NorthWest Biotherapeutics Inc (NWBO)

[sentiment_stocks]

Linda LIau's December 15, 2016 Presentation at Seattle Science. :)

https://www.youtube.com/watch?v=pkcoo5Uab00
Note: the link is under the intro section under "Videos" - which is right above "webcasts".

I'd mentioned that I'd transcribed it, so I'll post that now before I forget (or it gets taken down again) in response to Xoma's post.

I think it's very interesting to read/view this video with almost two years having passed since she gave her presentation. I know I certainly didn't want to think about the trial being taken out a full 3 years for every patient back then (end of 2016!). But this was also around the time that we as a board were starting to delve into the idea of psPD having possibly occurred in the treatment patients and what that might mean for the PFS end point. So viewing it in retrospect, it does seem that LL too was facing the possibility of a three year window for every patient as well.

I begin transcribing at minute 3:30 as that's when she really begins. Apologize for any typos or misses in transcription.


Immunotherapies for GBM: Tumor Vaccines by Linda Liau, M.D., Ph.D., M.B.A.

3:30
So as far as just as an overview and a background for immunotherapy, there are kind of two main forms of immunotherapy - what we call “active” immunotherapy and “passive” immunotherapy. Passive immunotherapy has to do with, for instance, the adoptive transfer of T cells to the host where you activate the T cells and then you infuse it back to the host. And these can be in the form of CAR-Ts and other forms of activated T cells. That’s not what I’m going to talk about today. What I’m going to talk about is active immunotherapy. And active immunotherapy actually requires the host to mount his or her own immune response. There are some potential advantages to active immunotherapy, some of which are that it can induce a de novo response to antigen. So you actually present an antigen to a patient and they mount their own immune response. It can induce systemic immunity, and importantly, it can induce dynamic immune recognition via epitope spreading. So for instance, if an immune cell kills a tumor cell, or targets an antigen on the tumor cell, the tumor cell dies off, other immune cells come in and pick up other antigens that is not necessarily the antigen you initially targeted. But the holy grail for cancer, and for cancer immunotherapy is really the last one, the ability to induce long-term immunologic memory.

As all of you know, glioblastomas and brain cancer are easy tumors to treat initially. You can take them out, you can do radiation and chemo. The problem is that they keep coming back. And that’s really the issue in terms of trying to induce some sort of long term immunologic memory for to prevent them from coming back.

So over the years, there have been hundreds and hundreds of pre-clinical studies and early phase one, phase two clinical trials. But there have been three tumor vaccine approaches to glioblastoma that have kind of advanced beyond phase one/two and two multi-center phase two or phase three trials. One of which is the EGFR mutant peptide vaccine called Rindopepimut. And the issues with that is that it does require mutant EGFR status for the tumor. So that’s only relevant in about thirty to forty percent of glioblastomas. There was a phase one [sic. three] trial, and I’ll talk a little about that trial, which was terminated earlier this year due to being unlikely to meet primary overall survival endpoints. But there are some issues in terms of trial design that I think we’ve learned some lessons from.

There are peptide-loaded dendritic cells. So these are dendritic cells loaded with, you know, three or four candidate peptides. It does require pre-selection of patients with the appropriate HLA types because there are only certain antigens that have been identified for certain HLA types. There was a phase two trial that also did not meet its primary endpoint, although they did identify some subgroups that may have had some benefit.

6:39
And then the trial that I’ve been involved in for a very long time now is tumor-lysate loaded dendritic cell vaccines, or DCVax-L, which is what the company called it after we licensed it to them. And what that is is it involves the patient’s own tumor tissue as the antigen source to pulse dendritic cells. And the advantage of that is you don’t actually need to know the antigen because, my personal feeling is that for glioblastomas they are so heterogenous that it’s not going to be one or two or three or even six antigens. It’s actually hundreds to thousands of antigens that we really need to identify to mount an effective immune response.

