Clinical Endocrinology (2014)
doi: 10.1111/cen.12692
ORIGINAL ARTICLE
Value of sonographic features in predicting malignancy in thyroid nodules diagnosed as follicular neoplasm on cytology Chiaw-Ling Chng*,†, Tom R. Kurzawinski‡ and Tim Beale§ *Department of Endocrinology, Singapore General Hospital, Singapore, †Department of Endocrinology, Royal Free Hospital, ‡Centre for Endocrine Surgery University College Hospital and §Department of Radiology, University College Hospital, London, UK
Introduction Summary Background The cytological diagnosis of follicular neoplasm (Thy3F) remains a diagnostic challenge. The main aim of this study was to stratify the risk of malignancy in thyroid nodules diagnosed as Thy3F on cytology (Thy3F) using thyroid imaging reporting and data system (TIRADS). Methods A database of thyroid nodules with Thy3F cytological results from ultrasound-guided FNA (US-FNA) between January 2007 and March 2014 was studied retrospectively. Information on patient demographics, ultrasound characteristics and final histology of the nodules was collated. The number of suspicious US features of each thyroid nodule was counted based on TIRADS. The malignancy rate of each of the TIRADS category was also calculated based on the final histological outcomes of the nodules and compared to that calculated using a recently proposed thyroid malignancy risk prediction model. Results The overall malignancy rate of Thy3F cytology was 243%. There were significantly higher percentages of malignant nodules with irregular margins (200% vs 0%, P = 0000), hypoechogenicity (743% vs 514%, P = 0013) and taller-than-wide morphology (171% vs 09%, P = 0001) when compared to benign nodules. The risk of malignancy increased with advancing TIRADS score: TIRADS 4A (143%), TIRADS 4B (231%), TIRADS 4C (875%) and TIRADS 5 (100%). The malignancy rate calculated using the prediction model similarly increased with advancing TIRADS score: TIRADS 4A (62%), TIRADS 4B (325%), TIRADS 4C (799%) and TIRADS 5 (90%). Conclusion Thyroid nodules with TIRADS scores 4C and 5 should be considered for single definitive surgery in view of the high malignant rate. (Received 5 November 2014; returned for revision 21 November 2014; finally revised 1 December 2014; accepted 4 December 2014)
Correspondence: Chiaw-Ling Chng, Department of Endocrinology, Level 3 The Academia, 20 College Road, Singapore 169856, Singapore. Tel.: +65-63265916; Fax: +65-62273576; E-mail: chng.chiaw.ling@ sgh.com.sg © 2014 John Wiley & Sons Ltd
Ultrasound-guided fine needle aspiration (US-FNA) has an established role in the evaluation of thyroid nodules. However, an indeterminate cytological diagnosis from the FNA remains a significant management challenge to clinicians. Thy categories (1–5) have been introduced in the United Kingdom as numerical definitions at the end of the descriptive report to clarify the management of thyroid lesions following from the FNA cytology report.1 They are similar but not identical to the US Bethesda Thyroid Cytology Reporting Categories (TBSTCR).2 Each of the categories has a definition and a recommended management. For example, Thy3 category is subdivided into: Thy3A and Thy3F. In most instances, a repeat ultrasound with or without FNA is recommended for a Thy3a cytology result.3 Thy3F is used when a follicular neoplasm is suspected. The histological possibilities may include a hyperplastic nodule, follicular adenoma, follicular carcinoma or a follicular variant of papillary thyroid carcinoma. These cannot be distinguished on cytology alone and surgery (hemithyroidectomy) is required for diagnosis with a completion of thyroidectomy recommended for those with a confirmed thyroid cancer.3 This treatment approach is not ideal, as the risk of malignancy from a Thy3F cytological diagnosis is between 15 and 30%,1 and the majority of these patients may not require surgery. In addition, algorithms to identify those at highest risk of malignancy may allow the surgeon to perform a single definitive operation, saving time, reducing patient morbidity and reducing costs. It has been reported that certain sonographic features of a thyroid nodule are associated with an increased likelihood of malignancy, although no single predictor has been found to have a high positive predictive value for cancer. Hence, guidelines with various combinations of ultrasound features that are both sensitive and specific for predicting the presence of cancer have been defined in several studies1,3–11 Thyroid imaging reporting and data system (TIRADS) was first proposed by Harvath et al.6 This scoring system was based on breast imaging reporting and data system (BIRADS) defined by the American College of Radiology12 and stratifies the risk of malignancy of thyroid nodules based on ten ultrasound characteristics. However, this system was considered to be too complex for implementation in clinical practice and have as modified by several other authors.9,11,13,14 1
2 C.-L. Chng et al. Kwak et al. demonstrated that the following US features showed a significant association with malignancy: solid component, marked hypoe-chogenicity, microlobulated or irregular margins, microcalcifications and taller-than-wide shape.9 They scored TIRADS as 4A (one suspicious feature), 4B (two suspicious features), 4C (three or four suspicious features) and 5 (five suspicious features) and demonstrated the fitted probability and risk of malignancy increased as the number of suspicious ultrasound features increased.9 The advantages of this scoring system are its simplicity and ease in implementation in clinical practice. A recent study by the same group proposed a thyroid cancer risk prediction model using features of thyroid nodules found on US and was validated based on multicentre retrospective data.15 The main aim of this study was to stratify the risk of malignancy in thyroid nodules diagnosed as follicular neoplasms on cytology (Thy3F) using TIRADS. We also compared the risk of malignancy calculated based on the prediction model and the actual malignancy rate in these nodules in each TIRADS category.
Materials and methods A database of thyroid nodules with Thy3F cytological results from US-FNA performed by a single radiologist between January 2007 and March 2014 was studied retrospectively. The data were collected from three different institutions where the radiologist routinely performs head and neck US and FNAs. All the US were performed and reviewed by the same radiologist with extensive experience (close to 20 years) in head and neck radiology. Ultrasounds were performed using a Toshiba Aplio 500 (Toshiba America Medical Systems, Inc., Tustin, CA, USA) and Philips HDI 5000 (Philips Healthcare, DA Best, The Netherlands). US-FNA was performed only on thyroid nodules with suspicious US features. FNAs were performed free hand using a 25-gauge needle using nonaspiration technique, under direct US guidance. The direct smears were air-dried, sent to the respective pathology laboratory on the same day, stained with May-Grunwald–Giemsa methods and interpreted by cytopathologists with experience in thyroid cytology. Information on patient demographics, ultrasound characteristics and final histology of the nodules was collated by the main author of this study, such that the performing radiologist was blinded with respect to the final histology of the nodules. Static ultrasound images and original ultrasound reports of the nodules biopsied were retrospectively reviewed for the following features: size (based on the largest diameter on US), solid component, hypo-echogenicity, microlobulated or irregular margins, taller-than-wide shape (ratio of anteroposterior to transverse diameter ≥1), the presence of microcalcifications and increased intranodular vascularity obtained via routine colour Doppler assessment. All relevant parameters were compared between the benign and malignant histological outcomes of these nodules. The number of suspicious US features of each thyroid nodule was also counted based on the TIRADS scores proposed by Kwak et al.9 In this scoring system, a solid component, hypoechogenicity, microlobulated or irregular margins, microcalcifications and a taller-than-wide shape were counted as suspicious
US features. Thyroid nodules without suspicious features were classified as TIRADS category 3. Thyroid nodules having one, two, three or four or five suspicious features were classified as categories 4A, 4B, 4C or 5, respectively. On the basis of the recent thyroid cancer risk prediction model proposed by Kwak et al.,15 we applied a score of 2 to hypo-echogenicity, a score of 6 to marked hypo-echogenicity, a score of 1 to a taller-thanwide shape, a score of 5 to microlobulated or irregular margins and a score of 2 to microcalcifications. The total risk score of each thyroid nodule was calculated by summing the scores of each suspicious US features, and the percentage chance of malignancy was obtained according to the total risk score. The malignancy rate of each of the TIRADS category was also calculated based on the final histological outcomes of the nodules. The thyroid cancer risk is calculated based on the prediction model and the actual malignancy rate in these nodules diagnosed as follicular neoplasm on cytology was compared. The study protocol complies with the local ethical guidelines. The data were analysed using SPSS version 220 (SPSS Inc., Chicago, IL, USA). Baseline continuous data were expressed in mean + standard deviation, and categorical data were expressed as percentages. Univariate associations between the presence of cancer and discrete variables were evaluated using chi-squared test. Independent Student’s t-test was used to compare between categorical and continuous variables. P < 005 was considered statistically significant.
