 Good morning. It's a thank you for the invitation. It's a pleasure to be here The last time I was here I was sharing earlier was when my wife was in the ICU for seven or eight days for Streptococcal pneumonia you everybody took great care of her So I have good memories of the suburban hospital I'm going to talk about the genetics and genomics of thyroid neoplasm and Some of the changes that have occurred that I think As I understand most of your physicians in practice should know so first. I'll review the Epidemiology of thyroid cancer that has significantly changed and what that means and then go over the genetic changes that lead to thyroid cancer initiation and progression and then finish off by Discussing the application of this new knowledge Since I understand most of your clinicians, I thought I'd make my presentation sort of case-based and not too many pathways So I have Two patients. I'm going to share with you both of which to care of or taking care of One was a patient. I took care of when I was in San Francisco 55-year-old woman who Presented with a eight-year history of a neck mass As you can see her ultrasound of the thyroid is really unimpressive You've got this almost one centimeter hypocholic thyroid nodule so She got it biopsied and it was inconclusive Unfortunately for her she didn't really have Adequate or proper follow-up you could say and presented with advanced lymph node metastasis as you could see in the image down below and the patient on the Your left-hand side is a 63-year-old patient with a pretty significant family history of thyroid cancer two of her daughters and her son and Had a screening ultrasound that identified a thyroid nodule But I'm not gonna she had a biopsy of the thyroid nodule which showed atypical cells I'm gonna stop with that case because this is a patient. We just recently saw in January and Use these two cases to really emphasize How things have changed Anyways the 55-year-old woman went on to have a thyroidectomy. You see her specimen This is the internal jugular vein very bulky advanced disease Unfortunately for her because of the delay in diagnosis she ended up with stage three thyroid cancer Probably doesn't project well here But you could see the overall survival is not nearly as good as stage one disease. So perhaps a delay in diagnosis due to the limitations of the biopsy cytologic examination So for those of you not familiar with thyroid cancer It's actually one of the ten most common thyroid cancers and five fifth most common in women and It is expected that will have over 60,000 thyroid cancer diagnosis this year What has been surprising though is as you could see in the figure that I borrowed from article by McLeod in the Lancet in almost every continent that you see in Europe Italy the Asia Australia New Zealand the incidence rate has almost doubled or tripled in some countries and This may be due to incidentally detected thyroid nodules, but I'll go into that a little bit more and maybe discuss some possibilities for this Obviously the most common thyroid cancer is papillary thyroid cancer it accounts for 80 percent or more than 80 percent of cases Unless common are follicular thyroid cancer. I just have that table up There just to remind you when I discuss the genetics I'm primarily going to be focusing on the common type of thyroid cancer of follicular cell origin so the Epidemiology of thyroid nodules have also changed many patients are Diagnosed of the thyroid nodules subclinically either on screening ultrasound for Occlusive carotid disease as you can see on the top image of the ultrasound here as subset a meter thyroid nodule that ended up being thyroid cancer or With the increased use of functional imaging studies for example a PET scan Initially to stage a patient with colon cancer might detect the thyroid nodule or during follow-up The prevalence then of thyroid nodules we used to say is about three to four percent on palpation But now with the increasing use of imaging studies It's actually upwards of 50 percent for people that are 50 years of age and older and actually when you have a family history that Patient that I shared with you the 63 year old woman the rate of thyroid nodules is They present 10 years younger with 10 percent higher prevalence of thyroid nodules Why is this important? Well, the rate and increase in thyroid cancer relative to the patients that have a thyroid nodule is Much more dramatic and about one third of these patients here in the green line Are going to end up having a thyroid nodule biopsy that's inconclusive So this represents significant health care expenditure and actually accounts for nearly 150 to 175 thousand thyroidctomies being done in the US completely to exclude a thyroid cancer diagnosis So why has the incidence of thyroid cancer increased? There's a nice paper analysis of this Using the NCI sero database by Davis and Welch and they looked at exactly what types of thyroid cancer were increasing No surprise the most common type of thyroid cancer papillary thyroid cancer accounted for most of this increase What was perhaps most surprising is the fact that most of these cancers were Less than two centimeters 80 percent of the increase was accounted for cancers that were less than two centimeters So based on their analysis, they thought the