Inside the GENOME

Myriad Live - Let’s Talk Paired Tumor-Normal Testing for Treatment

May 24, 2021 Myriad Oncology Season 1 Episode 14
Inside the GENOME
Myriad Live - Let’s Talk Paired Tumor-Normal Testing for Treatment
Inside the GENOME
Myriad Live - Let’s Talk Paired Tumor-Normal Testing for Treatment
May 24, 2021 Season 1 Episode 14
Myriad Oncology

Myriad Oncology Live episodes are recordings of an open-forum webinar hosted by Dr. Thomas Slavin. The opinions and views expressed in this recording do not necessarily represent those of Myriad Genetics or its affiliates. To participate in a future recording, visit for a list of dates, times, and subjects.

Show Notes Transcript

Myriad Oncology Live episodes are recordings of an open-forum webinar hosted by Dr. Thomas Slavin. The opinions and views expressed in this recording do not necessarily represent those of Myriad Genetics or its affiliates. To participate in a future recording, visit for a list of dates, times, and subjects.

0:00:11.5 Thomas Slavin: Welcome. This episode of Inside the Genome is a recent recording of Myriad Oncology Live, a webinar hosted by me, Dr. Thomas Slavin, Senior Vice President of Medical Affairs at Myriad Oncology. The opinions and views expressed in this recording did not necessarily represent those of Myriad Genetics or its affiliates. To participate in a future recording please visit Myriad Oncology for a list of dates, times and subjects. I look forward to exploring the world of genetics with you all.


0:00:39.9 TS: Hello everyone, welcome to Myriad Oncology Live. I am T. J. Slavin, Dr. Slavin. I'd see, we're a little smaller audience, but I think we'll just... We'll kick it off. I didn't mean to stop sharing my screen though, so I'm gonna re-share my screen really quick so I can go through a little housekeeping. So thanks for coming on, and I assume... Yeah, we'll get probably more people joining over a little bit of time. But house-keeping, if this is your first time, yes, you can ask literally anything you want. These are theme-based discussions, totally open door. Whatever is on your mind, feel free to ask. We do try to theme-base them, today's is really talking about tumor-normal testing for treatment, so if you have questions, we'll go through different things, but yeah, feel free to derail the conversation, ask whatever is on your mind and I can hopefully get it answered.


0:01:37.8 TS: We also have... Next week, we're actually on a break week, and then coming back May 4th, we're gonna talk about hereditary cancer in the LGBTQA+ community. And then the schedule is built out into July at this point, but it is not yet posted, so hopefully those can get posted pretty soon. I need to write a note to myself to hopefully get those up. And then also before I forget, which I did last time until the very end, we do record these now just for people that can't make them, so if that plays in to your willingness to speak, I apologize. You can always send chats to Shelly Cummings who is on and she can read any questions that anyone may have. And I don't think any of them are actually posted yet, but... And I don't even know where we're gonna post on, we'll probably put them on the education page ultimately of Myriad Oncology just for any future reference.


0:02:45.5 TS: And then we've been doing some podcasts and stuff, people are looking for fun, some people are starting their commute back to work fortunately, or unfortunately, however you wanna look at it. But we had a raising awareness for colorectal cancer, we just proved out another two, just a really good podcast with Karen Hurley, who also I think will be on our... From the Cleveland Clinic, who I think will also be on our LGBTQA+ webinar in a few weeks. So yeah, these are all up, and keeps growing.


0:03:25.3 TS: Alright, let me stop there really quick, make sure there's no questions just to start, but I did wanna show... Oh, and actually, let me take a step back, we have a special guest today as well, and I have not met you Kristen, I just saw you on. I didn't know if you were gonna be able to make the first half or the last half. And so if you wanna unmute yourself, we can absolutely work around your schedule, so if you have time now we can just jump into...


0:03:58.7 Kristen: Sure, that's great, I appreciate that. Can you hear me okay? 


0:04:02.2 TS: Yes, yes, thank you so much for coming on. This is great.


0:04:05.6 Kristen: Thank you for inviting me. And actually, I finished in clinic early, so I can go any time but...


0:04:09.8 TS: Oh, fantastic...


0:04:10.9 Kristen: Since I am talking...


0:04:14.0 TS: Then why don't we start with...


0:04:14.5 Kristen: Diana reached out... Yeah, asking me just to describe my role a little bit with Aurora's oncology precision medicine program, and then just briefly summarize some data we presented last fall at NSGC. So I do have a couple of slides. I don't know if I have the capability to share...


0:04:29.0 TS: Yeah, absolutely, you should be able to share. If you look at the bottom...


0:04:32.6 Kristen: Yeah, let me just try that.


0:04:37.7 TS: Yep...


0:04:38.8 Kristen: Great.


0:04:39.9 TS: Perfect. And it's up to you, if you wanna put it in presenter mode, you can. But you don't have to, this is fine.


0:04:51.3 Kristen: Okay, well I was trying to do that, but it's giving me a error message there...


0:04:56.8 TS: Sometimes it's a little finicky...


0:04:58.0 Kristen: Yeah, sorry about that.


0:05:00.3 TS: Yeah, this is fine. Oh no, no.


0:05:02.2 Kristen: Okay, let me just move the video here. I don't know why that's not working. That would be ideal, but... Okay, we'll just move forward anyway.


0:05:10.7 TS: Yeah, this is completely fine.


