 Good morning. Thank you for the invitation to present our work on immune checkpoint therapies, which anti-CTLA4 is the first in class. Anti-PD1 comes next in this class of immune checkpoint therapies. And I'm going to go through some of the background and I apologize if it's a little bit repetitive from what you may have heard already this morning. So one of the big questions I guess we've been asking, and this has been ongoing for years, this is nothing new, I think centuries for the basic immunologists and those of us in the clinic who have been trying to apply immunotherapy, why immunotherapy? And for a long time the room would empty out as somebody else mentioned whenever immunotherapy was mentioned because we really didn't have clinical efficacy to talk about, but why were we so insisting on pursuing it? And one of the things we know clearly is that the T-cell responses are pretty powerful. So if you get a viral infection, your T-cells will mount a response, it will eliminate the virus, and the next time the virus even comes into your body it will mount that response before you even get symptoms. So it's capable of recognizing the antigens, it's capable of producing memory cells, and it's capable of producing a response to another antigen. So if you get a second virus that's not even related or a bacterial infection, it can recognize multiple millions of different antigens. So why is it not recognizing the tumor antigens was really the big question. So if we knew how to harness the power of the immune system, which we already do in the infectious disease process, if we could do that in oncology we would really have a home run winner here. So this is just an electron micrograph showing you that these little T-cells can actually recognize the big tumor cells and eradicate them. So the rationale really for developing immunotherapy treatment for cancer patients, we have many pieces of evidence that we've been putting together through the years. We know that tumor infiltrating lymphocytes associate with good prognosis in a majority of tumor types. So if you look at colorectal cancers of ovarian cancers or melanomas, you know that clearly the larger number of T-cells, CD8, T-cell, CD4 T-cells, these tend to portend good prognosis. We also know that there's expression of tumor antigens on many of these tumor types that the T-cells can recognize. If you take the T-cells in vitro, they will recognize the cancer cells in vitro. Why isn't that happening in vivo? We also know that for cancers, you really need mutations, right? I mean that's how cancers are able to escape all of these therapies. They have multiple mutations in the cancer cells that allow it to grow and survive. These mutations really represent new antigens. They're no longer self-antigens. They're new antigens. So why isn't the immune system capable of now taking all of these new antigens and eradicating these tumor cells? That really is what it should be able to do. And again, it doesn't matter if it's a driver or passenger antigen, right? I mean, we're having this big discussion in the world of mutations. Is it the driver, a mutation, or a passenger? The T-cells don't care. It's a new antigen and the T-cells should be able to recognize it. I think the problem that we really have to face and deal with was that we were initially just focused on one T-cell receptor. I'm going to try, oh, sorry. We were really just focused on this one piece of how T-cells function. We really thought that it was all about the T-cell receptor recognizing the antigen in the context of MHC. Unfortunately, this is just one signal. So for those of us who spent years and decades just given a vaccine, a peptide vaccine antigen to our patients, hoping that that peptide vaccine was going to initiate a T-cell response because the T-cell receptor could recognize it in the context of MHC, I think we all read the papers. Those vaccine clinical trials did not lead to clinical benefit. And that's because we really did not understand the basic biology of T-cell responses. So it's not sufficient to give the antigen. That will get the T-cell receptor going, but that's all. You will not get anything else going. You really need to now figure out how you get the second signal, which is co-stimulation in CD28. It is a basic response. Every time you get a viral infection, it's T-cell receptor and CD28 that work together to help eliminate that infection. So those two signals are what are important. But the other thing that we have to recognize, and we only recently recognized, is that every time there's an on signal, the on signal drives an off signal. So in your T-cell, my T-cell, and everyone's T-cell, every time we get an on signal, there's a transcriptional regulation that then allows for the off signal, which is CTLA4. So your T-cells have a programmed amount of time to do their job. And then there's homeostasis and return to balance, where the off signal tells the T-cells, we're done. You've had enough time. This may work very well for the flu infection. It may not work so well in large tumor settings, where the tumors are continuously trying to evade the immune response. And now the T-cells have lots of CTLA4 on the surface, because they've seen the antigen for period of time. And CTLA4 is the only thing now that's preventing these immune responses from eradicating the tumor. There are other off signals. And I think this was the biggest point we learned in basic immunology. They're