 Welcome to noon conference hosted by MRI online. noon conference connects the global radiology community through free live educational webinars that are accessible for all and is an opportunity to learn alongside top radiologists from around the world. We encourage you to ask questions and share ideas to help the community learning grow. You can access the recording of today's conference and previous noon conferences by creating a free MRI online account. You can also sign up for a free trial of our premium membership to get access to hundreds of case based micro learning courses across all key radiology subspecialties. Today we're honored to welcome Dr Sunmi Chah for an update on imaging and 2021 WHO CNS tumor classification. Dr Chah completed her neuro radiology fellowship at New York University Medical Center. She is a neuro radiologist with special interests and expertise in brain tumor imaging at University of California San Francisco Medical Center where she also serves as the program director of diagnostic radiology residency and the vice chair of education. At the end of the lecture please join Dr Chah in a live Q&A session where she will address questions you may have on today's topic. Please remember to use the Q&A feature to submit your questions so we can get to as many as we can before our time is up. With that we're ready to begin today's lecture. Dr Chah please take it from here. Thank you for the nice introduction. I will share my screen. Is it okay to share now? Yes. Okay. This will stop other screen sharing. Do you want to continue? Yes I want to do that. And then if you could just let me know if the screen looks okay from your end. Looks great. Okay great. Well thank you so much for having me today. Hello everyone out there. I'm going to talk about some brain tumor imaging in the context of WHO CNS tumor classification. I don't have anything to disclose. So objective today is to I'm going to highlight a couple of key points from the 2021 WHO CNS tumor classification. And really morph that into how is that relevant to neuro radiologists or general radiologists who are looking at brain MRIs or spine MRIs for patients with brain tumor. And I'm going to just illustrate some of the correlation of molecular genetic markers in terms of what are you looking for on imaging because we're not neuro pathologists we're not basic biologists. But we must keep abreast with this explosive knowledge coming from the biological side because it does have implication for imaging. So the three things that I'm going to highlight is WHO CNS classification and neuro imaging techniques that most of you are very familiar with, but put that in the context of the molecular generic era of brain tumors. And then I'm just going to show you case by case how things are relevant imaging and neuro pathology neuro molecular genetics are intertwined. So first, let's start with CNS WHO classification. So some of you are already familiar that World Health Organization has been supporting classification of not just brain tumors, but all tumors for CNS tumors the first version came out back in 2079 and 2000 up to 2007 and these four classification schemes were purely based on histological and then something magical or something really groundbreaking happened in 2016 classification where molecular genetic information became part of the official classification system. So no longer just a histopathologic features of tumor was used to classify tumor. Now, there's just very, very advanced technique looking at molecular genetics but more interesting to us radiologists is that for the first time ever since 1979 MRI image made it to the cover of this book. And then fast forward five years later, 2021, WHO classification, the fifth edition was published, and you could see that now we have two MRI images. So, imaging is now really gaining attention to our neuro pathology, neuro pathology and neurobiology colleagues that imaging plays a such an important role of how we actually look at tumor so we are well on our way to becoming a very important we were already were but now we are pushing towards being part of the classification of CNS tumors. So the nutshell of the WHO 2021 is that there are molecular markers everywhere. And this is the cartoon I made. And it seems like it's we're just touching the tip of the iceberg. And these are some of the idea of molecular markers that are part of now ordinary conversation and tumor board, but they're actually many, many more to come. And they're already here. And next version of CNS classification which will be coming in maybe five to seven years. We are going to see even more. So what does this mean for radiologists, we just have to make sure that we know what is changing the field of classification of CNS tumors, so that we could keep up with how we interpret imaging. So here's the nutshell of CNS who there are so many molecular markers and we're going to, I'm only going to touch on the several really important ones today. And at UCSF, instead of just getting a histopathological report, we get something like this, this is called UCSF 500 gene panel, where we actually get not just a histological diagnosis, but we get IDH status. But in addition to that, we get a whole host of additional information. And this is really becoming a part of our standard of care for brain tumor pathological diagnosis. So I'm going to show you a couple of cases and we'll go over this at the end. You may not know the answer now, but I could assure you at the end of this 50 minute talk, you'll be an