TI-RADS Cases: Challenges in Thyroid Nodule Analysis - HD

Introduction to TI-RADS Challenges

After the discussion, maybe we could argue all Ty Rad cases are challenging to some degree, and I think that's true.

We are just having the conversation in the back.

And in one hand this is this is a great opportunity because really you will use your expertise to analyze the features of thyroid nodules, to be the decision maker about what goes forth in this followed or biopsied and worry less about the size, which was some of the conversation we were having in the previous sessions.

And this talk actually overlaps a fair amount with Franklin.

For those of you who have been here since the day began, you'll see some repetition.

Perhaps that's helpful. For those of you just on board, welcome.

I have no relevant financial relationships.

When I on an everyday basis I'm sort of thinking about how to use rads, I think I can boil it down to sort of three really challenging areas where I said, we're gonna have to use our expertise and that is correctly classifying epigenic foci and burn this word into your mind.

Try not to say calcification and try to think of dots as echogenic foci as a lead point to entertaining the possibilities really, and frankly mentioned this mixed cystic and solid nodules.

What is sponge formm? What is not sponge formm?

And what are some of the nuances of mixed cystic and solid nodules to pick the cancers out of the crowd?

And then margin features, margin features are very subjective and require a lot of detailed analysis.

And I think we'll boil it down to maybe these three categories.

We saw this before. This is a CR rads and it's full glory.

And Franklin showed a whole lot of fancy laminated pictures that he has all over his ultrasound lab to make sure that everyone is familiar.

With these, I don't think your technologists and the technologists that are in the audience, you have to commit this to memory.

What you do need to do is have in your mind, what are the suspicious features I'm looking for?

And the rest will take care of itself.

This may be a little bit blurry, I just magnified it.

Right from the chart.

Echogenic Foci

And we're gonna talk about echogenic foci and echogenic foci come in a couple of different flavors.

We can see that there are none or a large cele artifacts.

And we'll look at some examples. Macro calcification.

Macro calcification means an echogenic focus associated with distal acoustic shadowing, just like anywhere else in the body.

Peripheral rim calcifications, we're not gonna get into Franklin covered that nicely.

And then this punctate genic foci, which may and in fact more often are not microcalcifications.

And what we'll talk about that.

And then this little boxes that are at the bottom are incredibly important.

Large comal artifacts, it's defined to them V-shaped greater than a millimeter macro calcifications, as I mentioned, they cause distal acoustic shadowing.

And then punctate echogenic foci may have these small comal artifacts.

This is a slide that actually shows what Sam bodies are, within papillary cancers of all types.

And as a matter of fact, in some papillary non-cancerous lesions, for example, meningiomas, what Sonoma bodies are are in essence sort of a sloughed dead papilla that has developed dystrophic calcification for whatever reason.

And these have a very high specificity for a papillary thyroid cancer.

When they are histologically seen, you have a Sonoma body on the slide, it therefore almost always is a cancer.

Some unusual cases of Hashi can have it and so forth.

But in pathology, when you see a Sonoma body, you know, you have a papillary thyroid cancer, now microcalcifications, or at least what we think are microcalcifications, are supposed to be aggregates of these Sam bodies.

And microcalcifications were defined as punctate, less than a millimeter in size, no distal acoustic shadowing.

And in thyroid literature have a reasonably high specificity for papillary thyroid cancer.

The key is they are almost always in solid and hypoechoic nodules because papillary cancers are almost always particularly classic forms, solid and hypoechoic.

I have a couple of examples here.

Notice how these are solid hypoechoic and they actually have a variety of calcifications, which is very common among papillary thyroid cancers as well.

But again, emphasizing in solid and hypoechoic nodules or portions of nodules.

This paper was referenced before, these are actually cancers.

These are all papillary cancers that have echogenic foci within this cohort of cancers.

If you see an echogenic dot, is it a micro calcification?

The answer was no.

Even in proven cancers, the minority of these were Sonoma bodies.

They were other things that cause bright dots.

Other types of calcifications, colloid back walls of cystic spaces.

So even in cancers, those dots are usually not oma bodies.

And we're gonna talk about in benign nodules, the other echogenic FSA that we see.

