This article reviews the most current literature on thyroid nodule evaluation, with particular attention to the problem of the incidentally identified thyroid nodule. Although traditional risk factors for thyroid cancer, such as age, gender, and familial syndromes, are still important, the manner in which a thyroid nodule comes to attention is of great importance these days when considering how to proceed in a workup. Most thyroid nodules today are discovered through radiologic imaging tests performed for other reasons. This article covers the key considerations that are vital in balancing the risks and benefits of thyroid nodule workup and treatment.
Key points
- •
Thyroid nodules are extraordinarily common; by age 90, virtually everyone has nodules.
- •
Ultrasound is the most valuable imaging study for making decisions about which nodules to biopsy.
- •
Nodules that are greater than 2 cm in size, that are entirely solid in composition, and that have microcalcifications are most likely to harbor a cancer.
- •
Molecular markers can help predict the presence of malignancy in cytologically indeterminate nodules, but markers do not accurately predict aggressiveness of cancers.
- •
Small papillary thyroid cancers can be safely observed in selected patients; discussions with patients should incorporate this option.
The thyroid nodule—scope of the problem
Thyroid nodules are extraordinarily common. A key challenge for clinicians is to decide which ones require evaluation and intervention. Half of people age 50 or over with clinically normal thyroid glands and thyroid function have thyroid nodules, and by age 90, virtually everyone has nodules. Thyroid cancer is commonly found at autopsy in individuals who have died of other causes, never having been detected in that patient’s life. Estimates of cancer prevalence at autopsy are quite variable and depend largely on the method used to detect the cancers and geographic location, but range from a low of about 4% to a high of 36%. Thus, thyroid cancers can be clinically insignificant for many patients. The workup and treatment can potentially expose the patient to the risks of treatment without the likelihood of any benefit. This challenging aspect of thyroid nodules has been recognized for some time, but the problem has been compounded in recent years by advances in and proliferation of imaging technology.
Advanced radiologic imaging rates (computed tomography [CT], magnetic resonance imaging [MRI], nuclear medicine, and ultrasound) have increased 3-fold since 1996. These scans commonly reveal small, nonpalpable thyroid nodules, which in the past would never have been identified because they were too small to detect by palpation, and too small to cause symptoms to the patient. Because so many of these incidental thyroid findings are now being uncovered, a dramatic increase in the observed incidence of small thyroid cancers is being experienced.
The increase in thyroid cancer incidence caused by this phenomenon is a problem for several reasons. First, patients are exposed to harm from what is ultimately unnecessary treatment. Second, these incidental findings unnecessarily create “patients with cancer” with all the attendant anxiety, surveillance needs, and financial ramifications. Last, these patients affect the validity of studies designed to understand and mitigate the risks of death or recurrence from thyroid cancer by serving to falsely improve the results of clinical trials. With this in mind, the chief challenge to clinicians today is deciding which nodules require workup, and how aggressively to treat them. What follows is a review of the current evidence related to the approach to the patient with a nodule.
The thyroid nodule—scope of the problem
Thyroid nodules are extraordinarily common. A key challenge for clinicians is to decide which ones require evaluation and intervention. Half of people age 50 or over with clinically normal thyroid glands and thyroid function have thyroid nodules, and by age 90, virtually everyone has nodules. Thyroid cancer is commonly found at autopsy in individuals who have died of other causes, never having been detected in that patient’s life. Estimates of cancer prevalence at autopsy are quite variable and depend largely on the method used to detect the cancers and geographic location, but range from a low of about 4% to a high of 36%. Thus, thyroid cancers can be clinically insignificant for many patients. The workup and treatment can potentially expose the patient to the risks of treatment without the likelihood of any benefit. This challenging aspect of thyroid nodules has been recognized for some time, but the problem has been compounded in recent years by advances in and proliferation of imaging technology.
Advanced radiologic imaging rates (computed tomography [CT], magnetic resonance imaging [MRI], nuclear medicine, and ultrasound) have increased 3-fold since 1996. These scans commonly reveal small, nonpalpable thyroid nodules, which in the past would never have been identified because they were too small to detect by palpation, and too small to cause symptoms to the patient. Because so many of these incidental thyroid findings are now being uncovered, a dramatic increase in the observed incidence of small thyroid cancers is being experienced.