expert at it. So here's a young 19 year old came with a diagnosis of stroke because of the ADC and DWI appearance and end up having these molecular features when pathology was performed. Can you think of tumor type this could be? Second case, here are three different very different histologic tumor types, pylocytic astrocytoma, ganglia glioma, and pleomorphic xanthoastrocytoma. Can you think of a molecular marker that these three tumors often, not always, but often share? So what is the molecular marker? Third, here's a tumor with very interesting apparent calcification. Does not enhance, has this heterogeneous T2 and diffusion abnormality. Can you think of a genetic marker that defines this tumor? How about this one? Here are six different patients with a large midline tumor, all pediatric patients. Can you think of a diagnosis and a molecular marker that all these six different patients share? And how about this one? Two different patients with an extra axial tumor kind of look like meningioma, but they do not have the typical dual tail. This one looks very destructive. Can you think of a molecular marker that can unify these two, a very different type, but now we know that they're very related. How about this one? Two different patients with posterior fossa ependymomas. Do you know the molecular genetic difference between the patient up at the top versus patient up at the bottom? And here are four different patients with four different types of medulla blastoma. Can you name the four main subtypes of medulla blastomas? So we will go over the answer at the end of this talk. So neuroimaging of 2023, still the most important technique is the structural MRI. We cannot interpret physiologic or any other fancier imaging technique without actually seeing what the tumor looks like on structural imaging. We also do physiology based MRI to assess for their vascularity, their metabolism. And another very important type of technique that we use is this hybrid imaging called PET CT or PET MR. And many institutions are beginning to use this technique to look at, is this a recurrent tumor or is this a radiation necrosis? So today I'm only going to highlight a couple of the structural and a couple of physiologic MRI. And as I said before, structural MRI, post-con, pre-con T1, T2, flare, multi-modality, multi-planar imaging, this is the bread and butter of what we do, and this will never go away. But we put additional tests to look for hypervascularity, whether they're leaky permeability, and whether there is a hypercellularity or where there's high-colonial metabolism. Physiologic MRI gives us a lot of insight into a very non-invasive way of glancing at their tumor biology. It's not as good as obviously actually looking at pathology, but it's a really powerful non-invasive tool. And this is what we do at UCSF, and it's pretty standard at most institutions, pre-post T1, T2, FSC, flare, and DWI and ADC, ADC and DW, this is a must. And we also do SWI and ASL profusion imaging. SWI is becoming more and more important in brain tumor imaging because we use this primarily to look for areas of blood products, especially after radiation therapy. And we use this for assessing where the extent of micro and macro hemorrhages, and also vascular lesions that are mimicking brain tumors, and primarily in the brain tumor arena, we use this to assess for the extent of radiotherapy related injuries. So here are three different patients with susceptibility weighted imaging. You could see this patient as literally innumerable punctate dots of susceptibility or micro hemorrhage. This is a patient who received whole brain radiotherapy for medulla blastoma 15 years prior. Here's a patient with very peculiar looking branching pattern of SWI. This is a person with a venialitis. This whole thing was removed, and it's not glial blastoma. This is a venialitis or veins that are partially thrombose. And this is a patient with very clear, large vascular mass, and that's cavernous malformation. So SWI very helpful. Here's an example that we saw a tumor where a patient had a rim enhancing, very aggressive looking, right brachium pontis mass in the posterior fossa. But if you look at patients SWI, there are innumerable punctate micro hemorrhages. This is a telltale sign that patient probably had a radiation therapy. And lo and behold, we got the history after the fact that patient had a nasal pharyngeal carcinoma and a pituitary tumor that were radiated twice before. We did not have the radiation field, but with that history and with that SWI appearance, we feel very comfortable calling this radiation necrosis and patient was treated for steroids to control some of the edema related mass effect and pay this lesion slowly disappeared. DWI very important sequence we use this to assess for acute infarct, abscess, cellular tumor, and actively demyelinating lesion. So here are three different patients. Here's the DWI imaging without even looking at structural imaging when you see this homogeneous in reduced diffusion with an irregular mass like this. This is a intracranial abscess until proven otherwise. Here's a patient you can barely make it out. The lesion on DWI kind of disappears. This is what diffuse glioma looks like on diffusion. Here's a patient with two lesions have a leading edge reduced diffusion. This is pretty classic for non neoplastic, usually inflammatory, actively demyelinating type of lesion and this young patient was biopsied and that's a tumor factor of demyelinating lesion. So diffusion is a must sequence when you're