0:05:12.9 Kristen: My name is Kristen, I'm a senior genetic counselor with Advocate Aurora Health. I've been with Advocate since August of 2018, and I primarily practice as a clinical genetic counselor, provide cancer genetic counseling, but I'm also the representative, if you will, to our oncology precision medicine program. So as part of that, I am part of the team that reviews all of the somatic test results that are done throughout the Aurora system. So I just wanted to review a little bit of information about our process and then about the data we presented. I have to get back to the Zoom call 'cause it's not working for me to share...


0:05:58.7 TS: You can't...


0:06:00.6 Kristen: I can't advance the slides...


0:06:00.7 TS: Yeah, maybe re-share. Or maybe it's that Enable Content thing that was popped up in the PowerPoint... Might be preventing...


0:06:07.6 Kristen: Yeah, let me... I don't know why it's doing that. That's a bummer.


0:06:15.0 TS: Yeah, you can try to re-share. Sometimes that fixes it.


0:06:19.6 Kristen: I'm getting it even when I'm in PowerPoint, it's not letting me click on it.


0:06:23.1 TS: Yeah, it's probably that. Do you see that Enable... You had some error message at the time...


0:06:26.5 Kristen: Yeah, I do, and I try to enable the content and that's not working. So let me just try one more time to share, but...


0:06:43.9 TS: Oh, do you have windows open in the background? 


0:06:47.8 Kristen: I do.


0:06:48.8 TS: Yeah, maybe close those other windows out. It may help.


0:06:51.9 Kristen: Okay.


0:06:52.4 TS: 'Cause I think that's what that little error message is saying.


0:06:55.4 Kristen: Okay, maybe let me play with this and restart. I don't know if you wanna talk for a little bit and I can come back after I close everything out and restart the PowerPoint.


0:07:04.9 TS: Sure, and I do have some... The reason I brought you up was I had some slides from, I think, your NSGC talk, and I'm happy to share those. Honestly, I can't even remember where I got those from. I was just looking in the folder. I put good things to talk about on these webinars in folders, and apparently in, yeah, sometime last year, or I don't know, a couple months ago, I put your talk in my folder. Do you want me to pull that up? 


0:07:38.6 Kristen: Yeah.


0:07:38.7 TS: It's probably the same slides or similar.


0:07:40.2 Kristen: It's very similar, I cut it down some and took out the timings. And I was gonna show our dashboard, but PowerPoint has just completely frozen, I don't know what the problem is, I can't close out of it.


0:07:51.2 Shelly Cummings: You can email them to me and I can load them up, if you wanna try that, Kirsten.


0:07:55.9 Kristen: Okay, let me try that, Shelly.


0:07:58.9 SC: Not my Perspectives email.


0:08:00.6 Kristen: Yeah, let me... What is the...


0:08:03.4 TS: Yeah, Shelly is our tech whiz at Myriad. [laughter]


0:08:06.6 SC: That is not true.




0:08:09.0 SC: S-C-U-M-M-I-N-G, no S at the end, at myriad dot com.


0:08:15.9 Kristen: Okay.


0:08:18.2 TS: Yeah, and then, in the interim, for those on the line, I do wanna talk about this... Well, since you have time, we can always come back once that gets worked out.


0:08:29.3 Kristen: Absolutely.


0:08:29.5 TS: But let me share this, 'cause this was another good paper if people have not seen it. I don't know if people are familiar, can everybody see my screen? Yes, okay, we're getting nods. So I think this is a great paper. So this is my buddy, Steve Lincoln, over at Invitae and friends, and they put together a nice paper that was in... What was this in? JAMA Open Network, last year. Let me see when it was actually published. Yeah, October of 2020. I don't know who on the call is familiar with it, but it's a nice paper, kinda getting at the tumor-normal space. And so in a nutshell, what they did was they said, "Okay... " They just kind of did a descriptive analysis, so just reported out data on people sending in cases that already... So on the TRF, they wrote, "Something was in this person's tumor, can you look for it in the germline?" And really using any sort of germline test, then they reported what they found. And so when I say any report type of germline test, it could be anything from the standard 40 gene panel to single-site testing to a larger panel, whatever, they didn't get specific or restrict anything. And the results are pretty interesting, and here's kinda like table one, characteristics, it was kind of split, male, female, kinda standard US breakdown for the most part of self-reported race and ethnicity.


0:10:18.9 TS: These people had cancer, which is... So they were found to have something on a somatic report, so that was how you got included in the study, so everybody had some type of cancer, some people had multiple primaries. And then you see kinda the breakdown of the age of diagnosis, when the germline test was done, number of genes, so they did write out... A lot of people just had single-site, most likely, or a very small subset of genes, but yeah, some people had larger panels. And this is kinda starting to look at some of the data, so what this is, it's... And it's all common cancers, so it's... Here's all these different cancer types, and what do we see, so when it comes to the germline finding? So again, these are people... So you can read this as, "Of people with the APC tumor mutation reported on a TRF," so this is 112 people had APC REN, and APC is a common gene in certain cancers, you'll see it a lot in colon, other types of cancers, mutate in the tumor. A lot of times it's a passenger, not so much a driver, but I think we're still trying to understand its role. Only 2% of the time, if you had it in your tumor, was it found in somebody's germline. So that's kinda how you read this chart, and then you can kind of look through the rest of these.


