Browse the corpus

Thyroid nodules are common in clinical practice with a prevalence of up to 60%. The majority is benign and the risk of malignancy is between 7 to 15% (). Fine-needle aspiration biopsy (FNAB) of the thyroid gland is an accurate diagnostic test used routinely in the initial evaluation of nodular thyroid disease (-). Epidemiologic studies suggest that nodular thyroid disease is a common clinical problem, with a prevalence of 4% to 7% in the adult population in North America and an annual incidence of 0.1%, which translates into approximately 300,000 new nodules in the United States (). In patients with a single palpable nodule, additional nodules can be detected in about 20-48% by ultrasonography (). A survey of clinical members of the American Thyroid Association revealed that most endocrinologists (96%) perform FNA for diagnosis of thyroid nodules (). In addition, FNA with ultrasonographic guidance (US-FNA) is used routinely in follow-up surveillance of patients with thyroid cancer. Therefore, the importance of FNA biopsy in thyroid practice cannot be overemphasized. This chapter describes biopsy techniques, cytologic diagnosis, complications, FNA results, diagnostic pitfalls, and other information that may be useful to clinicians who manage patients with nodular thyroid disease.

The diagnosis of thyroid nodules by needle biopsy was first described by Martin and Ellis () in 1930, who used an 18-gauge–needle aspiration technique. Subsequently, cutting needle biopsy with Silverman or Tru-Cut needles was used for tissue examination. None of these techniques gained wide acceptance because of fear of malignant implants in the needle track, false-negative results, and serious complications. However, Scandinavian investigators introduced small‑needle aspiration biopsy of the thyroid in the 1960s, and this technique came into widespread use in North America in the 1980s (). For FNA biopsy, most use fine or thin (22- to 27-gauge) needles, most commonly 25 or 27-gauge needles. As the name indicates, the biopsy technique uses aspiration to obtain cells or fluid from a mass. In contrast to percutaneous large‑needle biopsy, which obtains tissue specimens and requires histologic fixation, aspiration biopsy offers cytological examination of the specimen. Another technique, fine‑needle non-aspiration (FNNA) biopsy, also referred to as capillary technique, avoids aspiration but obtains representative cytologic samples.

Although the FNA technique appears simple, considerable time and experience are required to acquire and maintain skillful biopsy technique. Debate continues about who is best qualified to perform FNA biopsy, but the best results are obtained if the person performing the biopsy has appropriate technique and volume. In the opinion of the authors, individuals performing biopsies should have appropriate knowledge of thyroid pathologies in order to relate the findings to the clinical context.

The basic equipment needed to perform FNA biopsy is simple and relatively inexpensive (, , ) The following items are necessary to have but many are not essential ().

The thyroid gland should be palpated carefully and the nodule(s) to be biopsied identified under sonographic guidance. The consent process is important. It should fully explain the procedure including risks and benefits, and address all questions raised by the patient. The procedure can be done with local anesthesia or without. Per nodule, typically 3 informative passes are obtained, in particular if rapid on-site evaluation (ROSE) by a cytopathologist is not available. In experienced hands, no serious complications are expected. Mild pain, minor bleeding or infection can occur but are unlikely to happen. Very rarely, transient hoarseness due to recurrent laryngeal nerve injury has been reported. With the appropriate caution, the procedure can usually be performed in patients with blood thinners.

The procedure is performed in the outpatient setting with the patient lying on an adjustable examination bed or chair. The procedure can be performed by a single person, but the presence of a nurse or clinical assistant is encouraged, if available. The patient may be seated or supine; we prefer the supine position. The patient is placed supine with the neck hyperextended to expose the thyroid; for support, a pillow is placed under the shoulders (). The patient is asked not to swallow, talk, or move during the procedure. It is best to talk to the patient and keep them informed of the progress of the biopsy. Once completed, firm pressure is maintained on the biopsy site(s) then the area is cleaned and a Band-Aid applied. It is best to observe patients for a few minutes for any side effects. This time can be used to counsel on after-procedure care instructions and potential results, if available including written instructions. Subsequently, the patient can be discharged.

