This episode examines a validation study of a highly sensitive NGS-based microsatellite instability (hs-MSI) assay for diagnosing constitutional mismatch repair deficiency (CMMRD). The assay was tested on blinded blood cohorts and CMMRD-associated tumors, compared with a low-pass whole-genome LOGIC/MMRDness score, and evaluated for gene-specific MSI indel patterns that identify PMS2 biallelic carriers.
0:10White Coast, bright screens, midnight on repeat. Counting tiny stars in the DNA DM, but so quiet in tumors that shout, looking for the crack with the... Welcome to Base by Base. The paper cast that brings genomics to you wherever you are.
0:24Thanks for listening, and don't forget to follow and rate us in your podcast app. Imagine for a 2nd um, that the cellular spell checker in a child's DNA like, the system that catches and fixes microscopic typos as cells divide, was just completely broken from birth.
0:39Right, not just struggling a bit, but entirely offline. Yeah, entirely offline. What actually happens when those genetic typos are just allowed to pile up, unchecked day after day. Well, you end up with, um, a catastrophic level of genomic instability.
0:55The whole fundamental architecture of the sales instruction manual just degrades, mutating way faster than the body can handle. And the stakes there are incredibly high. For a child born with this broken spell checker, there is a near 100% chance of developing aggressive, life-threatening cancers before they even reach adulthood. So the question is, how could a simple blood test not only catch this broken spell checker before it's too late, but actually read the specific pattern of the typos to find out exactly which genetic part is missing?
1:23Answering that, I mean, it changes the entire landscape of pediatric oncology. We're talking about moving from educated guesses to highly specific functional genomic blueprints. Exactly. So today we're taking a deep dive into this groundbreaking news study that totally flips how we diagnose the syndrome.
1:41Our mission today is to figure out how finding a specific flavor of genetic typo is actually saving children's lives. And before we get into the mechanics of that, we really need to acknowledge the massive collaborative effort here.
1:55Oh, absolutely. scale of this is huge. Yeah, today we celebrate the work of the researchers at the hereditary cancer group at Idabelle in Spain, the hospital for sick children in Toronto, and the international IRRD and C4CMMRD consortiums, who have advanced our understanding of childhood cancer predisposition.
2:12I mean, it takes global teamwork to crack rare mysteries like this. It really does. And to understand the magnitude of what they've done, We should probably define the clinical problem for everyone listening.
2:20The condition is called constitutional mismatch repair deficiency, or CMMRD. If you mentioned, it's extraordinarily rare and highly penetrant. And highly penetrant in genetics just means... It means that if you have the mutation, the likelihood of expressing the disease is almost absolute.
2:38You are going to develop it. Right. And we really need to emphasize why getting this diagnosis right is a matter of life and death for these kids. By age 20, patients with CMMRD have a greater than 90% chance of developing severe cancers.
2:51We're talking high grade brain tumors, hematological malignancies like lymphomas, early onset colorectal cancers. The mean age of onset is, what, just 7.5 years old? 7.5, yeah. It's devastating. The mechanism driving this is rooted in these biolic germline pathogenic variants.
3:11Okay, let's break that down a bit Sure. So Jern Ryan means the mutation is inherited, right? It's present in every cell from conception. And biolaic means the child inherited a defective copy of a mismatch repair gene from both parents.
3:23So they have 0 backup copies. Exactly. No backup. These genes, specifically MLH1, MSH2, MSH6, and PMS2, are the pillars of the DNA and mismatch repair system. And what's fascinating here is that PMS2 is actually the most frequently mutated gene in this group, but it's famously difficult to diagnose.
3:44Why is it so hard to spot? Because of something called pseudo genes. They're basically genomic copycats that just confuse the sequencing machines. Okay, let's unpack this because the clinical reality of what happens when those pillars fall is huge.
3:57When a child with CMMRD develops a tumor, the standard instinct is to hit it with traditional alkalating chemo, right? Like timisolamide, which is standard for things like gleoblastoma. But those drugs fail spectacularly here.