on signals were what we were trying to drive the entire time. The off signals actually are very, very important. We have to learn how to block them. CTLA4 is only the first that's been recognized. There are other ones, such as PD1, as we've learned. So now how do we start blocking the off signals? And how do we now allow the immune system to really take off to eradicate the tumor cells? Jim Allison's group was actually the group that identified the T-cell receptor protein structure. It was also the group that identified CD28 co-stimulation, and also the group that identified CTLA4 as an inhibitory molecule. All basic immunology. Now we can take these pieces and put them together in the clinic. His group proposed that if we used an antibody to block CTLA4, we would allow the on signals to continue without any off signals. And that would allow for tumor eradication. And so they developed this antibody against CTLA4, and they did a lot of beautiful preclinical models, which I'm not going to show you, and published the data and mouse models that any tumor type could be rejected just by blocking this one molecule. The antibody was then developed by Metarex and Bristol Meyers Squibb and was taken into clinical trials. And in the phase one clinical trial, one single dose of anti-CTLA4 was given initially. And so this is a patient from the phase one clinical trial, where you can see there's metastatic melanoma in the lung, and the patient received one dose on that phase one trial, and the disease was eradicated in that swan patient. The patient never got another therapy. This is now 10, 13 years later. This patient is perfectly fine without any recurrence. Again, speaking to the IDA that if you can get the immune response going, there is a memory component that can keep this disease under control for years. So anti-CTLA4 does not work on the tumor cells. It works on your T cells. Every patient's T cell, your T cell mind, they all have CTLA4. So if you get anti-CTLA4, it's working on your T cell. That means you can use it to treat any kind of disease. Okay, you can use it to treat prostate cancer, melanoma, kidney cancer, ovarian cancer, pancreatic cancer. Technically, it's about getting the immune response going. So here I'm showing in a prostate cancer patient this large prostate mass, and the patient also had metastatic disease. This patient participated on a clinical trial where they received anti-CTLA4. And again, this patient had a complete response where all disease was eradicated. The phase three clinical trial when anti-CTLA4 was completed in melanoma. And the one thing, although everybody is very excited about this increase in median overall survival, thing I really want to point out to you is this tail at the end of the curve. Everyone else in the control is dead. The tail at the end of the curve after the patient's finished therapy continues. So now four years, five years, six years out, there are still patients alive who had metastatic melanoma and received treatment anti-CTLA4. They're not receiving any additional therapy because they completed the trial, but they're still doing very well and surviving. So in summary, we found that CTLA4 blockade has a consistent anti-tumor response rate of approximately 10 to 20%. There's durable partial and complete regression of disease. The survival benefit was now seen in two phase three clinical trials in patients with metastatic melanoma. FDA approval of the drug was provided for melanoma in 2011. So this is a kidney cancer meeting. So just to let you know, there has been data in the phase two setting for patients with metastatic kidney cancer, and this was published previously in 2007 where they did see by research criteria these partial responses. But the important and exciting thing is that we have an understanding of the basic immunology. And once you have the basic science, there's so much you can do with it now, right? So CTLA4 is the first. The next one to come along is PD1 because now we know we have to start looking for the inhibitory pathways. Now we understand that. It's not just about driving the T-cell receptor with a particular antigen. You have to now block the inhibitory pathways. So when we blocked PD1 in patients with metastatic kidney cancer, this is what we saw. There's a patient, this was published data from Julie Grahamers' group, and you can see that the metastatic disease goes away. And not only metastatic disease in visceral and soft tissues, here's a bony met. And this resolves with anti-PD1 therapy. And as they obtain biopsies, you can see that the patient goes from not having a lot of T-cells in the tumors to having tons of T-cells within that tumor tissue to allow for the tumor eradication. This is the same type of data that we saw with anti-CTLA4. So we know that anti-CTLA4 increases the T-cell infiltration and T-cells produce interferon gamma. Interferon gamma, it turns out, actually up-regulates the PD1 and PDL1 pathways. It up-regulates PDL1 expression. So now if you block with anti-CTLA4, not everybody responds, only some patients. But the reason not everyone's responding is because a second inhibitory pathway comes into play. So now in the mouse models, Jamalison's group again showed that if you use anti-CTLA4 and combine it with anti-PD1 or combine that with anti-PDL1, the mice do much better. And now you're able to reject even more tumors, you're even able to cure more mice. This was then taken to a clinical trial where we now