ll right so we'll go ahead and get started and then let everyone continue to join uh so good evening everyone today it is my honor to introduce dr linda liao who really needs no introduction but i'm thrilled to recount her many accomplishments for the group anyways uh dr liao is currently professor and chair of the department of neurosurgery at the david geffen school of medicine at ucla where she is the co-director of the brain tumor center pi and director of the nci designated brain tumor spore as well she's had a truly remarkable career starting with her undergraduate studies in biochemistry and political science at brown where as a fun fact she was named by time magazine as one of the top 100 college students in the united states which i didn't even know that list existed but that's wonderful she then ventured to warmer climates receiving her md at sanford and a phd in neuroscience at ucla dr liao then completed her neurosurgical residency and a fellowship in neurosurgical oncology at ucla after which she joined the faculty with a rapid acceleration in her clinical and research pursuits she's had continuous nh funding for her work over the past 25 years which focuses on translational immunotherapies for brain tumors and biomarkers she's very well known for pioneering dendritic cell-based vaccines for glioblastoma and has really helped to launch the investigation of immune-based therapies for glioma in our field in a real testament to her contributions to science she was elected to the national academy of medicine in 2018 and her work has resulted in over 200 peer-reviewed papers chapters and a foundational textbook she's on innumerable scientific committees and boards leading our field both nationally and internationally she served as editor-in-chief of the journal of neuro-oncology from 2007 to 2018. also serving first as board director of the american board of neurological surgery and then becoming its first woman chair from 2019 to 2020. during this time because of course she was not busy enough she also received an mba from ucla going on to become a tenured professor of neurosurgery in 2007 and chair of the department in 2017. she's mentored dozens of medical students residents postdocs and junior faculty really leaving an indelible impression on the next generation uh dr liao represents the rare triple threat in our field and has further been a real trail blazer for women in neurosurgery and neural oncology dr liao we are just so delighted to have you here today well thank you so much for that very kind introduction thank you so much for having me um it's uh really nice to kind of see um kind of some of the some of my old friends at least on on the zoom name screen and um it's really a pleasure to be here um so uh today i'm just gonna i'm gonna talk about targeting immunotherapy-induced resistance and glioblastoma and um what's a little bit different about this is you know we've talked a lot about you know resistance to treatments um and there's certainly resistance mechanisms chemotherapy but uh now we're learning that there's actually resistance that occurs from our therapies basically our therapies are causing these tumors to uh create even further uh resistance to to uh to treatments and um and when i started in this field you know almost 30 years ago i thought we would have a cure to gbm by now but unfortunately we don't and i think the the complexities of this disease are just um uh you know getting uh just now getting to be better elucidated so um i'm gonna just uh let's see if i can move forward so oh these are my disclosures and then um so just as an introduction uh in terms of fda approved treatments for glioblastoma um as we all know you know radiation and chemotherapy is still the standard care for this treatment radiation has been around since the mid 70s and um really nothing happened for about 20 years um the only treatment available to our patients was radiation and surgery bcnu and ccnu was fda approved at the time but the clinical trials really didn't show any added efficacy uh over radiation so for about 20 years or so radiation was was the treatment um um actually for about 25 to 30 years that was kind of the main mainstay of treatment um you know in the late 90s uh then the gliado wafer was approved and then as was temozolomide and then in 2005 with the stoop protocol the standard of care for glioblastomas uh it became radiation with concomitant temozolomide and and really that's kind of been the standard of care for the last 15 years and there have been a few other things that have been fda approved before the system for recurrent jbms and then the tumor treating fields the optune device but other than that really nothing has been fda approved for uh for many decades for this disease and that's not for lack of trying um and as a uh you know uh a comparison uh in the last five years there have been over 50 uh cancer immunotherapy drugs approved for glioblastoma none yet for for other cancers but none yet for glioblastoma and then yet for brain cancer and and these are all the different cancers where there are now immunotherapeutic drugs that are available for patients um even pancreatic cancer i mean in some small subsets of pancreatic cancer there are some drugs that are fda approved for these very small subsets of patients um so what what is making glioblastoma so difficult uh to treat uh in you know particularly with uh with standard treatments but also with immunotherapy um and again it isn't for lack of trying uh you know as as you know many uh checkpoint inhibitors have been fda approved for other uh cancers and there have been several large clinical trials of uh checkpoint inhibition for glioblastoma and this is one of the uh you know the initial ones uh nivolumab versus specific patients with current glioblastoma we're currently a blessed german as you can see the the two survival curves are essentially the same um and this was published in you know just a little over a year ago um and you know in uh i guess