Context.—Fine-needle aspiration (FNA) biopsies have been an important component in the preoperative evaluation of thyroid nodules. Until the introduction of the Bethesda System for Reporting Thyroid Cytopathology (BSRTC) in 2008, individual institutions had developed their own diagnostic categories. The BSRTC proposed 6 categories in an attempt to standardize reporting of thyroid FNA.

Objective.—To present a 15-year experience of thyroid FNA at one institution, including data before and after introduction of the BSRTC. The risk of malignancy is compared with the meta-analysis of high-quality published data.

Data Sources.—Data sources were PubMed, a manual search of references, and institutional data.

Conclusions.—The diagnostic categories developed at our institution were similar to those proposed by the BSRTC, with best fit into the 6 categories easily accomplished and reported in the final 2 years of the study. Significant differences were noted in the frequencies of cases in diagnostic categories Benign (II; P = .003), Suspicious for follicular neoplasm/Follicular neoplasm (IV; P < .001), and Malignant (VI; P = .003) after the introduction of the BSRTC. Eighteen published articles met the criteria for inclusion in the meta-analysis. The risk of malignancy in each category in our institution was similar to that determined in the meta-analysis, except for Insufficient for diagnosis (I; 20% versus 9%–14%). Meta-analysis showed an overlapping 95% CI of risk of malignancy between Atypia of undetermined significance/Follicular lesion of undetermined significance (III; 11%–23%) and Suspicious for follicular neoplasm/Follicular neoplasm (IV; 20%–29%), suggesting similar risks of malignancy. The use of newer molecular tests for these indeterminate cases may further refine risk assessment.

Fine-needle aspiration (FNA) of the thyroid gland has proven to be an invaluable procedure for evaluating patients with thyroid nodules. It has been in use for more than 50 years, first advocated in Sweden.1  However, not until the late 1970s was its use advocated in the United States.2  As a method for triaging patients to distinguish those who require surgery from those who do not, institutions developed their own cytologic diagnostic categories without consensus among reporting institutions. Institutions reviewed their experiences with FNA of the thyroid to determine its usefulness and predictive value for malignancy.35  In 1996, the Papanicolaou Society of Cytopathology proposed guidelines and 10 diagnostic groups for the cytologic interpretation of thyroid FNA.6  In 2007, the National Cancer Institute hosted the NCI Thyroid Fine Needle Aspiration State of the Science Conference. The intention was to develop a system of classification of thyroid FNA similar in scope to the Bethesda Classification for gynecologic cervical cytology. By the publication and acceptance of what has become known as the Bethesda System for Reporting Thyroid Cytopathology (BSRTC), it was hoped that there would be more effective communication among provider stakeholders, greater facilitation of cytologic-histologic correlations, and the easy and reliable sharing of data between different laboratories.7  Further, the committee proposed morphologic criteria for each diagnostic category.8  It was hoped that the proposed label of “Bethesda” Classification would lend credibility to the classification by reference and similarity to the gynecologic cytology classification. The BSRTC has 6 diagnostic categories, including Nondiagnostic or Unsatisfactory (I); Benign (II); Atypia of undetermined significance or follicular lesion of undetermined significance (AUS/FLUS; III); Suspicious for follicular neoplasm or Follicular neoplasm (SFN/FN; IV); Suspicious for malignancy (V); and Malignant (VI).

Beginning in 1999, our institution embarked on a program of FNA sampling of thyroid nodules. The program was a coordinated effort among a small group of physicians, including a pathologist, interventional radiologists, endocrinologists, and thyroid surgeons. A system of classification based on the Papanicolaou Society's report6  was agreed upon by the providers. The system that developed was similar to that later proposed by the BSRTC, which enabled a smooth transition to that classification system. In this report, we have summarized our 15-year experience, with comparison to the BSRTC and an emphasis on the cytologic-histologic correlation. We also conducted a meta-analysis of high-quality published studies for comparison to our data. Our major aim was to reappraise the value of BSRTC and seek areas of potential improvement or update.

