Chapters Transcript Update on the Management of Thyroid Nodules Course: Updates in Endocrinology 2024 By way of disclosures, I have no financially, um, I am gonna talk about the American Thyroid Association treatment guidelines, and I do wanna say that we reserve the right to change our minds. These are, in fact, just a draft. Um, so, um, so if things look a little different on the final paper, don't, don't be shocked, but this is kind of where we're, we're thinking we're gonna head and we hope sincerely that this will, um, come to a close soon and that you'll get these guidelines in your hands this year. So I wanna talk about uh thyroid nodule evaluation from a holistic standpoint. I wanna talk about the sonographic risk stratification and the cytologic evaluation of nodules and how these various features that we think about nodules and silos interplay with each other to modify the risk of malignancy in a given nodule and we're gonna talk a little bit about molecular testing and patient demographics and how those impact um outcomes. And interwoven in there, I'm gonna have a few cases, and we're gonna talk about the nodule guidelines and how it uh comes into play with these various factors. So there are a number of sonographic risk stratification systems available, um, so the ACRIRADs, the ATA system, we have KIRADs, and then EU tires, and I'm often asked which, which one do you use? Which one should I use? And this is a survey of five societies, um, asking providers which Uh, which of the sonographic risk stratification systems that they use and what they found is that it was largely dependent upon two factors, what your job title was and where you lived. So radiologists were more inclined to use ACR tires, whereas endocrinologists were a little more flexible in what they used, but it was largely determined by geographic factors. Ultimately, when we look at these various risk stratification systems, they really fall into two major categories. One is pattern-based recognition and the other is uh points-based. So this is ACR tires, we know this is a points-based system. Um, most of the systems in use today look at pattern-based recognition, and the reason for that. Is largely because it's much quicker for us to see the pattern and quickly assign a risk to it, whereas with these point-based systems, it's a little more cumbersome to add up the individual points and then come up with a score. So, so we as a, as a clinician, I like something that I can rapidly look at it and say I know what this is, um. The downside of the pattern-based system though, is that not all permutations of each nodule can be captured with these pattern-based systems. So the ACR tires, in contrast, you can capture every single nodule with this risk stratification system. And when people ask me which system do I use, I say I use both. I use both the ATA and the ACR tires, and it really just depends on what is most convenient for me in that moment. And whether or not I need to decide to poke that nodule or not, and I'm gonna get more into that as we go forward, um, but many of us will use multiple, uh, risk stratification systems in our clinical practice. So if we look at, you know, we spent a lot of time parsing out which of the system performs better in terms of sensitivity, specificity. Um, and, and what I wanna just say is that these systems largely overlap very nicely, especially at the low end of risk and the high end of risk and nodule stratification. So those, they tend to agree, yeah, these benign nodules, these simple cysts or these really ugly nodules, we do a pretty good job and they all kind of overlap in terms of their description of these various types of nodules. And the real difference between these systems, though, is what is the threshold for FNA. And so that's where we get a lot of our information about how the diagnostic performance of these individual sonographic risk stratification systems perform. And so to demonstrate that, this particular study looked at the ATA and the ACRIAD systems, and we know with the ATA system, it's associated with a high sensitivity for detecting malignancy, but a lower specificity. And in contrast, the the ACR system is associated with a lower sensitivity, but a higher specificity. So if we take a nodule, And we classify it using the ATA system, but we apply the size thresholds for ACR to that nodule. What we see is that the diagnostic performance of the ATA system mirrors that of ACR tires, meaning we, the real thing that depicts how these systems perform is where is the size threshold for FNA. So if we modify those size thresholds, then we can alter the diagnostic performance of these um systems. Now the old ATA system, which we've, I, I've been using since it came out, has been really um uh not criticized, but folks have really kind of been concerned about the fact that a number of nodules are not are not adequately classified with the ATA system from 2015. Uh, so the nodules that largely fall into this category are isoechoic nodules with any suspicious feature, especially