4:09They do, and they fail because of how those alkalating agents actually function biologically. These drugs are designed to intentionally inflict massive damage to the cancer cell's DNA. But the drug itself doesn't kill the cell.
4:20It relies on the cell's own intact DNA repair system to recognize that catastrophic damage and say, okay, this is beyond fixing and trigger apoptosis basically cellular suicide. Wow. So if a child has CMMRD, their repair system is totally offline.
4:34The chemo bombards the tumor with damage, but the cell has no spell checker to even notice the damages there. Exactly. It just ignores the massive toxicity and keeps dividing. The chemo is essentially useless, but you're still subjecting the child to severe systemic toxicity.
4:50But that's awful. It is. However, knowing a child has CMMRD allows oncologists to immediately pivot to a treatment that exploits this broken system, which is immune checkpoint inhibitors. Immunotherapy.
5:03Because the tumor cells are accumulating thousands of typos, they start making weird mutated proteins that normal cells don't make, right? Spot on. The immune system sees those strange proteins as foreign invaders.
5:14We just had to give the immune system the right drugs to take the brakes off so it can attack. So the diagnosis directly dictates the treatment. We have to diagnose CMMRD fast. But historically, our tools for that have been pretty bad.
5:25I mean, think of it like this. If your DNA is a 1000 page book, standard microsatellite instability tests or MSI tests only look at 5 specific words to see if the entire spell checker is broken. It's an incredibly tiny sample size.
5:38Right. And it was originally designed for adult cancers. Those standard 5 marker tests miss CMMRD tumors entirely, most of the time, they yield only an 18% sensitivity rate. And the hurdle gets even steeper with PMS too, like I mentioned, because of those pseudogenes.
5:53I've heard pseudogenes described as like static on a radio, but that doesn't really explain how they trick a sequencer. It's more like, um, trying to find a crucial document on your computer, but there are 10 fake shortcut files with the exact same name that lead nowhere.
6:10The sequencer keeps opening the wrong files, right? Yeah, there's a much better way to look at it, yeah. Suder genes are these evolutionary relics. They look almost identical to the real PMS2 gene, but they're completely nonfunctional decoys.
6:21So when our next gen sequencers try to read the gene, they accidentally read the decoys, the data is just an unreadable mess. Plus, you have variants of unknown significance, where we see a change, but we don't know if it actually breaks the gene.
6:33So we've got an aggressive childhood cancer, a tiny window to act, a test that only looks through a keyhole, and a genome full of decoy files. How did the Ida Bell team fix this? Well, they obviously needed a wider lens, so they developed and validated something called highly sensitive assessment of MSI, or the HSMSI assay.
6:56Instead of 5 markers, it targets a custom panel of exactly 192 microsatellites. Okay, let's define what a microsatellite actually is for anyone not living in a genetics lab. Fair enough. Microsatellites are short, repeating sequences of DNA letters, like A-T-A-T-A-T over and over.
7:13Because they're so repetitive, the cellular machinery that copies DNA easily slips or stutters there. It's like trying to type the same word 50 times in a row without making a typo. Exactly. It's tough.
7:24So if the mismatch repair system is offline, these repetitive regions are the very 1st places where errors will accumulate. makes sense. So the researchers ran this 192 marker panel on a blinded cohort, 66 blood samples and 24 CMMRG tumor samples.
7:38But they didn't just test it in a vacuum. They pitted the HSMSI scores against logic scores, which were also called MMR Dinas. Right, and logic sequences the entire genome at a low pass level to live for instability across 1000000s of micro satellites, not just 192.
7:56It's the ultimate stress test for a new tool. But I have to push back here. Wait, if the logic test already looks at the entire genome, why are we trying to shrink our view down to 192 markers? Doesn't more data mean a better diagnosis?
8:07Why not just read the whole genome every time? I mean, that is the constant tension in modern genomics comprehensiveness versus clinical utility. Whole genome sequencing is incredibly thorough, but it takes massive computational power and significant time.
