ONCAlert | Upfront Therapy for mRCC

Case 3: Tissue & Plasma Sampling in Stage IV NSCLC

Targeted Oncology
Published Online:12:40 PM, Wed September 26, 2018

Anne S. Tsao, MD: I’d like to turn back to the case. Greg, what are your thoughts on this patient, who really doesn’t have full immune profiling from adenocarcinoma.

Gregory Riedlinger, MD, PhD: Apparently, the tissue is always the issue. Most laboratories generally have a cutoff for 20% tumor content for any of these types of sequencing-based assays with the idea that with a heterozygous mutation, you can then detect at 10%. Most of these assays are validated down to at least 5%. Obviously, we’re sometimes trying to push the limit in these cases, when we’re getting tissue and deciding what gets sent. We see it relatively routinely with Foundation Medicine. They’ll call some of the mutations, but they’re not able to calculate the tumor mutational burden or the microsatellite instability. They’ll kind of put it out as a qualified report. You should be aware that you’re potentially missing some driver mutations.

The other thing that we published on, as well as Memorial Sloan Kettering Cancer Center, this year, is that when you’re doing this sequencing, you’re not only sequencing the tumor. You’re sequencing everything. You’re sequencing the stroma and the hematopoietic compartment. Especially with a lot of these older patients and patients who have been on chemotherapy before, they often have clonal hematopoiesis of indeterminant potential. That can be picked up on these sequencing assays. If you see something like DNMT3A, that’s something to be aware of. But other things, like p53—you don’t necessarily know where it’s coming from. You’re also picking up germline alterations. Sometimes, with Foundation’s algorithm, you know they’ll fail certain specimens. But occasionally, if something on their germline-calling algorithm pops up that they think is potentially in the tumor, they might put that out as a real unqualified report. That’s another thing to be aware of when you have to interpret these reports.

Anne S. Tsao, MD: Ben, in this case, when you have no genotyping, what do you do?

Benjamin P. Levy, MD: Unfortunately, this is an all-too-common scenario. In much of my clinic, these are second opinions. We’re trying to get the tissue over. Or, sometimes we’re doing a biopsy and we don’t have enough. I have routinely used a plasma genotyping platform in treatment-naïve patients when we have the tissue that’s made the diagnosis of lung cancer but there’s not enough tissue for molecular interrogation. Of course, we can’t get PD-L1 [programmed-cell death ligand 1] off the blood, but I think there is a role for plasma genotyping in these patients for whom we have a diagnosis of lung cancer but the tissue isn’t sufficient. I think there’s a role.

The turnaround time for these tests is anywhere from 5 to 7 business days, and you’re trying to get going on a therapy. I think the presence of a mutation, when you find it in the plasma, essentially rules in, given the high specificity and the high positive predictive value. The absence doesn’t tell us much. I think we need to default to tissue. When tissue is insufficient, and if this were my patient, I would default, even though the patient’s a current smoker and unlikely has an actionable mutation. I would consider a plasma genotyping test to try to identify a targetable mutation.

Anne S. Tsao, MD: Paul, what are your thoughts? When would you test?

Paul K. Paik, MD: I agree. Increasingly, I tend to test more often. I tend to do things simultaneously now, because there really is no downside. Part of the benefit, at least at Memorial Sloan Kettering Cancer Center, is that we have an inhouse ctDNA [circulating tumor DNA] test that we can rely on. Preferentially, we’ll try to get a tumor biopsy done for complete testing. But, in a case like this, where it’s clear that the patient has rapidly progressive disease, there’s not going to be a lot of time to refer them to IR [interventional radiology] to get a biopsy done. Plasma genotyping really has a big role here.

Anne S. Tsao, MD: So, she’s gotten carboplatin/pemetrexed/pembrolizumab for 4 months and has rapid progression. Do you rebiopsy? Do you do plasma testing? What are your thoughts?

Benjamin P. Levy, MD: If the patient hadn’t had plasma genotyping or any molecular interrogation upfront, I think there’s a role for… It’s tough to do a rebiopsy on a patient who’s rapidly progressing. I think we all know that. I would probably do it if I hadn’t done it from the start. I would probably order a plasma genotyping test.

Paul K. Paik, MD: I would as well. And then, I’d hope that you’d get some period of disease stability later on with second-line therapy, to be able to get a formal biopsy if the plasma genotyping is a B negative.

