Good afternoon and welcome to the gritstone bio kalo webinar on oncology product candidates granite and slate. At this time all attendees are in listen only mode. A question and answer session will follow the formal presentations. If you’d like to submit a question, you may do so by using the q&a textbox at the bottom of the webcast player or by emailing your questions to questions at lifestyle advisors comm as a reminder, this call is being recorded and a replay will be made available on the gritstone bio website following the conclusion of the event. I’d now like to turn the call over to your host Dr. Andrew Allen, co founder president and chief executive officer of gritstone bio, please go ahead, Andrew.
Thank you, Tara. Well, good afternoon. Good morning, everybody, depending upon where you’re located. Pleasure. pleasure to be with you today. As many of you know, my name is Andrew Allen. I’m the CEO, one of the founders of Greystone buyer base today in the California Bay Area. It’s my pleasure to introduce two speakers who are joining us today both live in Paris and actually in the same room. First of all, we have Dr. Daniel catenacci, who gave the meteorol presentation on the granite data earlier this morning. Dan is the assistant professor of medicine at University of Chicago. And he is joined by Professor Terry onpay, who is professor of medical oncology at the San Juan Hospital in Paris. And they’ll both be speaking to you today. I’ll be making some forward looking statements during the course of this webcast. Our agenda today is to go through five elements. First of all, I’m going to briefly give an overview of our new antigen immunotherapy. I’ll cover this swiftly. Many of you have heard this before, and some of the relevant and technical data are published in in the scientific literature. Then I’m going to cover this slate off the shelf neoantigen cancer vaccine product. And I’ll be sharing with you some data from our version one of this product, and then describing the optimized version two that we’ve built using translational medicine data from the first study from the human data. And I explained what the version two looks like and where we’re taking it. Then I’ll hand over to Dan and he will describe with more time or with some more depth the granite data, which he presented in part earlier today at the meteorol then we’ll hand over to TRV who will cover the unmet need in colorectal cancer. He’ll describe current treatment options and describe the persistent unmet need that obviously we are seeking to address. And I will then close by describing the two randomized phase two studies that we’ll be running the beginning of next year in the setting of microsatellite stable colorectal cancer. But unlike the data you’ll see today, which is an N stage patients, the phase two studies will be further upstream in the patient’s course. So let me begin with an overview of our new antigen immunotherapy programs. As many of you know, the benefit of immunotherapy appears to be primarily limited to subjects who have an existing new antigen specific CDA T cell responses baseline. These tumors typically have high tumor mutational burden, they’re usually infiltrating CTA T cells in the tumor baseline, and PDL. One expression is often upregulated. And these patients can do extremely well on checkpoint inhibition with very substantial long term benefit, which obviously has really transformed the field of solid tumor immunotherapy over the last several years. But unfortunately, the very common solid tumors that kill most people often are cold tumors that have low tumor mutational burden, low PDL, one expression, and often no infiltrating CDA T cells. And if those patients are treated with a checkpoint inhibitor, unfortunately little useful happens apart from sometimes the development of autoimmunity. The therapeutic hypothesis that we’re pursuing is that we can perhaps bring those patients into the light as it were by providing them with new antigen specific CDA T cells. And we do so by taking a tumor biopsy predicting Neo antigens, putting those into a potent vaccine we use Prime boost the so called heterologous Prime boost with a chimpanzee and the viral prime and a self amplifying RNA boost. They contain the same set of Neo antigens. And for granted, those are 20 new antigens selected for that patient. So the product is made for that patient. Whereas for slate, it’s a fixed set of quote off the shelf Neo antigens. And we look for patients who have the right tumor mutations for therapy with slate. And then hopefully these will induce CDA T cells, which will then traffic to the tumors and remain active by virtue of concomitant administration of systemic checkpoint inhibitors. So that’s the therapeutic hypothesis that we’re pursuing. And today we’re going to show you data addressing that hypothesis.
Now, first of all, new antigen identification has been historically challenging. And we’ve spent several years now building a prediction model that starts just with formalin fixed paraffin embedded tissue. We wanted to have a system that did not require complex live cells or fresh tumor biopsies, but something that will be useful in the community where the common currency for us typical solid tumor patient is a former infects formalin fixed paraffin embedded core needle biopsy. That means we can really only use sequencing data as the input. And so we trained a model using human tumor samples, where we did identify surface presented peptides. And we correlated those with the genomic features of those same tumor samples. Based on DNA and RNA sequencing. We built a machine learning model that we called edge that can now take us just from sequence data alone to high probability prediction of relevant Neo antigens with over 70% positive predictive value. This work was published in Nature biotech at the end of 2018, when we’d have issued us patents covering this work. That’s half of the puzzle. Now you’ve identified the Neo antigens, you need to put them into a vaccine that will elicit strong CDA responses in humans, which is challenging again, most vaccines do not elicit CDs. And so we’ve developed this so called heterologous Prime boost system, we prime with a chimpanzee adna viral vector that drives a very strong CDA response, as well as CD fours and antibodies where that’s appropriate. And then we boost with a self amplifying RNA vector. Again, we don’t have time to go into this today. But this has some meaningful differences from mRNA vectors, which of course we’re all familiar with, given their dramatic success to generate neutralizing antibodies against cells COVID to self amplifying RNA, potentially drives stronger immunogenicity with CDA responses at a potential lower dose than mRNA. And that’s why we selected to use sound it’s given in a lipid nanoparticle. And we then take these two platforms and put them into patients. Again, we have the two overall programs granted, which is the individualized program and sleep the off the shelf. We’re going to show you today data today demonstrating that these are well tolerated. The only vaccine related adverse events are those associated with strong immune responses. We have early signs of clinical efficacy, including objective molecular responses. And further we’ve learned a fair amount about how to optimally dose these products as we move forward into our randomized studies. So let me describe slate. Slate is our off the shelf product we’ve demonstrated and you’ll see the data again today that this is well tolerated with no Stone, no dose limiting toxicities. And the phase one program will show evidence the non small cell lung cancer patients, all of whom had progressed on prior immunotherapy, sometimes with chemotherapy as well, that we’re seeing evidence of activity here with the slate product. That is probably the greatest in the six patients that we’ve treated, who have the Kara’s g 12. c mutation presented by HLA a Oh 201. And in the subset of these patients who had measurable ctdna circulating tumor DNA, we observed molecular responses into thirds, which correlated with clinical benefit. And we’ve seen one recessed response and unconfirmed response in one patient. And I’ll show you these pictures and images today. And as I mentioned earlier, we’re bringing a version two forward into the clinic. And that first patient has just been dosed with version two of slate as we announced this morning. So here’s the basic idea. You have several different k Ras mutations within the vaccine, but it’s not enough just have the mutation, you do need to have the relevant HLA class on a leil to present that mutation. So patients are screened for the presence of both the appropriate pair care as mutation plus h layer leil. And this adds up to about 10 to 15% of lung adenocarcinoma and around the same fraction of colorectal carcinoma patients. being eligible for treatment with this first product.
