Myriad Live - Let's Talk HRD

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 myriad-oncology.com/myriad-oncology-live for a list of dates, times, and subjects.

Transcript

[music]

0:00:11.5 Dr. 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 do 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.

0:00:40.5 DS: Hello everyone, welcome to Myriad Oncology Live. It's a great day and I'll stop sharing my screen, I do wanna share it back, [chuckle] so I can show a little bit of housekeeping. But thank you for coming on, if you haven't been on before, welcome. Congratulations for making it. And if you have any questions whatsoever, you can ask them, this is a completely open forum, so we're here for you. If anything comes up, just feel free to answer. These are theme-based. Today's theme is on HRD. I do, before I forget, because I've been forgetting pretty much every week consistently, I do wanna say that these will be recorded because we're recording them. We've had just a lot of interest to try to get them up in some fashion.

0:01:29.9 DS: I do have more clarity, though on the recordings this week, so it looks like we're gonna just do them as a podcast, just audio only, so these are just gonna be as part of the Inside the GENOME podcast ultimately which you can find at the bottom of the regular Myriad Oncology Live page. So this is a podcast that I started when I came to Myriad. Shelly has helped a lot with this podcast as well. It's a good podcast in general. "Hey", I would say, "Go look at it, enjoy." They're about 15-20 minute episodes with... We usually try to get a national expert, advocacy group, had a lot of interesting ones here. We have a new one actually probably being posted today with Annie Parker, so she was one of the first BRCA1 carriers tested in the United States. She had a movie after her, Annie Parker Decoded, which you can find on pretty much most streaming services, so really interesting. So that one should be up today, hopefully.

0:02:38.0 DS: And then so we're just gonna put these in here, so I think we're gonna title them something differently like MOL or Myriad Oncology Live or something, but at least it'll be here in case there's some content that you wanna go back to, yeah. We tend not to show a lot of slides in general, but there might be a way we can tack them along or something, so we just haven't quite sorted everything out, but hopefully you'll feel comfortable still asking questions. If not, you can always send any question to Shelly, who is running the chat. Thank you, Shelly. And today, yes, we're talking about HRD. We had a really good one last week, so... Yeah, and actually, I think most of these are gonna be put up at one time, so a lot of the recent one should be all up in a few weeks, but yeah, these last two have been very good.

0:03:34.6 DS: We had Dr. Eduardo Vilar last week, who's at MD Anderson and he was talking about his research with chemo prevention and Lynch syndrome. We also discussed a bit about vaccine development to prevention of Lynch syndrome, really interesting stuff. Today's HRD, next week, we're gonna do a 6:00 PM So, for our Eastern colleagues, hopefully post-clinic and keeping kind of a broad topic, just to make sure everyone's questions are answered in the space, so again, you can bring up anything you want. If you feel like you wanna ask about PALB2 carrier risk today, that's fine, but the main theme is HRD. And special guest today is Diana Turco. Diana, I don't know if you wanna introduce yourself.

0:04:26.6 Diana Turco: Sure, my name is Diana Turco, I'm a genetic counselor and a medical science liaison here at Myriad, and I work specifically with companion diagnostics and our tumor testing, so I'm here to ask questions, answer questions, help TJ out.

0:04:45.1 DS: Yeah, yeah, that's good. So I had to start. I had Diana Mayo. [chuckle] So I think you're trying to see if some investigators were going to be able to present some data. I don't know if you got the okay to show some of the data.

0:05:03.1 DT: I don't have their data, but I do have a different institution. So one big question that comes with HRD and it being a tumor test, is how do you incorporate it into your clinic and how do you handle germline testing as well, because I feel like I should take one step backwards and say, whenever we're talking about HRD, I default into talking about ovarian cancer, that's where it's been studied the most, that's where we're looking at it, and that's where it has the FDA approvals, so we can talk about other tumor types later too, but right now we're talking about ovarian and all women with ovarian should at least be offered, if not perform germline testing, so how do you do the germline test and an HRD test for those patients so that nobody falls through the cracks? And historically, I know some clinics are much better than others, but germline testing... I think it was January of this year, there was an article in JCO saying that the national average was somewhere around 30-33% of women were doing germline testing with ovarian cancer, so not a great number. And even clinics that are doing really well are not anywhere close to a 100%, so.

0:06:22.3 DS: Yeah, I think that was Allison Kurian kind of follow-up paper from the... I don't know if that was the California Georgia linkage through SEER or the larger... I think it was The California-Georgia linkage, so.

0:06:34.8 Shelly: It was, and it just hit JCO today that showed it was 34%, which is up a whole whopping 2% from the last... Last time, so, we have ways to go.