Results There were 497 FNAs with Thy 3, 4 and 5 cytological results performed on 465 patients between January 2007 and April 2014. There were 239 (481%) cytological results that were classified as Thy3F (suspicious of follicular neoplasm). Ultrasound and histological correlation were available for comparison in 144 cases and constitute the results of this study. Four incidental micropapillary thyroid cancers were excluded from the study. The overall malignancy rate of Thy3F cytology was 243%. The distribution of histology results in these 144 cases is presented in Table 1. The majority of nodules with benign histology were follicular adenoma (523%). Fifteen (429%) of the nodules with malignant histology were papillary thyroid cancer, of which four were follicular variants of papillary thyroid cancer. Sixteen (457%) of the malignant nodules were follicular thyroid cancer. Comparison of patient demographics and sonographic characteristics between thyroid nodules with benign and malignant histology is presented in Table 2. As only the solid component was biopsied in all thyroid nodules in our practice, this sonographic feature was not compared between benign and malignant nodules in our study. There was no significant differences in the age or gender distribution between patients with malignant or benign nodules (P = 0061 and P = 0153, respectively). In the comparison of sonographic characteristics, there were significantly higher percentages of malignant nodules with irregular margins (200% vs 0%, P = 0000), hypo-echogenicity (743% vs 514%, P = 0013) and taller-than-wide morphology (171% vs 09%, P = 0001) when compared to benign nodules. There was © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
Predictors of malignancy in thyroid follicular neoplasms 3 Table 1. Histological diagnosis in the 144 cases
120 100
N (%)
Benign Follicular adenoma Hurthle cell adenoma Multinodular goitre/Nodular hyperplasia Hyperplastic nodule Intrathyroid parathyroid adenoma Malignant Papillary thyroid cancer Follicular thyroid cancer Both papillary and follicular thyroid cancer Medullary thyroid cancer
109 (757%) 57 3 21 27 1 35 (243%) 15 16 3 1
Table 2. Comparison of patient demographics and sonographic characteristics between thyroid nodules with benign and malignant histology Benign (n = 109) Demographic characteristics Mean age (years) 441 149 Male 23 (676%) Female 86 (782%) Sonographic characteristics Nodule size (cm) 28 15 Irregular margins 0 Hypo-echogenicity 56 (514%) Taller-than-wide 1 (09%) morphology Presence of 5 (46%) microcalcifications Intranodular 55 (505%) vascularity
Malignant (n = 35)
P-value
496 157 11 (324%) 24 (218%)
0061 0153
32 15 7 (200%) 26 (743%) 6 (171%)
0222 0000* 0013* 0001*
3 (86%)
0302
17 (486%)
0500
*Denotes statistical significant results. Table 3. Malignancy rate based on TIRADS category in the 144 cases with diagnosis of follicular neoplasm on cytology TIRADS category
No of cases (n)
Benign (n, %)
Malignant (n, %)
Malignant rate (%)
4A 4B 4C 5 Total
56 78 8 2 144
48 60 1 0 109
8 18 7 2 35
143 231 875 100 243
(440%) (55%) (09%) (0%)
(229%) (514%) (200%) (57%)
no significant difference in the percentages of malignant nodules with microcalcifications or increased intranodular vascularity when compared to benign nodules. The malignancy rate based on TIRADS categories is shown in Table 3. There were no nodules scored as TIRADS 3 in our study. The risk of malignancy increased with advancing TIRADS score: TIRADS 4A (143%), TIRADS 4B (231%), TIRADS 4C (875%) and TIRADS 5 © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
100
Malignancy rate (%)
Histological diagnosis
90
87·5 79·9
80 Malignant rate based on histology
60 32·5
40 20 0
14·3 6·2 4A
Malignancy rate based on risk predicon model
23·1
4B
4C
5
TIRADS category Fig. 1 Graph comparing the malignancy rate of various TIRADS category calculated by the histology results of our study and malignancy rate calculated using the risk prediction model proposed by Kwak et al.15
(100%) (Table 3). The malignancy rate of each TIRADS category (calculated based on the recently proposed prediction model)15 similarly increased with advancing TIRADS score: TIRADS 4A (62%), TIRADS 4B (325%), TIRADS 4C (799%) and TIRADS 5 (90%). A comparison of malignancy risk based on the risk prediction model and histological outcomes in each TIRADS category is shown in Fig. 1.
