Thyroid nodules are a common clinical presentation, with

Original Research Endocrine Surgery Thyroid Ultrasound-Guided Fine-Needle Aspiration Cytology Results: Observed Increase in Indeterminate Rate over the Past Decade Otolaryngology Head and Neck Surgery 2017, Vol. 156(4) 611 615 Ó American Academy of Otolaryngology Head and Neck Surgery Foundation 2017 Reprints and permission: sagepub.com/journalspermissions.nav DOI: 10.1177/0194599816688190 http://otojournal.org Amy M. Manning, MD 1, Huaitao Yang, MD, PhD 2, Mercedes Falciglia, MD 3, Jonathan R. Mark, MD 1, and David L. Steward, MD 1,3 Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. Abstract Objectives. To evaluate changes in distribution of reported thyroid nodule fine-needle aspiration (FNA) cytopathology results since implementation of the Bethesda classification and revised 2015 American Thyroid Association (ATA) guidelines for selecting nodules for biopsy. Study Design. Retrospective review. Setting. Tertiary academic medical center. Subjects and Methods. Evaluation of ultrasound (US) guided thyroid FNA by a single surgeon using 2015 ATA nodule selection criteria and Bethesda reporting on 211 thyroid nodules in a 1-year period (2015). Comparison is made to an earlier sample wherein any nodule.1 cm underwent US FNA with cytology reported prior to Bethesda consensus (2006). Results. The current cohort involved mostly women (79%); nodules ranged from 1 to 7 cm (mean 6 SEM, 2.4 6 0.07 cm). Mean 6 SEM age was 53.5 6 1.1 years. Bethesda reporting yielded 6% nondiagnostic, 57% benign, 3% malignant, and 34% indeterminate (27% atypia of undetermined significance [AUS]/follicular lesion of undetermined significance [FLUS], 4% follicular neoplasm [FN]/Hürthle neoplasm [HN], and 2% suspicious for malignancy [SFM]). The malignancy rate in indeterminate nodules was 26% (18% AUS/ FLUS, 33% FN/HN, and 80% SFM). Age, sex, or nodule size did not correlate with indeterminate cytology. The comparator sample of 447 nodules had significantly different distribution, with 7% nondiagnostic, 80% benign, 5% malignant, and 8% indeterminate (P \.00001). Conclusion. We observed a significantly increased proportion of indeterminate cytology and corresponding decrease in benign nodules compared with an earlier sample, predominately from an increase in AUS/FLUS. Multiple factors are likely involved, including selection of sonographically suspicious nodules for biopsy based upon 2015 ATA guidelines coupled with cytopathological interpretation by a new generation of cytopathologists trained in the era of Bethesda reporting; further study is required to make a definitive conclusion. Keywords thyroid nodule, FNA, cytology, Bethesda, ATA guidelines, ultrasound needle biopsy, thyroid cancer Received June 30, 2016; revised October 19, 2016; accepted December 16, 2016. Thyroid nodules are a common clinical presentation, with palpable nodules present in approximately 5% of women and 1% of men in iodine-sufficient areas of the world. 1 In addition, high-resolution ultrasound can detect incidental thyroid nodules in 19% to 68% of randomly selected individuals. The greatest concern is the presence of malignancy in thyroid nodules, and the gold standard for diagnosis is fineneedle aspirate biopsy (FNAB). Ultrasound guidance of FNAB has improved diagnostic accuracy by decreasing the frequency of nondiagnostic or inadequate samples. 2,3 Cytopathologic results of FNAB are used to guide the decision of whether surgery is indicated and the extent of the recommended surgery. 1 Department of Otolaryngology Head and Neck Surgery, University of Cincinnati (UC) College of Medicine, Cincinnati, Ohio, USA 2 Department of Pathology and Laboratory Medicine. University of Cincinnati (UC) College of Medicine, Cincinnati, Ohio, USA 3 Division of Endocrinology, Diabetes and Metabolism, Department of Internal Medicine. University of Cincinnati (UC) College of Medicine, Cincinnati, Ohio, USA This article was presented at the 2016 AAO-HNSF Annual Meeting and OTO EXPO; September 18-21, 2016; San Diego, California. Corresponding Author: David L. Steward, MD, Department of Otolaryngology Head and Neck Surgery, University of Cincinnati (UC) College of Medicine, 231 Albert Sabin Way, MSB 6504, Cincinnati, OH 45267-0528, USA. Email: stewardd@ucmail.uc.edu

612 Otolaryngology Head and Neck Surgery 156(4) Current cytopathology results are reported according to classifications established by the Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) guidelines published in 2009. 4 In 2007, the National Cancer Institutes (NCI) hosted the NCI Thyroid FNA State of the Science Conference in Bethesda, Maryland, the goal of which was to establish a uniform reporting system to improve communication between cytopathologists, surgeons, endocrinologists, and other health care providers regarding the malignant potential of thyroid nodules. Prior to the Bethesda guidelines, thyroid nodule FNAB results were reported as benign, malignant, nondiagnostic, or indeterminate, with the indeterminate category traditionally requiring surgery for definitive diagnosis. 