Update on Recommendations for the Biopsy and Follow-up of Thyroid Nodules - HD

Introduction to Thyroid Nodules

Hi, Joe Langer, and I'll be talking about new guidelines
for biopsying thyroid nodules.

Before we get started, I wanna review a couple
of concepts, which seemingly are simple,
yet when you think about them are slightly more complex.

First of all, what is a thyroid nodule?
Secondly, how common are thyroid nodules in adults?
And third, how common is thyroid cancer?
To give us a perspective.

Definition of a Thyroid Nodule

One of the concepts
that comes about quite frequently is exactly
what do we mean when we see,
when we say we see a nodule on ultrasound?

A nodule is defined as an area of the thyroid
with a different architecture than the
surrounding parenchyma.

And as it turns out, fortunately,
most thyroid nodules are not tumors,
but are just hyperplastic regions of the thyroid.
And I'll demonstrate that to you in just a few minutes.

And also, most thyroid nodules are now detected
incidentally, that means the patient does not have symptoms,
does not have a palpable abnormality on their neck,
but is undergoing some sort of an imaging examination,
ultrasound or another type.

Prevalence and Detection of Thyroid Nodules

And in fact, we see a nodule when we talk about
how common thyroid nodules are.
It very much matters how they are detected.

As you can see with this chart here,
which indicates prevalence of thyroid nodules on this side
and then the age of the patient along the bottom.

If we use palpation or feeling a nodule on physical exam
or the patient feeling a nodule themselves, you can see
that the incidence increases with age.

And if you take a 50 to 60-year-old patient,
approximately five to 10%
of them would have a nodule felt on physical
exam on the thyroid.

But if we look at either autopsy
or ultrasound data, we can see that in fact,
thyroid nodules are incredibly more common
because of the sensitivity of these techniques
as compared with palpation.

If we take that same 50 to 60-year-old adult, we can see
that nearly 50% of patients would have thyroid nodules
and those over the age of 60,
the majority would have thyroid nodules if we look for them
by ultrasound or autopsy.

Differences Between Palpable and Non-Palpable Nodules

Now, what is the difference between nodules
that are felt on physical exam and those that aren't?

Certainly the size matters,
and as you would imagine, very small nodules,
smaller than a centimeter in size.
The majority are not felt on palpation,
but are detected by imaging.

But in the one to two centimeter range
and greater than two centimeter range, still many,
many nodules are missed by palpation.

And this could be because they are in the back
of the thyroid gland.
The patient has a large neck
and the palpation of the thyroid is limited.

The key question would be,
is there anything different about thyroid nodules
that are felt or not felt in terms of the likelihood
that they're going to be a cancer?

And multiple studies have shown no difference.
So being palpable
or non palpable does not
influence our decision making about whether
that nodule could or could not be a cancer.

Biopsy Outcomes and Types of Thyroid Cancer

We know that nodules do come to biopsy
that the vast majority of them are benign.

As I mentioned before, at least 70%, about 10% are a tumor,
but a benign tumor, a thyroid adenoma,
and somewhere between five and 12% prove to be cancer.

Of those that we have selected for biopsy, the most common,
common type of cancer is papillary thyroid carcinoma,
which includes pure papillary and mixed papillary
and follicular cancers.

Pure follicular cancer, about 10 to 15% medullary cancer,
five to 10%, and anaplastic less than 1%.

So when most people are talking about thyroid cancer,
they are predominantly concentrating on these papillary
thyroid cancers because these are,
those are the most common.

Some unusual things will also undergo biopsy.
For example, an area of thyroiditis that was not suspected
to be such and therefore did undergo a biopsy.

Hyperplastic Nodules

Now, as I mentioned before, I was going
to describe the hyperplastic nodule,
which is the most common type of nodule we see.

If we look here at this histology slide,
we can see the normal thyroid
architecture in the bottom here.

And then this is what a hyperplastic nodule represents.
In essence, it's normal thyroid tissue,
but in a disorganized fashion
or what I like to call an architectural distortion,

the normal thyroid has a very fine reticular network
of colloid, which is the protein
that thyroid hormone is made of, outlined by a matrix
of follicular cells.

The hyperplastic nodule has exactly the same thing,
but the regions of colloid are very large and
therefore become evident by ultrasound.

This very tight reticular pattern shows up
as the normal hyperechoic background of the thyroid,
whereas the thyroid nodule is usually visible
because of this colloid and fluid, and
therefore we are able to see it.

So these are the most common type
of thyroid nodule that we see.

And if you notice, they're not a mass lesion,
they are just an architectural distortion.

That is important because when we apply ultrasound analysis
to these nodules, some of the rules that we follow
for mass lesions will not be applicable.

The Epidemic of Thyroid Nodules

This discovery of all
of these thyroid nodules, not only by ultrasound
but also other imaging exams such as CT
and MRI of the neck has been called the epidemic
of thyroid nodules.

As we discussed before, in adults over the age of 50,
the vast majority
of them would have thyroid nodules if we looked
for them with ultrasound.

Fortunately, thyroid cancer is relatively uncommon.
About 65,000 cases are diagnosed each
year in the United States.

So the issue is what are we going to do to reconcile all
of these thyroid nodules
that we see when we image against the very small number
of them, which we know will be cancer?

