 The topic of my talk is WHO GYOMA classification update in genomics, what every radiologist needs to know. I'm going to start with showing a case over here of 55-year-old female presenting with word finding difficulty. The initial MRI shows a non-enhancing infiltrative signal of novelty involving the left temporal lobe as well as the left insular and if I ask my trainees usually what is the differential diagnosis and that's what they will usually start with you know it could be an infarct, infaction, encephalitis and then most of them will respond by saying because it's a non-enhancing tumor it could be low grade new plasma such as a glioma. However I want to emphasize this that this is what my impression should look like that I'm thinking that this is a mildly expansile infiltrative even though non-enhancing tumor but I do think that this is going to be an aggressive tumor such as an IDH wild type infiltrated glioma and the reason for that I'll come back during my lecture why I would write an impression like that and there's a reason to do that because these are aggressive tumors and you can see but then the tumors follow up you know the tumor is enhancing and it's showing much more aggressive imaging features now at that time you know the tumor was taken out partially and the integrated histological diagnosis showed that this is a WHO grade 4 glioma, IDH wild type right so that becomes an important aspect of why we should be dictating these cases in a manner where we have the genomic information also integrated into our reports. So the objectives of my talk I'm going to talk about some of the important genetic mutations in the gliomas IDH 1P19Q I'm going to talk about how WHO glioma classification update 2016 should be used in our reports imaging reports and how we can incorporate some of this molecular information for example IDH 1P19Q there are etc in our radiology reports. So if you look back the older WHO 2007 classification of CNS tumors divided the gliomas into astrocytic morphology and these were labelled as either grade 2, grade 3 or grade 4 or oligodendrodiomas which were graded as grade 2 or 3 and then there was a category which was ambiguous morphology which the neuropathologist will call as oligoastrocytoma whether in grade 2 or 3. So one of the biggest issues with this kind of histology based classification system was that it was very high in terms of variability no two neuropathologists agreed with each other most of the times and then also inability of this kind of classification system to predict clinical outcome and what has happened in the last 10 years or so with the discovery of IDH mutation there is a paradigm shift in classification and to some degree also management of these gliomas. One of the earliest papers in science showed that if you have IDH mutated glioblastoma there is a much better survival compared to IDH wild type glioblastoma. So even though you have a histological diagnosis is the same you know the glioblastoma but if you have IDH mutation these patients do much better than IDH wild type patients. The same concept was also shown in another paper in 2009 which showed that this beneficial effect of IDH mutation is not just seen in glioblastomas but also seen in anaplastic astrocytomas we have an anaplastic astrocytoma which is IDH mutated they are going to do much better than IDH wild type anaplastic astrocytomas. The same paper also showed that these IDH mutations whether IDH one or two they occur in majority of the lower grade gliomas that's one thing and this paper also showed that majority of the secondary GPMs they also have IDH mutation and only a small percentage five percent of the primary GPMs have IDH mutations. So majority of the primary GPMs are going to be IDH wild type bad tumors but then there is a small percentage of primary GPMs which are IDH mutated and they are the ones which are going to do better as far as survival. This is another paper in 2015 in NEGM from the TCGA group which looked at lower grade gliomas 2 and 3 and divided these gliomas based on the IDH mutations so majority will have IDH mutation and the next thing you look at is 1P19Q correlation and a big chunk of these tumors IDH mutated tumors also have 1P19Q correlation and these were the ones which were classified as molecular polybode anaplastic. If you have IDH mutation but don't have 1P19Q correlation you have ATRX and these are the ones which are molecular astrocytomas. Another important thing which this paper showed that almost 20 to 25 percent of diffuse lower grade gliomas which are grade 2 or 3 do not have IDH mutation. These are the ones which are IDH wild type and they are the ones which are going to do poorly and these are the ones which people have also called them molecular GPM so 3GPM and that's an important differentiation this paper showed that if you have IDH wild type lower grade glioma it's going to do much worse than an IDH mutated lower grade glioma. So what is IDH mutation isocytrate dehydrogenase? It's basically five genes encoding for three human IDH catalytic isozymes. Majority will have either IDH 1 or 2 but some cases also have IDH 3 mutation. Most of the times you can detect IDH mutation but even on the stroke chemistry detects almost 85 to 90 percent of the cases however to have 100 percent accuracy you need to have pyro sequencing or nexin sequencing to really detect the IDH mutation in a tumor tissue. IDH mutation occurs early in glioma genesis changes the function of enzymes and causing them to produce too hydroxygluparate at possible oncometablyte and not to produce NADPH which kind of protects the cells from some of the aggressive features. Exactly how IDH mutation contributes to