the punctate ecogenic fossa, and then we didn't do a very good job of describing uh the risk associated with cystic cystic nodules and in particular cystic papillary thyroid cancers. And then finally, anything that had a a coarse calcification other than an uninterrupted rim calcification was really just not captured with the old ATA system. But ultimately, the question is how often is that really an issue? And um studies vary in how, how often that you see a non-classifiable nodule, uh, but a pooled meta-analysis found that it was just under 8% of nodules are considered to be non-classifiable with the ATA system. So then the next question then becomes, well, what is the likelihood of cancer in those non-classifiable modules? And again, you can see some variability, but generally speaking, that risk of malignancy is somewhere between 10 and 20%. And again, in a pooled analysis, the, the risk of malignancy was a little bit on the higher end of that range, and that probably had to do with the fact that there was an increased um inclusion of studies that had surgically managed nodules as opposed to just all comers. So you're gonna have a little bit of a potential for bias there toward that higher end of the risk spectrum because maybe the more suspicious looking nodules were sent for surgery. So those are kind of falling into that category of what we used to call sort of the intermediate risk of malignancy nodules with the ATA system. If you look at that, that range of malignancy. So as we went forward with creating a new sonographic risk stratification systems, there was a number of goals that we wanted to achieve. So of course we wanted to be able to qualify and classify all modules, um, and in a way that made it not cumbersome for folks to remember. We still want to be able to achieve that quick recognition factor that we have with the pattern-based system, but we wanted to be able to include all these nodules that weren't previously categorized. Um, specifically, we wanted to be able to better classify nodules that had cystic content within them, and then we wanted to um include sort of an updated analysis of the existing literature that had come out since the last iteration of the guidelines so that we had a better appreciation of those individual features and where that put the nodule within the risk of malignancy continuum. And then the final thing you can see on the table there is the 2015 risk of malignancy assignment for nodules had big gaps in it. Uh, and so we wanted to eliminate those gaps so that there was more of a continuum across the spectrum of nodule appearance, um, that, that was without those gaps. So the, uh, low risk of malignancy and intermediate risk of malignancy then had a higher or wider range of malignancy associated with it. And then those high risk nodules, um, have a greater than 50% risk of malignancy instead of 70% to 90%. So the way that we approach this is we're taking an algorithmic approach. So the thinking about the highly suspicious features, and this is one of the concerns that we had with applying individual features an assigned point with the point-based system, is that punctate eogenic foi or taller than wide shape may mean something in one nodule, but not another type of nodule. So thinking about those highly suspicious features like taller than wide, um, uh, punctate ecogenic foci, applying those specifically to solid nodules because when it's in a solid or a hypoechoic nodule, those features have a higher risk of malignancy than when we see them in cystic nodules. So applying those specifically to the type of nodule that the, the patient has. And then limiting certain characteristics. The big one that we see with over calling with ACR tirades is the spongiform nodule. I'm sure we've all seen someone come into our office with a tirades 5 nodule and we look at the ultrasound images and it's a spongiformum nodule. So we wanted to try to mitigate some of that um concern that we are seeing in our clinical practice. And we wanted to continue to have this photo album or the Atlas, so that rapid recognition of nodules um that will allow us to rapidly use this within our clinical practice, um, without, um, not without a uh uh ignoring certain types of nodules but not making it so cumbersome that you can't use it. And then we took out those um nodules that were associated with very highly suspicious features. So if a uh a limb, a nodule regardless of its size has a highly suspicious lymph node, that's taken out of the atlas and we know what we need to do with those, um, so those kind of falsely drive up the rate of malignancy in that high risk group of nodules on the top tier. Um, so we remove those from the algorithm. We don't have a size threshold for those. We know those are cancer and we know those need to be treated appropriately. Um, and then if we see clear evidence of extra thyroidal extension, that should really, uh, also increase our level of concern for that nodule. So this is one way that we are looking at this algorithmic approach. We're