The increase in thyroid cancer incidence caused by this phenomenon is a problem for several reasons. First, patients are exposed to harm from what is ultimately unnecessary treatment. Second, these incidental findings unnecessarily create “patients with cancer” with all the attendant anxiety, surveillance needs, and financial ramifications. Last, these patients affect the validity of studies designed to understand and mitigate the risks of death or recurrence from thyroid cancer by serving to falsely improve the results of clinical trials. With this in mind, the chief challenge to clinicians today is deciding which nodules require workup, and how aggressively to treat them. What follows is a review of the current evidence related to the approach to the patient with a nodule.
Patient presentation
When a patient comes to the office with a thyroid nodule, the mechanism of detection is of paramount importance and will determine what next steps should be taken ( Fig. 1 ). A patient who presents with symptoms of tracheal or esophageal compression should raise concern for a malignancy, although large goiter and Hashimoto thyroiditis can also cause these types of symptoms. The clinician should inquire regarding symptoms related to change in ease of breathing, swallowing, and speech quality. In contrast, a mass that was first noticed by the patient but that is not otherwise causing symptoms will have a much broader differential diagnosis ( Table 1 ). Although a rapid increase in size can signal malignancy, it can also signal hemorrhage into a benign neoplasm.
| Benign | Malignant |
|---|---|
| Multinodular goiter Hashimoto thyroidtis Simple or hemorrhagic cyst Follicular adenoma Subacute thyroiditis |
|
Today, many thyroid nodules come to attention through radiologic imaging studies. The nodules are subclinical—not causing any symptoms for the patient and generally not apparent on the physical examination. There are 2 main pathways of such radiologic detection—the first is through incidental detection on radiologic imaging studies done for other reasons, such as chest CT (obtained for cough, for example), neck MRI (performed after motor vehicle accident/suspected whiplash injury, for example), or carotid ultrasound. The second pathway is detection through a “diagnostic cascade” of nontargeted and sometimes inappropriate initial testing—for example, when a thyroid ultrasound is inappropriately ordered as part of a more general workup for weight gain, fatigue, or hair loss—and a thyroid nodule is found but is unrelated to the patient’s presenting symptoms. There are emerging data suggesting that traditional thyroid cancer risk factors, such as age, may be different for subclinical nodules that turn out to be papillary thyroid cancer on evaluation than for clinically apparent cancers, so obtaining a complete history is of even greater importance than in the past.
On physical examination, a mass that is hard and fixed may indicate malignancy, but a mass may also be hard in patients with Hashimoto thyroiditis. The presence of palpable adenopathy should raise suspicions of malignancy. Lymph nodes may be biopsied if radiographically indeterminate as part of the initial workup.
Assessment of risk factors
Traditional risk factors for thyroid cancer in a thyroid nodule include age, gender, history of radiation exposure, family history of thyroid cancer, and cancer syndromes. Data refining the significance of these risks continue to evolve ( Box 1 ).
Risk factors
Symptoms
Persistent hoarseness, dysphagia, or dysphonia
Lump is growing
History
Radiation exposure before age 20:
Therapeutic irradiation
Advanced medical imaging (eg, CT scan/PET scan)
Family history
Multiple endocrine neoplasia type 2a and 2b
Medullary thyroid cancer
Nonmedullary thyroid cancer
Familial syndromes
Gardner syndrome (familial adenomatous polyposis)
Cowden syndrome (a PTEN hamartoma syndrome)
Pendred syndrome
Werner syndrome
Carney complex
Physical findings
Male gender
Firm or hard consistency of the mass/nodule
Adenopathy of neck
Fixed to surrounding structures (rare)
Obvious voice, respiratory abnormality
Visible vocal cord abnormality
Age
It is often stated that the risk of finding cancer in a nodule is greater among patients who are less than age 30 or greater than age 60. However, in one commonly quoted study, patients were only included if they were operated on, which limits its generalizability to the broader population of all people undergoing thyroid nodule evaluation. Furthermore, to be operated on, patients had to have both a cold nodule on scintigraphy and also a suspicious needle biopsy. Thus, the data from this study are not useful for predicting cancer risk in a patient who presents to a clinician without having had any workup performed.
