Endocr Pathol DOI 10.1007/s12022-014-9301-3
Follicular Variant of Papillary Thyroid Carcinoma: Accuracy of FNA Diagnosis and Implications for Patient Management Berrin Ustun & David Chhieng & Manju L. Prasad & Elizabeth Holt & Lynwood Hammers & Tobias Carling & Robert Udelsman & Adebowale J. Adeniran
# Springer Science+Business Media New York 2014
Abstract Follicular variant of papillary thyroid carcinoma (FVPTC) creates a continuous diagnostic dilemma among pathologists because of the paucity of nuclear changes of papillary carcinoma and overlapping features with benign and other neoplastic follicular lesions. Current guidelines for the management of thyroid nodules recommend surgery for confirmed PTC, suspicious for PTC, and follicular neoplasm cases, while further immediate diagnostic studies or treatment are not routinely required if the nodule is benign on cytology. This study is designed to determine the accuracy of cytology in the diagnosis of FVPTC, based on the Bethesda classification system, and determine the implications for patient management based on the current recommendation. Based on a retrospective review of cytologic diagnoses between January 2008 and December 2011, thyroid fine needle aspiration (FNA) cytology specimens with subsequent surgical intervention and a final diagnosis of FVPTC were selected. The cytologic diagnoses were compared with the final diagnoses, and the percentage of cases contributing to the final diagnosis of FVPTC was calculated for each diagnostic category. Triage efficiency and diagnostic accuracy were calculated. One hundred and fifty-two cases with histologic confirmation of B. Ustun : D. Chhieng : M. L. Prasad : A. J. Adeniran (*) Department of Pathology, Yale University School of Medicine, 310 Cedar Street, CB 510A, New Haven, CT 06510, USA e-mail: [email protected] E. Holt Department of Internal Medicine, Yale University School of Medicine, New Haven, CT, USA L. Hammers Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT, USA T. Carling : R. Udelsman Department of Surgery, Yale University School of Medicine, New Haven, CT, USA
FVPTC were identified (representing 128 patients—101 female, 27 male). All patients had undergone either lobectomy with completion thyroidectomy or total thyroidectomy. The cytologic diagnosis of “positive for malignancy” accounted for only 27 % of the final histologic diagnosis of FVPTC, while suspicious for carcinoma, follicular neoplasm, follicular lesion of undetermined significance, and benign accounted for 11, 23, 23, and 16 % of the final diagnosis of FVPTC, respectively. Only 18 % of the 55 cases tested were positive for BRAF mutation. The subtle nuclear features of FVPTC pose challenges for an accurate diagnosis. Therefore, a better approach is to triage these cases for surgical intervention and/ or further evaluation of the particular nodule. Our triage efficacy for FVPTC was 84 %; however, the diagnostic accuracy of PTC was 38 %. A negative diagnosis on FNA has diagnostic and management implications for up to 16 % of cases because they may have no further immediate diagnostic studies or treatment. BRAF mutation analysis provides minimal effect on diagnostic accuracy. Keywords Follicular variant of papillary thyroid carcinoma . Fine needle aspiration . Diagnostic accuracy . Triage efficiency
Introduction Palpable thyroid nodules are detected in about 4 to 7 % of the adult population. Only a small portion of these detected nodules are malignant [1]. The very high prevalence of the thyroid nodules in the general population and low rate of malignancy underscore the need for a reliable preoperative evaluation of any given nodule. To this end, the procedure of choice for triaging patients with palpable thyroid nodules is fine needle aspiration (FNA) biopsy [2]. Papillary thyroid carcinoma (PTC) is the most common type of malignancy of the thyroid gland, accounting for 70 to
Endocr Pathol
80 % of all thyroid cancers [3]. Numerous histological variants of PTC have been described in the literature, and some are known to have prognostic significance [4]. The follicular variant of PTC (FVPTC) is the most common variant after classic PTC, making up 10 to 15 % of all PTC [5, 6]. FNA biopsy is very sensitive for the diagnosis of the conventional PTC. On the contrary, reported sensitivity of FNA in FVPTC is significantly low [7–11]. Given this low sensitivity, FVPTC poses a significant diagnostic challenge to pathologists, endocrinologists, and surgeons [12]. The diagnosis of FVPTC is made by finding the typical nuclear features, which are seen in classic PTC. In theory, the rate and ease of diagnosis of FVPTC should be similar to classic PTC since the diagnostic criteria are identical. However cytological diagnosis of FVPTC is challenging and poses a diagnostic problem for the following reasons: (1) nuclear features of PTC are less obvious and focal, (2) absence of papillary groups or large syncytial sheets, and (3) predominance of follicular architecture with rare nuclear alteration and presence of variable colloid component [13–15]. Another important reason for the difficulty on cytomorphology is the overlap in histologic appearance between FVPTC and other lesions in the follicular neoplasm category, namely, follicular adenoma, follicular carcinoma, and benign nonneoplastic follicular lesions [16–18]. For these reasons, it is not uncommon for the diagnosis of FVPTC to be missed on FNA. In the literature, the studies looking into clinical behaviors and long-term survival similarities and differences between FVPTC and classic PTC reported different results. A recent study with long-term follow-up showed that FVPTC generally has favorable clinicopathologic features, which translates into less nodal metastasis and extrathyroidal extension. However, the mean survival for both groups of patients was not significantly different [19–22]. This current study was designed to determine the triage efficiency and diagnostic accuracy of cytology in the diagnosis of FVPTC, based on the modified Bethesda classification system, and to determine the implications for patient management based on the current recommendation for intervention. Given that the diagnosis of FVPTC is challenging, we propose that the goal of FNA diagnosis should be to improve the triage efficiency of patients with FVPTC for surgery or repeat FNA. This will ensure that patients with equivocal diagnosis have appropriate follow-up.
