Immunotherapy and Clinical Trials for Brain Tumors | UCLA Neurosurgery
March 30, 2019 - uploaded on April 30, 2019 https://www.youtube.com/watch?v=bZiwsLblLmk&feature=youtu.be (Thanks to Longfellow for his transcription program help!)
I’m gonna talk about brain tumor therapy, well clinical trials in general, but specifically focused on brain tumor immunotherapy trials. So these are my disclosures. I have research funding from some of these companies. I don’t have any stocks or stock options.
So the current standard of care for glioblastoma really has evolved over the last couple decades. Back in the 70s and 80s, really all we had is radiation. And that was pretty much it up until the mid-90s when we started to have some new things that were FDA approved. The BCNU wafers, the GLIADEL wafers were approved in the mid-90s, and then temozolimide then came into first line treatment in 2005 after the Stupp paper. And since then, there have been some approvals, bevacizumab or avastin, for recurrent glioblastoma, and the Optune device, tumor treating fields. Although standard of treatment really is still radiation and temozolimide. These are options that even though they are FDA approved, they really haven’t been widely adopted.
Over the last there or four years, there’ve been twenty cancer immunotherapy drugs that have been FDA approved for various cancers. Melanoma, lung cancer, and so forth. So I think there’s a lot of promise out there in the cancer area for immunotherapy, but we have yet to have an immunotherapy drug approved for glioblastoma. It’s a very complicated, heterogeneous disease so it’s been very difficult to get approvals for that particular drug.
With that being said, I do think there is some promise out there in terms of the promise of immunotherapy for this particular disease.
So I just want to start with a case report. This is a thirty three year old right handed man. He basically presented with new onset seizures and speech aphasia. A typical case that we see quite often at UCLA. This was his lesion, left temporal lob tumor, he had speech aphasia. We actually did the whole functional mapping, awake surgery, as Dr. Iverston (?) was talking about earlier. Did an unblock resection, got a gross total resection in this patient. Went on to do standard treatment. Here’s the pathology. So it was glioblastoma, basically a very text book presentation for glioblastoma pathologically. Ki67 index 25 - 30%, GFAP+; positive for EGFR; MGMT methylated. So even with MGMT methylation, based on the Stupp protocol, average median survival is probably about 24 months at best. So nothing too out of the ordinary there. This was actually before IDH1 was discovered, so we did not at the time know his IDH1 status. He actually turned out to be IDH1 wild type, which is actually the poorer prognostic group. But his post op treatment consisted of radiation and temodar, and then he went on to a clinical trial. And at the time, he enrolled in our earlier phase clinical trial of an autogolous vaccine made of tumor lysates and dendritic cells.
And this is kind of the trial protocol. The patients underwent surgery, and after the surgery, we took their tumor tissue, and then we co-cultured the tumor lysate with these dendritic cells that we actually developed from the patient’s blood. And then that was injected back into the patients as a vaccine at 0, 14, and 28 days. Post-op they then went on to standard treatment in addition to this.
So one thing, you know, it was an interesting case, and actually a case that I remember quite well because his wife was 7 months pregnant at the time he was diagnosed with glioblastoma, and he was told he only had a few months to live, and at best, a couple of years. So after surgery, he recovered well, and he actually and his baby, and he made it to the local news. And one thing we talk about in the glioma field, is that it’s always very heart warming to get Christmas cards from your patients. But unfortunately, you don't see more than one or two Christmas cards because of the duration of survival. So I got the first Christmas card from him, and this is Brad Jr., at one year old, when he was a year old. And I actually had a baby at about the same time. So it was quite heart-warming. But then I got a five year old Christmas card from Brad Jr. And then ten years! And the dad, Brad, was doing quite well at ten years. And then not recently, we just saw Brad a few months ago,. He’s nearing fifty years old now, and he has a 16 year progression free survival. He’s healthier than I am. He’s actually a water polo coach, and much more in shape.