7:24
We have been in a phase one [sic. three] clinical trial. It actually just closed enrollment earlier this month, actually last month, November. The enrollment included 331 patients and we’re just waiting for the number of events in order to do the data lock in analysis. So hopefully we’ll get that data soon. (lol)

7:47
So just briefly on the EGFRV-III peptide vaccine which may have been heard about or participated in, the subsequent, the phase three trials were really predicated on these earlier phase one, phase two trials, which did show, as you can see, the median survival of the treated patients versus the institutional controls. There was a difference between fifteen months and twenty six months, which is quite impressive in this work done by John Sampson at Duke. So that led to a phase three study of this EGRF-III vaccine in newly diagnosed glioblastoma patients which was called Act IV. this was an international, randomized, double-blinded, placebo-controlled phase three study. Newly diagnosed, surgically resected, EGFRvIII-positive glioblastomas. Half got the vaccine, half got control with KLH and temozolmide. It was opened in 200 clinical trial sites across 22 countries and there were 745 EGFRvIII positive patients enrolled. So a fairly large trial. And the overall endpoint was overall survival. They completed enrollment in December 2014, and then in March of 2016, the interim analysis was conducted and it was deemed by an independent DSMB, that it was, that this trial was unlikely to meet its safety [sic. primary] endpoint. So actually, the subsequent trials that were planned were stopped.

So why was that? What happened? And I think this is a lesson for us; and then I’ll talk about some lessons I’ve learned with our own trials as well. But one of the issues is obviously selection bias from early phase trials. A lot of these trials, as much as you want to not select… obviously you do have this propensity to select patients who may do better and those tend to be the patients that get on clinical trials. So for instance when this phase two data, when you compared treated patients, open label treated patients, to controls, you actually do see a significant difference, even if you do historical matched controls. But then when you, and you can see at the median survival there’s a significant difference between two years and almost four years in that cohort.

But the issue sometimes is that median progression free survival and overall survival may not capture the true efficacy of immunotherapy. This is actually a trial done in recurrent GBM and you can see if you look at median survival, that’s really the time at the fifty percent point, when these two curves are analyzed, and you can see it’s not really that much different. You know, it’s somewhere between 14 months and 15 months, and so you can get a non-statistical significance there.

11:15
But I think the true benefit of immunotherapy is really out here. It’s really the tail end of the curve. Cause what a lot of us are seeing is that there are twenty, to almost even thirty, forty percent of patients that do live significantly longer. Meaning three years or more with immunotherapy.

11:35
So the problem with the current design for FDA approval is that you submit a trial. The endpoints are difference, for instance, in overall survival, or median survival. You see some difference, and if you’re lucky, they’re statistically significant. But you’re not really capturing this tail end.

But in reality, when you treat patients, if you give them the option, you tell them if you can live three more months with this therapy, versus, you have a twenty to thirty-five percent chance of living five more years, most patients would actually opt for that thirty percent chance of living five more years than okay, I’ll do this and I’ll give you three more months.

12:15
But that’s not really captured in our clinical trial design, because it’s really designed to measure differences in chemotherapy and other therapeutics. And immune therapy is… it’s a different field. And that’s something, I think, where we’re struggling with now in the field to really, you know, get these drugs approved given what we’re seeing in terms of what the true efficacy for these agents are.

12:35
So the lesson learned from this trial I would say is yes, there is a selection bias for early phase trials, and we have to understand that. So when you power your subsequent phase trials, power them to an extent where you can hopefully overcome that.

12:50
But also there is a need for relevant endpoints. You know, median survival, overall survival at one year, or even two years, is really… may not be the meaningful endpoint for these trials. It’s really that long term tail end. The hazard ratio at three years or four years is really what we should be computing for these trials.