interpreting a brain mass. Here's another example. This patient came to us with a preoperative diagnosis from elsewhere right frontal glioblastoma. I think that's not a bad diagnosis. There's a lots of mass effect. There's edema crossing the corpus callosum with central necrosis, rim of enhancement. But once you see the DWI and ADC, you know that there's homogeneous reduced diffusion within the necrotic tumor. So that is a very unusual appearance of a glioblastoma. So this is more classic appearance for pyogenic abscess and indeed pathology proved that this is pyogenic abscess. Our surgeons going in knew that this was going to be pus because we told them and they end up just doing a little burr hole and sucking that pus out and patient did great. Here's a young woman that I showed a little bit earlier. This young woman was diagnosed with stroke at an outside hospital and you could see why. Why? Because there is actual homogeneously reduced diffusion and it's very dark on ADC. But I think most of you would also notice that that shape is not a good shape for a territorial infar. But nonetheless, patient was fine, the workup was negative, was transferred to our hospital and patient underwent surgery and this is a hypercellular, unfortunately, what's called a molecular glioblastoma. Perfusion, we use it to look for hypervascularity, hypervascular tumors. We also use perfusion for to detect recurrent tumor, assess for glioma grade, and sometimes post-dictal changes. Here's a patient who came to us with a homogeneously enhancing right cerebellar mass and you could see that DWI is not reduced. There's a little bit of a rim of susceptibility but not much else. So the question is, is this a metastasis or something else? Patient did undergo whole body workup and there was no mass. And if you add ASL perfusion, you could see that the whole lesion is very, very vascular in the cerebellum. And this is pretty classic appearance for a hemangioblastoma. And that's indeed what it was on pathology. Here's a person who's been coming to us for serial imaging after patient had a subtotal resection, but you could see here that we don't know where the recurrent tumor here is. Patient had a seizure, they controlled the seizure. So after they controlled the seizure, we brought the patient back and did a perfusion imaging. And you could see that there is a clear unmistakable lump of hypervascularity associated with non-mass like flare here. So our neurosurgical colleagues went in and resected this hyper-perfusion area. And the whole thing was a live recurrent diffuse glioma, IDH wild type. Spectroscopy, we use this tool now as a problem solving tool. Here is a normal, single voxel spectroscopy. Normal, NAA, creatine, and colon. This is what you want to see. And we've done many, many spectroscopic studies, both 2D, a single voxel, 2D and 3D. But I still find this single voxel very powerful. And here's an example of some of our patients that we did on spectroscopy. Oh, by the way, a single voxel only requires about less than three minutes of imaging. So it's a really powerful tool that does not take up much in terms of additional imaging. And we've now kind of developed four different types of spectroscopic appearance of an abnormal lesion. So what we call the proliferative where there's high, high colon, hypoxic profile where there is clear lactate peak, infectious profile where we see amino acids, alanine and acetate, and the necrotic pattern where we see predominantly very high lipid and lactate. And here's an example of how we use this single voxel, two minutes of additional imaging. So this patient came to us with a left frontal glioblastoma as the preoperative diagnosis. It does look like glioblastoma with central necrosis. But once you get DWI, you know that inside of that rim enhancing lesion, there's a clear reduced diffusion that looks like pus. So we brought the patient down. And with a single voxel spectroscopy using TE of 35 milliseconds and 288 milliseconds. And we saw all the metabolites that classically seen in pyogenic abscess such as amino acid, lactate, acetate, and the colon, which is not a tumor marker, it's a membrane turnover marker is very, very low. So our confidence bring the together with the diffusion that this is going to be a pyogenic abscess was near 100%. And our neurosurgical colleagues just did a very small burr hole and sucked out the fluid and lo and behold, there is that yellowish, purulent material, and this is a path proven pyogenic abscess. So let me now focus more on the brain tumors based on the molecular genetics and I'm going to start with three different types of pediatric tumors, and go on to adult tumors using the imaging techniques that I just described to you. So pediatric brain tumors in the WHL scheme from 2016 to 2021. Many different classification changes have been made. The first is these two tumor types one, mezzoloblastoma, the other appenduoma. And you could see that without knowing anything about the tumor. So here's postcon T1. You could see that the this patient has a tumor that is relatively homogeneously reduced on diffusion. So this is going to be some type of a cellular tumor patient. On the other hand, this tumor, both are midline enhancing lesion you could see that DWI is not reduced. So DWI is single most helpful sequence, after looking at post contrast imaging. So we already know this patient has a hyper vascular a hyper cellular tumor. And that's a mezzoloblastoma. And this patient on the right. This is a patient with appenduoma. And DWI is really