oncogenic transformation remains controversial. So based on the IDH mutation you know you can divide a terit diffuse infiltrating gliomas into the ones which have IDH mutation as I said majority will be IDH 1 or IDH 2 mutation and the next thing to look at is if the tumor has 1P19Q co-deletion and these are your molecular band polygons and gliomas if they don't have 1P19Q co-deletion and they have ADRX and TP53 these are your molecular astrocytomas and more than 50 percent of adult diffuse gliomas they don't have IDH mutation these are called IDH wild type and they have third mutation and these are the ones which are your majority of them are your primary GBMs. I'm not going to talk too much about that but you know the next molecular marker you can look at is MGMT and if these IDH wild type tumors have MGMT methylation and they do a little bit better than the tumors which don't have MGMT methylation and as I said you know more than 50 percent of primary GBMs these are your IDH wild type tumors. The reason it's important to segregate these tumors based on these genomic markers because the survival in each group is fairly different you can see oligodendrogliomas surviving almost 15 years the median survival for astrocytomas is seven years whereas if you have an IDH wild type glioma the chances are you know the survival is going to be much less compared to the IDH mutated tumors and if you have MGMT methylated IDH wild type glioma they do a little bit better than IDH wild type unmethylated gliomas. Another two different groups which I'm going to discuss again these are mostly pediatric tumors but there is some overlap with adult gliomas so I'm going to discuss that is your DRAF driven tumors they have either DRAF fusion or V600 evacation and majority of these are your previously called WTO grade 1 gliomas and the important aspect is that these tumors usually have very good survival long term survival. On the other end of the spectrum there is another tumor which has been added to the WTO glioma classification update is your diffused midline H3K27 mutant gliomas and these again occur in young children or young adults and the important aspect of this tumor is that the survival is really poor and that's the reason you know it's important to discuss these tumors along with the adult gliomas. And the majority of these tumors these subcategories I'm going to show you how these are really clinical syndromes because the majority of them have different age groups and the presentation is kind of also more specific to the group. This is the paper which WTO published discussing the 2016 classification update if somebody wants to go in more detail but one of the things which came out of this WTO classification update was that now the neuro pathologist will have to have three-layer diagnosis so rather than the diagnosis just purely based on histology and grade now we'll have to incorporate the molecular markers for example if the tumor is an oligotin ruplioma and the grade is grade 2 but the neuro pathologist will also have to do IDH mutation status and 1P19Q and if this tumor is IDH mutated and 1P19Q correlated the integrated diagnosis should look like this where you have oligotin ruplioma on grade 2, IDH mutated and 1P19Q correlated. What is pre-science and pre-science is the fact of knowing something before it takes place for knowledge and I always believe that you know this is really very important for glioma imagers and two of the things I always tell the trainees to look at before you even look at the image and I call it pre-image is the age of the patient of course that's very important the reason it's very important for some of these glioma subclasses is majority of the young kids they are going to have either BRF or K27 mutant gliomas young adults majority will have IDH mutated or some will have K27 mutated diffused midline gliomas if you're more than 40 years old the chances are it's going to be an IDH wild type glioma so that's important and then the second thing which is really important again from a glioma molecular subclassification system is the clinical presentation as majority of the IDH mutated tumors which are occurring in young adults they're majority asymptomatic whereas BRF tumors which occur the WHO grade 1 tumors majority of them are occurring in young kids they usually present with seizures and majority of the IDH wild type gliomas presenting in older patients they usually present with subacute neurological deficit and that's these are important aspects to keep in mind before you even start looking at the images for example you know this is an NRI of a patient which shows a really ugly ring enhancing necrotic mass in the right frontal low but fusion imaging shows the tumor is showing remarkably increased blood volume the pathology, histopathology is very classic showing endothelial hyperplasia, dormerolite formation and then proliferation of the blood vessels however the most important thing to me is is the age of the patient you know this is an older individual and then the second important thing is what is the clinical presentation this patient has been having subacute neurological deficit as you can see one month history of left side weakness and difficulty speaking before the MRI was obtained and these two important things along with imaging and histopathology of course confirms that this is a pure blastoma but more importantly also confirms the molecular something I think as I'm going to discuss so a few years back majority of the discussion at our tumor board used to look like this you know there is discussion about MRI imaging perfusion, histopathology and then some of the immunosteins but now majority of the discussion at