8:19With a rapidly progressing childhood cancer. You just don't have weeks to wait for a supercomputer to crunch 1000000s of data points. Ah, okay. So the 192 marker panel is like a proxy, a way to get the same answer without the bioinformatics bottleneck.
8:32Yes. It's flexible enough for standard sequencing platforms in clinical labs, has low computational requirements and a fast turnaround time. Okay, so let's look at how well that proxy actually performed.
8:43The researcher set a predefined score threshold of 4.57%. So out of the 192 genetic spots. If more than about 8 or 9 showed typos, the sample was flagged for CMMRD. And using that, the HS MSI test hit 98.5% sensitivity, and 100% specificity in detecting CMMRD directly from a blood draw.
9:05Which is incredible. 100% specificity is massive. It means 0 false positives among the controls, which even included patients with other complex cancer syndromes like lynch or lyfromony syndrome. didn't get confused.
9:17Wow. And the sensitivity was nearly perfect, too. 31 out of 32 CMMRD individuals correctly identified. The one case that missed is actually super interesting biologically. It scored a 3.70, so just below the threshold, but still higher than healthy controls.
9:30Right. That patient was clinically atypical. They had a hyperamorphic variant. Meaning the broken gene wasn't completely destroyed, just severely impaired. Exactly. It retained a tiny flaction of its original function.
9:40So this specific patient didn't develop their 1st tumor until they were 40, which is practically unheard of, for CMMRD. The assay accurately picked up on that slightly lower level of instability. And when they correlated the HS MSI scores from the blood directly with the whole genome logic scores, they got an R value of .89.
10:01To put that in perspective, looking at just these 192 spots gives you almost the exact same picture of systemic instability as spending weeks scanning the whole genome. It's like predicting a national election by polling one perfectly representative neighborhood.
10:15And it wasn't just blood. The assay successfully detected instability in all 24 tumor samples, whether they originated in the brain, colon, or blood. Okay, but we still have a major lingering problem. The assay proves the spell checker is broken by finding typos.
10:30But to finalize the diagnosis, clinicians still need to know which specific gene is mutated. And if it's PMS 2, we still have all those decoy files. How do they solve the pseudogene problem if they aren't reading the gene itself?
10:41This is the best part. They stopped trying to read the blueprint of the broken machine and started looking at the specific footprint the machine left on the factory floor. They analyze the actual types of typos, the indels, or insertions and deletions happening at these 192 markers.
10:57Right. Here's where it gets really interesting. When the cellular copying machine stutters, it usually drops a letter. So CMMRD blood usually has a massive proportion of one base pair deletions, like 77% of the errors were deletions on average.
11:11Yes, for patients with MLH1, MSH2 or MSH6 mutations, they follow that standard pattern. The copying machine was dropping letters. But the patients with biolic PMS 2 mutations, the ones hidden by decoys, showed a completely different biological signature.
11:26Instead of dropping letters, the system was constantly adding an extra letter, a significantly higher proportion of one base parent insertion. Exactly. This specific flavor of genetic typo became a definitive diagnostic marker.
11:38This insertion signature identified bileic PMS2 carriers with an accuracy of .997 is practically flawless. It's just a brilliant workaround. You bypass the decoys entirely by looking at the functional consequence of the mutation.
11:51Does the blood show the insertion signature? Yes. Then they have a PMS 2 mutation definitively. You don't even have to sequence the tricky gene. If we connect this to the bigger picture, the clinical implications cannot be overstated.
12:04A fast, definitive diagnosis diverts these kids away from toxic chemo and straight toward immunotherapy. It also tells us how aggressively we need to screen them. And they even looked at predicting the timeline of the disease.
12:18Yeah, they found a negative correlation between the instability score and the blood and the age of the 1st tumor diagnosis, and our value of negative .43. So the higher the typo rate in the blood, the younger the patient was when cancer appeared.
12:30Which tracks biologically. A higher baseline mutation rate means you're going to roll those disastrous genetic dice much faster. But we have to be careful not to present this as a crystal ball. The study is very transparent about confounding factors.