We’ve treated now 26 patients with version one and all the data I’m showing you today from version one of slate. You can see here that half of those patients were non small cell lung cancer patients, all of whom had received prior anti PD one therapy, and indeed a median of two priors, typically, of course, chemotherapy as well. We treated some patients with microsatellite stable colorectal cancer, and pancreatic cancer and a couple of other rarer types. And in terms of the safety on the side, on the right, you can see that the common treatment related adverse events are fever, as is often seen with potent vaccines, fatigue, and some injection site reactions, all of which are transient and self limiting. And the grade three for toxicity was typically seen with checkpoint inhibitors. This is the traditional resist assessment. And you can see the partial response that I described in a non small cell lung cancer patient. This was not confirmed, and I’ll show you those data shortly. But of course, we are innovating here and doing something that has never been done before in solid tumor patients to treat them with a potent new ancient vaccine. And we do not know how efficacy will be manifest. And tumor shrinkage, of course, is desirable, but it’s not necessary for clinical benefit. And I think there’s a little bit of a syllogism floating around at the moment that says, research responses are good new action vaccines don’t have lots of Reese’s responses, therefore they are no good. And of course that is irrational. That’s a false conclusion. And I think we need to open our minds to the fact that we have T cell infiltration into lesions, shrinkage may not be what we see different from what we’ve seen with prior therapies. And that’s why the molecular responses are particularly particularly important here. And of course, we’ve seen evidence of this phenomenon playing out with other products and most notably immuno core with their bi specific products have demonstrated in a subset of melanoma, that clinical benefit was profound with an overall survival has a ratio of around point five. And yet research responses were single digit and barely higher than in the control arm, suggesting indeed that there is now a wide separation between true clinical benefit. And what we may be seeing using traditional research response developed for cytotoxic therapies. So here are the molecular response data. Here we’re measuring the amount of mutant k wrasse DNA in plasma at baseline. And these essays are sensitive, but some patients have low concentrations, which are not necessarily measurable. So not all patients will be informative. When you’re looking for a single mutant DNA form. You have a much higher sensitivity when you’re looking for multiple forms, which is why generally, most of our granite patients are positive, because they were looking for at least 20 different mutations in the blood. But here we say we’re just looking for K Ras. And you can see on the left here, all of the non small cell lung patients where we screened for CT DNA, and five of them had detectable baseline CT DNA. And over on the right, you can see that three of those five had molecular responses that correlated with treatment for over 12 weeks. focusing in on the K Ras g 12 c sub set. Here we had six subjects, and we’re looking now at the time on study time on treatment compared to their immediate prior treatment. And this is the famous PFS to PFS one ratio. The general idea being that as your cancer progresses, you will spend less time on subsequent your second and third lines of therapy compared to first so in other words of PFS to one ratio over one is soft suggestion of benefit. And if you look on the right, you can see that clearly we have that suggestion has benefited in several subjects. This is in fact the last patient be treated with sleep version one, a 53 year old man with metastatic non small cell who presented in the summer of 2020 and was treated with standard Pember lism ab plus chemotherapy, but developed rapidly progressive disease after just three months. He came into our study and began sleep immunotherapy in January of this year. And you can see that he had around 17% shrinkage at the first scan that then turned into a 47% shrinkage at four months. And this was associated with what is now a fairly common ctdna pattern for us. We have an initial spike in CT DNA, which we presume is tumor lisis followed by clear molecular response radiologically you can see that this patient had several clearly measurable target lesions. And obviously these all shrank as you’d expect given the data I just showed you. Unfortunately at around five months he developed relapse in a virtual body with evidence of acquired resistance and breakthrough in a vertical body, which lead to disease progression.
This was the second patient we treated with slate version one and 84 year old lady who previously progressed on Pember lism ab therapy. She didn’t tolerate chemo well came into our study and resolved for around six months with around 20% tumor shrinkage and a molecular response, as you saw on previous slide, she did not tolerate the fatigue associated with checkpoint inhibitors well, however, and came off of all therapy at around six months for improved quality of life. And then finally, this interesting case of 55 year old man with metastatic non small cell carriers g 12. c, again, progressed on pemra chemo after around eight months of therapy, and has been on study now for two years, actually, he’s just come off all treatment, he’s never been ctdna positive. So we don’t have that readout for him. As you can see that he’s had this very persistent, roughly 20% tumor shrinkage, he’s asymptomatic in good health and now off of all treatment. Now, the version one product that we had included some p 53 mutations, as well as K Ras and we learned that P 53. was immunodominant and probably crushing the magnitude of the K Ras specific responses. And you can see this here in an HLA a 11, transgenic mouse, where t 12 V and G 12 d can be presented that with the version one cassette looking at those dark colored dots, there’s really very little measurable T cell response. With version two, we’ve removed p 53. From the cassette, it’s pure K, whereas we have multiple copies of the carriers, mutants. And you can see now we get strong T cell induction. And this is the product that’s gone back into patients. And we’re dosing three cohorts firstline colorectal maintenance, third line colorectal, and another group of patients who are with who have non small cell lung cancer and have progressed on prior immunotherapy. So we’re waiting to get those data obviously, and we’re anticipating having some data roughly mid 2022. Let’s turn to granite. What Dan will show you is that again, we have good safety and tolerability with the granite personalized new action vaccine. He’ll describe the efficacy with the induction of strong CDA T cell response that’s associated with ctdna molecular responses in four out of our nine colorectal patients where we have measurable ctdna. And that correlates with clinical benefit. You’ll also see that the 12 month confirmed progression free survival rate is around 30%, which compares very favorably with historic controls. And then we’ll be moving this forward into randomized studies, which I’ll describe at the end of the presentation. And with that, let me hand over to Dan, thank you very much, Stan, over to you.