the first step towards holding down into the molecular or biologic feature of their tumor. So let's start with mezzoloblastoma. Mezzoloblastoma now genetically are divided into four main types. And these some of you are already familiar with the wind sonic hedgehog and group three and group four. And our colleagues from Stanford published this beautiful paper. And this was already nine years ago, showing the imaging difference between the four sub types, the wind, the sonic hedgehog group three and group four. And it turns out that the wind type of mezzoloblastoma are almost always off midline and they're not in the fourth ventricle only. They almost simulate a CP angle. That's the wind type sonic hedgehog are usually the hemispheric tumors with this multi nodular solid component. And there's two different types of it but we are only going to just mention that this is a sonic hedgehog. And then these are the more common pediatric babies and infants can get this type of tumor where the tumor is in the dead midline. And some enhances avidly and some don't. And it turns out that the fourth ventricular midline tumor of mezzoloblastoma with avid enhancement are more likely to be great group three. And the less enhancing sub type tends to be group four. I find this fascinating that imaging, although it's not 100% can give us a glance into potential genetic and molecular sub types. So here's the fourth main sub types. How about ependymomas? So this is an interesting tumor too. So most of you are familiar that usually supratentorial ependymomas are intra parenchymo, the fourth ventricular or posterior fossa ones are inside the ventricle, and the core one is in the intra medullary. And I often wondered why we don't often see ependymomas right in the, in the middle of third or fourth lateral ventricle. Most of the ependymomas supratentorially that I've seen, they're all in the supratentorial parenchymo compartment. But then, now all these genetic information is coming out. So 2016 and 2021, WHO now are classifying supratentorial ependymomas, particularly in pediatric age group, based on this very sophisticated molecular markers. So the relay fusion positive supratentorial ependymomas are one of the most aggressive ependymomas more commonly occurring in pediatric age group, and they look even worse than some of the more really aggressive glioblastomas. And most of these ependymomas, particularly the relay fusion ones that I've seen, they tend to be intra parenchymo. But then when they do recur, they can recur all along the dual surface. So their biology is very different than the typical ependymomas that I've seen. So there are two subtypes that the WHO has clearly defined in the 2021 are the relay fusion and yet one I have yet to see a yet one molecular altered ependymomas. But like I said before, the supratentorial ependymomas in a young child, they tend to be this very aggressive molecular variant called relay fusion is the most common one that I've seen. And that's something that I want to stress that really ependymomas, when they do recur, they can recur along the dura. So please, if you're, if you know the molecular feature, please look very closely at the dual margin, because it may be the first sign of recurrence. Intraventricular posterior fossa. So now we know clearly intra parenchymo supratentorial ependymomas and posterior fossa intraventricular ependymomas are genetically molecularly completely different. They may look similar on histopathology, but they are not related at all in terms of molecular genetics. So what are the two subtypes that WHO defined in the 2021 and the two subtypes are posterior fossa ependymoma type A posterior fossa ependymoma type B. So what are the type A versus type B so the type A looks like this they almost look like a Cp angle or lower posterior cistern tumor going out into the foramen lusca often, and this is called the PFA, or I call it the PF ependymoma asymmetric because it's off to the one side. And this is an awful prognostic ependymoma and unfortunately much more common in pediatric age group. So here's this subtype. The group B are the ones that sits usually in the midline kind of simulate the appearance of a major low blastoma group three because they are midline enhances. But remember, ependymomas are not reduced on diffusion. And this is what I call the posterior fossa ependymoma that looks like a ball. And you could see that they are actually very different one is in the midline one is asymmetric. And it also turns out that they're very different genetically so PFA. I call it a stand for asymmetric group B B stand for balls so let's look at the tumor types a little different more carefully so here's group a ependymoma type a asymmetric off midline. They tend to have more necrosis hemorrhage, and these are much more aggressive component, and they have a unique genetic and molecular marker, very different from PFA, a PFB, excuse me, and this is much more common in pediatric age Here's the PFB group, more like ball shaped in the midline, and not as much as necrosis, little less hemorrhage, the midline location and this is the type of ependymomas that tend to occur older children or adult patients. And their main presenting symptoms they tend to present earlier because of the obstructive hydrocephalus, and they're very rarely invasive or infiltrative at all compared to the PFA variant. So here's PFA the awful, much more prognostically worrisome type of ependymomas. And this unfortunately is much more common in pediatric age group. And they have