0:06:51.2 DS: Yeah, for sure. And for those that don't know, I mean, we have been collaborating this effort too. So, it's essentially us, [0:07:00.5] ____, Amber A and Jean Deaux, contributing data to California-Georgia linkage, SEER linkage, so it's everyone that has cancer in those states has their hereditary testing data linked. We're looking at expanding it, it's a really exciting project and really incredible data has been coming out of it. So if you Google Kurian, Allison Kurian, who's at Stanford, you'll see a lot of her recent papers are using this massive... I mean, this is definitely mega-data, at its finest, so it's really a large aggregate data sets and you're able to pull some pretty nice findings out of it.

0:07:48.7 DT: Yeah, so after the germline are trying to figure out how to incorporate that, it's, who's initiating it? Is it starting with a genetic counselor, is it starting with the Gyn-onc team or what's the best way to do it? And, I think the answer is, whatever works for your clinic. [chuckle] So I live in Minnesota, I've talked to Mayo Clinic a lot, they weren't able to join us today but they went through a whole quality improvement process and hopefully in the future we can have them join at a time that's good for them to kinda walk through what they did. But they decided to have their women with ovarian cancer go see a genetic counselor and if that germline test was negative, then they'd come back to the Gyn-onc team with that result so that then, a my choice or an HRD test could be ordered to determine if they would be eligible for PARP or if they would preferentially benefit from a PARP inhibitor. So that's what works for their clinic.

0:08:52.9 DT: Another clinic that I've talked to a lot is Duke, and so I can pull up a poster from Duke because they did... They presented this at SGO this year, so I'll zoom in when we talk about the different parts. But this was from Katie Watson, a Gyn-onc at Duke and Doctor Secord, and what they have been doing is that they've done HRD and germline testing for several years. And part way through, or about a year and a half, two years ago, they implemented a process, and I think this poster is really nice, it makes sense to a lot of us, like hey, if I put a checklist in place or if I put a process in place, I hope that I'm gonna improve. And that's what they were able to show. But there was some benefits that came along with it that I'm not sure they expected. So, they looked at pre-implementation of the process and post, and what they found was that... I'm gonna zoom in here. So, 74% of women before the process were getting germline testing. After the process, there was 89%. So, the main point of this process was to focus on their somatic testing, but their germline benefited too. And so, Duke is doing a great job, they're clearly way above the national average for germline ovarian.

0:10:26.6 DT: And somatic testing went way up from 39% of women to 76. Some of this was part of timing and FDA approvals happened for more PARP inhibitors, and so that was the very natural increase, but also a nice way to see that they were helping out the women that should be offered testing. One other thing that helped was that they decreased the time, the turnaround time. So their previous timing was 6.2 months for the turnaround time of germline and it decreased to 3.7. And some of that is also because of PARP inhibitor approvals. They used to be late line, and now they're front line maintenance. And so, the genetics team worked really closely with the Gyn-onc team, just saying, "Hey, you guys need this earlier," so they improved or changed their process so that they could have that information for the Gyn-onc team to make that treatment decision. So it was a really nice, just picture of working together.

0:11:42.6 DT: And then the final part that I like to see is that, down at the very very bottom, the genomic instability, unable to analyze, was at 16% and it dropped to 7.5%. And being unable to analyze is a very common thing with tumors. Sometimes, in the past, I think I saw Kirsten on... We've gotten just a blank block of paraffin, so we can't analyze that. Or it could be that when we're in the recurrent setting, maybe they've already used pieces of tumor and we're getting kind of the ends of it, or maybe the edge or something that's not quite as easily tested, but by being able to move it up front because of their treatment decisions and just work together with pathology and genetics and Gyn-onc, they decreased that unable to analyze, just by putting that process in place. So, that was nice to see.

0:12:43.6 DT: And then, another thing that we see throughout these results is that they're identifying women... The golden goose, sorry I can't think of a better term, is to find a woman with a BRCA mutation. That's the golden ticket. That's what I meant, [chuckle] to get access to any PARP inhibitor, any line of therapy. But they were finding women with somatic BRCA mutations that weren't present in the germline, they were finding women that were somatic and germline BRCA negative, but genomic instability positive, which is still unlabeled and we know that those women will preferentially benefit from PARP. So, having that extra piece of information was great for their patient care.