increase is likely due to diagnostic scrutiny thyroid incident alomas or over diagnosis But using the same data set three years later Amy Chen and colleagues at Emory then looked at this data They show that the increase was actually more dramatic in women But there was also a small but real increase in thyroid cancer related mortality mostly in men When they looked at the age groups in which it increased as you could see those that were older than 45 were the ones That had higher even rates of thyroid cancer, which suggests that it is actually a real increase in disease burden And then they went on to look at the thyroid tumor size and found that even those that are larger than four centimeters Were also increasing so they suggested that this could not be just diagnostics scrutiny and there must be something else Well the environment has a lot of carcinogens Radiation exposure External beam that was used for therapy in the 50s and 60s and known established risk factor for thyroid cancer But we don't use it anymore so the other radiation exposure is the Chernobyl meltdown and We've learned important lessons from that. Yes, the rate of thyroid cancer increased in Eastern Europe predominantly in children and adolescents and That latency period was much shorter it used to be 20 30 years, but in those cohorts it was actually 8 to 10 years We don't know what the result of the exposure from Fukushima related to the earthquake in Japan could entail But certainly ionized radiation and fallouts have not happened worldwide where that really could account for this increased incidence of thyroid cancer we were puzzled by this and looked at the genetic landscape Or changes that are common in thyroid cancer over a 15-year period from 1991 to 2015 and we looked at BRAF mutations, RAS mutations, PTC rearrangements We'll all go into detail later on and what was surprising to us is the rates went from 68% to almost 92% of the tumors that is a patients diagnosed in 2000 to 2005 had somatic mutations specifically the rate of BRAF mutation had also increased The group at Pittsburgh also followed up this result or the study and did their own analysis and they primarily focused on a histologic subtype of papillary thyroid cancer follicular variant and they also found RAS mutations Were fairly more common in the later time period So I think this data tells us that this increase in incidence is multifactorial Certainly can't just be diagnostic scrutiny might be environmental and that there is a genetic basis For this increased incidence and that probably represents a real increase in disease burden Lastly it is the fastest growing cancer diagnosis in the US But I also want to point out although it's a small increase in mortality. That's been observed It's only one of three cancers for which overall mortality is actually increased So then the question becomes who dies of this disease? Well Thyroid cancer is really one of the few cancers that has a staging system with the age in it And as you can see from this figure Basically if you're older than 45 your risk of dying from thyroid cancer is higher Now I'm going to start sharing some genetic data with you We've been fortunate to be part of the working group of the cancer genome atlas in focused on papillary thyroid cancer and The state is not published yet But what I have is a figure of whole exome sequencing of the tumor samples And on your y-axis is the number of non silent mutations This is an over this isn't almost in 500 tumor samples. What's intriguing is Older patients with papillary thyroid cancer in their primary tumor have a greater number of non silent mutations So the higher mortality rate that we see might actually be related to this and I'll also talk about the implications of BRAF and more aggressive thyroid cancer later on So what about the genetics of thyroid cancer? This is just a very brief review of a working Multi-step progression model that most people have accepted you have a normal follicular thyroid cell here The most common type of thyroid cancer abbreviated here PTC is papillary thyroid cancer and there are three distinct Mutations that lead to thyroid cancer initiation a BRAF mutation RAS mutations, which are most commonly the K-RAS mutation and then ret PTC rearrangements These occur in about three-fourths of papillary thyroid cancer We think it's an early event because you could detect these mutations in very small occult papillary thyroid cancer The second most common type of thyroid cancer follicular thyroid cancer has a slightly different genetic profile RAS mutations have a much higher prevalence and then packs a PPR gamma rearrangement is another common Genetic alteration in this tumor. What's intriguing is though Usually there's a follicular adenoma, which is thought to be pre malignant that progresses to follicular thyroid cancer And this is based on the follicular thyroid cancer is having higher RAS mutation rates or packs PPR gamma rearrangements and then I won't cover her whole cell carcinoma, which some people think is very similar to follicular thyroid cancer So but these mutations don't only occur in thyroid cancer there's only two that are exclusively present in thyroid cancer and this is the BRAF mutation and the ret PTC Rearrangement and when I discuss molecular testing for diagnosis in