It may be more useful to consider age as a risk factor for death due to thyroid cancer. Understanding the risk of death allows one to decide about the utility of workup of a small nodule, because it allows assessment of the competing risks of death for a particular patient. Using this as a guide, the prognosis of someone diagnosed with follicular cancer is worse if they are age 60 or over, and for medullary cancer, prognosis also begins to decrease precipitously at age 60. Age is not a strong determinant of prognosis for papillary cancer. Anaplastic histology has a very poor prognosis generally, but is much worse among people diagnosed at age 40 or over.
Gender
The risk of thyroid cancer has historically been stated to be higher in men than in women presenting with a thyroid nodule. Indeed, autopsy rates of subclinical thyroid cancer are slightly higher in men than in women. However, the detected incidence of thyroid cancer is 3 times higher in women than in men. This incidence is likely, in part, because men access health care less frequently than women in general, giving them fewer opportunities for a nodule to be discovered incidentally through physical examination. In addition, male necks can be harder to examine than women’s necks, so nodules are likely to be found when they are larger or more noticeable. Last, women at almost every age have slightly more nodules than men, making them more likely to undergo workup than men. Thus, for a given individual presenting with a clinically apparent thyroid nodule, the likelihood of finding a cancer in the nodule is greater in a man than a woman.
Radiation Exposure
Data from long-term follow-up studies of Hiroshima and Nagasaki nuclear bomb survivors in Japan have recently been published. These studies show that the increased risk of thyroid cancer after a single large radiation exposure persists for at least 50 years. People exposed to 1 Gy or more from the bombings who were under the age of 20 have a 29% greater risk of developing thyroid cancer than for people of the same age who were not exposed. The younger people were at the time of exposure, the greater the risk of developing a thyroid cancer; however, individuals who were older than age 20 at the time of the bombings have not demonstrated an elevated risk of thyroid cancer. The risk of thyroid cancer from lower radiation doses received over a long period of time (eg, from medical imaging) cannot be directly inferred from these data, because the mechanism of action is different. However, rates are expected to be lower than that described above.
In a recent cohort study of Australian children ages 0 to 19 followed for a mean of 9.5 years after one CT scan (∼4.5-mSv radiation), there was a 24% increased risk of solid, leukemic, and lymphoid cancer in exposed children compared with children not so exposed. Subsequent CT scans incrementally increased the risk further, and the risk was greater for people exposed at younger ages, similar to the Japanese Hiroshima and Nagasaki data on one-time large exposures from the bombings. A complete history should thus include a query about exposure to advanced medical imaging—such as CT or positron emission tomography (PET) scans—during childhood and adolescence.
Family History and Cancer Syndromes
Family history has long been recognized to be a risk factor for medullary thyroid cancer. It is also now increasingly recognized to be a risk for the other thyroid cancer histologies. Two recent studies, one from the United States and the other from northern Europe, suggest about a 3-fold increased risk of developing nonmedullary thyroid cancer if a first-degree relative has had the diagnosis.
Several familial and syndromic cancer disorders are associated with increased thyroid cancer risk. Persons with familial adenomatous polyposis (Gardner syndrome) are reported to have a 1% to 2% lifetime risk of developing thyroid cancer, with a mean age at diagnosis of 28 years. This risk is just slightly greater than the current 1.1% lifetime risk of developing thyroid cancer in the in the United States. More serious is Cowden syndrome, one of the phosphatase and tensin homologue (PTEN) hamartoma tumor syndromes, which confers increased risk of breast and thyroid cancers. The risk of thyroid cancer in patients with this syndrome is increased 38% compared with people without the syndrome. In multiple endocrine neoplasia syndrome, types 2a and 2b are associated with medullary thyroid cancer. Patients with Pendred syndrome (deafness, goiter, and iodine organification abnormalities), Werner syndrome (premature aging and abnormal growth), Carney complex (myxomas of heart and skin, lentigines of skin, and endocrine overactivity), isolated hyperparathyroidism or pheochromocytoma, marfanoid habitus, and mucosa neuromas are also potentially at increased risk of thyroid cancer.