Materials and Methods A retrospective review of the files from Yale-New Haven Hospital’s Department of Pathology for the period from January 2008 to December 2010 identified a total of 152 FNA biopsy cases in which the patients underwent thyroid resection and the histological diagnosis showed FVPTC. Of the 152,
eight were excluded from this study due to nondiagnostic samples. Of the 144 thyroid FNA specimens, 74 specimens were processed in-house at Yale-New Haven Hospital. The inhouse thyroid FNAs were performed by a radiologist or endocrinologist using a 25-gauge needle. Two to four passes were obtained in most of the cases. For each thyroid nodule, direct smears (air-dried Diff-Quik-stained slides [Dade Behring, Deerfield, IL, USA] and alcohol-fixed Papanicolaoustained slides) were obtained. The needle was then rinsed in Cytorich Red solution (Thermo Fisher Scientific) for ThinPrep liquid-based preparation (Hologic, Marlborough, MA, USA). Seventy FNA specimens were prepared at outside institutions, and they were submitted with one or more of Papanicolaou-stained slides, Diff-Quik stained-slides, and ThinPrep liquid-based preparations. Depending upon submitting institutions, occasional cell-block slides were received. Cytologic interpretation and classification was based on the modified Bethesda System for Reporting Thyroid Cytopathology (BSRTC). At our institution, transition to the BSRTC took place in January 2008. The BSRTC categories were composed of the following: nondiagnostic/unsatisfactory, benign, indeterminate (AUS/FLUS), follicular neoplasm, suspicious for malignancy, and positive for malignancy. As a result, after the exclusion of the nondiagnostic cases, a total of 144 thyroid FNA cases from 128 patients were included in this retrospective study. The final cytologic and histologic diagnoses were identified from the pathology report. The cytologic diagnoses were rendered by seven board-certified cytopathologists with a range of years of experience from 3 to 10+ years. The triage efficiency in this study considered the FNA diagnoses that would potentially trigger surgical intervention or repeat FNA biopsy, and this included the following: positive for PTC, suspicious for PTC, follicular neoplasm, and indeterminate (AUS/FLUS). Diagnostic accuracy was considered as making the accurate diagnosis of PTC and that included positive for PTC and suspicious for PTC.