And as you would imagine even less likely that in a cystic nodule, that same is going to be any sort of a calcification.

This case was referred in a nodule with microcalcifications, and my eye is drawn to a number of things.

It's drawn to these very long and linear echogenic areas, which are the back walls of these cystic spaces.

And then these other echogenic foci, which clearly have reverberation artifact.

And previously people were using the term micro calcification.

If I had to point to one single major advantage of all of a CI ti rads, it's the introduction of this term genic foci.

Do not think or say calcification, think genic foci.

And then you're gonna subcategorize from there.

Of course not. This was a benign hyperplastic nodule.

Architectural Distortions and Spongiform Nodules

One of the concepts that I think is really important about thyroid nodules is the vast majority of nodules are not mass lesions.

They're actually architectural distortions within the thyroid parenchyma.

Many of us being diligent radiologists wanna apply all that we've learned about mass lesions to thyroid nodules.

For example, if we had a renal lesion that had internal flow, we'd be very concerned and rightfully that it's a renal cell carcinoma.

But if you instead say, most thyroid nodules are not mass lesions, they're architectural distortions, a lot of those rules go away.

The most typical appearance of a architectural distortion is something we call a sponge offor nodule.

Here's a histologic slide of a thyroid.

Here's the normal follicular architecture with colloid.

And here's a nice big sponge offor nodule.

It's an architectural distortion. It is not a mass lesion.

It's fine to have internal flow, it's fine to have ill-defined edges, and we're gonna talk about that.

It's fine to have a lot of the things that are not okay in mass lesions, but that is not the analysis that we wanna do.

And they're not uniformly benign.

We saw a case of this in Bill's literature that he quoted.

Very, very rarely a papillary thyroid cancer can have this appearance.

But really all the guidelines are willing to let these go because they're relatively uncommon.

Sponge offor nodules, you have to set your own mind's eye about where you wanna call sponge form nodule, roughly speaking, the definition that most people use is a nodule that has at least 50% interspersed cystic spaces, hence looking like a sponge or a Swiss cheese.

And I think four really great examples of sponge offor nodules.

Many of them have these large echogenic foci with reverberation artifacts from inspissated colloid within these nodules.

This is important because if you think you're dealing with a sponge offor nodule, you are done.

You have zero points.

You do not assign any other points for any other findings.

Really where you draw the line in the sand, sponge form or not, is in fact very important.

Let's talk about these now.

Echogenic foci, they are no shadowing, and they come in a variety of different shapes and forms.

The pioneering work on this was done by Mike be and up brown.

And he said, let's sort of define these.

Let's give this a language.

And he divided them into large C tail with an inverted triangle behind linear without a clear triangle.

Look at how big these are.

Definitely longer than a millimeter.

And then round and punctate. A little bit less specific.

Here's the classic appearance of the large CMA tail in a large colloid cystic nodule.

And in real time, this is much better appreciated.

We can see these little reverberation artifacts as inverted triangles behind the nodule.

And in this pioneering work, what they found in a small series is that these two on the top were uniformly benign.

And the ones on the bottom, a minority of them were cancers as the highest being in these very punctate small and round epigenic foci.

And so these are the potentially worrisome ones and the top, really not so much.

They went on and he teamed with harsh on Malley at UCLA.

And they did this really fabulous study with over a thousand patients that were coming for biopsy.

And they assessed epigenic foci into five different categories.

And I'll show you examples, punctate without artifact, echogenic foci was smallest than one millimeter com tail large com tail, clump like peripheral, with shadowing.

And they recorded the number and types of echogenic foci for all of the nodules that were undergoing biopsy.

And this is right from their paper.

Punctate, no artifact, small C tail physically measured here at less than a millimeter, our friend, the large oid reverberation, and then clump like with shadowing.

This would be a macro calcification.

The other examples of other types of echogenic foci.

And what did they find? Almost 75% of the nodules had echogenic foci.

Papillary thyroid cancer is rare.

The incidents in their cohort of cancers was less than 15%.

Clearly all of these echogenic foci that they were seeing were neither microcalcifications nor a slam dunk for being a cancer.