We need to look at some sort of a mechanism
for distinguishing between these incidentally detected
nodules in low risk patients without symptoms
and figure out who should undergo further diagnosis
and management to find the very few cancers.

And this brings up the concepts of over-diagnosis and
overtreatment, knowing
that these are asymptomatic patients,
who we are dealing with.

So what we're gonna spend the rest
of the time talking about is
what guides us in deciding which nodules should undergo fine
needle aspiration when incidentally detected, I'm not going
to talk about high risk patients
or congenital syndromes and so forth.

And we're going to focus in on how malignancy can be
stratified by risk using the ultrasound
appearance of these nodules.

Sure Signs of Malignancy

Now, what are sure signs of malignancy?
These are probably intuitively obvious to you.

Certainly if we see evidence of metastatic disease
and on neck ultrasound, this would be available to us
by looking at cervical lymph node metastases
or if we were to see direct invasion of that thyroid nodule
beyond the thyroid into the adjacent soft tissue,
something called extra thyroidal extension.

Extra thyroidal extension is actually relatively rarely seen
when we do neck ultrasound.

Metastatic disease
to cervical lymph nodes may be occult on ultrasound,
but also may be visible as long as we remember
to include an assessment
of the lateral cervical lymph node chains
while we're doing the scan.

And this is recommended for all neck ultrasound now
and as part of the AIUM guidelines.

Examples of Extra Thyroidal Extension

So here's an example of a thyroid nodule whereby ultrasound,
we can see direct invasion through the capsule
and into the overlying muscle.

The thyroid cancer is outlined with the electronic cursors.
Over here we see normal thyroid parenchyma
and these white arrows show us the capsule,
this nice thin white line of the thyroid.

As we move right over the nodule here,
we can see not only a bulge in the capsule of the thyroid,
but actually some of the thyroid
malignancy is extending itself, disrupting the capsule,
which we cannot see at all,
and also growing right into the muscle
overlying the thyroid.

So this would be evidence
of direct extra thyroidal extension, a sure sign
of malignancy when we're able to see it by sonography.

Examples of Metastatic Lymph Nodes

Another indication
of metastatic disease would be our ability
to see metastatic lymph nodes.

Here we have a transverse view of the right neck,
and in the right lobe of the thyroid,
we can see this papillary thyroid carcinoma.

But in the lateral cervical lymph node chain,
we can actually see a metastatic lymph node.

So when we have this finding, we can be sure we're dealing
with a thyroid cancer as well as metastatic disease.

So here we have a thyroid cancer, relatively small,
only about 11 millimeters.

And for those of you with very sharp eyes,
you may have questioned whether
or not there's a slight amount of extension
beyond the thyroid capsule.

This would raise concern for a malignancy,
but if we look out in the lateral cervical lymph node
chains, in fact, we can see a metastatic lymph node.

So we are sure of it.

The good news is that this very minimal amount
of extra thyroidal extension would actually usually not
change the surgical approach to this patient as opposed
to if there were gross extension.

But certainly putting the possibility
of this in the ultrasound report
to the surgeon is very possible
to alert them in case they would require additional imaging
prior to surgery.

Just as a reminder, papillary thyroid carcinoma metastases
can have a variety of appearances regardless of
what the primary tumor looks like.

They may be solid, they may be hyperechoic or hypoechoic.
This particular lesion is hyperechoic as compared
to most thyroid cancers and contains calcification.

This lymph node metastasis is mixed with both a cystic
and a solid component with some specular reflectors,
which could represent either colloid or calcification.

And this lymph node metastasis is a nearly
entirely cystic lesion.

And in adult patients who present
with a new cystic lesion in the neck,
metastatic papillary thyroid carcinoma is one
of the major considerations we need to talk about.

Microcarcinomas

Another thing that is quite common with thyroid cancer,
and that is the concept of a micro carcinoma.

A micro carcinoma is defined
as a cancer under 10 millimeters in size.

These commonly are detected on surgical
and autopsy specimens.

This was a patient who had had a FNA
aspiration of this nodule here, which was a follicular adenoma.

The patient had surgery which confirmed that diagnosis
that this was a benign lesion,
but also the pathologist noted this three millimeter
focus of thyroid cancer.

As it turns out, this is quite a common finding
for the pathologist to see in autopsy studies.

These papillary micro carcinomas have been reported
to occur in as high as 36%
of specimens in surgical series as high as 22%.

This may raise a concern that are these precursor lesions.
In other words, are these small cancers designed
and destined to grow up to be a larger
biologically significant cancer?

Overwhelmingly, the literature has shown
that this is not the case.

We have both epidemiology studies that show
that the countries that have the highest incidences
of these small micro carcinomas have no increased risk
of having a larger cancers.

There's a very different genetic predisposition for these.

And there's an equal gender prevalence
of these small tumors,
which is quite different from regular larger thyroid cancer,
which is much more common in women.

Pathologists consider this an incidental finding that is
that the vast majority
of these micro carcinomas are dormant lesions,
which only come to detection
because the patient had surgery for a larger lesion
or an autopsy series
because the pathologist was looking for these lesions.

Incidence and Mortality Trends

So this ties in with a very interesting result
that we've seen over the last several decades in the
United States, and that is that the incidence
of thyroid carcinoma is clearly increasing over time.