0:11:50.4 TS: And so there's some that really jump out, so AXIN2 here is the flip side of that, which is, it's not really a common tumor mutation, and the people that had it in their tumor, of those two people that had it in their tumor, both had the AXIN2 variant in their germline, and these are pathogenic variants. And so BRCA1 and 2, which a lot of people are probably ears up about, was about 40% here, so that's interesting. So overall, if you find a BRCA1 or 2 mutation in a tumor, actually a lot of times, then it's not germline, however... So 60% of the time, you could think it's not germline, however, a good chunk of the time it actually is germline. And yeah, you can just kind of look through all this.


0:12:39.0 TS: So it's an interesting table in itself, I mean MITF, high germline rate. So a quick way to look at this is, what are the genes with a really high germline rates? CHEK2, again, not the most common tumor mutation, but we all know 1100 LC, very common, and other CHEK2 mutations, 0.5 to 1% of the population, especially Europeans. And then MITF, I'm assuming this was a lot of the E308K variants you see here. Most of the time, yeah, it's in the germline. PALB2, this one was, I felt, pretty interesting. This is almost more interesting to me than BRCA1 and 2 in a sense, because it shows that, yeah, if you find a PALB2 mutation in anything, it's a very high likelihood that it's germline. So that's pretty shocking actually to find it 64 times on reports and then 46 of those, so 72% actually it was in the germline. Also, the succinate dehydrogenase complex genes, that's pretty high, 50%.


0:13:54.2 TS: So again, probably rare to have mutations overall in the tumor, and these mutations tend to be in more neuro-endocrine and paraganglioma type tumors, and then when you see them, yeah, it's pretty high likelihood to be germline. So RAD51C and D, look at that, that's interesting, very rare tumor mutations, which is what we've been seeing with our pharma trials and things looking at tumors, we've been seeing the same. I mean, you just don't see these mutated that often. They're a common homologous recombination repair gene and yeah, a lot of times it's due... There's germline mutation. So I'll stop there for any questions. There's a couple other really interesting figures and tables that I can walk through. And I can also bounce back to this. I don't know, Shelly, do you have Kristen's... Have you worked your magic? 


0:14:47.2 SC: I have it and she was... Yeah, she was also able to... You started on her end, so maybe we can flip back to her to...


0:14:55.9 TS: Yeah, why don't we do that? And then I'll come back to this to close it up.


0:15:00.0 SC: Yeah. And if it doesn't work on her end, I have them so [0:15:02.3] ____.


0:15:02.3 TS: Yeah.


0:15:03.8 Kristen: Great, thank you. Let's try again. Now it seems to be working. Can you see it now? 


0:15:17.2 TS: I do not. Are you sharing your screen? 


0:15:19.7 Kristen: Oh, let me... There we go.


0:15:34.0 TS: There you go.


0:15:35.0 Kristen: How about that? 


0:15:36.4 TS: Great. Yeah, perfect.


0:15:45.2 Kristen: Okay, wonderful. So as I was saying, I'm part of the precision medicine team. And at Aurora, all of the somatic testing that's done, at least the large-scale genomic profiling tests that we send out, are ordered through a centralized ordering process, which allows the OPM team and pathology to work together to find the best sample, order the best test. And then all the results come back to a centralized location. So we actually can track a lot of metrics that way and make sure that we're reviewing all of the genomic profiling that's being done within the system. So that's something that's reviewed by one of our pharmacists and by myself, and then some of that's done administratively. If there's things that are straight forward or if there's a report where there's no findings, we just make our recommendations back to the physician. But if there's anything interesting to discuss or if there's multiple findings and we wanna talk about sequencing or things like that, then we bring that to our weekly molecular tumor board and discuss those more interesting or complicated cases every week on Friday morning.


0:16:50.1 Kristen: So my role really is specifically related to any potential germline findings. I review every patient that comes through the precision medicine group and I do a basic chart review, looking at the diagnosis, any reported family history, to try to capture those patients who may qualify for germline testing that haven't yet been referred. But then I also do review those somatic test results, looking for any findings that we suspect may be germline, whether it's due to the allele frequency or the phenotype or some founder mutations that show up. We see a lot of CHEK2 variants that I can pretty much guarantee will be germline findings just by looking at them.


0:17:31.0 Kristen: So I review all of that and then track those recommendations, make the recommendation back to the physician, track those in a real-time spreadsheet. So this is just an example. These are some of our metrics from 2021 thus far. And this does need to be updated a little bit since we're now well into April, but it just kind of shows we keep track of the number of cases reviewed, why we're recommending referral for a patient, whether it's due to their history, their family history, or the test result, or all of the above, if they've been referred, if they've been seen.


0:18:06.8 Kristen: And then I think what's interesting is, for those who are referred based on some sort of somatic finding, what are the most common reasons that I'm recommending the patient see a genetic counselor? And it's not unexpected, as you can see in that chart there. The top reasons are things that we might expect, BRCA genes, ATM, CHEK2. So those sort of fit with why patients should be coming to see us anyways. And then this has evolved over time. We used to just kind of track the number of patients and whether they were referred and maybe why, and then we started really trying to track the test results themselves and comparing the somatic and germline findings, looking at concordant and discordant results, and that is essentially what prompted the data we presented last year. Based on all of the patients reviewed in 2019, we were looking at those germline and somatic results and trying to make some comparisons.