Workup of the thyroid nodule
TSH and Other Laboratory Tests
After obtaining a history and physical examination, a decision will be made about whether to further pursue workup. The first step, if that workup is to be undertaken, is to obtain a result of thyroid stimulating hormone (TSH). In a UK study, the rate of hyperthyroidism (15%) and hypothyroidism (2.3%) was higher in people undergoing nodule workup than the general population, so TSH assessment will allow identification of people with thyroid dysfunction requiring treatment regardless of the rest of the workup. In this same study, if the TSH was greater than 5.5, the odds of finding a thyroid malignancy on needle biopsy were 11 times greater than if the TSH was in the normal range. A low TSH result should be followed by thyroid scintigraphy for workup. A normal TSH should be followed by needle biopsy if appropriate, and a high TSH should get a workup for hypothyroidism in addition to needle biopsy when indicated.
Obtaining a calcitonin level is not recommended in the United States for several reasons, although there is not uniformity in US guidelines. Three major reasons given to omit calcitonin assessment are as follows:
- 1.
There is a high false positive rate (59% or more) in some studies.
- 2.
Pentagastrin is not available in the United States for further stratifying cancer risk if a high basal calcitonin result is found.
- 3.
The serum calcitonin cutoff levels for sporadic medullary thyroid cancer have not been agreed on.
Measurement of serum antithyroid peroxidase antibodies is not necessary in the routine workup of a thyroid nodule. Although it may help define the patients with autoimmune thyroiditis, it does not necessarily obviate fine-needle aspiration (FNA) if a discrete nodule is present. Serum thyroglobulin measurements are also unnecessary. These levels can be elevated in both benign and malignant disease and vary by gender and age.
Scintigraphy
Scintigraphy has been largely replaced by ultrasound, but still has at least 2 roles: identifying hyperfunctioning nodules when a low TSH is found on initial testing, and to a limited extent, determining which nodule to sample in patients with multiple nodules. The latter indication has now been largely replaced by ultrasound assessment.
Scintigraphy can be performed using 123 I or technetium-99m pertechnetate. However, radioiodine is preferred, because about 5% of thyroid cancers will be missed by pertechentate in people with normal TSH. Nonfunctioning (“cold,” uptake less than surrounding tissues) nodules may warrant needle biopsy, depending on size and ultrasound characteristics. Autonomous nodules (“hot,” uptake more than surrounding tissues) do not require needle biopsy, as they rarely harbor cancer. If the nodule is functioning but not making enough thyroid hormone to suppress TSH, it may appear indeterminate on scintigraphy.
Indeterminate nodules may occur for the reason outlined above, but can also occur because, as a 2-dimensional image, abnormal tissue may have normal tissue superimposed on it. This abnormal tissue with superimposed normal tissue may be particularly true when nodules are less than 2 cm in size. For the purpose of biopsy decisions, indeterminate nodules should be treated as though they are nonfunctioning.
Ultrasound
Thyroid ultrasound is the first choice of imaging studies for thyroid gland evaluation. Nodule size, detailed characteristics, anatomic location, and condition of nearby structures are all clearly delineated. Ultrasound has been shown to be more accurate than physical examination in detecting nodules.
Ultrasound qualities of a nodule in isolation are not diagnostic of a malignancy, but they do indicate which nodules are more likely to harbor malignancy, and so can inform decision-making about nodule selection for biopsy.
Several ultrasound findings have been found to be associated with malignancy among patients brought to surgery after biopsy ( Table 2 ). However, in a recent well-done case control study, only 3 of the commonly cited characteristics were found to be the most useful:
- 1.
Microcalcifications
- 2.
Size greater than 2 cm
- 3.
Nodules that are entirely solid composition