our tumor boards is kind of centered around this genomic analysis also known as methylation arrays and methylation analysis and copy number 5 we are the pneumopathologists are looking at how different genes and genetic pathways are either amplified or suppressed in the tumor tissue for example this particular tumor is turned out there was an IDH wild type glioma and GBM also was MGMT un-methylated EGFR was amplified as you can see from the methylation and the copy number 5 and this is really what makes the diagnosis as an IDH wild type glioblastoma grade 4. I consider my GBMs to be good, bad and mad and this is what I considered as a bad GBM and will go to the good ones you know when we reach the IDH mutation mutated tumors category so majority of the IDH wild type gliomas these are third mutated GBMs they occur in older individuals these are three different examples over here they are necrotic enhancing masses usually have low prognosis and more importantly presenting with subacute symptomatic neurological deficit another example over here is 43 year old female left frontal lobe tumor if you look at the tumor you know there is not much of enhancement there is non-enhancing flare expansion signal of normality there is some restricted diffusion in the medial part of the tumor and some increased blood volume also in that medial part of the tumor over the tumor looks very innocuous and it looks almost like a low grade glioma will be the patient undergoes partial resection and the histopathology comes back as anoplastic astrocytoma of age of grade 3 but more importantly this tumor also had IDH wild genomic markers so the IDH mutation was absent and that becomes really an important feature despite the tumor looking a non-enhancing and low grade on imaging and you can see within two months after resection and therapy there is a recurrence nodular enhancing recurrence at the margin of the surgical resection cavity which was resected again and this time the histopathology came back as an oblastoma and not just a grade 3 tumor and you can see the patient doing poorly 11 months after the initial diagnosis and 14 months after the initial diagnosis the follow-up MRI showing that the tumor continues to progress and the patient did poorly now this is the category we talked about from the NEGM paper that these are your lower grade gliomas despite having a great free diagnosis but these are the tumors which don't have IDH mutation and these are your tumors which are going to do poorly also have been called as molecular GBM or pre-GBM by various researchers and this led to WHO coming up with this C impact now update 3 which which categorize that you know that you know if you have a lower grade glioma for example a nanoplastic astrocytoma but if this is an IDH wild type tumor by genomic markers it actually should be labeled as diffuse astrocytic glioma IDH wild type with molecular features of the oblastoma and WHO grade 4 the important thing is that these are the tumors which are going to do poorly and they should be treated aggressively and almost treated as a gay oblastoma from the ghetto rather than treating them as as new grade glioma now moving on to the IDH mutated gliomas this is an example over here a well-defined bifrontal load large mass without any significant edema but the more important thing is that this is a younger patient 28-year-old male clinically all this patient has is headache so there is no real neurological deficit and this is you know the SWI and it's being on this patient showing that this patient the tumor is also partially calcified as you can confirm with CT scan also this partial calcification and turns out this is an oligodendrogyoma grade 2 IDH mutated more importantly 1B19 co-related hence this integrated diagnosis of a molecule and oligodendrogyoma now molecule oligodendrogyomas as we know they have median survival of 15 years this is another patient which was diagnosed with a non-enhancing tumor in the left and proximal region in 2011 did not undergo any therapy the patient was just an observation and you can see that the tumor is very slowly progressively increasing in size and even in 2018 even though it has increased in size there is no contrast enhancement and at that time the patient underwent resection and the tissue came back as oligodendrogyoma grade 2 IDH mutated in 1B19 co-related so these are you know tumors which are going to do rather well despite any therapy despite any radiation you know these these tumors progress rather slowly and have a longer median survival compared to an IDH mutated once these genomics upclassification system kind of became a little bit more clear in the last seven eight years we have been looking at some of these some of these morphologic imaging features for example we discovered a feature we started calling it as T2 fluid mismatch sign which was published in clinical cancer research in 2017 then the important aspect of this feature is that these are tumors which are occurring in younger patients clinically intact these are very homogeneously bright on T2 and and dark majority of the tumor is in fact hypo intense or dark signal on the flare images except the peripheral remover here non-enhancing tumors very low blood volume but more importantly what we realized is that these tumors majority of these turned out to be IDH mutated in 1B19 co non-co-related so these are your molecular astrocytoma so we published this sign as I said in clinical cancer research T2 fluid mismatch sign the important thing was the sign was 100% positive predictive value it's been validated in multiple studies by different other researchers since then and is a is a sign very specific for IDH mutated astrocytomas