But, so, there are these long term survivors over the years, and I have a couple dozen of these now in these trials. So there must be something about Brad, and these long term survivors, that allows them to live such a long time. And as many of you may know, this is very unusual for glioblastomas. So we, this is actually the focus of our research lab at UCLA. We’re trying to figure out how do we detect and select for these patients that may actually do quite well with immunotherapies. So when he reached the five year point we actually thought, oh wow, you know, he’s lived a long time. So we actually looked for certain kinds of t-cell responses in his blood. So one of the… you know… we were looking for immune markers, and one of the actually control kind of antigens is CMV. Yeah, it's actually a viral antigen. And what was interesting in his case was you know he actually had a lot of CMV in his tumor. It's actually been reported that this does occur in glioblastomas, but it's really kind of controversial as to whether it's just a, you know, bystander or an actual, you know, target or even a, you know, something that causes the etiology of this. But what was interesting is before therapy… I don't know if you could see… he had 0.19% of his t-cells were targeted to CMV. After the vaccine 4.3% percent of these t-cells in the blood were targeted to CMV. It may seem really small but if you have 4.3% of your t-cells targeting a specific target, that's actually quite significant. And it really what it shows us is that it does actually suggest that the vaccine was creating an immune response. Whether this was kind of leading to his own long-term survival, we actually don't know for sure. Other things we found is that the gene expression signatures or the molecular profile of the tumor may correlate to better prognosis with this, you know, immunotherapy. Or nowadays, we're finding that for all sorts of therapies, and and I think as we move to the world of precision medicine, we're going to have to like really save the tumor tissue, analyze the tumor tissue, and then stratify patients based on what the tumor tissue shows. In this particular case, we looked at, you know, the the whole population of the initial clinical trial cohort, and what we found was that the long-term survivors actually had this this basically mesenchymal profile, (illegible). And and these people… whereas the proneural group, which many of these are IDH1 mutated which technically do better… so actually, if you look at the data here, they actually did pretty well. You know, 2000 days is actually, you know, not bad. But these people, they actually about half of them, continue to live beyond five years at the time this was reported. And Brad is actually in this group and four of those five are actually still alive today. So that’s a part of this 24 people cohort that data that we have over the years.
And you know, we’ve also looked at the kind of things that may predict response… and so oftentimes in immune therapy not just with vaccines but also now with the checkpoint inhibitors and all the other new immunotherapies coming up, you see this increased enhancement after treatment and then over time it kind of goes away on its own. So it's this, you know, kind of a phenomenon that we call pseudo progression. So sometimes after treatment… and this also occurs after radiation and after temozolomide… you do see a period of time where things actually look like they get worse before they get better.
So with that there's a lot of work now that we're doing at UCLA in terms of imaging and they're trying to determine, you know, what pseudo progression and what's real progression and you know. That’s you know whole whole talk in and of itself. But that's actually an area that's quite you know quite heavily investigated as well.
So based on those early phase studies, we did go on to a phase three multi-center clinical trial and 3331 patients. And I just wanted to give you an update of that trial as of late 2018.
So this trial isn't unblinded yet, but there was a crossover arm so I must say the majority of people actually got the vaccine at some point either early, or at crossover. So if you just look at the whole intent-to-treat population in this trial, there was a 28.2% three-year survival which is relatively good for glioblastoma. I remember when I was a resident, three-year survival at 5% was considered great, you know, and and I don't think anyone else making it more than five years in these cohorts. But it's interesting. It's not just three year survival, it's three year… the majority of these people actually have progression-free survival. So they're like Brad.
And we you know if you kind of just look at the top 30%… so if you look at the top 30 versus the bottom 30… there’s this one, you know, of the 330 patients, you know, that the top 100 patients, their estimated survival, their kind of predicted survival, is almost, you know, 60 months. So that is almost five years for median survival. And when we compared it to other types of trials that been done, you know, the the standard treatment, which is the Stupp protocol, the prognostic factors were all very similar.
And then this is kind of the survival rate so so the, you know, the 3-year survival with standard of care was 16 percent. And these trials failed so because they did not show an increase between controls and treated. Although if you look at the two year survival rate, there was an increase, and I'll explain some caveats about clinical trial design as to why that potentially could be. But Optune, which is actually an FDA approved device, actually has a three year survival of 26 percent … whereas this trial actually the whole the intent to treat population is 28 percent. So that's actually the control and the treated so if you kind of imagine if you take that whole population… hopefully the treated would be higher, but we actually don't know that yet.
But interestingly, you know, now we kind of looked at well, why did some people, you know, live so much longer. And you know, one of the things that many of you may know MGMT methylation does suggest a better prognostic response and better survival for these patients. Better prognostic response to Temozolomide, but they also tend to do very well on these immunotherapy trials as well. So if we stratified out MGMT methylated of… there's a three year survival of 49%. So almost half of these patients are in that methylated group and are living to 3 years. And if you're unmethylated, you're a little closer to what the the standard protocols would be. So it could be that in the future we would stratify people with certain markers to get immunotherapy, and people without those markers to get something else. And that's probably what's gonna evolve into the future. And I think one of the problems with our immunotherapy trials is that the response rate is usually about 20-25 percent. So if you take two groups and you look at median survival, which is usually kind of what the, you know cutoff is … some difference in median survival… if 50% of the patients have no response then that median survival won't be any different between control and treated. So I think there there's going to be some changes in the way we think about how how we approve these treatments based on that.
So with that, you know we've been working on various clinical trials to kind of enhance this effect. So we do have an upcoming clinical trial where we actually are going to be using this vaccine in combination with a checkpoint inhibitor.