Results A total of 144 thyroid FNA samples from 128 patients were identified and reviewed. Sixteen patients had a repeat FNA biopsy. The clinical, pathologic, and molecular characteristics are highlighted in Table 1. The patient population was predominantly female (101 female and 27 male) and ranged in age from 14 to 80 years (median age, 51 years). Forty-three patients (representing 34 %) were aged 45 years and below, while the majority of the patients (85; 66 %) were older than 45 years. This is clinically important, given the significance that age has on the staging of these tumors. All patients had confirmed FVPTC on histology. Of the 128 patients, 28 had
Endocr Pathol Table 1 Clinical, pathologic, and molecular characteristics of FVPTC
Table 2 Distribution of cytologic diagnosis
Characteristics
Patients, number (%)
Cytologic diagnosis
Subdiagnosis
Number of cases
51 43 (34) 85 (66)
Positive Suspicious Follicular neoplasm
PTC PTC
39 (27 %) 16 (11 %) 33 (23 %) 10
Age Median (years) ≤45 >45 Sex Female Male Tumor size (cm) Median ≤4 >4 Vascular invasion Present Absent Extrathyroidal extension Present Absent Central nodal metastases Present Absent CBA Thyroid surgery Lobectomy Total thyroidectomy ± CLND BRAF V600E mutation Present Absent CBA
Hurthle cell neoplasm Microfollicular-patterned neoplasm FVPTC
101 (79) 27 (21) Indeterminate 1.3 122 (95) 6 (5) 6 (5) 122 (95) 5 (4) 123 (96) 22 (17) 67 (52) 39 (31) 28 (22) 100 (78) 10 (8) 45 (35) 73 (57)
CBA cannot be assessed
undergone lobectomy with completion thyroidectomy, 48 had undergone total thyroidectomy, 38 had undergone total thyroidectomy with lymph node dissection, and 14 had undergone total thyroidectomy with intraoperative frozen section. On histologic examination, vascular invasion was identified in 6 cases (5 %), while extrathyroidal extension and central lymph node metastases were present in 5 (4 %) and 22 (17 %) cases, respectively. Overall, final pathologic staging on surgical resections were as follows: pT1 (104/144, 72 %), pT2 (24/144, 17 %), and pT3 (16/144, 11 %). All the patients with lymph node metastasis had metastasis to central neck lymph nodes (N1 disease). The overall mean size of the resected thyroid nodules was 1.3 cm (range 0.1 to 11.3 cm). By using the modified 2007 National Cancer Institute (NCI)/ Bethesda guidelines, FNA diagnosis resulted in the following distribution of cytological diagnosis: positive for PTC (39/ 144, 27 %), suspicious for PTC (16/144, 11 %), indeterminate (33/144, 23 %), follicular neoplasm (33/144, 23 %), and negative (23/144, 16 %) (Table 2).
Microfollicular patterned (INa) Nuclear atypia (INb) Negative Goiter Lymphocytic thyroiditis All cases
11 12 33 (23 %) 2 31 23 (16 %) 19 4 144 (100 %)
Of the 23 negative cases, the cytologic diagnosis was predominantly rendered as hyperplastic nodule (goiter) (19/ 23, 83 %) while lymphocytic thyroiditis accounted for the remaining 4 cases. Of these 23 nodules with a negative cytologic diagnosis, 16 (70 %) measured less than 1 cm in the greatest dimension on surgical resection. The mean size of the resected nodules in the negative diagnostic category was 0.7 cm (range, 0.2 to 2.5 cm). Sixteen of these 23 cases were in-house cases, and a review of the cytology slides revealed that 12 of the cases (80 %) displayed no atypical nuclear features and 4 (20 %) displayed subtle nuclear clearing and rare nuclear enlargement and nuclear groves in small subsets of follicular cells which were unrecognized by the pathologists at the time the negative cytological interpretation was rendered (Figs. 1 and 2). These cases were stratified into encapsulated and nonencapsulated FVPTC based on the presence or absence of a capsule on histological evaluation of the
Fig. 1 Follicular variant of papillary thyroid carcinoma. Cellular aspirate showing crowded groups. Most of the nuclei are round and uniform, but there is a small subpopulation with nuclear grooves. (Papanicolaou stain, original magnification ×200)
Endocr Pathol Table 3 Clinical, pathologic, and molecular characteristics of FVPTC cases diagnosed as negative on fine needle aspiration based on stratification into EFVPTC vs. NEFVPTC Characteristics
Fig. 2 Follicular variant of papillary thyroid carcinoma. Crowded groups showing nuclear enlargement and rare nuclear grooves. Scant sticky colloid is present. (Diff-Quik stain, original magnification ×400)
tumors (Table 3). The two histologic subtypes are identical in terms of median age and number of patients aged 45 years and below. There were no significant differences in terms of vascular invasion, extrathyroidal extension, and central lymph node metastases between the two groups. The sonographic characteristics of these 23 cases were also evaluated to determine if there were unique features that could help in better triage of these negative cases (Table 4). The nodules were predominantly solid (91 %) and hypoechoic (65 %). Other sonographic findings included absence of microcalcifications (65 %) and presence of well-defined, regular margins (61 %). The shape of the nodule was insignificant as almost half were oval while the other half had nodules which were taller than wide. At our institution, the