Numerous reports, reviews, and even textbooks provide detailed descriptions of various FNA biopsy techniques (-). Although most reports agree on the principles of the technique, variations have been described to improve results. It is important to position the patient correctly, identify and locate the mass, provide adequate light during the biopsy, and have a clinical assistant available for help if needed. The physician performing the biopsy should be positioned at the patient’s side, preferably contralateral to the lesion. The nodule(s) to be aspirated is identified, and the overlying skin is cleansed with alcohol. The use of povidone-iodine or sterile technique is not necessary but encouraged. A retrospective study across multiple sites of the Mayo Clinic compared ultrasound-guided thyroid FNA using no anesthetic, subcutaneous injectable anesthetic, and topical anesthetic to compare the degree of pain/discomfort (). The study found patient discomfort associated with FNA was comparable during and after the procedure regardless of the use of anesthetic or the type. However, others reported that FNA-associated pain is frequent and that the use of local anesthesia is beneficial ()

In the method illustrated here, a 10‑mL plastic syringe is attached to a Cameco syringe holder and held in the right hand by a right‑handed operator (). Two fingers of the free (left) hand firmly grasp the nodule while the other hand holds a pistol-grip syringe holder (). The needle is then inserted through the skin and into the nodule. Once the needle tip is in the nodule, gentle suction is applied while the needle is moved in and out within the nodule vertically (). This maneuver allows the dislodging of cellular material and easy suction into the needle. During this period of 5 to 10 seconds, suction is maintained, and as soon as fluid or aspirate appears in the hub of the needle, the suction is released, and the needle is withdrawn. Of note, many operators do not use the handle; in this case, simply using a syringe with needle may suffice using the same maneuver.

Nowadays, ultrasound evaluation identifies cystic and partially cystic lesions. The appearance of fluid suggests that the nodule is cystic; suction is maintained, and as much fluid as possible is aspirated. It is important to release the syringe plunger and remove the vacuum before withdrawing the needle; this allows the aspirate to remain in the needle and not be sucked into the syringe. Next, the needle is detached from the syringe (), and 5 mL of air is drawn into the syringe (). The needle is reattached to the syringe, and with the bevel facing down, 1 drop of aspirated material is ejected onto each of several glass slides (). It is important that all slides be labeled and placed in order on a nearby table before the aspiration begins. Smears are prepared by using a second glass slide in a manner similar to that of making blood smears (). The slides for wet fixation should be placed immediately in 95% alcohol for staining with the Papanicolaou stain. For Giemsa staining, air-dried smears are necessary, and prepared slides are left unfixed and transported to the laboratory. Rapid on-site evaluation (ROSE) of cytology slides to evaluate for adequacy by cytopathologists may lower non-diagnostic rates and the number of necessary needle passes (). This may, however, add more time to complete the procedure.

Usually, 3 to 6 aspirations are made (, ), some authors suggest at least 6 (). Additional passes might be needed for molecular testing if decisions were made to obtain those. However, needle washouts typically provide sufficient materials for molecular analyses. Preferably, the aspirates should be obtained from the peripheral areas and different parts of the nodule in a sequential manner to ensure representative sampling (). For larger nodules, the deep center of the mass should be avoided because it is more likely to contain degeneration and fluid, decreasing the chance of a diagnostic specimen. For cystic lesions, the fluid should be completely aspirated and FNA attempted on residual solid tissue if present. Aspirated fluid should be placed in a plastic cup and saved for cytology evaluation. For each pass, one should use a new needle and syringe. The needle can be inserted in parallel or perpendicular way to the ultrasound probe. In the parallel approach, the US image can see the entire needle path. The perpendicular approach is generally simpler and the needle travels shorter distance; however, only needle tip can be seen in the nodule. Both techniques are equally effective.

Generally, minor procedures such as thyroid FNA, can be safely performed while patients are on antiplatelets/anticoagulation (). Special attention should be made to use a smaller gauge needle (27-gauge needle), minimize the number of passes, and monitor the patient for a few minutes for signs of bleeding. In certain situations, INR might need to be less than 2- 2.5 at the day of the procedure per certain operator preferences. Overall, the risk of stopping anticoagulation for the purpose of FNA procedure must always be weighed against the risk of complication related to discontinuing the medication for few days. The overall data suggest low risk of bleeding when doing the biopsy on anticoagulation and antiplatelet therapy with hematoma only occurred in 0.89% ().