Prospective Correlation Study

Beginning January 1, 1999, through December 31, 2013, all results of thyroid FNA interpretations at the University Medical Center of Princeton at Plainsboro, New Jersey (formerly University Medical Center at Princeton, Princeton, New Jersey) were recorded prospectively into a database. Thyroid nodules were localized with palpation and ultrasound. The aspirates were prepared as direct smears and were alcohol fixed. Excess material from the aspirates was placed in Saccomanno or formalin (10%) fixative and a cell block prepared, when possible. The smears were stained with the standard Papanicolaou stain. Immediate diagnosis of adequacy was not provided. The material was submitted by clinicians from their office practice and from radiologists performing the procedure in the hospital outpatient setting. Interpretation of the aspirates was performed by pathologists only, not cytologists. During the 15 years of the study, 3735 of 5574 nodules (67%) were interpreted by one pathologist, 1616 (29%) by another, and 223 (4%) by a third. No retrospective review was performed on cases prior to analysis of the data.

The cytologic diagnoses were classified into 9 categories with the following criteria:

  • 1.

    Unsatisfactory. Fewer than 6 groups containing at least 10 well-preserved cells on each of at least 2 slides. Colloid only and hypocellular aspirates from cysts were considered unsatisfactory with a descriptive comment.

  • 2.

    Benign thyroid nodule. Variable numbers of unremarkable follicular cells in flat sheets, large clusters, and macrofollicles. Colloid and histiocytes were a common accompaniment. The differential diagnosis for the group included hyperplastic nodule, adenomatous follicular nodule, and macrofollicular adenoma.

  • 3.

    Chronic thyroiditis. Generally cellular specimens with abundant lymphocytes or lymphoid tangles and follicular cells showing a variety of changes in the same nodule, including variable oxyphilic change, nuclear pleomorphism, and nuclear anisocytosis. Histiocytes were common. Colloid was variable.

  • 4.

    Follicular lesion. Hypercellular specimens with colloid and variably sized follicles, predominantly microfollicles. Fewer cellular specimens with atypical nuclei, including nuclear enlargement, mild nuclear pleomorphism, and pale nuclei. The differential diagnosis included a benign nodule, a follicular neoplasm, and papillary carcinoma. Generally, the features of the specimen were insufficient to reach a more definitive diagnosis.

  • 5.

    Follicular neoplasm/suspicious for follicular neoplasm. Hypercellular specimens with a predominance of microfollicles in 3-dimensional, overlapping structures. Abundant isolated, intact, individual cells with syncytial microbiopsies. Little colloid was present. Cytologic nuclear changes of papillary thyroid carcinoma were lacking. The differential diagnosis included benign thyroid nodule, follicular adenoma, and well-differentiated follicular carcinoma.

  • 6.

    Hürthle cell lesion/suspicious for Hürthle cell neoplasm. Variable cellularity with predominantly follicular cells showing abundant granular and eosinophilic cytoplasm. Enlarged nuclei with or without prominent nucleoli. Variably sized follicles or sheets of follicular cells. Absence of lymphocytes. Colloid variable, but generally scant. The differential diagnosis included hyperplastic nodule, Hürthle cell adenoma, and Hürthle cell carcinoma.

  • 7.

    Suspicious for neoplasm. Generally hypercellular with cytologic features suggestive of papillary carcinoma, such as enlarged nuclei, atypical nuclei with irregular nuclear membranes, nuclear grooves, overlapping and crowding of nuclei, nuclear molding, and fine granular or powdery nuclear chromatin. These features were quantitatively and qualitatively insufficient for a more definitive diagnosis of malignancy. The differential diagnosis included benign thyroid nodule and papillary thyroid carcinoma.

  • 8.

    Suspicious for papillary carcinoma. Generally hypercellular with cytologic features suggestive of papillary carcinoma, such as enlarged nuclei, atypical nuclei with irregular nuclear membranes, nuclear grooves, overlapping and crowding of nuclei, nuclear molding, fine granular or powdery nuclear chromatin, and intranuclear pseudoinclusions. These specimens usually lacked 1 or 2 of the cytologic features of papillary thyroid carcinoma, preventing a definitive diagnosis. The differential diagnosis included benign thyroid nodule and papillary thyroid carcinoma.

  • 9.

    Malignant. Generally hypercellular with cytologic features qualitatively and quantitatively definitive for papillary carcinoma as described in Suspicious for papillary carcinoma. This category also would include anaplastic carcinoma, medullary carcinoma, lymphoma, squamous cell carcinoma, and other carcinomas, such as metastatic carcinoma.

In November 2011, our institution modified its reporting categories to include the BSRTC category with a best fit to the categories that had been in use.

This study has been approved by the Institutional Review Board at our institution.