13:10
And then also I think, as with any treatment, there’s going to be a subset of patients that benefit, and some that do not. Cause obviously with these immunotherapy trials, twenty-five, thirty percent have that long tail survival, but obviously the other seventy-five percent doesn’t. I think the next goal is really to figure out who those twenty-five percent of patients are, either by expression analysis or whether it’s extent of resection, or GBM subgroups, those are the kinds of things that I think will help move the field forward in glioblastoma.

13:44
So with that in mind, I just wanted to give an update on the current status of our dendritic cell based vaccine trial which is called DCVax-L. Just as an overview for those not familiar with what dendritic cells are… they are professional antigen-presenting cells. They’re actually normal cells in your body. They come from the bone marrow. They’re recruited to the peripheral organs, whether it be the skin, the gut, or anywhere where your body would capture antigen. And they reside there. They’re in relatively small populations, but that’s where they live. Once they capture an antigen, let’s say a virus or bacteria, then they’re activated and then they migrate to secondary lymphoid organs like the lymph nodes and the spleen where they activate T cells. And it’s activated T cells that actually are targeting the antigen, and in our case, we would want them to target tumor antigens.

14:44
So this is the concept, you know the dendritic cell picks up tumor targeted proteins. The caveat in brain tumors is these dendritic cells aren’t circulating around in large numbers in the brain per say, so the thought, or I guess the idea that we came up with in terms of this vaccine was that we take out the patient’s own dendritic cells, take out the tumor, have them meet ex vivo, and then inject them back into the patients en vivo. So we would take out dendritic cells, load them with tumor antigens, and there are various different types of antigens that you can load these tumors with. And believe, me, we’ve tried many, many different types, and we still come back to the autologous tumor antigens just simply because we haven’t found a good cocktail of antigens that really would work for this. And I think it’s because of the heterogeneity of glioblastomas. So once they encounter t Cell, the t cell gets activated as I mentioned, the activated t cells then proliferate and divide, there’s this clonal expansion, and then they migrate to the tumor site. And they are able to migrate into the tumors. They are able to cross the blood brain barrier into tumors, and we actually know that for a fact now.

16:02
Interestingly, no one believed that about twenty years ago. So this is actually our first paper kind of showing that… that you could actually, and this is in a mouse model, that you could actually vaccinate against a target tumor antigen. And we did show t cell infiltration following vaccination into brain tumors. This is actually very well known now, but back in the mid-nineties, I remember applying for grants for this and everyone thought, there’s no way t cells get into the brain. It’s privileged. So this is, I think, a change in the concept of how we think of immunotherapy.

16:42
So actually in 1999, 98 actually, but we didn’t get the paper out until 99, at 2000, we did the first human injection of DC based vaccines into a human patient. And what that involved was taking out peripheral blood monocytes from the patient, culturing it with GM-CSF and IL4, co-culturing it with tumor lysates taken from the patients. And over the years, we’ve tried different other tumor antigens to do this, and then injecting it back into patients.

17:19
That first single case study led to a phase one clinical trial. And this is essentially the concept that I was describing. We got some funding to do that phase one clinical trial and what we found was that yes indeed, we are able… now in humans.. we are able to induce a t cell response, t cells getting into the tumor following this vaccination.

Interestingly, the t cell infiltration seemed to correlate with survival, so, for instance, the patient with the longer survival, 30 months, had more t cell infiltration post vaccination then the patients who did not have t cell infiltration. We also looked for other factors that correlated with response, and I can tell you we looked for many, many, many factors, and the only one that we found that was significant was TGF-beta. So the more TGF-beta you had in your tumor, the less likely you were getting t cell infiltration into the tumor, and the lower the survival. And that’s probably because TGF-beta is acting in an immunosuppressive capacity there.

18:29
So that led to a phase one/two clinical trial. And this was an open label trial. We saw good, just like other trials, a good extension of median survival, compared to institutional, historical controls. But again, this wasn’t randomized, this was a phase one with historical controls… although the three year survival percentage in our trial was 47 percent compared to our institutional control matched for age, extent of resection, all that, which was only 21 percent.