The FNNA, also referred to as fine-needle capillary (FNC) sampling technique, has been described by several authors (, ). This technique is thought to minimize trauma to thyroid tissue and to reduce blood contamination. For this technique, patient preparation is similar to that for FNA. However, no syringe or suction is necessary. The hub of a 27-gauge needle is held in a pencil-grip fashion, and the needle is gently inserted into the nodule and then moved in and out for 5 to 10 seconds. Aspirate flows into the needle through capillary action, and as soon as aspirate appears in the hub, the needle is withdrawn and attached to a syringe with air inside. Next, the plunger is used to expel the material onto glass slides. The procedure is repeated several times, and the slides are prepared as described above for FNA. Other techniques such as Core-needle biopsy ( ) or Large-needle aspiration biopsy ( ) are used less commonly as first-line procedure. They can be associated with a higher risk of complications such as pain and bleeding, but they have some utility with repeatedly non-diagnostic FNA or when lymphoma, which requires additional analyses such as flow cytometry, is suspected.

After the biopsy has been completed, firm pressure is applied to the biopsy site(s) with a 4×4-inch gauze pad. Once bleeding has stopped, an adhesive bandage is placed on the puncture site(s), and the patient is observed for a few minutes. If there are no problems, the patient is allowed to leave.

Thyroid FNA biopsy, particularly using US-FNA, is very safe. No serious complications such as tumor seeding, nerve damage, tissue trauma, or vascular injury have been reported (-). Needle puncture may cause slight discomfort and possible skin ecchymosis at the aspiration site(s). However, even a minor hematoma is uncommon. Patient use of anticoagulants or salicylates does not preclude FNA biopsy. Needle track implantation of thyroid carcinoma is extremely rare and appears to be an exceptional complication (). Post aspiration hemorrhage within a cystic lesion is uncommon.

Aspirates from normal glands often have scant thyroid follicular cells and colloid. Wet‑fixed smears are usually prepared with a modified Papanicolaou stain, which shows nuclear details. Air-dried smears are often prepared with a Romanovsky stain. May‑Grünwald-Giemsa is a modified Romanovsky staining procedure that is sometimes used in thyroid cytologic preparations. The six Bethesda Classification System categories of cytology results include: Non-diagnostic (unsatisfactory), Benign, Atypia or undetermined significance (or follicular lesion of undetermined significance), Follicular neoplasm or suspicious for follicular neoplasm, suspicious for malignancy, and malignant ().

Aspirates obtained from multinodular goiters, a benign microfollicular adenoma, or normal thyroid are referred to as colloid nodules and show loosely cohesive sheaths of follicular epithelium, colloid, blood, and rare macrophages. Colloid nodules are the most common cytology and contain an abundance of colloid with sparse follicular cells. There is considerable variation in the number of cells, as well as the type and amount of colloid present (). Another benign diagnosis is Hashimoto thyroiditis. It has typically a characteristic pattern on FNA smears, showing hypercellularity with lymphocytes, Hürthle cells, and minimal or no colloid (). Subacute (granulomatous) thyroiditis is a rare condition with a benign aspirate. Typically, the smear shows multinucleated giant cells, epithelioid histiocytes, and scattered inflammatory cells.

Papillary carcinoma, the most common thyroid malignancy, is readily diagnosed by FNA. Typically, cytology shows large irregular nuclei, and nuclear grooves. Psammoma bodies may or may not be present, but if present, they are highly suggestive of papillary thyroid carcinoma (). Medullary thyroid carcinoma accounts for 5% to 10% of thyroid cancers and may present as a thyroid nodule or neck mass. Typically, aspirates from a medullary thyroid carcinoma are hypercellular, composed of large, poorly cohesive cells, and predominantly spindle shaped. Amyloid is often, but not invariably, present, and there is no colloid (). High-grade carcinoma can be diagnosed cytologically, but distinguishing between primary and metastatic cancer is not easy.

