Schnaiter et al. pooled Myriad MyChoice HRD+ CDx results from four cohorts (4,943 HGOC tumors) to test whether tumor HRD-related genomic instability scores (HRD-GIS) provide evidence for BRCA1 and BRCA2 variant classification under ACMG/AMP criteria.
0:00Welcome to Base by Base, the papercast that brings Genomics to you wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. Thanks for having me. great to be here.
0:10Okay, so I want you to imagine for a moment taking a genetic test to understand your risk for hereditary breast or ovarian cancer. You know, you swab your cheek or maybe you give a blood sample, send it off to the lab, and you wait.
0:22And you're expecting a really definitive answer, right? Yeah, you want to clear, say, for, you know, a clear at risk result. Exactly. So you can plan your life, structure your healthcare, maybe take necessary preventive measures, but instead you get a giant question mark.
0:38And that question mark, in the world of clinical genetics. That is known as a variant of uncertain significance. Or a VUS. The V US. It means the lab found a mutation, like a spelling change in your DNA, but the global scientific community simply doesn't have enough data to know if that specific typo causes cancer.
0:59Or if it's just like a completely harmless, benign quirk of your personal biology. And getting that result is, it's incredibly common, isn't it? Oh, very calming. Because in major genetic databases, like Clinvar, up to 44% of the reported variants in the crucial BRCA one and BRCA 2 genes are classified as variants of uncertain significance.
1:21Which is a huge number. Right. And in a recent Danish cohort, analyzing high risk patients, it was 34%. So, to put that into perspective, getting a VOS is, it's like a warning light flashing on your car dashboard, you know?
1:36You take the car to the mechanic, they plug in their little diagnostic tool, but then they just shrug and tell you they have never seen that specific error code before. That is a perfect analogy. You're just left to drive them around, not knowing if the engine is about to completely explode, or if it is just a faulty sensor.
1:52Yeah, and because the clinical impact is entirely unknown, neither the presence nor the absence of that variant allows you or your doctor to make informed decisions. You can't plan treatments. No, you can't plan preventive surgeries, you are just essentially trapped in this clinical limbo.
2:10Which raises an incredible question for our deep dive today. How could this agonizing diagnostic mystery change if we stopped just looking at the genetic curve itself and instead looked at the physical damage the tumor leaves behind?
2:24Like, what really happens when we use a tumor's own physical scars to decode the DNA? And that is the exact puzzle researchers are solving in the material we are analyzing today. They're shifting the focus from just reading the blueprint of the genetic code to actually examining the physical wreckage left behind in a cancer cell.
2:44When its molecular machinery breaks down. Exactly. Well, today we celebrate the work of Simon Schneader and the team at the Medical University of Innsbruck, along with the Enogimma consortium, who have advanced our understanding of how tumor genomic instability can classify BRCA variants.
2:58It's a fantastic piece of work. They took a massive pool of real-world clinical data, and they actually found a way to read those physical scars. Yeah. But, um, to understand how they achieved this, we 1st have to reframe how we think about BRCA one and BRCA two, right?
3:13Yeah, we do. Because we usually think of them strictly as, you know, cancer risk markers, like red flags. Right. But their actual day-to-day job in the cell is high fidelity DNA repair. Because our DNA is just constantly under attack.
3:26Just going about our normal metabolic processes or, you know, being exposed to background radiation, it causes our DNA strands to break. And the most dangerous type of damage is a double strand break. That's where both rails of the DNA ladder are just completely snapped in half.
3:41And when that severe break happens, the cell relies on a mechanism called homologous recombination, and BRCA1 and BRCA2 are the crucial proteins that orchestrate this repair. Right, essentially they act like site managers, they find the matching backup copy of that DNA sequence on the sister chromosome, and they use it as a flawless template.
4:03To stitch the broken ladder back together. Exactly. Without losing a single litter of code. But when a patient has a pathogenic BRCA variant, meaning the protein is fundamentally broken, the cell loses access to that high fidelity repair system.
4:17It enters a state called homologous recombination deficiency. Or HRD. HRD, right? And here's the catch. A cell with HRD still has to fix those broken DNA strands, otherwise it just dies immediately. It panics, it resorts to these sloppy backup repair pathways.
4:33And these backup systems, they don't use a flawless template. They just kind of grab the broken ends of the DNA and blindly jam them back together. Which I imagine introduces massive errors. Huge errors.
4:45Those sloppy repairs scramble the chromosomes completely. Chunks of genetic code are lost, whole ends of chromosomes are like fused to the wrong partners. Wow. Yeah, the entire genome becomes deeply unstable.