Discussion The diagnosis of follicular neoplasm (Thy3F) remains a diagnostic dilemma. The distinction of follicular adenoma from follicular carcinoma requires the presence of capsular or vascular invasions but such characteristics cannot be assessed on cytological evaluation. In our study, the majority of the nodules with Thy3F cytology were benign, which were mostly follicular adenomas on histology. Although there were a significant proportion of papillary thyroid cancers among the malignant nodules, four of these were follicular variants of papillary thyroid cancers. It is well known that the cytological diagnosis of this entity is often challenging to make, and it can be misdiagnosed as follicular neoplasms on FNAC.16 The current British Thyroid Association guidelines recommends repeat FNA for Thy3A cytology results, but a diagnostic hemithyroidectomy remains the next step in the evaluation of Thy3F cytology outcomes.3 The overall malignancy rate of nodules with Thy3F cytology result was 243% in our study, which is similar to the malignancy rate for this cytological category based on the current guidelines.1 Hence, the fact that majority of these cases have benign disease on postoperative histology justifies the effort to better select candidates for surgery. Ultrasound is an important diagnostic tool in predicting thyroid malignancy and selecting thyroid nodules that should be evaluated by FNA. Known suspicious US features include hypo-echogenicity, microlobulated or irregular margins, microcalcifications and a taller-than-wide shape. Although no single US feature can reliably predict the risk of malignancy, a combination of these features is known to provide better
4 C.-L. Chng et al. diagnostic accuracy17,18 In our study, there was higher percentages of malignant nodules with irregular margins, hypo-echogenicity and taller-than-wide morphology, although size of nodule, the presence of microcalcifications and intranodular vascularity were not significantly different between the two groups. In particular, irregular margins have the highest positive predictive value for malignancy (100%) as all the nodules that exhibit this sonographic characteristic were malignant. This US feature was also the strongest predictor of malignancy in the study by Maia et al., on 80 patients with indeterminate thyroid cytology who were surgically treated at a single centre.19 The ‘taller-than-wide’ morphology was first proposed by Kim et al., as an US feature of malignancy, which was based on the fact that growth of most benign nodules has been found to remain within normal tissue planes, whereas malignant nodules grow across normal tissue planes.8 This sonographic feature was also found to have the highest diagnostic odds ratio among all the analysed ultrasound features in predicting malignancy by a meta-analysis by Brito et al.20 The same study also confirmed that nodule size is not an accurate predictor of thyroid cancer across different cut-offs, a finding that holds true across all the various cytology subgroups.20 Although the presence of microcalcifications and increased intranodular vascularity were found to be predictive of malignancy in some studies,20,21 these results were not reproduced in others.19,22 There was no significant difference in the age between subjects with benign or malignant nodules. Age was also not a predictor of malignancy in several studies specific to the cytological diagnosis of follicular neoplasms.23–25 There was also no male or female predominance in the malignant nodules, even though the incidence of thyroid nodules is higher in women.26 Thyroid imaging reporting and data system (TIRADS) was introduced to improve communication between radiologists and physicians through a standardized reporting format.6 As its introduction many variations of TIRADS have been proposed, some of them including additional US features such as elastography and vascularity criteria.7,9–11,13,14 The TIRADS used in our study is simple, convenient, accurate, and stratifies the malignancy risks. This TIRADS predicts malignancy risks according to just the number of suspicious US features. The malignancy rate in our study sample of thyroid nodules with Thy3F cytology results increases with increasing TIRADS score and reached 100% when all 5 suspicious features were present. A recent study, which assessed data from 136 indeterminate thyroid nodules, for which histopathological, cytological and ultrasound data were available, classified the nodules according to both the Bethesda system and a version of TIRADS that included vascularity criteria.14 On the basis of their findings, they suggested that for patients whose nodules have a TIRADS score of 4B or 5 and also have a Bethesda classification of either IV or V it may be advisable to proceed directly to definitive surgery on account of the high malignancy rates (700–857%) that are associated with these features.14 Our results concur with their findings, although the TIRADS system employed in our study differs from their study where nodule vascularity was not a criteria, and we used the Thy system for cytological classification instead
of the Bethesda system. However, results of both studies imply the higher risk of malignancy in nodules that are suspicious of follicular neoplasm which have more suspicious ultrasound features. Although TIRADS has been demonstrated to be accurate in predicting thyroid cancer risks in several studies, the system proposed in the studies lacked reproducibility, because each was developed based on the individual institution. This lack of uniformity with regards to the implementation of TIRADS has resulted in different reported values for its accuracy. The main drawback of the TIRADS proposed by Kwak et al., (used in our study) was that each US feature is considered with the same weight without considering different probabilities of malignancy for each US feature.9,15 The same authors subsequently designed a simple diagnostic prediction model using US features of thyroid nodules based on a multicentre data collected retrospectively which took this limitation into account.15 The results were also reproducible when used by less experienced radiologists.27 We applied this malignant risk prediction model to our study and showed increasing malignancy rates with increasing TIRADS score. The malignancy rates obtained for each TIRADS category were similar to the malignancy rate based on the histological outcome of these nodules (Fig. 1). Notably, the risk of malignancy increases exponentially in TIRADS 4C and 5. Our results suggest reproducibility of this malignant risk prediction using the model suggested by Kwak et al.,15 and its potential utility in the clinical management of thyroid nodules with Thy3F cytological outcomes. The main limitation of our study is the retrospective design and small numbers. Forty per cent of our patients with Thy3F cytology did not have histological results. The main reasons were refusal to surgery and lost to follow-up. In two patients, the decision was not to proceed to surgery due to the unrelated terminal diseases. In United Kingdom, a diagnostic hemithyroidectomy is recommended as the next step in the management of thyroid nodules with Thy3F cytological results based on the British Thyroid Association guidelines.3 Hence, a selection bias for surgery is unlikely as all the patients with Thy3F cytological results in the study were counselled for diagnostic hemithyroidectomy as a standard practice by their referring doctors, and this decision was not based on the ultrasound characteristics of the nodule. The strength of our study was that the ultrasound examinations were performed by the same radiologist, which reduced diagnostic variability. However, the TIRADS and risk prediction model used in our study is simple and has shown good reproducibility among less experienced radiologists, allowing more widespread use among clinicians who may not be specialist radiologists. The results of our study demonstrated that nodules with Thy3F cytology results and TIRADS score of 4C and 5 have very high risk of malignancy, and these patients should be considered for single definitive surgery. However, the absolute number of nodules exhibiting three or more suspicious US features in our study is small, with most nodules demonstrating one or two suspicious characteristics. Nonetheless, the malignancy risk of the nodules with one or two suspicious sonographic features is non-negligible (Fig. 1) and should be communicated to the patient when planning treatment. © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
Predictors of malignancy in thyroid follicular neoplasms 5 US-guided core-needle biopsy (US-CNB) provides a larger tissue sample that retains its cellular architecture and may enable a more precise histologic diagnosis.28 However, this procedure is not commonly used in the routine assessment of thyroid nodules, due to its perceived higher risk of complications and technical difficulties in smaller lesions.29 Recent evidence suggests the use of US-CNB to be an effective complementary diagnostic tool in the further evaluation of thyroid nodules which are indeterminate on initial FNA cytology.30,31 Thus, the application of the TIRADS score to thyroid nodules with Thy3F cytological results may help to identify selected cases where an US-CNB may be useful. More prospective studies, exploring the complementary role of US-CNB or combining new molecular diagnostic testing32 with this US-based malignant risk stratification model may help to delineate the malignancy risk of nodules with less than three suspicious US features, or further enhance the accuracy in predicting malignancy in nodules with TIRADS scores of 4C and 5 to reach 100%.
7
8
9
10
11
12
Conclusion The Thyroid imaging reporting and data system (TIRADS) is simple to use and accurate in malignant risk stratification of thyroid nodules diagnosed as follicular neoplasms (Thy3F) on cytology. Thyroid nodules with TIRADS scores 4C and 5 should be considered for single definitive surgery in view of the high malignant rate.
13
14
Disclosure statement The authors declare that no competing financial interests exist.