Yet this line here shows that the mortality
of thyroid cancer has been really stable across here.

Diagnostic ultrasound became very popular in the seventies
and thyroid needle biopsy in about the eighties.

And if we look very carefully here, we'll see
that the diagnosis of thyroid cancer started
to increase about 1991.

And this is largely thought to be due to the tools
of ultrasound detecting many, many more nodules.

But if you note without any change in mortality,

and if we look at this data a little bit more clearly,
and that is at the size of thyroid cancers
that are being detected, we can see
that the larger cancers have not undergone increased
detection neither in the two to five centimeter range,
nor the larger than five centimeter range.

But really what accounts
for this upswing in detection is really the smaller tumors
one to two centimeters in size,
and tumors are less than a centimeter in size.

As we just discussed,
the pathologist largely consider this an indolent finding
without any biologic potential to harm the patient.

In the vast majority of instances, so many have noted
that this lack of change in mortality is
because all we are doing is detecting smaller
and smaller cancers,
but these cancers would not play a significant role in the
mortality of any of those patients.

And this leads into that concept of over-diagnosis
and overtreatment of an indolent finding.

Over-Diagnosis and Overtreatment

If we look at the thyroid cancer statistics in the
United States, we can see
that thyroid cancer has the highest increase in diagnosis
of all cancers, nearly a threefold increase from 1975
to 2009.

And as I mentioned before, the large majority
of this increase is because of these indolent cancers.

We can see again that stable mortality rate over the
last three decades.

And again, this is due
to a disproportionate increase in the small cancers which
have a very low biologic aggressivity index.

And so what we are doing is creating more
and more people with the diagnosis of thyroid cancer,
but not at all changing the mortality statistics.

This particular concept of over-diagnosis has led
to a number of very interesting discussions in the
literature about, again, over-diagnosis and overtreatment.

This particular author actually suggests
that removing the word cancer,
because the word cancer creates a lot of stress not only
for the patients who are told they have a small cancer, but additionally the physicians who care for them
who feel compelled to act the concept of a small lesion
with indolent course, a papillary lesion
with indolent course, a click was one of the terms
that was suggested that if the patient
and the referring physician were told the patient has a
papillary lesion of indolent course by removing
that word cancer, they may feel more comfortable not to
refer every single patient with this diagnosis for surgery.

Certainly some small cancers do have an aggressive course.
This is the minority of lesions.

Here's about a five millimeter cancer
that has both metastatic lymphadenopathy
to the lateral cervical lymph node chains as well as
to the cervical lymph node chains.

This particular cancer,
although small in size, is acting aggressively
and everyone agrees this type
of micro carcinoma should be treated.

We do know that there is a very distinct linkage
between incidence of metastatic lymph nodes
and the likelihood for these cancers to be aggressive.

This was data that was published in the World Journal of Surgery, which looked at patients
with micro carcinomas that were proven surgically
to have negative lymph nodes in the blue as compared to
positive lymph nodes in the pink.

There were no deaths
and no distant metastasis in patients
who were lymph node negative and definitely morbidity
and mortality in those patients who were lymph node positive
for local recurrence.

There are a small number of patients
with lymph node negative disease at the time of surgery
who will have recurrence,
but about 20%
with lymph node positivity will have lymph node recurrence.

So evaluation of the lymph nodes
becomes very important when you know
or suspect that you're dealing with a small cancer.

The other thing that we would look
for would again be extra thyroidal extension.

This was a biopsy proven
and surgically proven small papillary thyroid cancer only
seven by five by seven millimeters.

But again, this was a small amount
of extra thyroidal extension that was proven at surgery.

This patient had surgery, was lymph node negative,
and now is disease free 10 years down the road.

But this type of finding would also spur you
to recommend a more aggressive approach.

So in summation, all these micro carcinomas are important.
A very small percentage of them are biologically aggressive,
and we look for that by local aggression
or invasion into tissues as well as metastatic disease.

If there is no evidence of these findings,
the patients prove to have a normal life expectancy.

And so you'll see a threshold of eight
or 10 millimeters for recommending fine aspiration in the
absence of these findings in almost all schema
that look at recommending biopsies.

Ultrasound Features Indicating Increased Risk of Malignancy

So we looked at some
of the features which are surefire signs of malignancy.

The rest of these features
that we're gonna talk about now are things
that have been associated with an increased risk
of malignancy, but not 100%.

So these will be the things that we'll be looking
for when we do our thyroid ultrasound to try to distinguish
between the vast majority
of benign nodules from the small percentage
that will be a malignancy.

Calcifications

Now calcifications are very important
and we can talk about calcifications in terms
of micro calcifications, macroscopic calcifications
and peripheral calcifications.

We know we're dealing with a calcification when we see
distal acoustic shadowing, such
as in this papillary thyroid cancer.

But another type of calcification
that's very common in thyroid cancer is so-called
micro calcification.

These are tiny calcifications,
less than a millimeter in size, which don't shadow.

And these are often seen in combination.
It's common to have multiple calcifications in the
same thyroid nodule.

Another type of calcification
that we're gonna talk about is peripheral
or eggshell calcification versus peripheral
interrupted calcification.

Most thyroid cancers that have a thin
and regular eggshell type
of calcification are benign, but not all.