5 The Bethesda guidelines further subcategorize the indeterminate category into atypia of undetermined significance (AUS)/follicular lesion of undetermined significance (FLUS), follicular neoplasm (FN) or suspicious for follicular neoplasm (SFN), and suspicious for malignancy (SFM). Each category carries a different likelihood of malignancy, 4,5 which can be used to better guide the decision of whether to operate. In addition, the advent of molecular genetic testing for indeterminate fine-needle aspiration (FNA) specimens, introduced in 2009, allows for the possibility of supplementary information regarding the likelihood of malignancy and thus an additional way to rule out or rule in malignancy prior to proceeding to the operating room. 6-8 This may also contribute to an increasing proportion of indeterminate results, as cytopathologists are aware of this new possible alternative to proceeding directly to surgery. In 2006, the American Thyroid Association (ATA) published guidelines for the management of thyroid nodules and differentiated thyroid cancer (DTC), which were revised in 2009 and again in 2015. 1,9,10 These guidelines discuss the use of ultrasound for clinical examination of thyroid nodules and have identified high-risk ultrasound features, including the presence of microcalcifications within the nodule, shape taller than wide, irregular margins, hypoechogenicity, extrathyroidal extension of the nodule, and suspicious appearance of cervical lymph nodes. Based on ultrasound features, the ATA has stratified risk of malignancy in thyroid nodules by sonographic appearance and has made recommendations regarding when FNAB is indicated. 1 Prior to 2006, at our institution, FNAB was recommended by our surgeon for any nodules.1 cm in diameter. To date, no investigation has occurred to our knowledge examining the effects of utilization of the Bethesda classification and the ATA guidelines for selection of nodules for biopsy based on size and sonographic appearance on the distribution of thyroid FNA cytology results. It is the investigators hypothesis that rates of nondiagnostic and malignant FNAB results would remain roughly equivalent, with higher frequency of indeterminate results and a concordant decrease in the number of benign results. To test this hypothesis, we compared a recent prospective cohort of patients undergoing ultrasound (US) guided FNAB by a single surgeon at an academic medical center during a 1-year period with an earlier cohort of patients having undergone FNAB prior to utilization of the ATA guidelines and Bethesda classification and prior to the availability of cytomolecular testing. 11 The cytopathologists reviewing the specimens changed between the time of collection of the initial cohort and the current cohort, such that cytopathologists involved in the current cohort completed their thyroid cytolopathology training after implementation of the Bethesda classification. Methods After obtaining approval from the University of Cincinnati Institutional Review Board, the authors prospectively enrolled consecutive subjects over a 1-year period in 2015. The senior author (D.L.S.) performed all US examinations and FNABs in a tertiary care, university practice setting. All specimens were reviewed by cytopathologists trained in the post-bethesda classification era. All subjects had a comprehensive ultrasonographic examination of the neck, which included the central and lateral compartments. Ultrasoundguided FNABs were performed as indicated by the 2015 ATA guidelines, 1 such that sonographically high-risk nodules were selected if.1 cm, intermediate- and low-risk nodules if.1.5 cm, and low-risk nodules if.2 cm. No sonographically benign simple cysts were biopsied. When multiple very low-risk nodules were present, the largest.2 cm and any high-, intermediate-, and/or low-risk nodules were biopsied as recommended in the ATA guidelines. Along with demographic information, the ultrasound results and cytology results were recorded, according to the Bethesda classification. 5 Surgery was recommended for all FNABs with malignant or suspicious for malignancy results. In the event of an AUS reading, the cytopathology was reviewed by a second cytopathologist. If there was disagreement between the first 2 cytopathologists interpretations, a third cytopathologist reviewed the case, and the majority opinion was reported as the final pathologic diagnosis. Patients with FN/Hürthle neoplasm (HN) and AUS/FLUS cytology results were offered the options of observation, repeated FNA with or without cytomolecular testing, or surgical excision, and the majority (70%) underwent surgical removal. In addition, surgery was performed for a subset of nodules that were benign by FNAB results but were large or symptomatic and those that were cytologically malignant or suspicious for malignancy. Final surgical pathology of resected thyroid nodules, where applicable, was recorded. FNAB cytopathology results of the 2015 group were then compared with the results of an earlier cohort who underwent US-guided FNAB between 2003 and 2005. 11 This earlier data set comprised patients who underwent US-guided FNAB for nodules.1 cm in diameter without specific consideration of sonographic risk pattern, with cytology results reported as nondiagnostic, benign, malignant, or indeterminate. The data of the2cohortswerecomparedviax 2 analysis. All ultrasound examinations and FNABs were performed by the same surgeon (D.L.S.) in both cohorts. The cytopathologists differed between the 2 cohorts, however, and the current cytopathology staff completed their training after implementation of the Bethesda classification.