true form we always hear about these trials from the uh the uh pharma companies oftentimes before they actually get published so that trial that uh that was published in jam on colleges checkmate 143 that was just alone and then there's also been checkmate 498 that's an ebola map with radiation and then the wall back with the volume and these are still um these are trials that uh you know unfortunately have not reached their uh primary endpoints but there are now you know different combinations of these types of checkpoint inhibitors with other treatments so um you know and for a while you know it's thought that perhaps you know immunotherapy you know at least checkpoint inhibition wasn't a you know a viable treatment for uh for glioblastoma um but then you know a couple years ago uh this paper from uh from our group at ucla as well as you know patrick lynn at the brigham um as well as some other groups that have kind of published similar studies showed that perhaps it's the timing of these treatments so this is a paper that showed that if you gave anti-pd1 inhibition neoadjuvantly meaning before surgery there was some increase in survival in these recurrent glioblastoma patients um and granted the numbers are very small we're talking only about 16 patients there's 19 patients in the other arm there it was a statistically significant difference um and interestingly the um this was one of those window of opportunity studies so we actually took the patient's tumor samples after uh the um administration of neoadjuvant uh anti-pd1 inhibitors and then looked at uh you know uh um immune signatures and there was an increase in gamma interferon uh signatures uh in the um um in the neoadjuvant group versus those who did not get neoadjuvant vampirism up suggesting that an upregulation of the interferon gamma pathway may play a role in this but um at around the same time there are other studies that came out that suggested that um if you give pd-1 blockade in a subprimed setting basically the t cells are not yet primed against your antigen you can actually induce dysfunctional uh pd1 positive cells and and and and anti-pd1 resistance so actually it's count you know basically makes uh makes the tumor um more resistant to immunotherapy if you actually uh give a checkpoint inhibitor without activating the t cells and that that's always been a problem with glioblastomas it's actually an immunologically cold tumor for the most part you're not you know many studies have shown that although yet there are some t cells in glioblastoma there aren't very many and perhaps the um the results that we're seeing with neoadjuvant treatment of uh of these uh tumors is that for those patients that have t cells in the tumor you are perhaps able to uh induce activation or of those t cells against the antigen however this would suggest that if you have a subprimed t cell the the use of a checkpoint inhibitor may actually make your um tumor even more immune resistant um and then some work done by uh rob prinzen in our group at ucla what he found was that in glioblastoma patients uh the expression of pd1 by t cells uh in in in these tumor samples actually uh reflect uh exhausted t cells and this is just a study looking at different uh markers of t cells they're the activated t cells they're the um um memory t cells the activation t cells and the exhaustion t cells and in glioblastoma patients uh both in the um well mostly in the tumor infiltrating cells most of those cells are are exhausted or in that path to exhaustion so so the the question is well how do we uh perhaps revive those t cells that so that they can be tumor specific and actually attack uh you know attack these tumors and and really drive this to the memory t cells which is actually what we want um and uh patients do have memory t cells in the periphery but these t cells are just not getting in uh into the tumors when they're you know infiltrating into the tumors um so um to that regard we went back and looked at um the basically the tumor samples from the patients that were enrolled in this in the neoadjuvant pd1 blockade trial and looked at their tumor infiltrating t cells and then comes to see well what was found was that these patients um you know perhaps had an increased survival there was an increased interferon gamma signature in these tumor samples and then uh the the question the next question was well what's going on with those t cells so we uh so from the resected tumor tissue um uh we isolated the cd45 cells and then did scitov mass spectroscopy or single cell genomic sequencing and what uh and muscle's work was done by uh you know rob prince and uh and his group uh along with uh the people working um in uh in our labs collectively um and this actually was part of our spore project um as part of our ucla brain cancer sport but what we found was that the um neoadjuvant anti-pd1 treatment um it did increase the uh perhaps the number of t cells but these t cells were actually uh progenitor like exhausted t cells they actually had exhausted t cell markers uh for instance in this cluster this l4 cluster here as well as these there's an increase in t red cells which are also immunosuppressive t cells so um i guess as a a feedback mechanism that these tumors kind of uh probably would were um expressing was that when we treat it with these anti-pa1 inhibitors these teeth it actually drove these t cells further to uh exhaustion and that could be one mechanism for uh for resistance because if you have t cells that you have a number you know small number of t cells in the tumor perhaps you are increasing the the activation of those t cells but then you're actually um over stimulating them to a point of exhaustion that they can't really attack the tumors but actually the the the bigger immunosuppressive effect of antipedin1 therapy or immunotherapy prior um uh to uh t cell activation is this uh infiltration of immunosuppressive macrophages and myeloid cells so what we found was that anti-p1 treatment in a population