Role of Tissue Versus Circulating Tumor DNA in Molecular Testing in mCRC

Transcript:
Kristen K. Ciombor, MD: Dr Kasi, we talked about circulating tumor DNA [ctDNA] a little, mostly in next-generation sequencing [NGS]. But what about minimal residual disease or molecular residual disease, MRD? Where do you think that will fit, either now or in the future, in our colorectal [cancer] practice?

Pashtoon M. Kasi, MD: You bring up an important point. In terms of the field of ctDNA and liquid biopsies as a whole, it depends on the context. We were earlier referring to advanced metastatic settings, where it’s mainly NGS platforms. In the context of minimal residual disease [MRD] or leftover DNA after curative intent therapy, those are different platforms looking at different methods. At ASCO GI [American Society of Clinical Oncology Gastrointestinal Cancers Symposium], the studies that have had some readout…and some of the data in terms of looking at timing, there are so many variables in terms of not just the assay. In addition to looking at the assay, you’re also looking at the timing of collection, biology, shedding, and the platform. We’ve gone beyond concordance. It’s not a question of whether the assay is positive. It’s more about what to do with the results of those assays.

Kristen K. Ciombor, MD: That’s a great point. Yes, we still have a long way to go. At the colorectal day at ASCO GI, are there any good studies that you think will be helpful? The ctDNA?

Pashtoon M. Kasi, MD: In 1 study presented, which we’re part of, we tried to look at this question of can you do a circulating tumor DNA–MRD assay earlier in the postoperative period as the patient is being seen? The GALAXY study from Japan just read out in Nature. If you look at the methods, they collected ctDNA at week 4 or later. But clinically, that’s when we want to start doing chemotherapy. You want to get information on MRD status earlier. Looking at real-world workflow, patients are being seen within a week or 2 of their operation. Ideally, we’re tagging along with the surgical colleagues at that visit or pre-op.

The question then comes up, will the background noise of cell-free DNA muffle the signal where would you have patients who are falsely negative identified but may still have disease? The presentation is going to outline that not only were we able to identify, but we identified more in the first 2-week time point that we looked at. The positive is that in terms of the proportions found in this large cohort of available data, we were able to show that it didn’t impact the proportion of individuals who were detected with the status. You still have to keep in mind the negatives because of the low volume of shed, but some of the earlier work looking at different methods and platforms may or may not be applicable.

From a kinetics noninvasive nature of the liquid biopsies, we’re presenting some work on the clearance of ctDNA. There are different cutoffs. [Kanwal] Raghav and his group have looked into the earlier disappearance of ctDNA. Different researchers have come up with different cutoffs of clearance. You can’t fake the ctDNA going away within 30% or 50% clearance. Overall, it seems like as if a very strong prognostic early readout, not just for response but for PFS [progression-free survival] and OS [overall survival]. In our cohort and the cohort from the group in Italy, the median overall survival wasn’t even reached in patients who had cleared ctDNA as early as 2 to 4 weeks of field evaluation. [We compare that with] a PFS and an OS that were no more than a year and a half and highly effective in the first-line setting.

Kristen K. Ciombor, MD: That seems like a very powerful tool. We’re still figuring out how best to use it.

Joel R. Hecht, MD: Are you also being inundated by the fact that it’s not just 1 or 2 companies that are doing this? Basically, every genomics company has its own assay. Even the original assay that 1 company has may change over time.

Pashtoon M. Kasi, MD: Absolutely.

Joel R. Hecht, MD: It’s not like IHC for loss of DNA mismatch or PER proteins. The pathologists have worked this out. The problem is that for the patient who’s in the clinic today, which assay? Is today’s assay the same as last year’s? Do you have any thoughts about that? Sometimes I get lost in this.

Pashtoon M. Kasi, MD: As they say, you don’t want to do cross-platform comparisons. You don’t want to compare the platform or the timing. If you look at the researchers looking at MRD detection rate after stage IV resections, with different cohorts there are different groups. If you look at the methods, it varies across platforms. It might be an in-house, plasma-only, 20-gene panel assay that’s reporting out whatever percentages. Some of the same platforms have gone from 70 to 100 to 500 [genes]. The same platform, which is a good problem, is improving over time. Across platforms, there are different lenses with which they’re looking at the cancer. As we’re digesting this information, it’s important to have some idea or at least look at the methods now more than ever before with this rapidly evolving technology. And it’s not just ctDNA of the methylated platforms. You’re looking at fragmentomics moving into early detection. It’s a good problem. The space is getting crowded, but now more than ever you need to look at the methods and keep an eye on that.

Transcript edited for clarity.