They actually found that there was no statistical significant difference in identifying any of these types of echogenic foci in the cohort of malignant nodules as compared with benign nodules.

I just wanna show you a little bit of the data from their paper because I think it's very into very instructive.

Large cele artifacts, which means a colloid is not uniformly benign.

There were two malignancies, and we're gonna look at one from the paper and I'll show you another example.

But all of the other types of echogenic foci, the risk of cancer was about 15 to 20%.

Slightly above baseline, generally in a busy biopsy scenario, about 10% to 15%.

But not statistically significant from the benign nodules that were referred to their institution for biopsy.

The other thing they did was they looked at the parenchymal pattern, whether the nodule was predominantly cystic and if the solid part was hypoechoic.

And again, you can see here that being predominantly cystic was a very strong predictor of being benign.

And having hypo echogenicity was a very strong predictor of being malignant.

And that is what we should be again, concentrating on is, is this echogenic focus in a mixed cystic and solid nodule or one that's all solid hypoechoic, or the solid component is hypoechoic.

Let's take a look at these two abberant nodules, at least one of them where there was large comal artifact.

Notice how in both instances it was in a portion of a nodule that was hypoechoic.

And again, this is right from their paper.

This is a lobulated ill-defined markedly hypoechoic nodule, and absolutely I can point to and they can point to an area where we see large comal artifact, but all of the other features of this nodule outweigh that.

Some people think comal artifact equals colloid equals benign.

No, usually it does, but it can occur in a malignancy.

Many malignancies have areas of colloid.

It's always all of the different features.

If you see comal in a solid and hypoechoic portion of a nodule, cannot emphasize this enough, that's the suspicious feature.

Additionally, they found that the number of echogenic foci per nodule increased the risk of cancer as an increasing number of any kinds were found.

If you just had one of the different types of echogenic foci, your risk was relatively low.

But if you had more than two types, you had at least a 25% risk of being a cancer.

This is a nodule that we had recently in our lab.

It's about a 2.8 centimeter nodule.

And there's all kinds of echogenic foci that catch my eye here.

And I'll be showing you a cine in just a moment.

We have some reverberation artifact, we have some cystic spaces, but I am concerned because this is a predominantly solid and hypoechoic nodule.

Let's take a look at real time.

Absolutely there's reverberation artifact that's coming from here, but in several other portions of this nodule, there are other types of echogenic foci.

And so that is what is suspicious about this particular nodule.

Let's see if my clip will play.

Here's the color doppler.

Notice how none of the criteria talked about vascularity.

I'm not going to get into that.

But vascularity by itself is not a predictor.

It's only a predictor of malignancy if it occurs in a package with other suspicious findings, particularly in the United States.

And this was a papillary thyroid carcinoma.

And again, it is the solid and hypoechoic nature of this nodule with the echogenic foci.

That is the concerning finding, just a po of micro calcifications in papillary cancer.

Here's one, predominantly solid.

And here's a very large cystic one where the area, the solid component is what we feel concentrate on and we'll get into that as well.

This sort of tough like looking area literally looks papillary and is announcing itself to the world.

I think look at me, even though I'm mixed cystic and solid, I am a papillary thyroid cancer.

These are the trickier ones, these ones that are linear without reverberation artifact.

But notice how this is in a solid component of the nodule, and again, a teeny tiny small C tail, but it's within the solid component of this nodule, not really a mixed cystic and solid nodule.

We see a case like this about once a month, where people once to call this area in the center of this nodule calcifications.

The nodule sort of looks like it has irregular margins in its hypo coic.

But look at how bright this area is and how large it is.

There's no distal acoustic shadowing, and this is what the nodule looked like three months earlier.

Just a pitfall of a colloid containing nodule, which can spontaneously regress or after it's been biopsied can involute.

And as that colloid sort of aggregates in the center, it can fake you out as being a type of calcification.

Don't fall for that. A very interesting article that I actually came across just recently, I apologize, it's not in your handout, but a very interesting point about ultrasound features and how we perceive them with respect to nodule size and features of papillary cancers larger than a sonometer.

These are our odds ratio, irregular margins, which we'll talk about 38 times, microcalcifications 18 fold and so forth.

And no vascularity was a risk factor.