Biopsy samples of thyroid nodule this will be important in looking at the data so how do these mutations the common mutation ret PTC RAS and BRAF cause thyroid cancer All of these mutation act on the map kinase pathway the matted mitogen activated protein kinase pathway So for example the red PTC is a transmembrane tyrosine kinase receptor However, when you have the rearrangement, it's constitutively activated. It does not need a ligand to bind it to signal downstream similarly, if you have a RAS mutation you also have activation of this pathway and then as I mentioned the BRAF mutation which also lead to Increased activation of this pathway all of which lead to thyroid cancer initiation and progression There have been nice transgenic mouse models demonstrating these three mutations Are involved in initiation and progression of thyroid cancer So to go back to the 63-year-old woman with a strong family history of thyroid cancer who had a biopsy Showing a typical cells I said stop because we're evaluating her here in this era Now that we have this information. So how are we going to use this information? the genetics of thyroid cancer I Have a very basic Diagram up patient with a thyroid nodule FNA stands for fine needle aspiration biopsy fortunately 80% of these biopsies are going to be interpreted benign and Defaults a negative rate for that is less than six five percent very reliable tests in that category Less than ten percent will also be malignant About a third though will be inconclusive and that's not because the cytopathologist are not good It's just the cells of a malignant tumor and a benign tumor look very much the same that you need to see Capsular invasion or angioinvasion to determine whether it's malignant or not and that's why so many people need a Diagnostic thyroidectomy the risk of cancer is anywhere from five to fifty percent really know better than a coin toss So this is a group of patients that I spend a lot of time Discussing removing half of the thyroid gland and then saying well if it ends up being cancer Then we would need to remove the other half of the thyroid gland To emphasize how a significant problem this is I just have 10 studies It's sort of a meta review of the data from 10 different institutions over 8,000 thyroid biopsy results And as you could see on the y-axis is the risk of malignancy on the x-axis is the different Categories of cytologic diagnosis by based on the thyroid nodule biopsy result Well, you see there's fairly good agreement when it's benign and when it's malignant This is across multiple institutions What is alarming though is let's say what's considered a non-diagnostic sample at some institutions Ends up being cancer only in five percent But at other institution it ends up being cancer in fifty percent of the cases and this follows for the other Cytologic groups follicular neoplasm and even when it's suspicious for cancer You could see it's not really uniform risk of malignancy anywhere from a little over 40 percent to over 80 percent So it's certainly a limitation and this is why the NCI Bethesda Classification system was established so that we are all speaking the same language when you look at these biopsy results but again you see The inconclusive categories even with this established classification if a patient has atypia their risk of cancer is anywhere from five to fifteen percent If it's a follicular neoplasm anywhere from fifteen to thirty percent and these patients need an operation Just for the sole purpose of a diagnosis So this brings up. How are we gonna? Is there any data supporting using genomics and genetics to make the diagnosis? There's been the literature has been replete with multiple studies identifying multiple markers for thyroid cancer diagnosis Fortunately, we're at the point where there's actually two commercial assays I won't necessarily talk about the commercial assays, but rather the data that Others and we have generated that have led to these commercial assays being developed and Two of them have been deemed FDA Approved we really don't know the cost benefit of using these assays and practice however yet So I went through and described the common genetic changes that occur in thyroid cancer and There have been studies including ours That have looked at a panel of these in a hundred and ten Indeterminate or inconclusive thyroid nodule biopsies. I'm just using the review article by Pashkin colleagues that looked at the four largest studies and That did the mutation testing in clinical thyroid nodule biopsy samples as you could see The specificity is fantastic 98 to 100 percent what is limiting is the Sensitivity the number of times it's positive in a patient with the disease and in our study It was as low as 38 percent and as high as an 86 percent in the Italian group study So what was important though is that you don't just do your somatic? Mutation testing for one gene, but rather do the panel of genes and this clearly shows you that your sensitivity improves in that scenario so This didn't really become or available until the study out of the University of Pittsburgh looking at this in over 900 clinical thyroid FNA samples and I'm sharing this data with you because I think this is one of the better studies that really Crystallizes the impact in patient management. So this