of patients induces upregulation of immunosuppressive macrophages and myeloid cells so even though you you're getting more t cells in and perhaps the t cells are more activated you know the t cells may be getting exhausted and then on top of that you're recruiting this whole population of immunosuppressive cells that are actually coming in to uh fight the t cells so there's really this kind of battle going on in the tumor microenvironment that is actually induced uh by applied by our treatments and so i think in order to really uh get effective treatments for glioblastoma we really need to understand what our treatments actually cause and and the timeline for that you know when when when the tumor you know when certain treatments should be given to mitigate some of these uh immunosuppressive responses so how do you get more t cells into the tumor so because that's really uh and i still believe that is the first step the problem with um and probably one of the reasons that immune checkpoint inhibitors don't work as a single agent for glioblastoma is that we don't really have a lot of t cells in the tumor and even if they're in there they're exhausted um so you really need to recruit new t cells from the periphery because um as i showed in the um in this slide here oops the glioblastoma patients do have memory t cells and activated t cells in the periphery but how do we get them into the tumor um so uh one of the best ways to actually get t cells into the tumor is is through active vaccination um and uh and this was actually um work that i you know did as part of my you know my my first uh koa award many many years ago and what uh we were the first group to find that with vaccination we could actually get t cells to cross the blood-brain barrier and go into brain tumors um i didn't realize that at the time but apparently no one else had found this uh before this point um and it was because at the time it was thought that the uh the the brain was immune uh immune privileged so that there was no uh way to get activated t cells or at least tumor specific t cells to migrate from the periphery into brain tumors or at least it's not that that doesn't happen now of course people know that it does uh and it actually happens in several different pathologies besides brain tumors um as well but uh but now you know but now i think we understand that there are ways to get t cells into tumors either by active vaccinations sometimes you know oncolytic viruses um do that as well as you know you could actually just give adoptive therapy of of car t cells or other types of t cells directly into the tumor but the first step needs to be getting those kind of fresh non-exhausted t cells into the tumor um and then once uh and the those initial studies that we did using dendritic cell vaccination for instance uh did show that t cells got into the tumor and then uh we did see some increased survival uh in uh patients in the early trials um particularly in patients with uh the mesenchymal subtype of glioblastoma and uh and these studies subsequently led to other phase one phase two and and uh and phase three uh clinical trials which um i'm not gonna talk about uh in this talk but what i wanted to focus on were the resistance mechanisms that we learned uh from from all these studies uh over the years um so as i mentioned you know this was a paper we published 10 years ago what we found was that when we used um dendritic cell vaccination the mesenchymal gene expression signature was associated with the increased tumor tumor infiltrating lymphocytes and we often see this in you know increased contrast enhancement after treatment which subsequently went away on its own and in these patients when we were able to get the post treatment specimens we saw um you know this kind of large infiltration of cd8 cell t cells which you know what wasn't uh visible uh or we didn't see in for instance the pro-neural subgroup of glioblastoma patients so this mesenchymal subtype tends to be more immuno-responsive you know these uh it tends to uh react more to immunotherapy in the sense that you could get t cells into these tumors so um we then went to look at well what kind of you know what can we do to make glioblastomas more immune responsive to you know help turn these kind of relatively cold tumors to more hot tumors and uh some adjuvants that could be used are things like toe like receptors tlr's recognize pamps pathogen-associated molecular patterns and it's really an uh you know an important uh element in the innate immune system and there happened to be you know for instance uh fda approved tl7 agonists uh there's a drug called uh amico mode and recyclamod that's been fda approved for warts um and then tl3 agonist uh there's a drug called poly iclc that's been used in the past for glioblastoma and clinical trials it actually was negative as a single agent but um but i think in combination with vaccination or with some sort of t cell activation signal these uh tlr and agonists could be of benefit so we did a very a small clinical trial to test that we looked at um a group of patients who just got that well that well all these patients got dc vaccination so we looked at a group that got dc vaccination plus placebo dc vaccination plus poly iclc which is a tlr3 agonist or dc vaccination plus a mechamod and what we found was that um with the tlr agonist the poly iclc there was an increased um uh effector uh effector pd1 positive t cell response so and uh and there was also an increase uh you know number and and activation of uh t cells uh suggesting that that this particular uh uh adjuvant was was driving these cells more towards activation and less towards the immunosuppressive phenotype um and we also looked at this with a single cell sequencing again looking at the t cell cluster showing you know drive of these t cells to active memory t cells as opposed to the um the more kind of naive anergic t cells and we did uh gene expression profiling and then and showed that this particular group