This is why vascularity will never stick.

But for capillary cancers that were diagnosed under a soter in size like they do in Korea, notice how microcalcifications was only just under a fourfold risk.

And I think the reason is in very, very small nodules, it's extremely hard to differentiate.

Is it a mixed cystic and solid or a term I think of compact sponge offor or are there truly micro calcifications?

And my advice is resist the urge to jump on that three points, hold back and just say it's indeterminate.

And I sort of say if I'm unsure, I always choose the lower of the two point scales because most cancers have several features that are going to announce themselves as the from the same paper that shows when we look at large cancers, the number of malignant features that they chose, really the majority of them had many, many features in all instances, it's pretty atypical to just have one quote malignant feature and truly be a thyroid malignancy.

Take endpoint echogenic foci are most concerning when in the solid component of a nodule, and especially if in combination with other features such as mar typo echogenicity and other calcifications.

They're most commonly misinterpreted to be microcalcifications in those small nodules.

And large com tail artifact does not always mean benign be on your guard if they're occurring in combination with other suspicious features.

Composition: Spongiform and Mixed Cystic and Solid Nodules

Okay, composition generally straightforward, but a couple of nuances.

We already a little bit tackled sponge formm or not, but we're gonna revisit that a little bit and where you draw your line between all solid and predominantly solid.

Just refreshing our memory, these are the sponge form nodules, which in essence means game over.

You are confident it's a sponge form nodule.

If it looks like this, you don't apply any more points for echogenic foci literally game over it's sponge form done.

You don't need to do another thing.

And where you draw the line, as we mentioned before, is somewhat subjective.

And this was the nodule I showed you before.

Nodules that are solid or predominantly solid, it's interesting, there's a little bit of a difference in opinion between the lexicon that was published and then that user's guide that bill and Franklin and Jenny wrote.

And they sort of say about 5% or less cystic space, you can call the nodule predominantly solid.

If it's 49% cystic and 51% solid, you don't wanna be calling that predominantly solid, and where you draw the line is going to be subjective.

I've seen people call a nodule like this predominantly solid, and I've seen people call a nodule like this predominantly solid.

I personally would call both of them mixed cystic and solid.

And that's a point difference between those two.

If you were to be using the a TA, this would be low suspicion, this would be very low suspicion.

I think our endocrinology friends got it right by having these at lower levels of suspicion.

We touched upon this before the mixed cystic and solid nodule.

About 6% of papillary cancers are predominantly cystic, which is defined as at least 50% cystic content.

And actually the a TA atlas that we saw before, Susan Mandel, who's the major author of that, is from my institution, and we share slides.

I had a laugh because a lot of the images in there are from our shared slide decks.

This was a benign nodule to Franklin.

Doesn't need to worry, we biopsied it, but the cancers usually announce themselves that they have tough like areas, they have calcifications, they really look quite different than this nodule, which if you were to suck the fluid out of here would just look like a normal little island of thyroid tissue.

Again, that histology side is always in my mind, sort of a disruption of architecture and so forth.

And this was a really excellent paper in the American Journal of Neuroradiology, which sort of walked through the complexity of mixed cystic and solid nodules with an ascending risk of cancer as the solid component becomes less concentric and more eccentric and certainly heightened when that solid component has other features such as tuft it, irregular margins, calcifications and so forth.

And Franklin alluded to that, that the next version of RADS may be adding some of these components to the analysis of mixed cystic and solid nodules.

Most cystic papillary cancers, I think, you can pick out of the crowd.

Here's one that has this very large solid component, very vascular here, calcifications OID galore.

And here it is in a asinine loop, where really the solid component looks just like the papillary thyroid cancer I showed you earlier, just in this case, living in the middle of a cystic nodule.

And just again, a popery of cystic papillary thyroid cancers.

Hopefully you would pick these out of the crowd.

And again, notice how the suspicious epigenic foci are in the solid components, not in the mixed cystic and solid components of these nodules.

Most mixed cystic and solid nodules, in fact ar benign.

Remember to analyze the solid components of the nodule and a question that comes up about reporting nodules.

You use the whole size of the nodule when using the ti rads, even if the solid component is just a small little expressions in it, it's the outer diameter that we're going to use.