15
References 1 British Thyroid Association. (2009) Royal College of Pathologist: Guidelines for reporting of thyroid cytology specimens. Available at http://www.rcpath.org/ Resources/RCPath/Migrated%20 Resources/Documents/G/g089guidanceonthereportingofthyroidcytologyfinal.pdf. Accessed September 19, 2014 2 Cibas, E.S., Ali, S.Z. & Conference, N.C.I.T.F.S.o.t.S. (2009) The Bethesda System For Reporting Thyroid Cytopathology. American Journal of Clinical Pathology, 132, 658–665. 3 Perros, P., Boelaert, K., Colley, S. et al. (2014) Guidelines for the management of thyroid cancer. Clinical Endocrinology (Oxf), 81 (Suppl 1), 1–122. 4 Gharib, H., Papini, E., Paschke, R. et al. (2010) American Association of Clinical Endocrinologists, Associazione Medici Endocrinologi, and European Thyroid Association Medical guidelines for clinical practice for the diagnosis and management of thyroid nodules: executive summary of recommendations. Endocrine Practice, 16, 468–475. 5 American Thyroid Association Guidelines Taskforce on Thyroid, N., Differentiated Thyroid, C., Cooper, D.S., Doherty, G.M. et al. (2009) Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid, 19, 1167–1214. 6 Horvath, E., Majlis, S., Rossi, R. et al. (2009) An ultrasonogram reporting system for thyroid nodules stratifying cancer risk for © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
16
17
18
19
20
21
22
clinical management. Journal of Clinical Endocrinology and Metabolism, 94, 1748–1751. Hambly, N.M., Gonen, M., Gerst, S.R. et al. (2011) Implementation of evidence-based guidelines for thyroid nodule biopsy: a model for establishment of practice standards. AJR. American Journal of Roentgenology, 196, 655–660. Kim, E.K., Park, C.S., Chung, W.Y. et al. (2002) New sonographic criteria for recommending fine-needle aspiration biopsy of nonpalpable solid nodules of the thyroid. AJR. American Journal of Roentgenology, 178, 687–691. Kwak, J.Y., Han, K.H., Yoon, J.H. et al. (2011) Thyroid imaging reporting and data system for US features of nodules: a step in establishing better stratification of cancer risk. Radiology, 260, 892–899. Park, J.Y., Lee, H.J., Jang, H.W. et al. (2009) A proposal for a thyroid imaging reporting and data system for ultrasound features of thyroid carcinoma. Thyroid, 19, 1257–1264. Russ, G., Bigorgne, C., Royer, B. et al. (2011) The Thyroid Imaging Reporting and Data System (TIRADS) for ultrasound of the thyroid. Journal de Radiologie, 92, 701–713. American College of Radiology, B.-R.C. (2003) ACR BI-RADSâultrasound. In: V.A. Reston ed. ACR BI-RADS breast imaging and reporting data system: breast imaging atlas, 4th edn. American College of Radiology, Reston, Virginia, 1–86. Russ, G., Royer, B., Bigorgne, C. et al. (2013) Prospective evaluation of thyroid imaging reporting and data system on 4550 nodules with and without elastography. European Journal of Endocrinology, 168, 649–655. Maia, F.F., Matos, P.S., Pavin, E.J. et al. (2014) Thyroid imaging reporting and data system score combined with Bethesda system for malignancy risk stratification in thyroid nodules with indeterminate results on cytology. Clinical Endocrinology (Oxford), doi: 10.1111/cen.12525. [Epub ahead of print]. Kwak, J.Y., Jung, I., Baek, J.H. et al. (2013) Image reporting and characterization system for ultrasound features of thyroid nodules: multicentric Korean retrospective study. Korean Journal of Radiology, 14, 110–117. Manimaran, D., Karthikeyan, T.M., Khan, D.M. et al. (2014) Follicular variant of papillary thyroid carcinoma: cytological indicators of diagnostic value. Journal of Clinical and Diagnostic Research, 8, 46–48. Moon, W.J., Jung, S.L., Lee, J.H. et al. (2008) Benign and malignant thyroid nodules: US differentiation–multicenter retrospective study. Radiology, 247, 762–770. Tutuncu, Y., Berker, D., Isik, S. et al. (2014) The frequency of malignancy and the relationship between malignancy and ultrasonographic features of thyroid nodules with indeterminate cytology. Endocrine, 45, 37–45. Maia, F.F., Matos, P.S., Pavin, E.J. et al. (2011) Value of ultrasound and cytological classification system to predict the malignancy of thyroid nodules with indeterminate cytology. Endocrine Pathology, 22, 66–73. Brito, J.P., Gionfriddo, M.R., Al Nofal, A. et al. (2014) The accuracy of thyroid nodule ultrasound to predict thyroid cancer: systematic review and meta-analysis. Journal of Clinical Endocrinology and Metabolism, 99, 1253–1263. Gul, K., Ersoy, R., Dirikoc, A. et al. (2009) Ultrasonographic evaluation of thyroid nodules: comparison of ultrasonographic, cytological, and histopathological findings. Endocrine, 36, 464–472. Dutta, S., Thaha, M.A. & Smith, D.M. (2011) Do sonographic and cytological features predict malignancy in cytologically
6 C.-L. Chng et al.
23
24
25
26 27
indeterminate thyroid nodules? Annals of the Royal College of Surgeons of England, 93, 361–364. Sorrenti, S., Trimboli, P., Catania, A. et al. (2009) Comparison of malignancy rate in thyroid nodules with cytology of indeterminate follicular or indeterminate Hurthle cell neoplasm. Thyroid, 19, 355–360. Castro, M.R., Espiritu, R.P., Bahn, R.S. et al. (2011) Predictors of malignancy in patients with cytologically suspicious thyroid nodules. Thyroid, 21, 1191–1198. Calo, P.G., Medas, F., Santa Cruz, R. et al. (2014) Follicular nodules (Thy3) of the thyroid: is total thyroidectomy the best option? BMC Surgery, 14, 12. Eng, C.Y., Quraishi, M.S. & Bradley, P.J. (2010) Management of Thyroid nodules in adult patients. Head & Neck Oncology, 2, 11. Ko, S.Y., Lee, H.S., Kim, E.K. et al. (2014) Application of the Thyroid Imaging Reporting and Data System in thyroid
28
29 30
31
32
ultrasonography interpretation by less experienced physicians. Ultrasonography, 33, 49–57. Screaton, N.J., Berman, L.H. & Grant, J.W. (2003) US-guided core-needle biopsy of the thyroid gland. Radiology, 226, 827– 832. Lo Gerfo, P. (1994) The value of coarse needle biopsy in evaluating thyroid nodules. Thyroidology, 6, 1–4. Nasrollah, N., Trimboli, P., Guidobaldi, L. et al. (2013) Thin core biopsy should help to discriminate thyroid nodules cytologically classified as indeterminate. A new sampling technique. Endocrine, 43, 659–665. Choi, W.J. & Baek, J.H. (2014) Role of core needle biopsy for patients with indeterminate, fine-needle aspiration cytology. Endocrine, 45, 1–2. Yip, L. (2014) Molecular diagnostic testing and the indeterminate thyroid nodule. Current Opinion in Oncology, 26, 8–13.