term “indeterminate” corresponds to the NCI 2007 guidelines category “Follicular cells of undetermined significance/Atypia of undetermined significance (FLUS/AUS).” This category is further subclassified into two diagnostic subtypes: (1) FNA specimens showing scant cellularity with a predominantly microfollicular architecture and scant colloid, which raises the concern for follicular neoplasm (INa) and (2) FNA specimens showing subtle atypical nuclear features, which raises the concern for papillary thyroid carcinoma (INb). The INa category accounted for 2 of the 33 indeterminate cases (6 %), while the INb category accounted for 31 (94 %). Follicular neoplasm category comprises three diagnostic subtypes as follows: (1) cellular FNA sample showing predominantly microfollicular architecture (MF), (2) some features suggestive of but not diagnostic for papillary carcinoma follicular variant (FL), and (3) cellular FNA specimen showing predominantly Hurthle cells with scant colloid (HCN). The distribution of cytologic diagnosis of cases in the follicular neoplasm category is as follows: MF (11/33, 33 %), FL (12/33, 37 %), and HCN (10/33, 30 %). Sixteen of the 144 cases (11 %) had a cytologic diagnosis of “suspicious for PTC,” while 39 (27 %) had a cytologic
Age Median (years) ≤45 >45 Sex Female Male Tumor size (cm) Median ≤4 >4 Vascular invasion Present Absent Extrathyroidal extension Present Absent Central nodal metastases Present Absent CBA Thyroid surgery Lobectomy Total thyroidectomy ± CLND BRAF V600E mutation Present Absent CBA
Encapsulated (n=8)
Nonencapsulated (n=15)
45 5 3
49 4 11
7 0
11 3
0.8 8 0
0.4 15 0
0 8
0 15
0 8
1 14
1 5 2
3 7 5
1 7
3 12
1 3 4
1 2 12
diagnosis of “positive for PTC.” The calculated triage efficiency (potential for triggering surgical intervention or repeat FNA) is 84 % (121/144). The diagnostic accuracy is 38 % (54/ 144) (Table 5). The in-house cytologic specimens with the diagnoses of indeterminate, suspicious for PTC, and positive for PTC were triaged for diagnostic and/or prognostic BRAF mutation analysis. Ten (18 %) of the 55 cases tested harbored the BRAF V600E mutation. (Table 1)
Discussion Classic PTC and its variants represent the most common malignancy, with definitive preoperative diagnosis of classic PTC being around 94 % [2, 14]. The definitive preoperative cytological diagnosis of FVPTC is generally low. Most of the studies have reported sensitivities ranging from 9 to 37 % [5,
Endocr Pathol Table 4 Sonographic findings in FVPTC
Characteristics
Number (%)
Consistency Solid Cystic Mixed Echogenicity
21 (91) 0 (0) 2 (9)
Hypoechoic Isoechoic Microcalcification Present Absent Margin Regular Irregular Shape Oval Taller than wide
15 (65) 8 (35) 8 (35) 15 (65) 14 (61) 9 (39) 12 (52) 11 (48)
7–11, 17]. Tielens et al. [19] reported a sensitivity of 75 %; however, the results of the study are misleading as patients with suspicious FNA were included as true-positive results, leading to an apparent increase in calculated sensitivity. The diagnostic dilemma in the preoperative detection of FVPTC is not only in enhancing the specificity of preoperative diagnosis but also to simultaneously increase the sensitivity of detection so that patients with malignancy are discovered soon enough to allow effective therapy. In our present study, the definitive preoperative diagnosis was 27 % (39/144). But our instructional diagnostic accuracy was 38 % (54/144), which included both diagnostic categories of positive for PTC as well as suspicious for PTC. One might argue about the inclusion of the suspicious category in the diagnostic accuracy. We defined diagnostic accuracy as the ability to identify nodules requiring immediate attention. This includes the categories of positive and suspicious for malignancy. More so, the category of suspicious is added because in most cases when a diagnosis of suspicious for malignancy is made, there is a reasonably high inclination towards Table 5 Triage efficiency Cytologic diagnosis
Number of cases (%)
Current intervention
Positive
39 (27)
Suspicious Follicular neoplasm Indeterminate Negative All cases (ex-nondiagnostic)
16 (11) 33 (23) 33 (23) 23 (16) 144 (100)
Total thyroidectomy Lobectomy Lobectomy Repeat FNA No intervention
malignancy but there is a holdback usually due to scant evidence. Many studies have shown that precise recognition of FVPTC is low due to the rarity of the nuclear features, which are necessary to making a definitive diagnosis. Separate studies by Jain et al. and Logani et al. concluded that majority of the cytology specimens from patients with FVPTC were interpreted as suspicious for malignant lesion or follicular neoplasm [23, 24]. In this present study, 23 % (33/144) resulted in an indeterminate result. The indeterminate category in our institution comprises the lesions in the FLUS/AUS category of the Bethesda classification. Majority of the cases were called indeterminate due to nuclear atypia (31/33, 94 %), and rare cases (2/33, 6 %) were called indeterminate due to architectural atypia with the low cellularity and predominantly microfollicular pattern. Only one of those cases with an indeterminate call had a repeat FNA biopsy, which