The Meta-analysis

We searched PubMed with the term “cytology AND thyroid AND Bethesda AND accuracy” in May 2014. Additional articles were also found or removed by manually reviewing the full text and references of the matched articles. Two of the coauthors (M.M. and L.Z.) cross-checked the accuracy and relevance of the articles. Only studies meeting the following criteria were considered to have high-quality data and to be relevant to the subject, and were included: (1) the studies were indexed in PubMed; (2) the studies reported correlations between thyroid cytology Bethesda classification and respective surgical pathology diagnosis of thyroid nodules; (3) the studies included both classes II and VI, and one or more additional other categories, or included categories III, IV, and V. The criterion 3 is particularly important for ensuring that the included data represent a wide spectrum of pathology between benign and malignant cases. This criterion in our view would significantly reduce the selection bias introduced by the studies that included only one or two classes.

We used the best-fit approach to extract data and reclassify cases, mostly for the works published prior to the publication of BSRTC. Specifically, the acceptable alternative terminologies for AUS/FLUS (III) category were indeterminate for neoplasm, indeterminate cells of undetermined significance, atypical cellular lesion, atypical cells of undetermined significance, atypical follicular cells, abnormal, indeterminate, indeterminate follicular lesion, and inconclusive or indeterminate. The acceptable alternative terminologies for SFN/FN (IV) category were neoplasm, follicular/Hürthle cell neoplasm, suspicious for neoplasm, suspicious for follicular or oncocytic neoplasm, Hürthle cell, indeterminate Hürthle cell, and Hürthle cell neoplasm.

The statistical analyses, including meta-analysis, were carried out by using STATA version 12.0 (Stata Corp, College Station, Texas) as described previously by Zhang et al.9  Random efforts were applied to the meta-analysis. A random-effect model was used for meta-analysis when I2 > 50% or P < .05 in the heterogeneity test; otherwise, a fixed-effect model was used. P < .05 was considered statistically significant.

Between January 1, 1999, and December 31, 2013, a total of 5574 nodules were sampled during 4338 patient encounters. Table 1 shows the distribution of diagnoses before and after the inclusion of the BSRTC category, with the descriptive diagnosis for each FNA. There was significant frequency difference between before and after inclusion of the BSRTC (P < .001), particularly within the Benign (II) and Suspicious for malignancy (V) categories (P < .001 for both). No frequency difference was found between before and after inclusion of the BSRTC in the categories Insufficient for diagnosis (I), AUS/FLUS (III), SFN/FN (IV), and Malignant (VI; Table 1).

Table 1.

Distribution of Diagnoses Before and After Implementation of the Bethesda Classification

Distribution of Diagnoses Before and After Implementation of the Bethesda Classification
Distribution of Diagnoses Before and After Implementation of the Bethesda Classification

Of the 576 FNAs with “Insufficient for diagnosis” (which included cysts), 193 patients (42%) had reaspirations, of whom 165 (85%) resolved into a more specific category. Of the 238 “Follicular lesion” diagnoses in 236 patients, only 6 patients (2.5%) had repeat aspirations of the nodules. All 6 (100%) had a diagnosis of “Follicular lesion” on the repeat aspiration, which was separated from the first aspiration by as much as 6 years. Of the 343 “Suspicious for follicular neoplasm/follicular neoplasm” diagnoses, which included those nodules with Hürthle cell features, in 324 patients, only 6 patients (1.8%) had repeat aspirations of the nodules. All 6 (100%) had a diagnosis of “Suspicious for follicular neoplasm/follicular neoplasm” on the repeat aspiration, which was separated from the first aspiration by as much as 7 years. Only 1 of these 6 patients (16.6%) had Hürthle cell features, which was present on both the original as well as the repeat aspiration.

According to the BSRTC, although Hürthle cell lesions/neoplasms are included in the category SFN/FN (IV), it was also recommended to specifically note the presence of Hürthle cell changes. At our institution, Hürthle cell lesion/neoplasm has always been considered a distinct category, and although indicated as BSRTC SFN/FN (IV), it is presented in Table 1 as a separate line category.

Similarly, “Suspicious for papillary carcinoma” had always been considered by us as a separate category, and it is presented as a BSRTC category V lesion, “Suspicious for malignancy.”