is a table I followed from that paper and you have the Three groups that are somewhat of a clinical dilemma and people that see patients with thyroid nodule you get a biopsy result That's atypical or a US or it's a follicular neoplasm. So you can't say it's benign or malignant on cytology Or it's suspicious for cancer But if you look at using cytology only Verses if you look at these panel of mutations the RAS mutations Detected where both H-RAS K-RAS and N-RAS the risk of malignancy if the Nodule was negative for any of these mutations for each of those three categories is almost down in half So based on this primarily for those biopsy results that are atypical or Follicular neoplasm if it's positive the patient can have one definitive treatment and doesn't need Multiple operation. So if it's positive, it's great because it makes a significant impact in the Management of this patient. The problem is the sensitivity is fairly low So, how could we improve the sensitivity? Are there more mutations in thyroid cancer that are just not detected? This is a nice analysis of whole exome sequencing Data, it's in over 3000 tumor samples compared to their counterpart normal thyroid tissue in this case as well and as you could see on the y-axis is the number of Sematic mutations and then on the x-axis are the different cancers Well, you see thyroid cancer there has a fairly low mutation rate Compared to melanoma lung cancer and bladder cancers, which are much more aggressive this is again data from the cancer genome atlas project and I share this with you because the previous analysis would suggest that it's a fairly low frequency Mutations events that is in thyroid cancer But when you look at not only point mutations, but gene fusions or when you look at single copy number amplifications of genes or focal deletions in genes you could actually in Nearly 95% of thyroid cancer samples detect any one of these mutations or genetic alterations So it's not far-fetched with the advancing technology of next generation sequencing that one could develop an assay that Specifically looks at a panel of these mutations Which would allow the sensitivity of molecular testing in thyroid nodule biopsies improve tremendously The next area that has been developed into a commercial assay is genome-wide expression profiling I just have four studies up there that identified anywhere from a three gene signature to over 600 gene signature Primarily looking at the thyroid nodules that are difficult to diagnose on biopsy and This is just a heat map and as you could see you don't have to be a molecular biologist or a bioinformatician You could tell there's pretty good separation of the benign tumors from the malignant tumors so this is nearly 10 years ago and It was great promising accuracy Overall accuracy of a hundred percent greater than 90 percent sensitivity and specificity But it wasn't until this large study that first we could even find out whether Genome-wide expression analysis could be done in clinical biopsy samples This is a large study both academic and community hospitals 49 clinical centers And they looked at over 4800 thyroid nodules Unfortunately, though only a little over 400 of those were the inconclusive type So much of the data I share with you is in 265 of those patients that ended up having an operation so 85 of those patients ended up having thyroid cancer Now This was a promising assay. You look at the sensitivities 95 percent. I think or rather 92 percent Which is I think acceptable however, the specificity is low and when you look at the Categories in which this analysis would be most helpful would be the atypia of undetermined significance With a negative predictive value. That is when it says it's negative. It's negative 95 percent of the time So this ended up being an assay that tells us oh this thyroid nodule is likely to be not Cancerous doesn't really help you tell you whether it is cancerous But I think it's also informative to look at the study a little bit more closely because when you look at the benign samples It's really Almost split down the middle. It's only in 50 percent of the cases That it ended up saying that it was benign for a benign tumor So that's not really helpful and then lastly There were some false negatives so seven patients that were would have been classified as not having cancer And one of which ended up having a fairly aggressive type of thyroid cancer. So there are limitations to this assay So where are we with diagnostic markers and application of genetics and genomics? Clearly, I have a curve here. Let's say this is the benign thyroid nodules And these are the malignant the inconclusive being that sort of overlapping area. We have great Molecular testing to detect mutations and tell us if it's positive. It's cancer and then we have a reasonably good gene expression classifier analysis that tells us half of the Thyroid nodules that are not cancerous or not cancerous with that inconclusive result so What we need is additional Genomic analysis, and I think the data that's going to be Coming from the cancer genome Atlas project, which we're looking at both the methylation of genes the micro RNA both of which are Very important in gene expression regulation as well as the copy number analysis But I would like to share with you some