Thyroid Fine Needle Aspiration (FNA) and Cytology - HD

Introduction

Thank you so much for inviting me.

I'll apologize in advance.

This is not a very image rich lecture.

I'm gonna cover something that I think is very straightforward and that is thyroid fine needle aspiration.

A very nice and easy technique and something that I think is very complicated and that's the understanding of cytology and what it means for patient management.

I have no relevant financial relationships.

Technique for Thyroid Fine Needle Aspiration

Just like to highlight a couple things about technique in our lab, we do this as low key as possible to minimize patient anxiety.

We just use aseptic technique and clean the skin with chlorhexidine liquid Hiba cleanse, it's a wonderful antiseptic soap and a wonderful ultrasound coupling agent.

So there's actually no need for gel.

Very rarely do we give anesthesia.

We generally use 27 gauge needles and it's been proven that, the finer the needle, generally the, the better the specimen.

Occasionally you may have a non-diagnostic aspirate.

We have the benefit of onsite cytology or a very, a very fibrotic nodule, in which case we may go to a 25 or a 23 gauge needle.

But the specimen is all about how you're moving the needle in the thyroid, nodule per se.

And you wanna really move this very briskly.

It's the cutting bevel of the needle that's acquiring the specimen.

We're not truly aspirating.

What we do is pull the plunger back in the syringe, which is just a handle to about one cc, and that gives a little bit of capillary action and brings the material into the needle.

The key thing is targeting suspicious areas, particularly calcified and solid mural nodules.

And if you have larger nodules, nodules larger than four cm, making sure you sample all different parts of the nodules to avoid a, a sampling error bias.

And here's just a predominantly cystic nodule with a small solid expressions, you can beautifully target that solid area.

And this was a cystic papillary thyroid carcinoma.

Dwell time. How long do you stay in about two to five seconds.

You should move your hand back and forth at about three oscillations per second.

And we very, very rarely use core biopsy in our lab, but others have some success with that.

Cytology and Molecular Marker Analysis

I'd like to move on to the cytology and ultimately molecular marker analysis, which I think is quite nuanced.

And as you'll see ever changing with some new things that literally happened last week at the American Thyroid Association meeting.

Pre-2008 Cytology Diagnosis

So prior to 2008, the cytologist basically could render a diagnosis of non-diagnostic, not enough cells, and notice how the risk of cancer is not zero.

This is not a benign aspirate. This is non-diagnostic.

And usually you repeat this.

Benign means benign in the vast majority of cases, but there's a small false negative rate.

Malignant should be malignant in nearly all cases.

But this middle ground, the so-called indeterminate biopsy, was quite a dilemma for managing the patient with reported incidents of cancer all over the place, 15 to 75%.

Not very helpful. Patients would typically be referred for surgery, what's been called a diagnostic lobectomy, in essence, surgery to remove that indeterminate nodule with the vast majority of those being benign.

So clearly you're exposing these patients to the risk and cost and so forth of surgery only to wind up with a benign diagnosis in the vast majority.

But certain patients were being referred for, referred for a total thyroidectomy in our institution as our cytologists.

Were noticing that that in that that indeterminate category probably could be further broken down.

And in fact, one of our pathologists that our institution led the the Bethesda classification, which was adopted in 2008 to do just that.

Bethesda Classification System

So Bethesda one is non-diagnostic, two is benign and six is malignant.

We've discussed those, but they added these categories.

Bethesda five suspicious for malignancy, Bethesda, three pH follicular lesion of uncertain significance.

And then Bethesda four follicular neoplasm.

Suspicious for malignancy means there's features very worrisome for papillary thyroid carcinoma, but not outright.

Three means there's more atypia than you usually see.

Most of these will be benign.

A handful will be a cancer and most will be a follicular variant of papillary cancer.

And Bethesda four monotonous sheets of follicular cells.

Most of these are benign adenomas.

So the beauty of this is it now aligns with a very much more precise risk of malignancy based on these classifications in order to guide management.

Management Based on Cytology Results

Putting this into play, as I mentioned, non-diagnostic, you repeat the biopsy benign you observe, again, pointing out the risk of cancer is not zero, but it's low and malignancy, you do a total thyroidectomy.

And so we're left with these, again, indeterminate Bethesda three through through five with these newly assigned risk of cancer.

What to do in our institution. We repeat flus times one.

This is not the practice everywhere.

And let me show you why this is data from Vander Linein, but our institutional experience is very similar.

