The main reason for specifying adequacy of thyroid FNA specimens is to avoid false-negative diagnoses. To reduce the risk of false negatives, the cytopathologist should be able to identify the tissue appropriately. This, however, involves multiple parameters including the operator, slide preparation, proper fixation and staining, and also the inherent characteristics of the nodule itself, such as solid vs. cystic components, hemorrhage into the lesion, degenerative or necrotic changes and amount of sclerosis, calcification, or ossification involving the nodule. Therefore, there is no single criterion for the adequacy of thyroid FNAs. It should also be noted that the adequacy discussion applies to specimens that would otherwise be reported as benign; in other words, if a specimen is considered for any diagnosis other than benign, it should not be reported as nondiagnostic or unsatisfactory, but instead findings should be communicated in an appropriate manner in the pathology report. In this context, the value of detailed verbal or written communication cannot be overemphasized. While in certain practices, a terminology of nondiagnostic implies the features and findings are not “diagnostic for a specific entity,” and the term unsatisfactory is used when there is insufficient material for proper evaluation; the two forms a single diagnostic category in the Bethesda system.

One of the earlier reports on quantitative criteria for adequacy was from Dr. Goellner at Mayo Clinic giving actual numbers of follicular cells necessary for adequacy [33]. While “adequacy” reflects much more than the number of follicular cells on glass slides, this proposal by Goellner has remained useful for reporting thyroid cytology for decades and was also included in the Bethesda terminology. For this purpose, six groups of well-visualized cells, each with ten follicular cells, should be considered an “adequate” specimen for evaluation of thyroid nodules in the appropriate setting. This means that the cytologic specimen should be sufficient to identify the “lesion,” with clinical and preferably radiologic correlates.

The exceptions to the quantitative requirements for adequacy are those that would identify the lesion in the thyroid as anything other than benign or otherwise guide the clinical or surgical management of the patient. Examples include colloid nodules or inflammatory processes where the follicular cell component may not be well represented or not present at all in the aspirate smears.

Cyst contents without sufficient follicular epithelial cells are considered nondiagnostic. The main concern for these cases is a cystic papillary thyroid carcinoma. In such instances, the aspirates are reported as nondiagnostic with a statement that the FNA shows “cyst contents only.” Still, such smears have a very low risk of malignancy particularly for nodules smaller than 4 cm in size and those that shrink after the FNA procedure [34].

Similarly, obscuring blood, preservation and/or fixation artifacts, and staining problems can render the specimen nondiagnostic even if the cellularity is quantitatively “sufficient.”

While there are wide variations in the literature for the nondiagnostic category, overall it averages around 10 % [14, 15, 35–40]. In a meta-analysis including a large series in the post-Bethesda era, Bongiovanni reported an average nondiagnostic rate of 13 %, ranging from 1.8 to 23.6 %, in over 25 thousand FNAs [20].

The risk of malignancy for nondiagnostic specimens is difficult to assess in small series without sufficient follow-up, because the majority of these cases do not lead to surgical intervention. The studies that report a malignancy risk for this group with surgical follow-up overestimate the malignancy risk because of selection bias, i.e., the patients with surgical follow-up usually have additional indications for excision, such as increasing size, clinical symptoms, or abnormal or suspicious findings on imaging that skew the risk stratification for these patients. Overall, the malignancy risk with nondiagnostic specimens is actually very low. While it ranges from 0.6 to 39 % in different series, depending on how the data is collected, the malignancy risk is especially low for nodules without suspicious radiologic findings and smaller lesions [15, 35–37, 39, 41, 42]. In a study including 393 cases with an original nondiagnostic FNA but with adequate cytologic, surgical, or ultrasound follow-up, only 2.3 % were associated with malignancy [41]. In this series, the risk increased significantly with each 1 cm increase in any dimension of the nodule [41].

The overall inadequacy rate may decrease with ultrasound guidance [2, 3]. On-site evaluation of thyroid FNAs, with or without USG, may also prove helpful in further reducing the nondiagnostic rate of thyroid FNAs [38, 43, 44]. However, it should be emphasized that more important than any USG or aspiration technique is the experience and competency of the operator performing the procedure and also the cytologist evaluating the specimen [45]. The Bethesda system recommendation for nondiagnostic aspirates is a repeat FNA but “no sooner than 3 months later,” preferably with ultrasound guidance and rapid, on-site adequacy evaluation. While ultrasound guidance is likely to reduce the nondiagnostic rate, similar to on-site evaluation, there is no convincing data in the literature that requires a specific time interval for a repeat aspirate. Actually, recent studies that looked into this recommendation did not find any basis for a 3-month period in their series [46, 47]. Furthermore, no contraindication is proven for immediate repeat aspirate, either. On the other hand, it seems reasonable to allow the tissue repair to prevent overinterpretation of reparative/reactive changes as an atypical or neoplastic process particularly by an inexperienced cytopathologist. However, additional factors, including patient compliance, clinical and ultrasonographic findings, and operator experience should all be considered in deciding the most appropriate follow-up. This is particularly evidenced by studies that showed clinical and radiologic follow-up was as acceptable as a repeat aspirate for initially nondiagnostic thyroid FNAs, particularly in the absence of suspicious radiologic findings [41, 48, 49].