Thank you very much. It’s a pleasure to be chatting with you today. And building on the talk that I gave earlier at the Mini World. So we conducted as you heard in this Granick study, a phase one dose escalation study with adaptive design, holding the chat vector dose constant and increasing doses of the SAM boosts along with subcutaneously paluma map and IV volley map given monthly doses for six months and then every three months thereafter. You can see in the phase two expansion cohorts, which remain ongoing, and microsatellite stable colorectal cancer gas was off July no carcinoma non small cell lung cancer expansion cohorts and the primary objective of the study, safety and tolerability overall response rate by classic resist and recommended phase two dose because we’re selecting for new antigens, one question is are we enriching for higher tumor mutation burden or other biomarkers that would suggest that that we’ll have a response just to immune checkpoint inhibitors alone. And what you see here on the left, compared to the TCGA cohorts, historical cohorts is that that’s not the case. So the tmb is remains quite low and the patients enrolled in our studies you can see there by color coding of the specific tumor types and the PDL one also lower than than the reference TCGA chords and on the right, similarly, interferon gamma signatures are not enriched or upregulated any way in the granite patients compared to the TCGA. So with that, we see the patient demographics that were enrolled on the study on the left, typical median age range and you can see the numbers per tumor type, a large group were enrolled or MSS, colorectal cancer. And on the right similar to the slate, you can see that treatment related adverse events about 5% were most common. We were fever injection site reaction that were self limiting, and the remainder were two About what you’d see with immune checkpoint inhibitors alone. Very importantly, I think we see here that the vaccine strategy consistently induced new antigen specific T cells. And that the the baseline samples demonstrating a lack of a T cells in these patients really reflective of their general poor immunogenicity. So here are the resist data. And so you can see here that of the 22 year old patients that we had one complete response and a number of patients with stable disease with disease control rate of 27%. And so we’ll be focusing now on the MSS colorectal cancer patients because classically, you know, immunotherapies are not known to be effective in this disease. So we’re going to look at these patients here with stable disease and also patients who had progressive disease but were treated beyond progression. And so again, you compare to other tumor types colorectal cancer, that’s MSS specifically which again is the majority of the disease more than 95% of stage four diseases MSS. You can see here with PD one therapy PD one and ctla combination blockade really shows no benefit from these therapies. PFS on on the left and o s on the right. So that’s the baseline from where we’re coming from. And also in this cohort, there were some MSI high tumors here in at least two of the patients but not all patients had a known MSI high status and known microsatellite instability status. So we know that MSI high patients have clearly different potential for response from these immune checkpoint inhibitors alone.
And so in the patients treated in the grand study with MSS colorectal cancer, you can see here on the swimmer plot that patients are doing quite well better than the benchmark that’s that’s shown there. What would be expected from patients in this setting from historical control, and some patients remaining on therapy even beyond far beyond this. So this is promising. And when we’re looking at seat responses by resist and classic resist, where you can see that in the blue diamonds, where the resist base progressions are not necessarily predicting the longer term benefit, as as was being explained earlier. And so based on that, we also looked at other parameters to measure response other than classic resist, including immune resist to allow for at least a second confirmed progression. But also other things like ctdna dynamics, as you heard. And so this is showing the ctdna dynamics of evaluable patients that were able to get pre and post draws so far, some are not at the time point yet to get the next draws. And what we saw here, as was described, some of the patients had a transient uptick and ctdna. But then a nice decline with with significant decline. And you can see here, four of the nine patients are 44% of the patients had a ctdna response. And as has been demonstrated in many, many studies so far that patients that are having ctdna response really will predict longer term outcomes, even within two to four weeks of starting various types of therapy. And so this was shown here in this in this data set as well where those patients were having CTD in response or having improvement in overall survival. So you here you see here, the progression free survival by classic resist and also ipfs by immune resist, and particularly in the eye and ipfs where again, this is defined as progression at second scan to confirm progression, you can see particularly at long term, landmark at one year here, compared to historical control of what one might expect in a setting where there’s no one, you know, alive at that time without progression. We’re seeing a promising amount of patients here on the order of 25 to 30% this year, which is quite quite outstanding, and microsatellite stable colorectal cancer. So now a few details on a few of the patients one of the patients g8 as you can see here as a 50 year old female with metastatic colorectal cancer MSS. You can see that they’ve been through two lines of standard chemotherapy folfox Bevis ism AB full fury Bevis ism AB and continued on maintenance five, if you Bevis ism AB and you can see there the the characteristics of the biomarkers starting at baseline at the time of vaccination, showing elevated CEA elevated liver enzymes for liver disease, elevated ctdna and not upstanding or high extremely exceedingly high tumor mutation burden and a negative PDL one expression stain. And as has been described, and is becoming quite typical of what we’re seeing the tumor markers and ctdna. In this slide, it’s the tumor marker CEA shows an increase initially, but then a nice decline. And then ultimately has several months later, starting to see emergence of resistance. And similarly, with the CTD, now showing that exact same trend, and by ctdna, you can see a number of different mutations, some of which are here, new biallelic tap one mutations, suggesting that this might be the mechanism of acquired resistance through reduced tumor antigen presentation at this this gene is involved with. And again, you see here even on resist looking by row at different lesions, and by column at different time points, you can see that at baseline to the first scan, you can see a swelling of the of the lesion. And then as you continue to follow, you see actually decline. And this is really very in line with what we were seeing by the tumor marker and the CT DNA. And so again, resistance doesn’t capture this at one time point.