that got treated with the poly iclc actually both the groups that got treated with poly iclc as well as were recycle mod they did show some changes in myeloid cell differentiation uh gene expression patterns as well as lymphocyte gene expression patterns more so in the poly iclc group than the um than the recipromite group and they that these these adjuvants really helped to elicit a very strong type 1 type 2 interferon response um and then what was even more interesting was that uh with poly iclc actually the patients live longer um this particular group albeit the the numbers are very small again um there but it was statistically significant uh the group that got dendrite cell vaccination plus poly ice celsius uh had a 50 survival rate and now we're you know the majority of these patients are reaching 100 months uh and uh with not only survival but but really um no tumor recurrence so it does suggest that there is some um added benefit uh to adding these um these adjuvants to to t cell activation signals and as i mentioned i think one reason that single agent checkpoint inhibitors aren't working in glioblastoma is because the t cells are not getting in and what we do know is that uh at least in in our animal studies and and several other studies is that you do need an activation signal to get the peripheral t cells to go into the brain tumors um sometimes we already see uh glioblastomas with t cells so there probably was some activation signal that got those cells in there but for the most part the majority of these tumors don't really express a lot of activated t cells but the dendritic cell vaccination is able to get these t cells infiltrated into the tumor and these are just uh schematics of of what what happens um if you just use the anti-pd1 antibody alone we don't really see a significant increase in t cells in the tumor if you use the dendritic cell vaccination we do see increased t cells into the tumor but uh but in terms of if we combine the two then we we see not only increase in numbers but also the increase in uh perhaps percentage of activation uh signals and activated t cells um so so the the thought is that uh you know again pd when blockaded alone uh did not show a response in our uh animal models uh if we gave the dendritic self-vaccination alone we did see a response to some of the animals but it was uh not in all of the animals and uh part of that was because the the percent of t cells that were activated was relatively low compared to when we actually used both agents alone however what we also found um was that in when we used um dendritic cell vaccination plus pv1 inhibition uh again you get more t cells into the tumor but then you also get a uh a yes a compensatory migration of these immune suppressive cells uh that are actually coming in from the you know from the periphery to try to basically suppress that immune response that is being activated by the activated t cells plus checkpoint uh inhibition um so uh so our uh you know initial uh clinical trial that we started uh as part of our spore five years ago was to look at um basically the effect of neoadjuvant pd1 blockade with dendritic cell vaccination this is a group of 20 patients in each arm and then uh we the the end point this is a window of study uh window of opportunity study so then point was really to look at the tumor tissue uh after the neoadjuvant blockade and then also look at overall survival in the group that got a dc vaccination plus uh pd-1 blockade versus the group that got dc vaccination plus placebo um and uh this this study is still ongoing so i don't have the results of this study yet we actually um after all the pre-clinical studies and ind enabling studies who got fda approval in november 2019 uh the trial opened in 2020 but had some fits and starts because of covid um but we've already enrolled 20 patients we're halfway through so anticipate we should have the results of that study uh you know in the next year or two but one thing that was interesting you know in some of these patients um was that just like we saw in our animal studies we saw an enhanced immune response with autologous tumor lysate and uh and the antip one uh antibody but uh we we saw this uh very actually very profound immune response in in some of these patients very similar to like the um uh the side effects that you see in people who have gotten you know cartier therapies um so you know this is an example of one particular patient he got the uh autologous tumor lysate dendritic cell vaccine plus pembro uh as part of the trial at the window of opportunity trial we did surgery took out his tumor uh he you know he clinically did better we were measuring uh crp and other kind of you know immune or inflammatory markers that actually went down and then um subsequently uh he he had uh injections of of his uh subsequent injections and every time he had an injection these markers went up uh to a point where um in this particular case he was getting so much inflammation that we had to treat him with uh an il6 inhibitor of toaster leucine and then eventually we were able to control the inflammation with bevacizumab but in this particular study this is actually a patient where his post uh you know pembroke vaccine treatment caused so much inflammation we had to go back in and resect some of the tumor tissue um and what we found was uh was actually quite interesting um so following just neoadjuvant pd1 uh inhibition we saw increased uh t cells this is a cd3 t-cell marker the red dots are t cells um but then when we you know when he had the vaccinations with the pembroke um pembrolizumab he had increased t cells and this is probably what led to some of that inflammation that he had that required subsequent uh debulking of the tumor so but this this actually probably was not the the uh i guess the reason he had so much inflammation it was probably because in addition to the t cells coming in um he had uh after immunotherapy there was a huge migration of immunosuppressive myeloid and