But as you can see, the difference is not necessarily, in retrospect, it’s not the median survival, it’s really that tail end. So that tail end in this particular case was almost thirty percent. And I can tell you a lot of those patients are actually still alive.

So that’s actually what intrigues me the most. Who are these tail end patients? And how can we identify them?

19:34
So with that, you get to thinking, well what is different about these patients? I’ll tell you about two of the first cohort of patients. So this is patient number one, and this is actually patient number three in the clinical trial. Both had left temporal lobe tumors, both underwent gross total resection, so you know, they had the minimal disease prior to initiating this trial. They both had standard treatment with radiation and temazolomide. And then they subsequently enrolled in this clinical trial. And you can see the demographics are very similar between the two. You know, the first patient was 39 years old. The second patient 34. Both male. Good Karnofsky performance score. You know, they both had small kids when I met them. It was actually very easily similar type of patient. Histologically, if you looked at their pathology, they looked identical. You really couldn’t tell anything on histology that was different between these two patients. But you can see the overall survival was significantly different.

20:42
So the first patient, his overall survival was 33 months, almost 34 months. Whereas the second patient, was, at the time we published this paper, over eight years. And actually, this paper was published quite a while ago and he’s still alive now, going on fourteen years. And we’ve had several of these patients. And this particular patient actually had activation of a CMV directed t cell response, and I’ll show you that.

But what was curious is what was different about patient one versus patient three. And one thing we did look at was gene expression profile. For those in the field, you know, glioblastomas are now classified into different subgroups. You know, pro neural, classical, mesenchymal. So one criticism of this trial is well maybe you just enrolled all the pro neural patients, you know the good sub group patients, and that’s why you’re getting longer survivals in these patients. So we looked at that, and interestingly, that wasn’t actually the case. So patient number one was in that top cohort of patients. He lived thirty-three months, which was actually higher than the normal, you know than median for glioblastoma. But it was actually not that much different then would be expected for his subgroup. Pro neural subgroup, IDH1 mutated, MGMT methylated, you know those good prognostic factors. He kind of lived as long as probably would have been expected, given his molecular subgroup.


22:13
[color=red]Interestingly, the patients who actually saw the most benefit from immunotherapy was in the mesenchymal subgroup. And that’s actually the poorer [performing] subgroup. Those are the ones that are IDH1 wild type, MGMT not methylated. Well methylated or nonmethylated, it was actually in both populations. [/color] But those are the ones who typically have, you know, a median survival of about 14 to 16 months. So in that group, the vaccine actually did seem to show a significant difference. And actually, all our long term survivors were in the mesenchymal subgroup. (is that what we'll find to be the case in the 100 or so longer term survivors?)

So there is something about that particular subgroup that you know, I think is, you know, inducing a perhaps better immune response for these patients. And that’s kind of what we’ve been studying, you know, for the last five or six years now.

So things that, you know, seem to correlate with this is that they do seem to be increased with increased tumor infiltrating lymphocytes. In our initial trial, we saw that those with increased lymphocytic filtration do better. So it seems to be the case for this.

23:20
But it wasn’t actually just the lymphocytic infiltrates after vaccination, because we didn’t, you know, many of these patients actually didn’t recur, so we actually have never gone back to take out tumor. But it’s actually the lymphocytic infiltration BEFORE vaccination. So patients who tended to have some T-cell in their tumor before entering the trial actually seemed to do better. So, this is probably a more immune responsive subgroup. You know, a subgroup where, you know there is some initial immune response, maybe it just wasn’t enough, and immune therapy could make a difference in this particular subgroup.