12:43Oh, absolutely. The age when the blood is drawn matters, because typos accumulate over time. The specific gene affects the baseline too. Like MSH6 mutations tend to have lower overall instability scores than PMS 2 mutations.
12:57And the paper also dives into the sheer chaos inside the tumors themselves. The blood gives a stable germ line picture, but the tumors acquire secondary somatic mutations, like poly or peel D1. Wait, let me try to translate this.
13:12If the mismatch repair system is the proofreader. Genes like pulley and POD1 are the typist putting letters on the page, right? That is exactly the dynamic. They are polymerases, the enzymes physically copying the DNA, and they have their own built-in proofreading ability.
13:26So if a tumor starts with a broken proofreader from CMMRD, and then randomly mutates to break the typus 2. You've got a blindfolded typus and a sleeping proof reader. The mutation rate goes into overdrive.
13:37We call that ultra hypermutation. When a tumor gets a coal mutation on top of CMMRD. The sheer volume of chadic typos can actually mask that specific PMS2 insertion signature within the tumor tissue. The noise just browns out the signal.
13:50Which perfectly highlights why a highly sensitive blood test is so vital, you read the baseline signature in the blood before the tumor scrambles everything. Exactly. Because of all these variables, we can't use a single blood score as an absolute timeline yet.
14:04We need large prospective collections of samples over time to see if a sharp spike in the score predicts an imminent tumor. Right. Right. Watch the trend line, not just one data point. But even so, as a pure diagnostic tool today, this is a monumental leap.
14:18I mean, look at the suspected cases in their cohort where standard genetics were hopelessly ambiguous because of the pseudogenes. This essay came in, flagged this systemic functional instability, and provided undeniable evidence for a life-saving diagnosis.
14:32The ability to definitively say, yes, this child has CMMRD without spending weeks fighting through genomic static. It opens the door to precision medicine when they need it most. So to summarize all this, the new HSMSI essay uses a targeted panel of 192 genetic markers to accurately detect CMMRD from a simple blood draw.
14:52And beyond just diagnosing this syndrome, the specific pattern of genetic insertions that uncovers definitively pinpoints the tricky PMS2 gene, bypassing those sequencing hurdles entirely. It's a fundamental shift.
15:04We're moving from reading the code to reading the consequence of the code, and leaves us with a really fascinating question. What does this mean for the future of diagnosing other complex genetic syndromes?
15:13Could the specific flavor of DNA errors, these unique functional fingerprints become the ultimate diagnostic shortcut across all the genomics? This episode was based on an open access article under the CCBY 4.0 license.
15:25You can find a direct link to the paper and the license in our episode description. If you enjoyed this, follow or subscribe in your podcast app and leave a five-star rating. If you'd like to support our work, use the donation link in the description.
15:39Now stay with us for an original track created especially for this episode and inspired by the article you've just heard about. Thanks for listening and join us next time as we explore more science based by base.
16:01White Coast bride screens Midnight on repeat. Counting tiny stutters in the DNA beating, but so quiet in tumors that shout, Looking for the crack where the truth leaks out. A missing repair can slip through the scenes, but patterns don't lie in the date, the streams, one gene leaves a fingerprint sharp and clear, turn the noise to signal.
16:19bring it near Running high, is it light up the dark, micro satellites spark. When this match, repair falls out of tune. We wear the scars like a rising. And it's P and it's too. The one base shows a clean little answer.
16:36With a doubt used to grow, roll, roll. I used to grow. On the table, 192 lines. Every repeat, question, every caller sign, scores that climb higher when the clock starts young. One and in the rhythm before words are sung, not every clue is wide.
17:00Sometimes it's a single step where the truth will hide. So we met the indoors that the numbers talk. Got them watchful days down a steady or up. All it high, sensitive, light up the dark. Micro satellite spur.
17:12From blood to tissue, we can name that. So the next move's chosen, our guest alone, and if it's Pete and Mr. But one day shows a clear little answer. Now the future knows.