4:58And that chaotic instability is a central driver of oncogenesis. you know, the creation of cancer. Especially in high grade ovarian cancer, right? Exactly. But that exact same chaos also creates a really unique therapeutic vulnerability.
5:12How so? Well, if the cancer cells primary BRCA repair is broken and a doctor treats the patient with a PRP inhibitor drug, which blocks those sloppy backup systems, the cancer cell essentially accumulates way too much unrepaired damage and it just self-destructs.
5:27Oh, that makes sense. Yeah, that concept of synthetic lethality is the driving force behind HRD testing today. Oncologists routinely test ovarian tumors to see if they have this genomic instability. So they can prescribe the PRP inhibitors.
5:41Right. If the tumor is highly unstable, it proves the cell is deficient in homogous recombination, and the patient gets the drug. Okay, lets unpack this. If a VUS means we just don't know if a variant is harmful or benign.
5:54Does that mean doctors are basically flying blind when choosing a treatment for these patients? If we connect this to the bigger picture. While HRD testing is already used to guide therapies. It's potential to actually classify the underlying genetic variant itself, hasn't been fully utilized until now.
6:10Because there's a disconnect. A massive disconnect, yeah. The American College of Medical Genetics and Genomics. along with the association for molecular pathology. The ACM GMP guidelines. Right, those guidelines.
6:21They have a highly rigorous statistical framework to classify variants. But historically, that framework relied heavily on, you know, population frequencies, computational predictions and family pedigrees.
6:32So the tumor's physical state was used to treat the immediate cancer, but that localized tumor data wasn't formally translated into the math required to update the germaline genetics. Exactly. It didn't make it to the global databases.
6:46So the variant remained a V US for the patient's entire family. Because measuring genomic chaos is, I mean, it's incredibly difficult to standardize, right? You can't just look under a microscope and say, yep, looks messy.
6:58You need hard replicable numbers. Which brings us to how Schneider's team actually bridged this gap in their study. Right. They needed massive amounts of uniform data to prove a statistical connection between the genetic variant and the physical scars.
7:13So they conducted an extensive literature review. Yeah, identifying 4 independent cohorts of high grade ovarian cancer. They use data from the Marburg cohort, the NHS cohort, study 19, and the NoVA trial.
7:26That gave him an impressive data set, almost 5000 tumors. right? 4943 tumors, to be exact. And for every single tumor, they knew the patient's exact BRCA variant status. And to ensure they were comparing, you know, apples to apples across all those tumors, they restricted their analysis to tumors assessed by one specific diagnostic tool.
7:47Yes, the myriad my choice, HRD plus CDX assay. That standardization is critical. It is. The myriad assay is this really fascinating piece of technology. It physically sequences over 50,000 single nucleotide polymorphisms. SMPs. Right, SMPs, evenly distributed across the entire genome.
8:07It basically maps the topography of the cell's DNA to quantify exactly how much sloppy repair has actually taken place. And it calculates a composite genomic instability score, an HRBGIS. Based on 3 specific types of biological wreckage.
8:21Let's go through those. First, it looks for a loss of heterosigosity. Yeah, so we usually inherit two different versions or alleals of every chromosomal segment. When the cell gets desperate to fix a break using those sloppy backup systems.
8:32You might just delete a massive chunk of one alleal entirely. Leaving only the other one. Right. And second, the assay measures telameric alalic imbalance. Telomeres are the protective caps on the ends of our chromosomes, and the backup repair systems often cause uneven chaotic breaks that extend all the way into these telomere regions.
8:50Throwing the genetic balance completely off. Exactly. And finally, it measures large scale state transitions. This is where massive chunks of chromosomes snap and just get glued back together to completely different chromosomes.
9:03Wow, creating these massive structural fusions. Yeah. The assay adds these 3 metrics together to produce a single score. And in the clinical world, a score of 42 or higher is considered G shy. GIS High.
9:18That is the established threshold where a tumor is officially recognized as having homologous recombination deficiency. So if the BRCA gene is the mechanics supposed to fix the DNA, the HRDGIS is like measuring how many dens are left on the car, and a score 42 means it's heavily dented.
9:33I mean, yes, basically, a score 42 means the vehicle has been absolutely totaled by an incompetent mechanic. And that works perfectly for the initial concept. If the variant is genuinely pathogenic, the mechanic is broken, and you expect to see that massive structural damage.
