The Riddle of Thyroid Nodules

Thyroid nodules are really, really common—the Starbucks of our necks. It is estimated up to 75% of the population may have at least one thyroid nodule, and most of us will never know we have them. Why are we even talking about them? It turns out somewhere between 7-15% of these nodules are actually thyroid cancers.

Given this information, you might think the case of thyroid nodules should be open and shut; if you find a nodule that is cancerous, just remove it. Oh, how I wish it were that simple. Gray areas and controversies abound in the management of thyroid nodules, and their prevalence in the population exacerbates the associated costs and headaches.

Welcome to the world of thyroid nodules.

Lumpy bumpy thyroid

Thyroid nodules reside in (surprise!) the thyroid gland, a small butterfly-shaped organ which sits in the front of the neck. Medical illustrations of the thyroid—draped around the upper part of the trachea (windpipe), just below the adam’s apple—remind me of the “facehugger” from the movie Alien. But rather than implanting a horrific beast that subsequently ruptures from your chest, the thyroid secretes hormones that affect metabolism throughout the body.

A thyroid nodule is simply a lump which appears separate from the rest of the thyroid gland on imaging. Occasionally these nodules may be palpable (you can feel it), but more frequently they are discovered incidentally when your neck is imaged for some other purpose, e.g. MRI for neck pain, or chest CT for pneumonia. Radiologists cutely call these surprise lesions “incidentalomas” because they are incidentally discovered. Get it? They happen all over the body, but thyroid nodules are probably the most common type.

We’ve got a problem

Once a thyroid nodule has been discovered, it should be evaluated on thyroid ultrasound. Ultrasound is considered the gold standard for visualizing thyroid nodules. The thyroid gland lies just below the skin surface, and we can see even the tiniest nodules.

Sample longitudinal (up/down) ultrasound image of the thyroid. Red arrows indicate the normal thyroid gland, and the blue arrow localizes a nodule at the lower end of the thyroid.

So what’s the problem? Despite some clues that ultrasound gives us (to be discussed below), we still cannot say with certainty if a thyroid nodule is cancerous based on ultrasound alone. In practice, this translates into tons of thyroid nodules in which the risk of cancer is uncertain.

To further complicate matters, many small thyroid cancers may be indolent, i.e., they will never spread beyond the thyroid and cause death of the patient. One study found 36% of individuals had (previously unknown) thyroid cancers discovered at autopsy, meaning they died with thyroid cancer but not from thyroid cancer. Another found that thyroid cancer rates tripled between 1975 and 2009, but mortality from thyroid cancer was the same over that time period. In other words, we are finding and removing more thyroid cancers, but it’s doing nothing to improve survival rates from the disease.

To summarize the problem: We can’t tell if some thyroid nodules are cancer, but in some cases it might not even matter!

Oh, brother

How do doctors deal with these issues? In short: it’s complicated. The rest of this post is dedicated to walking you through the current approach to thyroid nodule management. As we wade into the swamp, keep in mind that our primary aims are to determine if these thyroid nodules are malignant (cancer), and to decide on a treatment option.

Question 1: Is the nodule hot?

An accepted first step in evaluating a newly-discovered thyroid nodule is to determine if this nodule is “hot.” If you are giving the screen a weird look right now, allow me to explain.

Because of genetic mutation, some thyroid nodules secrete an increased amount of hormone compared to the rest of the thyroid gland. These hyperfunctioning nodules are much less likely to be cancer, and are treated differently than other thyroid nodules. When a thyroid nodule is first discovered, blood tests are obtained to see if thyroid function is out of whack, suggesting a hyperfunctioning nodule may be present. If the bloodwork looks off, a special imaging test called a radionuclide thyroid scan is performed, which can help distinguish between a hyperfunctioning nodule and other causes of increased thyroid function. A hyperfunctioning nodule will appear “brighter” on this type of scan, and we call this a “hot” nodule.

normal thyroid and hot nodule
Radionuclide thyroid scan illustrating a hot nodule. Left image: Normal butterfly-shaped thyroid gland. Note the diffuse, uniform appearance. Right image: Hyperfunctioning nodule stands out due to accumulation of more radioactivity, referred to as increased uptake or increased activity.

