 I'm Karen Kenoski, and welcome to this month's edition of kidney cancer news. I'm going to talk to you today a little bit about genomics, the signaling with a focus on non-torque one based on the previous presentation from Borgi that you heard, pharmacogenomics. We'll have a slide on Sunitinem and Pasoponem, and then I'm going to finish up with a case. So we heard very exciting data from the TCGA, and this was a slide that I got from Kim. And obviously we are interested in classifying runosal carcinoma, and we can classify it based on a number of different factors, whether it be at the DNA level, mRNA, microRNA, or protein. And ultimately what we are all interested or primarily interested in is isolating groups of patients that respond differently to therapy. So we are interested in being able to identify actionable classifiers, and these fall into two different categories. As you all know, they can either impact treatment or they can impact outcome. And this is what we refer to as predictive and prognostic markers. But of course I think these two are not equal, and we may be particularly interested in those that affect treatment. And in this context mutations are to me, and perhaps all of us, of most interesting that they may be predictive and open up new opportunities for therapeutic development. So obviously that is prescient, as was reviewed by Dr. Limehan yesterday, for how genes have transformed the field of runosal carcinoma therapy with the original identification of VHL mutations in monheapalinda patients and subsequent finding of VHL inactivation in probably close to 90 percent of clear cell runosal carcinomas. This is light that I got from Kim that shows multiple bars, each of them representing a clear cell runosal carcinoma. And as you can see, the mutation frequency, as she pointed out, is two mutations per megabase. And given that there are 3,000 megabases in the human genome, what we are looking at here is about 6,000 mutations per tumor. So this is not going to be an easy problem. I believe in red, at the top of each bar, you have mutations in protein coding genes, and what you are seeing, there is over 50 mutations in protein coding genes, again illustrating the complexity of the problem. This also a slide that was courtesy of Dr. Radmell, looking at the mutations that were associated with clear cell runosal carcinoma at a statistically significant p-value after a false discovery rate correction, you can see that the p-value is there. Not surprising, at the top of the list, we see VHL and PBRM1, note there were 49 and 27 mutations. This is expected, this is the expected ratio, but perhaps more concerning is that all the other genes are mutated at a rather low frequency. I believe this would correspond to a mutation frequency of about 5 to 10 percent. So the conclusion, at least my conclusion from this data, which we have also observed in our smaller analysis of genome or exome sequencing of nine tumors that we have done in the laboratory, and I believe a Japanese group has also found, is that most genes that are mutated in clear cell runosal carcinoma are mutated at a low frequency. I'm moving on to the mTORG1 pathway, and this was the focus of Dr. Gan's talk. This is obviously a pathway that my laboratory is very interesting. I'm not going to review the details, as he has already reviewed them. Suffice to say that we recently reported the transcription factor EB, which was shown yesterday by Jeff Dom to be an important transcription factor, as we all know in translocation carcinomas, is regulated by mTOR complex 1, which suggests that mTOR complex 1 inhibitors, such as the rapalox, may be a good approach for the treatment of this rare group of malignancies. We had previously shown, along with Greg Samensa and Bob Abram, that the hypoxia inducible factor is regulated by mTOR complex 1, and obviously this is the target of EHL, and he regulates this protein red 1. So as you all know, the mTOR complex 1 plays a critical role in the regulation of cell growth, and it integrates positive signals from growth factors, but it also integrates negative signals, such as from AMP kinase, which is an energy-sensing kinase. So in the situations of low energy, in a TSE1-TSE2-dependent manner, largely, mTOR complex 1 is inhibited. Similarly, in the situations of hypoxia, we have the regulation of hypoxia inducible factor alpha, increased expression of red 1, which is directly regulated by both HIF1 and HIF2, and TSE1-TSE2-dependent inhibition of mTOR complex 1. Now this poses a bit of a paradigm, or a bit of a paradox, I should say, because obviously in renal cell carcinoma, we have constitutive activation of HIF, and this would be expected to induce red 1 and down-regulate mTOR complex 1, but mTOR complex 1 is activated. We've examined these recently in the laboratory. Here we are looking at two gene expression arrays, one from the Sanger, the other one from the Dana-Farber, and as you can see, red 1 is consistently induced in clear cell renal cell carcinoma. We find these also by real-time PCR as well as by immunohistochemistry. So what is blocking this negative feedback loop? We don't know. This has led us to discover somatically acquired mutations in TSE1, which would uncouple this negative feedback loop, but this is only in a small group of patients. We're very interested in red 1 and recently reported the crystal structure, and we have also looked at a mouse with a beta-G insertion in intron 2 that is the fission of red 1 function, which has told us that in some cell types, energy signaling occurs through the LKB1 and AMB kinase pathway. Now the focus