0:19:04.0 Kristen: So the purpose of that really was determining how often the patients had a germline finding that was not reported on the tumor test, and then exploring the trends among those results. And this could be germline findings of any variety. We made the decision upfront that anything reported on a germline report that was not reported on the somatic report would be considered a discordant result, although we know some of those variants of uncertain significance may not have clinical implications. But just for simplicity's sake, that's how we decided to categorize things. And then there's somatic findings where we were suspicious. They may be germline in origin, but then upon doing the germline testing, we did not confirm them in the germline. So again, some of those, it was maybe just a poor judgement call and they're truly somatic findings, but kind of tracking when something didn't fit quite what we were expecting to find.


0:19:56.2 Kristen: So again, this is our data from 2019, and in that year's time, there were 677 tests that were performed. 202 patients had both tumor and germline test results to compare in order to allow us to look at this. So that was then the population that we used for this study. And among those 202 patients, there were 55 discordant results in 49 unique patients. So about a quarter of the cases had some sort of finding discordant for some reason that we wanted to look into a bit further.


0:20:34.4 Kristen: So if we break it down, those 55 discordant results, there were 14 where the gene simply wasn't analyzed by both tests. So we found something on one test not on the other because we didn't look, and so obviously that's a little difficult to determine whether that comparison is really accurate or fair because the gene wasn't assessed by both tests in the same way. There were 26 cases where there was a germline finding that was not identified or at least not reported, I should say, on the somatic test itself. And then the remaining cases, 15 cases where there was a somatic finding that we were interested in, we thought might be germline, recommended confirmatory testing, but then it wasn't ultimately confirmed by germline testing. Of these, then I really wanted to step back and say, "Okay, but which ones might have clinical implications?" So which ones were classified as likely pathogenic or pathogenic and ultimately might have some treatment or screening implications for the patient or family members? And there were 14 total that were classified as pathogenic or likely pathogenic variants that we went on to look at.


0:21:42.3 Kristen: So this just shows of those 14 discordant results that were pathogenic or likely pathogenic which genes were involved. And again, this sort of fits with the reasons we were recommending referral, genes that we would expect to find, about a third was due to BRCA1 and 2, about a third to CHEK2, and then a variety of other genes made up the remainder. But I do think what was interesting here is that these are very moderate to high penetrance, actionable genes. It's not that we were identifying a lot of carriers, like MSH3 carriers or MutH carriers. These were truly genes where finding one pathogenic or likely pathogenic variant might have implications.


0:22:31.3 Kristen: As far as the type of variant that was identified, I listed them here, and these are not unexpected at all. I think this makes perfect sense to all of the genetic counselors who looked at this data. We know that somatic testing's not optimized to detect large deletions, duplications, rearrangements. Depending on the coverage, some intronic or splice types variants may not be reported, and so this really fits with that. But I think it was helpful to us to try to give some examples to our referring physicians about why the somatic testing does not necessarily replace the germline testing because they're not equivalent, they're not always looking for the same thing, and so there are pathogenic mutations that may be missed if we don't do both.


0:23:17.8 TS: So Kristen, these are the missed only then? 


0:23:21.2 Kristen: These are the missed, correct. These were germline variants that were identified that were not reported on the somatic test, exactly.


0:23:29.2 SC: The patient population again was all-comers? 


0:23:33.2 Kristen: So all patients who had had some sort of somatic test and then germline test that we could compare it to, correct. And it's across all tumor types...


0:23:43.0 SC: That's what I meant.


0:23:43.1 Kristen: The vast majority of patients were stage four metastatic disease, but we did... There were patients across all stages.


0:23:53.1 TS: Yeah, that's really interesting and like you said, completely... [chuckle] It shows nicely the need for copy number variation and going into introns and things like that.


0:24:02.5 Kristen: Absolutely, which we discussed, I feel like weekly at tumor board, because some physicians totally understand that and others are like, "Well, but I did this test. Isn't that good enough? We would have expected to find something if it was there." So this shows, again, about a quarter of the discordant results involved a germline pathogenic or likely pathogenic variant, so something that was actionable, that would have been reported when the germline testing was done. Overall, in our population of patients, this means that about 7% had a germline finding that would have been missed if we just relied on that tumor testing, about a third because the gene wasn't included on that semantic panel, and then about two-thirds because the variant was not detected.


0:24:47.0 Kristen: Again, looking at the genes that were involved, these were actionable findings. Some of these genes have treatment implications, certainly many of them have implications for screening prevention strategies within the family overall. And then this is just the tagline. Like I said, I feel this comes up every week, and we're still trying to convince people that tumor testing certainly has value, we are big proponents of that, but it's not a replacement for germline testing and there are things that may be missed, and so it is important to make sure we're doing both types of testing whenever we're concerned about hereditary cancer syndrome.


0:25:22.8 TS: Yeah, no, that's excellent. I don't know if anyone has any questions. Yeah, there's multiple reasons, missing is one, re-classifications is another big one, so. And not that all labs that do germline testing do re-classifications, so good to use labs that do re-classifications actively. But yeah, that's one that, yeah, I don't know of, honestly, any somatic lab that's doing re-classifications over time, so they're not really invested in the life cycle of that patient, is kind of the way I would tend to think about it. Yeah, and you pointed out the rearrangements, I mean complex things like Boland inversion and MSH2, the pseudogene and PMS2, these are all very difficult things that you have to have a lot of expertise. And then just variant calling is a big one. If you're calling things with germline, you're usually using ACMG type algorithm and if you're calling tumor, a lot of people might not know, it is a different calling through AMP criteria generally, AMP, which is also in conjunction with ACMG, but it's a different rule calling, so there's some nuances and, yeah, there's...