a very specific molecular markers that are very different than PFB, the one that looks like a ball shaped in the midline posterior fossa. This is a ball shape. It's a better prognosis and affects adults a little bit more than pediatric age group. Moving on to second type of tumor that pediatric patients might get is the diffuse midline glioma. So these are tumors that are disorders primarily of a histone. And I think some of you know that each human cell DNA is about 1.8 meters, but thanks to histone, which winds down the DNA into 90 micrometers. So you could see that if there is a histone related abnormality, it could lead to just devastating tumors like this, particularly the histone age three K 27 M locus tends to cause tumors of this gigantic midline glioma. So comes the name of diffuse midline glioma age three K 27 M altar. So this is in 2021 version change its name and it affects these midline structures and they look like this. But just terrible tumor in the midline patient is sometimes very minimal symptomatically altered but it's a tough tumor surgical resection is not a possibility because they tend to involve the deep phylamid nuclei like this, it's just an awful tumor. Here's the six different patients with the same awful diagnosis and they're all even though histologically they may look very different, but they have histone age three K 27 M altered, and they tend to occur in the midline, and hence the name diffuse midline glioma. And some of the spinal cord tumor, and most of them now are called a diffuse midline glioma only when they could our pathologist can actually definitively identify the histone alteration particularly age three K 27 M. And here's an example of a spinal cord, we used to call this spinal cord astrocytoma or glioblastoma, and that's that's not wrong. But once they get a tissue and our pathologist look for this particular mutation. And that's how we know that this is a diffuse midline glioma age three K 27 M altered. Here's a patient I saw, but what does it mean for radiologists so we're not the one who's going to diagnose age three K 27 M molecular features. I want you to remember that these are really terrible tumors, they sneak around, they actually can spread all on the CSF, as if they are metastatic pineal blastomas, or measure the blastomas, and keep an eye on brain, any lesions in the brain cannot be ignored. Let me show you this is one of our patient for many years ago. When we first to our pathologist start to test for the age 20s. age three K 27 M after radiation therapy our radiation oncologist did a great job. Some of these enhancing tumor looks better. The flare looks better too but remember this original flare imaging of the brain. You're not sure whether this was really a real finding or is that a tumor, but five months later, you could see all of those areas are now nodular and patient recurred. And unfortunately, the real down the tumor that they couldn't control was not the spinal cord diffuse midline glioma. And this is the CSF ependymol leptomeningio spread of the original tumor. And this is how diffuse midline glioma age three K 27 M altered tumor behave on follow up imaging. So please make sure you get brain imaging to make sure we don't miss this very subtle lesions that show up on flare imaging alone. These areas may not enhance at all. So I'm just going to show you some of interesting molecularly conjoined tumor types. One is this. So many in Geoma and he mangeo pericytoma back in the day when molecular genetics was not the not a thing in terms of our daily conversation. I used to think that many in Geomas and he mangeo pericytomas were related and just that the he mangeo pericytomas were much more aggressive and angry or looking and destroying the bone. And it turns out that molecular genetics have proven that these two tumors are not related at all. But it turns out that solitary fibers tumor, which is a pretty rare extra axial tumor that we see but elsewhere in the body to is intimately associated with he mangeo pericytoma. Even though they look very different. This is what's called the grade one and he mangeo pericytoma is grade two or three, and they are related by this particular nuclear expression called stats six. So, WHO now lumps solitary fibers tumor and he mangeo pericytoma as a one single tumor type with a varying degree of aggression so SFT is usually grade one, he mangeo pericytoma is usually grade two and three. And he mangeo pericytomas can recurred anywhere else in the body, but both these tumor types have stats six nuclear expression, and that is really needed to make the diagnosis of SFT and he mangeo pericytomas in the brain. And with a circumscribed tumor, and if you look at these three very different histologically tumors have a nodule and assist nodule and assist nodule and assist. Allocytic astrocytomas ganglia gliomas and pleomorphic xantho astrocytomas, and some of the super tentorial of these three types share a molecular marker called be wrapped V 600 E mutation at our institution some of these patients with this particular patient they have a anti be wrapped therapy they're in clinical trials for, and we are getting pretty good response to therapy. They're again, three very different histologically, but image wise, I always knew they kind of look similar because they share this shape. They have a pattern of nodule and assist, and a lot of these super tentorial pylosidics ganglia and pxas have be rough V 600 E mutation. Very interesting. This is a paper from more than decade ago that already looked