in the process of making a simplified version of this, um, but we're going back and forth with the publishers on how to make that look. Um, but for now, this is what we're using. So first thing that we're looking at is the composition. So at the top of the um of the algorithm, so is the nodule solid, greater than 90% solid, or is it greater than 10% cystic? And then we cannot assess the composition because of a coarse calcification obscuring the ecogenicity of the nodule. So then within the solid group, it's broken down into markedly hypoechoic nodules, so those that are as dark as or darker than the surrounding musculature, um, those that we are calling non-markedly hypoechoic, and we know that that's a bit of a mouthful, but um that's, that's how it's gonna be called, um, so those are hypoechoic, but they're not quite as dark as the others. And then isoechoic or hyperechoic nodules. And then from there it's just the presence of any suspicious features whether um it's 1 or more or 2 or more um and then down from there based on the various ecogenicities of the nodule. So more of an algorithmic approach, um, and we will still have an atlas for this. And then when we look at the cystic nodules, they're largely broken into two groups. So, um, with partially cystic nodules that are not spongiform, they're either high risk or low risk. And so the high-risk nodules are those classic papillary thyroid cancers that are cystic. So these have um uh a jagged interface with the of the solid component with the cystic area or an uh an acute angle interface with uh either a coarse or a punctate calcification within. And then everything else falls under the low suspicion group and this again has a risk of malignancy of 3 to 20%. We did classify irre the uh spongiform nodules with a coarse calcification as being, we elevated it from very low risk to low risk. The data in this area is very um is, is of, of, it's of low quality, um, but we didn't want to ignore the, the potential for an increased risk until further data can be. Uh, captured in this group of nodules. So for now, if they have an irregular margin, an irregular halo, or macrocalcifications within a spongiform nodule, that elevates it slightly uh in risk. And then, um, there's no change to the pure cysts, we consider those to be benign. Uh, and then those nodules which have a course calcification that precludes, uh, classification, uh, what we're really looking at is whether or not there's an interruption in that calcification. And if there is soft tissue extruding through that, then that classifies it as a higher risk. If it's irregular with an interrupted calcification, but no soft tissue extrusion, then that's intermediate and then The low suspicion are those eggshell calcifications that are uninterrupted. So if we look at these individually with the solid component, again, the high suspicion, these are markedly hypoechoic with one suspicious feature or uh so over here, or non-markedly hypoechoic with two or more suspicious features. So in this example it has irregular margins and um punctate ecogenic foci. And then markedly hypoechoic nodules like this example with no suspicious features have been um classified as intermediate suspicion. If it's hypoechoic with one or more suspicious features like this example, then that becomes um an, an intermediate suspicion module, um, previously non-classifiable. um, and then in this example, this is a hypoechoic nodule with one suspicious feature with an irregular uh lobuated margin. This is not an acagenic focus. This is just posterior enhancement with a um posterior to a cystic space, so it's not a, it's not a microcalcification. And then the low suspicion nodules are any uh solid nodule that lacks suspicious features and is not markedly hypochoic. And then the partially cystic nodules are those that are greater than 10% cystic. And again, this classic cystic PTC picture here, uh, where we have a solid component that has suspicious features within it, the punctate ecogenic foi, um, or, uh, macrocalcifications, and then the, the acute angle interface here between the solid and the cystic. Um, you can see that again here. Uh, and then this one has sort of that lobuated appearance with the punctate ecogenic foci. And then the low suspicion we recognize those um are gonna be a a bit of a broad range of malignancy risk, 3 to 20%, and there is some gradation in there. We didn't for simplicity's sake, want to include that in the algorithm for folks that have more interest in that, they can refer to the text to see what individual features will push a nodule toward that higher end of the spectrum versus the lower end of the spectrum. And then the spongiform nodules that are low have been elevated to a low suspicion, and these are the macro calcifications or an irregular uh margin or an irregular halo, and then the classic spongiform nodules still remain very low risk. And some examples of the inability to determine the ecogenicity of the nodule due to the calcifications. You can see the soft tissue extrusion