0:13:34.6 DT: And the last thing, I'll say here is too, is that Mayo clinic do a lot of big institutions that are doing these dual processes, they have a lot of genetic counselors, so it's great to be able to call up and get one of a dozen of a team of genetic counselors, there are other institutions that might just have one that's doing all cancer types, they're doing breast, GYN, GI, so to ask them to rush a GYN or ovarian patient and add them on is not a feasible process, and so I've seen other institutions work together with their genetics team to say, "Hello, how can we help educate any of the PAs or NPs that sit in GYN to help order that testing, and if it's positive or if that patient has a strange family history or whatever else, then call in the genetics team." So more of a post-test counseling scenario, but then the GYN team is still getting the information that they need for treatment in that first line setting.

0:14:47.3 DS: Yeah, no, thank you.

0:14:48.3 DT: Yeah.

0:14:48.7 DS: Diana. Unfortunately, I don't have my polling feature, I don't know, we set up a new round of invites, so I lost that. I'm gonna have to figure out what happened to it, but usually we have a largely genetic counselor biased audience, and I was just wondering to mean if anyone wants to share their experience of how this is being handled at their institution? I think it would be helpful for everyone. Probably a lot of different models, whether it's, you know again, we're talking here and largely ovarian just because that's where HRD probably most Jews, but you know, it depends on the terminology of HRD because you know, there's kind of the cause and the consequence, and when it comes to the cause, we're now using PARP inhibitors in you know, all kinds of cancer, so it's becoming much more common place in pancreatic and prostate, you're seeing it in breast, we're gonna maybe even as of tomorrow, actually be able to look at some data that's gonna be presented to ASCO, on Olympia.

0:15:57.6 DS: So everybody is really waiting with baited breath for that, so that's looking at PARP inhibitors for first line breast cancer and her to negative breast cancer. So that could really change the dynamic because that would move the need to know if you have a BRCA1 or two germline mutation at least to upfront, you know, right after your initial chemotherapy, even in breast cancer. So first line maintenance. So that's really the cause, the consequence before I forget, Shelly and I'll come back...

0:16:30.5 Shelly: Go ahead.

0:16:31.5 DS: But the consequence is really that genomic scarring you know, kind of assay, you know, looking at loss of heterozygosity or telomeric allelic imbalance or large state transitions and things. So, yeah, go ahead Shelly sorry.

0:16:42.6 Shelly: Yeah, sorry to interrupt you. I think there was an interesting poster from SGO that Diana can give a high picture overview of the differences with the mutations and the HRD pathway and how they don't all act the same, which has consequences for the decisions that are made. Diana do you wanna pull that up? 

0:17:10.1 DT: Yeah, so can you see it sort of...

[overlapping conversation]

0:17:15.6 DT: Yeah, So like TJ mentioned that catch phrase of cause versus consequence, and so we think about the causes of HR deficiency, of BRCA mutation or other gene mutations, BRCA promoter, methylation. And then we think about the consequences, you know, example is the Mayo choice genomic instability score or LOH, the different biomarkers that are looking at the downstream effects of HR deficiency, and so the question has been in the field, what should you do? How do you measure HRD? And if... The joke has been, if you ask five people, you'll get 10 definition, but this was a really nice piece from SGO it was presented, this is data from the PAOLA trial and AstraZeneca presented this, so you can see on the bottom or the list of a bunch of different genes in the HR pathway, and then on the y-axis are the genomic instability scores for those patients.

0:18:19.9 DT: And everything that's highlighted and pink, the average of those scores was above the cut-off of 42, which is the cut off for positive versus negative, and the BRCA is not represented on here, but it goes along with what a lot of people were suspicious of that RAD51C and D, BRIP1 and PALB2 were all going to be big players in the HR pathway, but all the ones on the right, we're also a bit what we were suspicious of, that CHEK2 ATM, BARD1, NBN, those are not good targets for PARP therapy, these women in the trial, from what we know of the PAOLA trial is if they had a low genomic instability score, they did not benefit from the addition of PARP... Of Olaparib [0:19:12.9] ____ and the women with high or HR deficient scores. They did benefit from Olaparib. And so we really have felt the field shift a bit from about a year ago, I was thinking, well, what if we looked at each individual cause, and it seems over the last year, it's really swung to say, look at the consequence.

0:19:40.0 DS: Yeah, and I would still say different in different cancer types. I mean, prostate right now, looking at the profound FDA label, it really is just for the HRR genes, and it's a laundry list, it's probably this list actually verbatim.

0:19:55.8 DT: Yeah. [chuckle]

0:19:57.8 DS: And it was interesting because if you go back and look at that paper, the data doesn't look extremely exciting, and so it's nice to see this was such a beautiful representation because you can really see the individual genes. I think this is very needy, and they really actually, in this talk at SGO, if anyone listens to it, they really didn't get into I think as much as they even could have. And one thing that looks really interesting to me, and others feel free to chime in, and Kirsten's looked through some of this as well, so if any other thoughts, definitely speak up, but I was talking with Kirsten and Jerry, our Gene Scientific Officer, and when you look at Bloom... If you could, I don't know if you could go back to that. Thanks, Diana. But if you look at Bloom, it's like, "Well, what's going on there?" I mean, that's... Bloom, and really actually Bloom is not part of the top of my head of even that profound label.