of the work we've done We've sort of done our own integrated analysis we looked at the methylome of thyroid cancer and follicular variant of thyroid cancer, which is very particularly difficult to diagnose and Compared it to normal thyroid tissue. So some of you might not be familiar with methylation it's essentially methylation at the CPG nucleotide usually it resides in the promoter region of a gene So if that region is methylated or hyper methylated it generally silences a gene and Vice versa in some cases if the methylation is low or Hypomethylated it might turn on a gene and it's been implicated in carcinogenesis So it made great sense to look at it and we looked at Over a half almost a half a million CPG sites And this is a principal component analysis And I don't know if it projects well enough for you to see but you could see the red samples are those follicular variants that are sometimes difficult to diagnose I shared with you and The green are papillary thyroid cancers and you could see they cluster Fairly well together and in blue you have the normal samples, but some of the follicular variants are all over the place We then asked well We're interested in finding out the ones that are Mutation negative because you could test for the somatic mutation in the tumor sample or in the biopsy sample and tell if it's cancer But if it's mutation negative are you able to distinguish it or classify it as a malignant lesion? And this is a heat map that Shows a comparison between here normal and these are wild type tumors. These are tumors without B-Rap mutation RAS mutations or ret PTC rearrangements and these are ones with B-Rap mutations or RAS mutations and their methamlone or they're distinctly different from normal pretty much at similar sites would suggest that Selecting some CPG sites and doing a methylation analysis for specific genes might be helpful in distinguishing these types of tumor then the last part of my talk is going to focus on prognosis and Thyroid cancer is one of those cancers that has a fairly good prognosis. It's only 10 to 15 percent of patients that have aggressive disease and There's been almost a cottage industry of better stratifying it because you want to identify that 10 or 15 percent and Treat those patients aggressively, but yet treat the other patients not aggressively So do less invasive operation? Remove less lymph nodes not give them radio iodine if they don't need it And I just have a table here as you could see over eight different clinical and pathologic classification system all based on clinical information or available or pathology and They all work great. This is a study comparing all of these classification systems includes it in the TNM staging system which is what commonly used and They predict overall outcome terrific here is the Low-risk intermediate risk high-risk the TNM staging system the Ames classification system What's striking though is the stage one patients. That's about 80 percent of all patients They have essentially near normal life expectancy 20 years out So are we treating the 80 percent of patients like the patients with stage 2? Which has a markedly different overall prognosis the stage 3 and 4 we know are advanced disease So how could we better identify these patients that are going to have? aggressive cancers and that warrant more aggressive treatment so people have looked at the somatic mutations and Evaluated whether the presence of this mutation is associated with more aggressive thyroid cancer and this was the first study by Ming Xiao zing from Johns Hopkins that looked at this and As you could see in the disease-free survival curve here those that were had tumors that were The B graph mutants the most common type the v600e had a worse disease-free survival This study though was criticized for not being a Multivariate analysis and the follow-up was not inadequate and they also included more aggressive types of thyroid cancer That perhaps influenced this result So we looked at this question specifically in patients with just garden variety conventional papillary thyroid cancer for which most patients do well from The presence of the mutation was associated with older age Also with higher rates of lymph node metastasis and distant metastasis and not surprisingly They also had a higher stage of cancer. What was most probably compelling to us regarding this association was when you look at those patients that had stage 1 papillary thyroid cancer mind you That's the group that has a near normal life expectancy over 20 years The rate of recurrence Depending upon what system you used TNM staging or the risk stratification system the risk of recurrence was anywhere from two to three times higher and those would be raft mutations I have a question in the audience go right ahead Yeah, so well, I think the question you're asking is is thyroid cancer monoclonal or polyclonal No, so these are driver mutations Obviously in a tumor the cell population is relatively heterogeneous, but these are driver mutations So people have done Analysis and micro dissecting and even in multicentric papillary thyroid cancer The rate of positive is fairly high So you're not going to lose or it's not going to be a genetic drift so to speak in the tumor itself Unless it's acquiring an additional mutation So we believe