Manning et al 613 Table 1. Distribution of FNA Biopsy Results in Past (2003-2005) and Recent (2015) Groups. Category Results 2003-2005 (n = 447), No. (%) 2015 (n = 211), No. (%) P Value Nondiagnostic 33 (7.4) 13 (6.1) Benign 357 (79.9) 120 (56.8) Malignant 20 (4.5) 6 (2.8) Indeterminate 37 (8.3) 72 (34.1) \.00001 AUS/FLUS 58 (27.4) FN/HN 9 (4.3) SFM 5 (2.4) Abbreviations: AUS, atypia of undetermined significance; FLUS, follicular lesion of undetermined significance; FN, follicular neoplasm; HN, Hürthle neoplasm; SFM, suspicious for malignancy. US-guided FNABs were performed on 211 nodules from 204 different patients in the current cohort. In total, 161 (79%) of patients were women, with patient age ranging from 18 to 85 years (mean 6 SEM, 53.4 6 1.06 years). Nodule size ranged from 1 to 7 cm (mean 6 SEM, 2.4 6 0.07 cm). Nodule size, patient sex, or patient age did not correlate with indeterminate cytology. The comparator cohort comprised 90% women and a total of 447 US-guided FNAB samples. Nondiagnostic specimens occurred in 13 of 211 specimens (6%), similar to the 7% rate reported in the prior data set. Bethesda reporting of the current sample resulted in 57% benign, 3% malignant, and 34% indeterminate (27% AUS/FLUS, 4% FN/HN, and 2% SFM). The comparator sample of 447 nodules had a significantly different distribution, with 80% benign, 5% malignant, and 8% indeterminate (P \.00001) (Table 1). Neither nodule size nor sonographic features were specifically reported in the prior data set. Final surgical pathology of the current data set (where available) yielded a malignancy rate of 26% for indeterminate nodules (16% of AUS/FLUS, 30% of FN/HN, and 80% of SFM). A single nodule that was benign by FNA cytology had a final malignant pathology result, consistent with the 0% to 3% expected false-negative rate of Bethesda reporting. Compared with the Bethesda-defined risk of malignancy of indeterminate nodules (5%-15% of AUS/FLUS, 15%-30% of FN/HN, and 60%-75% of SFM), 5 malignancy rates of both AUS/FLUS and FN/HN in our current data set are consistent but at the upper limits of the expected range; however, malignancy rates of SFM of the current data set are slightly higher (by 5%) than expected by Bethesda. Discussion We have observed a significant shift in the distribution of thyroid FNA biopsy results at our institution over the past 10 to 15 years. In that same time period, there have been multiple advances in the management of thyroid nodules, between the original ATA guidelines for management of thyroid nodules and DTC and its 2 subsequent revisions, 1,9,10 the Bethesda classification of thyroid FNAB result reporting, 4 and cytomolecular testing. 6-8 Each of these factors has altered our practice, from which nodules are biopsied to the reporting of these biopsy results to the final recommendation for or against surgery and the extent of surgery performed. Of specific clinical interest is the subset of FNAB results described as indeterminate. In 2015, Hirsch and colleagues 12 reported an increase in malignancy rate of final surgical pathology results in patients who had indeterminate FNAB cytology results who then underwent thyroidectomy, when comparing pre- and post-bethesda criteria implementation. This was despite a lower rate of thyroidectomy for indeterminate nodules, suggesting that use of the Bethesda category increases the diagnostic accuracy of indeterminate nodules. Previous groups have also noted a shift in their cytopathology reporting on the basis of Bethesda. Chen and colleagues 13 noted a decrease in FN/HN nodules with an overall increase in the proportion of indeterminate nodules and a decrease in benign nodules. They expressed concern that the AUS/FLUS category would be overused in lieu of a more definitive diagnosis. A study by Richmond et al, 14 however, noted an overall increase in the proportion of FNAs classified as benign and a decrease in the proportion making up the indeterminate group, when pre-bethesda cytology specimens were retrospectively reviewed and reclassified according to the Bethesda criteria. Similarly, Theoharis and colleagues 15 also noted a significant reduction in indeterminate results and a concomitant increase in benign results. Clearly, further investigation is warranted to better characterize any shift in cytology reports when comparing pre- and post-bethesda implementation. Our data indicate a significant shift in the distribution of FNAB cytology results at our institution, with striking differences observed in rates of benign FNAB results (57% in current data set vs 80% in