And similarly, in the same patient and Oliver, having two lesions at baseline, with one of them actually having a response initially from the gecko, and one of them just being stable without growth. You can see though, that the one that was stable, before and after on PET scan on the last row, showing a clear response by pet. So again, the limitations of looking at just the tumor sizes are emphasized here by looking at other parameters that also matter. And then in a second patient g 16, is a 63 year old female with metastatic MSS CRC and you can see their their treatment trajectory with first line folfox, chemo and progression, then panitumumab, and then progression, and that’s tuckson map. And again, their baseline biomarker results showing elevated ca 99 ctdna. And low tumor mutation burden, PDL, one was not done in this patient. And you see, you know, the same sort of phenomenon where it takes a little bit of time, by row and each of these different lesions that are being shown to to, to to shrink, but you’re seeing that response over time. Another thing to point out, I think, is that the response here overall by resist is you know, minus 16% minus 19%. These by classic resist, of course, are considered stable, because it’s not, you know, below a certain threshold that’s arbitrary, in some ways of less than 30%. But you can see clearly here that these are shrinking. And again, other things that I think are important are long term survival that we’re showing here on the tail of these survival curves. So with that, I will pass the baton on to Dr. Andre, and he will take it from here.
Yeah, pleasure to be with you in Paris with my French accent. And just a little overview of the situation of metastatic colorectal cancer, with a meter surgical corrector. Cancer is a frequent disease with its 9% of, of cancer in us and it’s not our country, and 9% of deaths for men and, and, and women. And it’s different stage, the local stage, the regional disease when it’s some nose stage recurrent colorectal cancer. And when the patient that metastasis is very difficult to cure the disease, the only way to cure the disease is to remove this metastatic metastasis by surgery. And it’s not chemo responsible of some long shot shot survivors. It’s a surgery. Sometime it’s possible to have surgery and the way to, to avert some patient to with those things is for IPS, it’s now some standard of care in first line, it’s five a few days a few will be imaging based chemo president especially mama it’s with oxaliplatin, or an or irinotecan and it’s with TJ finds his way tight and left side. And if the tumor is his right side, it’s replaced by by Bev and now and for dnn alpha nsi. Correct, I can throw the stone healthcare establishment on TP one and I will show the applications or curve the median of our survival is between 24 and 36 months The median PFS is between eight and 12 months and not a lot of progress in tenure we we with the progress of chemo have already started to therapy. The median PFS is still the same since a long time. In first line in seven lines of standard of care is to switch to chemo irinotecan or senator Tina and to do blood with Auntie vgf. And for patient with that prognosis with mutation of the half the 62 e, it’s a new standard of care. It’s with the face rays combination at city map and they have nidito agafay. The median survival in single line is 12 months as the median PFS is between four and six months we have the opportunity to assert line with some some some choice that suits his choices with very low median PFS to most and very low median Wes we have to go off and eat we have to create interfaces with delivering and in US and around the world. It’s opportunity to rent reduce and TJ far in in some patient to accelerating it was some facilitating retribution and as a new new IRA is to targeting to better diagnose everything and just just clinical trial that in third line was poor with poor whites and poor poor PFS just to see is a DFS DFS with duplexes to start off care. One of the standard of care is the PFS is increased by devices now but this study was published in in a long time ago now and the median GFS is still the same at with orders. Maybe a little bit better if if we use it to three convenience on the first day you know, express was bad, but not really fancy, strong thing that animals is our tradition is really important to consider the fact that we have three kinds of humor we have a call tomorrow we have a tumor with literally new immune infiltration and we have some tumor with huge filtration in metastatic rectal cancer, it’s only five or 6% with the
hot tumor in this tumor is in the large majority MSI and Holly emiko steady, steady, steady stable with some rare mutation like a poor imitation, the measure eg a cold or with amino acid tration The hope is a son of medical oncologist concerning the metastatic colorectal cancer was steady cannot 177 it’s important it’s only focused on MSI emitted steady correct cancer dmmr metastatic colorectal cancer and turbo improved the TSS is a p value is highly significant as our issue is all six barrels of chemo chemo, it works a little bit less in MSI patients. And it’s really interesting to see the beginning of the curve. It’s a racist iteration and one of the reasons I was across the curve is because we have some sort of progression. And if it’s racist progression, it’s an event in the curve. The result of overall survival was presented at a school overseas improved by editing persons the p value is borderline because measuring your patient receiving no couraging signal line and Tom who is now also on health care and this is for physicians MSI patients raise it as a copy of nivolumab and in every prep prep, prep treated patient was amazing. is a curve of PFS as you girls have never seen the inpatient every pre treated and now it’s it’s clear that some patients are cured and it’s renewed in the area of metastatic colorectal cancer without surgery. There is a curve of what what what was the past and what what what is your activity right now the the curves of film okay imaging is all patient die except Some with surgery and to PFS six months with ortho targeted therapy and the combination of folfox or furphy. We have improvement. It’s clear that the median PFS increased the median is increased, but the majority of patient died. And the leader may not realize that three years were engineer with surgery of metastatic disease. The blue curve is together families ma is the dream of the oncologist and we have to do better but it’s with Babylonia. It’s made it possible to do better with combi srtp when an artist GTFO and we have a plateau and we have patient two or three years we have now some patient with surgery with a complete pathological response in cc tuition. What about checkpoint inhibitor in microsatellite stable disease is don’t work. We have a study done by horsh. With Addison’s disease lab, we it was administered in Magnus after injection to maturity and no improvement by an anti PDL. To conclude it’s clear now we have two different correct cancer. We have NSI it’s 15% of correct cancer, but it’s only 5%. Because the majority of stage two or three, cure with surgery. And it’s it’s patient with Meiko Saturday testability is sporadic. It’s a whole the elderly population. It’s also the engine room. It’s more prevalent interproximal localization it’s with fragance 25% of mutation of DNR v 62. But immuno worse, their tumor mutation is very high with immune Effie tration and huge efficacy of checkpoint inhibitor. Examples that
checkpoint inhibitor can working in ngi. And in cells around we have the majority of the correct eye cancer 85% of metastatic colorectal cancer with it’s a it’s a proficient mismatch repair. It’s an average man with loss of marriage energy. And it’s prevalent in registered organization with more frequent mutation of chaos, low tumor burden and no caveats yet at this time, and is a reason it’s a huge unmet need and with the efficacy of immuno oncology in any size a dream is true to other research agencies in this population with immuno oncology. In conclusion, growing cancer is the single most common cause of cancer related this after the Durham cancer outcome with with janosik of metastatic disease is for only 14% at five years. At this time in more in relation with surgery chemo for metastatic disease is palliative for the vast majority of patients. We have insert line to PFS and why is immunotherapy give benefit only on MSI patient and impatient with mutation with rare, poor imitation and early detection of diseases are key focus currently and improved therapeutic approach into adjuvant and firstline sitting. Really important. Thank you for your attention.