Plus on the first biopsy, the patient comes back about four to six weeks later, we repeat the biopsy almost half are now benign.

We handle those as benign and we just follow them.

About 28% will come back flus the second time.

And interestingly in their series and in our series, this is a pretty high risk of cancer.

Many of these are follicular variants of papillary cancer.

And so we think about flus times two differently than flus times one.

So back to that grid.

Here we have flus times one flus times two, follicular neoplasm and suspicious for malignancy.

We're a little stuck here because now we have to recommend observation and surgery and if we recommend surgery, half the thyroid or the entire thyroid.

So there was a need to go beyond the cytology.

There can be tests that can be added, which can be considered either as a rule out test or a rule in test.

A rule out test is a test where you wanna better predict a benign outcome for your patient because you're going to suggest observation and not surgery.

A rule in test is you know, you're recommending the patient have surgery, but wouldn't it be nice to know if you should refer them for a hemithyroidectomy or a total thyroidectomy?

Molecular Testing Overview

So enter molecular testing, molecular testing comes in a variety of a different, very innovative, methods including gene expression based gene mutation based oncogenes micro RNA or a combination.

And let me just highlight a few.

Affirma Gene Expression Classifier

A test that we have very commonly used, A so-called affirma.

Affirma is a gene expression classifier, which basically takes the gene, the material from the thyroid nodule and looks to see what genes are being expressed.

Failure to express genes associated with with cancer.

In cytologically, indeterminate nodules has a very high negative predictive value.

So if you are thinking that this particular patient would be an observation candidate, a negative gene expression classifier adds a lower, confers, a lower risk of malignancy when it's negative.

And reaffirms this, this plan of observing the patient.

This was the landmark paper that was published in the New England Journal of Medicine by Eric Alexander out of the Brigham.

And they showed for a TIA plus Bethesda three and Bethesda four, A very high negative predictive value with the gene expression classifier.

But notice that the negative predictive value is not high enough in this suspicious for malignancy category to avoid surgery.

So a very good rule out test, this was put into play, this is a paper published by Dan Deic who's in an endocrinologist, and he looked at the impact from his practice upon recommending surgery versus recommending follow up for patients who had a benign affirma.

Prior to that 74 of his 368 patients would've been recommended for surgery.

But now he was only sending eight patients with a benign affirma.

And you may be thinking, why is he sending these eight patients?

And that's because Dan is a super smart guy and a great ultrasound uh, operator.

And there may be features that made him refer patients to surgery even with a negative affirma.

And I'll highlight that in just a moment.

Now, there's a big misconception about what does a suspicious GEC mean the re, the results come back as benign.

No genes expressed or su or suspicious quote genes expressed.

Well, if we look at the best 11 papers that have looked at this and the risk of malignancy with a positive or a suspicious test, the risk of malignancy of patients that went, excuse me, the rate of malignancy these patients were all operated on based on a positive test is all over the map.

The New England Journal actual incidence is low 38% and 37% and can be as low as 15%.

So you wanna remember that a suspicious firma is not suspicious.

Cytology, you have the cytology in hand, usually flus or fn, and now you know that some genes are expressed, but still the vast majority of these are not cancerous.

That's a common misconception. And why is this?

Well, remember that these genetic tests are modifiers.

The cytology drives this.

And then within a cytologic category, there are people who have a higher risk and lower risk.

And that is influenced by a variety of things, including your cytologist at your institution, which are giving you the feedback on site about which tests to do the cancer rate in your specific population of patients, which may be quite different than that New England Journal article.

I've heard clinicians say it's always a 5% negative predictive value, and that's not true.

That was true for the whole cohort at the New England Journal with their cancer prevalence and their cytologists.

And the one thing I'd like to highlight specifically is ultrasound appearance.

So both of these nodules came back with a Bethesda three or plus cytology result.

This one on the top, solid hypoechoic, calcified, funny margins, and this one on the bottom fairly innocuous looking.

Do these nodules have the same risk of cancer?

Of course not. The nodule on top has a very high risk of cancer.

And as a matter of fact, a number of studies show that if you have just one suspicious feature, your risk of cancer with a plus cytology is over 70%.

Again, many of these are follicular variant of papillary thyroid carcinoma.

So this particular patient's risk is well over 70%.

This patient is a surgical patient.

Even if that affirma came back negative, this patient is still surgery.

There's no role for a rule out test, we've already decided to do the surgery.

Let's make pretend this is a hundred year old patient with comorbidities and a very innocuous looking nodule, a very good candidate to lower that patient's specific risk.