Whereas those that have interrupted
or irregular calcifications are actually quite
suspicious for thyroid cancer.

Here we can see this nodule
that has some calcifications interrupted
and then continuing.

And this proved to be a follicular thyroid cancer.

This nodule has some regular calcification here,
but very irregular calcification.

And if you look carefully,
there's soft tissue disrupting the rim here.

This type of visible soft tissue extension through the rim
of calcification is actually highly concerning for a cancer,
usually a papillary thyroid cancer.

So we look very carefully at papillary thyroid cancers.

Just again, a couple of examples
of very irregular peripheral calcifications.

And then we look very carefully at the rim.

And if we can see any soft tissue extending beyond the rim
that's highly suspicious for a thyroid cancer.

This is a nodule here
where we can see there are some peripheral calcifications.

They're a little bit irregular,
but the key finding is that about five
to seven o'clock here, this expression of soft tissue coming right beyond the confines
that is highly concerning for papillary thyroid carcinoma
and actually provides a very nice site
for targeting your fine needle aspiration.

Echogenicity

Now, echogenicity of thyroid nodules is also very important.

And when we talk about echogenicity, we talk about the echogenicity
relative to the adjacent thyroid parenchyma.

Hypoechoic means darker than the adjacent
parenchyma, and we can describe that as mild, moderate,
or markedly hypoechoic
with markedly hypoechoic representing echogenicity
that is darker than the overlying strap muscles.

When we think about thyroid cancer,
most thyroid cancers are hypoechoic
and many are markedly hypoechoic,

but benign thyroid nodules being
so much more common than thyroid cancers an uncommon
presentation of a common thing outnumbers the cancers.

So if we just know
that a thyroid nodule is hypoechoic statistically,
it's still more likely to be benign such as this nodule.

But if we see hypoechogenicity, it raises concern
that it may be a thyroid cancer,
since most thyroid cancers are hypoechoic.

So what's going to happen is,
you'll see when we talk about biopsy, is that
we will acknowledge,
particularly when they're solid like these two will always
come to biopsy, but we will know
that definitely some will be benign
and some will be malignant.

Now what's interesting is one of the features
that we're gonna talk about when we talk about guidelines is
that combination
of features becomes the most important thing.

And when we start to add other features concerning
for thyroid cancer, the more features we add,
the more concerned we are.

So if we have a nodule that's solid
and hypoechoic, we're moderately suspicious,
but if we add another feature such
as marked hypoechogenicity calcification
or aggressive margins, then our concern
for cancer rises exponentially.

And we'll talk about that later on.

Margins

Evaluation of the margins
of a thyroid nodule is actually something
that is very important,
but really takes a high level of expertise
to do this very carefully and very thoroughly.

We can divide margin analysis into a number of features
that margins can be smooth
and defined throughout the entire thyroid nodule smooth,
but ill-defined in one area, or it can be ill-defined
and irregular.

And the irregular part is important here.
Jagged spiculated,
irregular margins are the concerning kind.

They can also be microlobulated or macrolobular.

One thing that's very interesting
and unique about thyroid nodules is
this category ill-defined.

For those of you who are familiar with breast ultrasound,
for example, you may know that an ill-defined margin
anywhere along a breast lesion is highly concerning,
and that is because we're trying to distinguish
between a benign breast tumor versus a malignant tumor.

Remember when we talked about thyroid nodules
and most of them are not tumors,
but just architectural distortion?

Well, because architectural distortion
commonly has an ill-defined margin, this feature
by itself is not important for thyroid nodules.

These features of aggressive irregular spiculated,
jagged microlobulated
and lobular in fact, are the ones that we concentrate on.

So just some examples,
spiculated margins in a papillary thyroid cancer here
as well as here, we can see these tentacle like extensions
into the soft tissue, very jagged and irregular margins in this anaplastic cancer,
which is growing right into the neck musculature, as well
as this papillary thyroid cancer here.

Notice how all of these are also solid hypoechoic,
and several of these also have micro calcifications as well.

So analysis of the margins is very operator dependent.

And when you look at the literature, the inter observer variability, it has been very high.

And the specificity for thyroid cancer, very variable again,
because I think it depends on how dedicated
an imager you are to analyze this sometimes
very tricky finding.

An example of a lobulated margin here,

and I should mention that all
of these diagrams have come from this
very excellent website.

It actually come from the breast literature talking about this.

But here we can see a thyroid nodule that's hypoechoic
and again, has this lobulated margin that's proved
to be a papillary thyroid cancer.

Here we have a macrolobulated margin,
and here we have some irregular margins, part
of which are tentacle like,
and also part of which are lobulated.

And again, for those of you that are the eagle eyes,
this is approaching
and very minimally disrupting the thyroid capsule.

And both of these were papillary thyroid carcinomas.

Other Features: Halo and Taller Than Wide

Now, people have talked about a dark rim
or a halo around the nodule.

People have described this as partial or complete.

The problem is that this finding can be seen in many tumors
such as adenomas.

It's lacking in most carcinomas,
but it's also lacking in most benign nodules.

It can also occur in some very longstanding benign nodules.

So it's really not a terribly helpful feature.

Here's a benign nodule that lacks a halo
and actually has a partially ill-defined border,
but is your very typical benign nodule.