Great, thank you theory. So let’s close with where we’re taking the granite program now. So as you’ve heard, colorectal cancer is a huge unmet need. And with our immunotherapy, we’ve clearly seen signals suggesting that we might be able to make a difference there. Now, what’s helpful from a trial design perspective is that we do not need to run a third arm in a randomized trial of checkpoint inhibitor alone, because as you saw from the Moodle data with the Tesla lism app, that question as to whether checkpoint inhibitors are active as monotherapy or actually sorry, maintenance therapy in this disease has been answered. And the answer is unfortunately not they’re not active. And that makes it an easier trial for us to contemplate. And so immunotherapy generally is believed to work better as you move upstream. Patients have better immune systems, less complex disease, and they typically have time to mount the immune response that we obviously need to see. And some of the patients that we’ve treated as you’ve seen have progressed extremely quickly. And unfortunately, I think we just won’t be able to help them given the rapidity with which their disease progresses because it takes somewhere between eight As sometimes as much as 12, or even 16 weeks for the T cell response to really hit its peak. So for this reason, we are moving upstream in colorectal cancer in microsatellite stable disease. And the study that we’ve designed first study 10 is a randomized phase two three study, in the first line maintenance of metastatic microsatellite stable colorectal cancer. This is designed as an integrated phase two three randomized trials. It’s a single protocol, but it has two discrete elements and open label phase two, followed by a pivotal phase three. So the inclusion criteria is shown on the left it says untreated stage four disease so either newly diagnosed metastatic de novo or a patient who previously was treated for earliest stage disease, but unfortunately their disease has has recurred and progressed. These are microsatellite stable patients B RAF wild type, which eliminates a small group of patients for whom there is a good alternative therapy as you heard, and patients who plan for treatment with what is in the United States the typical standard of care for Fox plus Bevacizumab. Patients will offer us a piece of their tumor tissue, and they’re screened for their Neo antigens. Roughly 60% of patients we anticipate will be accepted into the program 40% will be rejected for insufficient predicted new antigens. Those who are taken in are then randomized one to one. Those who they all receive exactly the same induction chemotherapy with folfox plus services about and during that induction chemo, we’re making products for the patients who are allocated to the active treatment arm when oxaliplatin is completed, and no later than 24 weeks, although most patients actually stopped a little sooner because of neuro tox, then the maintenance phase begins. And this of course, is where then the treatments differ between the active arm receiving immunotherapy plus five a few perspective versus the control arm five, a few bed alone for the phase two portion open label an end of between 40 to 80. So we have some flexibility written into the protocol. And the primary endpoint for that phase is molecular response as we’ve discussed. And then the phase three is approximately 200 patients. And the primary endpoint there, as written in the protocol is disease progression through ipfs criteria using IO resist for the reasons that we’ve discussed. Now, we took this trial to the FDA to discuss this. And they agree with the notion of running this as a not operationally combined single protocol, phase two three trial but to be very clear that phase phase three is statistically distinct from the phase two they are unrelated. ip ipfs per irises will be the protocol specified primary endpoint as the protocol is launched. But the FDA will review the phase two data before we initiate the phase three components of the trial, they wish to see more data, obviously comparing resistant irises. Importantly, as we’ve discussed, and for reasons you’ve seen a third treatment arm of checkpoint inhibitor alone is not required in this setting. And finally, an important nuance is that it’s acceptable to diagnose the MSS population, basically by taking all comers and then subtracting those who are diagnosed as MSI high based on local testing. So this study will begin in early 2022. Now, we wanted to move upstream further still. And as many of you I’m sure are aware, the utility of ctdna as a stratification tool is rising dramatically. And actually, in particular, in colorectal cancer. And the company natera has done some very nice work using their tumor informed CT DNA assay as a patient selection or stratification tool. And so the notion here is that you take patients who have stage two three disease, who have definitive surgery, of course, everyone hopes that they’re cured by the surgery. But sadly, somewhere between 25 to 30%, unfortunately, are not cured by surgery. And you can see that very early because they have persistent CT DNA in plasma. So you do as an important blood draw post op, you look for mutant DNA, and you’re looking for the mutations that you know are there because you sequence the the original primary tumor, so there hence the term tumor informed? If ctdna is present, then the patient is screened for new antigen sufficiency, as I mentioned earlier for the study 10. And if the patient does have sufficient Neo antigens, they’re then randomized to receive either immunotherapy as maintenance after adjuvant chemo, or simple observation after adjuvant chemo. The endpoints for this study will be CT DNA and disease free survival with overall survival of CA a remote endpoint. And CT DNA is very powerful to enable this kind of study, because the data from the terrorists suggests that patients who are ctdna positive post op will experience disease recurrence at around 12 months. So pretty much everybody has recurred by around 1415 months, which means this trial is efficient, everybody in the trial gives an endpoint. It’s also an important draw because these patients need help, they are going to recur, the standard of care, unfortunately, is unlikely to be sufficient. Hence the rationale for adding in immunotherapy in this patient group. And obviously, this is as early as we can go at this point, but obviously very interesting design. Okay, with that our presentation is complete. We have 15 minutes left if we want to finish the hour. So let’s hand this over to q&a. Back to Tara, please.
