 So, setting the scene for treatment of kidney cancer in 2014 is probably something very difficult because we still have a lot of unsolved questions. And what I'm going to do in the next 20 minutes is to give you my idea about what we know about this question we have about predictive factors, role of surgery, role of observation, choice of first line. When to switch from first line to second line, and we still have an open question here, how we should choose second line and finally what's are going to be the new targets. And of course, I mean, you will hear about all this question in the symposium, but what I'm going to try is to give you an idea of what I think here. So first of all, I mean, you are going to hear later on today, I mean, about genes and things are changing very rapidly in kidney cancer. We have a lot of new genes coming and these genes certainly have prognostic values. So one of the most important thing is that we now have new genes that probably some of you really very well know and some of you may be here for the first time. And if you look at CTD2, BAP1, PBRM1, one of the interesting thing is that all these genes are chromosome-free, same chromosomes than VHL gene. And that's for me a very interesting observation because when you look at these genes, they seem to be quite mutually exclusive. So this is an interesting paper that was published by the Jim Brugarodas group showing that either you have a mutation on PBRM1 gene or BAP1 gene, but having both of them mutated is quite exceptional here. And that's interesting because these genes have a prognostic value. So when you look at the survival of patients, depending on whether they have a mutation on BAP1 or PBRM1 gene, and these two different cohorts, one from TCGA and one from UCLA, which validated this observation, PBRM1 gene is a good prognostic gene. So if you have mutation on PBRM1 gene, you will have a better survival. If you have BAP1 mutation, then you have a worse prognosis here. So certain and interesting observation, and this is what you have when you put all these two genes together. You can see that in the very rare situation when you have both genes mutated, and I told you it's very rare here, the expectancy of life is very, very short. So of course, these genes are probably going to become targets for new drugs in the future, and we still are waiting about knowing exactly the function of these genes. But as you see here, they are very prognostic here. We probably also have genes that are predictive of recurrence, for example, and there was a very interesting publication at ASCO some years ago, which was recently published in Lancet oncology showing that CMET genes might be predictive for recurrence here. And one of the questions is, are we going to get a signature of recurrence in these patients here? And I certainly encourage you to come at ASCO. I'm going to present, I mean, a validation study of what Brian really showed some years ago at ASCO, showing that we can take some genes and putting some genes from angiogenesis immune response and some cell adhesion gene, we can predict probably the risk of recurrence. So I will show you at ASCO that this is probably something that is going to become a very interesting test, I mean, to predict and select patients for a given trial, for example, and also for the question of monitoring patients here. So moving to surgery, I mean, there is still an open question about cytoretically nephrectomy. I think we heard very recently from the Hang and Tony Shoray consortium very interesting data at ASCO GU showing that, of course, I mean, if you look at a large database, I mean, cytoretically nephrectomy seems to be better than non-doing cytoretically nephrectomy. So I always think that this is bias, and I'm sure this is bias, but the interesting observations that they did in this presentation is that, of course, I mean, the more you have risk factor in this setting, the more the benefit is becoming lower and lower. So if you look at risk factor 4, 5, and 6 in the IMDC classification, you can see that the benefit of cytoretically nephrectomy is probably no present here. Well, if you see one of two risk factors, it seems that there is a benefit in terms of survival here, 30 months versus 22 months, or 20 months versus 10.2 months. So that's certainly a very interesting retrospective observation. I still think that in the presence of this type of patients, I don't know whether we should do nephrectomy or not, and you can see that this is a relatively big tumor, but still, I mean, the tumor burden is quite high. I'm not sure that we know exactly if we should do nephrectomy in this patient, and I still think that the carmenotrile, which is ongoing, is going to be very important to answer this question, and I hope we will finish it in the next two years from now, showing whether, I mean, nephrectomy followed by sunitinib will be better than sunitinib alone or if it will be equivalent in terms of outcome. Third point I want to speak about is observation. So what we have put in the guidelines