0:26:47.5 TS: And my mind's changed over the years on this, but I don't know. People may remember, when tumor testing first kind of hit, I would say there was this weird dichotomy of, "Well, tumor testing is totally different than germline." It was strange, it was like two years where people were kind of treating them as 100% different in variant classification, where even if something was known... A known TP53 mutation in Li-Fraumeni Syndrome, if it was found in the tumor, people would say, "Oh, we don't know what that means because it could have different connotations," things like that. And those worlds have definitely merged over the last 15 years or so since all this has been out. So now I think it's nice to see that, yeah, if things are hot spots, you see them in the germline, I think there's way more cohesiveness to variant calling in general between tumor and germline. No, that's really good work. So what are you gonna do with all of this? Are you planning to...


0:27:52.0 Kristen: That's a great question. Well, now we have all of our 2020 data compiled, so I'd like to put it together and update the numbers and hopefully publish the data. But I don't have any dedicated time for that, so it's tricky to squeeze it in between clinic and between meetings, but that would be the goal to get that information out there. Although, it's one of those things I think... I feel like we're moving ever closer to paired tumor-normal. That's just gonna become what we do ideally, so that we don't have to try to guess, what's in the germline? What's in the tumor? How does this all play together? So I feel like I need to do it fast or it's gonna be outdated because we'll just be doing paired testing on everybody, but...


0:28:33.8 TS: And that's a good segue, I'm gonna launch...


0:28:34.6 Kristen: Apparently it was interesting.


0:28:36.0 TS: Yeah, I'm gonna launch this poll. What do people tend to order? We're on our call today. Let's see. But yeah, I wholeheartedly agree with you. I think that's where the field is moving for a multitude of reasons. One, I think one trial that is gonna be very interesting that will likely be read out at ASCO is gonna be the Olympia Trial. I don't know if people are familiar with that, but it's looking at Lynparza for first line maintenance in breast cancer and it's looking at germline mutation carriers, and it's Lynparza verse placebo. Here, let me share these.


0:29:29.9 TS: So yeah, a lot of people do both, so that's good. But what that means is when you start looking for first line maintenance, it really says you need to know at least BRCA1 and 2 mutation status, and really at the time you're trying to make a decision on somebody with HER2-negative breast cancer, which is what the trial's on. So, which is 70% or 80% of patients. So we're kind of getting it into the world where we, just using that as an example, where I think we're gonna need to know germline status for, at least for breast cancer for 80% of people right off the bat, just from that trial, because they've already reported that the results look good. They just haven't actually shown the results yet. But I know it's all getting pulled together now. Well, good, so yeah, we have a lot of people doing both types of tests and... Can you walk us through a little bit at your institution, Kristen? Who does the testing and how it all works, the tumor and the germline? 


0:30:30.9 Kristen: Sure, so right now, the medical oncologists are ordering the tumor testing, but again, they just put in sort of an oncology genomic profile order through Epic, so it's all one centralized order-able that then goes to the precision medicine team and they actually place the order with the lab, coordinate with pathology. So I'm very removed, I don't order any of those tests, although I do review all the test results. And then we've got a department of, I believe we're up to 11 genetic counselors now. So the oncologist typically orders the somatic testing and then refers to the genetic counselors and we order the germline testing. We have a few doctors who are doing point-of care testing for specific tumor types. So for instance, our pancreatic cancer patients, the oncologist upfront will order both the somatic and the germline testing and then kinda loop us in on the back end if anything is found that they want us to discuss with the patient. I feel like that's another area where we're shifting and how our clinic models work and operate and trying to make sure that patients have access, are getting it done in a timely manner... With precision medicine, these patients are really sick and they need the information as soon as possible, so trying to streamline that as much as we can.


0:31:52.4 TS: Yeah, I don't know if anybody else is willing to kinda share how their institution is doing tumor or normal. Clearly, we have some people on the line that are doing both. And I'm seeing different models pop up and from everything, from surgeons or oncologists are ordering all the tumor, but the germline is still completely going through the more traditional genetic counseling pathways, to the providers, surgeons, oncologists, APPs are just ordering everything to... Kind of what you just described where there's a decision that something's needed and then it goes through some sort of specialized team that's doing the ordering... I think genetic counselors nationally right now are trying to figure out where they fall in the whole somatic space, and it really does vary by institution right now. And I don't know, are genetic counseling programs right now starting to teach more on the somatic side? 


0:33:01.4 Kristen: That's a great question. I don't know the answer, I'm not directly involved with the program that has curriculum about it, but it certainly would have been helpful. It was completely learn on the job for me.


0:33:14.0 TS: Yeah, yeah, no, for sure. Well, no, thank you so much for presenting, that was great. And feel free, yeah, you can keep chiming in and I'll show the last couple tables 'cause I think that really hits home on some of the concepts you were talking about too.


0:33:31.2 SC: Kristen, can I ask you my question? 


0:33:32.6 TS: Oh, sure.


0:33:34.1 SC: So, Kristen, if you had... You've been doing this in your institution, but if you had a way to optimize this in a way that pulled in all healthcare providers that are involved in the patients coming in at the right time, where would you plug the genetic counselor into this process... Or the advanced practice nurse who's talking to patients? Where's the ideal positioning for that since a lot of places are doing it differently, given you have some experience with that? 