23:58
This also tends to be the group, and we’re seeing this NOW in other trials, this tends to be the group that tends to show more T2 changes following vaccination OR IMMUNE ENHANCEMENT following vaccination, and more likely what we call PSEUDOPROGRESSION FOLLOWING TREATMENT. So there is some, you know, immune, I guess, component to this particular subgroup of patients. And interestingly, and more RECENT data where we’ve actually done genetic sequencing of these tumors, these happen to be the most highly mutated. So interestingly, they’re probably not, these patients are probably not doing as well because their tumors are so mutated, but they actually turn out to be a better target for immune therapies because the more mutations you have, the more targets you have for immune based therapies.

So this is kind of the quantification of what we saw in terms of tumor infiltrating lymphocytes. So it does seem that tumor infiltrating lymphocyte content, you know, pre-treatment, does tend to be a biomarker for survival post-treatment.

25.07
Question asked.

Answer:
CD8. Yeah. CD4 there was some increase, but this was specifically CD8. So it’s the t cell, the killer t cell infiltrates.

25:20
And this is work done by Rob Prins, who has worked with me through all of this actually for the last twenty years, showing that you know, cause right now we’re looking for biomarkers, to figure out who that twenty five to thirty percent of patients are. So this actually shows that an increase in TCR, t cell receptor, overlap in the tumor versus the peripheral blood, is also a potential biomarker for increased survival.

So what this means, what we did was, actually in these patients, we isolated the t cells from the tumor, and we isolated t cells from the blood, and then sequenced the t cells, and looked for their t cell receptor repertoires. So if we saw… the patients that did the best were the patients who had the greatest overlap between the two… so basically with the blood t cell receptors, we’re targeting… you know, the blood and the tumor t cell receptors were overlapped, or similar. So the thought is obviously you vaccinate to an antigen, and it’s an antigen that is relevant in the tumor, because obviously why would you have t cell receptors to a particular antigen in the tumor unless there is something there that led the t cells to be there. And if your blood t cells had a greater overlap that could predict a better immune response. And there were patients that were significantly…. this is a confusing graph… but essentially what it means is the colors, the yellow is more overlap, and the dark, you know, hollow spaces meant that there was less overlap.

Question: So the left is the good side?

LL: Right. The left is the good. Right, it’s the colors… ideally I should have made this into a 3-D graph, it would have been easier, in terms of the overlap.

But what we’re looking for is something that we can monitor in patients, that allows us to monitor their response to treatment.

27:23
So you can imagine, if you take out the t cells, sequence the t cells, figure out the repertiore, t cell repertoire of the tumor infiltrating t cells, and then if you get an increase in that overlap following vaccination, you could actually monitor that. And if that increase goes down, then, you know, that’s the time, maybe you need to revaccinate or boost that immune response to this t cell repertoire.

Other things we’re looking at right now is ways to enhance the t cell response. Because obviously… you know the… one issue is that the t cells may get in, but they’re non-functional because of, you know, immune checkpoints and things like that. And what we found, you know, or as you may have heard, things like these PD1 inhibitors are now actually FDA approved for other cancers like melanoma and lung cancer. There have been trials in glioblastoma, but so far, nothing substantial, I think, as a single agent. And I think the reason is that glioblastomas aren’t inherently that immunogenic. They’re not getting a huge population of t cells there. So if you don’t have t cells in the tumor, then unblocking the block, ‘cause that’s what these checkpoint inhibitors do, they actually block the immune suppression… unblocking the block doesn’t really help ‘cause there’s no traffic going through. So that’s kind of probably why these check point inhibitors in itself does not work.

28:55
So we actually have a clinical trial that we just got FDA approval for - we’re probably starting right after the new year - to do DC vaccination followed by nivolumab, which is a PD1 inhibitor. And this is just based on some clinical, some pre-clinical studies, in animal models showing that if you combine the two, we do get increased survival, and probably for that reason, even though the vaccine can get t cells in, still the t cells are immunosuppressed by the local immune environment.