© 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
doi: 10.1111/cen.12692
ORIGINAL ARTICLE
Value of sonographic features in predicting malignancy in thyroid nodules diagnosed as follicular neoplasm on cytology Chiaw-Ling Chng*,†, Tom R. Kurzawinski‡ and Tim Beale§ *Department of Endocrinology, Singapore General Hospital, Singapore, †Department of Endocrinology, Royal Free Hospital, ‡Centre for Endocrine Surgery University College Hospital and §Department of Radiology, University College Hospital, London, UK
Introduction Summary Background The cytological diagnosis of follicular neoplasm (Thy3F) remains a diagnostic challenge. The main aim of this study was to stratify the risk of malignancy in thyroid nodules diagnosed as Thy3F on cytology (Thy3F) using thyroid imaging reporting and data system (TIRADS). Methods A database of thyroid nodules with Thy3F cytological results from ultrasound-guided FNA (US-FNA) between January 2007 and March 2014 was studied retrospectively. Information on patient demographics, ultrasound characteristics and final histology of the nodules was collated. The number of suspicious US features of each thyroid nodule was counted based on TIRADS. The malignancy rate of each of the TIRADS category was also calculated based on the final histological outcomes of the nodules and compared to that calculated using a recently proposed thyroid malignancy risk prediction model. Results The overall malignancy rate of Thy3F cytology was 243%. There were significantly higher percentages of malignant nodules with irregular margins (200% vs 0%, P = 0000), hypoechogenicity (743% vs 514%, P = 0013) and taller-than-wide morphology (171% vs 09%, P = 0001) when compared to benign nodules. The risk of malignancy increased with advancing TIRADS score: TIRADS 4A (143%), TIRADS 4B (231%), TIRADS 4C (875%) and TIRADS 5 (100%). The malignancy rate calculated using the prediction model similarly increased with advancing TIRADS score: TIRADS 4A (62%), TIRADS 4B (325%), TIRADS 4C (799%) and TIRADS 5 (90%). Conclusion Thyroid nodules with TIRADS scores 4C and 5 should be considered for single definitive surgery in view of the high malignant rate. (Received 5 November 2014; returned for revision 21 November 2014; finally revised 1 December 2014; accepted 4 December 2014)
Correspondence: Chiaw-Ling Chng, Department of Endocrinology, Level 3 The Academia, 20 College Road, Singapore 169856, Singapore. Tel.: +65-63265916; Fax: +65-62273576; E-mail: chng.chiaw.ling@ sgh.com.sg © 2014 John Wiley & Sons Ltd
Ultrasound-guided fine needle aspiration (US-FNA) has an established role in the evaluation of thyroid nodules. However, an indeterminate cytological diagnosis from the FNA remains a significant management challenge to clinicians. Thy categories (1–5) have been introduced in the United Kingdom as numerical definitions at the end of the descriptive report to clarify the management of thyroid lesions following from the FNA cytology report.1 They are similar but not identical to the US Bethesda Thyroid Cytology Reporting Categories (TBSTCR).2 Each of the categories has a definition and a recommended management. For example, Thy3 category is subdivided into: Thy3A and Thy3F. In most instances, a repeat ultrasound with or without FNA is recommended for a Thy3a cytology result.3 Thy3F is used when a follicular neoplasm is suspected. The histological possibilities may include a hyperplastic nodule, follicular adenoma, follicular carcinoma or a follicular variant of papillary thyroid carcinoma. These cannot be distinguished on cytology alone and surgery (hemithyroidectomy) is required for diagnosis with a completion of thyroidectomy recommended for those with a confirmed thyroid cancer.3 This treatment approach is not ideal, as the risk of malignancy from a Thy3F cytological diagnosis is between 15 and 30%,1 and the majority of these patients may not require surgery. In addition, algorithms to identify those at highest risk of malignancy may allow the surgeon to perform a single definitive operation, saving time, reducing patient morbidity and reducing costs. It has been reported that certain sonographic features of a thyroid nodule are associated with an increased likelihood of malignancy, although no single predictor has been found to have a high positive predictive value for cancer. Hence, guidelines with various combinations of ultrasound features that are both sensitive and specific for predicting the presence of cancer have been defined in several studies1,3–11 Thyroid imaging reporting and data system (TIRADS) was first proposed by Harvath et al.6 This scoring system was based on breast imaging reporting and data system (BIRADS) defined by the American College of Radiology12 and stratifies the risk of malignancy of thyroid nodules based on ten ultrasound characteristics. However, this system was considered to be too complex for implementation in clinical practice and have as modified by several other authors.9,11,13,14 1
2 C.-L. Chng et al. Kwak et al. demonstrated that the following US features showed a significant association with malignancy: solid component, marked hypoe-chogenicity, microlobulated or irregular margins, microcalcifications and taller-than-wide shape.9 They scored TIRADS as 4A (one suspicious feature), 4B (two suspicious features), 4C (three or four suspicious features) and 5 (five suspicious features) and demonstrated the fitted probability and risk of malignancy increased as the number of suspicious ultrasound features increased.9 The advantages of this scoring system are its simplicity and ease in implementation in clinical practice. A recent study by the same group proposed a thyroid cancer risk prediction model using features of thyroid nodules found on US and was validated based on multicentre retrospective data.15 The main aim of this study was to stratify the risk of malignancy in thyroid nodules diagnosed as follicular neoplasms on cytology (Thy3F) using TIRADS. We also compared the risk of malignancy calculated based on the prediction model and the actual malignancy rate in these nodules in each TIRADS category.
Materials and methods A database of thyroid nodules with Thy3F cytological results from US-FNA performed by a single radiologist between January 2007 and March 2014 was studied retrospectively. The data were collected from three different institutions where the radiologist routinely performs head and neck US and FNAs. All the US were performed and reviewed by the same radiologist with extensive experience (close to 20 years) in head and neck radiology. Ultrasounds were performed using a Toshiba Aplio 500 (Toshiba America Medical Systems, Inc., Tustin, CA, USA) and Philips HDI 5000 (Philips Healthcare, DA Best, The Netherlands). US-FNA was performed only on thyroid nodules with suspicious US features. FNAs were performed free hand using a 25-gauge needle using nonaspiration technique, under direct US guidance. The direct smears were air-dried, sent to the respective pathology laboratory on the same day, stained with May-Grunwald–Giemsa methods and interpreted by cytopathologists with experience in thyroid cytology. Information on patient demographics, ultrasound characteristics and final histology of the nodules was collated by the main author of this study, such that the performing radiologist was blinded with respect to the final histology of the nodules. Static ultrasound images and original ultrasound reports of the nodules biopsied were retrospectively reviewed for the following features: size (based on the largest diameter on US), solid component, hypo-echogenicity, microlobulated or irregular margins, taller-than-wide shape (ratio of anteroposterior to transverse diameter ≥1), the presence of microcalcifications and increased intranodular vascularity obtained via routine colour Doppler assessment. All relevant parameters were compared between the benign and malignant histological outcomes of these nodules. The number of suspicious US features of each thyroid nodule was also counted based on the TIRADS scores proposed by Kwak et al.9 In this scoring system, a solid component, hypoechogenicity, microlobulated or irregular margins, microcalcifications and a taller-than-wide shape were counted as suspicious
US features. Thyroid nodules without suspicious features were classified as TIRADS category 3. Thyroid nodules having one, two, three or four or five suspicious features were classified as categories 4A, 4B, 4C or 5, respectively. On the basis of the recent thyroid cancer risk prediction model proposed by Kwak et al.,15 we applied a score of 2 to hypo-echogenicity, a score of 6 to marked hypo-echogenicity, a score of 1 to a taller-thanwide shape, a score of 5 to microlobulated or irregular margins and a score of 2 to microcalcifications. The total risk score of each thyroid nodule was calculated by summing the scores of each suspicious US features, and the percentage chance of malignancy was obtained according to the total risk score. The malignancy rate of each of the TIRADS category was also calculated based on the final histological outcomes of the nodules. The thyroid cancer risk is calculated based on the prediction model and the actual malignancy rate in these nodules diagnosed as follicular neoplasm on cytology was compared. The study protocol complies with the local ethical guidelines. The data were analysed using SPSS version 220 (SPSS Inc., Chicago, IL, USA). Baseline continuous data were expressed in mean + standard deviation, and categorical data were expressed as percentages. Univariate associations between the presence of cancer and discrete variables were evaluated using chi-squared test. Independent Student’s t-test was used to compare between categorical and continuous variables. P < 005 was considered statistically significant.