here. This nodule has a very irregular margin, but no soft tissue poking through, and then this is our classic eggshell calcification. So if we over, if we superimpose this on the old system, where are the changes? So, the first change to the high risk is the addition of that cystic, that classic cystic papillary thyroid cancer. Uh, we downgraded the hypoechoic solid nodule with a regular margin to low risk from intermediate risk. Uh, we added the regular calcifications with no super, uh, no protrusion, and then the spongiform nodule with calcifications were added to the low risk. And then the one that really changed the most, uh, is the intermediate group, and this is largely composed of those previously uh non-classifiable nodules, um, that with the old system, and then the very low and the benign did not change. OK. So now all nodules are classifiable. Um, the risk of malignancy is more inclusive of existing and emerging literature. Um, really only one nodule was changed from the 2015 system. We just added more to the system. So the one nodule type that was reclassified were those non-markedly hypoechoic nodules without suspicious features. We downgraded those to low risk from intermediate. So thinking about it in that term, it's really that intermediate risk category you kind of need to get familiar with, um, and then the cystic PTC in the high risk group. So I wanna go through a couple of cases here. So this is a 64 year old female. She has a 1.4 centimeter nodule. It's solid, it's mildly hypoechoic or non-markedly hypoechoic. Um, and it has one suspicious feature. It's got this coarse calcification here and you can see that it's a calcification cause it's got posterior signal dropout. So what is her classification? If we look at our, our diagram here, it's solid. Um, it is non-markedly hypoechoic. It has one suspicious feature, so it is an intermediate suspicion nodule with a 20 to 50% risk of malignancy. So the question is, what should we do with this nodule. It's intermediate suspicion, do we need to poke it or not? And with the old system, if it was over 1 centimeter, we recommended an FNA for intermediate suspicion nodules. Um, and now with this new system, one of the things we felt was really important to do was to include a threshold range for FNA rather than just a single cut point, because we all know clinically. There are times that we don't wanna have to biopsy a 1.4 centimeter intermediate risk nodule, but it's hard to justify in our current litiginous medical system that we live in. Um, so this gives us flexibility to say, you know, this nodule is of little concern to you in light of your other medical conditions, we can watch you. Or on the contrary, we may say, you know what, you're young, you're healthy, and I'm worried about this. I'm gonna stick it on the lower end of that spectrum of size. Um, so these are the size ranges that we have come up with for recommendation of FNA. And for low suspicion um nodules, the threshold is up to 2.5 centimeters. So this was largely why I was using two different uh systems, the ATA and the ACR tirades, because the ACR tirades had a little more of a sort of conservative approach to biopsying things and that less things needed to be biopsied. Um, and now with this, um, I can use one system and, and still have that flexibility. Um, and again, we still, for spongiform nodules, can, um, we can biopsy them, but we do not have to. Um, if we are going to biopsy them in certain clinical scenarios, if it's greater than or equal to 2.5 centimeters would be the recommendation. And again, if we have a patient who has those abnormal lymph nodes or clear evidence of extrathyroidal extension, or perhaps personal risk factors, family history of thyroid cancer, Um, then in that scenario, a size less than 1 centimeter may be used. So back to our patient, she's got an intermediate suspicion nodule to, uh, risk of malignancy is 20 to 50%. Our threshold range for FNA is 1 to 2 centimeters. So we can or can't, we don't have to poke this. So switching gears and thinking about the cytologic classification of nodules, um, and last summer, the um new version of Bethesda came out. Um, to provide a new range of malignancy risk and then also to alter some of the names. So the nomenclature for these categories has changed, um, whereas we used to call these AUS Flus, Bethesda 3 is now just AUS. They dropped the flus, um, and then the SFN or suspicious for. Follicular neoplasms was dropped as well because it was getting confused with SFM or Bethesda 5. and then the implied risk of malignancy is inclusive of iFP because that is surgically managed and so most of us were looking at that as if it was considered part of the risk of malignancy anyway. Um, so, so these ranges are inclusive of that. The management. Um, you can see is, is recommended there. So, um, with follicular neoplasms, the recommendation is either molecular testing or diagnostic lobectomy. Um, repeat FNA is not a part of that equation and surveillance is not a part of that equation like we see with