This is a true tumor of the thyroid, a follicular adenoma.

And the halo here often represents compressed blood vessels
as this benign tumor is slowly growing within the gland.

A feature that came right from the breast imaging literature
was a taller than wide appearance,
and that means that on a transverse view the nodule,
it's higher or growing more in the AP dimension than in the
transverse dimension.

And this represents very aggressive growth.

This finding is something that is very specific
that is when you see it, it highly predicts a cancer
but not relatively sensitive.

This was a very interesting paper published in JAMA in 2015,
which exactly looked at this
and noted that this feature very much varies
with nodule size.

So if you take a very small cancer, zero
to five millimeters,
and you see this finding very specific
for thyroid cancer and relatively sensitive.

But if you look at thyroid cancers over a centimeter in
size, again, if you see the feature, very predictive
of malignancy, but you can see that the sensitivity,
meaning the number of cancers
that would show this feature drops dramatically.

Sensitivity and Specificity of Features

If we look at sensitivity
and specificity for a wide variety of features, again,
we can see microcalcifications very specific.

If you see them nine times out of 10, you're dealing
with a cancer but not very sensitive.

Only roughly 43 out of a hundred cancers are going
to show that feature.

Again, irregular margins, very operator sensitive,
and so all over the place here,
taller than wide, we just talked about.

And vascular flow within a nodule,
which I'll talk about in just a minute.

Also all over the map, the pattern of features
as we've mentioned before, helps us.

And that is when you start to combine one
or two features that are highly specific for thyroid cancer,
your specificity goes up,
but your sensitivity drops, meaning less
and less cancers fit that pattern.

And so this is why we can't just screen for thyroid cancer
by the presence
or absence of these particular features.

Yet we can certainly use ultrasound to pick nodules
that are very likely
to be a cancer when they have these highly specific features
alone or in combination depending at what the case may be.

The reason why we miss many thyroid cancers
by these techniques is that about 20 to 30%
of thyroid cancers have a follicular element,
whether they are that mixed papillary and
follicular type of cancer
or a pure follicular cancer, they look slightly different.

They are much more commonly iso
or hyperechoic to the thyroid.

They may have calcifications that are peripheral,
they tend not to have microcalcifications, and they may
or they may not have a halo.

And this appearance can overlap with adenomas
that we just looked at.

So this is one of the things where we can say
that thyroid ultrasound is very specific
but not very sensitive
for picking all thyroid cancers out of the crowd.

A very nice summation
of this was shown in these two articles,
which compared pure papillary thyroid cancer in orange
versus mixed papillary and thyroid cancer.

In the green, for example,
microcalcifications are present in about 40%
of pure papillary cancers,
but almost always absent in thyroid cancers
that have follicular histology being solid.

Certainly it's a concern for both
and irregular margins for both.

And again, the presence and the absence of the halo.

Many cancers in the mixed follicular type will in
fact have a halo.

Intranodular Flow

Now, intranodular flow is very interesting,
and again, as we talked about
before, many benign nodules, both true tumors,
adenomas as well
as hyperplastic nodules will have internal flow.

Most cancers, the typical pattern would be central flow.

But again, since benign nodules are
so much more common than malignant nodules,
just knowing there's vascular flow
as an isolated feature is not helpful in making the
distinction between benign and malignant.

In Korea where the vast majority
of thyroid cancers are pure papillary, they looked at this
as an isolated feature.

And what this group did is that said
that if you take away nodules that have any
of these features of malignancy such as marked hypoechogenicity, aggressive margins, microcalcifications
and taller than wide shape,
and you just look at vascularity
as an independent predictor,
it was not predictive at all in distinguishing
benign and malignant.

But in Korea, as I mentioned, the vast majority
of cancers are papillary thyroid cancers.

If you look at a number of papers
that have been published in other countries,
most commonly Europe
and in Italy, they have found vascularity
to be a helpful feature.

And this gets back to histology.

In Europe, it's very common to have follicular cancers up
to 25% of the surgical series.

So follicular cancers do tend to be vascular
as opposed to Korea.

Where follicular cancer is incredibly rare,
it's not a helpful feature.

The United States is somewhere in between with a
rate of follicular cancer
of approximately five to 10%.

Benign Ultrasound Features

So the literature has commented on all
of these features which are highly predictive of malignancy
that spend a little bit of time on features which are
predictive of a benign nodule.

And at the foreground of this is a nodule
that appears almost entirely cystic to us as long
as we see no solid elements with the exception
of these bright echoes,
which is actually an ultrasound artifact caused
by the colloid within these nodules,
a so-called comet tail artifact.

These nodules are considered absolutely benign
by all investigators.

There's another pattern that is not always benign,
but very highly predictive of benign histology,
and that is the so-called spongiform nodule.

This is defined as interlaced cystic spaces in a nodule
with no atypical features.

It's allowed to have some vascularity
because remember, our model here
of the hyperplastic nodule is
that there is some thyroid architecture in here,
but it does not tend to be very vascular.

Note how this is sort of a spongiform appearance
and there's no mural nodules.

And typically these nodules lack calcification,
and the vast majority are small under two cm.

That pattern has a very high prediction
of being benign.

And in the Moon series, only one of 360 cancers had this appearance.