The distribution of several diagnoses showed a significant change after the introduction of the BSRTC classification. Benign (II) showed a modest but significant increase in percentage of cases, whereas SFN/FN (IV), Hürthle cell lesion/neoplasm (IV), and Suspicious for papillary carcinoma (V) all showed significant decreases.

A total of 556 patients underwent a partial or total thyroidectomy. Among them, 634 nodules were identifiable based on the clinical or ultrasonographic description, and were correlated with their cytologic diagnosis. The distribution of the cytologic-histologic correlation of the nodules within the excised thyroid glands is presented in Table 2. The 2 diagnostic categories Hürthle cell lesion/Neoplasm and Suspicious for papillary thyroid carcinoma are presented as separate line categories.

Table 2. 

Cytologic-Histologic Correlation 1999–2013

Cytologic-Histologic Correlation 1999–2013
Cytologic-Histologic Correlation 1999–2013

There were 241 malignant nodules identified, of which 19 (7.9%) were follicular carcinoma, 16 (6.6%) Hürthle cell carcinoma, 203 (84.2%) papillary thyroid carcinoma, 2 (0.8%) medullary carcinoma, and 1 (0.4%) squamous cell carcinoma. Within each category, except Hürthle cell lesion/neoplasm, papillary thyroid carcinoma was the dominant type of cancer identified, including 10 of the 14 cases (71.4%) within the follicular lesion category and 15 of the 18 cases (83.3%) within the follicular neoplasm category.

Table 3 presents the cytologic-histologic correlations charted in Table 2 as risk of malignancy (ROM) given the cytologic diagnosis. Malignancy included follicular carcinoma, Hürthle cell carcinoma, papillary carcinoma, medullary carcinoma, and squamous cell carcinoma. Although other carcinomas can be seen in the thyroid gland, these were the only malignancies identified in our patients. The ROM was calculated by dividing the number of malignancies detected by thyroidectomy, by the number of thyroidectomies, in a diagnostic category, × 100. The ROM was higher for insufficient (I; 20%) and AUS/FLUS (III; 25%) than that reported in the BSRTC.8  Of those insufficient cases which on subsequent aspiration resolved into more definitive categories (n = 165), 16 (10%) came to surgery. Of these 16, an additional 5 (31%) malignancies were identified. If these cases had surgery on the first “Insufficient for diagnosis,” the ROM of “Insufficient for diagnosis” would have been 23%. In Table 3, these 16 cases were categorized in their more definitive categories, not in the “Insufficient for diagnosis” category.

Table 3. 

Cytologic-Histologic Correlation Risk of Malignancy

Cytologic-Histologic Correlation Risk of Malignancy
Cytologic-Histologic Correlation Risk of Malignancy

Of the 628 articles identified in PubMed, 7 articles met the inclusion criteria. An additional 11 qualified articles were identified by searching the references of those articles and personal article collections. A total of 18 articles therefore were included for the meta-analysis and subject to the heterogeneity test (Figures 1 through 7; Table 4). Our meta-analysis found that AUS/FLUS (III) and SFN/FN (IV) had similar ROMs, as shown by overlapping 95% CIs (average [95% CI], 0.24 [0.16–0.32] and 0.25 [0.19–0.31], respectively; Figures 3 and 5 and Table 5). The ROM of the combined AUS/FLUS (III) and SFN/FN (IV) was 0.24 (0.19–0.30; Figure 4 and Table 5) and did not overlap the 95% CIs of Benign (II) or Suspicious for malignancy (V). The pooled ROMs of categories Benign (II), Suspicious for malignancy (V), and Malignant (VI) were significantly different from each other, and from AUS/FLUS (III) and SFN/FN (IV) or combined from AUS/FLUS (III) and SFN/FN (IV). Of note, the pooled ROM from meta-analysis was slightly different from the crude malignancy rate in cases combined because weighted ROMs from the included studies were used for calculation (Figures 1 through 7 and Table 4).

Figure 1. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification I for thyroid fine-needle aspiration (I2 = 42.1%, P = .04). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 1. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification I for thyroid fine-needle aspiration (I2 = 42.1%, P = .04). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 2. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification II for thyroid fine-needle aspiration (I2 = 93.6%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 2. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification II for thyroid fine-needle aspiration (I2 = 93.6%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Table 4. 

The List of the Studies Qualified and Included in the Meta-analysis

The List of the Studies Qualified and Included in the Meta-analysis
The List of the Studies Qualified and Included in the Meta-analysis
Figure 3.