at over 1000 tumors. PXA pylosidic and ganglia and has shown that up to 60 70% of these tumors will harbor be wrapped V 600 E mutation. So these work has been going on for decades, even though it now just made it to the 2016 and 2021 WHO. The work has been going on for decades to come to the fruition and make it into the classification scheme. So molecular glioblastoma. Before I talk about molecular, let me tell you the classic glioblastoma is here eight patients that I've known in the past, have the classic central necrosis, very aggressive irregular remit of enhancement. These are all glioblastomas. And the classic GBM now have all these extra molecular genetic diagnosis that our pathologist is testing for, and some of them will be MGMT methylated and most some of you know the implication of MGMT hyper methylated GBM. These are actually the ones that respond quite well to team is all my chemotherapy, but there's all these other things that they test for. But the unifying one is that all glioblastomas are IDH one wild type. So there's no more what's called the glioblastoma IDH mutant does not exist anymore. All glioblastomas are IDH wild type. Now, what is molecular glioblastoma these have a genetic mutation the three component here to promote our mutation EGFR gene amplification trisomy seven and monosome 10. And on imaging, they do not look like the classic glioblastoma I just showed you. This is that 19 year old who was originally misdiagnosed as a stroke. If you look at her post con imaging, it does not enhance at all. And on DWI, it has really hyper intense DWI signal intensity and very low ADC. This is a path proven molecular glioblastoma with these three molecular alterations third promoter. EGFR trisomy seven and monosome 10. And this is new. This is a separate tumor, but nonetheless molecular glioblastoma are just as aggressive as classic glioblastoma. This young lady who received TPA and cerebral angiogram looking for a source of stroke elsewhere came to us. This was resected. This is a molecular glioblastoma. If you look down, patient has a second focus of additional tumor. And this is really bad news. Additional lesion. And that is the lesion that turns five months later into a Frank glioblastoma. And this is a terrible prognostic situation. Diffuse gliomas not to be confused with diffuse midline gliomas diffuse gliomas are what we used to call astrocytomas oligo dendro gliomas or all as oligo astrocytomas so we don't call oligo astro anymore diffuse gliomas are non glioblastomas and they're infiltrating either astrocytomas or oligo dendro gliomas. And these are the three molecular markers that define diffuse gliomas IDH being the most important. It signifies that it's going to be a lower grade. One P 19 Q chromosome number one chromosome number 19 locus of co deletion. This is a disease defining chromosome marker for oligo dendro glioma. ATRX is a marker defining astrocytomas. So at our institution almost all diffuse gliomas on imaging and initial pathology histologic analysis will undergo IDH for sure, and one P 19 Q of to make sure that it's not an oligo. And if one P 19 Q is intact, then they will go ahead and do ATRX to prove that this is an astrocytoma. So this is how it's divided diffuse gliomas is either mutant or wild type. I already told you wild type diffuse gliomas are defective molecular glioblastoma. In the muting variant, they will undergo one P 19 Q co deletion testing. If that is deleted, that's an oligo dendro glioma period and if it's one P 19 Q intact, then ATRX is lost. And that's an astrocytoma. The wild type. These are much more aggressive tumors. They if they have a third promoter EGFR amplification trisomy seven and monosomy 10. This is called the molecular glioblastoma and diffuse astrocytoma IDH mutant was newly graded. This is done grade two, three and four on WH for 2021. And this, the most aggressive variant is now called the grade four, but it has to have this particular molecular deletion called CDK and to AB homozygous deletion. Looking at the image and you could tell that this tumor is already trying to enhance and there's some areas of central necrosis, but doesn't really look like a Frank glioblastoma. You should think about grade four diffuse astrocytoma IDH mutant. Our institution has the capability to do this molecular analysis, CDK and to AB homozygous deletion. This confirms that this is a grade four diffuse astrocytoma IDH mutant. It's the king right now of determining the fate of a glioma, whether it's going to glioblastoma route or diffuse lower grade glioma it was discovered in 2008. And if there is a IDH mutation, it leads to this particular molecule to hydroxyglutarate accumulating. And it's much, much more common in lower grade gliomas. And in primary GBMs it's almost never seen, but in some cases of a glioblastoma that de differentiated from lower grade, you may actually detect this but in classic glioblastoma, it's never seen. So what does this mean that we at least have to know what IDH1 mutation, 1P19Q co-deletion ATRX loss means that I just already told you. So here's some of the tumors that this is from literature where T2 is super bright flare it becomes darker. This has been called a T2 flare mismatch. And this is not a 100% rule but this has been described as T2 flare mismatch meaning T2 is super bright flare gets darker. This has been a mullet imaging correlate of a diffuse astrocytoma IDH mutant ATRX loss, not a 100% rule, but you could assess at least guests before the surgery. Now, here's a patient with diffuse