macrophage cells so even though we there are the red cells the t cells getting into this tumor there is this there are these immunosuppressive cells that came in and really there's this battle going on in the brain between the tumor cells and the t cells and then this new population of of myeloid cells that were actually induced by the immunotherapy because we don't actually see this uh as much in people who didn't get the neoadjuvant pd1 inhibition so um so this is kind of uh you know what what it looked like so this was the initial surgery where you know the patients only got the pd1 um antibody and uh and there was you know some increased uh um the tumor cells are in blue the the t cells are in red and then the green cells are the immunosuppressive macrophages and then after the combination treatment you could see you do get more of the t cells which are the the red cells but many many more of the green cells so this this whole lesion was actually being overcome by the these green cells the green cells the immunosuppressive macrophage has actually outnumbered even the tumor cells uh and the t cells so um and and this is you know this is these green cells are fighting the the red cells and over time it allows the blue cells the tumor cells to outgrow this population so that led to kind of our the design of our upcoming clinical trial because i think it's it's not going to be enough to just you know what we do know is it's not enough to just use a checkpoint inhibitor you need to get the t cells in so uh so that's why we combine the activation signal the vaccine plus you know to get the t cells in plus the checkpoint inhibitor but then as a result of that we get this uh compensatory or response whereby um immune effect that combination of immunotherapy creates this uh huge influx of these myeloids um macrophage tumor suppressor immune suppression of cells so this next uh trial is is really a combination of dendritic cell vaccine plus poly iclc to do activate that danger signal that innate immune response immune checkpoint inhibition to to really block the immune checkpoints and then csf inhibition which actually is an agent that blocks that those screen cells those myeloid macrophage cells and that's kind of the the upcoming trial design and uh this is you know the the uh the design where we're gonna try to um add add those uh other components to to this uh uh current uh pd1 uh plus um uh dendritic cell vaccine trial um one other thing that i think is lacking in the field is uh is the ability to monitor response or or to actually have biomarkers to determine which patients will have a response um and one thing that we've been looking at uh over the years is the um i guess the use of tcr uh you know t t-cell receptor sequencing and the old overlap of tcr sequencing um uh antigens you know the antigens for which these t cells are responsive to um to look at that overlap between the peripheral blood and the tumor because if if you can imagine if the there is the the peripheral blood has antigen specific t cells that actually can get into the tumor that match the antigens within the tumor then the hope is that there will be this kind of perpetual replenishing of t cells from the periphery that could get into the tumor and attack the tumor rather than have relying on the infra tumor t cells which uh as we have shown are for the most part you know tired and exhausted and get even more tired and exhausted when you give people inhibition so um so this is you know a potential blood marker that we could uh potentially monitor uh over time to see if there was any response to immunotherapy in our prior dendritic cell vaccine trials what we did find was that people who had a high overlap um pre-op and a high overlap post-op for the most part they all have very good responses meaning long-term survivals of of more than five years or so and people who had low overlap before treatment there was a group where they they were able to have high overlap after treatment um some of those had had very good responses and some did not and we're still trying to figure out what dictates uh the difference between those that respond and those that did not and then if you're you don't have any overlap either before or after treatment for the most part we're not seeing the t cells uh you know expanding that will will actually attack the tumor antigens in the tumor and those tend to have poorer survival um so i'm just going to spend the next you know a few minutes just uh talking not so much about the science or and and the immunology um but but also about but more so about the the clinical trial design and uh and this is actually the other reason i think we are we don't have any fda approved treatments for immunotherapy right now um as i mentioned the there's the science behind it because there's the i think the need to uh use combination immunotherapies and combinations in the right order and the right uh rationale uh for using these treatments and uh and that's certainly a big part of kind of the um the bottleneck that that we're having in in the field but i think another um a problem with uh with getting these these you know treatments to patients is really the clinical trial design and as as many of you know um there are challenges to randomize clinical trials uh particularly in glioblastoma and it's so hard to get now patients who want to be the placebo arm of any trial uh you know and uh and that you know becoming more and more difficult uh you know for for these larger phase three trials because you know most patients would rather go to a phase one or phase two trial where they're actually guaranteed the treatment even though it hasn't shown uh efficacy in those early phase studies so patient recruitment and retention is difficult uh there is a lack of active treatment you know um the lack of an active treatment arm in the control arm does you know have some ethical uh issues as you can imagine to really keep patients on these trials when they you know that they're only