So it's not a 0% chance of cancer, but far less than 1%.

And so these are typically thought of as highly likely to be benign.

Another way to think about this is Bayes Theorem,
which is a predictive model,
and it basically says that in a given group, we know
that about seven to 9% are cancers.

So for every thousand nodules
that lack those very suspicious features, we know
that less than 1% will be a cancer.

And remember, some of this data comes from papers in which they were heavily weighted towards
papillary thyroid carcinomas.

Comet Tail Artifacts vs. Microcalcifications

Now, something that I see quite commonly is people see
bright hyperechoic dots or lines in a thyroid nodule
and want to call these microcalcifications.

We have to be very careful about over calling
microcalcifications because true microcalcifications carry a
very high weight in terms of risk of malignancy.

So instead what we wanna look for is to see if we can tell
for sure that these are comet tail artifacts.

This is a cartoon that I'm going
to show you the data from the paper in just a second.

But in these small mixed cystic
and solid nodules, we wanna look for this inverted triangle,
the inverted triangle.

It's highly predictive
that these are benign colloid containing
hyperplastic nodules.

We now have a new language
for these non shadowing echogenic foci.

And when we see the inverted comet tail, we wanna call
that out because that is highly likely to be benign.

But these other non-shadowing echogenic foci
or hyperechoic foci should be distinguished.

The long and linear ones are almost always benign.

These often represent the back wall
of cystic spaces within thyroid nodules.

These very punctate fine ones should be called
microcalcifications, and they are the ones
that are commonly associated with malignancy.

In this particular paper, you could see that nine out
of 13 were benign and four were malignant,
but these were mixed cystic and solid nodules.

Punctate echogenic foci in solid hypoechoic nodules are
almost always micro calcifications.

When nodules are of mixed echogenicity, it may be hard to tell
and often a biopsy is warranted.

But we don't wanna over call either
of these micro calcifications without a
further sub-analysis.

When these punctate dots become a little bit larger,
this may be a little bit more of a challenge.

And in this particular paper here, one
that had this pattern were thought to be a cancer.

Cystic Papillary Thyroid Cancers

Now the concern is that we're going
to miss a cystic papillary thyroid cancer.

Fortunately, they're relatively rare.

They comprise less than 6% of all papillary thyroid cancers,
and I think they have a relatively specific appearance,
thyroid cancers that are predominantly cystic.

The solid part is the key to the analysis.

The solid part, instead
of looking like the hyperplastic nodule that we looked
before, tends to look more frontlike.

It tends to have shaggy irregular borders as opposed
to geographic sharp borders.

It tends to have an eccentric component bulging into the
center of the nodule as opposed
to a geographic configuration.

And it tends to have increased blood flow
for its small size.

And so papillary cystic papillary thyroid
cancers are uncommon.

But those are some of the features in a very excellent paper
by Kim and colleagues in 2010, they did the sub-analysis
and painted the picture for us of assigning risk
of cancer based on those features.

The lowest risk is when we have concentric soft tissue
that lacks any suspicious features
and then coming to the highest risk
where we have a big mural nodule with calcification
and with irregular edges.

And so it's the solid part of mixed cystic
and solid nodules that should be analyzed.

And again, just showing you some cystic papillary thyroid
cancers, every cancer that I have seen has had one
of these features, whether it be frontlike elements,
exuberant vascularity and
or calcifications here
with shadowing in those types of cancers.

Summary of Ultrasound Features

So in summary, basically we've looked at a number
of features and we know that no individual
feature is sensitive enough
to identify all malignant nodules.

But on the other side of the coin, we know
that there are certainly absence of features which
confer a very high negative predictive value
for underlying malignancy.

Guidelines for Biopsy in Low-Risk Patients

So when we talk about guidelines, which we're getting
to in just a moment, remember we
as imagers may not have the full picture
on the medical side.

The endocrinologist may say the patient is not at low risk,
and therefore the guidelines that we're about
to talk about don't apply to them.

And several things may put them at increased risk
such as genetic syndrome such as Cowden syndrome,
familial adenomatous polyposis,
and the MEN syndrome, strong family history,
which is defined as two
or more family members nodule detected in a patient
of young age, a particular personal history of head
and neck radiation as a child,
particularly children who are treated
with whole body irradiation or exposure to nuclear disasters
or a few things, or knowing that the nodule was detected
as being positive on a PET scan confers about a 35% risk.

So some of these things may pull the patient out of general guidelines.

SRU Guidelines (2005)

The first attempt at coming out
with general guidelines were tackled by the SRU in 2005,
and at that time, endocrinologists were recommending biopsy
for all nodules larger than 10 millimeters regardless of
how they looked on ultrasound.

And this was the first attempt to sort
of marry the ultrasound appearance
with the risk for malignancy.

So hopefully, based on the knowledge
that we've built through this lecture,
you can understand the rationale microcalcifications one
of our strongest predictors if the nodule was 10 millimeters
or larger so
that we were not over biopsying microcalcifications.

But you'll note down here the guidelines definitely said
apply clinical judgment.