Results There were 497 FNAs with Thy 3, 4 and 5 cytological results performed on 465 patients between January 2007 and April 2014. There were 239 (481%) cytological results that were classified as Thy3F (suspicious of follicular neoplasm). Ultrasound and histological correlation were available for comparison in 144 cases and constitute the results of this study. Four incidental micropapillary thyroid cancers were excluded from the study. The overall malignancy rate of Thy3F cytology was 243%. The distribution of histology results in these 144 cases is presented in Table 1. The majority of nodules with benign histology were follicular adenoma (523%). Fifteen (429%) of the nodules with malignant histology were papillary thyroid cancer, of which four were follicular variants of papillary thyroid cancer. Sixteen (457%) of the malignant nodules were follicular thyroid cancer. Comparison of patient demographics and sonographic characteristics between thyroid nodules with benign and malignant histology is presented in Table 2. As only the solid component was biopsied in all thyroid nodules in our practice, this sonographic feature was not compared between benign and malignant nodules in our study. There was no significant differences in the age or gender distribution between patients with malignant or benign nodules (P = 0061 and P = 0153, respectively). In the comparison of sonographic characteristics, there were significantly higher percentages of malignant nodules with irregular margins (200% vs 0%, P = 0000), hypo-echogenicity (743% vs 514%, P = 0013) and taller-than-wide morphology (171% vs 09%, P = 0001) when compared to benign nodules. There was © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
Predictors of malignancy in thyroid follicular neoplasms 3 Table 1. Histological diagnosis in the 144 cases
120 100
N (%)
Benign Follicular adenoma Hurthle cell adenoma Multinodular goitre/Nodular hyperplasia Hyperplastic nodule Intrathyroid parathyroid adenoma Malignant Papillary thyroid cancer Follicular thyroid cancer Both papillary and follicular thyroid cancer Medullary thyroid cancer
109 (757%) 57 3 21 27 1 35 (243%) 15 16 3 1
Table 2. Comparison of patient demographics and sonographic characteristics between thyroid nodules with benign and malignant histology Benign (n = 109) Demographic characteristics Mean age (years) 441 149 Male 23 (676%) Female 86 (782%) Sonographic characteristics Nodule size (cm) 28 15 Irregular margins 0 Hypo-echogenicity 56 (514%) Taller-than-wide 1 (09%) morphology Presence of 5 (46%) microcalcifications Intranodular 55 (505%) vascularity
Malignant (n = 35)
P-value
496 157 11 (324%) 24 (218%)
0061 0153
32 15 7 (200%) 26 (743%) 6 (171%)
0222 0000* 0013* 0001*
3 (86%)
0302
17 (486%)
0500
*Denotes statistical significant results. Table 3. Malignancy rate based on TIRADS category in the 144 cases with diagnosis of follicular neoplasm on cytology TIRADS category
No of cases (n)
Benign (n, %)
Malignant (n, %)
Malignant rate (%)
4A 4B 4C 5 Total
56 78 8 2 144
48 60 1 0 109
8 18 7 2 35
143 231 875 100 243
(440%) (55%) (09%) (0%)
(229%) (514%) (200%) (57%)
no significant difference in the percentages of malignant nodules with microcalcifications or increased intranodular vascularity when compared to benign nodules. The malignancy rate based on TIRADS categories is shown in Table 3. There were no nodules scored as TIRADS 3 in our study. The risk of malignancy increased with advancing TIRADS score: TIRADS 4A (143%), TIRADS 4B (231%), TIRADS 4C (875%) and TIRADS 5 © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
100
Malignancy rate (%)
Histological diagnosis
90
87·5 79·9
80 Malignant rate based on histology
60 32·5
40 20 0
14·3 6·2 4A
Malignancy rate based on risk predicon model
23·1
4B
4C
5
TIRADS category Fig. 1 Graph comparing the malignancy rate of various TIRADS category calculated by the histology results of our study and malignancy rate calculated using the risk prediction model proposed by Kwak et al.15
(100%) (Table 3). The malignancy rate of each TIRADS category (calculated based on the recently proposed prediction model)15 similarly increased with advancing TIRADS score: TIRADS 4A (62%), TIRADS 4B (325%), TIRADS 4C (799%) and TIRADS 5 (90%). A comparison of malignancy risk based on the risk prediction model and histological outcomes in each TIRADS category is shown in Fig. 1.