Thank you, Andrew. At this time, we’ll be conducting a question and answer session. If you’d like to submit a question you may do so by using the q&a textbox at the bottom of the webcast player, or by emailing your questions to questions at lifestyle advisors comm to our analysts, we invite you to join PSU please hold while we pull for questions. Our first question comes from Elizabeth Webster at Goldman Sachs. Elizabeth, please go ahead and unmute your line. You’re muted Elizabeth.
Probably not hearing anything. If you can’t get the microphone to Wordsworth, perhaps you could submit the question in written form. We can answer it through that medium. Yep. Let’s move to the next question. Please. Time.
Okay, great. The next question comes from Thomas Schrader at btig. Thomas, please go ahead and unmute your line.
Do you hear me? Okay.
Can you find some good afternoon.
So thanks for all the detail. And I have to congratulate the company on doing a huge amount of heavy lifting and in a difficult area. So it’s good stuff. But I have a question for Dr. Ken Dynegy. I think it’s pretty appealing that CT DNA changes predict some sort of bigger benefit. How quickly do you think you would trust ctdna data to say to a patient, you should really be patient here, you’re likely to benefit?
I think it’s a great question. And, you know, I think you see the trajectory that they’re that in this setting, there is an uptick. And so getting too excited early, I think would be premature. I think that, you know, certainly by a couple of months later, you would be seeing response by ctdna, objectively, and as I showed you that that typically predicts longer term benefit, and when you compare to patients who aren’t having such a response.
Okay, thank you. And then if we can return to syllogisms, now that I’ve looked them up, is ctdna really different than a tumor response? Or is resist just too high a bar? Does a patient better? Or can you? Are you confident that a patient will benefit if they have a 10% reduction in tumor volume? And is ctdna just more sensitive? Or is it really different? And I think he had one patient where the tumor got bigger and CT DNA went down? Is that just a top spot in time? Or is is that an interesting patient? Yeah, I
think I think that you’re getting it exactly right, where I showed the patient that actually had a concordance increase in size of the mass and increase in CT DNA, and then both eventually drop. And so you know, so the the first part of your question, in terms of you do, I believe the ctdna, more than the changes in the mat and the size on the scans? You know, this this concept of pseudo progression with immunotherapy is a known phenomenon now like, and there are a number of publications melanoma was is, you know, the primary side of that, where patients are looking like they’re progressing on scans classically, but cpdna doesn’t live. And there are publications that show that patients who are showing, showing decrease or discordances, with what you’re seeing on scan where you’re seeing by ctdna, the CCNA predicts the reality. And so as opposed to those patients were truly progressing. You’ll see the CTD continually rise. And so I think CCNA has been demonstrating to be very important and instrumental and helping to differentiate that. And I think that, you know, this is the tip of the iceberg here now with vaccination, because, you know, this is yet another variable that I think takes even a little bit more time as you’ve been hearing to generate that response. So you made even more accentuated, I think, this initial uptick before you see response, and patience, I think is going to be required. Great, thank you. Thanks, Tom.
Thank you for the questions. Tom. The next question comes from Sean At HCA right, Shawn, please go ahead and unmute your line.
Good afternoon, guys. And thanks for taking my question. My first question is on the patient’s relative T cell response compared to their cpdn level and essentially survival. Have you seen any positive correlation between these?
There’s not a strong correlation. I think we need more data to answer that question for sure. It’s interesting that some of the early patients, the ones, number g 123, if you go back and look at that T cell response, like they had some of the best T cell responses is interesting. Actually, they were dosed with the lowest dose of self amplifying RNA. And parenthetically, it looks as though that actually is the better dose. Self amplifying RNA is an unusual beast, because it does make copies of itself. And therefore unlike pretty much every other drug we give where you get what you’re given the self amplifying RNA, you get something more than you’re given. And we’ve been figuring that out. Some of those patients did remarkably well. And one of those patients g two is actually one of cataratas patients who had a very nice T cell response. Do you want to speak to that patient? patient? Nine, I think in
Yeah, so he had an excellent response, and was doing very well for for some time, before eventually developing brain metastases. But before that, he was doing very well, for longer than then, like, a year, year and a half before progressing.