So if it was a small cancer that was aggressive
with metastatic lymphadenopathy, for example,
certainly biopsy would be recommended if it was a two
millimeter micro carcinoma that,
and the patient had no comorbidities and
or had no evidence
of metastatic disease biopsy was not always indicated solid
and coarse calcifications at 15 millimeters
and consider biopsy if it was a mixed cystic solid nodule
or was the type of cystic nodule that had a mural nodule.

And then we'll talk about growth,
but certainly if substantial growth was noted over time
to consider biopsy.

So the guidelines have evolved quite a bit from there.

ATA Guidelines

And the one that is most commonly in use now was
published by the ATA.

And for those of you that are interested,
you can download it at this website.

And coming very soon will be the American College
of Radiology's version of TI-RADS, which is not yet released.

So I can't discuss it in full, hopefully next time.

I can give a supplement
and I can give you all the rules,
but it's just about to be published.

The very first attempt at TI-RADS was back in 2009,
and this is how it was presented.

And again, you can see sort
of this pattern approach was applied.

The problem with this particular guideline was the very
vast risk
of malignancy in this four B group.

And so many decision making had
to be done here without a lot of further guidance.

KK and colleagues revised this a bit
and sort of firmed up the ultrasound features
and in essence sort of had an accounting
that if you had one ultrasound feature,
you were at this level, two, three to four and so forth.

But there was no weighting to any
of these particular features at all.

And, but it did help refine the risk of malignancy.

The ATA guidelines used a variety of things that
were revolutionary,
not only in aligning sonographic pattern to the risk
of malignancy and assigning a specific size threshold,
but also came forward with some recommendations for,
and I think just as importantly, lack of follow up
for certain ultrasound appearances.

And they also gave the strength of the evidence
to support each recommendation.

There is evidence that's lacking for some of these,
and they clearly say
that those decisions are based on expert opinion.

And as we move forward
and people are collecting prospective data on nodules,
which have been followed for years
and years, these guidelines certainly
will be able to change.

So this is an overarching look at the atlas
that the ATA has put forth, which is in that document.

But we're gonna break this down into some specific risk
categories and look at the features.

So if we look at that high suspicion category, which is 70
to 90% risk, these are the features
that we've just talked about.

These include microcalcifications, solid consistency,
hypoechogenicity, and aggressive margins.

Of course, whenever we see metastatic lymphadenopathy,
that right away tells us we're dealing with a cancer
and then types of calcifications with
that interrupted peripheral rim.

So these are all the features that we've talked about,
and you will notice that for each of these nodules,
there's usually a minimum of two
or three features in combination that makes
that nodule high suspicion.

The intermediate suspicion would be those hypoechoic
and solid nodules that lack additional features such
as micro calcification.

Aggressive margins are taller than wide shape.

Most of these will be benign,
but about 10 to 20% of these will turn out
to be a thyroid cancer in low suspicion.

We have nodules that are solid and are either isoechoic
or hyperechoic to the surrounding parenchyma,
but lack additional features such as calcifications
or irregular margins.

The vast majority of these will be benign,
many will be benign adenomas,
but in the United States, about five
to 10% will be those follicular
cancers that we talked about before.

Partially cystic nodules are those that have this eccentric solid component would be in this low suspicion category,
and very low would be our spongiform hyperplastic
nodules that we talked about.

And outright benign would be an entirely cystic appearing
nodule with the risk of cancer well under 1%.

And we covered that
before only the comet tail
reverberation artifact is seen.

They then took these suspicions
and aligned them with a size.

So high suspicion
and intermediate suspicion should be
biopsied at a centimeter.

The strength of the recommendation, you'll notice strong,
but the quality of the evidence actually low for these.

There's not a lot of controlled
trials that have looked at these.

If you are somebody who's at the low suspicion about 1.5
centimeters is when the biopsy is recommended.

And again, low here on both the strength of recommendation
and the quality of evidence.

If you are a very low suspicion nodule,
these would be those spongiform appearing nodules.

Note that either at two centimeters you can biopsy
or you can recommend ultrasound follow-up
because the risk of malignancy is quite low.

And if you are a benign nodule that is entirely cystic,
no biopsy is needed at all.

They do note that FNA is typically not recommended
for any nodules under a centimeter in size for any
of these in the low risk patient.

But again, a high risk patient
with a nodule under a centimeter may not necessarily fall
into these guidelines when we have multiple nodules,

what you should do is look through and see if any of the nodules fall
to a higher suspicion.

If any of the nodules do that,
you should biopsy regardless of size.

But if all of the nodules are similar
and they all fall into the low
or very low suspicion categories such
as a true multinodular goiter, we know the risk of
malignancy is very low.

And either the largest nodule bigger than two cm should be biopsied or the patient should undergo
ultrasound surveillance.

TI-RADS Overview

Now what about nodules that have undergone FNA
and you're going to follow them with ultrasound,
the strategy of the follow-up
and the timing of the follow-up should be based
on the sonographic pattern.

This is a brand new concept that was first introduced here
by the ATA, noting
that all those nodules smaller than a centimeter
with very low and entirely cystic appearing nodules require
no ultrasound follow up.

So if you have a nodule that's very low suspicion,
if it's less than one centimeter,
it does not need follow up If it's larger than one
centimeter, they actually say that the utility
of ultrasound follow up is not known.

They could neither recommend
that you do it or you don't do it.

But they did say if you choose to do it,
the interval should be a minimum of 24 months.