Discussion The diagnosis of follicular neoplasm (Thy3F) remains a diagnostic dilemma. The distinction of follicular adenoma from follicular carcinoma requires the presence of capsular or vascular invasions but such characteristics cannot be assessed on cytological evaluation. In our study, the majority of the nodules with Thy3F cytology were benign, which were mostly follicular adenomas on histology. Although there were a significant proportion of papillary thyroid cancers among the malignant nodules, four of these were follicular variants of papillary thyroid cancers. It is well known that the cytological diagnosis of this entity is often challenging to make, and it can be misdiagnosed as follicular neoplasms on FNAC.16 The current British Thyroid Association guidelines recommends repeat FNA for Thy3A cytology results, but a diagnostic hemithyroidectomy remains the next step in the evaluation of Thy3F cytology outcomes.3 The overall malignancy rate of nodules with Thy3F cytology result was 243% in our study, which is similar to the malignancy rate for this cytological category based on the current guidelines.1 Hence, the fact that majority of these cases have benign disease on postoperative histology justifies the effort to better select candidates for surgery. Ultrasound is an important diagnostic tool in predicting thyroid malignancy and selecting thyroid nodules that should be evaluated by FNA. Known suspicious US features include hypo-echogenicity, microlobulated or irregular margins, microcalcifications and a taller-than-wide shape. Although no single US feature can reliably predict the risk of malignancy, a combination of these features is known to provide better
4 C.-L. Chng et al. diagnostic accuracy17,18 In our study, there was higher percentages of malignant nodules with irregular margins, hypo-echogenicity and taller-than-wide morphology, although size of nodule, the presence of microcalcifications and intranodular vascularity were not significantly different between the two groups. In particular, irregular margins have the highest positive predictive value for malignancy (100%) as all the nodules that exhibit this sonographic characteristic were malignant. This US feature was also the strongest predictor of malignancy in the study by Maia et al., on 80 patients with indeterminate thyroid cytology who were surgically treated at a single centre.19 The ‘taller-than-wide’ morphology was first proposed by Kim et al., as an US feature of malignancy, which was based on the fact that growth of most benign nodules has been found to remain within normal tissue planes, whereas malignant nodules grow across normal tissue planes.8 This sonographic feature was also found to have the highest diagnostic odds ratio among all the analysed ultrasound features in predicting malignancy by a meta-analysis by Brito et al.20 The same study also confirmed that nodule size is not an accurate predictor of thyroid cancer across different cut-offs, a finding that holds true across all the various cytology subgroups.20 Although the presence of microcalcifications and increased intranodular vascularity were found to be predictive of malignancy in some studies,20,21 these results were not reproduced in others.19,22 There was no significant difference in the age between subjects with benign or malignant nodules. Age was also not a predictor of malignancy in several studies specific to the cytological diagnosis of follicular neoplasms.23–25 There was also no male or female predominance in the malignant nodules, even though the incidence of thyroid nodules is higher in women.26 Thyroid imaging reporting and data system (TIRADS) was introduced to improve communication between radiologists and physicians through a standardized reporting format.6 As its introduction many variations of TIRADS have been proposed, some of them including additional US features such as elastography and vascularity criteria.7,9–11,13,14 The TIRADS used in our study is simple, convenient, accurate, and stratifies the malignancy risks. This TIRADS predicts malignancy risks according to just the number of suspicious US features. The malignancy rate in our study sample of thyroid nodules with Thy3F cytology results increases with increasing TIRADS score and reached 100% when all 5 suspicious features were present. A recent study, which assessed data from 136 indeterminate thyroid nodules, for which histopathological, cytological and ultrasound data were available, classified the nodules according to both the Bethesda system and a version of TIRADS that included vascularity criteria.14 On the basis of their findings, they suggested that for patients whose nodules have a TIRADS score of 4B or 5 and also have a Bethesda classification of either IV or V it may be advisable to proceed directly to definitive surgery on account of the high malignancy rates (700–857%) that are associated with these features.14 Our results concur with their findings, although the TIRADS system employed in our study differs from their study where nodule vascularity was not a criteria, and we used the Thy system for cytological classification instead
of the Bethesda system. However, results of both studies imply the higher risk of malignancy in nodules that are suspicious of follicular neoplasm which have more suspicious ultrasound features. Although TIRADS has been demonstrated to be accurate in predicting thyroid cancer risks in several studies, the system proposed in the studies lacked reproducibility, because each was developed based on the individual institution. This lack of uniformity with regards to the implementation of TIRADS has resulted in different reported values for its accuracy. The main drawback of the TIRADS proposed by Kwak et al., (used in our study) was that each US feature is considered with the same weight without considering different probabilities of malignancy for each US feature.9,15 The same authors subsequently designed a simple diagnostic prediction model using US features of thyroid nodules based on a multicentre data collected retrospectively which took this limitation into account.15 The results were also reproducible when used by less experienced radiologists.27 We applied this malignant risk prediction model to our study and showed increasing malignancy rates with increasing TIRADS score. The malignancy rates obtained for each TIRADS category were similar to the malignancy rate based on the histological outcome of these nodules (Fig. 1). Notably, the risk of malignancy increases exponentially in TIRADS 4C and 5. Our results suggest reproducibility of this malignant risk prediction using the model suggested by Kwak et al.,15 and its potential utility in the clinical management of thyroid nodules with Thy3F cytological outcomes. The main limitation of our study is the retrospective design and small numbers. Forty per cent of our patients with Thy3F cytology did not have histological results. The main reasons were refusal to surgery and lost to follow-up. In two patients, the decision was not to proceed to surgery due to the unrelated terminal diseases. In United Kingdom, a diagnostic hemithyroidectomy is recommended as the next step in the management of thyroid nodules with Thy3F cytological results based on the British Thyroid Association guidelines.3 Hence, a selection bias for surgery is unlikely as all the patients with Thy3F cytological results in the study were counselled for diagnostic hemithyroidectomy as a standard practice by their referring doctors, and this decision was not based on the ultrasound characteristics of the nodule. The strength of our study was that the ultrasound examinations were performed by the same radiologist, which reduced diagnostic variability. However, the TIRADS and risk prediction model used in our study is simple and has shown good reproducibility among less experienced radiologists, allowing more widespread use among clinicians who may not be specialist radiologists. The results of our study demonstrated that nodules with Thy3F cytology results and TIRADS score of 4C and 5 have very high risk of malignancy, and these patients should be considered for single definitive surgery. However, the absolute number of nodules exhibiting three or more suspicious US features in our study is small, with most nodules demonstrating one or two suspicious characteristics. Nonetheless, the malignancy risk of the nodules with one or two suspicious sonographic features is non-negligible (Fig. 1) and should be communicated to the patient when planning treatment. © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
Predictors of malignancy in thyroid follicular neoplasms 5 US-guided core-needle biopsy (US-CNB) provides a larger tissue sample that retains its cellular architecture and may enable a more precise histologic diagnosis.28 However, this procedure is not commonly used in the routine assessment of thyroid nodules, due to its perceived higher risk of complications and technical difficulties in smaller lesions.29 Recent evidence suggests the use of US-CNB to be an effective complementary diagnostic tool in the further evaluation of thyroid nodules which are indeterminate on initial FNA cytology.30,31 Thus, the application of the TIRADS score to thyroid nodules with Thy3F cytological results may help to identify selected cases where an US-CNB may be useful. More prospective studies, exploring the complementary role of US-CNB or combining new molecular diagnostic testing32 with this US-based malignant risk stratification model may help to delineate the malignancy risk of nodules with less than three suspicious US features, or further enhance the accuracy in predicting malignancy in nodules with TIRADS scores of 4C and 5 to reach 100%.
7
8
9
10
11
12
Conclusion The Thyroid imaging reporting and data system (TIRADS) is simple to use and accurate in malignant risk stratification of thyroid nodules diagnosed as follicular neoplasms (Thy3F) on cytology. Thyroid nodules with TIRADS scores 4C and 5 should be considered for single definitive surgery in view of the high malignant rate.
13
14
Disclosure statement The authors declare that no competing financial interests exist.