0:20:54.6 DT: I don't think so.

0:20:55.8 DS: No. So it's interesting that it's here and look how high the genomic instability is, it's off the chart, so that dotted line, anything above that dotted line would be positive, and this is... They used our assay here, so this is 42 and above is positive in ovarian cancer and 42 and down is really negative on our test, but if you look at the Bloom, that's really interesting, but here's what I think is actually going on with some of these and arguably, maybe even BRIP1, and even potentially RAD51 C and D and PALB2. It may be that there's two hits going on, so they didn't elaborate into that, but I bet you, if you really dug in for Bloom, those cases, even though it's only two cases, I bet you both those cases have both alleles knocked out because Bloom is a well-known radio insensitivity. Bloom syndrome is a well-known radio insensitivity syndrome.

0:21:50.2 DS: If you remember from genetics boards, kids get it, they can get all kinds of different cancers, leukemia is really a big hallmark. Absolutely cannot give them radiation. But if you remember your genetics, the Fanconi anemia and the radio insensitivity syndromes, they do have a hallmark of genomic instability, and actually one older assay that we used to use was chromosome breakage studies. And so, there's increased chromosome breakage in children with Bloom, similar to Fanconi anemia. So that's why I just think it all fits that those are probably two alleles not working correctly, and yeah, probably same thing for all the ones in red. And PALB2 has really become... I mean, outside of BRCA one and two, the studies are really starting to look pretty strong for PALB2, as a definite gene that maybe conducive to PARP inhibitor, treatment with a PARP inhibitor. There was a recent paper as well from Judy Garber's group...

0:23:03.1 Shelly: Nadine Tung.

0:23:05.2 DS: And Nadine Tung, yeah. Looking at that, and they had a nice waterfall plot in it, showing that the PALB2 carriers actually had pretty good response, objective response.

0:23:14.9 DT: But I... This is, I call it my cynical aspect of...

0:23:20.0 DS: Yeah, yeah. [chuckle]

0:23:20.9 DT: I think it's great. I think in that paper, there was maybe only eight PALB2 carriers, or maybe it was 50 I could be mixing up, but it was small, and I just don't see people breaking down clinical trials to say, "Let's do a whole clinical trial focused on just RAD51 or just PALB2," when they know that they can cast a wider net instead of looking at causes, they can cast a wider net and look at consequences.

0:23:53.6 DS: Right, right.

0:23:54.2 DT: Maybe we will get smaller studies like this or that give us a little bit of a breakdown, but I don't see any major trials running off of the causes anymore.

0:24:06.7 DS: Yeah, no, I agree. I think, yeah, you need to know what's going on with the consequence and what's interesting, ATM there on the far right, those may just be people with germline mutations in ATM or something. I mean, ATM is not a generally well-thought-of driver for ovarian cancer, it's not like people with ATM mutations are necessarily higher risk for getting... Germline ATM mutation carriers are not at a higher risk of getting ovarian cancer, at least based on current data, thoughts, and we have a decent amount. So those might... They may aid... We know ATM germline mutation carriers are common in the population, so you would expect, looking at 800 or so, that you're gonna have a few ATM carriers in there, and they're probably heterozygous, so they probably still have some working copy of ATM and I think that's why that's not really up a similar maybe with CHEK2.

0:24:57.3 DT: Yeah.

0:25:00.3 DS: But lots of study. It's very exciting.

0:25:02.0 DT: Yeah.

0:25:02.9 DS: Yeah, well, thank you for showing that. So any questions on HRD or... Yeah, how to implement it, how people are implementing in clinic or who's doing the testing or whatever. Just take a break for any questions. Anything else you wanted to add, Kirsten? On any of that data. Thanks for coming on.

0:25:25.6 DS: No, no, no additional comments.

0:25:29.1 Shelly: So I had a follow-up comment... Question, so Diana and TJ, you just talked about how all of the genes act differently in response, and I wondered if there's different efficacy for PARP inhibitors, or are all PARPs the same? 

0:25:48.2 DS: I don't know, Diana, if you wanna... Do you wanna take that one? 

0:25:51.0 DT: Sure, I know TJ and I have talked about this in the past, there's this really... In depth article from Science, that goes through each individual PARP and the mechanisms, but the short answer is no, they are not all the same. One of the biggest reasons that I understand, so I like to break it down a bit easier than the science paper... [chuckle]

0:26:18.5 DS: I can... I don't think I can't even understand that paper well.