AUS. So the, when we evaluate a patient with a thyroid nodule, we perform FNA whatever our diagnosis is, the biggest concern we have is what is the likelihood that I missed the diagnosis of cancer. Um, and so if we take that with each of the categories and look at non-diagnostic cytology, Uh, what we can see is that a nodule that has non-diagnostic cytology, its actual risk of malignancy is dependent upon its sonographic appearance. So this is where I want you to start thinking about nodules kind of more holistically. And thinking, OK, I have a very low risk looking nodule. It's cytologically non-diagnostic, that risk of malignancy is very low, but as we get more suspicious features, that risk of malignancy increases significantly, and that may change our uh diagnosis and our, our treatment approach. So the new guidelines will recommend consideration. Of lobectomy for patients who have repeatedly non-diagnostic cytology if the sonographic pattern is highly suspicious. On the other hand, if a nodule has a very low uh sonographic appearance and it's repeatedly non-diagnostic, the question we would ask is why are we poking it if it's low risk? And and this gives us the option to say, I don't have to do anything with this nodule other than watch it, um. So that was actually a strong recommendation, which we don't have a lot of uh in this guideline, unfortunately. So back to our patient, uh, the 1.4 centimeter nodule, she's within the threshold for FNA, um, and we perform the FNA and it's benign. So what is the likelihood, um, or what is the appropriate follow-up interval for this patient? Well, the appropriate follow-up is really not so much based on the fact that, um, you know, we feel like we have to follow this forever because we know malignant transformation doesn't exist within thyroid nodules, but we're, it's really driven by, did we get the diagnosis right on FNA. So again, looking at the sonographic appearance, that helps us to predict the risk of missed malignancies in these nodules. So as we acquire more suspicious features, the risk of missed malignancy or a false negative FNA increases significantly. Um, and so this is important for us as we think about when do we need to think about taking it out in spite of a benign FNA or do we need to repeat the FNA. So our guidelines for follow-up of nodules with an initially benign cytology are based upon the sonographic appearance of the nodule. So a nodule that has a benign FNA with a very low risk of malignancy, there is no need for ongoing surveillance. So that is a change. Um, if the nodule has a low risk of malignancy, we repeat the ultrasound follow up, um, if it has a benign FNA in 3 to 5 years. When do we repeat that FNA if the nodule changes to a more suspicious pattern? Um, and that holds true for the intermediate risk nodules as well. You can see the threshold or the the surveillance interval with a benign FNA intermediate sonographic appearance is 18 months to 3 years. And again, repeat the FNA if it changes to a more suspicious pattern. We felt really strongly that the size change was not as important of a um. A trigger for us to repeat the FNA because the data does not support that as a predictor of missed malignancy. Multiple studies have demonstrated that growth alone is not a good predictor of missed malignancy, but rather the sonographic appearance changing is a better predictor of missed malignancy. So within the text, we will say you can perform a repeat FNA for nodule growth, but it is not essential. Um, and then if a nodule has a, a benign FNA and a high risk of malignancy based on the ultrasound appearance, then the repeat FNA should be performed within 12 months. So, um, what is the likelihood that a nodule will change to a more suspicious pattern when we follow up? Studies suggest less than 4% of nodules actually change their sonographic pattern when we follow them up. So this gives us a lot of flexibility to say we probably don't need to do multiple biopsies on these patients, um, especially not for growth alone. If we do repeat an FNA on a nodule, and the second FNA is benign, the likelihood of a missed malignancy is exceedingly low and consequently repeat surveillance imaging is not necessary. Uh, and then if we've done an FNA, it's benign, and we've watched it for 5 years and it hasn't changed in size or appearance, then we no longer need to follow up, uh, this nodule with ongoing surveillance, and I would say this is my favorite guideline of all. Um, because I am still seeing patients who somehow get back into my clinic 10 years after their FNA, and they just want reassurance that the nodule is benign and it hasn't changed. Um, and so I'm using this as my, my, uh, sword to say no more. So, um, ongoing surveillance is not needed. If there are reasons to suspect growth or change or development of a new nodule, then repeat ultrasound is needed, but that's not surveillance, that's for diagnostic purposes. OK. Case number 2, this is a 44 year old female with a 1 centimeter nodule with no family history, no radiation