resulted in subsequent indeterminate cytologic diagnosis. In this particular study, the majority of the cases in the indeterminate category did not have a repeat FNA as recommended by current Bethesda guideline but rather had surgical intervention because of other concomitant risk factors. If they did have a repeat FNA, this would upgrade or possibly downgrade the original cytological diagnosis of indeterminate and consequently would impact the triage efficiency and diagnostic accuracy. Twenty-three percent (33/144) of our cases were in the category of follicular neoplasm, which should trigger immediate surgical intervention (lobectomy) according to the thyroid cancer guidelines [25]. For follicular neoplasms, most investigators believe that intraoperative evaluation does not provide accurate definitive diagnosis and a unilateral thyroid lobectomy is the standard initial surgical intervention [5]. The NCI Thyroid FNA State of the Science Conference held in 2007 proposed six basic diagnostic classifications for the reporting of thyroid FNA specimens [26]. At our institution [27], we use the modified Bethesda classification system for classifying thyroid FNA specimens. We further subclassify the follicular neoplasm category into three subcategories: (1) microfollicular-patterned neoplasm, (2) follicular lesion with some features suggestive of but not diagnostic of FVPTC, and (3) Hurthle cell neoplasm. Of the cytological diagnosis of follicular neoplasm, 37 % (12/33) exhibited enough atypical nuclear features which are beyond the indeterminate (AUS/ FLUS) diagnostic category but not enough to place those cases in the suspicious for PTC category. In our experience, there are a number of cases in which some nuclear features of PTC (most commonly nuclear clearing and rare grooves) are present. The frequency of these nuclear features varies from case to case but generally not enough nuclear changes to place those cases in the suspicious for PTC category. At our institution, our experience has been that the subcategory of follicular lesion with some features suggestive of but
Endocr Pathol
not diagnostic for FVPTC (FL) allows the pathologist to characterize follicular lesion one step further and gives more information to the clinician as to the level of concern that the lesions in this subcategory elicit. It is to ensure that patients with lesions in this subcategory are ultimately referred for surgical consultation. Two of the 12 cases diagnosed as FL on cytology had frozen sections performed at the time of surgery and were upgraded to positive for PTC and had total thyroidectomies based on the intraoperative consultation. The rest of the follicular neoplasm diagnoses were simply called either Hurthle cell neoplasm (10/33, 30 %) or microfollicularpatterned neoplasm (11/33, 33 %), which in turn puts these cases in the gray zone which will include adenomatous goiter, follicular adenoma, follicular carcinoma, Hurthle cell neoplasm, and, less likely, FVPTC [28, 29]. The reported rate of malignancy with the cytological diagnosis of follicular neoplasm varies but generally ranges from 8 to 30 % [30, 31]. As a result, even though the Bethesda guidelines recommend lobectomy for a cytological diagnosis of follicular neoplasm, there are still a good number of institutions where the Bethesda classification is not implemented [32]. Terms such as follicular lesion are used; hence, it may not be a routine practice in these institutions to perform immediate lobectomy when the term “follicular lesion” is used in the place of “follicular neoplasm.” In our experience, any cellular follicular neoplasm with subtle nuclear atypia should be assumed to be FVPTC until proven otherwise and careful search for the subtle nuclear features of PTC should be performed so that FVPTC can be distinguished from other lesions in the follicular neoplasm category (Fig. 3). This is the goal of subclassification of follicular neoplasm, and in this way, the pathologist will deliver more useful information to the clinician. We evaluated common sonographic findings of FVPTC in all cases with a negative cytologic diagnosis in our present study. The small sample size is admittedly a limitation, but it
Fig. 3 Follicular variant of papillary thyroid carcinoma. Morphologic appearance of follicular architecture and subtle nuclear features such as clearing, enlargement, and grooves (H & E stain, original magnification ×400)
