 All right, so we'll just pull up Danny Heng's slides there. Fantastic. All right, so I'm tasked with talking about prognostic factors in metastatic renal cell carcinoma and why are they relevant in practice. So thank you for all being here. So as you know, there are many different prognostic factors available and they can help restratify our patients X, Y and Z and their patient factors, the most important one out of all of any prognostic factor system of course is performance status and whether or not they have symptoms. There's some tumor burden factors such as prior nephrectomy, sites of metastases, bone metastases, bone metastases, sometimes a refractory, some of the treatments that we have, liver metastases, LDH, anemia, calcium, sodium. There are pro-inflammatory markers you've heard of IL-6 today, the ESR, neutrophilia is a marker of inflammation, thrombocytosis as well and of course C-reactive protein which are really prominent in Asian prognostic factors. There are also treatment related factors. What did you have for your prior therapy? Did you need radiation beforehand? What's your disease-free interval? Because if you had a long disease-free interval, there's some more indolent disease. What's your diagnosis-to-treatment interval? So similar to the disease-free interval. And of course, as you voted in the room, this is a very commonly used prognostic system profile. This is the Memorial Sloan Kettering Cancer Center prognostic profile and it uses calcium, LDH, hemoglobin, time-to-diagnosis-to-treatment, and Karnovsky performance status as the five prognostic factors. And you can see there's a favorable risk group, intermediate risk group, and poor risk group. And this was developed in the era of immunotherapy. So these median overall survivors here reflect the immunotherapy era. Sponsored by the KCA actually here. The international kidney cancer working group also developed a set of prognostic criteria. They're more complex. They use a lot more different prognostic criteria, so it's a little bit harder to use. But again, you can see that there's a favorable intermediate and poor risk population. In the dotted lines here, this was in the targeted therapy age. So the solid lines are the immunotherapy age and the dotted lines are the targeted therapy age. You can see that across all subgroups, there's a difference in overall survival. We're lucky to have the International MRCC Database Consortium. Currently, it includes 3,700 patients from 25 institutions. Back when it was six institutions big, we developed the International MRCC Database Consortium, the IMDC criteria, as I prefer to call them, the prognostic factors. And as you know, there are two clinical factors, the Karnoski performance status of less than 80%, a diagnosis to treatment interval of less than one year, and four laboratory factors, anemia, hypercalcemia, neutrophilia, and thrombocytosis. All of those are each individually, independently associated with the poor overall survival. If a patient has zero of those risk factors, they're in the favorable risk group. If they have one or two of those factors, they're in the intermediate risk group. If they have three to six of those factors, they're in the poor risk group. And this is how they stratify out. So we externally validated in a group of another 1,000 patients that we haven't studied before, and this is the external validation of all patients treated in the targeted therapy era, so not in the immunotherapy era. And you can see the favorable risk group is 43 months. In the intermediate risk group, it's 23 months. And in the poor risk group, it's eight months. We should point out that these overall survivals are definitely improved from the immunotherapy era. Before the favorable risk group was about 37 months. Intermediate risk group was around 17 months. And the eight month group, the poor risk group, was about four months. So we've definitely made strides with targeted therapy. It's a testament to the efficacy of targeted therapy. We could be a little bit more specific now about the different patient populations, because what I just showed you are prognostic factors that we could use in the first line, second line, and non-clear cell carcinomas. We've externally validated it. But now we can look at specific benchmarks as well. And although you might not use these benchmarks for everyday patient counseling, they're certainly important in clinical trial design and for statisticians to use in sample size calculations for upcoming clinical trials. So for example, you can identify progression free survivals in the IMDC of first line therapy, specific first line therapy such as intermediate and poor risk groups with a diagnosis treatment interval less than one year similar to the ADAPT inclusion criteria. So this was actually used in a sample size calculation. And you can see that the progression free survivals and overall survivals in the different clinical trials such as T01, Intersect, Gold, they're actually fairly similar to the clinical trials that were already reported. Prognosis is not static though. It's actually a dynamic process. So for example, if we have a patient sitting in front of us and we say the median overall survival