exposure, and a normal TSH. So this nodule is markedly hypoechoic, um, and solid and may have some punctate ecogenic foci. So our nodule on our on our algorithm, this, uh, would be solid. Markedly hypoechoic, and then one or more suspicious features with its taller than wide shape, ecogenic foci, irregular margins, uh, potential for um extra thyroidal extension anteriorly, there's, you name it. This is a funky looking module. So high suspicion, risk of malignancy is probably greater than 50%. So our threshold for FNA is 1 to 2 centimeters. Um, and, um, I would say probably all of us would agree that, uh, this should be poked, um, given its appearance, and the FNA is AUS. So I want to talk a little bit about molecular testing. We know that molecular testing has really kind of come a long way, and um these two tests that we use most commonly commercially, um, are associated with an excellent negative predictive value, uh, giving us confidence when a nodule has a benign result that we can follow that as if the cytology were benign. We felt that it was important to talk about molecular testing that we state that a a clinician's confidence in using a molecular test should be based on the fact that validation studies have been performed and that real world validation studies uphold the performance of that test. Um, so I think that's an important component as more and more of these tests are being developed, that we make a statement to say validation studies are important so that we know how these perform in a real world setting. So the randomized trial, uh, recently at the um UCLA Health Systems, um, did a monthly block randomization study comparing. Uh, the GSC and the thyroic version 3 to all patients that came through their door with an indeterminate cytology. So patients were randomized for 1 month on one of the tests and then the next month on the other tests, they didn't get both tests, but they wanted to see how their performance, uh, stacked up next to each other. Um, and with a decent number of patients, almost 350 patients, what they saw was that these tests had very similar performance parameters, and the authors concluded that these two tests performed very similarly. Whichever one we choose to use is going to be based on our personal preference, uh, and ease of use or insurance coverage potentially. But really, the question really is, is when should we be using these tests? Uh, we know they perform well, when is the right time to use them? Does the ultrasound retrospectively modify the risk of malignancy in a given nodule once we have that cytology? So this is a compilation of multiple studies looking at the sonographic appearance on the um on the left is ACR tirades on the right is ATA ultrasound pattern. And what you see is at the very extremes of um the ultrasound appearance, so low risk and high risk, the um the risk of malignancy is pretty clear cut either very low or very high, and then those intermediate risk patterns. Um, have a more indeterminate malignancy risk, suggesting that we should potentially at either end of the spectrum, just the sonographic appearance alone can really inform our decision making about whether or not the next, what the next step should be in the evaluation of that nodule. Um, so unfortunately, um. What happened? There we are. Most nodules that come back with a Bethesda 3 or 4 cytology result are gonna fall into those intermediate sonographic patterns. But when you do see those very low risk sonographic patterns or those very high risk sonographic patterns, they can be informative. And so the, the UCLA group took this a step further in their randomized trial and looked at how did this test perform for predicting the pre-test probability of malignancy with ultrasound alone. And then they superimposed the prediction of malignancy risk when they added the molecular testing results on. So with the Affirma GSC, Bethesda 3 and 4 nodules that had a low or intermediate ultrasound appearance, um, the risk of malignancy in that group was kind of on the low end of the spectrum. When you added the molecular result onto that, the suspicious read gave you an increased risk of malignancy, sufficient to justify sending the patient on to surgery. But if you look at the high risk group, just the ultrasound alone gave you a risk prediction of around 78%. And then when you add on a suspicious molecular result, the risk of malignancy went up, but you pretty much have already made your decision right here. You know, you can feel confident with a risk of malignancy of 78%. A patient needs to have surgery. Um, and so even a benign result with this, with this test wouldn't necessarily reassure me, um, and I would want that patient to have surgery. Now, unfortunately, most of the nodules, you can see, um, had, most of the nodules had a lower intermediate sonographic appearance, but when that sonographic appearance was high risk, they were, most of them ended up being suspicious on the, on the test. So the test performed well for predicting malignancy or suspicious nodules, um, but