does indicate that solid lesions, hypoechogenicity, absence of microcalcifications, and presence of well-defined regular margins are regular features seen in FVPTC. Unfortunately, these individual features are also seen in benign thyroid lesions to variable degrees; hence, the presence of a single individual sonographic characteristic is insufficient to suspect FVPTC. All the sonographic characteristics must be evaluated in conjunction with the clinical impression, and the intervention is based on the overall assessment. One major limitation in this study was our inability to describe the pattern of vascularity in these lesions. Even though this is not a peculiar feature of malignancy, marked intrinsic hypervascularity with increased flow in the central portion of the nodule relative to surrounding thyroid parenchyma is the most commonly described vascularity pattern in thyroid malignancies [33]. We stratified the 16 cases of FVPTC with benign cytologic diagnosis into two—encapsulated and nonencapsulated—to determine if there was a difference in the biologic behavior of the tumors in both categories. Even though it does not appear that there is a difference in biologic potential, the small sample size is a major limitation as one cannot reasonably draw a meaningful conclusion based on this. However, studies have shown that the nonencapsulated variant behaves more like classic PTC, with a significantly higher rate of BRAF V600E mutations and nodal metastases [34]. BRAF is mutated in a small percentage of FVPTC. Most of these mutated cases are usually recognized by the pathologist either as suspicious or positive for PTC on FNA. As a result of this, BRAF analysis has a limited value in preoperative diagnosis of FVPTC [35]. In this study, 18 % of the 55 cases harbored the BRAF mutation. In turn, BRAF molecular test neither improves triage efficacy nor diagnostic accuracy. There are several studies that have utilized ancillary studies to improve the morphologic evaluation of indeterminate lesions, enabling a better selection of patients for surgery. The expression of HBME-1, cytokeratin 19, and galectin-3 has been used to classify cytologically indeterminate lesions into high-risk (those that need immediate surgery) and low-risk (those that can be followed up) groups [36, 37]. The addition of RET proto-oncogene to HBME-1 and galectin-3 has been found to increase the overall diagnostic accuracy of cytology in follicular proliferations to 97 % [38]. These immunohistochemical stains can be performed on liquid-based preparations. It should be noted, however, that many of these markers are also focally expressed in benign nodular goiters, so they are used in conjunction with one another and with careful cytologic and morphologic evaluation. In summary, the diagnosis of FVPTC on FNA biopsy can be very challenging for the pathologist. The definitive diagnosis of FVPTC is not always possible. However, FNA biopsy plays an important role in identifying these patients and triaging them appropriately. While evaluating any follicular lesion, consideration should be given to FVPTC and strict
Endocr Pathol
cytologic criteria should be applied to identify these patients. Pathologists should develop a very low threshold for detection of FVPTC, and careful search for the presence of subtle nuclear atypia is critical to achieving this. These features can be seen focally and admixed with colloid and benign follicular epithelium. Our study suggests that emphasizing triage efficiency (i.e., FNA classification that triggers surgical intervention or repeat FNA), rather than diagnostic accuracy, might improve the clinical management of patients with FVPTC. The ultimate goal should be to minimize the negative diagnosis so that patients would not be lost to follow-up. The triage efficiency in our study was 84 %, an indication that up to one in six cases of FVPTC could still be missed and these cases could potentially have no intervention. In the 16 % of FVPTC where a benign diagnosis was rendered on FNA, most were microcarcinomas and no specific clinical or radiologic characteristics distinguished these cases from the other FVPTC cases that were appropriately triaged by FNA.
Funding This study was supported by funding from the Department of Pathology, Yale University School of Medicine, New Haven, CT. Financial Disclosures The authors have no financial disclosures.
References 1. Ramacciotti CE, Pretorius HT, Chu EW, et al. Diagnostic accuracy and use of aspiration biopsy in the management of thyroid nodules. Arch Intern Med 144:1169–1173, 1984. 2. Wang HH Reporting thyroid fine needle aspiration: Literature review and a proposal. Diagn Cytopathol 34:67–76, 2006. 3. Adeniran AJ, Zhu Z, Gandhi M, et al. Correlation between genetic alterations and microscopic features, clinical manifestations, and prognostic characteristics of thyroid papillary carcinomas. Am J Surg Pathol 30:216–222, 2006. 4. Lam AK, Lo CY, Lam KS. Papillary carcinoma of thyroid: A 30-yr clinicopathological review of the histological variants. Endocrine Pathol 16:323–30,2005. 5. Kesmodel SB, Terhune KP, Canter RJ, et al. The diagnostic dilemma of follicular variant of papillary thyroid carcinoma. Surgery 134: 1005–1012, 2003. 6. Shih SR, Shun CT, Su DH, et al. Follicular variant of papillary thyroid carcinoma: Diagnostic limitations of fine needle aspiration cytology. Acta Cytol 49:383–386, 2005. 7. Lin HS, Komisar A, Opher E, Blaugrund SM. Follicular variant of papillary carcinoma: the diagnostic limitations of preoperative fineneedle aspiration and intraoperative frozen section evaluation. Laryngoscope 110:1431–1436, 2000. 8. Caraway NP, Sneige N, Samaan NA. Diagnostic pitfalls in thyroid fine-needle aspiration: a review of 394 cases. Diagn Cytopathol 9: 345–50, 1993. 9. Powari M, Dey P, Saikia UN. Fine needle aspiration cytology of follicular variant of papillary carcinoma of thyroid. Cytopathology 14:212–215, 2003.