at ESMO 2012 is that some kidney cancer are very in the long course of disease, and we all know that this is the case in some of our patients, and certainly justify a period of observation. And we have data from prospective trials. This is a target trial where we have shown that even when we delay giving seraphenium in this setting, I mean, we still have tumor shrinkage and PFS, which is very equivalent. And we also have the same data from the record one trial showing that the PFS in the placebo patients are exactly similar when you cross over to the drug, so at least we can wait for a certain period of time. So question for me is how long should we wait? And that's coming for one observation I have done recently when I recently presented data of long observation. I would like to share with you two cases which really trigger my question here. So this is a 58-year-old male. He had surgery for a clear cell gastrinoma, and he had small lung nodules in 2012. And as I always do, I mean, I suggested to do observation first. And this is a scan in 2012, and this is a scan in 2012-2013. So very, very little increase in the tumor burden here. So I still consider observation, and I had to see him three months later. But in the meantime, I mean, he had this bone mass which occurred, which I never suspected to be. First observation, so some patients do that. Second observation, I mean, same kind of man, I mean, 59-year-old right nephrectomy, great free tumor. He had a rejection of a lung nodule, and he developed lung nodules in March 2012. So very slow-growing disease. In October, I mean, he had not really significant change in size of these nodules. And in March, I mean, started to tell me, I mean, I would like to go to treatment. And in fact, I mean, the increase in lung nodules was very minimal at that time. And I said, okay, we're going to start treatment, and I tried to enroll this patient in a clinical trial. And then I did a brain CT scan, and that was a brain CT scan. It was absolutely asymptomatic. So it was, for me, a question of when should we stop to observe patients here. And one question is, is tumor burden a potential predictive factor? And I think we have done a very interesting work with Roberto Jacovelli when he was in my institution, looking at tumor burden as predictive factor of prognosis here. So what we did with Roberto is to look at patients in Gustavoci who were screened for two big trials, the record one trial and the target trial. And we look at whether tumor burden was prognostic or predictive. And interestingly, when you look at this, at this court of patients, first of all, I mean, of course, I mean, tumor burden is becoming higher when you are in higher risk of classification by memorial classification. Of course, tumor burden is higher when you are PS1 versus PS0. But interestingly, it still is an independent prognostic factor here. And when you have a big tumor burden, you have PFS and OS, which is absolutely different. And what we've shown in this paper is that every one centimeter increase in tumor burden increase the risk of death of 5%. So that's certainly a very important observation. Does it mean that there is some of ratio of tumor burden where we should start? We still don't know. But I think we need some prognostic factors to determine, I mean, when we have risk to develop new site of disease, when we have any risk of rapid progressive disease. And certainly, I mean, I would push you to go to Brian Rainey's presentation tomorrow and certainly at ASCO, because he had a prospective observation study, which is very interesting. Let's move to first choice, to first line treatment. And of course, I mean, we have three drugs which can be used in first line. And I'm not going to spend time about the best map with Sint-Affron, because if you wake up tomorrow, I mean, you will hear about my view about the best map tomorrow. But what about Sunitil and Pazopaneel? And of course, I mean, we have some comparative studies which have just been shown and you are all aware of the compare study and of the PICE study, compare showed that these two drugs are very equivalent. And I think the other ratio of 1.047 is quite convincing for me that these two drugs are equivalent in terms of PFS. And then we learned this PICE study showing that when you look at patient preference in patients who receive these two drugs one after the other one, I mean, there is a strong preference for Pazopaneel in this situation. Question is, I mean, these studies are not perfect. And you certainly have heard a lot of contradiction and discussion about these two studies and their interpretation. And question is, compare as bias? I would say yes, they have biases. Certainly they have. But still, I mean, overall survival is not biased. I mean, survival is absolutely similar in more than 1,1100 patients. So that's very convincing for me that these two drugs are very similar here. Some people would say PICE is as biased, I would say no, that's not bias. It has limitation. And of course, I mean, we compare 10 weeks of