getting the placebo um you know and glioblastomas are rare they're it's a heterogeneous disease compared to other cancers and the numbers uh that we need to do these trials are are just not there and then given the heterogeneity and the sub stratification uh that you know you need to uh have in order to show efficacy even in small subgroups uh that that's just incredibly difficult and not only is there stratification of the tumor there are also characteristics of the host and the host response that you know i didn't really talk about but you know the host um also plays a role in this heterogeneity and then you know as i mentioned we have of the lack of biomarkers for for response uh to to a lot of these treatments and then by the time you finish a trial there's rapidly progressing new knowledge uh that could change the the the course of how you evaluate um treatment response for instance what you know and uh a lot of the way we evaluate treatment responses by imaging and uh and as you know that that has changed over time as well um so there is more of a movement towards the use of external control arms and uh this has been you done in other fields outside of brain cancer and you know for other cancers and and i think it is gaining popularity and it's something that i think our field really needs to consider uh more deeply simply because of the problems with you know getting uh good data from randomized controlled trials and keeping patients on these trials so um and this is just a schematic uh that that was uh you know that shows how some you know some of these trials can be designed whereby you use external data with some propensity matching and you know get your kind of virtual treatment versus control arms um there's there's actually some guidelines in terms of you know quality checks for what what would serve as kind of good data sources for uh selection of external control arms and many of these um you know have to do with like for instance getting data from large well-controlled randomized clinical trials because the database and the data collection is is relatively good in these larger kind of well-controlled trials you know you have to look at the similarity of data sets some synthetic control methods um as well as the relevance and reliability and um and i'll talk to touch a little bit about that on that a little bit later but um but so so with that in mind i mean if you just look at the um for instance the last you know seven seven to ten years um there have been several large randomized control trials that have been done for newly diagnosed glioblastoma and many of these trials have control arms so so this is the so this is the control arm in in these studies so there's you know in this particular study there's 411 229 it if you add all these control patients all these control patients got the exact same thing they got radiation and temozolomide the stupid protocol so you're you know if you just look at these controlled patients you already have 1300 control patients who got the exact same thing and if you look at the control the kaplan meyer curves of these trials they they absolutely overlap um so so you know one question is do we need to do another you know do we need to have control arms for all our future clinical trials will enroll another thousand thousands of patients onto the control arm if we have a robust data set that shows pretty much the same uh survival among all these patients who got temozolomide and radiation um and uh you know one uh thought was that well how uh i guess how how rigorous is the use of uh of external controls would a study that a a true randomized control study that was negative be erroneously um i guess statistically tested as they or come out as positive when we use these external control arms so this was just an exercise that showed um you know for instance the um the cell death study with rinda pippa matt and temozolomide using that as the experimental arm and testing that against the all these external controls that trial actually turned out to be negative as you know similar to what the con you know the controlled study showed as did dose dense temozolomide the only trial that actually was positive in this comparison was actually the um the tumor treating field study which actually was also positive when uh when that um when the actual uh randomized control trial was done um and then you know when we think about randomized control trials as i mentioned there's the issue about you know having patients on the placebo arm um and uh and then the ability to to really kind of you know uh rigorously do this in a in a such a heterogeneous population of patients such as those who have glioblastoma um another consideration is that well are these randomized control arms really um reflective of the real world you know population you know what we do in the real world for our patients and uh and one thing that uh that knows you know in the data you know from pulling the data from these uh trials that have been done over the last decade or so the majority of these patients that were enrolled in these trials you know up to 94 i mean 87 was the average 87 of these patients were white um so we really are lacking in diversity uh in in our you know randomized controlled trials at uh in neuro neuro-oncology in our academic centers so that's something that i think we really need to think about more deeply i mean is one um you know our randomized control trials the best you know way to test new therapeutics even if it is scientifically is it the best representation of what we will actually do um you know clinically for for our patients because the representation in these trials is really not reflective of what our real the real world is um at least not not uh not in los angeles and new york i think so in conclusion um you know things to think about uh how to accelerate the translation uh and clinical development of nucleoblastoma therapies i i think we need to develop better immunocompetent animal models that better