29:30
So going back to, you know, so we’ve also looked at these long term survivors, you know what is it about them that makes them such long term survivors. And this was actually patient number three in our phase one cohort. And what we did was we actually… and he happened to be a HLA A2 positive… so we have, you know, antigens that, candidate antigens, that we kind of tested his t cells against… so we just tested for a panel of known candidate antigens. And interestingly, for a lot of these patients we actually used viral antigens as the control to test for whether there was an immune response. And as it turns out, this patient actually mounted a very significant immune response following vaccination to CMV. And then when we went back and looked at his tumor tissue, he actually had very strong expression of CMV on immunohistochemistry. And then if you look at the graphs before… so before treatment, these are tetramer analysis of t cells populations… so before treatment, he had .19 percent, you know, very small percentage of his t cells were recognizing CMV. After treatment, that went up significantly to 4.36%. And that’s huge. To have 4.36 of t cells throughout your body targeted to one particular antigen. So there was a huge increase in the immune response to CMV following vaccination. And then it waned over time… at six months it was low. And then what… we started doing booster vaccinations, and as we did boosters, it goes up again, wanes in six months, and goes up. And we kept doing that until we actually ran out of vaccine for this patient.

31:20
But this paper was actually published in 2008, and he was a five year survivor at the time. He’s still alive today [it’s Brad Silver], and it’s been six year since then. So he, actually it’s nine years since that time, so he’s fourteen years out now. And his scan now, looks exactly like this. He actually has no recurrence. And you know when I first met him, his kid, his son was very small, now his son’s in high school, and he’s hoping to go to his high school graduation soon.

31:53
So this is obviously the issue that we want to learn more about. And over the years there have been many clinical trials using dendritic cell vaccines throughout the world. And this was a meta analysis done by Rick Komotar back in 2013, and that time, there were already 21 published studies, 403 patients in five continents, and overall, the conclusion of this met analysis was that dendritic cells loaded with autologous tumor cells did increase overall survival in glioblastoma patients, pretty much in all these trials. It showed some survival advantage. Granted these were all non-randomized, single arm studies. So you have to take these results with that in mind.

32:40
So based on all the earlier trials and the pre-clinical work, we went on to embark on a phase three multi center randomized clinical trial which actually just closed enrollment last month. And what that involved was, you know, we do surgery on these patients. So the patients underwent surgery. Following surgery, they got a leukepheresis to isolate their white blood cells. We would culture them with GMSF and IL4, make them dendritic cells. And during that time, they would undergo standard treatment which was radiation and temodar. It takes a few weeks to make the cells, and then after that, you know the patient would have their post radiation MRI scan, and then they would randomize to either placebo or injection in this clinical trial. So there are a couple of things that, you know… we designed this trial about seven, almost eight years ago now… so when… there are a couple of things that I think we’ve learned from this trial that may effect the final results. And I think it has helped us to kind of learn how to design future trials in the future.

33:55
One issue was that at this post radiation MRI scan, if the patients had progressive tumor, they were actually excluded from the trial. They were put into this separate informational arm. And the thought about that was that… you know, in our prior experience with the phase one trials… was that if you had big, bulky disease or actually not even bulky disease but progressive disease, those patients did not respond as well. Because if your tumor’s growing, you know, exponentially, you don’t have enough time to mount an immune response to the tumor. And those patients, you know at least what we saw, did not benefit. So that was the rationale for excluding the, what we call, early progressors in that trial.

34:40
Unfortunately what we’ve learned over the years is that determination of early progression is, is, is… difficult, because of the issue of pseudo progression. So there may be some patients that were excluded who probably should not have been excluded, or vice versa.

35:00
So the trial was held in fifty sites throughout the United States, Canada, UK and Germany. It was actually well over fifty million at this point, and you can see your site here, that’s that little star up in Washington way up North. And we just recently closed the enrollment. So what this was was this was an international, randomized, double-blinded, placebo-controlled study. The randomization was two to three, randomly assigned to DCVax versus placebo controlled. And a third just got autolous PBMC, no dendritic cells, But they did get injection of their own blood. The primary end point was progression-free survival, overall survival is the secondary endpoint. We’ve had 331 subjects enrolled, and currently, the trial is still blinded until data lock. And data lock will occur when 75% of the patients have an event, meaning death, or progression or death.