again, it didn't really help you in your decision making. And the same thing was true with thyroy, those nodules with a lower or intermediate ultrasound appearance had a risk of malignancy around 24%, which increased with a suspicious read to 66%. And again, the ultrasound alone had a risk of malignancy of 80%. I think all of us would feel confident sending those patients on the surgery without the molecular test. So, The message here is look at the nodule. Does the nodule already look ugly to you and you have an indeterminate cytology, that's a high risk nodule. Those need to go on to surgery, those nodules that are intermediate risk, that's where we get our biggest informative piece, our biggest piece of information about where to go with this nodule. If it's low risk ultrasonograph by ultrasound, then it has a suspicious result. That's when it helps us to know we need to go on to surgery. So our guidelines for the management of AUS cytology, um, and this is not in any particular order, um, we can for very low suspicion nodules, even in the presence of an AUS cytology, perform surveillance sonography, um, in lieu of any additional workup. Um, we can repeat FNA, we can do molecular testing, uh, we can get a second opinion on cytology. And if the nodule is sonographically suspicious, or if the nodule is grown, the patient has compressive symptoms, or they strongly want this thing out, we should just send them straight to surgery. So, again, in no particular order of priority, um, but those are our options. And if we look at Bethesda 3 versus Bethesda 4 nodules, uh, the Bethesda 4 nodules, we recommend second opinion cytology, molecular testing, or surgery. Uh, repeat FNA is unlikely to be helpful, um, in this scenario, and ultrasound follow up would not be recommended. All right. So back to our case. 1 centimeter nodule AUS cytology, we referred her for a right hemithyroidectomy and she had a 1.1 centimeter papillary thyroid cancer. So, um, the nodule looks ugly and, and it is. So why does the nodule look ugly? Why does the papillary thyroid cancer look like that? Um, and so just buckle in. I'm gonna talk some physics here, but I think you guys can handle it. So what we're doing when we do an ultrasound, the ultrasound waves go through the tissues and at various points in their migration through the tissue. Some of the ultrasound wave is reflected back to the probe. When that happens is when the tiss the ultrasound wave travels through tissue of different densities. So it goes from um, from muscle to fat. There's a difference in density or acoustic impedance in those two tissues, and some of the ultrasound wave is reflected back to the probe. So in the normal thyroid histology without a nodule, there's all these acoustic interfaces between the follicular cells and the colloid where that is in the middle of those lakes of of follicular cells. So the result is the ultrasound hits the follicle and the the the difference in acoustic impedance between the follicular cell and the colloid shoots the ultrasound back up to the probe and the thyroid having all these lakes of colloid shows up as bright. So if we think about papillary thyroid cancer histologically, it's a densely cellular tissue that's packed with cancerous cells and the result is the ultrasound wave just goes through it, but it doesn't encounter anything with a different acoustic impedance. We don't see those lakes of colloid in there and so the wave is not shot back to the probe and it shows up as dark. Follicular cancers because they're composed of follicular units will still show up as bright because they have those lakes of colloid within them. So that's why follicular cancers look bright and the papillary thyroid cancers look dark. Now, if we think about papillary thyroid cancers, they, uh, from the TCGA and looking at almost 500 papillary thyroid cancers, what they found is that thyroid cancers kinda came in two flavors, the BRAF-like and the RAS-like tumors. And so those differentiations were also held true when we looked at the differentiation score, how many follicular fractions there were, and the underlying histology. So tall cell variant versus um classical variant, and then the follicular variants and the RAS-like uh group. So if we look at the sonographic features based on the molecular characterization of the nodule, a nodule that has a BRAF mutation is more likely to be on that upper row of the ultrasound characterization, those high risk nodules with irregular margins, dark eiteexture, lymph node metastasis, um. And then if we superimpose the TCGA with the sonographic appearance, you can see the BAF-like tumors are more likely to be hypoechoic, and the RAS-like tumors are more likely to be isoechoic, and cytologically that holds true as well with what we see with the follicular component and the colloid, um, and versus the densely cellular material. So I only have a minute left, so I'm gonna try to breeze through this. Um, 75 year old female with a 1.7 centimeter thyroid nodule. It's solid, it's isochoic, and