Transcript edited for clarity.

Anne S. Tsao, MD: I’d like to turn back to the case. Greg, what are your thoughts on this patient, who really doesn’t have full immune profiling from adenocarcinoma.

Gregory Riedlinger, MD, PhD: Apparently, the tissue is always the issue. Most laboratories generally have a cutoff for 20% tumor content for any of these types of sequencing-based assays with the idea that with a heterozygous mutation, you can then detect at 10%. Most of these assays are validated down to at least 5%. Obviously, we’re sometimes trying to push the limit in these cases, when we’re getting tissue and deciding what gets sent. We see it relatively routinely with Foundation Medicine. They’ll call some of the mutations, but they’re not able to calculate the tumor mutational burden or the microsatellite instability. They’ll kind of put it out as a qualified report. You should be aware that you’re potentially missing some driver mutations.

The other thing that we published on, as well as Memorial Sloan Kettering Cancer Center, this year, is that when you’re doing this sequencing, you’re not only sequencing the tumor. You’re sequencing everything. You’re sequencing the stroma and the hematopoietic compartment. Especially with a lot of these older patients and patients who have been on chemotherapy before, they often have clonal hematopoiesis of indeterminant potential. That can be picked up on these sequencing assays. If you see something like DNMT3A, that’s something to be aware of. But other things, like p53—you don’t necessarily know where it’s coming from. You’re also picking up germline alterations. Sometimes, with Foundation’s algorithm, you know they’ll fail certain specimens. But occasionally, if something on their germline-calling algorithm pops up that they think is potentially in the tumor, they might put that out as a real unqualified report. That’s another thing to be aware of when you have to interpret these reports.

Anne S. Tsao, MD: Ben, in this case, when you have no genotyping, what do you do?

Benjamin P. Levy, MD: Unfortunately, this is an all-too-common scenario. In much of my clinic, these are second opinions. We’re trying to get the tissue over. Or, sometimes we’re doing a biopsy and we don’t have enough. I have routinely used a plasma genotyping platform in treatment-naïve patients when we have the tissue that’s made the diagnosis of lung cancer but there’s not enough tissue for molecular interrogation. Of course, we can’t get PD-L1 [programmed-cell death ligand 1] off the blood, but I think there is a role for plasma genotyping in these patients for whom we have a diagnosis of lung cancer but the tissue isn’t sufficient. I think there’s a role.

The turnaround time for these tests is anywhere from 5 to 7 business days, and you’re trying to get going on a therapy. I think the presence of a mutation, when you find it in the plasma, essentially rules in, given the high specificity and the high positive predictive value. The absence doesn’t tell us much. I think we need to default to tissue. When tissue is insufficient, and if this were my patient, I would default, even though the patient’s a current smoker and unlikely has an actionable mutation. I would consider a plasma genotyping test to try to identify a targetable mutation.

Anne S. Tsao, MD: Paul, what are your thoughts? When would you test?

Paul K. Paik, MD: I agree. Increasingly, I tend to test more often. I tend to do things simultaneously now, because there really is no downside. Part of the benefit, at least at Memorial Sloan Kettering Cancer Center, is that we have an inhouse ctDNA [circulating tumor DNA] test that we can rely on. Preferentially, we’ll try to get a tumor biopsy done for complete testing. But, in a case like this, where it’s clear that the patient has rapidly progressive disease, there’s not going to be a lot of time to refer them to IR [interventional radiology] to get a biopsy done. Plasma genotyping really has a big role here.

Anne S. Tsao, MD: So, she’s gotten carboplatin/pemetrexed/pembrolizumab for 4 months and has rapid progression. Do you rebiopsy? Do you do plasma testing? What are your thoughts?

Benjamin P. Levy, MD: If the patient hadn’t had plasma genotyping or any molecular interrogation upfront, I think there’s a role for… It’s tough to do a rebiopsy on a patient who’s rapidly progressing. I think we all know that. I would probably do it if I hadn’t done it from the start. I would probably order a plasma genotyping test.

Paul K. Paik, MD: I would as well. And then, I’d hope that you’d get some period of disease stability later on with second-line therapy, to be able to get a formal biopsy if the plasma genotyping is a B negative.

Transcript edited for clarity.
Copyright © TargetedOnc 2020 Intellisphere, LLC. All Rights Reserved.