0:34:05.1 Kristen: Yeah, it's a great question. What I kind of wish we could do is have a genetic counselor in each clinic and do more of that multi-disciplinary model that if the patient is coming to the breast MDC, the genetic counselor's on-site and can have that discussion and can place the order when everything else is going on at the same time. Or similarly, if the patient's coming in to see their oncologists, that if we're embedded in the clinics, we could help assist with that ordering and explaining to the patients at that point of care. Because I do think the Point-of-Care testing has huge advantages in reaching more patients. We've proven that in our institution, way more patients with pancreatic cancer are getting tested now than they were when the doctor just placed a referral and the patient had to make a separate appointment and come see us. And even if they did follow through, it delayed the process, but many of them didn't even get to us.


0:34:56.2 Kristen: So I do think it needs to happen more upfront. What's tricky, of course, for us, is that with such a huge system, we can't all be in all places at once, and so we're still sort of a separate department, but trying to figure out how to embed ourselves in those clinics or train the people that are in those clinics, like we've done with the pancreatic team about the appropriate patients to test, agree upon a panel, get them to set up to do that ordering upfront and then be available for questions, concerns, any issues that come up.


0:35:31.2 Michelle: Hey TJ it's Michelle. We run NOLS in Missoula. I have a question and also some comments. So I recently got referred to... I do mostly germline and I get a lot of referrals from Providence, and this particular patient was a uterine cancer patient, I can't remember her exact pathology, but she met criteria for testing... For germline testing. So I tested her and sent it off and then went to try to find the pathology report and there was never any tumor testing done. So my question is, since I've already tested her, do I need to go ahead and request like MSI and IHC? And I guess part of my frustration with this case is it never really got to oncology because the surgery was done by OB-GYN, so I just feel like it'd be a great place to start some kind of protocol that if that's the situation in pathology, that the pathologists could automatically order something like that, I'm just wondering if you have any comments and do I still need to do tumor testing on this patient? 


0:36:49.2 TS: Yeah, good question. So it sounds like probably a gynec or maybe a gynecologist did the surgery, probably that she was early stage, I'm assuming, or she would have ended up needing some chemotherapy, is that? 


0:37:01.9 Michelle: Yeah, I think she was stage one, but I can't quite remember...


0:37:06.1 TS: Yeah, which is pretty common for endometrial and maybe I missed it. Did she have a Lynch syndrome mutation or a germline mutation? 


0:37:15.2 Michelle: The results are pending.


0:37:16.9 TS: Oh, okay, they're pending now...


0:37:18.3 Michelle: I just tested her last week.


0:37:20.2 TS: Yeah, so I have some personal thoughts on this, not the views of Myriad, I should say, but I really have not been blown away that if germline testing is done looking for Lynch syndrome, that we're missing a bunch of cases of Lynch syndrome. So meaning that germline testing seems, in my mind, very adequate to diagnose Lynch syndrome and really as is a necessary thing to diagnose Lynch syndrome, in the sense that, I can't even think of an example... And I'm sure there's maybe some very rare case reports, but where someone had in today's era of EPCAM deletion and the four main Lynch genes... If someone did not have a mutation in one of those five genes, let's call it, still was diagnosed with Lynch syndrome. And I'm talking about a true mutation. I'm not talking about if there was a VUS or something like that that needs some exploration. That's kind of a different thing. There's gonna be some data coming out in that regard, probably over the next couple of months, that we've had a little early snippet access to, but it really does kind of confirm that same concept, that I really think that for colorectal and endometrial patients, if the question is whether they have Lynch syndrome to me, germline testing is a great way to answer that question.


0:38:49.7 TS: Now, so that kind of gets at your question of, does this person still need MSI or IHC after they have germline testing? So assuming that germline testing comes back negative, the likelihood that this person still has... Also, assuming there's not some Amsterdam I or II criteria or something that just really looks striking for Lynch, at least, yeah, I don't think there's a lot of utility at that point of doing MSI or IHC. I mean, unless it wasn't being considered to use for treatment because if it comes back normal, you're like, "Yeah, it's normal," and if it comes back showing MSI high or something, now you kinda get into this diagnostic odyssey of trying to explain that away when you already know the person doesn't have a germline mutation or maybe she'll come back with a germline mutation.


0:39:38.2 Michelle: Okay.


0:39:38.8 TS: It's kind of my two cents. There's not really guidelines around this, so I will also preface. So I don't know if others have thoughts there. But also when you do a multi-gene panel, now you actually open up the door. Yeah, does this person have PTEN, other things that could potentially also predispose to endometrial cancer? So it sounds like you're working it up right. And then getting back to your, I guess, your second part, the protocol, Michelle, I think a lot of people, a lot of centers are setting up protocols not only around colorectal, but around endometrial in some way shape or form. I mean, a lot of the bigger academic centers are just doing universal MSI and/or IHC for any colorectal and endometrial at this point. And you could argue that that's used not only for germline, but also for treatment considerations and things. I've seen different protocols for endometrial, sometimes people just do a under 50 or something. If someone needs treatment, obviously, especially for endometrial, knowing if somebody has high tumor mutational burden or MSI, these are really important things for chemotherapy selection in today's era.


0:40:58.2 Michelle: Okay, cool. Thanks.