it has a bit of an irregular halo, so that makes it a little suspicious. So the ATA classification for this nodule, uh, it's solid, hypo isoechoic, and one suspicious feature, so it's intermediate suspicion. Unfortunately, our patient also has metastatic breast cancer and she's on chemotherapy. So, um, she's got an aggressive cancer. So this is one of my favorite studies looking at patients over the age of 70 and just examining what their risk of dying for thyroid cancer is. So, They looked at the entire cohort of patients. They all had an FNA, and what they found is that only a minority of patients had a significant risk thyroid cancer. So approximately 1.5% had what was considered to be an aggressive cancer. It was uh medullary, poorly differentiated anaplastic or widely metastatic thyroid cancer. When they followed them for 48 months of follow-up, what they found is that only less than 1% of patients with 4 years of follow-up died from thyroid cancer, and all of those patients who died had a nasty thyroid cancer. Then none of the patients died from run of the mill, regular low risk thyroid cancer. All of the people that died from thyroid cancer were in this high risk group. And when they, when they examined them initially, they were able to tell who all those patients were with those nasty high-risk thyroid cancers. They were sonographically evident, they were cytologically evident or by other radiology. They were obviously the nasty cancers. There were no surprises in here. They were able to predict who all these people were in this high risk group right at the get-go. When they looked at who died from non-thyroid causes, they found that people who had other primary malignancies or coronary disease were 2 times more likely to die than their peers. So because of that, we wanted to consider and account for these factors when we're evaluating a patient with a thyroid nodule and making that decision about whether to stick a needle in it, their comorbidities, their age, their personal risk factors, if they had radiation as a child, family history, a cancer syndrome, um, if it's a PET positive, and then we also included patient preference in that as well, so the patient can also be a part of the sort of shared decision making. This is my second favorite guideline, and those folks who have a reduced potential for benefit, a more conservative use of FNA, uh, is appropriate based on the risk of greater risk of harm than benefit, meaning stopping um chemotherapeutic agent for a patient with widely metastatic breast cancer so that we can evaluate a low risk thyroid cancer. Um, so with that, I will quit and I'm sorry I went 2 minutes over. I just wanna say thank you to all of you for hanging with me, and this is the team of folks that I've spent the last 5 years of my life working on these guidelines with, so I want to acknowledge their hard work as well. Uh, I know it wasn't part of your talk and it would be a whole other talk, but the use of molecular markers to guide the surgical approach, is that gonna be included in the guidelines? I know you didn't get to that today. Perfect question. Um, it, the, the question is, will the use of molecular markers be in the guidelines in the terms of guiding. The surgical approach. And the answer is it will be, um, but not in our, it's gonna be in the cancer guidelines. So they are addressing that. So one of the questions is if uh for all high-risk nodules that are benign on FNA, we recommend a repeat FNA at one year regardless of stability and size and appearance. I think um we recommend a repeat FNA. Within 1 year, it doesn't have to be at 1 year. I think most of us, if we're highly suspicious, we'll repeat it within 3 months. Um, the clinical situation may dictate that though, if we've got somebody for whom they've got other cancers going on or other reasons to delay that, uh, diagnosis or the patient opts for active surveillance or something, we may not repeat the FNA, um, even if it was to come back as cancer. So I think that's a clinical case by case basis, um. No, that the question about, um, not sure I understand the question necessarily. So, yeah. OK. All right. Thanks. For one short question, yeah, you know, there's several papers published about the locations of cell nodules and noticing one of the slide measure isomers and posterior. What do you think about the the cell nodule locations? Yeah, I, I, I'm sorry, I, I couldn't, I am the, the nodules locations, you know, superior inferior, and isthmus. There's also anterior, posterior several papers published. So I think you. Asking about the nodule location as a favoring a higher or lower size threshold for FNA, and there is some limited data supporting um the fact that nodules in the isthmus may have a higher risk of malignancy and sometimes location too if you're worried about the potential for spread um externally to the thyroid, then that may make you use a lower size threshold as well. So, yeah. Thanks. Published March 22, 2024 Created by Related Presenters Jennifer Sipos, MD
CME Knowledge Hub