treatment and it would have been better to compare one year of treatment, but that's not feasible in fact. Of course, I mean, not all patients were eligible for primary and point. But still, I mean, these two studies show that for me, I mean, Pazopaneel is not inferior to Sunitine. And that patients do prefer Pazopaneel. So that's for me an important question. But what's going to happen with the new schedule of Sunitine, which are certainly very attractive one. And I'm certainly confident that this is going to change the scene. And that's actually an open question and I still don't know the answer to that. So next question for me about treatment of kidney cancer is when we should switch to second line. And that's a very open question and we don't have any recommendation in the guidelines. I'm sure that racist progression is not a good choice. I think when we interrupt VGF inhibition, we have tumor growth increase. And the question is, can we look at this tumor growth increase? And this is something that we did with one of my fellow, which is Charles Ferté, and we published this observation recently in European neurology, is to look at how treatment changed tumor growth. And how interruption of treatment changed tumor growth here. So the idea here was to try to look at tumor growth rate. So you look at tumor burden and then you look at change in tumor burden over time. Then you look at what happened during treatment. And then you look at what happened at the time of progression. And when you stop progression here. So of course, it's looking at change in tumor growth here. And what we observe, this is with Soraphine, but that's the same with Verolimer, that's the same with Sunitine. Is when you give treatment here, so this is placebo, this is before treatment here. On treatment, you have change in tumor growth. And then when you start to progress, you have, of course, an increase in tumor growth. But when you stop the drug, although it was progressing, you have an increase in tumor growth here. And this is very significant at the time you stop treatment here. Which means that if you interrupt VGF inhibition, then you are going to see the tumor growing faster than if you continue on tumor inhibition. So for me, the consequence of that is when you switch, you have to propose a second line, which is active enough to do better than slowing the tumor growth. And of course, I mean, you have to define some patients where you have no question. I mean, primary refractory disease, I think for me, is a good reason to change. New site of disease with progressive resist is certainly a good reason to change. Outside of that, I still don't know whether we should switch or not or continue on. So if we switch, which second line should we use? I think after cytokines, I mean, every TKI is active. Certainly, acetylib induces a longer PFS. That's certainly what we observe with the access study. But interestingly, overall survival is not better than with seraphimib. The question is, I mean, is PFS important here? Or is OS the most important in second line? I still don't know, but certainly you can use every TKI you have in your hand after cytokines here. So what about after VGF inhibition? We have in the recommendation and the guidance, I mean, two drugs which are recommended. One is everolimus. One is acetylib. And you certainly know the data from these trials here. So how could you select? Of course, I mean, the idea would be to have a direct comparison between acetylib and everolimus, and we don't. So although we should not do indirect comparison, I did it. And that's what I show on this slide, looking at acetylib PFS, OS, seraphimib PFS, OS, and everolimus PFS and OS. So this is the access trial. This is from the record one trial. So if you look at post-soonitini patient, which is the only situation where we have TKI data in this setting here, you can see that the PFS, 4.8 months, is certainly better than with seraphimib. It's probably not so much different from everolimus here. So 4.6 is a retrospective analysis on pure second line patient from the record one trial. If you look at overall survival here, you can see that's really in the same range. And even seraphimib, although it's not significant, has a higher number than all the other one. So very difficult to have a good idea here. And of course, I mean, everyone has been looking at this interesting study presented by Tom Hudson and recently published in J.C.O. looking at 10 serolimus versus seraphimib in second line patients. And of course, it's not everolimus. It's not acetylib, but it's an mTOR, and it's a TKI. Only soon it treated patients, 500 patients, so quite a large number of patients. And that's the data which comes from this study here. PFS was very similar, but OS was statistically different in favor of seraphimib. So very interesting data. Should this comparison be put exactly the same for acetylib and everolimus? I think probably not. But in other hand, I mean, we have here some data which supports that TKI after TKI is active and might be a good choice