entry onto the trial are pretty consistent. So, that’s good. It helps us with that.

But one thing I wanted to clarify, because I think people get confused about the HRD test: what you said, the genes that we’re testing in HRD are different. So, we all know that the HRD process involves a series of proteins at multiple steps and they have different roles. But they’re not all equal. Some of them, if they’re lost, it doesn’t matter. There are work-arounds. The synthetic lethal screen is very important right now, and it’s also identifying those opportunities for us to figure out new ways that not only PARP, but also all kinds of drugs, might have synthetic lethality that takes advantage of mistakes. The other thing is that this loss of heterozygosity gets us away from looking at specific genes.

Matthew A. Powell, MD: Signature?

Robert L. Coleman, MD: Right. It just says, “Hey, listen, the smokes there. We figure the fire’s probably there.” We don’t know where it is and exactly how to classify it.

Bradley J. Monk, MD: I like HRD for PARP, but I like these other molecular alterations at the time of recurrence that might inform my next clinical decision, including HRD and PARP inhibitors.

Gottfried E. Konecny, MD: But Rob makes a good point that if we focus too much on this HRD signature, it doesn’t always give you the most accurate pinpoints to those genes that are relevant. Because once you have a genomic scar, there are studies to show that you keep that. And you can have reversion mutations or other problems, and it then misleads you and suggests you have sensitivity to a drug that you may not have.

Bradley J. Monk, MD: Matt, you were telling me about cell-free DNA at the 2017 ASCO Annual Meeting. It’s an opportunity where it’s a liquid biopsy, right? You don’t need to stick a needle in. Tell us what was learned at the 2017 ASCO Annual Meeting from cell-free DNA.

Matthew A. Powell, MD: It was an interesting abstract looking at the patients who had plasma and tumor specimens collected at the time of their initial diagnosis and their interval debulking. They evaluated pathways of interest. It was quite interesting, and they actually found more mutations in the cell-free DNA than they found in the tumor biopsies.

Bradley J. Monk, MD: Right, and it controls for heterogeneity.

Matthew A. Powell, MD: Exactly. So, it’s certainly identifying additional areas where targets are important. And to your point, when we look at patients at the time of relapse, a more convenient biopsy may be a liquid biopsy.

Bradley J. Monk, MD: They were paired, though, and there wasn’t a lot of drift to your point.

Robert L. Coleman, MD: Right. But I think that we’ve still got to go back and think a little bit about what we’re getting out of cells for DNA. So, what is it? That’s the one question. We know it’s DNA fragments that are found in the blood, and we can characterize them. They’re nice because they’re dynamic. We can follow a dynamic process. You get a cycle of chemotherapy; you can get a blood test. After another cycle of therapy, you can get a blood test. It does process that, but we need to understand how that relates to tumor biology directly. I think that’s what the point of this was: that you know there is a pretty good overlap—it’s not 100%. And then, over time, you see this narrowing of clonality, which might be important. You might see leveraging of new mutations. But the point is that we have to also know whether or not that is actually reflecting the ongoing tumor biology.

Bradley J. Monk, MD: I think we’re getting better with that. We’ve clarified the role of a germline testing panel. We’ve clarified somatic BRCA. We’ll talk about an FDA-approved PARP inhibitor that is approved in somatic BRCA. We’re understanding the ramifications of an HRD signature, and we’re also understanding that there are other targetable genes. So, we’ll keep learning. You will keep working. This is your area of interest.

Matthew A. Powell, MD: Brad, one last thing you mentioned was the Lynch syndrome genes. Obviously, the MSI-positive cancer indication with immunotherapy is quite exciting. So, we have now one more thing to look for in our cancer, although it’s quite rare.

Bradley J. Monk, MD: It would be exciting to give a patient pembrolizumab on-label if you could find a microsatellite unstable patient.

Robert L. Coleman, MD: Right

Matthew A. Powell, MD: Again, that Lynch syndrome cohort would probably be one that we would target.

Bradley J. Monk, MD: I actually have one patient with it that I’m treating.

Transcript Edited for Clarity