15
References 1 British Thyroid Association. (2009) Royal College of Pathologist: Guidelines for reporting of thyroid cytology specimens. Available at http://www.rcpath.org/ Resources/RCPath/Migrated%20 Resources/Documents/G/g089guidanceonthereportingofthyroidcytologyfinal.pdf. Accessed September 19, 2014 2 Cibas, E.S., Ali, S.Z. & Conference, N.C.I.T.F.S.o.t.S. (2009) The Bethesda System For Reporting Thyroid Cytopathology. American Journal of Clinical Pathology, 132, 658–665. 3 Perros, P., Boelaert, K., Colley, S. et al. (2014) Guidelines for the management of thyroid cancer. Clinical Endocrinology (Oxf), 81 (Suppl 1), 1–122. 4 Gharib, H., Papini, E., Paschke, R. et al. (2010) American Association of Clinical Endocrinologists, Associazione Medici Endocrinologi, and European Thyroid Association Medical guidelines for clinical practice for the diagnosis and management of thyroid nodules: executive summary of recommendations. Endocrine Practice, 16, 468–475. 5 American Thyroid Association Guidelines Taskforce on Thyroid, N., Differentiated Thyroid, C., Cooper, D.S., Doherty, G.M. et al. (2009) Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid, 19, 1167–1214. 6 Horvath, E., Majlis, S., Rossi, R. et al. (2009) An ultrasonogram reporting system for thyroid nodules stratifying cancer risk for © 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
16
17
18
19
20
21
22
clinical management. Journal of Clinical Endocrinology and Metabolism, 94, 1748–1751. Hambly, N.M., Gonen, M., Gerst, S.R. et al. (2011) Implementation of evidence-based guidelines for thyroid nodule biopsy: a model for establishment of practice standards. AJR. American Journal of Roentgenology, 196, 655–660. Kim, E.K., Park, C.S., Chung, W.Y. et al. (2002) New sonographic criteria for recommending fine-needle aspiration biopsy of nonpalpable solid nodules of the thyroid. AJR. American Journal of Roentgenology, 178, 687–691. Kwak, J.Y., Han, K.H., Yoon, J.H. et al. (2011) Thyroid imaging reporting and data system for US features of nodules: a step in establishing better stratification of cancer risk. Radiology, 260, 892–899. Park, J.Y., Lee, H.J., Jang, H.W. et al. (2009) A proposal for a thyroid imaging reporting and data system for ultrasound features of thyroid carcinoma. Thyroid, 19, 1257–1264. Russ, G., Bigorgne, C., Royer, B. et al. (2011) The Thyroid Imaging Reporting and Data System (TIRADS) for ultrasound of the thyroid. Journal de Radiologie, 92, 701–713. American College of Radiology, B.-R.C. (2003) ACR BI-RADSâultrasound. In: V.A. Reston ed. ACR BI-RADS breast imaging and reporting data system: breast imaging atlas, 4th edn. American College of Radiology, Reston, Virginia, 1–86. Russ, G., Royer, B., Bigorgne, C. et al. (2013) Prospective evaluation of thyroid imaging reporting and data system on 4550 nodules with and without elastography. European Journal of Endocrinology, 168, 649–655. Maia, F.F., Matos, P.S., Pavin, E.J. et al. (2014) Thyroid imaging reporting and data system score combined with Bethesda system for malignancy risk stratification in thyroid nodules with indeterminate results on cytology. Clinical Endocrinology (Oxford), doi: 10.1111/cen.12525. [Epub ahead of print]. Kwak, J.Y., Jung, I., Baek, J.H. et al. (2013) Image reporting and characterization system for ultrasound features of thyroid nodules: multicentric Korean retrospective study. Korean Journal of Radiology, 14, 110–117. Manimaran, D., Karthikeyan, T.M., Khan, D.M. et al. (2014) Follicular variant of papillary thyroid carcinoma: cytological indicators of diagnostic value. Journal of Clinical and Diagnostic Research, 8, 46–48. Moon, W.J., Jung, S.L., Lee, J.H. et al. (2008) Benign and malignant thyroid nodules: US differentiation–multicenter retrospective study. Radiology, 247, 762–770. Tutuncu, Y., Berker, D., Isik, S. et al. (2014) The frequency of malignancy and the relationship between malignancy and ultrasonographic features of thyroid nodules with indeterminate cytology. Endocrine, 45, 37–45. Maia, F.F., Matos, P.S., Pavin, E.J. et al. (2011) Value of ultrasound and cytological classification system to predict the malignancy of thyroid nodules with indeterminate cytology. Endocrine Pathology, 22, 66–73. Brito, J.P., Gionfriddo, M.R., Al Nofal, A. et al. (2014) The accuracy of thyroid nodule ultrasound to predict thyroid cancer: systematic review and meta-analysis. Journal of Clinical Endocrinology and Metabolism, 99, 1253–1263. Gul, K., Ersoy, R., Dirikoc, A. et al. (2009) Ultrasonographic evaluation of thyroid nodules: comparison of ultrasonographic, cytological, and histopathological findings. Endocrine, 36, 464–472. Dutta, S., Thaha, M.A. & Smith, D.M. (2011) Do sonographic and cytological features predict malignancy in cytologically
6 C.-L. Chng et al.
23
24
25
26 27
indeterminate thyroid nodules? Annals of the Royal College of Surgeons of England, 93, 361–364. Sorrenti, S., Trimboli, P., Catania, A. et al. (2009) Comparison of malignancy rate in thyroid nodules with cytology of indeterminate follicular or indeterminate Hurthle cell neoplasm. Thyroid, 19, 355–360. Castro, M.R., Espiritu, R.P., Bahn, R.S. et al. (2011) Predictors of malignancy in patients with cytologically suspicious thyroid nodules. Thyroid, 21, 1191–1198. Calo, P.G., Medas, F., Santa Cruz, R. et al. (2014) Follicular nodules (Thy3) of the thyroid: is total thyroidectomy the best option? BMC Surgery, 14, 12. Eng, C.Y., Quraishi, M.S. & Bradley, P.J. (2010) Management of Thyroid nodules in adult patients. Head & Neck Oncology, 2, 11. Ko, S.Y., Lee, H.S., Kim, E.K. et al. (2014) Application of the Thyroid Imaging Reporting and Data System in thyroid
28
29 30
31
32
ultrasonography interpretation by less experienced physicians. Ultrasonography, 33, 49–57. Screaton, N.J., Berman, L.H. & Grant, J.W. (2003) US-guided core-needle biopsy of the thyroid gland. Radiology, 226, 827– 832. Lo Gerfo, P. (1994) The value of coarse needle biopsy in evaluating thyroid nodules. Thyroidology, 6, 1–4. Nasrollah, N., Trimboli, P., Guidobaldi, L. et al. (2013) Thin core biopsy should help to discriminate thyroid nodules cytologically classified as indeterminate. A new sampling technique. Endocrine, 43, 659–665. Choi, W.J. & Baek, J.H. (2014) Role of core needle biopsy for patients with indeterminate, fine-needle aspiration cytology. Endocrine, 45, 1–2. Yip, L. (2014) Molecular diagnostic testing and the indeterminate thyroid nodule. Current Opinion in Oncology, 26, 8–13.
© 2014 John Wiley & Sons Ltd Clinical Endocrinology (2014), 0, 1–6