historic set) as well as the rate of indeterminate results (34% vs 8%). The rates of nondiagnostic results were similar in both the current and historic data sets (6% vs 7%, respectively), as were the rates of nodules called malignant via FNAB (3% vs 5%, respectively) or suspicious for malignancy or malignant (5% vs 5%, respectively). A segment of the decrease in the proportion of benign results could be due to improved selection criteria for performing FNAB in the first place, as implementation of the ATA guidelines should theoretically decrease the rate of FNAB performed on benign-appearing nodules. This indicates a possible selection bias based on ultrasound examination and recommendations by the ATA guidelines. In addition, subclassification of the indeterminate category has allowed for division of this category into 3 subcategories (AUS/FLUS, FN/HN, and SFM), allowing for stratification of risk of malignancy within the indeterminate category. As previously noted by Chen and colleagues, 13 our data suggest that perhaps pathologists are using this category at a much higher rate than they were prior to the standardization and subclassification of the category provided by the Bethesda guidelines. However, as the

614 Otolaryngology Head and Neck Surgery 156(4) rate of malignancy in the indeterminate category in our data set is congruent with that expected by the Bethesda report, 4 this would suggest that the salient contribution to the shift we noted in distribution of FNA cytology may be selection of sonographically suspicious nodules for biopsy. A prior study by Sacks et al 16 retrospectively compared thyroid FNA cytopathology, rate of surgery, and malignancy rate in pre- vs postintroduction of the Afirma gene expression classifier assay. This group noted a significantly increased rate of indeterminate cytology and decreased rate of benign cytology when comparing FNABs done prior to and following the widespread availability of the assay at their institution, without a significant change in the rate of thyroidectomy or in the rate of malignancy on final surgical pathology. While the authors note that their data are correlative, as are the observed data in our current study, they raise the possibility that the availability of cytomolecular testing may prompt a cytopathologist to hedge by reporting an indeterminate result when a benign diagnosis is uncertain. This pattern of shift in cytopathology results is possibly consistent with that seen in our data; it is a limitation of the retrospective nature of our study that we cannot discriminate among the multiple potential causes of the observed pattern; it is likely that several factors (implementation of Bethesda reporting, ATA ultrasonography guidelines, and availability of cytomolecular testing) all play a role. Another potential confounder of the observed shift in FNA biopsy results at our institution, as well as a limitation of this study, is a change in the cytopathologists between the 2 time periods. It has been postulated by others that cytopathologists may use the indeterminate category when they are unsure of diagnosis. 13 In addition, previous research published by our institution 17 indicates that high-volume cytopathologists, who read.50 thyroid FNABs per year, are less likely than lowervolume cytopathologists to report indeterminate cytology and more likely to classify FNAs as benign. The cytopathologists at our institution in the recent 2015 group are high-volume cytopathologists who completed training in the time since the implementation of the Bethesda criteria. This supports the notion that cytopathologists at are institution are not likely to be simply relying on the indeterminate category in lieu of providing a more definite clinical diagnosis. In addition, it is the policy at our institution s pathology department for a second or sometimes third cytopathologist to review cytopathology samples initially classified as AUS, in an effort to reduce erroneous use of the AUS category. Conclusion Utilization of ATA thyroid nodule selection criteria for US FNA, Bethesda consensus cytopathology reporting, and availability of cytomolecular genetic testing have significantly altered the diagnosis and management of cytologically indeterminate thyroid nodules in the past decade. Our data show that the distribution of benign and indeterminate cytology reporting has also changed significantly at our institution, predominately from an increase in the AUS/ FLUS category and a decrease in benign category. The proportion of this effect due to each of these factors