0:41:00.6 TS: Yeah. And someone texted me, "Are germline mutations always a cancer driver?" So that's a great question. I don't know if anyone has any thoughts there. I would say no. [chuckle] And I don't know... Kristen, based on your data, what have you been seeing? 


0:41:25.1 Kristen: We talk about this a lot at Molecular Tumor Board when there is something that we suspect is germline. So it shows up and it's at a high variant frequency, but it's one of those things that, is it sort of an incidental finding that we're uncovering the germline mutation in the tumor, or is it actually relevant to the tumor itself? And I think our feeling too is that quite often it's not the driving mutation. It's just along for the ride, and we uncover it because we did the testing. One thing we've been talking a lot about, just maybe a little feedback for you guys, although I know this is not necessarily Myriad specific, but especially when that's something like a BRCA1 or BRCA2 mutation, we would really love to order HRD testing. Sometimes we find it on a non-ovarian tumor and kind of to try to see, does there seem to be that same phenotype that maybe would suggest there is some relevance to it within the tumor versus just being kind of an incidental germline finding. So that's come up.


0:42:28.9 TS: Yeah, you're saying if you see a somatic BRCA 1 or 2 mutation...


0:42:32.1 Kristen: Correct.


0:42:33.2 TS: Yeah, yeah...


[overlapping conversation]


0:42:34.4 Kristen: Yeah. To try to figure out how important it is, to know should we be trying to target it, should we not be trying to target it? So that's just come up a few times among our doctors kind of off the cuff, like, "I wish we could get further testing on this," but it's not always of the tumor type where we have access to HRD testing at this point.


0:42:54.4 TS: That's a good question. A couple thoughts there. The way our test actually works, if you have a BRCA1 or two mutation identified in the tumor, which in these cases you would, our HRD test then reports out as positive, period. And then there's the genomic instability score. Now, what I will say is, our threshold for ovarian cancer is really set to pick up about 95% of BRCA1 and 2 mutation carriers at the genomic instability score of 42. It should be going down, in some instances down to 33, which is actually the 1%, the first percentile threshold. So long story short is if you see a BRCA1 or 2 mutation in the tumor, yeah, the genomic instability score is probably just gonna be high at that point. So yeah, in that sense, it's almost... Across cancers, you would always just get a positive test in our [0:43:57.3] ____.


0:43:57.4 Kristen: What about something like a PALB2 or something else in the pathway, but not BRCA1 and 2? 


0:44:01.6 TS: Yeah, we are looking at all of that right now. We're also looking at it in non-ovarian tumors and we have some clinical trials in motion and different things across all kinds of different cancer types. And so, yeah, hopefully over time we can really expand out genomic instability across all cancers, would be the goal. There was actually some data presented at a SGO, which was interesting. I have it on my computer, but I probably shouldn't show it, but it looks good. And it was looking at gene... And we actually... It was using some of... Well, it was using our data through AstraZeneca studies, and it was looking at exactly what you just brought up with PALB2. So it actually looked at multiple... So the genomic instability scores across different mutations in the tumor. And it was interesting, RAD51C and D... There were only a few events, so they looked at about 800 tumor samples from ovarian cases, and I think most of these were from the 1st-line maintenance trial, PAOLA.


0:45:15.4 TS: And in that, they identified... Yeah. Rarely you see a homologous recombination, other gene mutation outside of BRCA1 and 2. When you see it, it's kind of... It was kind of a toss-up. RAD51C and D actually looked like... Tended to go with a pretty high genomic instability score. Bloom, interestingly, BLM, had a very high genomic instability score, which... I'll come back to that in a second. PALB2, I don't think there were really enough cases to make any inference. But then a lot of genes like ATM and CHEK2, they actually had really low scores, and that's probably because those are mutations in the germline, a lot of times. What we just saw with that, Steve Lincoln paper, where it's more likely that if you find a ATM mutation or CHEK2 mutation in the tumor that it's probably gonna be germline in that case, and it's not really a tumor driver, getting back to this whole thing about, "What is a cancer driver, and how do you determine that?" Bloom, getting back to Bloom then, things like that, and even RAD51C and D, and... A lot of these, they're not... Bloom as a case example, that we don't tend to think of it as a HRR gene, but actually, if you have autosomal recessive Bloom syndrome, it is a radio sensitivity syndrome and actually you use chromosome breakage studies to help with the diagnosis, like fanconi anemia.


0:46:47.4 TS: And so there's a huge amount of genomic instability, and my gut tells me that the case... There was only two or three, but it was off the charts on the genomic instability, and I bet you in those cases that the other allele was also broken. And so there was probably no working Bloom in that tumor, and hence you just had a massive amount of genomic instability. So I think... And that was just literally presented a couple weeks ago, so I think there's a lot going on to better understand how genomic instability or scarring and tumors and PARP inhibitors relates to single-gene mutations. I am a fan overall of knowing that genomic scarring, is it present or not, like the genomic instability, just because I think it gives you more information than just a... Well, I know it does, we've even shown some data on this. It certainly gives you more information to have the genomic instability scoring if you're trying to make that determination of PARP inhibitor therapy in particular, over the HRR gene mutations, like if you took 15 or 20 HRR genes, just because your catchment, your net is just so much bigger to look at, you're looking at the end result, you're looking at the consequence of dysfunctional homologous recombination, and so instead of trying to figure out the cause at that point.