Margin Features

Okay, margins. Here is margins and we have ill-defined and ill-defined and irregular are not the same and particularly the Korean papers.

You wanna be careful, these are very different.

And we'll analyze that in a minute.

Ill-defined does not carry weight in thyroid world.

And we will talk about why, irregular is a synonym for jagged and spiculated when you read the literature.

That's how the Asians use that term an extra thyroidal extension.

Lobulated means protrusions into the adjacent tissue, irregular, jagged spiculated, sharp angles, not ill-defined, ill-defined is not a big deal for the reasons we discussed before.

If you think about an architectural distortion in a thyroid, it almost by definition will have ill-defined margins because it's not a mass lesion, it's just a patchy ill-defined area of architectural distortion.

I borrowed this from the breast literature, which defines all of the margins that people look for when doing breast ultrasound analysis.

And all of them hold true as a risk for thyroid malignancy, again, with the exception of the ill-defined margin, not in thyroid land.

The breast does not have ill-defined patchy areas.

It has mass lesions.

We're gonna depart from our mass lesion analysis when we think about that.

Just a popery again, of examples, speculated margins, tentacle like areas extending into the parenchyma, jagged irregular margins.

Here's a nice taller than wide nodule with course calcifications.

Again, tentacle like areas and a commonality, again, solid hypoechoic.

Most thyroid cancers have several abnormal features, very unusual to just have one that grabs your eye.

The evaluation of margins is very operator dependent and really requires very careful analysis of all of the margins.

For those of you that can capture SI highly recommend it particularly for this feature.

I think it's very helpful to try to catch some of these more subtle marginal features and so forth.

Lobulated margins, whether they're macroscopically, lobulated or just small loation.

Again, architectural distortions tend to lack that.

If you see a little lobulated area in a nodule that should be viewed with suspicion.

Micro loation as well.

And again, solid and hypoechoic nodules.

A theme here for all of these that are papillary thyroid cancers.

Back to the poorly defined margins, here is a very benign appearing nodule.

And I would describe this as poorly defined.

I could not take a pencil and clearly tell you where the nodule begins and ends, but that is not a concerning feature and that's why it has zero points in the scoring system.

That was a question that came up earlier.

And again, just emphasizing that there's a very ill-defined margin here, but it matters not when we're looking at thyroid nodule because the vast majority of these are not mass lesions.

Again, this was one of my slides that made it to one of the papers.

This was an anaplastic cancer and I think just about everyone would be able to look at this and know it's a very bad actor.

It's growing right through the outer so-called capsule.

It's not a true fibrous capsule, but the outer white genic line that defines the edge of the thyroid parenchyma into the muscle.

Here's a little bit of normal thyroid here for the eagle eye.

In the audience, you can actually see tracheal evasion, which is something if you do very careful ultrasound, you can sometimes pick up.

What are we looking for with extra thyroid extension?

We're gonna be looking for this white line that goes over the top of the thyroid here.

And if there's a mushroom or a protrusion of the nodule beyond that, we can with certainty, say we have extra thyroidal extension.

Here's the sagittal view. I can see the white line.

It is disrupted here and it picks up again here.

And let's look at that with a cine loop so you can hopefully appreciate this little mushroom of tissue.

And we have another very important clue.

We have metastatic lymphadenopathy, both posterior to the thyroid back here as well as pre laryngeal and as well as lateral cervical lymphadenopathy.

This was a fairly aggressive cancer demonstrating all of those features.

Here's a closeup of the extensive central compartment and a little bit of the lateral compartment lymphadenopathy, but more often than not, we don't have gross invasion.

And instead we may be just looking at a thyroid nodule that abuts the capsule.

There is literature to show that if the abutment is at least 50% of the diameter there, there's a very high risk of some sort of invasion.

It might be microscopic, it may be macroscopic.

But that is in fact what we are looking for.

And if you zoom up here again, I think maybe the eagle eyes in the audience could appreciate there is this little mushroom of tissue extending beyond the contour of the thyroid, which is very suspicious.

Now sometimes we're gonna have a false positive.