36:11
So so far that hasn’t happened. But it’s so, the whole group is doing better… you know, the treatment and the control group. They seem to be doing better than the historical, I guess, analysis for this. In a way, it’s good. It’s very good for the patients, but it may not be so good for the trial, because you kinda, you want to see a difference between the two groups, and the fact that the whole group seems to be doing better, although we don’t know who’s who, you know, that could be an issue in terms of FDA approval.

36:48
But we all do believe that there is, you know, efficacy in this, and it’s really comes down to what, you know, potentially happened with the EGFR vaccine trial. You know in terms of the endpoints and the trial design, and how to best elucidate the differences here.

37:08
So this was the trial design. So the patients underwent surgery, and then during that recovery period post surgery, they had leukapheresis. Then they had six weeks of radiation and temodar while the vaccine was being made, and then they had their baseline MRI. So at that baseline MRI, if they had progression, they were put into something called the informational arm. They still got the vaccine because we made it for them, and it would be unfair to not give it to patients. But they weren’t included in the trial analysis.

So the trial was randomized… and that’s probably why it took a little longer to approve, because we had some attrition at his point. The trial was randomized two to one, and that was the intent-to-treat population. They either got DCVax or placebo. And then this is the issue with why we think perhaps the whole group is doing better. There is a crossover arm. So if the patients got a progression, or was deemed to have progression on the MRI scan, they were allowed to cross over and get the true vaccine if they were on placebo. The patients love this. You know, it’s much easier to enroll patients with a cross over arm than to say, you know, if you randomize to placebo, too bad. But it really does kind of muddy the trial results. Because then essentially, we may be just comparing DCVax early to DCVax late.

38:37
So here’s um, so here’s the treatment groups, again… we’re still blinded to the results… but what we do know is that two thirds were uh randomized to DCVax and a third to placebo. So all the boxes in green got DCVax at some point. So obviously the top part of this chart, those were the patients, um, who randomized to the treatment. And at this point, and looking at their analysis, they either had no progression, they either had progression and recurrence and went on some other treatment, or they had progression and recurrence and then just went on standard treatment and continued with their vaccine. Um, the placebo arm, uh, there WAS a percentage of those patients… like I said, because of the cross over arm, that got progression, um, that got the DCVax at recurrence. And it turns out that the green boxes ACCOUNTED for 86% of the patients. So a significant percentage of these patients actually did get DCVax at some point.

39:40
But… and I think it’s… So we went back and we looked at this Informational Arm. So these are the rapid progressors… you know these are the people that got taken out early because their tumors were growing despite radiation and chemo… and they got enrolled into the Informational Arm or the pseudo progression arm, and we kinda called it slightly different things… um… during these two time points… between 2008 and 2012. And then subsequently, we felt that some of these were pseudo progressors.

But looking at this arm, there are 55 patients. And so this is almost like a separate little single arm study. What we found is that 20 patients had further progression on subsequent MRIs. So 20 were probably true progressors. So they’re the ones that… they had the progression on the MRI scan, and the next scan showed more progression and more progression. So those are true progressors.

25 had stable disease at the two month scan. So you’ve got that first post-radiation scan, and then you had another scan and actually it was stable. So those patients, in retrospect, were probably pseudo progressors, or people that were, you know, were read by radiologists as progression, but were probably pseudo progressors. And the issue with these multi-center trials is that, um, they’re all centrally reviewed… in terms of the radiology review… that’s to, you know, keep it unbiased. But because of the central radiology review, you don’t have the benefit of the clinical picture to decide if this is true progression or pseudo progression. So it’s all based on, you know, a certain percentage of increased mass on the MRI scan.