The accumulated experience of the past 4 decades has confirmed the reliability and usefulness of FNA as a diagnostic test (-, -, -). The role of FNA biopsy in the evaluation of thyroid nodules is now firmly established, and FNA has become the initial test because it is both safe and cost effective. In most clinics, FNA has become a standard test, performed most often by an endocrinologist.

An adequate specimen of good technical quality is considered diagnostic or satisfactory and may be benign, suspicious, or malignant. A benign (negative) cytologic diagnosis is reported for 50% to 90% of the specimens (average, 70%) (, , , ), 10-30% of FNA cytologic specimens may be suspicious for malignancy (indeterminate) (average, 20%) (, ). A malignant (positive) cytologic diagnosis is made in about 1% to 10% of cases (average, 5%). For example, Caruso and Mazzaferri () reported the following results from 9 series that included more than 9,000 patients: benign, 74%; malignant, 4%; inadequate, 11%; and suspicious, 11%. We reviewed more than 18,000 specimens from 7 large series and obtained similar cytologic results: benign, 69%; malignant, 4%; suspicious, 10%; and nondiagnostic, 17% ().

False-negative results reflects missed malignancy. False-negative rates generally occur in 1.5% to 11.5% (average, <5%) (, , , ). The false-negative rate is defined as the percentage of patients with a benign cytology in whom malignant lesions are later confirmed on thyroidectomy. The frequency of false-negative cytologic diagnosis depends on the number of patients who subsequently have surgery and histologic review. In most retrospective series, less than 10% of patients with a benign cytologic diagnosis subsequently undergo thyroid surgery, suggesting that false-negative rates should be interpreted with some skepticism (, ). Despite this note of caution, most authorities agree that the true false-negative rate is less than 5% if all patients undergo thyroid surgery. False-negative rates are lower in centers experienced with the procedure and with cytologic interpretation by expert cytopathologists.

False-positive rates vary from 0% to 8% (average, 3%) (, , ). A false‑positive diagnosis indicates that a patient with a malignant FNA result was found on histologic examination to have benign lesions.

Interpretive or sampling errors account for false diagnoses (, , ). Hashimoto thyroiditis is probably the most common cause of a false‑positive cytology. Misclassification of follicular and Hürthle cell adenomas as papillary carcinomas accounts for other errors. FNA biopsy of thyroid lymphomas may yield lymphocytes that can be interpreted as Hashimoto thyroiditis, accounting for a false‑negative diagnosis. Inadequate or improper sampling accounts for some false-negative errors. For example, nodules smaller than 1 cm may be too small for accurate needle placement, and nodules larger than 4 cm may be too large to allow proper sampling from all areas, thereby increasing the likelihood of misdiagnosis. Finally, the cytopathologist should establish and observe criteria to exclude a diagnosis of malignancy (, , , ).

Hypercellular specimens from follicular or Hürthle cell lesions may have features suggestive of, but not diagnostic for malignancy (, , , ). Thus, the cytopathologist labels these suspicious for malignancy because cytologic features neither confirm nor rule out malignancy. Histologic examination is necessary for definitive diagnosis. Hypercellularity may be seen with non-neoplastic lesions, and Hürthle cell changes may be seen in patients with lymphocytic thyroiditis. The diagnosis of follicular neoplasm is indicative of an underlying malignancy in 14% of cases and Hürthle cell neoplasm in 15% (, ). Many pathologists maintain that benign and malignant follicular tumors cannot be distinguished on the basis of aspirated cells only, and the lesion must be removed for histopathologic examination (, , ). However, Kini et al () believes that follicular adenomas and follicular carcinomas usually can be differentiated on the basis of nuclear size.

Several authors have discussed the problem of follicular neoplasm. In a study of 149 patients with the cytologic diagnosis of follicular neoplasm, Tuttle et al. () reported that risk of malignancy was higher in men, solitary nodules, and nodules larger than 4 cm. In a study of 219 patients with follicular neoplasm, Schlinkert et al. () showed that nodules are more likely malignant in younger patients, in men, if the nodule is solitary, and if it is larger than 4 cm. Baloch et al. () studied 184 cases of follicular neoplasm and reported that risk factors for malignancy included male sex, older age (>40 years), and larger nodules (>3 cm). Overall, they found that 70% of these lesions are benign. Molecular studies are now routinely used with thyroid nodule evaluation and biopsy to help further asses the likelihood of malignancy in thyroid nodules in nodules with undetermined cytology with reasonable sensitivity and specificity [see section below].