of Dr. Gan's talk was on a family of transcription factors, the foxtrotranscription factors, which he proposed serve to impose a checkpoint in tumors that have considerably active mTOR complex 1. So he showed data of how mutations in TSE1 result in the development of cysts. This is not something that we understand, but as you saw before from Dr. Linihan's talk, it's also observed in BHD deficient kidneys. However, in the context of TSE1 heterocyclicity, and this is borrowed from him, you can see that the loss of the foxtrotranscription factors results in a substantial increase in the development of renal carcinomas by comparison to TSE1 heterocyclic kidneys. Of course, the challenge here is that TSE1 is not lost except for a small percentage of renal carcinomas in patients. So one way to tackle this problem of FOXO would be by targeting simultaneously both mTOR complex 2 and mTOR complex 1 using catalytic inhibitors. As you all know, the RAPA logs are allosteric inhibitors. They bind to the FRB domain. The mTOR catalytic inhibitors bind to the kinase domain and would simultaneously block mTOR complex 2, which is important for AKT activation, which in turn phosphorylates FOXO resulting in 1433 binding and sequestration in the cyroplasm and would also inactivate mTOR complex 1. And obviously there is a lot of interest in this class of drugs. So I'm going to discuss very briefly what are the implications for patients of the biology and how we may use some of this information even if we don't have level 1 or level 2 or level 3 evidence. So this is a patient with tuberculosis complex who presented with a recurrent epitheliorangio angiomyelipoma. This is not your run-of-the-mill angiomyelipoma. This is actually a very aggressive tumor that metastasizes. And this was in a 24-year-old gentleman. He presented with a hemoglobin of 3.8. As you can see, very poor performance status. This was a recurrent tumor and embolization fail. He was requiring a lot of units of pag red blood cells. And the tumor was deemed to be inoperable. And this is by our surgeons who are very good. In particular, this patient was being taken care of by Dr. Raj, who is actually rather aggressive. We don't have established therapies. And what the team recommended for this patient was hospice. However, and this is just to show the histology of the epitheliorangio myelipoma, because the patient had tuberculosis complex and had a germline mutation in T.C.1 or T.C.2. We perceived that it would be helpful to consider serolimus. That T.C. complex is a proximal inhibitor of mTOR complex 1. And this is what we did. We recommended, I recommended, that the patient be treated with serolimus empirically. This is what the hemoglobin looks like when the patient presented with hemoglobin of less than 4. These are the transfusions that he was getting. And this is what happened after we started the mTOR complex 1 inhibitor. You can see the patient had one transfusion, the hemoglobin stabilized, and this is what the scans look like. So this is from February to May. It is, of course, an anecdote, but it is, of course, a very meaningful anecdote for the particular patient who was going to be sent to hospice. So I like to move now on to pharmacogenomics. And I'm just going to show you a couple of slides, one slide, which is adapted from Dr. van der Veld. Again, looking at sunetin of metabolism and illustrating some of the genes that they have looked at. So here you have in red genes in whom polymorphisms have been associated with toxicity and in green genes in whom polymorphisms have been associated with efficacy. And what is not worthy, and I would like for her to comment during the Q&A period, is that not all of the genes are associated with both, as would be expected, for changes polymorphisms that affect sunetin of metabolism. This is a slide that I believe it's been shown perhaps twice, once by Danny, the other one and the other time by Brian. This is from the actual publication in the GCO, and that came out not that long ago. Again, showing that there are polymorphisms on passuponib. In this case, we are focusing on pharmacodynamic polymorphisms in IL-8, which been others have shown it's important for the angiogenesis scape in patients treated with sunetinib as well as H1-alpha. I should point out, as Brian also mentioned, that these p-values, even though statistically significant, they are not corrected for multiple comparisons, and in fact, after corrections, they are no longer statistically significant. So there are a number of questions about polymorphisms. The first one, should these SNPs be incorporated clinically? In my view, it's still premature. They require perspective validation and evaluation after a false discovery of rate correction. Is there a class effect? Obviously, in as much as there is shared metabolism and mechanism of action, one would expect a class effect. Would outcomes be improved by increasing drug exposure? That's really the key question that comes from this analysis. Now, my time is up. So what I'm going to do is I'm going to ask for a show of hands, even though I cannot say anything, and if it would be of interest, I can tell you about a case report. This is a patient for whom we did whole genome sequencing, and what we found and how we tried to act on that information. I think this is something that we are going to encounter clinically, or we can move to the Q&A. So whoever, those of you who are interested in hearing this case presentation, please raise your hand, and if you are not interested, we'll move on to the Q&A. Okay, all right. So it sounds like there is interest, so I will