glioma. And how do I know it's more likely to be IDH mutant or wild type? The DWI is going to be one of the most helpful technique. You could see that the tumor kind of disappears in DWI. This is more likely to be IDH mutant, the better prognostic glioma. This is that patient with molecular glioblastoma, not enhancing, but look how very much aggressive the DWI looks, very reduced on DWI, very dark on ADC. This is not a lower grade tumor at all, despite the fact that it hardly enhances, has apparent circumscribed border. This is not a good tumor. This is a IDH wild type. This is molecular glioblastoma. How about this one? This patient has a serpiginous looking calcium and that calcifying marker is pretty good, not a 100% rule for a specific type of tumor. And you could see that the fluorescent in situ hybridization, that's what that is, shows you that 1P19Q co-deletion, you could see that there should be two pink and two green, our pathologist confirmed for me, but there's one missing. So this is so called a 1P19Q co-deletion, and this is a disease defining marker of an oligo dendro glioma. We don't do CT anymore to confirm the presence of calcification. We may do SWI, but CT to confirm calcification is no longer a standard of care practice. This is a patient. This is the last case I'll share with you. He came back in 2003 and this was discovered after he had a car accident and CT detected a low density lesion, so he was completely asymptomatic. I looked at this and I said, hmm, I'm not sure if it's tumor. So why don't we just seek a serial imaging. So he came back every year, and then it started to grow in about three years. Here's 2012. So nine years. This lesion has almost doubled. So I scratched my head and said, does lower grade gliomas grow this slowly? Probably could. So after 2020 patient called, emailed me and said, my, can you take a look at my MRI and I look at it and I look at from the 2003. So these are 17 years apart. And I explained to the patient, this is growing. We no longer can just sit around and do nothing. Neurosurgical colleagues, after hearing my brain tumor talk about 2HG, he said to me, why don't we get a 2HG MRS? And I thought, really an idea. So we brought the patient to our research scanner and our outstanding postdocs and PhDs there helped us to do a 2HG scan. And lo and behold, this patient within this lesion had an unmistakable, not an artifactual, a 2HG peak that you can only really see in IDH mutant gliomas. So this pushed us over the edge. Patient went for surgery, gross, totally resected. And this is a gross, totally resected IDH mutant 1P19Q co-deleted oligodendroglioma. And this is the first time that I was just stunned at how this noninvasive technique really helped us. So this is already three years ago now. And patient is doing very well that 2HG is indeed a marker for IDH mutant gliomas. And this happens to be an IDH mutant oligodendroglioma. So let's go over some of the cases that I showed you earlier. Patient has these three molecular markers altered. So what is the tumor type? I showed you this case already twice with this really profound, reduced diffusion, young lady who initially thought to have a stroke. Unfortunately, this is a molecular glioblastoma. How about these three tumor types? Pylocytic, ganglion, PXA, they all share sometimes this nodule and cystic appearance. What would be the defining molecular markers? The BRAF V600E mutation. How about this tumor that I just showed you? Calcification. So one P19Q co-deletion of an oligodendroglioma. So if you see a tumor and you happen to have a CT and you see calcium and putting together with MRI features, you could be with reasonable confidence, tell that this is going to be an oligodendroglioma. And if you're ever doubt, you can actually do a 2HG MRS scan and see if you can detect 2HG. That will be a slam dunk that this is an IDH mutant tumor. How about this one? Midline, just awful looking, expansile tumors in pediatric age group. What would be the histologic marker and diagnosis? This is diffuse midline glioma. H3K27M altered. These are histone-butated, really aggressive tumors. How about these two tumors? Extra-axial tumors, not a meningioma, but they are joined together. SFT and hemangioparasite tumor. What is the molecular marker that defines these two tumor types into one? That's the STAT6 nuclear expression. We already discussed the two different types of posterior fossa ependymomas. They're completely different genetically. One is called the PFA, the asymmetric type. The other one is called the PFB as in ball-shaped in the midline. A is asymmetric pediatric age group, and it's an awful prognosis. The one in the midline looks like ball-shaped. These are better prognostic of, and very rarely these will be infiltrated. So the two different types, subtypes of a posterior fossa ependymomas are posterior fossa ependymoma type A. The other one is posterior fossa ependymoma type B. B is better ball-shaped. PFA is asymmetric, and it's an awful prognosis. The four main subtypes of medulla blastomas, we already discussed this, the one that is off midline, kind of looks like CP angle or lower cranial nerve schwannoma. That actually is one of the best prognostic subtypes of the medulla, and that's the WINT. Hemispheric, multi-nodular, solid, aggressive-looking reduced undiffusion. That is going to be the sonic hedgehog. Two midline tumors, reduced undiffusion, one enhances more adequately, the other one is less enhancing. The one that enhances more is the group 3, and the less enhancing midline