That should be opening in the next couple months. And this just kind of illustrates the timeline it takes to get something to, you know, to that point. You know we did the first in man dendritic cell vaccination for glioblastoma back in the late 90s. And then did the phase 1, you know, clinical trials. This is kind of when Brad was enrolled. We've been following these patients for, you know, 16 or plus years now. And now, you know, we finally just completed enrollment of the
Phase three, and now we're going to combination trials starting in the next couple months. Other clinical trials we have at UCLA… we do have a gene therapy trial, and some of you may have heard the Tocagen virus, the Atoka Virus (?), and what it is is actually it is a replicating competent retrovirus. And in the… basically. it's injected into the tumor after resection or, you know, or for a non-resected tumor, it’s injected in there. And then patients… then the virus theoretically gets into only dividing cells. So the tumors divide unfortunately… your brain cells don't divide that much as an adult. Once you know the virus gets in, then you treat it with a basically an anti, you know, viral drug, an antibiotic, to kill off the virus but not the cells within them.
So in terms of the timeline for this, this is actually also a therapeutic developed at UCLA. Nori Kasahara developed this in 2009, or in the mid-2000s when he was at UCLA. And actually we were the first to inject this into humans. And then that's going on to phase 1 phase 2, and now, you know, phase 3 trials for for this particular therapeutic. And and then the future iterations of this will probably include this virus with also immune modulators. Because one thing we're finding is that viral therapy, or gene therapy is actually really immune therapy. The long term survivors in those trials, whether it be the toca virus, the polio virus, or you know, these different kind of viral therapies that you may be hearing about… it turns out that they are getting an immune response to the virus, and that's probably what's dictating their long-term survival.
We also just recently… Tim Cloughesy, who's our neuro-oncologist, and our group, recently published a paper in Nature Medicine just last month on the use of these checkpoint inhibitors. And another concept that came out of this recently is that the timing of treatments makes a difference. So this was just a very simple study. It was using one drug, Nivolumab, it’s a checkpoint inhibitor. And the two arms… the difference between this arm and that arm was just the timing of when you gave it. So basically, one group got the checkpoint inhibitor before surgery and one group got the checkpoint inhibitor after surgery. So one group got it before, and then after… meaning neoadjuvant and adjuvant… and one group just got adjuvant. These checkpoint inhibitor trials have been done previously, and they actually all failed as a single agent for glioblastoma. But it did seem to extend survival if you did it before surgery. And the reason we think immunologically that is is because you actually need the antigen present in order to induce an immune response. So if you take the tumor out before you induce the immune response to the tumor, you actually don't get the effect. So that's why giving it before surgery while you have the tumor there to induce your, you know, immune response, your t-cell response, and then giving it subsequently after, actually makes a difference. So this is kind of really you know actually renewed interest in these kind of checkpoint inhibitor and immunotherapy trials. So, you know, so it kind of for a while when all those trials failed people used to think well they're not going to work for a glioblastoma.
I think they can work… it’s just a matter of timing, and also maybe some of these need to be used in combination with vaccines, because one thing the vaccines do do is they get the T cells in the tumor.
So that's the whole concept of getting a cold, what they call a cold tumor, non immunogenic tumor, to a hot tumor and and you know our group was also very, you know, instrumental, and kind of getting this to, kind of renewed clinical trials with these drugs. And then another thing that I think, you know, UCLA's really made an impact on, are, you know, this other category of glioblastomas, with the IDH1 mutation. The IDH1 mutation is… IDH1 is an enzyme in the kreb cycle for, you know, if you remember biology. And it was actually a mutation that was found in glioblastomas in 2009, so it's a relatively recent, and you know, we and a company called Agios, also in 2009, discovered that these IDH1 mutated gliomas secrete a metabolite called two hydroxyglutarate. And based on that, we’ve actually, you know, gone on again to do a Phase one clinical trial with this drug. And this is kind of… what those trials have shown… basically these are more the non enhancing low-grade tumors, because the IDH1 mutated gliomas tend to be lower grade, or the secondary glioblastoma, so these drugs just show that if you take a non-enhancing tumor, give the drug, you actually decrease the growth rate of the tumor. So it doesn't make the tumor go away, it's not cytotoxic, but it’s, you know, it's potentially cytostatic. So for a low-grade tumor, I guess the the benefit of this is that in the future you may be able to avoid radiation for these low-grade tumors, in which typically do occur in younger patients. And, you know, and you heard from Dr. Kim what the kind of long-term potential neurocognitive effects could be for radiation.
So with that being said, I’m sorry the writing is so light, but there are some websites that you could go to if you want to learn more about our clinical trials. And then I just do want to end with this kind of statement that now, because the tumor tissue is so important for deciding what to do after surgery... It does make a difference, you know how you save the tumor tissue, and what you do with it.
Sometimes we use it to make the vaccine, so we actually need the actual tumor tissue fresh at the time of surgery. But there are a lot of trials who don't, that don't necessarily need the tumor fresh, but what they do need is the genomic analysis of the tumor and the genomic sequencing. And you know at UCLA we do all the, you know, we have someone who comes down to the OR 24/7 to collect tumor tissue for this purpose.
But if you're in the community, and you don't have that, there are these commercial places that do this. Foundation medicine actually does genomic sequencing, and Caris does, and these are now becoming paid for by insurance so this is something that I think this is something that a lot of patients are going to be asking for. And probably in the future, it will be pretty standard
, so thank you for your time.