9:48But to prove this holds up to the strict ACM JMP guidelines, the researchers had to do some math. They had to calculate likelihood ratios. They needed to see how reliably a high dent score predicted a truly pathogenic variant versus a harmless wild type variant.
10:05And when they applied this math to the 5000 tumors, they found that out of the tumors known to be driven by genuinely pathogenic, broken BRCA variants, an overwhelming 91% were GSI. 91%. They had that heavy genomic scarring above the score of 42.
10:21which completely confirms the biological mechanism. But the wild type tumors, you know, the cancers that developed in patients with completely normal, healthy BRCA genes, they provided a crucial baseline.
10:32Yeah, in those tumors, 30% were also DA shy. Wait, if the patient has a perfectly healthy BRCA gene, why would their tumor still be riddled with 30% of these massive genomic scars? Because biology always finds a workaround.
10:44You know, homologous recombination relies on an entire pathway of proteins, not just BRCA one and BRCA 2. Oh, I see So a tumor could have a mutation in a completely different gene within that same pathway, like, uh, RD 51C or Pelb2.
11:00Or even more commonly, the tumor utilizes epigenetic silencing. Ah, okay. So the BRCA gene sequence is perfectly fine, no typos at all. But the cancer cell attach chemical locks, like methyl groups, to the genes promoter region.
11:15Exactly. It essentially taped the blueprint shut. The protein is never manufactured, so the cell still suffers from the exact same deficiency and accumulates the exact same scars. The mechanic is fully qualified, but has been like locked out of the garage.
11:28That's it. And because of this biological complexity. Finding hygenomic instability doesn't guarantee the BRCA gene itself is mutated, which is why the math is so vital. Right. So the researchers calculated that the likelihood ratio for a BRCA variant being pathogenic, if it's found in a GI shy tumor, is 3.03.
11:45Meaning a variant in a highly squared tumor is roughly 3 times more likely to be a true cancer driver. And under that rigorous ACM jamp Asian framework, a likelihood ratio between 2.08 and 4.33 translates to an official evidence strength of supporting pathogenic.
12:04Which is massive. Yeah. It means if a patient has a VS, but their tumor shows high instability. Geneticists can mathematically add a supporting pathogenic weight to the scales. Moving that variant out of limbo and closer to a definitive, dangerous classification.
12:18And the math works in the inverse direction too, providing even stronger evidence. It does. The likelihood ratio for a variant being pathogenic in a G's low tumor, so a tumor without heavy scarring was just .13.
12:30And under the guidelines, a ratio between .053 and .23 translates to moderate benign evidence. So a lack of genomic scarring strongly suggest the BRCA protein is functioning perfectly, pushing the variant heavily toward a harmless classification.
12:44Here's where it gets really interesting, but wait, if a broken BRCA gene causes this instability, why on Earth did 9% of the pathogenic tumors have a low score? Did the test fail? What's fascinating here is the binary threshold of 42 is rigid.
12:59First, we have to look at how that threshold of 42 was even established. Right. How do they pick that number? It was originally said to capture 95% of tumors with known pathogenic variants in earlier studies.
13:10But is a very strict binary cutoff. When the researchers analyze the individual cohorts here, many of those low scores were sitting right on the boundary. So they weren't actually zero. Far from it. In the Novier cohort, for example, all of the pathogenic tumors that were classed as low actually had genomic instability scores above 33.
13:31Oh, wow. Yeah, they exhibited significant structural damage. They just missed that rigid clinical cutoff of 42. So if the clinical guidelines eventually evolve to allow for continuous variables or maybe multiple categories like high, medium, and low.
13:46The statistical power of this asset could be even stronger. Definitely. But threshold math doesn't explain everything. There's a deeper biological phenomenon happening in that 9%. And that requires stepping away from the car analogy for a 2nd to look at the fundamental rules of genetics.
14:01Right. In the Marburg cohort. Several of the tumors that had pathogenic BRCA variants, but paradoxically low instability scores were of endometrioid histology. Which is a very specific subtype of ovarian cancer.
14:14It is. And in these particular tumors, the inherited BRCA variant was likely just a harmless bystander. A bystander. How can a broken DNA repaired gene just stand by while cancer forms? It comes down to Knudson's 2 hit hypothesis.
14:28Right, the 2 hits. Yeah. We inherit 2 copies, or alleles, of the BRCA gene, one from our mother, one from our father. Inheriting a pathogenic variant means you were born with one broken copy in every single cell of your body.