I was thinking to jeetu, who had the even better response as well, the gastric cancer patient with the who had no of this disease at baseline? Oh,
okay. Not knowing them by the numbers, yes. So that that patient’s an interesting one that had stage four metastatic g junction cancer, that classically is a very bad outcome, less than 5% five year survival rate, and was doing very well, by the time that the patient’s vaccine was ready actually had a nice response to the induction chemotherapy that we were giving in the setting as a bridge to getting to the, to the vaccination. And actually we went to surgery because his baseline disease was was all ago, metastatic, so to speak. So the the primary tumor was removed. And the, the residual disease that was likely left behind in some metastatic lymph nodes was the concern. And that’s why we don’t typically do that, unless they have a dramatic response. But his expectation for recurrence was more than 80%. And so then the patient was vaccinated had extremely high T cell response and even to date now. Now, I think it’s been two years or more since surgery without evidence of disease, and a ctdna. Actually, postoperatively was a little bit elevated and now remains not detectable. So this is an example. I think that we’re alluding to moving this vaccination strategy to earlier settings, and as an adjective and to suppress a likely recurrence. And so far, this patient really fits that that situation, I think,
yeah, not a very nice T cell response. Very soft data showing we just need more data at a fixed dose, which obviously is coming next. Thank you for that that was helpful. My last question is on during the course of treatment, have you seen any specific mutations or types of epitopes that turn up that more often the responsive responders or the compared to the non responders? And or is there any other clinical biomarkers that you’ve seen, this has a positive prognostic value. So the obvious, negative prognostic biomarker is a very high CT DNA baseline. If you have a CT DNA, measured in 500, or greater, and this is in human genome equivalents per mil, which is the now standard units used for measuring ctdna. If you’re over 500, you’re going to progress extremely quickly and probably before our therapy has chance to help you. And again, this is one of the benefits of moving upstream is that we hopefully will not have those patients being very numerous in our sample set, because we probably can’t help those core subjects. So that’s the most obvious prognostic factor we’ve observed. There’s no other common mutations shared between patients and remember, of course, that with granite, pretty much every mutation is personal, right? These are your passenger mutations. There’s very little sharing between subjects. So there’s no common pattern that we’ve observed in terms of that mutational architecture. One thing that did pop up and Dan mentioned this, in his talk, just now was this very interesting observation of byli lick tap one mutation in that patient. g8, who was doing very well had a complete molecular response was was heading towards a year, and then around a year we started to see mutant DNA reappearing in her blood unfortunately, So we kind of knew that she was going to really logically recur, which she did at around 15 months with with a liver metastasis appearing on radiology. But what was noticeable is that we screen actually for some sort of candidate resistance mutations. And she developed a biologic tap one mutations, meaning she had one mutation in one tap one meal and a different mutation in the other tap one early on, which obviously, basically doesn’t happen by chance, and is a very strong evidence that indeed, it was evasion of the class one directed immune response that led to acquired resistance for that patient. So that was very interesting. And we’ve got some other data that will probably present at a scientific meeting, where we’ve seen similar phenomena in a couple of other subjects with delayed acquired resistance, which is obviously good news, bad news. The bad news, of course, is this resistance. The good news is that it does at least suggest when we see these recurrent alterations in class on antigen processing and presentation that our therapy is doing what it’s meant to do. Thank you for the additional color. And thanks again for taking my questions. Very good. Thank you, Shawn. Okay, I think we’re gonna move to written questions now. So I’ll read out the questions. The first couple of questions come from Mark frame at Cowen, his first question, do you see differences in the ctdna response for sequences included in the vaccine cassette versus sequences that were not included? As his first question, the simple answer is, generally No. All of the mutations traveled together. And if you go back at your leisure, and look at the slides, where we showed you the CT DNA data, for example, for patient g8. On the left on that slide, you can see a huge massive lines, those are all of her tumor mutations being tracked over time. And then on the right, were her 20 vaccine included mutations, and they basically all travel together. So we’ve not seen much evidence yet of subclone ality happening, meaning that you see some clones remaining suppress certain mutations remaining low, but others breaking through on a resistant clone. We’ve generally seen all the mutations traveling together. Then there’s next question, how did the CT DNA sequences that are declining, compared to the epitopes, against which you can detect meaningful T cell responses occurring?
It’s a little challenging to answer that question because we don’t get enough blood typically from a regular blood draw to actually measure single antigen specific ch T cell responses that just aren’t enough T cells in the standard 10 an old blood draw. So we have to measure using many pools of peptides covering multiple candidate Neo antigens. So can’t really answer that question currently. And then Mark’s final question, how does the evidence for ctdna responses being meaningful differ between immune therapies like this versus traditional therapies? So that’s an interesting question. I’ll maybe take a stab at it and then hand it back to Dan and Terry, obviously, with the new therapies where we’re heavily incentive to start looking for better biomarkers because radiology is failing us. We’re seeing a lot of evidence now that in particular, stable disease doesn’t seem to mean very much. It can be bad, it can be good. And ctdna helps us to associate those those two populations. immuno Core i mentioned earlier has shown that in their patients who actually were labeled as progressive disease by resist when they broke them into ctdna responders versus non responders. Even people who were labeled PD were actually did very well had good long term survival outcomes. So clearly ctdna immunotherapy is really helping provide illumination of this very large and rather useless SD stable disease population. And there’s less work being done I think conditional therapies because resist works okay for traditional cytotoxicity targeted therapies. So the incentive to do the work is not there. I think it’s there now with immunotherapy. But let me hand over Dan to any comments on this question.
to agree with your comments. It’s a clear relationship between the efficacy of chemo and cDNA and metastatic disease is not used because complex at this time because it’s no library with a lot of of passivity that international sitting it’s no no progress since a long time and really all the all the study not all words are measured yesterday use cnn to try to select the intricacies effect of authority and 83 a new stat and for for especially for for 5g Vantage two key points for metastatic and an immuno oncology is probably also a key point. Because pseudo progression is really a problem and it’s a way to better to better under stands as a sort of progression at the beginning. And it’s very important
to thank you. Okay, let’s move on. So Elizabeth Webster, we now have your question. I’m sorry, we couldn’t hear you live. But thank you for writing it in. This was question from Goldman Sachs for granted a molecular response is defined as a 50%, or greater reduction in ctdna. From baseline, could you provide some color on that threshold and how it was set? The simple answer is that now widely used in the literature, so that’s why we chose 50% as the threshold some people have used 30%. But I think most folks now settling on a 50% reduction as being standard. Dan Tarija, would you agree with that?
Yeah, I would agree with that. And I think it’s somewhat arbitrary. I certainly see that that’s the threshold that’s been adopted, and generally predicts the longer term benefit.