that's predicted for you is 27 months or 44 months, what happens if a patient actually lives beyond that? So what happens if the patient comes back to you 36 months later or 56 months later and said you were wrong? What do we do then? So it shows us that prognosis, although we have information from the baseline it actually changes as we keep on going. So this is the concept of conditional survival. Conditional survival is how long, how does survival change for each individual patient as you survive as we gain more information about how long you survive. And specifically for those patients that actually survive past the median in their risk group how does their conditional survival change? And so this is what's shown here, Lauren Harshman from our group published this and this showed, I'll take some time to explain it in the y-axis here. This is the probability of living another two years and on the x-axis these are the months on targeted therapy that a patient already has been on. And of course the favourable risk patients do better than the intermediate risk patients who do better than the poor risk patients at the outset, at baseline. But as time passes, three months pass, six months pass on targeted therapy, 9, 12, 15, 18 months pass on targeted therapy, you can see actually the poor risk patients, they actually exceed the intermediate risk patients. And so what this means is that what we identify at baseline it can change and of course we know that intuitively but now we can show this with data. This has been seen by other groups as well and we can see your based analysis. This was previously published and you can see at baseline this is someone's overall survival with stage 4 kidney cancer. If you've already lived 12 months you can see the curve is better. If you've already survived two years, three years, four years, five years, the more years you survive obviously the better conditional survival you have in the future. So I think we've really reached the ceiling with our prognostic factors. We've had a lot of different prognostic factor models, lots of clinical variables, there's a lot of publications on this but I think we've reached a ceiling and we're in desperate need now to make it better by using biologic markers. And so I hope that there will be more data coming out looking at biomarkers but not just on their own because they're not helpful on their own. We want to know how helpful they are if you add them to existing models which is the IMDC model. So if you add biomarker X and biomarker Y to the IMDC model does it actually improve accuracy? If it doesn't improve accuracy then it actually doesn't matter this biomarker because we already have a clinical model that works. So there are some candidate examples. So this was published by the Cancer Genome Atlas Research Network and there are different mRNA expression profiles, micro RNA expression profiles, protein and DNA methylation profiles that actually show favorable, intermediate, porous groups. Of course they haven't been combined with the clinical data because the clinical data wasn't available for this study but I think that would be an important next step. So I've told you all about prognostic factors for metastatic disease. Why are they important? Who cares? Do we actually use them in clinic? And I think the second part of the title that I was given. And so of course prognostic factors, they're important for patient counseling. All patients want to know is my prognosis measured in months? Is it measured in years? And there's only so much the look test can tell us. When we see someone come in our clinic there's only so much the look test can tell us we need to be a little bit more specific and so that's why prognostic factors are helpful for patient counseling. It's also important for clinical trial risk stratification and retrospective study adjustment methods. An example here is by the Italians Roberto of Covelli and their group looked at sequencing of targeted therapy, looking at VEGF VEGF MTOR versus the VEGF MTOR VEGF strategy in a subset of patients in patients who received three lines of targeted therapy and they showed that the hazard ratio was 2.59 after testing for prognostic factors because you can imagine group A, group B in a retrospective study would be in balance so you want to try to balance those prognostic factors by adjusting them using Cox proportional hazards regression modeling using our prognostic factors. So that's an example here of using prognostic factors in a retrospective analysis. There are caveats to this data of course you can't adjust for things that you haven't collected and in this particular study there was no Exitinib and it assumes that you make three lines of therapy. Other reasons why prognostic factors are important I think it's important for planning therapy, a patient's therapy. So for example for Temserolimus we can use Temserolimus in poor risk patients Temserolimus is not the only option for poor risk patients but you should use Temserolimus in poor risk patients. So that's an example of using prognostic factors for deciding on which treatment to use. Also for prognostic factors we can decide is active surveillance an appropriate strategy. So in very very