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification III for thyroid fine-needle aspiration (I2 = 93.2%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 3.

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification III for thyroid fine-needle aspiration (I2 = 93.2%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 4. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification IV for thyroid fine-needle aspiration (I2 = 91.0%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 4. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification IV for thyroid fine-needle aspiration (I2 = 91.0%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 5. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classifications III and IV (combined) for thyroid fine-needle aspiration (I2 = 94.2%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 5. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classifications III and IV (combined) for thyroid fine-needle aspiration (I2 = 94.2%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 6. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification V for thyroid fine-needle aspiration (I2 = 93.3%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 6. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification V for thyroid fine-needle aspiration (I2 = 93.3%, P < .001). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 7. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification VI for thyroid fine-needle aspiration (I2 = 25.1%, P = .24). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Figure 7. 

Forest plot of meta-analysis on the risk of malignancy of Bethesda classification VI for thyroid fine-needle aspiration (I2 = 25.1%, P = .24). The reference numbers of the included articles are listed in the parentheses. Abbreviation: ES, estimated.

Table 5. 

Meta-analysis of the Risk of Malignancy in the Bethesda Classes

Meta-analysis of the Risk of Malignancy in the Bethesda Classes
Meta-analysis of the Risk of Malignancy in the Bethesda Classes

In the past 40 years, thyroid FNA has been recognized as the most useful diagnostic tool in triaging those patients most at risk of harboring a malignant lesion. We have reported our 15 years of experience with the technique, ROM in each diagnostic category, and comparison with the pooled ROM through a meta-analysis on high-quality published data.

The results of this thyroid FNA study confirm the usefulness and comparability of this technique in our institution to the experience at other institutions.1017  The distribution of diagnoses is similar to that reported as expected when using the BSRTC and our meta-analysis results.8 

After the implementation of the BSRTC, the distribution of several diagnoses showed a significant change. Benign (II) showed a modest but significant increase, whereas SFN/FN (IV), Hürthle cell lesion/Neoplasm (IV), and Suspicious for Papillary Carcinoma (V) all showed significant decreases. The publication by Baloch et al8  of diagnostic criteria for each category aided us in refining our interpretations and may account for the change in distributions from 1999–2011 versus 2012–2013. However, our experience with reporting the BSRTC encompassed only 2 years versus 13 years. The distribution could regress to the results of the previous 13 years as time and the number of aspirates progresses forward.

The recognition that the BSRTC category of Papillary Carcinoma accepted a less than 100% ROM influenced us to progressively eliminate the category of Suspicious for papillary carcinoma, in which we had a 93% ROM, in favor of the more definitive diagnostic category Papillary Carcinoma.

Indeed, combining our Suspicious for papillary carcinoma category, primarily used before the introduction of the BSRTC, with our Papillary Thyroid Carcinoma category would have produced a ROM of 96%, similar to that reported in the BSRTC8  category Malignant (VI).

Of the 576 insufficient for diagnosis patients, 193 (34%) had repeat aspirations, with 85% of them resolving into more definitive diagnostic categories. A total of 41 patients went on to have partial or total thyroidectomy, with 8 (20%) of them showing malignancy. The meta-analysis also identified a 12% (95% CI, 9%–14%) ROM in this category. The higher rate of malignancy in our series may reflect other suspicious clinical features that prompted the election of surgical excision. It is noteworthy to recognize that insufficient for diagnosis does not mean “negative for malignancy,” and these patients should have close clinical follow-up.

The ROMs of our categories Suspicious for malignancy (V) and Malignant (VI) are similar to that of the meta-analysis. AUS/FLUS (III) is slightly higher (25%) and SFN/FN (IV) is slightly lower (17%) than the ranges seen in the meta-analysis (95% CIs, 11%–23% and 20%–29%, respectively). This likely reflects subjective differences in how we interpreted the criteria for these categories. This subjectivity was mirrored as well by the wide 95% CIs seen in the meta-analysis. It might be useful for laboratories to create a standardized performance measure (such as ratio of AUS/FLUS to total thyroid FNAs) and incorporate a consensus review for lesions falling into indeterminate categories (AUS/FLUS [III] and SFN/FN [IV]). It is interesting to note that there was a wide CI for the Suspicious for malignancy (V) category in the meta-analysis, indicating significant subjectivity here as well.