combine anti-CTLA4 with anti-PD1 and granted this is in metastatic melanoma, but it's also ongoing in kidney cancer where we're combining anti-CTLA4 and anti-PD1. But this data was published in melanoma and for those of you who are familiar with these spider plots and the waterfall plots, you can tell that this is really dramatic responses that we're seeing in stage 4 patients in a phase 1 clinical trial where these patients are having phenomenal clinical responses, 50% activity of this drug for these patients now. So what are the issues that we have to address? So we really need to understand the cellular and molecular mechanisms involved in the anti-tumor effect. Can we identify predictive, prognostic and pharmacodynamic biomarkers? What are the best standard of care therapies that we can combine with these immune checkpoint therapies? And then can we identify new targets? What are the other inhibitory pathways that we need to consider to help improve the responses? So for doing this, we really have to move efficiently between the clinical trials and the laboratory interrogation. And so just to remind everyone that in the lab hypothesis testing is wonderful in mouse models, okay? They're in bread and the disease is homogeneous. The question though is how do we now find out what's happening in the patients to really answer these questions that I posed on the previous slide. The patients are polymorphic and the disease is heterogeneous. So the most we're going to get out of the clinical trials is hypothesis generation. We're going to generate some hypothesis. It's correlative data. And the important thing to do is really go back to testing it in the lab so that you can now really say this pathway does this. And then how do you take it back between basic immunology and the clinic, which is really the important step that led us to having clinical benefit now. So to rethink the clinical trial design, we've now proposed that when we're doing these phase one, two and three clinical trials, which you're all aware of, phase one being the safety and dose escalation, phase two being the efficacy and phase three comparison standard to care, we've now integrated phase one and phase two eight trials. And these are trials that they're built really to answer the questions of biomarkers and mechanisms. They're not really built to answer the clinical endpoint question of is this drug going to give you a certain number of PRs or CRs. So for example, this trial was conducted and completed. It's a pre-surgical clinical trial with anti-CTLA4 because again, the pre-surgical part I want to stress, we need the tumor tissues if we're going to understand what the T cells are doing, what all those immune cells that you just heard about, what are they doing in the tumor microenvironment, how is the therapy affecting them and what can we come up with next for combination. So it's the tumor tissues that guide us in terms of the biomarkers. So here patients with localized bladder cancer receive two doses of anti-CTLA4 before they went to surgery and then we were able to collect all of the blood and tumor tissues for the analysis. Similarly, again I told you anti-CTLA4 is not specific for tumors. So here we did prostate cancer. These are patients with localized prostate cancer and because we're trying to identify ways to combine these immunotherapies with standard of care therapies, here we did a clinical trial where we combined anti-CTLA4 with hormonal therapy in these patients and so they got one dose of Lupron, two doses of anti-CTLA4 and then these patients went to surgery. This is a third trial. This is the pre-surgical trial in patients with kidney cancer and in this trial we have anti-PD1 alone, anti-PD1 plus Bevacizumab or anti-PD1 plus anti-CTLA4. So it's a three-arm trial in patients with metastatic RCC but all of these patients are scheduled for nephrectomy. So we're going to have all of those tumor tissues and then they can continue on their therapy after completing surgery. But now we have the tumor tissues and the blood samples for all of the immune monitoring that we need to do. And just to show you again we have access to all of the tumor tissues which are very important if you're going to get the infiltrating T cells because that's the only way to get them out is to actually have tumor tissues to sort them from. We have non-malignant tissues as comparison and the prostate again we're able to have tissues from both bladder and prostate. And this way we can then start with unbiased analyses. So for example we start with an aphymatrix analysis where we compare controlled tissues and again controlled tissues being stage match patients who go directly to surgery because that's a standard of care. They go directly to surgery, they don't participate in the clinical trial but that's our control and we sort the T cells again and we can sort the tissues and we can do gene array analysis. We can then ask for what's the overlap between an anti-CTLA for monotherapy plus an anti-CTLA for plus hormonal therapy or an anti-CTLA for plus anti-PD1. We can start to do all of these overlap to look for the immune response gene signature that's coming out with these different immunotherapy agents. And then we can rank these pathways and genes based on an ingenuity analysis and when we did this initially we came up with this list and you can see at the top of the list is this pathway called the ICOS pathway. So ICOS really stands for inducible