however the sensitivity is rather low this sign does not this this feature is not seen in all the IDH mutated astrocytomas it's only seen in 50 to 20 percent of IDH mutated astrocytoma so that's an important thing to understand so just to summarize IDH mutated gliomas they occur in younger patients majority of the time the patient is asymptomatic it's found either you know CT scan done for minor trauma or headaches majority of the times these are non these are non-enhancing or partially enhancing large tumors they might have have says also the prognosis usually is good and and again another important aspect majority of IDH mutated gliomas they occur in the frontal lobes so that's another important aspect I've shown you this oligodendropioma astrocytoma the T2 fluid mismatch sign not all the tumors which have calcification are going to be oligodendropiomas we have seen a few astrocytomas so the important thing is that you need to have one P192 co deletion status known to really classify them IDH mutated gliomas either in two oligodendropiomas or astrocytoma categories and majority of them are going to be lower grade grade two or grade three however there's a small percentage of higher grade or grade four gliomas which are going to be IDH mutated and and we discussed five percent of IDH five percent of primary GBMs are going to be IDH mutated and this is an example and we'll talk a little bit more about that in the next slide over here so as I said the IDH mutated GBM this is a patient younger patient 38 a right frontal lobe tumor and as you can see the tumor does show some central necrosis and peripheral enhancement but also notice that there is a large part of the tumor which is non-enhancing there is some mild edema there is mass effect due to the large size of this tumor and this patient underwent resection and the histopathology showed that this is a glioblastoma and the more important aspect is that this patient also had IDH mutation and GMT was methylated EGFR was negative all favorable genomic markers the patient underwent surgery temazolomide radiation therapy and DC Bax trial and this is a follow-up you know nine years after the tumor is the patient is doing rather well there's no recurrence seen and and that's an important aspect to understand because these are these are despite having a diagnosis of a glioblastoma this is a good glioblastoma to have and the important thing in this this particular patient is that these are your IDH mutated glioblastomas which are going to do well so a few facts about IDH1 and GMT majority of the lower period gliomas and only five percent of the GBMs primary GBMs have IDH mutation IDH mutation is seen more commonly in younger patients IDH mutated tumors tend to be far less aggressive than their IDH wild type grade matched counterparts T2 fluid mismatch sign if you see that 100 percent almost 100 percent positive addictive value or specificity for IDH mutated astrocythomas but the sign is not seen in all of the IDH mutated astrocythomas only seen in a smaller subgroup of around 50 to 20 percent of the cases and again another important aspect IDH mutation mutated tumors they have a preference for frontal lobe majority of the frontal lobe tumors are going to be IDH mutated tumors especially in younger patients now moving on to another category which is new in the WHO 2016 classification update is your diffuse midline gliomas which classically again occur in younger of younger adults of young children this is a 14-year-old presented with headaches and and you look at this white calamic tumor which is causing some massive fact and even hydrocephalus however the tumor does not show any enhancement there is no edema associated with this tumor and the imaging features look fairly low grade this patient underwent the follow up MRI two months later with increasing headaches and and now within two months the tumor has increased significantly in size more importantly also shows some enhancing areas and at this time the patient underwent a subject of the resection because of the deep location of the tumor the histopathology comes back as anaplastic astrocytoma of age of grade three and based on this you know the patient should be doing relatively well and should have a median survival of at least seven to eight years based on the data we know however more importantly this patient this young young child actually died within a year because of very bad leptomine yield spread of the tumor and and the important thing to understand why this happened because this is a tumor which was H3K27 mutant and has been now called as a diffused midline glioma in the WHO classification update and and and it's really an aggressive tumor to have and just to summarize H3K27 mutant gliomas they occur in children or young adults they are either in midline or paramidline classic example is thalamic tumors are in the brainstem quite a few of the DIPG we know they are now as H3K27 mutant gliomas the patients are usually symptomatic very poor prognosis majority will have some kind of leptomine yield spread and the median survival is around one year this is another example over here or an infiltrated tumor in the in the palms in the brainstem in a in a younger patient don't know this was a K27 mutant glioma and you can see the patient doing fairly well within a fairly poorly within less than a year you can see those tumor is recurring and there is a spread as well as leptomine yield spread of the tumor you can see enhancing tumor enhancement along the colus medullaris within the lower part of the fetal stack over here you can see the drop metastasis and leptomine yield spread of the tumor so these are the ones which are going to do