The usefulness of FNA of the thyroid is in its ability to aid clinicians in their decision on how best to treat patients. Thus, Benign (I), Suspicious for malignancy (V), and Malignant (VI) categories have sufficient ROMs, in our institution (4%, 66%, and 96%, respectively) and other reports summarized by our meta-analysis, as to give clinicians clear treatment directions.

The more problematic categories of AUS/FLUS (III), SFN/FN (IV), and Hürthle cell lesion/Neoplasm (IV) have ROMs that are in an indeterminate range, 17% to 39%, and therefore do not give clinicians a clear direction. In our series, very few of these cases had repeat aspirations, which have been reported in other series to resolve the dilemma in up to 50% of cases of AUS/FLUS (III).18  However, the ROM for an AUS/FLUS (III) diagnosis with a subsequent Benign (II) diagnosis on a follow-up FNA is not the same as a single Benign (II) diagnosis, but rather falls somewhere in between the two (15%–29%).19,20  This would seem to preclude the use of a second aspiration in cases of AUS/FLUS (III) for additional risk stratification. The presented meta-analysis shows overlapping ROMs for AUS/FLUS (III) and SFN/FN (IV). On the surface, AUS/FLUS (III) and SFN/FN (IV) have different diagnostic criteria, but the ROM is what drives a clinician's decision, not the terminology. Newer molecular tests have been developed and in particular assist in triaging these indeterminate (AUS/FLUS [III] and SFN/FN [IV]) cases.21,22  These initial papers appear encouraging, but further refinement and development of additional molecular markers may enable an even more accurate categorization. In its latest guidelines, the American Thyroid Association indicates that “investigations such as repeat FNA or molecular testing may be used to supplement malignancy risk assessment data, in lieu of proceeding directly with a strategy of either surveillance or diagnostic surgery” for AUS/FLUS (III), and “molecular testing may be used to supplement malignancy risk assessment data, in lieu of proceeding directly with surgery” for SFN/FN (IV).23  The use of these molecular markers as a reflex to initially indeterminate cytologic diagnoses may therefore lead to, in the future, the collapse of the AUS/FLUS (III) and SFN/FN (IV) categories into a single indeterminate category with a consistent and unified set of management recommendations. However, prospective studies and high-quality data are needed to validate our findings and proposals.