co-stimulator. It belongs to the same family actually as CD28, CTLA4, PD1, PDL1 and L2. It's actually only expressed on T cells and its expression is increased on activated T cells. It interacts with a ligand that's expressed on the antigen presenting cells and has a very diverse role that's been reported but not really a role that's been established in anti-tumor immune responses. So we sorted these T cells. First we looked at the T cells in the bladder and prostate here and you can see in non-malignant tissues versus untreated tumor tissues and then anti-CTLA4 treated tumor tissues. By flow cytometry you can see that there's very few ICOS positive CD4 T cells in the non-malignant already untreated tumor tissues. But after treatment in anti-CTLA4 there's this increase in the ICOS positive subset of cells and that was seen both in the bladder cancer patients as well as the prostate cancer patients despite this combination therapy which is telling you that it's really a result of the immunotherapy treatment. We sorted these ICOS positive T cells, we went back to the patient's tissues and here you can see that these T cells are in the tissues. These tumor tissues express this antigen known as NYES01 which is a tumor antigen and when we sorted these T cells out we could show that they really do respond to this tumor antigen. They produce interferon gamma, TNF alpha and MIP1 beta. So these are antigen specific T cells. We now had an idea from the tumor tissues what to focus our studies on in the blood and that's the important thing because lots of us collect peripheral blood but what are you really going to focus on with the peripheral blood? There's so many things you could do with it and then the data becomes really muddy, you don't know what to make of it. The tumor tissues allowed us to focus in on the blood what to look at and here we went back to looking at ICOS and so you can see in the pre-therapy sample versus after the first dose of anti-CTLA4 and the second dose of anti-CTLA4 there's this dramatic increase in the ICOS positive CD4 and ICOS positive CD8 population. So it's both CD4 and CD8 T cells that are increasing ICOS expression. So our changes in ICOS expression associated with clinical benefit and to do that we did a retrospective analysis in patients with metastatic melanoma on the phase 3 clinical trial that was completed and in the small retrospective analysis we could see that patients who had a sustained increase in the ICOS positive T cell subset did much better than those who did not. So now we've finished the clinical trial data and we've generated some hypotheses. The first is the ICOS-ICOS ligand pathways necessary for effective anti-tumor immune responses in the setting of anti-CTLA4. So to test this hypothesis we went back to the mouse models and in ICOS knockout or ICOS ligand knockout mice because you can knock them out in mice, not in patients. It's very nice to do it this way. You can see the wild type mice really have improved anti-tumor responses compared to these controls but if you knock out ICOS or ICOS ligand you can see now you have impaired anti-tumor responses indicating the importance of the pathway in generating this anti-tumor response. The second hypothesis we had was that this pathway could now be targeted to improve the anti-tumor responses and to target this path will we combine an agent of targeted ICOS and the ICOS ligand with anti-CTLA4 and here you can see any of the monotherapies really did not give you that dramatic an increase but once you combine targeting ICOS and anti-CTLA4 now you have improved anti-tumor responses and tumor rejection. If you do the same experiment that you did in the wild type mouse now do it in the ICOS knockout mouse you really lose all of these responses indicating the importance of this pathway. So we have another molecule that we could target the ICOS-ICOS ligand pathway but what I want to point out is that since we now have this new understanding of the immunology that's important for T cell responses here are all the other pathways that have been recently identified. So CTLA4 and PD1 yes they're important but as you can see there are a lot of other pathways and on this slide there's not a single tumor cell. This is a T cell antigen presenting cell B cells and K cells macrophages none of this is targeting the tumor cell. So it's now really going back to understanding how these pathways function and how do we target these pathways not just for kidney cancer but probably for a lot of other tumor types. What we're hoping to do is that we've all been used to this survival curve you increase the median overall survival but everybody dies at the end and that's not what you want. Anti CTLA4 gives you this sort of survival curve you can see about 20 to 25% of the patients with these long term survival and what we hope to do is improve that as you saw with the anti CTLA4 and anti PD1 combination so that we continue now to really target multiple immune therapy checkpoints and as well as multiple other standard of care therapies to give you this improved survival. So thank you there are lots of people to thank they're all listed on this slide because the work I showed was both the lab and the clinic and I'm very indebted to all of the surgeons pathologists and the clinical folks that we work with. Thank you and of course the patients who participate in these presurgical clinical trials to let us do these types of studies thank you very much.