0:26:22.6 DT: But the PARPs are, or these drugs have two mechanisms, so they can be a PARP inhibitor, that's the term that we use the most, where they're inhibiting the ability of PARP to call in all the right players to repair the DNA damage or it can be a trapper. So, most drugs are pretty equivalent in their ability to inhibit, but they can trap at different strengths, like how strongly does it hold it on to the DNA, and some drugs may be weaker trappers, and that's not always a bad thing, maybe that means that you compare it with an immunotherapy or a chemotherapy, and some drugs may be extremely strong or potent trappers, which is not always a good thing because it may bring along different toxicities, so there's strengths or there's good and bad to depending on how strong of a trapper it is, but they are all different.

0:27:27.4 DT: And I say that too, because... Oh Shelly found it, yeah... It only opens... Open at your own risk, it's real, it's real lengthy... But I say that because I think it's really important to look at the clinical trials and see what PARP was used and to follow that trial, because as we move forward, we've gone on a long time for PARP solo therapy or mono-therapies, and we're starting to move into a lot of trials that will be reporting, pairing it with [0:28:04.3] ____, immunotherapies, chemotherapies that you don't wanna pair the wrong PARP with that drug because of the differences between them.

0:28:16.2 DS: Right, yeah, it's gonna be an interesting future and I don't know what's going on with certain drugs like the LIPRAB right now that some people may use... It seems like that drug in particular, because to my understanding, the PARP trapping people are able to use it a little bit more with the chemotherapies, whereas some of the other ones are just much harder, like Niraparib and Olaparib, they seem to just not do well with other chemotherapies because of increased side effects and things, if you're combining them.

0:28:58.2 DT: Yeah, so I think that's a good way to evaluate a clinical trial too of... If it didn't... If it seems like a failed trial, and actually I liked that when I've been at meetings and they talk about failed trials, it's nice to know what didn't work out well, instead of always just seeing the only successes, but some of the failed PARP trials may have been with a drug that ended up not working out. I think a long time ago, and I might be messing up the name, but there was a drug called a Niraparib or something similar to that, and it never came to fruition...

0:29:34.1 Shelly: Yeah, the Niraparib...

0:29:36.0 DT: Yeah, and it seems like Niraparib is a little bit on that same track, some of the trials that haven't done as well, we're with the Niraparib. Yeah, Kirsten knows much better than I do.

0:29:49.1 DS: Yeah, and Kirsten... Just to introduce Kirsten Tim, so she's our Lab Director for myChoice assay so she's the one that is really the brains behind everything from the cause and the consequence testing. So if you have any lab detailed questions, she's a good person to ask. She might be having some issues at the moment with their audio though.

0:30:19.0 Kirsten: Yeah, my Internet's not good.

0:30:21.5 DS: It's okay. Well, good. And so I wanted to bring people up to speed thinking about that consequences and the causes, there's actually a current effort just to put on everyone's radar, I mean there's this group called Friends of Cancer Research, and they really do try to harmonize to the extent possible, different types of assays. So we're working with them right now. It's a big project to see if we can kinda compare and contrast assays a little bit, share more light on the field, so providers are more knowledgeable about all this, and part of the project, honestly, was just getting good definitions out in the literature, I mean even just the cause in the consequence language, I mean that's not common place right now, but understanding what genes are part of HR deficiency in general, pathways and things... This is not easy to find when you go to look it up. So I think ultimately, we just finished Phase One with them and there will be a... They apparently said there's gonna be posting a white paper that has a lot of this information on it, so it's gonna be a good resource if you wanna learn a little bit more about HRD and what's kind of the current cutting edge... What's available in the marketplace, what's being used on the research side, it'll be a really nice resource.

0:31:50.3 Kirsten: TJ, a few minutes on that one slide, you mentioned GIS, could you define that and then how is that related to LOH? 

0:32:02.0 DS: Yeah. So GIS is, I guess to start backwards, so lots of heterozygosity is just when you have lost at a specific site, so a lot of people think of it, it's kind of a confusing term... It's funny, in my genetic training, I always had trouble with that term, and at some point I just... It became a common place to me, and now I throw it around just like people who used to throw it around on the pathology side in the past, so it's good to bring it up, but...

0:32:36.2 DS: It's really... People may be familiar with it in the sense of, if you think of someone with this germline mutation carrier, for instance, like a BRCA1 germline mutation carrier who gets ovarian cancer or breast cancer, often if you look at the second allele, 'cause remember we have two alleles and BRCA1 lives on chromosome 13, so in that area on chromosome 13 where BRCA1 lives, you have two alleles... One came from your mom, one came from your dad. If you have a mutation in one and you develop cancer, often times you know really the vast majority of times that second allele is missing. And that's called loss of heterozygosity, and it's a weird term and it comes out of doing more PCR analysis.