1
Einhorn
J,
Franzen
S.
Thin-needle biopsy in the diagnosis of thyroid disease
.
Acta Radiol
.
1962
;
58
:
321
336
.
2
Miller
JM,
Hamburger
JI,
Kini
SR.
The impact of needle biopsy on the preoperative diagnosis of thyroid nodules
.
Henry Ford Hosp Med J
.
1980
;
28
(
2–3
):
145
148
.
3
Gharib
H,
Goellner
JR,
Johnson
DA.
Fine-needle aspiration cytology of the thyroid: a 12 year experience with 11,000 biopsies
.
Clin Lab Med
.
1993
;
13
(
3
):
699
709
.
4
Bisi
H,
de Camargo
RY,
Longatto Filho A. Role of fine-needle aspiration cytology in the management of thyroid nodules: review of experience with 1,925 cases
.
Diagn Cytopathol
.
1992
;
8
(
5
):
504
510
.
5
Baloch
ZW,
Sack
MJ,
Yu
GH,
Livolsi
VA,
Gupta
PK.
Fine-needle aspiration of the thyroid: an institutional experience
.
Thyroid
.
1998
;
8
(
7
):
565
569
.
6
Guidelines of the Papanicolaou Society of Cytopathology for the examination of fine-needle aspiration specimens from thyroid nodules: the Papanicolaou Society of Cytopathology Task Force on standards of practice
.
Mod Pathol
.
1996
;
9
(
6
):
710
714
.
7
Cibas
ES,
Ali
SZ.
The Bethesda System for Reporting Thyroid Cytopathology
.
Am J Clin Pathol
2009
;
132
(
5
):
658
665
.
8
Baloch
ZW,
LiVolsi
VA,
Asa
SL,
et al.
Diagnostic terminology and morphologic criteria for cytologic diagnosis of thyroid lesions: a synopsis of the National Cancer Institute thyroid fine needle aspiration state of the science conference
.
Diagn Cytopathol
.
2008
;
36
(
6
):
425
437
9
Zhang
L,
Chen
Z,
Fukuma
M,
Lee
LY,
Wu
M.
Prognostic significance of race and tumor size in carcinosarcoma of gallbladder: a meta-analysis of 68 cases
.
Int J Clin Exp Pathol
.
2008
;
1
(
1
):
75
83
.
10
Amrikachi
M,
Ramzy
I,
Rubenfield
S,
Wheeler
TM.
Accuracy of fine-needle aspiration of thyroid: a review of 6226 cases and correlation with surgical or clinical outcome
.
Arch Pathol Lab Med
.
2001
;
125
(
4
):
484
488
.
11
Arda
IS,
Yildirim
S,
Demirhan
B,
Firat
S.
Fine needle aspiration biopsy of thyroid nodules
.
Arch Dis Child
.
2001
;
85
(
4
):
313
317
.
12
Greaves
TS,
Olvera
M,
Florentine
BD,
et al.
Follicular lesions of thyroid: a 5-year fine-needle aspiration experience
.
Cancer Cytopathol
.
2000
;
90
(
6
):
335
341
.
13
Sidawy
MK,
Del Vecchio
DM,
Knoll
SM.
Fine needle aspiration of thyroid nodules: correlation between cytology and histology and evaluation of discrepant cases
.
Cancer Cytopathol
.
1997
;
81
(
4
):
253
259
.
14
Ravetto
C,
Colombo
L,
Dottorini
ME.
Usefulness of fine needle aspiration in the diagnosis of thyroid carcinoma: a retrospective study in 37,895 patients
.
Cancer Cytopathol
.
2000
;
90
(
6
):
357
363
.
15
Wu
HH,
Jones
JN,
Osman
J.
Fine-needle aspiration cytology of the thyroid: ten year experience in a community teaching hospital
.
Diagn Cytopathol
.
2006
;
34
(
2
):
87
88
.
16
Bakhos
R,
Selvaggi
SM,
DeJong
S,
et al.
Fine-needle aspiration of the thyroid: rate and causes of cytohistopathologic discordance
.
Diagn Cytopathol
.
2000
;
23
(
4
):
233
237
.
17
Al-Hureibi
KA,
Al-Hureibi
AA,
Abdulmughni
YA,
et al.
The diagnostic value of fine needle aspiration cytology in thyroid swellings in a University Hospital, Yemen
.
Saudi Med J
.
2003
;
24
(
5
):
499
503
.
18
Ratour
J,
Polivka
M,
Dahan
H,
et al.
Diagnosis of follicular lesions of undetermined significance in fine-needle aspirations of thyroid nodules
.
J Thyroid Res
.
2013
;
2013
:
250347
.
19
Renshaw
AA.
Does a repeated benign aspirate change the risk of malignancy after an initial atypical thyroid fine-needle aspiration?
Am J Clin Pathol
.
2010
;
134
(
5
):
788
792
.
20
Marchevsky
AM,
Walts
AE,
Bose
S,
et al.
Evidence-based evaluation of the risks of malignancy predicted by thyroid fine-needle aspiration biopsies
.
Diagn Cytopathol
.
2010
;
38
(
4
):
252
259
.
21
Nikiforov
YE,
Ohori
NP,
Hodak
SP,
et al.
Impact of mutational testing on the diagnosis and management of patients with cytologically indeterminate thyroid nodules: a prospective analysis of 1056 FNA samples
.
J Clin Endocrinol Metab
.
2011
;
96
(
11
):
3390
3397
.
22
Alexander
EK,
Kennedy
GC,
Baloch
ZW,