10. Yang J, Schnadig V, Logrono R, Wasserman PG. Fine-needle aspiration of thyroid nodules: A study of 4703 patients with histologic and clinical correlations. Cancer 111:306–315, 2007. 11. Miller JM, Hamburger JI, Kini SR. The needle biopsy diagnosis of papillary thyroid carcinoma. Cancer 48:989 –993, 1981. 12. Salajegheh A, Petcu EB, Smith RA, Lam AK. Follicular variant of papillary thyroid carcinoma: a diagnostic challenge for clinicians and pathologists. Postgrad Med J 84:78–82, 2008. 13. El Hag IA, Kollur SM. Benign follicular thyroid lesions versus follicular variant of papillary carcinoma: Differentiation by architectural pattern. Cytopathology 15:200–205, 2004. 14. Zacks JF, de las Morenas A, Beazley RM, O’Brien MJ. Fine-needle aspiration cytology diagnosis of colloid nodule versus follicular variant of papillary carcinoma of the thyroid. Diagn Cytopathol 18: 87–90, 1998. 15. Wu HH, Jones JN, Grzybicki DM, Elsheikh TM. Sensitive cytologic criteria for the identification of follicular variant of papillary thyroid carcinoma in fine-needle aspiration biopsy. Diagn Cytopathol 29: 262–266, 2003. 16. Williams MD, Suliburk JW, Staerkel GA, et al. Clinical significance of distinguishing between follicular lesion and follicular neoplasm in thyroid fine-needle aspiration biopsy. Ann Surg Oncol 16:3146– 3153, 2009. 17. Jogai S, Adesina AO, Temmim L, et al. Follicular variant of papillary thyroid carcinoma—A cytological study. Cytopathology 15:212– 216, 2004. 18. Dobrinja C, Trevisan G, Liguori G, et al. Sensitivity evaluation of fine-needle aspiration cytology in thyroid lesions. Diagn Cytopathol 37:230–235, 2009. 19. Tielens ET, Sherman SI, Hruban RH, Ladenson PW. Follicular variant of papillary thyroid carcinoma. A clinicopathologic study. Cancer 73:424–31, 1994. 20. Lin HW, Bhattacharyya N. Clinical behavior of follicular variant of papillary thyroid carcinoma: presentation and survival. Laryngoscope 120:712–716, 2010. 21. Chang HY, Lin JD, Chou SC, et al. Clinical presentations and outcomes of surgical treatment of follicular variant of the papillary thyroid carcinomas. Jpn J Clin Oncol 36:688–693, 2006. 22. Lang BH, Lo CY, Chan WF, et al. Classical and follicular variant of papillary thyroid carcinoma: a comparative study on clinicopathologic features and long-term outcome. World J Surg 30:752–758, 2006. 23. Jain M, Khan A, Patwardhan N, et al. Follicular variant of papillary thyroid carcinoma: A comparative study of histopathologic features and cytology results in 141 patients. Endocr Pract 7:79–84, 2001. 24. Logani S, Gupta PK, LiVolsi VA, et al. Thyroid nodules with FNA cytology suspicious for follicular variant of papillary thyroid carcinoma: follow-up and management. Diagn Cytopathol 23:380–385, 2000 25. Cooper DS, Doherty GM, Haugen BR, et al. Management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid 16:109–142, 2006 26. Baloch ZW, Cibas ES, Clark DP, et al. The National Cancer Institute Thyroid fine needle aspiration state of the science conference: A summation. Cytojournal 5:6, 2008 27. Theoharis CG, Schofield KM, Hammers L, et al. The Bethesda thyroid fine-needle aspiration classification system: year 1 at an academic institution. Thyroid 19:1215–1223, 2009. 28. Baloch ZW, Fleisher S, LiVolsi VA, Gupta PK. Diagnosis of “follicular neoplasm”: A gray zone in thyroid fine-needle aspiration cytology. Diagn Cytopathol 26:41–44, 2002. 29. Yang GC, Goldberg JD, Ye PX. Risk of malignancy in follicular neoplasms without nuclear atypia statistical analysis of 397 thyroidectomies. Endocr Pract 9:510–516, 2003. 30. Raber W, Kaserer K, Niederle B, Vierhapper H. Risk Factors for Malignancy of Thyroid Nodules Initially Identified as Follicular
Endocr Pathol
31.