For those that are in the intermediate
or low suspicion, they do recommend a follow-up in
approximately 12 to 24 months.

And if it's a high suspicion nodule
and it did not meet criteria, for example,
an eight millimeter nodule that was calcified and hypoechoic,
they recommended a shorter interval follow-up
of six to 12 months.

Again, noting that there's not a lot of evidence
to base these recommendations.

Okay, one of the problems with the ATA categorization is there's a fair number
of nodules which have not been fully characterized
by the atlas that they have laid out,
and they are already looking to revise that
and make it more inclusive moving forward.

Just some examples of nodules
that I've encountered in my practice,
which are not encompassed by
the ATA would be this mixed cystic
and solid nodule where there are hyperechoic foci.

In the solid part, this was a cystic papillary thyroid
carcinoma, a nodule here
that has coarse central calcifications,
a small eggshell nodule,
and then this sort of puckered looking lobulated marginated nodule.

This was a mixed papillary and follicular cancer.

So the ATA does not characterize all nodules.

Follow-Up Strategies and Growth Considerations

Many of you may be thinking about why are we doing follow-up
on nodules that have had a previous biopsy.

And the reason for that is that FNA even done under the best of circumstances has a small
but definite false negative rate, generally quoted
to be one to 2%.

And there's a variety of reasons for these,
including heterogeneity of the nodule
and just nuance in the cytologic interpretation of these.

One of the things that has been suggested
as a follow-up has been
to look at growth under the assumption
that growth would be a very good discriminator
between benign and malignant.

And I'll talk about how that's not the case.

We certainly know that most benign nodules grow,
but they tend to grow slowly.

So an accelerated pattern
of growth has been a concern for a malignancy.

The problem has been how to define significant growth,
and there's a number of definitions out there,
but there is no consensus.

The ATA in their guideline uses a 50% volume growth
with at least a growth of two millimeters in more than one dimension,
such that these smaller nodules are not unduly weighted.

And most importantly when you analyze a nodule is if the
solid component is growing in the mixed cystic
and solid nodule, it's much more
concerning than just if the fluid component is growing previously, all nodules
with growth were recommended to
be re-biopsied sort of endlessly.

And there was an endless follow up pattern,
but some of those things are changing as well.

One paper in the literature found a 10% false negative rate,
but this was done by palpation.

All of these papers, which were very large papers done
with ultrasound guidance, have generally shown between a one
to 2% false negative rate.

And the most important takeaway from those three papers has
been the following of 184 nodules, which were followed.

And at the end of the follow-up period,
which was a mean of seven years, they in fact found that
that nodule was a malignancy.

They had 77 of those such patients and they went back
and looked at whether the malignancy would have been
predicted by growth
or whether malignancy would have been predicted
by ultrasound appearance of the 17
for whom they had surgical pathologic follow-up.

Of the 77 17, I should say, had growth greater than 50%
and 16 were benign and one was malignant.

And this was a 2.2 centimeter follicular variant
of the nodule, which proved to be a cancer that did not have greater than a 50% growth.

Nine were malignant.

And it turns out that eight out
of the 10 would have been predicted
by ultrasound features despite the lack of growth.

And each of these papers has consistently shown
that it's the ultrasound features much more importantly than
the growth of the nodule that have been the predictor.

They all had the same conclusion.

And lack of growth cannot be reassuring of benign.

So if we look at the ATA guidelines in terms
of the follow up of patients with known cytology,
they recommend following those patients
and that the risk for determining if the patient needs a
repeat biopsy should be entirely based on
sonographic pattern.

So if you have a high suspicion nodule
and a negative biopsy,
they actually recommend not only repeating the ultrasound,
but also repeating the FNA
within a year.

And there's a nice body of evidence as we've just discussed, to support that

if it's an intermediate or low suspicion nodule
and you have a benign biopsy, the quality
of evidence here is quite weak,
but they recommend that you reassess that nodule
for growth again, greater than 50%.

Or if the nodule is developing new suspicious features such
as suspicious calcification,
it should be biopsied in the very low suspicion category.

There's not a solid recommendation as to whether
that particular nodule needs a repeat ultrasound at all
or a repeat biopsy.

They do say if you elect to re-biopsy a nodule in any
of these categories and you now have two benign biopsies in
hand, that can be considered strong enough
to consider the nodule overtly benign.

Future Directions

So moving forward, what should we look for?

Again, a revision of the ATA guidelines, since about 20%
of all nodules are not characterizable, specifically those
with macroscopic calcification
and mixed cystic solid nodules.

A reevaluation of the need for follow up of all
of the nodules that are a very low suspicion
and over a centimeter in size and potentially an age cutoff.

For example, there is an age cutoff in colonoscopy
screening above a certain age.

Also look for the ACR TI-RADS to be released.

The concept of this, that there will be points assigned
to each nodule feature with higher number of points assigned
to more suspicious features such as micro calcification
and lower or no points assigned
to non suspicious features such as spongiform appearance.

TI-RADS will break nodules into categories one through five,
and the goal is to categorize every possible nodule into
a TI-RADS category.

Each TI-RADS category will then have a specific size threshold for FNA
and a tiered follow up recommendation very similar to
how the ATA has strategized their follow up.

Thank you very much.