0:48:10.4 TS: So when you're looking at the consequence, you're pulling in everything from promoter mutations to, who knows what, any... Multifactorial reasons, the homologous recombination pathway is not working and everything in between. A good question. Let me show... I know we only have a few more minutes. I do wanna show the last couple little tables. Can everybody see this screenshot. So yeah, 'cause this is a nice paper and they really did walk through a lot of different things. This is kind of more general, if you look at the cancers that came in, colorectal, breast, lung, these were some of the more common cancers that people are sending in the TRF saying, "Yeah, this person had a mutation identified. Is it in the germline?" And then when you run the multi-gene panel, you see, this is more kind of standard epidemiology genetics, like what kind of distribution of mutations do you see across, so it's not really specific to finding the mutation that was noted on a TRF, although, a lot of times that was found, which is why this number is so high, 'cause these numbers are insanely high, if you think about it. Most of the time, colorectal cancer only has about a 10%.


0:49:33.9 TS: If you took all people with colorectal cancer, you'd only expect germline mutations in around 10%. To be at 25%, it's a little off-putting, but you have to remember how these people were brought into the study. They already had a known something or other found in their tumor, which gave them an A-priority risk to have a higher likelihood to find it in the germline. So that was interesting.


0:49:56.8 TS: And then this is... Yeah, getting back to Kristen, what you were also talking about, the reasons for the difference when you're looking at tumor testing. I thought this was a nice table and a little scary with the unknown reason here. So a lot of times, yeah, the gene on the germline panel just wasn't in the tumor test. That gets at... Yeah, that... If you're testing somebody for hereditary cancer, personally, yeah, use a hereditary cancer test. If your question is, "Does this person have a hereditary cancer mutation?" Don't try to answer that question with a somatic test. Try to answer it with a hereditary test because a lot of times they're just not on the assays. And even from some of my work in the past with Guardant, when we had our JCO paper and doing some of that work, they have...


0:50:50.2 TS: Guardant360 is just a perfect example. I don't know if they've changed it since, but we had a paper looking at the mutation spectrum of punitive germline mutations in the background of Guardant360. So as standard Guardant testing, which is cell-free DNA analysis, at 2% to 3%, you can see the circulating tumor mutations. But they came to, myself and Jeff Whitesell and some of us at City of Hope and said, "Can you help us sort some of this other stuff out, 'cause we're seeing all the stuff that's clearly germline, it's floating around 40% to 60% allele fraction?" And when we kinda dug into that test for Lynch syndrome, for instance, all they do is like MSH2 exon 12, so all the other... If your question is, "Does this person with colorectal cancer have Lynch syndrome?" and your only assay on that person is a Guardant360 test, unless they change something... This was circa 2017, 2018, then that was not a good test to try to even figure out...


0:51:55.4 TS: And even back then, they weren't reporting anything about germline results, and I mean... I know since that paper, they're at least doing BRCA1 and 2 and maybe they're alluding to germline results in other variants. But yeah, you can see how the breakdown of knowing what test, ordering the right test for the right situation, I guess is what I'm getting at. And then, yeah, getting to your other main point about the copy number variation, all the other nuances of the test, I mean that's all really summed up nicely here. Variant interpretation difference, that's a big one as we kind of already got into. And then unknown reason, which is a bit scary, and that could be for a lot of different reasons. When you have a germline mutation, especially if it's not driving the cancer itself... Cancer can be very disorganized, meaning that it does not need to have 46 chromosomes.


0:52:55.3 TS: It can have 80 chromosomes, it can have 38 chromosomes, it can have anything in between. And because of that potential for allele dropout, you could always potentially drop out alleles that do have the germline mutation. Again, that would be theoretically more in situations where you don't have a driver or where it's not... Where that germline mutation is not driving a tumor, because you would think, "If the germline mutation's driving a tumor, it should be in the tumor." So it's always been a concern and that might filter into some of these unknown reasons in my mind, so... Yeah, plenty to learn, this is a hot field, so definitely keep exploring it, Kristen. [chuckle] So any other questions? I don't know if there's anything else on the chat, Shelley? 


0:53:47.6 SC: Nope, the chat's been pretty quiet today.


0:53:48.1 TS: Yeah, no, yeah, we had some good discussion, and I just wanna say thank you so much, Kristen, for coming out and showing some of your data. Very exciting, keep working on it. Keep doing the good work that you're doing.


0:54:02.9 Kristen: Thank you. Thanks for inviting me.


0:54:04.5 TS: Yeah.


0:54:05.6 Kristen: This was really, really interesting. I did not know these talks happen, so I appreciate the invite.


0:54:09.0 TS: Yeah, yeah, and so if you ever wanna come to some other ones, Myriad Oncology Live, you can just Google it and you'll go right to it. And then we put out all the talks on that. May 4th is the next one. Rob is on, and I don't know, Rob, feel free unmute yourself if you're listening and give a little teaser if you want. Sorry, I put you on the spot.


0:54:35.1 Rob: That's okay. You kind of said it, we're just gonna talk about how hereditary cancer fits into the LGBTQ+ community and so we're gonna have a couple of guest speakers as well. So it should be an interesting talk.


0:54:49.1 TS: Yeah, that'd be great. And then, yeah, we have a lot of new topics coming, and again, building out through I think, early July, actually. So I'll get those posted. I made a note to myself, hopefully, we can get those up very soon so. And thanks Shelley as always for running the chat. But yeah, I hope everybody has a nice rest of their day.