Hyperfunctioning nodules are treated differently than other thyroid nodules. Rather than biopsy or surgery, they are killed with radioactivity. That may sound alarming, but it’s actually quite boring. A pill containing radioactive iodine—an isotope called iodine 131—does the trick. Iodine is normally used by the thyroid to produce hormone, and is preferentially taken up by a hyperfunctioning nodule; good for you, bad for the nodule.

As we move on, remember that the rest of this post refers to thyroid nodules which are not hot (although I’m sure they have a great personality).

Question 2: How big is the nodule? 

In addition to blood tests, all thyroid nodules get an ultrasound. Many of these nodules are first discovered incidentally on CT or MRI (anecdotally, I can tell you this is uber-common). But no matter how they were discovered, nodules should be characterized on ultrasound.

When it comes to thyroid nodules, size matters. Only nodules that are greater than 1 cm should be further evaluated, meaning those less than 1cm can go on their merry way.

Succinct and simple, right? Wrong. Some experts recommend nodules up to 1.5 cm in size can be ignored if the patient is older than age 35. Additionally, if a nodule less than 1cm is causing symptoms, or there are abnormal lymph nodes in the neck, then an ultrasound should be performed. In medicine, there are always exceptions to the rule.

Question 3 (aka the million dollar question): Does it need to be biopsied?

Despite some wrinkles in regard to nodule size, at this point the situation is fairly straightforward: we found a thyroid nodule, did some blood work, and performed an ultrasound (if the nodule was larger than 1 cm). Based on what the nodule looks like on that ultrasound, we next determine whether or not it needs to be biopsied.

The following imaging features can indicate a thyroid nodule is more likely to be cancerous (my post on ultrasound explains more about ultrasound appearance and terminology):

  • solid: as opposed to fluid-filled (cystic) nodules
  • hypoechoic: meaning they look darker on ultrasound compared to the normal thyroid
  • microcalcifications: these small calcifications appear hyperechoic (bright) on ultrasound
  • irregular borders
  • invasion outside the thyroid

Based on a combination of these criteria and nodule size, an approximate cancer risk can be estimated for a given nodule, and subsequently a recommendation made whether or not to biopsy that nodule. In other words, a larger nodule with several worrisome features should be biopsied, whereas a smaller nodule without those features should not (the exact sizes and criteria that doctors use are less important than the concept behind these decisions).

Nodules that do NOT meet criteria for biopsy should be followed with ultrasound. Unfortunately, we don’t have very good evidence-based data on how frequently these nodules should be imaged. Most doctors use common sense, basing their decision on how suspicious the nodules appear on initial ultrasound: highly suspicious nodules get repeat ultrasound in 6-12 months; less suspicious nodules get one in 12-24 months; and non-suspicious nodules need no follow-up.

Question 4: Surgery time?

Thanks to our mad nodule-categorizing skillz, we have filtered out the least concerning nodules, leaving only nodules that need to be biopsied.

Thyroid nodules are biopsied using a technique called fine-needle aspiration, or FNA. “Fine” refers to the small diameter of the needle, not its sexiness. Aspiration means that negative pressure is applied to the syringe as the needle biopsy is performed, in order to suck up cells and small bits of tissue.

[Editor’s note: The workup of thyroid nodules after biopsy is complex and, at times, confusing. Rather than lay out precise, detailed explanations of all the subtleties, I chose to emphasize a few important concepts. If you are interested (and a bit masochistic), you can find all 90+ pages of the 2015 American Thyroid Association thyroid nodule management guidelines here]

Biopsy samples are categorized using the thrillingly-named “Bethesda System for Reporting Thyroid Cytopathology.” Each of the 6 categories has a corresponding estimated cancer risk and associated treatment recommendations.