This is a very recent false positive where one of my very amazingly excellent sonographers had this very maged up view and she was very concerned about a break in the capsule here.

But that was just a feeding vessel.

These were not papillary thyroid cancers.

The patient had a cancer on the other side as it turns out.

And so that was just a rare false positive here.

We can define definitive extra thyroidal extension.

When we truly see the disruption of the so-called capsule, we can be very concerned about it if we see a lesion that has greater than 50% contact with the margin.

And we have a very high negative predictive value.

If you can clearly see normal intervening thyroid parenchyma over it as it stands now, the T reds considers just a bulge without any of these features suspicious.

And I would very respectfully disagree with that particular characterization.

I do not assign three points to a nodule that's just bulging the capsule into which I can actually see normal parenchyma going over it.

Here's definite extra thyroid extension, then you just say, this is certainly a cancer and you don't need to add points.

But again, this would have gotten all sorts of points.

If it's questionable, this is a suspicious nodule where I am uncertain, I will assign that three points.

But often a very benign hypoblastic nodule just happens to be in the isthmus or in a location where it's very innocuously bulging the capsule.

That is not suspicious in my mind.

Take home messages for these ill-defined margins are not suspicious, unlike in other disease processes.

Aggressive margins meaning infiltrative, jagged spiculated, and irregular is a synonym in that sense.

Ill-defined not irregular are not the same thing.

Definite extra thyroidal extension.

We know we have a cancer and beware of the bulge.

The bulge can be either associated with an innocuous nodule, in which case I don't even mention it.

Or if you have a nodule which already is looking suspicious, I will then upgrade one of the few times I will in fact upgrade.

Challenges with Current TI-RADS

In my opinion, I think that these are some of the challenges of RADS as it currently stands.

And for those of you that were in the last session, it is meant to be applyable guideline is meant and it is currently under revision with us, a number of sites reporting back our experience about how well it's working with the intent to either dial up or dial down the number system as it stands if we've over called or under called.

I think echogenic foci unfortunately are still being over called.

And it's my expectation that the next iteration will dial that down to one.

Because as we mentioned, most other nodules will be solid.

Hypoechoic have funny margins, they'll be well over the threshold.

They don't need those three points.

And the ones that are marginal, if they're only being boosted up to a higher category just because of people worrying that these are echogenic foci, we'll solve that in one fell swoop.

And bill Middleton told me that already the data miners are working on that fortunately, where you draw your line between all solid, predominantly solid, mixed cystic and solid and where you draw your line, there is very important.

And again, the marginal analysis, don't over call the budge if the nodule is usually not suspicious.

And as we mentioned before, this is not a system that's going to detect all cancers in the world.

I am getting some feedback from endocrinologists in our area.

This was feedback that I received.

This particular patient had her ultrasound somewhere else, but I was asked if I would agree with the characterization of this nodule as a T rats two.

And technically speaking it is, it's mixed cystic and solid.

The solid component is ISO or hyper coic.

And at least on the static images that I was provided, there were no calcifications that I could see in retrospect.

I'm wondering if I had a syn a if there may have been some very subtle microcalcifications in there, but because of its size it was biopsied and this was a papillary thyroid carcinoma.

Another challenge I think is trying to identify some of these very subtle features just on static images.

Again, I'll put a plugin for having your technologist acquire cna whenever possible, or if there's a question, mult multiple images.

And this was actually a pure papillary thyroid cancer.

There is a type of thyroid cancer, which is a follicular variant, in which it truly is a cancer under the microscope, but it's very non-aggressive in its appearance.

I thought it would be that it wasn't.

It was just a flat out regular papillary thyroid cancer that looked like that again, in the eye of the beholder.

Is this solid? Is it mixed cystic solid? What is it?

I may have described this as mixed cystic solid, to be honest, with you.

Some people may have said sponge formm and punchline, this was one of these follicular variants of papillary thyroid carcinoma.

I don't think that, I think as I said before, let's view this as an opportunity to add our expertise to thyroid interpretation by feeling comfortable identifying and tackling all of these subtleties.

And hopefully we'll have a deification over time on size and this will out value to us by being to really able to triage the suspicious features from the nons suspicious features.

And thank you for your attention.