recapitulate tumor and host heterogeneity because heterogeneity is still a very big problem um with uh with the studies that we do in this for this disease um we have to consider timing sequencing and the role of surgical resection in combination immunotherapies and it's not just a matter of throwing everything together at the same time i think there is uh you know some thought that needs to be put into how the how the immune system works with the how the immune response works and what you know how to time the the combination treatments to best fit that we need better biomarkers of response whether it be tumor or blood or csf biomarkers uh oh and or and or imaging biomarkers that could actually look look at response to treatment or stratify patients that would respond to certain treatments uh non-invasively um and then you know these previously large randomized controlled trials have been done in glioblastoma it would be good if these uh trials uh release their data um so that it could be mined for uh external controlled propensity matching and uh and also i think um the you know more of a movement towards the use of external control arms for registration on our trials or even going on to phase four trials uh with real world efficacy as opposed to having this kind of step-wise progression from you know from very strict randomized controlled trials that um you know could get to publications and fda approval but aren't really being used in the real world because they're just not either not practical or not reflective of the population that we're serving so with that being said um i just wanted to end with um kind of a an overview of our brain cancer spore at ucla um the theme of our overall sport is targeting resistance and what i talked about today was just project one it was this project uh that's led by myself and rob prince where we're looking at immunotherapy resistance and and how to target the tumor micro environment to improve immunotherapy resistance we also have a project led by tim klause and dave nathanson that's looking at resistance to chemotherapy and targeted inhibitors and how actually by giving targeted inhibitors it leads to uh this metabolic vulnerability state um and also uh kind of but this vulnerability state that could be uh i guess attacked by for instance bclx1 inhibitors but there are two blocks to apoptosis and the goal really is to block both of these signals uh when when uh when patients are treated with these inhibitors and then our third trial our third project uh is a a study on rate uh radiation resistance um this is led by uh frank pajama and it's really looking at you know why radiation fails and one thought is that it actually induces uh this population of glioma stem cells that actually uh then becomes you know non-radio-sensitive and uh and there's some you know uh interesting clinical trials uh based on pre-clinical data that we're starting uh with the use of dopamine receptor antagonists and statins for radiation resistant tumors so with that i just wanted to say thank you for your time uh and uh thank you for inviting me to to give this talk it's really an honor to be here hi sander that was great a tour de force for sure um i'm i'm so fascinated with the the the non-randomized clinical trial thing um what does that mean for fda approval if if you went that route well um i you know i i don't know i mean i think the fda uh you know uh hopefully would be more open to this you know i mean have they been for other cancers or is there a track record of them allowing for that type of data and giving cms approval to move forward to develop anything like that yeah not uh not final approval i mean i think for there have been some conditional approvals um i think uh other regulatory agencies are more open to this the european ones the canadian ones um fda is a little uh different in the sense that um they they don't you know for those of you who've done you know kind of submitted things to the fda it's almost like you submitted and if you don't hear anything for 30 days it's good right um so so you know we submitted things and you know there wasn't a negative response but but i think um as far as other treatments um you know i i don't know off the top of my head anything that's had you know final approval but but i think we're at a point at least in glioblastomas where you know i'm not sure how we're gonna get anything approved if we have to do big you know hundreds of patients randomized trials yeah i agree hi linda it's guy that was great thank you hi guy i had a quick question for you maybe i missed this because i i i had to be back and forth with something else but in your in your pv1 trial that you're running are you i i saw that it was for you know for for resectable recurrent tumors but is there any sort of screen you know because obviously there's been a bunch of papers looking at which tumors may be more or less pd1 sensitive and so how are you how are you pre-screening to up the ante on on potential responders yeah yeah um right now we're not pre-screening um because actually we don't know exactly which biomarkers are you know are are gonna be the the uh you know the potential responders um i i my personal opinion for for immunotherapy in general i i do think the um the the mesenchymal subtype are probably more are going to be more immunoresponsive and they've been other groups um you know mark gilbert and amy humber found out as well and and i think perhaps also the people at the mayo clinic but the problem is i i don't think there's any consensus as how to test for response you know whether it's white imaging whether it's by immunotypistic chemistry sequencing you know there's no test yet that we have thank you hi hi linda this is fabio very very nice talk with you in this issue of you know the pd1 inhibitor when given before you have a primed or or in a subprimed tumor do you have any idea when after for example giving the dendritic cell vaccine this tumor will be primed that then you know will be the best