this data speaks to you know, if it's biopsy positive, it's going to behave this way It's not going to lose the mutant B-Rap so more recently in over 1800 patients this data has been confirmed and this is a study across three continents and looking at B-Rap positive and negative thyroid cancer on the curve You see here. This is overall survival. I already told you. It's a relatively not aggressive type of thyroid cancer There is a real but small difference in overall survival in patients with B-Rap mutant tumors as compared to negative tumors and when you look at the Less aggressive type the conventional papillary thyroid cancer You also see this difference What's the problem with this? This is great It's one of the few examples that you could do a genetic test a single gene predict prognosis There are some gene express expression profiling Analysis for breast cancer. You could do a similar thing But this this is pretty remarkable for thyroid cancer because it's a low death event What the problem is is I have in a box down below It's the rates the number of thyroid cancers that are positive for this mutation It's nearly 50% So this would mean you would have to Overtreat nine out of ten patient with a B-Rap mutation to possibly Benefit that one patient that has a B-Rap mutant sample So this is really the major shortcoming of this so We have been trying to look beyond B-Rap because we recognize the limitation is That it's too prevalent of an assay and I'm going to just share with you unpublished work that we have submitted And we looked into this. This is all work done by a really bright medical student Ryan Ellis who did a medical research scholars program with us last year Why did we look at PDL one? well Lymphocytic thyroiditis is very common in patients with thyroid cancer in fact it is thought To initiate papillary thyroid cancer the link between inflammation and cancer it occurs at about 30% the specimens that you look at in a patient that has had a thyroid ectomy for papillary thyroid cancer What's intriguing though is there's also an association That is disease-free survival the patients that have lymphocytic infiltration specifically of their tumor tend to have lower risk of recurrent or persistent disease Well, we were familiar with the work in PDL one and PD one in Targeting it for immunotherapy for cancer PDL one is a ligand for PD one which is expressed in T lymphocytes and Binding of these two inhibits T cell Activation and there had been studies that show that papillary thyroid cancer actually have activated T cells invading the cancer cell So the first thing we did or Ryan did was look at our existing genome-wide expression data To look at whether PDL one or PD one expression was different and you ask why well there have been two studies in colorectal cancer and Breast cancer that showed that it might be a prognostic marker. So He did this and in this panel. You see this is a papillary thyroid cancer that was stained for the PDL one and Cancer samples were positive. There were some samples that were also negative and Then he asked whether the positivity of PDL one expression This is the messenger RNA expression of the gene was higher or lower based on whether there was Lymphocytic infiltration of the cancer as you could see on this H&E Staining of a papillary thyroid cancer and what he found was indeed the expression level for PDL one was higher in those that had lymphocytic infiltration of their papillary thyroid cancer and then he went on and a Subset of the samples to demonstrate many of these are indeed T cell positive that were PDL one positive so there had been increasing work that shows that these oncogenic mutations could also function in tumor evasion So the second thing we did was look at the relationship of PDL one expression relative to whether the Papillary thyroid cancer tumor sample was BRAF mutant versus negative And as you can see on this figure those that were BRAF positive had higher PDL one levels than the papillary thyroid cancers that were not BRAF positive and we looked for the other activating mutations the rats ret PTC rearrangement, but there was really no association The next thing to do is to really indirectly demonstrate what this is a function of BRAF activation or not that you have higher PDL one levels, so we took four thyroid cancer cell lines and Two of them the 8505 C and the SW 1736 Have BRAF mutations and then we used a compound that inhibits BRAF and as you could see You have in those that have BRAF activation good inhibition of the phosphorylation of arc Which is a downstream target compared to total arc But then as you could see in those mutant Cell lines when you use the plexicon compound to inhibit BRAF you have lower PDL one expression Then you do as compared to a vehicle and down here is just a densitometry level differences in the different mutants So we found PDL one to be up-regulated in papillary thyroid cancer Not surprisingly it was higher in tumor samples with lymphocytic infiltration And then lastly it was higher in BRAF mutant positive Samples and likely due to activation of this pathway as when you inhibit it you get lower levels of PDL one So what? the whole idea and looking at this was to look at prognosis and Given that in other types of cancers there had been a association with