here. Also, I mean, if you look at the data, and I added in the previous table the seraphimib data from the intersector trial here, you can see that seraphimib is certainly an absolutely acceptable option in this setting here. And you can see that the survival in these two studies very similar, 16.5, 16.6 months. So it raises the question whether, I mean, acetylib, everolimus are really better than seraphimib. I still don't know. I'm still not sure of that. And that's probably an interesting observation from this trial here. So to summarize second line option, I think everolimus and acetylib remain standard of care because that phase three data. I think seraphimib has no quite interesting data in second line which supports that it might be a good option here. I think probably differences in toxicity profile are some things that you have to take into account when you choose. And I think intersex study, although it should be interpreted with some caution, supports the use of TKI after TKI more than anything else here. So to hand, I want to speak about new targets and I think the reason why we all are interested in new targets is not only because it's new, but because we have the feeling that we have reached a plateau with our drugs. We have seven drugs, but the PFS in every new studies is plateauing at between eight and 11 months. So OS, I mean, has been very much improved in the past 10 years, but now we are in the range of 26 to 39 months. So we think that we need new targets, of course. So you are going to hear in this meeting about two targets. One is PD1, PDL1 pathway. No question. I mean, the activity is promising. This is a slight problem. Most of you have seen at least once, maybe 20 times from David McDermott about activity in phase one of PD1 antibody, and this is quite interesting. You hear this long, long response and quite important tumor shrinkage here. And this is one of the patients which really impressed me because patients who had received Sunitinib, Saurafinib, Hasidak inhibitor in a phase one study. And you can see here that a very important tumor shrinkage over time and this response lasted for a very long time from what I heard here. Also, if there is some suggestions that some biomarkers will arise. So probably you will hear about this question here, but at least, I mean, we have an indication that maybe PDL1 might be one indicator. It's probably not the only one. And I've heard from many people that some PDL1 negative patients will respond to the drug. But still, there is a strong trend that PDL1 positive patients have more chance to respond to these agents here. I think we'll have new biomarker, but this kind of biomarker are going to be very important in the future. And of course, I mean, we have an ongoing phase three and the enrollment has been completed here. Nevolumab has been compared with Averolimus and we are to all of you, I mean, expecting the data from this study where overall survival is going to be the primary endpoint. Probably next year or at the end of next year, we might have the data from this study here. Second pathway, which we are all interested in is a CMET pathway. There is a strong rationale, I mean, to use CMET inhibition. And the reason is that CMET is overexpressed in many clear cells, not only papillary tumors. You will hear about papillary tumors in this meeting, but in clear cell, especially when they go over time, I mean, CMET is overexpressed. CMET expression is induced by VJF inhibition. So if you put VJF inhibition on tumor, I mean, you will induce CMET expression. So strong rationale here. So there is a promising activity of one of the CMET inhibitors we have in our hand, which is Cabosantini band. This has been, I mean, presented at ASCO by Tony Schwiery. Very impressive activity in very highly treated patients. Of course, small phase two, but good enough and convincing enough to go to a large phase three and the phase three is actually recruiting patients. And probably some of you are part of this study called the Meteor Study, which is going to look at Cabosantini compare with Vavorolimus, exactly the same inclusion criteria as in the Nevolumab study. And the primary endpoint by difference to Nevolumab will be PFS. So to promising pathway in 2014, Nevolumab and Cabosantini, the question is, will we get a winner? And the question is, maybe we'll hear the data from these two studies next year or early 2016, and we will probably have a winner, which will be the first drug registered for second line. And that's probably going to be a question here. So to conclude my lecture, I think RCC in 2014 is a lot of drugs, is a lot of questions. It's also a lot of hope. And that's certainly something very important for someone like me who has been in the field of kidney cancer for more than 20 years now. I mean, we have a lot, a lot of hope and a lot of change in kidney cancer, of course. So what will be in 20, in 10 years from now, I think you will have the answers tomorrow. I'm sure that Martin is going to give all the answers tomorrow. Thank you very much.