poorly and as I said they occur in the midline we have seen K27 mutant gliomas even in the spinal cord this is an example of a younger patient who was suspected to have vertical sclerosis based on the peristesian and the numbness and the upper extremities and ended up having this MRI which showed a really ugly looking tumor in the spinal cord in the cervical spinal cord which almost looks like a GBM and turns out this was a GBM on histopathology but more importantly also was an H3K27 mutant glioma and was labeled as a diffuse midline glioma another example over here you know is not that common you know we we've seen a couple of cases of these tumors occurring even in the colus medullaris this is a again a younger adult had a tumor in the colus medullaris which was presumed to be a mix of apillary pandevoma how it turned out to be an H3K27 mutant midline glioma and a very aggressive tumor to have and that is why WTO glioma classification update in 2016 categorized these tumors as a separate category calling them as diffuse midline glioma H3K27 mutant why it's important to identify this tumor because this is a paper showing how pediatric GBMs correlate with genomic markers as far as their survival so you can see a smaller subgroup of pediatric GBMs which are ideated mutated doing very well compared to the ideated wild type pediatric GBMs but if you look at the K27 this tumor is actually so aggressive it's worse than ideated wild type bioglastoma so it's it's a category which has been now recognized as a separate category based on this mutation now the last category I'm going to talk about is your WTO majority of these are WTO grade 1 tumors occurring in younger patients, younger young children usually incidentally found or maybe presenting with seizures as you can see four different cases and majority of them either have cyst or cyst with a mural enhancing nodule and these are the ones we categorize them as WTO grade 1 tumors majority of biocytic astrocytomas and geniogliomas they have this PRF mutation and it's important to identify that because majority of these tumors are going to do really well and the long-term survival once resected completely is is fairly good. However not all DRF gliomas are lower grade there are some aggressive DRF driven tumors for example this one turned out to be a PXA grade 3 this one was histologically called a GBM grade 4 tumor but was a DRF driven tumor and that's important segregation. The reason it's important because our neurobiologist despite having a PXA grade 3 but the DRF B600 mutation an aggressively tumor of the partial resection did not treat this tumor with standards 2 which is Temizolovoid Invadation Therapy. In fact once we identify these PRF mutations now in any tumor these patients undergo a targeted anti-PRF and MEK inhibitor combination therapy similar to these PRF tumors in melanoma and lung cancer patients and this is a six-week follow-up where you see tremendous response to this kind of targeted therapy and that's an important aspect you know the reason to identify some of these genetic mutations is to have the ability to have some kind of targeted therapy for some of these genomic markers. Okay just to summarize a few take home points WHO classification glioma classification update 2016 is based on the molecular markers genomic markers as we see now that you know ideates and 1p19u correlation status they provide a better correlation with tumor behavior and biology than was being done by the older traditional histological classic patient system. IGH mutation is associated with practice revival and is seen in younger patients. IGH variety gliomas they usually have high blood volume I did not really show you an example over here but that's another important imaging marker we know now and as I said you know these molecular subclassic case classification what it has shown us is that these tumors really not just can be categorized as a broad category of gliomas these are actually clinical syndromes and the reason for that is majority of these molecular subclasses they occur in in specific age group and they have specific clinical presentation for example IGH mutation gliomas they occur in younger patients they're usually asymptomatic frontal lobe you know these tumors are usually large by the time they present and they have a lot of non-enhancing component they might have cysts and and as I've shown you we discovered this if you are mismatch sign which has been validated now with multiple other studies once we see that we know that this is going to be an IGH mutation non-co-deleted tumor or a molecular astrocytoma not seen in all the molecular astrocytomas but in a smaller subgroup because IGH wild type tumors these are the gliomas which are which are seen in older individuals presenting with subacute mineralogical deficit usually a large necrotic mass with hemorrhage edema and high blood volume and and these are the ones which are going to do poorly diffuse midline gliomas the H3K27 mutant gliomas offering mostly younger patients either young children or younger girls their midline paramedicline thalamus, brainstem, spinal cord, location and they usually have poor prognosis associated with lepromeningial stride whereas BRAF tumors these occur in young adults or children either asymptomatic or presenting with seizures multiple location of the tumor many of these tumors have cyst or assist with mineral nodule and that's an important aspect so I want to thank my collaborators and as this work you can see is not possible without a team effort and we have a great report including some of the eminent neurocologist, radiation oncologist, neurosurgeons and neurocologists at NYU thank you very much