prognosis so in 98 patients with papillary thyroid cancer, which is a relatively small cohort But with adequate follow-up of median follow-up of five years We looked at this should be disease-free survival and this is in months and When you use the median as a cut-off This is an independent sample than the initial genome-wide expression analysis You could tell that the patients with higher PDL one level had a worse Disease-free survival as compared to those with a lower level So perhaps if you add this information to those tumor samples Specifically that are BRAF mutant positive and then stratify those that have high PDL one levels You might be able to refine the subgroup of patients that are going to have more aggressive disease This is just overall survival as you saw even in 1800 patients It's difficult to demonstrate a survival difference, but there's actually we found a statistical Trend when you looked at overall survival So we were excited with this data and we're trying to get a large enough cohort of patients that we could actually look at The presence of BRAF mutation and PDL one expression and seeing if you could better stratify really those subgroups of patients That are going to have aggressive disease So to go back to that patient The 63 year old woman with a family history Her biopsy was atypia of undetermined significance. So we did molecular testing So for us at the NIH clinical center and actually when I was in San Francisco, UCSF this had become part of a ICD or routine diagnostic test that you would request is that you could request for BRAF mutation testing in the thyroid nodule biopsy sample and Since her risk was anywhere from five to fifteen percent and she had a significant family history We wanted to treat her appropriately in the initial treatment and not do a diagnostic procedure and She ended up having a mutant positive tumor and we're planning on treating her aggressively Removing all of the thyroid gland and the lymph nodes and she probably will go on to get radio iodine therapy Just based on that information both mutation positive and the family history I'm gonna finish off Just summarizing what I shared with you, you know the genetics and genomics data That's been generated over the last 10 15 years has really gone into clinical practice I think the somatic mutation analysis for the driver mutations is a very good way if it's positive To know it's cancer the gene expression or gene profiling analysis Tells us if it's benign. So there's clearly room for improvement And I think the additional data that's being generated by the cancer genome atlas might add to the accuracy of these markers And then I think some of this information is also going to bring up Important clinical questions that we never thought of What is a thyroid nodule that's biopsy that is predicted to be benign? Natural history We're using the genetics of it to say it's benign and I shared with you at least the multi-step Hypothesis of thyroid cancer that most of us have accepted with progression from a follicular Radinoma to follicular thyroid cancer So I think that brings up an important clinical question that we really don't know the answer to Probably until 10 or 15 years down the road and then lastly I think it's not clear if it's cost-effective I understand some of the Commercial assays are in the thousands of dollars, but I think with the advances in technology and the Next generation sequencing that's probably going to be reduced where I think will deem it cost-effective as it could alter the management of just avoiding an operation in hospital stay to doing it in one setting and then it also might influence the type of adjuvant therapy that patients get for example a Patient with a BRAF mutation papillary thyroid cancer that stage two Is more likely to get radii dinoblation than not I'm just going to finish up with that and thank you for your attention and I look forward to any questions you have Are there Yeah, so the Mutation analysis really could be done could be done off of the cytology slides And in fact a couple of our patients since there wasn't an extra biopsy sample The pathologist retrieved it from the slide smear So for the mutation analysis, that's probably adequate for the gene expression classifier. There's it's actually involved You need an additional biopsy of the thyroid nodule and there's a special media that's Use at least the commercial assay that is used for that No, you cannot go to our car specimen Sensitivity and specificity Yes, so that that's a terrific question and you know, I've been involved in some of these study designs, so I think Sure, the question was what is the gold standard for these indeterminate or inconclusive biopsy results so I think The when you're establishing a new marker the gold standard has to be histology has to be the tumor being removed and the pathologist looking at it Once you have a sensitivity and specificity established based on that in a relatively large number of patients then I think You will have to accept the performance of that test in subsequent patients without having the gold standard diagnosis so both of the larger studies I shared with you did precisely that so the molecular testing for the somatic mutation did that in over 900 thyroid nodule cases The gene expression classifier did