0:33:25.9 DS: Kirsten, I don't know if you remember the actual way it came about, but really looking at... I'm trying to blank... Like gel electrophoresis, I guess, is kind of where that term comes from. So looking at the protein bands and there would be times where they would see a loss of a band and they would say, "Oh, this person has loss of heterozygosity," meaning that one of their parental bands was missing, so it's an old term that stuck around today, but it's really like a deletion on the second allele, if you wanna think of it that way. But it can also encompass across the genome too, so when we say LOH and... Kirsten, I don't know, if your Internet is okay, maybe you could explain a little bit about how we measure LOH and LST and TAI. I think that could be useful.

0:34:17.8 Kirsten: Sure. So I hope you can hear me. So the genome [0:34:23.8] ____ we calculate... There's three components to it. One of them is an assessment of genome-wide LOH, and the way that we assess it is by looking for LOH regions of a certain size and then counting the number of regions that we see. So as long as they are above a certain size, they just count as one, it doesn't matter if they're 20 Megabases long or 200 Megabases long.

0:34:50.7 DS: A Megabase is a million bases.

0:34:51.8 Kirsten: And the score literally... Yeah, sorry.

0:34:55.1 DS: No, it's okay.

0:34:56.5 Kirsten: Yeah, so the sign is irrelevant, or the link is irrelevant in our counting system. And then we combine that with two additional scores to give our final GI score and genomic instability is just a name to describe the mutation phenotype that we see in tumors with defects in the homologous recombination pathway, and basically they are tumors that have accumulated damage due to double-strand DNA breaks, so we see amplifications and deletions of regions of the genome and sort of a shuffling around of the genome.

0:35:36.2 DS: Yeah. And then the other two components, the telomeric allelic imbalance, that's kind of verbatim what it means, just that the telomeres are out of whack, and so the alleles are imbalanced. So telomeres are the ends of the chromosomes, and then the large state transitions, those are... To my understanding, I mean just kind of like large fusion events across the genome where something's kind of chopped out and fused, but feel free to correct me or maybe go in to more detail.

0:36:05.9 Kirsten: Yeah, LST is actually kind of complicated. It's actually hard to explain, in a way. It's any... We count any break points, so rather than counting regions, we're counting break points now, so any break point or junction between two regions of the genome where there's any kind of change in a state. So there could be a change in copy number or a change of non-LOH to LOH, so any kind of rearrangement, each time we see a break point between two adjacent regions. And they could be normal to abnormal, like two copies to four copies, or it could be abnormal to abnormal, three copies to four copies. That's what LST is.

0:36:48.6 DS: Yeah, so our test does all three of those, and then, yeah, it gives a combined score.

0:36:57.7 Ralph: TJ, can I ask questions regarding this compliance score? 

0:37:02.0 DS: Sure.

0:37:03.4 Ralph: What is the weight of these three biomarkers? Are they equally weighted in this score or... Is one of the three more dominant? 

0:37:14.5 Kirsten: So, I'm gonna leap in and answer this one, if that's okay TJ.

0:37:18.7 DS: Oh yeah, I was gonna defer to you.

0:37:21.7 Kirsten: So if you look at the equation, they would appear to be equally weighted because all we do is add them together, but in reality, the LST score has a dynamic range that's about double the other two scores, so that means it ends up contributing more to the final score than either LOH or TAI because it tends to be a high number than the other two, and so where we actually end up is that about half of the final score comes from LST, and about a quarter from LOH and TAI, on average. And part of the reason that we were happy with that final set up and found that it performed the best is that in our early studies we found that LST was actually the best predictor of HR deficiency. So it is more informative than either LOH or TAI and it makes up half of the final score.

0:38:28.0 DS: And then the score is normalized to 100 then, from zero to 100...

0:38:33.6 Kirsten: No, we don't actually do any normalization, but what...

0:38:36.9 DS: So... That's literally a count. So like LST might have 22 events or something like that with... Oh, interesting.

0:38:43.6 Kirsten: Yeah, yeah, there's no normalization of the score.

0:38:47.0 DS: Okay. Yeah... So it could go over... Well, over 100, I mean it could go to whatever? 

0:38:53.2 Kirsten: Well, and when she ran some simulations at one point because we had the size limitations on the regions that get counted and don't counted, and as you get to really unstable genomes and more damages accumulating, you can get new damage, like it turns a region that was counted into a region, that's not counted. So we find, in theory, you could get as high as 300 in practice, we've never seen anything... I think the high school we have seen was 105. And that's based on probably 10,000 to 15,000 genomes at this point.