So in our particular lab, this is how we view it, that if it is a GEC negative nodule that's a low risk patient, no suspicious findings, a smaller nodule, and the patient is willing to undergo observation firma.

A rule out test is a very good test for those types of nodules.

Gene Mutation Based Tests

Let's talk a little bit about the gene mutation based test.

And these are the ones that are gaining traction.

These can be oncogene panels and now often in combination with the RNA component.

And I'll do my best to explain that a little bit as well.

These tests first splashed on the scene in about 2010 and the genetics of thyroid cancer is known very well.

A number of oncogenes and gene mutations have been identified that are found in thyroid carcinoma.

And I've listed the vast majority of them here.

What's interesting though is that these original tests, first generation had relatively low sensitivity, and I'll explain that and only moderate specificity.

The sensitivity part can be thought of as follows, that BRAF mutation, which is seen in papillary thyroid cancers, if you know you have a BRAF F mutation, you know it, you have a cancer.

But remember, we're not dealing with positive cytology, we're dealing with indeterminate cytology.

Many of these are either non-cancerous or they're follicular variant of papillary cancer, which do not tend to have the BAF mutation.

They more commonly have KRAS, nras and so forth.

So the incidence of having a BAF mutation in nodules that have indeterminate cytology is very low.

Additionally, we have overlap and that is some of these oncogenes are expressed both in adenomas and in follicular type cancers, the NRAS, the KRAS and so forth.

So it's a very, very complicated, very, very complicated background here.

The two most common oncogene panels that are out there now, start with the thyroid genex that we talked about, but interspace has added the micro r uh, MR NA as well.

And this is called thyroid, MIR.

Um, if both of these tests are negative, that negative predictive value is very high, but only a moderate positive predictive value.

So a very strong rule out test, moderately strong, strong rule in test thse version two, and this is important as I'll explain in a, in a moment, comes out of Nik AFOs lab at University of Pittsburgh, who's one of the leaders in this field.

And uh, this particular test in their population at Pittsburgh had wonderful sensitivity and specificity, but when it was tested in other populations that had a lower risk and excuse me, a lower rate of cancer in the cohort, it did a little bit less well, but the negative predictive value held up.

This was a paper that looked at the difference between Moffitt Cancer Institute and UPMC and you can see the black here is PIT and their population.

The positive predictive value for cancer when that test was positive was relatively high, but among flu lesions at Moffitt, not a very high positive predictive value.

However, hot off the press right from the A TA presented last week is the latest version of thi aeq soon to be released.

This will be thi Aeq version three.

So we all, now, if those of you who are using it are using version two that has this profile, but as the test gets better and better over time, the positive predictive value will be higher.

There was just a multinational, a multi-institutional study that was presented there and should be published soon.

So I believe that moving forward virus EQ three will have both a very high negative predictive value and a much higher positive predictive value and can serve as both a rule in and a rule out test.

So molecular testing is very complicated, but it's very useful in certain scenarios.

It's useful if you want to prescribe observation to a patient and you want to use the negative predictive value to do that.

If the surgery would be altered by the test, then it's worthwhile to do it.

But if your surgeon already knows they're going to do a complete thyroidectomy, it may not be worth the expense to do it.

NIFTP and Tailored Therapy

And let me finish up with either a further complication.

Not all cancer is created equal.

You may have heard of this entity called the nift PA non-invasive follicular thyroid neoplasm with papillary like nuclear features.

In essence, it is a cancer because the nuclear features say it's a cancer, but its biological behavior is very non-aggressive, much more like an adenoma.

Now some of the surgeons are very satisfied to only do a hemithyroidectomy and leave it at that.

You're gonna get a positive cytology, they'll do the HEMI there may not go back to do the completion.

So this is hot off the, the press of the THI EQ website that you can look at.

And now we can actually offer very tailored therapy based on the results.

So no mutations, the risk of cancer is low, very close to benign cytology.

If it's positive, but it's only RAF positive, it could be a cancer, but it could be a non-aggressive cancer, A lobectomy would be fine.

And if it's BRAF positive or has multiple mutations, it's probably a cancer and you can actually see with multiple, it's probably a high risk cancer and maybe that surgeon would in addition to do the, the total thyroidectomy do a prophylactic neck resection.

Conclusion

So I think it's a very complicated topic that's obviously evolving with the A TA, just presenting this data last week.

But I hope that my synopsis has been helpful. Thank you.