 So it's a pleasure to be here today to talk about work from the Stanford Kidney Cancer Research Group and I'm an instructor in medical oncology So we're all in this room because we need answers at every step of cancer care for kidney cancer We know there's something that happens a tumor initiation and very often that's loss of VHL Which leads to dysregulation of the hypoxia signaling pathway and increase in HIF VEGF and angiogenesis Then the tumor cells grow large enough so that they're diagnosed and usually that's incidentally nowadays on a CT scan Done for some other reason And then they go to surgery if they've localized disease But for the unfortunate patients who recur or those that are metastatic at Presentation we have to select our first treatment and when we do that They're now luckily seven approved drugs that target hypoxia and m-torchinus pathways and of the five that are first line We don't really know which one to give first and we choose it based on our best Gas and then something happens to the patient cells. They may respond. We may be turning down the pathway We're not really sure until the first imaging at eight to twelve weeks after treatment And sadly up to 25 percent receive no benefit from their treatment and they've had side effects without response So what we need are proteomic biomarkers to quantify targets and downstream Downstream signaling inhibition upon treatment initiations that we can try and find out much sooner rather than waiting eight to twelve weeks What's happening at the proteomic level for these patients and these patients are? receiving therapies as you can see on the figure that Either are inhibiting receptor kinases or small molecules that inhibit the kinases up and down the pathways And you can see some of the key players This is a highly simplified version of the pathways involved But you can see that we need to be able to measure downstream changes in PI3 kinase AKT m-tor as well as Ras-Raf through mech and urk To understand if the therapies are actually hitting their targets are not earlier Now there's several barriers that we're talking around the lunch table even that have inhibited the development of proteomic biomarkers First it requires invasive procedures to provide enough tissue for analysis the gold standard for Western is a Western blot Which usually requires about a million cells for one measurement And that's just not tenable to take a patient through multiple surgeries just to see what's happening with their tissue Things like that are standard in clinical practice like immunohistochemistry and flow cytometry to look at proteins require many processing and staining steps and each step can be due to Human error can be less and less precise and so you get a qualitative measure Not a highly quantitative measure of the proteins and lastly we're talking about looking at activation of proteins And we have to measure the phosphorylations and complex phosphorylations are very difficult to measure unless you have phospho specific antibodies And we don't have phospho specific antibodies for every single phosphorylation So to answer these questions, we've developed the use of a new nano amino assay, which I'll describe in the next slide It's abbreviated NIA to distinguish and quantify multiple proteins and their modifications in patient cells The NIA allows rapid analysis of very small amounts of tissue such as fine needle aspirates or cells from blood and it's capillary based microfluidic instrument that's commercially available called the nanopro 1000 and in this instrument the proteins are separated by charge So a very small amount of charge difference like a single phosphorylation Will focus in a different part in this micro capillary and what makes it nano is that the input requires only four nano liters of Wysate so you can take a fine needle aspirate flash freeze it minimizes processing steps and then process it for analysis So it decreases processing steps and it decreases human error Once we have the proteins lined up based on charge you use antibody detection And what you get is like on the bottom a wave form where the area under each curve is the amount That's present of phosphorylated and unphosphorylated isoforms So we get very high resolution and for the first time we can see date changes and even mono phosphorylated isoforms That we couldn't even see before because we don't have antibodies to them Luckily, it's automated. So it's near real time. We get a result in four hours So this is just a partial list of some of the assays that we've been developing and particularly for hypoxia signaling We've been looking at things like CA9, HIF, VEGFR as well as up and down the MAP kinase and PI3 kinase pathways And we've shown that changes in specific phospho isoforms occur with successful targeted treatment in two clinical contexts so far First we published in 2009 that a monophosphorylated IRC2 isoform predicts responsive patients with chronic myelogenous leukemia to a matinib and a second one is a was with a clinical trial of regercitib, which is a PI3 kinase PI3 kinase inhibitor and a polo like one kinase inhibitor and that's still in clinical trials We found in patients with