move on to the case. So this is a 67-year-old patient of mine who presented with painless hematuria. He had a CT scan that showed a mass embedded in the right kidney, standing into the renal vein. The metastatic workup was negative. He had a radical nephrectomy, which showed a 3.5-centimeter tumor, Forman grade 3, standing into the lumen of the renal vein. We have a program in the laboratory to systematically implant tumor from patients into mice. We implant the tumor orthotopically. These tumors, microscopically, look like Cleosalrinazole carcinoma, and microscopically also preserve the histology in the architecture. Today we've implanted over 100 tumors, and it's very exciting that it's not only the histology, but also the gene expression, molecular genetics, and, most importantly, treatment responsiveness that is preserved in these models. For this particular patient, he was followed with imaging studies. And 15 months later, routine imaging studies shown extensive adenopathy as well as a 4-centimeter Cetabular metastasis. The patient was not in a poor prognostic group. I can't remember whether it was in a good or an intermediate group. And soon it was begun. At that time, DNA from the primary tumor, as well as from peripheral blood monocular cells, was sent out for genome sequencing. These are the findings, which we reported in the Personal Genomes Conference at Colespring Harbor last year. We found approximately 6,500 mutations. There were 59 mutations in protein-coding genes. Some of these were synonymous, but there were also mutations in supply sites. Interestingly, the technologies got in so good that we were, in fact, able to validate by Sanger sequencing every single mutation that had been identified by Illumina. These are just the percentages of the mutant allele compared to the wild-type allele. This is not something I'm going to go into, but suffice to say that every mutation was validated. Interestingly, we found a mutation in TSE-1. A previous study had failed to identify somatically acquired mutations in Ruin and Salcarcinoma. We did find one, and we also found a mutation in ERB-V4. This data was integrated with copy number analysis, both pair copy numbers, as well as allele-specific copy numbers. And the TSE-1 gene, which is in the long arm of chromosome 9, you can see that there is a deletion here, as you can see here. So the wild-type copy had been lost. This was a point mutation affecting a supply site. And it resulted in an in-frame deletion of exon-5. So exon-5 was lost from the transcript, but these did not alter the frame of the protein. The mutation was accompanied by loss of the wild-type allele. We did reconstitution experiments in TSE-1 deficient cells and showed that the mutation was a loss of function and that they stabilized TSE-1. And not surprisingly, we found low TSE-1 levels in the tumor inactivation of mTOR complex-1. So the patient started on sunetinib and also on solyndronic acid. After two cycles, he had improvement in the lymphadenopathy, but progression of the acetabular metastasis. He had a resection, and this was stabilized. And then at that point, based on Bob Mozart's data, he was switched over to everalamus. Interestingly, on everalamus in the second line, he had a stable disease for 12 months. So this contrast to the progression after two cycles of sunetinib. And obviously, this led us to hypothesize that maybe TSE-1 mutations, as would be supported by the biology, might be a predictor or a responsiveness to mTOR-1 inhibitors clinically. How often are these mutations seen? We've done sequencing in 77 tumors, and we only found three additional mutations. By contrast, there was just one mutation in P10. So the patient, at that point, they started posopenib and progressed in three months and then started surafenib and progressed after two months. And I think this is what all of us encounter clinically, right? So what happens is we put these patients on the different agents, and they keep progressing, and then it gets to a point to, well, what do we do next? What options do we have? Themsyrolimus, but as you know, themsyrolimus and everalamus are both RAPLOX, so it's unlikely that this is going to give additive effect. And Bevacusumab interferon, which at this stage where the patient has progressed on three-vegeta receptor, two inhibitors, seems to me would have a little hope. Now, the patient did have a mutation on ERVB4, which, as you know, is a member of the HER2 and EGFR family. We did in silico-structural studies to ascertain whether the mutation was compatible with a gain-of-function mutation. We did in vitro assays, which unfortunately were inconclusive, but as you may know, ERVB4 mutations have been found to occur at high-frequency metastatic melanoma. So there is, in fact, a phase II clinical trial looking at LAPATINEP for the treatment of patients with melanoma in a documented ERVB4 mutation. So at that point, we decided we would try LAPATINEP. This is obviously off-label. Unfortunately, this did not work. The patient progressed when we did imaging studies. And obviously, this raises the question, was progression due to a coupling of MTOR complex one as a consequence of the TSE2 mutation? And should we have considered LAPATINEP in combination with an MTOR1 inhibitor, and there are such clinical trials? For other disease types. So at this point, we decided to switch over to epilimum map, which I think we will be hearing more about, and I think it's an interesting drug. And since it's available for melanoma, we were able to get it for our patients. So I'd like to finish, first of all, by thanking the speakers