medulla blastomas are the group 4. So here is your main four molecular genetic subtypes of medulla blastomas. And imaging, not as good as pathology or molecular genetics or UCSF 500 gene panels, but it does a pretty decent job estimating what the molecular genetic abnormality might be of a given tumor. So with that, I would like to summarize that I gave you some highlights of 2021 WHOC and its tumor classification, but I also want to highlight that this is a field that's evolving every year. So the next version, which I am told might come out in 2025 or 2026, may actually have even more molecular genetic markers, so please stay tuned about that. And I showed you some update on structural and physiologic MR, like diffusion being super important to differentiate abscess versus glioblastomas, cellular versus less cellular tumors, and MR perfusion to detect recurrent high-grade gliomas, and spectroscopy methods that you can actually detect some of the oncometabolite associated with IDH mutin such as 2-hydroxyglutarate. And the imaging correlates of molecular and genetic profiles of CNS tumors, an example being the four subtypes of medulla blastoma, and two subtypes of ependymomas that you could actually guesstimate, not 100% rule, but with a reasonable confidence you could predict their molecular subtypes without ever touching a tissue or doing a craniotomy. Imaging is such a powerful technique and it's non-invasive, so we are in a very interesting and very important field. And as I said, molecular generic era of CNS tumors is here, it's only going to get more complex and advanced, and imaging must keep up with its pace. So with that, I thank you for your attention, and I will stop sharing my screen and take any questions. So I think we got some in the chat Q&A. Oh, where does ATRT stand in this classification? So ATRT stands for atypical teratoid raptoid tumor. This is actually a tumor that is super aggressive, but they're not medullos, they're not ependymomas, and they have their own very specific molecular marker called INI1. So our pathologists can specifically test that molecular marker to differentiate because ATRT can look just like medulloblastoma, can just like aggressive ependymomas, the PFA. So we actually have the, our pathologists have the power and method to identify ATRT. There's no formal classification of ATRT yet. But it might be next time, next version might include that. But like I said, the molecular marker that defines ATRT is this enzyme called INI1 mutation. If that is altered, that's an ATRT and nothing else. Next question. Since gliomatosis survivors obsolete now, what is to be labeled according to recent recommendation? This is an excellent question. Even though WHO said don't use this, frankly at tumor board I use this term because some tumors, entire hemispheres all infiltrated with non-enhancing, flare-bright lesion, nothing is enhancing, nothing is reduced. So now we use, I use the term, this is the gliomatosis survivors pattern of diffuse glioma. But these are usually, when they do go for pathology, they are IDH mutant tumors and they're not oligos. They're never, I've never seen gliomatosis survivors tumors that are actually oligo or 1P9TQ code deleted. So yes, it's obsolete and yet we still use this term. Is the term relay fusion still used or is it outdated? Another good question. No, it's the WHO 2021 came up with this ZFTA, some other name, but in our tumor board we still use the term relay fusion because everybody understands what that is. So the molecular genetic term has become much more complicated, but when I'm reading out with a trainee, we look at the electro-medical record and if we see the term relay, we know that we have to really look for odd places for recurrence. So that's really what's important here, that relay fusion ependymomas tend to occur at least maybe half a dozen cases that I've seen in the dual surfaces. So the probably next version might end up just getting rid of relay altogether, but we still use it and that is still, that term is still exist in the very complex pathology reports so you can search for it as well. Is there any other question? Let me just do that. There's a chat here. There's a chat about your painting if you want to say what painting that is. So this is Georgia O'Keefe. This is not actual real painting, but if anybody would like to donate this to me, I will be forever grateful. Just kidding. This is one of the most beautiful paintings that I've ever seen. This is a Georgia O'Keefe. I don't know the name. I think she just named it Flower. But I've been using this for the entire pandemic. And thank you so much. So there's no molecular marker associated with this painting. Is there anything else that I could answer for our outstanding audience? I think that's it. Dr. Chah, thank you so much for your lecture today. Oh, thank you. Thank you so much, everyone. Have a wonderful rest of the week and thank you for this opportunity again. Bye bye. Absolutely. Thank you and thanks for everyone for participating in our noon conference. You can access the recording of today's conference and all our previous noon conferences by creating a free MRI online account. Join us next week on Thursday, September 21st at 12 p.m. Eastern. We're featuring Dr. Sheila chef for a noon conference entitled peripheral vascular ultrasound Venus Doppler and challenging arterial cases. You can register for this free lecture at MRI online calm and follow us on social media for updates on future noon conferences. Thanks again and have a great day.