requires further study, but despite an increase in the indeterminate category, our data are consistent with expected malignancy rates in each of the indeterminate subcategories, suggesting that selection of nodules for US FNAB based on ATA 2015 guidelines of sonographic risk pattern assessment rather than nodule size alone is a factor. Acknowledgments We thank Rebecca Reinert, clinical research coordinator in the Department of Otolaryngology Head and Neck Surgery, University of Cincinnati, for her assistance with data collection. Author Contributions Amy M. Manning, analysis and interpretation of data, drafting and revising of the manuscript, final approval of the version to be published, agreement to be accountable for all aspects of the work; Huaitao Yang, analysis and interpretation of data, revising of the manuscript, final approval of the version to be published, agreement to be accountable for all aspects of the work; Mercedes Falciglia, analysis and interpretation of data, revising of the manuscript, final approval of the version to be published, agreement to be accountable for all aspects of the work; Jonathan R. Mark, conception and design of the work, revision of the manuscript, final approval of the version to be published; David L. Steward, conception and design of work, acquisition, analysis and interpretation of data, revision of manuscript, final approval of version to be published, agreement to be accountable for all aspects of the work. Disclosures Competing interests: David L. Steward, Astra Zeneca, Inspire Medical Systems, Rosetta Genomics research grant. Sponsorships: None. Funding source: None. References 1. 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;26:1-133. 2. Carmeci C, Jefferey RB, McDougall IR, et al. Ultrasound-guided fine-needle aspiration biopsy of thyroid masses. Thyroid. 1998;8: 283-289. 3. Danese D, Sciacchitano S, Farsetti A, et al. Diagnostic accuracy of conventional versus sonography-guided fine-needle aspiration biopsy of thyroid nodules. Thyroid. 1998;8:15-21. 4. Cibas ES, Ali SZ. The Bethesda system for reporting cytopathology. Am J Clin Pathol. 2009;132:658-665. 5. 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;36:425-437. 6. Nikiforov YE, Steward DL, Robinson-Smith TM, et al. Molecular testing for mutations in improving the fine-needle

Manning et al 615 aspiration diagnosis of thyroid nodules. J Clin Endocrinol Metab. 2009;94:2092-2098. 7. Alexander EK, Kennedy GC, Baloch ZW, et al. Preoperative diagnosis of benign thyroid nodules with indeterminate cytology. N Engl J Med. 2012;367:705-715. 8. Ferris RL, Baloch Z, Bernet V, et al. American Thyroid Association Surgical Affairs Committee. Thyroid. 2015;25:760-768. 9. Cooper DS, Doherty GM, Haugen BR, et al. American Thyroid Association Guidelines Taskforce 2006 management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid. 2006;16:109-142. 10. Cooper DS, Doherty GM, Haugen BR, et al. 2009 Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid. 2009;19:1167-1214. 11. Bhatki AM, Brewer B, Robinson-Smith T, Nikiforov Y, Steward DL. Adequacy of surgeon-performed ultrasound-guided thyroid fine-needle aspiration biopsy. Otolaryngol Head Neck Surg. 2008;139:27-31. 12. Hirsch D, Robenshtok E, Bachar G, Braslavsky D, Benbassat C. The implementation of the Bethesda system for reporting thyroid cytopathology improves malignancy detection despite lower rate of thyroidectomy in indeterminate nodules. World J Surg. 2015;39:1959-1965. 13. Chen JC, Pace SC, Chen BA, Khiyami A, McHenry CR. Yield of repeat fine-needle aspiration biopsy and rate of malignancy in patients with atypia or follicular lesion of undetermined significance: The impact of the Bethesda System for Reporting Thyroid Cytopathology. Surgery. 2012;152:1037-1044. 14. Richmond BK, O Brien BA, Mangano W, Thompson S, Kemper S. The impact of implementation of the Bethesda System for Reporting Thyroid Cytopathology on the surgical treatment of thyroid nodules. Am Surg. 2012;78:706-710. 15. Theoharis C, Adeniran AJ, Roman S, Sosa JA, Chhieng D. The impact of implementing the Bethesda system for reporting of thyroid FNA at an academic center. Diagn Cytopathol. 2013;41:858-863. 16. Sacks WL, Bose S, Zumsteg ZS, et al. Impact of Afirma gene expression classifier on cytopathology diagnosis and rate of thyroidectomy. Cancer Cytopathol. 2016;124:722-728. 17. Houlton JJ, Sun GH, Fernandez N, Zhai QJ, Lucas F, Steward DL. Thyroid fine-needle aspiration: does case volume affect diagnostic yield and interpretation? Arch Otolaryngol Head Neck Surg. 2011;137:1136-1139.