While FNA diagnosis of thyroid nodules can be utilized for confirmation of malignancy or determination of the extent of surgery, the primary purpose of a thyroid FNA is to document that the nodule is benign and no surgical excision is necessary. As the overwhelming majority of thyroid nodules are benign, in most practices, at least 60 % of thyroid FNAs are reported as such [9, 11, 14, 15, 33]. Therefore, thyroid FNA has been an extremely useful tool in prevention of many unnecessary thyroidectomies. When a benign diagnosis is rendered on cytology, the nodule can safely be followed clinically and radiologically, and no further immediate diagnostic studies are indicated [50].

The benign diagnosis includes multiple entities, including benign follicular nodule, colloid nodule, and inflammatory conditions.

The main cytologic characteristic of a benign follicular nodule is presence of colloid and a mixture of bland follicular cells, commonly including Hürthle cells. Therefore, proper identification of colloid on cytologic material is very important. It is common for colloid to “wash off” with fixation. Therefore, it is best seen on stained air-dried smears as dark blue-magenta-colored material. Colloid may be thick, dark, and cracked, or it may be “watery” as clouds of bluish tinge on smears (see Fig. 5.4). To an untrained eye, it may be difficult to differentiate colloid from serum.

When specimens show abundant colloid, even in the absence of follicular cells, those cases are reported as “benign” or “colloid nodule” as the malignancy risk for such lesions is considered to be extremely low [51]. However, in practice, this is a relatively rare occurrence. These can be considered as one end of the spectrum of “macrofollicular lesions.” The term colloid nodule should be reserved for those lesions that are clearly dominated by definite colloid on smears. Additionally, the cytologic findings should be supported by the imaging characteristics of the nodule sampled. A specimen with abundant colloid should not be reported as benign or adequate if the ultrasonographic features are consistent with a solid lesion.

In addition to colloid, follicular epithelial cells are commonly seen in smears from benign follicular nodules (see Fig. 5.5). They may be seen as sheets or follicles of various sizes. It is important to note that a minor component of microfollicles can be seen in benign follicular nodules, and presence of microfollicles in such a background should not be interpreted as atypical or follicular neoplasm. Follicles show a range of sizes and three-dimensional intact follicles can be seen. As the size of the follicles decrease, it is more likely to see colloid in the center of the follicles. Depending on the aspiration technique and the size of the needle, occasional thick tissue fragments may be seen; however, in a fine needle aspiration specimen, three-dimensional groups of follicles (instead of occasional individual follicles) should not be seen. Cellularity may be low to moderate and occasionally marked; however, there is a mixture of follicular architecture, ranging from small to large macrofollicles, flat sheets, and occasional microfollicles.

It is at this point pertinent to mention what constitutes a microfollicle. So far, the best definition of a microfollicle is by Renshaw as “less than 15 cells, arranged in a circle that is at least two-thirds complete, and flat.” Microfollicles can also be seen as small, compact, three-dimensional “spheres” with colloid in the center.

Follicular epithelial cells are bland, with moderate to abundant cytoplasm. Cytoplasm may be smooth or granular depending on the amount of cytoplasmic organelles and the metabolic activity of the cells. During the aspiration and smearing of the specimen, the cytoplasm may be ripped off, and scattered naked nuclei may be present in the background. Nuclei of normal follicular cells have a very slight variation in size, shape, and chromatin pattern. They are round to oval, monotonous cells with smooth, homogeneous chromatin. The nuclear membrane is usually very smooth and regular, without indentations, grooves, or intranuclear inclusions. Occasionally, one or two small, inconspicuous nucleoli may be seen but without angulations.

Hürthle cells or oncocytes are commonly seen as a part of benign follicular nodules. While Hürthle cell terminology for oncocytic lesions of the thyroid is a misnomer as Hürthle cells are actually the C-cells of dogs [52], it has been well established in the literature and clinical practice to use the name Hürthle for oncocytic cells in this location. Hürthle cells have abundant cytoplasm filled with mitochondria, which gives a homogeneously granular appearance to these cells. Nuclei are round to oval, moderately enlarged, and usually with a single prominent nucleolus. Hürthle cells may also show marked nuclear enlargement, membrane irregularities, and hyperchromasia, which should not be interpreted as atypia or malignancy. It should be noted that there may be “early Hürthle cells” with features intermediate between bland follicular epithelial cells and Hürthle cells.

Background is usually dominated by colloid and may be bloody. Bloody background is a sign of high vascularity and more commonly seen with neoplastic nodules; however, needle size, aspiration technique, and medications such as blood thinning agents may be related to markedly bloody aspirates.

The benign follicular nodule category should not be considered a diagnosis of exclusion. A nodule should not be diagnosed as benign if features are not diagnostic for any specific lesion. The cytopathologist should identify the features of a benign follicular nodule for appropriate diagnosis.