14:40That is the 1st hit. But the cell can still function perfectly well using the one remaining good copy. The high fidelity repair system stays online because the backup alleal is doing all the heavy lifting.
14:51Yes, for a classic BRCA driven cancer to develop. The cell has to suffer a random localized 2nd hit. Like a spontaneous mutation. Or an epigenetic silencing event. Yeah. Something that knocks out that remaining good copy.
15:05And only when both copies are gone, does the cell plunge into homologous recombination deficiency? Triggering the massive scarring and driving the cancer? So in these specific endometrioid tumors, the patient inherited the 1st hit, the broken BRCA gene, but the ovarian cell never took the 2nd hit.
15:25The good copy just kept working. Instead, the cell became cancerous for an entirely different unrelated biological reason. Right. The BRCA system was still actively repairing double strand brakes during the tumors formation, which is exactly why the myriad assay didn't find any massive chromosomal wreckage.
15:42That is mind blowing. The genetic variant was literally a bystander at the scene of the crime. It was. The cancer formed through an alternative pathway, bypassing the BRCA vulnerability entirely. But unfortunately, the researchers didn't have comprehensive data to track those 2nd hits across all the cohorts, did they?
15:59No, and that is a primary limitation of retrospective studies like this. Standard clinical pathology reports rarely map out whether a 2nd hit like a localized loss of heterozygosity at the BRCAE gene itself actually occurred.
16:15Plus the data sets didn't always differentiate between germ line variants, the ones you inherit, and somatic variants, which are spontaneous mutations that only exist inside the tumor. Right. So we have to be really clear about the boundaries of how this deep dive into the data can be applied in the real world.
16:33Absolutely. The statistical power and the likelihood ratios generated in this study are strictly validated for high grade ovarian cancer. And crucially. They are validated specifically using the My Choice, HRD plus CDX assay.
16:47Because that was a standardized tool used across these 5000 tumors. Right, because different sequencing technologies or different mathematical algorithms might measure those chromosomal fusions and telomere imbalances differently.
16:58You can't just take the math from the myriad assay and apply it to a completely different diagnostic kit. No, unless an alternative assay undergoes rigorous validation to prove it aligns perfectly with this data set, variant curators have to be cautious.
17:13We also need extensive subsequent research before we can apply these exact likelihood ratios to other hormone driven cancers, like breast or prostate cancer. Because the biological pathways, and therefore the scoring thresholds, might look quite different in breast tissue compared to ovarian tissue.
17:29Exactly. So what does this all mean? If a listener has a family member with a VUS? Does this mean they need to demand this specific myriad assay? Or is this more about scientists pooling data behind the scenes to update the rule books?
17:42This raises an important question. While the assay limitation is a real boundary for individual clinical testing. The overarching goal of this research is systemic change. It's about translating physical tumor damage into the rigorous Besian math required by the ACMJM guidelines.
17:58Yes, this data empowers the clension enigma consortium to permanently upgrade the global variant databases. It's essentially crowdsourcing clarity from thousands of tumors to eliminate the uncertainty for future patients.
18:10Right. When varying curators reevaluate a VUS in a global database like Clinvar, they can now look at this research and mathematically bump that variant up to likely pathogenic or down to benign based on the physical evidence left in these 5000 ovarian tumors.
18:27And once that database is updated, that definitive classification becomes immediately available to genetic counselors and doctors everywhere, regardless of what diagnostic assay they use in their local clinic.
18:38It completely removes the paralyzing uncertainty for families trying to assess their hereditary cancer risk. Allowing them to confidently pursue targeted life-saving prevetic measures. Exactly. So, to summarize the findings from today's deep dive, the profound genomic scars left by a homologous recombination deficiency offer powerful statistically robust evidence to classified BRCA1 and BRCA2 variants.
19:03By integrating these physical tumor instability scores into the rigorous clinical frameworks, researchers now have a vital mathematical key to decode variants of uncertain significance. Moving them out of clinical limbo and into actionable categories.
19:15This research masterfully bridges the gap between the physical reality of a tumor's wreckage and the abstract microscopic code of our inherited genome. What does this mean for the future of personalized medicine?
19:27When the physical damage inside a tumor can be used to diagnose the very genetic flaws that caused it. a huge step forward. This episode was based on an open access article under the CCBY 4.0 license. You can find a direct link to the paper and the license in our episode description.
19:43If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating. If you'd like to support our work, use the donation link in the description. Now stay with us for an original track created, especially for this episode, and inspired by the article you've just heard about.
19:58Thanks for listening, and join us next time as we explore more science based by base.