myelitis plastic syndrome that a very specific phospho-IRC isoform Decreased in patients that had clinical benefit So now we're using NIA to profile kidney cancer and this is a photograph of one of the samples that we got So we get the kidney from the OR and we bivalve it and then we can directly visualize areas of normal kidney tissue like the blue dot and areas of hypoxia and areas that are very good Specimen and we take samples of all of them so we can start addressing what's different from the normal tissue and And how the tumor may have intra-tumoral Heterogeneity so the goal is to measure activation of hypoxia pathways to determine the presence of RCC and Characterize the degree to which different pathways are activated at the proteomic level and these ex vivo FNAs are obtained under direct visualization so this is just looking at our monthly case accrual from 2012 to 2013 and we've Accumulated samples from greater than a hundred patients to date and that's more than 200 FNAs several FNAs per patient And we can see that there was a little drop-off into in the summer of 2012 because we are transitioning from pilot funding to an R21 Which allowed us to hire our study coordinator Tommy Metzner who's in the audience and we can see that our sample collection has accelerated Now that very important questions are how reproducible is this well? We found that there are a hundred cells are adequate for analysis and it's highly reproducible and We can keep the specimens on ice for up to 60 minutes before we get them to the lab and pellet them So this is an example of looking at mech one phosphorylated and unphosphorylated isoforms, and they don't change over time We can start to use this in Overnight profiling. So this is a sample of 20 20 FNAs from RCC specimens and by the next morning when we come back to the instrument We've profiled eight different proteins and you can see that there are very different basal levels of activation of these proteins In fact if we look across 200 specimens for Erk one, for example We can see that I had to graph this on a log scale because basal set points for each tumor and each pathway can be very different So it's hard to choose this one set point. That's a threshold above which or below which they're going to respond to a treatment So this I think is some of the difficulties that we have in trying to develop proteomic biomarkers from one time point at surgery so what we find is that a paired analysis can be more effective and Profiling shows that RCC is highly phosphorylated compared with adjacent non tumor tissue So you can see a trace in the upper left corner of an FNA from kidney cancer It's a clear cell and the different phosphorysoforms are indicated with the peas We look at it profile the adjacent non-tumor. It's pretty clear that the phosphorylations are not there And if we look across panels of head and neck tumors kidney tumors adenos and Scar in sarcomas we can see that for the kidney tumors if you take the ratio of tumor phosphorylation to normal It's always high, but what you also notice is it's not always the same fold difference. It's a very different fold difference So FNAs of tumor tissue versus adjacent non tumor tissue are one way we're trying to look at the comparison of activation Now we have work in progress as well The goal is to define nanoproteomic signatures to monitor and predict therapeutic response to approved and novel agents And our approach as I mentioned is to profile cells from RCC patients before and much earlier during treatment than the 8 to 12 Timepoint and these are patients treated with TKI's mTOR inhibitors or building this into new clinical trials of novel agents And it would be ideal to use minimally invasive techniques to do this And so we've received an R21 to be able to do this with percutaneous biopsy as well as serial blood Specimens where we can analyze cells from blood and we've accumulated blood on greater than 60 patients to date So in summary what I've said is that we can use nanoscale technologies to measure the hypoxia signaling pathway in several Numbers of in small numbers of tumor cells from kidney cancer kidney cancer patients and other solid tumors as well as hematologic tumors And now we're poised to interrogate molecular mechanisms at the proteomic level of tumor initiation and tumor sensitivity and tumor resistance to therapy And The ultimate goals are to accelerate development of new drugs by selective patients likely to benefit We want to get a much earlier sense and not have to wait 8 to 12 weeks before we know if they're receiving benefit and to Define biomarkers of response earlier and ultimately to personalize therapy for patients whether it's in the neoadjuvant setting So we can be courageous and say we're going to put you on this because we can tell it's working as early as one week or In the metastatic setting as well So I'd like to thank all of our collaborators It takes a very large team of people to be able to do this difficult proteomic work Especially my mentor Dean Felscher as well as the kidney cancer research group at Stanford Which is spearheaded by John Leppert also in the podium and Sandy Sturtivost and Lauren Harshman Thank you