0:39:27.6 DS: Yeah, no, interesting. I wanted to bring up... Just with our remaining time, I went to talking about reversion because there's been a little bit of coming out there, and so for people that don't know, a reversion mutation is someone that has... It's a pretty broad spectrum, but it's tend to be used currently in BRCA1 and 2 primarily, but it would be someone with a mutation in, for instance, BRCA1 that gets another mutation just randomly in the tumor or some subset of cells that actually essentially fixes the ability of the protein to be made correctly from the BRCA1 in that example.

0:40:20.1 DS: And this is gonna be an interesting field to keep tabs on, it's been well known now, I would, for about I mean jeez at least five years, if not seven years or so, that people have known this is going on... I'm gonna share my screen. There was a recent meta-analysis that came out that I think is interesting, if people wanna take a peak, it's pretty meaty, but... And I'll show the... This is meta-analysis of reversion mutations in BRCA genes, it identify signatures of DNA and joining, repair mechanisms driving therapy, resistance [0:41:00.8] ____ is the first author, and it's an Annals of Oncology, we should... We should start probably reading these assuming they'll go into podcast form at some point.

0:41:11.2 DS: But the interesting thing was most of these were deletion, so when you actually look at what's going on, there are these small little deletions that happen around the mutation, and you know I thought that was an interesting part of this, so even though they're random events, they tend to be kind of these one to five or whatever over one base pair deletions being the primary event, they rarely affect the actual original mutation, so they're usually just somewhere in the realm. And I pulled my other paper in 'cause this is still the best picture I've ever seen of a reversion. So this was a paper that we worked on at City of Hope... I see Beta's on trying to see if you're an author, Beta... We're not an author. Sorry, but we had a lot of city Hope authors here on this. And Dr. Whites was the last author. There was some work we did with garden health, but this has a really nice picture, so we were looking at germline mutations in the background of their circulating tumor DNA assay.

0:42:24.9 DS: And you can see reversion really nicely in this picture here where... So this was a patient with ovarian cancer who had a BRCA1 germline single nuclear type variant, so this is their germ line variant of BRCA1 2071 deletion A, and it was seen in about 74% of their circulating tumor DNA. So that's really high, and people might be kinda confused there, because we tend to think of variant allele fraction are 50/50, but when you're dealing with tumor, it can be really disorganized, so if you're losing the other allele the allele, with the germline mutation can actually be enriched. We did some of that analysis here too, that shows that I won't really belabor it, but that's what it's getting at, as your allele fraction goes up or as your copy number variation of that locus goes up, your allele fraction goes up. That's genetics 102.

0:43:26.3 DS: But the... So then this person also had a second hit, which is this BRCA1 gene deletion again, most of the time that second hit is a loss of heterozygosity events where you're usually losing that whole gene, and then this is really interesting, so then essentially downstream of their germline mutation, and remember, this allele is lost, so now you see this other variant pop up and so, yeah, it's most likely on this allele, the allele with the germline mutation, and it was at a pretty low fraction 2%. And again, these fractions are so low, 'cause these are actual circulating tumor DNA, but if you see the germline mutation is a 2071, this a little bit downstream, 2079 to 2101 deletion. So it's exactly what that meta-analysis showed, this is over one base pair deletion and adjust the reading frame, so it's a frameshift by itself, but when you take one frameshift combining with another frameshift sometimes it can actually fix the reading frame.

0:44:34.5 DS: So you had... The reading frame was broken here in the DNA when it goes to RNA, so the reading frame would be off, meaning that you not get viable protein product from the germline mutation on that allele, but then if you go a little bit downstream, there's actually a fix, because now you have this other mutation coming in on the DNA, so the messenger RNA... Remember, DNA makes RNA makes protein, the messenger RNA actually fixes that reading frame, so now the nucleotides from the RNA are set up correctly again for your amino acids, so it does the correct amount of amino acids in the right order after that fix and apparently this little bit that was lost in the middle, or disruptive in the middle wasn't important enough for BRCA1 to function correctly, 'cause this restored the reading frame and this person became, was already having resistance of PARP inhibitors therapy, so it's gonna be really interesting over time to see where this whole field goes because we're definitely learning that if you get those second hits.