41:15
So I think what’s happened is, you know, there is a group of patients here that are pseudo progressors. And then one patient, um, had a, was a confirmed pseudo progressor, because his tumor actually totally went away. You know, so we knew that he was definitely a pseudo progressor. And then five patients we couldn’t really tell for sure with the subsequent MRI scans.

41:38
So these patients, because they were off trial, we were able to analyze their data early, and this is essentially what that data shows. Of the 25 patients that were pseudo progressors or possibly pseudo progressors, uh ten of them are, err, still alive over three years. They’ve progressed.. uh… they haven’t had any… not even alive… they actually… you know, most of these haven’t even had progression.

42:10
So, it’s actually quite exciting to us that, you know, we hopefully can identify a population of patients where this could have a significant benefit, meaning three year or more survival for these patients. So that’s that tail end of the curve. And it probably has to do with, you know, the patients that do have more pseudo progression, they actually do do better. Uh… we’ll obviously all do these sub group analyses once the bigger trial is unblinded and we have the cohorts identified. Um… there’s uh… you know, countless different subgroups that we were looking into, um, looking at. Um, and one of which is actually looking at CMV, to see, you know, if these long term… now that we have 331 patients, we have a population of those patients that are going to be long-term survivors to see what is it about their tumors or their immune response. We have blood on all these patients, through all these time points. So that will be a very interesting analysis, I think.

43:12
So the survival of this informational arm, these 55 patients, as you can see, the rapid progressors… so those were the, you know, 20 patients that had that initial read as progression so they got taken off the trial, subsequent scans showed further progression so those were kind of confirmed as rapid progressors. Those we kind of knew weren’t going to do as well. But the ones who were pseuoprogressors, actually did significantly better. They’re drawing out that tail end of the curve, you know 25% of them are now actually over four years. So that’s, you know, some intriguing preliminary data. And once we get the bigger randomized trial unblinded, we’d obviously like to look at that further.

44:00
So in conclusion, brain tumor vaccines are feasible and safe, and can potentially lead to a significant subgroup of long-term, meaning greater than five year survivor, progression-free survivors. So this is just not being alive, it’s being alive without progression. And I think, you know, for patients, that is meaningful. Granted, it’s not everybody, like I said, it’s about twenty-five percent, but if we could identify who those 25% of patients are to get on these trials, that may be meaningful.

The lessons learned over the years from our trial, and the other trials that have been going on in brain tumor immunotherapy and vaccines is that there is definitely a selection bias with early phased, non-randomized trials, so we have to account for that.

The endpoints that are relevant for immunotherapy are probably going to be different than the kind of accepted endpoints currently that are being established for FDA approval of chemotherapies and other agents. And then the heterogeneity of the tumor and the patient population needs to be taken into account.

So as I mentioned, you know, there, we, at least in early trials, we see some difference between the mesenchymal subgroup and the pro neural subgroup of glioblastoma. That’s obviously something we’re going to look at inner randomized trials when that gets unblinded. Things like MGMT methylation, which tends to portend to more pseudoprogression, that’s obviously something that we would look at. And that’s something that’s very easy to do. It’s done… everyone looks at the MGMT methylation now. So that’s actually a biomarker that potentially could be already universally applied. IDH1 mutation, 1p 19q co-deletion, TERT mutations, these are all the things that would play into these subgroup analyses. And then Ben Ellingson, who works at UCLA, he’s been doing a lot of work on image analysis, looking for imaging biomarkers for this and other types of treatments.

So I just want to say thank you for your time. And last night at dinner, we go to talking about how much tax I pay in Southern California. 13.9% state income tax. So I make 13.9% less than all of you every year. But the reason I think that we pay for that tax is that… this is my son and my daughter… we can go surfing and skiing on the same day. And it doesn’t rain much. Actually we need rain. We’re in a drought. So I wish it did rain. So thank you.