Inadequate specimens are labeled non-diagnostic or unsatisfactory and account for 2% to 20% of specimens (average, 10%) (, ). Several factors influence non-diagnostic rates for FNA results, including the skill of the operator, vascularity of the nodule, criteria used to judge adequacy of the specimen, and the cystic component of the nodule (-). Overall, a satisfactory smear contains at least 6 clusters of well‑preserved cells, with each group consisting of at least 10 to 15 cells. Re-aspirations with US-FNA yields satisfactory specimens in more than 50% of cases with a non-diagnostic initial FNA (, ). Chow et al. found a 7% malignancy rate in 153 patients with initial non-diagnostic smears (). Among 27 patients treated surgically, 37% had cancer. Re-aspiration with US-FNA was diagnostic in 66% and 56% without US; overall, 62% of re-aspirations were diagnostic. For patients with non-diagnostic FNA biopsy, FNA can be repeated few weeks later. A repeat ultrasound-guided FNA will yield a diagnostic cytology specimen in 75% of solid nodules and 50% of cystic nodules ().

If a patient has two nondiagnostic FNAs, ultrasound-guided core-needle biopsy can be considered as it has higher diagnostic yield compared to further repeating FNA (, ). However, the prevalence of cancer in non-diagnostic results is lower than in the general population. Nodules with repeatedly non-diagnostic FNA results typically have a low risk of malignancy and can be monitored if suspicious sonographic features are lacking ().

Analysis of the data reveals that the sensitivity of FNA ranges from 65% to 98% (mean 83%), and specificity ranges from 72% to 100% (mean 92%) (, , ). The predictive value of a positive or suspicious cytologic result is approximately 50%. The overall accuracy for cytologic diagnosis approaches 95% ().

Several guidelines and reviews have been published to help improve the adequacy and accuracy of cytology specimens (, -). FNA biopsy should be performed by individuals who have had training in both thyroid ultrasound examination and thyroid biopsy. Thyroid FNA in the hands of experienced operators achieves high diagnostic accuracy. Aspirates should be obtained from different portions of the nodule, preferably peripheral areas, in an organized and sequential manner. It is essential to ensure that an adequate number of follicular cells are present. A cytopathologist, preferably one with experience in thyroid cytology, should review and interpret the slides. If re-aspiration yields insufficient material, a core biopsy is the next step. In the event that the final result is still insufficient, surgical excision is warranted for nodules with suspicious features. Several reports have offered suggestions to minimize false-negative rates (, , , ). In a review on thyroid FNA, Belfiore and La Rosa suggested acquiring biopsy expertise, avoid suboptimal sampling, and repeat FNA during follow up especially if sonographically suspicious to reduce false-negative results (). To minimize false‑negative results, we follow the steps summarized in .

Published thyroid guidelines and reviews state that thyroid US should not be used as a screening test in the general population (, ). However, US is recommended for all patients with a single palpable nodule or a multinodular goiter or in a patient suspected of having a nodule () (). US machines are safe, easy to use, relatively inexpensive, have high resolution, and are widely available. It is important to note that US results are quite operator dependent.

Thyroid US is the imaging modality of choice for the evaluation of thyroid nodule(s). High-resolution neck US is safe and provides excellent imaging results with no significant cost, and can be done in the outpatient settings. Sonographic features used to characterize thyroid nodule risk of malignancy focus on five major sonographic criteria: echogenicity, composition, shape, echogenic foci, and margins (). Vascularity is debatable and therefor, not currently included in those criteria. US can also identify concerning lymph nodes. Features such as hypoechogenicity, irregular/infiltrative border, microcalcification, and a taller-than-wide shape in the transverse view, individually or in combination, are suggestive of malignancy especially if cervical lymph node involvement is seen. The diagnostic sensitivity and specificity of each of these features is variable and none of them alone can reliably distinguish malignant lesion from otherwise ().