that but in a significantly less number of cases 265 so probably not as rigorous as the Somatic mutation testing Yeah That's a terrific question, and I think what the surgeons deal in practice with yes the advanced cancer You know what to deal with but most of the patients 80% of the patients are going to have Subclinical papillary thyroid cancer or a thyroid nodule. That's one centimeter. Yes so No, I happen to agree with you so I think you will be happy to know that there was a meeting two weeks ago of people across the country surgeons and There is a big push that we need to be judicious and maybe not overtreating some of these one centimeter papillary thyroid cancer So believe it or not, there's a Proposal of active surveillance in those patients. There's wonderful Japanese data showing these patients don't progress in three to four percent actually regress So so that is where I think there'd be very good application of it And then I think the second application at least surgically that I could share with you is There is a big push to start doing lymph node dissection and that's not without morbidity to the patient So my opinion is you really shouldn't do it But if you're gonna do it do it in a BRAF positive tumor not a BRAF negative So I think those are two areas that the finer molecular analysis could probably alter the management the surgical management of the these patients Yeah, you know so the data is both some people think it's a finder bias So a patient with auto mean thyroid Disease is gonna get a neck exam is gonna get an ultrasound of the thyroid you're gonna check their antibody So some people think it's a finder bias your finding papillary The other thought is though is the rate as I shared with you know, 30% is pretty high So I'm not really sure about the association in regards to the antibody titer being positive as you probably know the chronic lymphocytic Thyroiditis waxes and wanes So it might be their antibody positive And if you had the opportunity to remove their thyroid gland and look at it that they're gonna have lymphocytic infiltration They just don't have The active disease at the time you're evaluating them is the way I would explain it and then anything specific to do for those Patients for their management. I'm not really sure if there is anything specific as it relates at least to the thyroid other than you know They're obvious risk of hypothyroidism and needing replacement. I don't know if I'm answering the last question That's an interesting question. I don't know You know If it's gonna be Predisposed and is it gonna be positive? I'm not aware of any study that looked at has looked at that But I think that's a fascinating question actually to pursue. Yeah Yeah So not to my knowledge I think in the 70s Norm Thompson Wonderful underconsurgent from University of Michigan Said substernial goiters have a high rate of follicular thyroid cancer and any patient with substernial goiters should have a thyroid Epidectomy for that reason. I but I think the follow-up data I think the risk of thyroid cancer whether a topic Substernial is in a patient with a multi nodular goiter is really similar So I typically quote patients three to four percent of a dominant nodule in a multi nodular goiter Whether it's substernial or ectopic, so I'm not aware of the rate being higher But there is dated literature that suggests that it might be higher Yeah, you know, it's certainly a risk factor. I myself sometimes in the OR you're using fluoro and you forgot to wear your thyroid shield It's an interesting question I would say the first thing to do the question was you know if your Radiation exposure field should you get screening? I think a simple neck exam and a thyroid ultrasound. I Would probably appropriate. I think you're more likely to find something incidentally of no consequence Yeah, so what those would have to be sort of germline genetic syndromes. I'm not aware of any Available testing that could predict the risk of thyroid cancer using blood samples Are there genetic reasons for the gender disparity between Yes, we've The higher rates of thyroid cancer in women and then when men get it the mortality rate is higher Or it's more advanced disease in men. We've actually looked at that and at least their genomic profile of the tumor seems to be different and As you could imagine, that's a complicated study You know, what time do you obtain a tumor sample from a woman? Is it during men's sees? So overall there's a difference. So what we did to answer that question whether there's really a gender difference is we use a transgenic mouse model of thyroid cancer and removed the ovaries from the female mice and In half of the mice we did a sham surgery and then we did a similar thing in the male mice Removed the testicles did a sham surgery and what was surprising to us is by virtue of removing the ovaries mind you this is a transgenic model of follicular thyroid cancer due to a thyroid hormone receptor beta in activation and The female mice if they didn't have ovaries their rate of developing thyroid cancer was lower and in men They had less or the male mice They had less aggressive less larger thyroid cancers if their testes were removed So it suggests there's some effect of the sex hormone, but specifically what is unclear Yes, that's a common