select patients where there's a very small bulk of disease you know maybe four lung metastases that cannot be resected, they're not actually growing on subsequent CT scans really quickly anyway and their favorable risk maybe it's worthwhile to do some active surveillance for a little bit and spare the patient the toxicities of targeted therapy but of course this is in a very highly selected group of patients. And finally something that's more recent that we've used our prognostic factors for is asking the question is cytoreductive nephrectomy appropriate? So I want to spend some time on cytoreductive nephrectomy. What is a cytoreductive nephrectomy? It's in the face of metastatic disease and you still have your primary intact should we take the primary kidney tumor out? We usually don't do that for lung cancer for example we don't take out the lung if you have bone and liver metastases we don't do that in colorectal cancer for example but in kidney cancer there are phase 3 trials that support the use of cytoreductive nephrectomy albeit in the immunotherapy era. So we wanted to look at this in the targeted therapy era so we looked at our database and at that time there were 3200 patients and about 80% of patients had an nephrectomy. We wanted to exclude the patients that had an nephrectomy prior to developing metastatic disease because that's not the question we're asking here. The question we're asking is for people with synchronous metastases with their primary still intact is a cytoreductive nephrectomy helpful? And so if you exclude those patients we're left with 676 patients without cytoreductive nephrectomy and 982 patients with cytoreductive nephrectomy and these are the two populations that we are comparing. And of course this is the median overall survival and there is a big difference between those 20 months versus 9 months but of course you have to stop and not really use this Kaplan-Meier curve because it's full of biases, right? You have to make sure that you adjust by all of our prognostic factors because maybe there's a reason why those cytoreductive nephrectomy patients actually got surgery. Maybe all the sick people didn't get surgery and so you can't pay too much attention to this curve. But this is the hazard ratio adjusted for IMDC criteria at 0.6 confidence circle of 0.52 and 0.69 suggesting that there is a favorable effect of cytoreductive nephrectomy in patients with synchronous metastatic disease. But do all patients benefit from a cytoreductive nephrectomy? Is there a way that we can choose who would benefit from a cytoreductive nephrectomy for now? And so this analysis was done at 4 o'clock in the morning before the UASCO deadline and this is what we looked at and so we looked at patients that had median overall survival of less than 3 months or 6 months, 9, 12, 18, 24 months and this is the difference in median overall survival for cytoreductive nephrectomy versus not and what I want to point out is actually the incremental benefit the difference between not getting nephrectomy versus getting an nephrectomy you can see if you don't live very long there actually is no difference not much difference even if your median survival is estimated to be 12 months you know your incremental benefit of 2 months that's pretty questionable whether or not you should subject someone to a full surgery and have to recover if their benefit is only about 2 months and the benefit really improves it increases as we have a longer projected median overall survival I think we all already knew this we all intuitively know not to do a cytoreductive nephrectomy in someone very very poor risk or very very poor prognosis but now I think we have data to support that and here are the adjusted hazard ratios you can see the hazard ratios touch unity over here but as we reach 18, 12, 18, 24 months 36 months it more approaches the hazard ratio of 0.6 that we see using the IMDC prognostic factors we also did this analysis looking at well can we predict how you'll benefit from a cytoreductive nephrectomy based on the number of criteria you have and so if you have no criteria actually we couldn't do that analysis because most patients with no criteria with favorable risk disease got a cytoreductive nephrectomy which reflects current day practice similarly if you had all 6 factors we couldn't do that analysis either because most patients didn't get a cytoreductive nephrectomy there are very few patients with all 6 factors so what about all those in between well if you have 1, 2 or 3 of those factors there was quite a difference in terms of median overall survival but if you have 4, 5 or 6 factors 4 or 5 factors there actually wasn't much of a difference so it might not be useful so cytoreductive nephrectomy perhaps is not appropriate for patients with a survival estimated to be less than 1 year and perhaps is not appropriate in patients with more than 4 or more adverse prognostic factors so in conclusion prognosis is important for patients counseling, study design and planning therapy prognosis is a dynamic process it's not just that baseline and finally prognosis needs to be improved with biomarkers and hopefully use these prognostic scores in your clinic thank you very much