0:45:45.4 DS: It can disrupt how you're responding to PARP inhibitor therapies. In this meta-analysis, they actually bring up that in the one cohort they looked at, it was like 25% of people that had a reversion mutation. But I just think there's a lot of unknowns, so it's a great area to study, we definitely need more information. I would imagine over time, we'll have a pretty good studies showing the consequence of picking up a reversion, the problem is it's only gonna be in some subset of cells, so what does that mean for the overall treatment? Is it gonna be enough to say... Some people are even using PARP inhibitors in people that don't have any deficiency in the HR pathways or HR proficient, and you have to wonder if over time, we'll have enough data to say, "Yes, we see a reversion floating around to try something else."

0:46:40.3 DS: But right now, at least in ovarian cancer, it's not like we have a whole arsenal of treatments, different cancers like breast cancer, there's clearly a lot of chemotherapeutic agents many that work, and you have a lot of different options, but ovarian cancer has always been a pretty limited cancer, it's a lot of platinum and taxing therapy in the beginning, and now PARP inhibitors are absolutely changing the whole face of ovarian cancer, doubling-tripling, people are even potentially saying some people are maybe cured and things like that with ovarian cancer, which is crazy, and no one's ever really said that. I would say before the last couple of years, but yeah, we are gonna have to keep close tabs on these reversion and what that is gonna ultimately mean for treatment.

0:47:28.0 Shelly: TJ, with these reversion mutations. Does that lead you... I know there's a lot of uncertainty, but does that lead you to believe that maybe we'll have to do serial testing to monitor the reversion mutations that could occur if like...

0:47:45.4 DS: Possibly, yeah. I think time will tell.

0:47:47.5 Shelly: Okay.

0:47:48.2 DS: Yeah, and Kirsten, I'm sure I remember and Diana, you might know too, we had some data on serial biopsies, and I thought the overall genomic scoring didn't get worse and it sounds like we haven't really reported out much on the reversion front. I don't know if anyone wants to elaborate there.

0:48:06.9 DT: Ask that Kirsten just... Oh, there she goes.

0:48:10.2 Kirsten: Yeah, we did do a study at one point looking at peered samples from a variant ovarian patient, so it was their primary resection, this is a biopsy taken it, first relapse from platinum treatment, and we didn't see any patients, whether they changed status in terms of myChoice status. And the scores were very concordant as well. In that study, we did identify one reversion, we have reported it out occasionally on reversions that we've seen, most of them have been similar to, but we haven't done a comprehensive report. Most of them are that similar to the scenario you described where you have primary mutation that's usually a deletion, second hit is usually loss of heterozygosity, and then what you see with the reversion is a subsequent endow that puts the gene back in frame. We did come across one tumor in one patient where they had a pathogenic substitution, so... And a minor acid change that was pathogenic, and those are pretty rare, and in that particular patient, post-platinum therapy, they developed a reversion and it was a reversion to wild type at that same pace, so that's the only time that we've ever seen a reversion wild type, and obviously, those are gonna be much more challenging to detect, especially if you only are looking at the post-treatment sample.

0:49:47.4 DS: Yeah. And in any of the clinical trials that we've done to date, have you ever looked at the sub-set of outcomes and people with versions... I'm just curious.

0:49:58.3 Kirsten: No. I mean, almost all of the clinical trials that we've done, even the late stage ones like Quadra, the sample, the tissue sample we've received has been from the primary resection. I know that there is increasing interest from pharmaceutical companies who are working on PARPs to look at reversions, so I expect that we will start to see clinical trials where the a reversion mutation is being assessed.

0:50:29.0 DT: They reported ARIEL 4. It was 20 or less than 20 patients with reversion, and it showed that in that late line setting, they did not respond to PARP.

0:50:47.0 Kirsten: Yeah. And I think that you kind of see that from the the response and clinical benefit you see in Quadras far lower than what you see in the frontline setting, and I suspect that we're gonna see these pharma companies wanting to try and find a way of treating those their substitute patients that are no longer PARP respondent.

0:51:11.0 DS: Yeah, yeah. It's really interesting. A lot to learn, for sure. Yeah, that's good. Well, any other questions? I do have a hard stop personally myself, so we may wrap it up at the top of hour, but next week, yes, 6:00 PM Eastern will be trying to get a special secret guest. We'll see, but I wanna do tumor normal anything in between so bring any sort of complex questions that you have and... Yeah, again, I'll just be open and... Yeah, whatever you wanna discuss, we'll definitely have some content to share though no question. So yeah, anything else from anyone? Did we answer all the questions? I see Robin Palmer he has a reversion of the same or different sites than the initial mutation or... Can it be either? Yeah. So we kind of addressed that, I think Robin. Yeah. Good. Well, great. Well, thanks everybody. Thanks, Diana, for coming on, thanks Kirsten. Thanks, Ralph. Thanks, Shelly, for running the chat, and I hope everyone has a great rest of their day.

[music]