Experience with Interpreting Somatic Reports

EP: 1.Defining Germline Sequencing and Incidental Germline Findings

EP: 2.Tumor-Normal Matched Sequencing and Platforms Utilized for Tumor Profiling

Now Viewing

EP: 3.Experience with Interpreting Somatic Reports

EP: 4.Identification of Patients for Germline Testing

EP: 5.Role of ctDNA in Germline Testing

EP: 6.Operational Processes and Challenges

Mark Robson, MD: I think a lot of people have been doing, at least up to this point, doing tumor only sequencing. I just wanted to show what we talked a little bit about this before the program started. These are data actually looking at the identification of germline variance or potential germline variance using a computational algorithm on tumor sequencing. And this is a group of about 850 patients who had, CGP probably stands for comprehensive genetic profiling or something, by a commercial enterprise. And then they do have this fairly complicated algorithm that they use to try to figure out whether or not something might be germline. These were the results of this. It's interesting that they found nearly, slightly over a third of the patients had a potential abnormality. And then they took these to frank germline confirmation with a separate sample, presumably. Ninety-six percent of the variations that they identified were indeed confirmed to be germline, not unconfirmed to be germline. This particular algorithm supposedly works particularly well without a separate tube but it's interesting. And they found a lot of these germline variance in tumor types, again, without testing recommendations or without universal recommendations. Essentially using the tumor as a population screening technique. And perhaps not surprisingly, a fair number of the BRCA2 variance were non-BRCA tumor types. So, this is the idea of using the tumor, essentially, you're doing opportunistic population screening in cancer patients in an attempt, as Dr Ross said, to kick it out. Dana, what do you think, I have a couple of issues, or a couple of questions. What's been your experience because somatic reports frequently don't do the same kind of variant curation that we do on germline reports, particularly tumor only? What's been your experience with confusion around that issue?

Dana Farengo Clark, MS, LCGC: First of all, just confusion in the EMR [electronic medical record]. Things being listed as the patient is a BRCA2 carrier and it's a somatic BRCA2, so there's a host of those sorts of issues. Just incorrect documentation of the actual etiology of the mutation. The other thing is just HDVS [Human Genome Variation Society], the standard nomenclature when you get a somatic report. FoundationOne has a lot of these issues of they're giving you not standard nomenclature. So, every time I get a report from our somatic testing lab, I vet it. I will look it up in the germline literature because obviously, the way people vet things in somatic testing reports is very different than how we look at them in germline. I'm kind of the genetic counselor who does that sort of work to see, A, has it ever been seen in the germline, and B, what does it consider? Because they will be switching over, but historically, I think somatic labs didn't ever call things likely pathogenic, so they kind of bumped it up or bumped it down based on their experience level. And we don't confirm VUS, or variants of uncertain significance, so those would never even reach me.it could've been something that was likely pathogenic that they downgraded to a VUS.I always make sure, if things look suspicious, to sort of re-read the report. Sometimes, for people that are out there that are getting reports that they don't understand, there's always someone at the lab that you can call to ask them for the correct nomenclature because some of these, if you go look them up in ClinVar to determine their germline significance, they're impossible to find because the nomenclature is just completely different. I would encourage anyone out there to lean on the labs and ask them. And I know our group at Penn Medicine has tried, "Please use HGVS nomenclature like everybody else does," so our somatic reports are a little bit better in that they report in HGVS. But again, just the classifications can be very, very different. And to speak to your other point about that abstract, when we looked, we sort of did an algorithm with looking at high-risk genes, so BRCA, Lynch, and we found that about 75% of those that actually wound up with a mutation in what we've called a high-risk gene in the tumor, that 75% of them did confirm. We sort of thought, "Our sniffers are OK." Without knowing anything else other than that this is a high-risk gene and high germline to somatic ratio, we did confirm a lot of those. But the somatic reports can be tricky. I always tell a new genetic counselor, "Just know what you're looking at. Know what genes were tested. Know the pluses and minus, del/dups, all that sort of thing. Make sure you really look at the report or have someone walk you through it."

John Henson, MD: And on that point, you know what I like to do, is I use VarSome, the website, and I'm sure you folks use this as well. But it will help to clarify those issues. In fact, it'll even give you synonyms. And I'll tell you, any VUS that seems like it might be relevant, I run it through that, and it's remarkable how many times you come away with that sense that you better think really hard about ignoring that VUS. And VarSome, I personally find that to be a very, very valuable site.

Mark Robson, MD: I always have a very difficult time with VUS because in the surveillance setting, it's pretty straightforward if you're telling people to do an MRI or not. But here, if you're talking to family members about having body parts lopped off, it gets a little anxiety-provoking, especially if you've identified it through basically population screening, unanticipated finding in the setting of tumor profiling. It's awkward.

Transcript Edited for Clarity