for their great presentations, and also thanking the people in my laboratory, as well as our colleagues at UT Southwestern and my funding sources. Thank you for your attention. Okay, I believe we have time for some questions. So I would invite any questions to the panel. Walter, can we get the mic? So, I guess it's on. One of the challenges here is just the sheer mass of data that's being generated, and I'm struggling with this a little bit. And let's just take, for example, to simplify the question, the pharmacogenomic data that's currently available with each one of these drugs. We simply cannot do a large prospective phase three study for each one of these, or even for a collection of the SNPs in a large population. I mean, we're never gonna finish these kind of trials. So how do we move forward? How do we take what is potentially interesting data and move that forward to clinical application? It's something that we've been struggling with. I don't expect an answer, but maybe some thoughts here from the panel. Excellent question, Brian. Yeah, I mean, we're obviously all struggling with the same question. And as I mentioned, it's non-overlapping candidate lists. And in the candidate lists are just our hypotheses about what's important, right? There's not necessarily, we're not necessarily including all the SNPs. So I think, I think with the work that's been done so far, there are some areas of overlap. As I mentioned, there's an effort from the major international groups to combine data and samples. Again, still retrospective, but hopefully can solidify those clues and tease out things that are just false discovery, et cetera. And then I think a much more discreet list can be tested in a prospective trial. So I don't know another way to do it. There are some areas of overlap that I think are promising and that make biologic sense, especially with metabolism genes and linking that to PK. So I think we can move it forward in that way. Astrid, would you like to comment? I agree with Brian that we have to combine our databases and first validate it retrospectively. Hopefully we can define a gene of polymorphism that are interesting for prospective studies. Anybody else would like to comment? And just a very quick comment. I mean, I think that we are in a big discovery phase and these things come cyclically. I think that things will come together and it won't be all on our disease group either. I think things that we discover from the other groups will eventually merge together so that it is more comprehensible. The only thing I would add is that it seems to me, as Brian alluded to in his talk, that it would be important to define all the polymorphisms that we think are interested in systematically look at them for all the endogenesis inhibitors at once. So I think that would be important. Okay. Hi, good morning. My question is a corollary. I think what I had to ask is for the TCGA data that has been collected, is there treatment information and accordingly, are we gonna be able to get predictive information also out of it, not just prognostic? Right, there is treatment information. These are not clinical trial patients. So these are patients that are treated with whatever standard of care was available at the time and not with the kind of rigor of a clinical trial. But those are questions that we can at least generate hypotheses with this data. Yeah, because it would be important for the extent of work and the large sample size. Absolutely. The first paper will only be descriptive, but there will be a lot of data. It'll be a tremendous resource for those kind of questions. Eric? Fantastic session. Really, really interesting. It's exciting to see how things are progressing. Question for Dr. Gann. You know, what's really interesting in the VHL patients is that we also see cysts and solid lesions in the kidney, and the question really is what's the interplay between those? So my question for you is, in your mice, are the solid tumors arising from parenchyma or are they arising from cysts? And the second question is, what's the status of GSK3 beta? Okay, so for the first question is the cyst. Yeah, so in the TSC1, we're still not so sure. Yeah, the renal tumor, whether it's developed from cysts or not, so that's some question we're trying to address. I don't know whether Jim have any thoughts on that. You probably also have the mice, so the renal tumor developing those TSC no-com mice. Yeah, they tend to arise in the context of cysts. There is cyst, yeah, but we're trying to address whether those renal tumors developed from the cyst or not. The second question is about the GSK. So when we look at the, so that's an interesting question. So first, in the TSC no-com mice, I think that was already been published by Brand Mining at Harvard Medical School. So typically, this feedback, if you consider feedback, those AKT downstream substrates would be, so it would be considering with the FOXO, the phosphination would be decreased in the TSC no-com mice. So this is only, well, only talk about the mouse and branding fiber balance. But interestingly, the GSK phosphination is upregulated, so it's opposite as what we expected, because GSK is also AKT substrates. So they identified actually in the TSC no-com setting, the S6 kinase can actually phosphornate GSK. So the AKT is done, but S6 kinase up and it takes the place of the AKT, so it phosphornates the GSK. So in the GSK, phosphination actually is up in the TSC no-com setting. So that's the opposite as what we expect. Join us again next month for another edition of KB Cancer News. I'm Keri Konoski, wishing you good health.