  1. Nondiagnostic/Unsatisfactory (cancer risk 1-4%): Not enough material was obtained during biopsy to make an accurate diagnosis.
    1. Recommendation: Repeat the biopsy. If still uncertain, surgery should be considered to remove the nodule.
  2. Benign (cancer risk 0-3%): You are in the clear.
    1. Recommendation: No treatment is necessary. These are sometimes followed with ultrasound in a manner analogous to nodules that do not meet biopsy criteria.
  3. Atypia of undetermined significance, follicular lesion of undetermined significance (cancer risk 5-15%): This mouthful of a category encompasses biopsy results which are not clearly benign or malignant.
    1. Recommendation: Unclear (unsurprisingly). Biopsy is usually repeated, and molecular testing may be used on the biopsy specimen in order to further stratify risk. Molecular testing looks for specific genetic mutations or molecular markers that can indicate a higher risk of cancer.
    2. If the risk is still uncertain after these tests, there are 2 choices: continue to follow with ultrasound, or remove surgically. The results of molecular testing can help guide the decision on which of these options makes more sense.
  4. Follicular neoplasm, suspicious for follicular neoplasm (cancer risk 15-30%): A subtype of thyroid nodule that presents a unique problem is a follicular neoplasm (neoplasm=abnormal growth of tissue); it can be benign (follicular adenoma) or malignant (follicular carcinoma). Unfortunately, there is no way to tell—via biopsy, ultrasound, or other clinical information—whether a given nodule is benign or malignant. The pathologist must look at the whole nodule to determine if cancer is present.
    1. Recommendation: Surgery. Nodule removal is usually performed by removing the affected half of the thyroid, called a lobectomy.
  5. Suspicious for malignancy (cancer risk 60-75%): Biopsy results are very suspicious but don’t quite meet the criteria for cancer.
    1. Recommendation: Surgery.
  6. Malignant (cancer risk 97-99%): Yup, it’s cancer.
    1. Recommendation: Surgery.

Further treatment after surgery for thyroid cancer will sometimes include radioactive iodine treatment with iodine 131, the same isotope used to treat hyperfunctioning nodules. In these cases, the thyroid gland is first completely removed, and then iodine 131 is administered to kill any remaining thyroid cancer cells and to kill all remaining normal thyroid tissue.

Since we have nuked what was left of the thyroid, patients will have to take thyroid hormone replacement medications for the rest of their lives. This may seem like a bummer, but it has one big advantage. Patients who have this treatment should NOT be producing any thyroid hormone on their own (endogenous thyroid hormone). If a doctor detects endogenous hormone, it can indicate the cancer has returned. Thus, a simple periodic blood test can monitor a patient’s thyroid cancer.

Special considerations

I want to briefly touch on two unique situations which are treated differently from what I have described so far.

  • Very large thyroid nodules: Nodules larger than 4 cm are often treated with surgery right off the bat rather than biopsy, for a couple of reasons. First, nodules of this size are inherently at higher risk for being a cancer. Additionally, when you biopsy a large nodule, you only get a sample of a relatively small part of that nodule. Even if the biopsy came back benign, you would still worry you missed a cancer in a different part of the nodule.
  • Multinodular goiter: It is quite common to have multiple nodules in the thyroid. If only a few nodules are present, each nodule should be treated separately based on the criteria above. However, sometimes the entire thyroid is enlarged and replaced with nodules—called a multinodular goiter (goiter=enlargement of the thyroid). There may be dozens of nodules, and it’s impossible to biopsy all of them. If one of those nodules stands out and looks highly suspicious, it should be biopsied. If none of the nodules look particularly suspicious, options would include biopsy of the largest (>2cm) nodule, or surveillance of all nodules with follow-up ultrasound.
cluster of grapes
These grapes remind me of a multinodular goiter. Radiologists just don’t see the world in the same way.

As you can see, the management of thyroid nodules is far from straightforward. Because of this uncertainty, oodles of thyroid ultrasounds and biopsies are performed, probably more than are truly needed. With continued research, the hope is that we can decrease unnecessary testing and imaging of these nodules, and focus on treating clinically relevant thyroid cancers.

Questions or comments? Please post ’em below.

Comments are closed.