timing to add yeah good question um in our animal studies it's about it's about two weeks because i mean theoretically you want to get t cells in so hopefully i mean if if t cells are going in and they're at you know and you know we we could take them out and be sure they're antigen specific if the t cells are in then theoretically they should be primed and it takes you know it's probably about you know two weeks in humans but we haven't done the study where we go in and biopsy you know we haven't taken out the tumor in the human to really know that another question what what do you think is the ideal um nice model for just studying immunotherapy because one of the issues of gl 261 for example it seems like everything works yeah i think that's one problem with immunotherapy there's just no good mouse model right um because you need a you need an immunocompetent um model um i mean the best kind of models for heterogeneity are the you know the xenograph models but they don't have an immune system so um so i think um that's why we you know we rely a lot on these you know window of opportunity studies where we actually you know just just test test these agents you know in humans and take out the you know the tumor and and and actually i think we probably need to move that into the newly diagnosed setting because you know a lot of trials now are done in the recurrent setting but um actually i think i think immune therapy works much better in the newly diagnosed setting before you know everybody's kind of failed all of their treatments um and but uh so i think one of our our next sport trials is actually to you know do do the the current trial but you know with with the newly diagnosed patients but with these window of opportunity types of designs so for instance we would give the immunotherapy um you know uh be before or or after radiation and chemotherapy and then have a you know maybe a 12-week set point where we go into and see what's going on thank you so much for such a wonderful and comprehensive talk um just a couple questions one you know curious about your thoughts on car t cells and sort of the the promises and pitfalls and the failures that we've seen you know we know just basic of course in what you're saying the activation is important we seem to obviate the need for that with these activated cars um there you know it seems like maybe the predominant issue is the myeloid suppressors i was just curious if you think there's still sort of hope hope there in that field um as an immunotherapy and then secondly with respect to the radiation resistance that we seem to be incurring is there do you think that there's some possibility that you know in the recurrent setting perhaps for methylated patients that there's a role for overcoming that with for example you know stereotactic radiosurgery you know where we can overcome some of the radio resistance is there you know any thought at um combination therapies uh with that yeah you know great questions um so as far as car t therapy um yeah we actually have you know one of our developmental projects in our sport is actually a uh by specific car t um and it actually is you know it's basically a car t to target the the antigen target but also it um it releases uh you know something to target pgf beta so so basically the the immunosuppressive signal that occurs when you actually put in adoptive t cells so so i think you know i think there are kind of innovative ways to to design uh you know car t cells that could hopefully get like because what you need to do is not just target the tumor cells you have to fight off the other stuff that comes with targeting yourself so i think that hopefully you know may have some promise it's still you know as you know it's still not um great for for solid tumors you know it's shown you know a lot of you know efficacy for for you know blood tumors but but where where that will be for solid tumors you know i don't know but i think it you know you need to move beyond just single target car t's if we're gonna try to tackle gpm and then in your in response to your answer regarding radiation um yes i think actually as i was mentioning you know what's the third project in our sport um is actually uh from one of our radiation you know oncology colleagues and it's looking at um so so what he did and i didn't get into this um but um what he did was he looked at cells that uh actually were you know non-uh basically grew out despite radiation basically we're radiation resistant um and if you just look at cells themselves they should you know gpm is pretty radiation sensitive but then why can't you kill you know why can't you kill him in the brain um so he basically took took those cells and ended this kind of screening panel to see what what actually prevents these cells from transforming from a radio-sensitive cell to a radio-resistant cell and interestingly um the agents that came up the most often were dopamine antagonists um so uh so actually uh you know we're starting a trial of quito pain which is seroquel for uh you know for uh for these tumors and um and i think there could be a role where for instance if you had a glioblastoma patient who got radiation or you know and then failed there could be a role where you actually re-irradiate and then give these kind of um agents like like i said you know what one what particular classes are the dopamine antagonists to block that transition from the uh immuno you know the radiation sensitive to the radiation uh resistant cells um because interestingly the the use of this drug like the the dopamine antagonist it doesn't work by itself you actually have to do it with radiation at the same time which i i think is um it's an interesting phenomenon interesting thank you so if that's all the questions we'll um maybe let dr lee i'll take a short break and then uh the residents can stay on uh for their meeting okay great thanks linda appreciate it thanks sander let me know when you want to come to l.a thanks very much great talk bye you