Right. Thank you. Okay, let’s move on. Next question is from Yuan g. v. Riley. Yuan asks, just like COVID vaccine can induce T cells that have activity towards a set of T cell epitopes. In your clinical trial, this slave vaccine Did you measure T cells from patients to see one if T cells are induced to how strong the responses towards the set of T cell epitopes? So yes, we absolutely did that you want and we were disappointed with the T cell responses in with version one, based on the granite data, and we will we started growing a little bit before slate, we were expecting to see strong CDA responses measurable in hundreds of T cells per 10 to the six pbmcs in the standard, overnight elispot assay. And we saw that in a couple of patients, we did not see that in the majority. And we were puzzled. And we went back. And interestingly, because we put p 53. into this version 1% and P 53. was presented by some common natural aliens, we can actually measure the P 53. specific response as a control. It was useless because the patients didn’t have p 53. mutations in their tumors, but it was actually quite a useful vaccine immunogenicity control. And what we learned was that B 53 was driving a very strong CDA T cell response. So we appear to be defining what’s now thought of as a hierarchy of immuno dominance within human shared new antigens, which obviously is a new area and no one knew this before. But it looks like p 53 is clearly immuno dominant over k Ras perhaps because there’s more p 53 HLA peptide complex on the surface of the tumor cell. And that’s what drives its apparent dominance. So for version two, we kicked out v 53. We repeated k Ras, and you observe I showed you the mouse data transgenic mouse data, suggesting that we should see superior responses in humans, which we anticipate will lead to better clinical effect. So we’re hoping to see a further amplification of the effects in non small cell and perhaps we’ll start seeing effects of benefit in colorectal. Okay, let’s keep moving. Next question is from Wayne Coker at Dover solutions? When can we expect to hear any further readouts regarding this program? So we’ll be launching the phase two randomized studies, as I mentioned in the first half of 2020 to the O 10. Study, the maintenance study is open label. And so we should have some data in 2023. So think about the relevant timeframe. For slate, we will expect that data from v2 in mid 22. So that’s a much more near term catalyst. Next question is from Sultan Beardsley at MSN. Does the data potentially support accelerated approval in third line, MSF CRC patients, particularly the ones with liver metastases? I think the simple answer today is No. Because the FDA has not accepted ctdna response as a surrogate for accelerated approval. Obviously, everybody’s watching this space closely. But to this point, the answer is no. Because we haven’t seen resist radiologic responses in colorectal then obviously, we would not expect to use resist response as the basis for accelerated approval either. That’s why I think the right path is to do a randomized study using alternative endpoints. Next question comes from fatty Qur’an. Do you expect any support from the government NCI or other foundations as you continue to lead the science on CDH cells for treating tumors? We’re always open to that. And James galley who leads the NCI tumor vaccine division is a member of our scientific advisory board. So obviously, we we speak and work closely with James and if there is an opportunity for us to work together, we obviously would really enjoy that I’m sure. Okay, next question from an attributed question. Are you looking for ctdna increases as a predictor of loss of response that is not included as a predicted new antigen that could inform a new vaccine? Okay, I think I get the question. Yeah, we certainly are not During those data, if we did start to see evidence that a particular mutation, particularly subclonal, was breaking through when other mutations were remaining suppressed. And we did not think that that was a good new antigen. That would be evidence, perhaps that we need to think again. So I think a simple answer would be yes, we are collecting those data. But again, because every patient is different, it is very hard to draw broad general conclusions, because all of the mutations essentially with granted are distinct. Next question for patients who do not progress rapidly, what is known about the mechanism of response? Okay, are you looking for ctdna increases as a predictor of loss of response? Okay, so I think that’s more of the same question. Really? I think we’ve covered that.
Okay, last question. I think, what from Nick Abbott at Worlds, why do you think the GA patient says as for gastric, gastric esophageal adenocarcinoma patients did so poorly, too much early progression? Question, Mark. Dan, you’re probably the right person to answer that question.
I think has been alluded to a lot of these patients are sick. And so by the time the vaccines ready, whether it be whatever tumor type it is, when we’re looking at later stage disease of these, these tumors, it’s too late. And it’s too late to mount the response. So in patients that we can get it to them early, I think that’s where we’re going to have a bigger impact. Pretty good.
Okay, those are the questions. Any closing comments, Terry? Dan, anything else?
Just to comment about the fact that two other induction therapy and 12 during induction therapy is the time to, to do and to prefer the the vaccine is very important. And for that correct, I cancer is really is really a good choice. And also, because it’s very ethics, because all patient procedures, not healthcare during the first part of the therapy, and it will be very easy to improve patient in Australia.
Very good. Thank you to you, then. Yeah, I
100%. Agree. And not much more to say other than I think that as as we use that as a beachhead. And if we’re starting to see, you know, what we’re wanting to see, then that can be expanded to other tumor types, like ask yourself Joe cancer, but an earlier sentence.
So the point is, as a genome project, because we have already to improve the scenes and CNA is is crucial. And we have the CNA patient with a very high risk of prolapse. And it’s not it will be not necessary to have a lot of patient to see some things like we had in the past with a huge phase three in Nigerian sitting, it’s really a way to select patient with bad prognosis. And it’s, it’s really not necessary to have huge trial but to to select patient with a risk. And with that probably needs to be rapid to other answer to the question is not a lot of patients. That bad
prognosis but low tumor burden, which is I think the time to come in with something like this to suppress in any tumor type. I think that’s going to be important.
Very good. Excellent. Well, thank you very much. First of all, cherry thank you for making us feel like we’re in Paris. We really appreciate it. I wish we could have been with you in person.
It’s more meeting is real.
That’s right. Thank you, Dan, for traveling. And thank you all for your interest for your questions. We appreciate it. And obviously you can reach us through Investor Relations, that gritstone bio if there are any further questions. And with that, I think we will close the meeting. So thank you, everybody.