et al.
Preoperative diagnosis of benign thyroid nodules with indeterminate cytology
.
N Engl J Med
.
2012
;
367
(
8
):
705
715
.
23
Haugen
BR,
Alexander
EK,
Bible
KC,
et al.
2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer: the American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer
.
Thyroid
.
2016
;
26
(
1
):
1
133
.
24
Banks
ND,
Kowalski
J,
Tsai
HL,
et al.
A diagnostic predictor model for indeterminate or suspicious thyroid FNA samples
.
Thyroid
.
2008
;
18
(
9
):
933
941
.
25
Blansfield
JA,
Sack
MJ,
Kukora
JS.
Recent experience with preoperative fine-needle aspiration biopsy of thyroid nodules in a community hospital
.
Arch Surg
.
2002
;
137
(
7
):
818
821
.
26
Bohacek
L,
Milas
M,
Mitchell
J,
Siperstein
A,
Berber
E.
Diagnostic accuracy of surgeon-performed ultrasound-guided fine-needle aspiration of thyroid nodules
.
Ann Surg Oncol
.
2012
;
19
(
1
):
45
51
.
27
Bozhok
Y,
Greenebaum
E,
Bogdanova
TI,
et al.
A cohort study of thyroid cancer and other thyroid diseases after the Chernobyl accident: cyto-histopathologic correlation and accuracy of fine needle aspiration biopsy in nodules detected during the first screening in Ukraine
.
Cancer
.
2009
;
117
(
2
):
73
81
.
28
Faquin
WC,
Baloch
ZW.
Fine-needle aspiration of follicular patterned lesions of the thyroid: diagnosis, management, and follow-up according to National Cancer Institute (NCI) recommendations
.
Diagn Cytopathol
.
2010
;
38
(
10
):
731
740
.
29
Jo
VY,
Stelow
EB,
Dustin
SM,
Hanley
KZ.
Malignancy risk for fine-needle aspiration of thyroid lesions according to the Bethesda System for Reporting Thyroid Cytopathology
.
Am J Clin Pathol
.
2010
;
134
(
3
):
450
456
.
30
Mufti
ST,
Molah
R.
The Bethesda System for Reporting Thyroid Cytopathology: a five-year retrospective study of one center experience
.
Int J Health Sci
.
2012
;
6
(
2
):
159
173
.
31
Nayar
R,
Ivanovic
M.
The indeterminate thyroid fine-needle aspiration
.
Cancer Cytopathol
.
2009
;
117
(
3
):
195
202
.
32
Oertel
YC,
Miyahara-Felipe
L,
Mendoza
MG,
Yu
K.
Value of repeated fine needle aspirations of the thyroid: an analysis of over ten thousand FNAs
.
Thyroid
.
2007
;
17
(
11
):
1061
1066
.
33
Poller
DN,
Ibrahim
AK,
Cummings
MH,
Mikel
JJ,
Boote
D,
Perry
M.
Fine-needle aspiration of the thyroid: importance of an indeterminate diagnostic category
.
Cancer Cytopathol
.
2000
;
90
(
4
):
239
244
.
34
Renshaw
AA.
Subclassification of atypical cells of undetermined significance in direct smears of fine-needle aspirations of the thyroid
.
Cancer Cytopathol
.
2011
;
119
(
5
):
322
327
.
35
Sclabas
GM,
Staerkel
GA,
Shapiro
SE,
et al.
Fine-needle aspiration of the thyroid and correlation with histopathology in a contemporary series of 240 patients
.
Am J Surg
.
2003
;
186
(
6
):
702
709
.
36
Smith
M,
Pantanowitz
L,
Khalbuss
WE,
Benkovich
VA,
Monaco
SE.
Indeterminate pediatric thyroid fine needle aspirations: a study of 68 cases
.
Acta Cytol
.
2013
;
57
(
4
):
341
348
.
37
Theoharis
CGA,
Schofield
KM,
Hammers
L,
Udelsman
R,
Chhieng
DC.
The Bethesda Thyroid Fine-Needle Aspiration Classification System: year 1 at an academic institution
.
Thyroid
.
2009
;
19
(
11
):
1215
1223
.
38
Wu
HH,
Rose
C,
Elsheikh
TM.
The Bethesda System for Reporting Thyroid Cytopathology: an experience of 1,382 cases in a community practice setting with the implication for risk of neoplasm and risk of malignancy
.
Diagn Cytopathol
.
2012
;
40
(
5
):
399
403
.
39
Yang
J,
Schnadig
B,
Wasserman
PG.
Fine-needle aspiration of thyroid nodules: a study of 4703 patients with histologic and clinical correlations
.
Cancer Cytopathol
.
2007
;
111
(
5
):
306
315
.
40
Yassa
L,
Cibas
ES,
Benson
CB,
et al.
Long-term assessment of a multidisciplinary approach to thyroid nodule diagnostic evaluation
.
Cancer Cytopathol
.
2007
;
111
(
6
):
508
516
.

Author notes

The authors have no relevant financial interest in the products or companies described in this article.

Competing Interests

Presented in part at the Princeton Integrated Pathology Symposium: Head and Neck Pathology; February 9, 2014; Princeton, New Jersey.