32.
33.
34.
Neoplasia by Fine-Needle Aspiration: Results of a Prospective Study of One Hundred Twenty Patients. Thyroid 10:709–712, 2000. Marhefka GD, McDivitt JD, Shakir KM, Drake AJ 3rd Diagnosis of follicular neoplasm in thyroid nodules by fine needle aspiration cytology: does the result, benign vs. suspicious for a malignant process, in these nodules make a difference? Acta Cytol. 53:517–523, 2009. Fu L, Chhieng D, Theoharis C, et al. Impact of Second Review of Thyroid FNA on the Implementation of Bethesda Classification: An Analysis of Cases on a Consult Service. Modern Pathol. 26(S2):89A, 2013. Yoon JH, Kim E, Hong SW, et al. Sonographic features of the follicular variant of papillary thyroid carcinoma. J Ultrasound Med 27:1431–1437, 2008. Gupta S, Ajise O, Dultz L, et al. Follicular variant of papillary thyroid cancer, encapsulated, non-encapsulated and diffuse: Distinct biologic and clinical entities. Arch Otolaryngol Head Neck Surg. 138:227– 233, 2012.
35. Proietti A, Giannini R, Ugolini C, et al. BRAF status of follicular variant of papillary thyroid carcinoma and its rela\tionship to its clinical and cytological features. Thyroid 20:1263– 1270, 2010. 36. Fadda G, Rossi ED, Raffaelli M, et al. Follicular thyroid neoplasms can be classified as low- and high-risk according to HBME-1 and Galectin-3 expression on liquid-based fine-needle cytology. Eur J Endocrinol. 165:447–453, 2011. 37. Cochand-Priollet B, Dahan H, Laloi-Michelin M, et al. Immunocytochemistry with cytokeratin 19 and anti-human mesothelial cell antibody (HBME1) increases the diagnostic accuracy of thyroid fine-needle aspirations: preliminary report of 150 liquidbased fine-needle aspirations with histological control. Thyroid. 21:1067–1073, 2011. 38. Rossi ED, Raffaelli M, Minimo C, et al. Immunocytochemical evaluation of thyroid neoplasms on thin-layer smears from fine-needle aspiration biopsies. Cancer. 105:87–95, 2005
Follicular Variant of Papillary Thyroid Carcinoma: Accuracy of FNA Diagnosis and Implications for Patient Management Berrin Ustun & David Chhieng & Manju L. Prasad & Elizabeth Holt & Lynwood Hammers & Tobias Carling & Robert Udelsman & Adebowale J. Adeniran
# Springer Science+Business Media New York 2014
Abstract Follicular variant of papillary thyroid carcinoma (FVPTC) creates a continuous diagnostic dilemma among pathologists because of the paucity of nuclear changes of papillary carcinoma and overlapping features with benign and other neoplastic follicular lesions. Current guidelines for the management of thyroid nodules recommend surgery for confirmed PTC, suspicious for PTC, and follicular neoplasm cases, while further immediate diagnostic studies or treatment are not routinely required if the nodule is benign on cytology. This study is designed to determine the accuracy of cytology in the diagnosis of FVPTC, based on the Bethesda classification system, and determine the implications for patient management based on the current recommendation. Based on a retrospective review of cytologic diagnoses between January 2008 and December 2011, thyroid fine needle aspiration (FNA) cytology specimens with subsequent surgical intervention and a final diagnosis of FVPTC were selected. The cytologic diagnoses were compared with the final diagnoses, and the percentage of cases contributing to the final diagnosis of FVPTC was calculated for each diagnostic category. Triage efficiency and diagnostic accuracy were calculated. One hundred and fifty-two cases with histologic confirmation of B. Ustun : D. Chhieng : M. L. Prasad : A. J. Adeniran (*) Department of Pathology, Yale University School of Medicine, 310 Cedar Street, CB 510A, New Haven, CT 06510, USA e-mail: [email protected] E. Holt Department of Internal Medicine, Yale University School of Medicine, New Haven, CT, USA L. Hammers Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT, USA T. Carling : R. Udelsman Department of Surgery, Yale University School of Medicine, New Haven, CT, USA
FVPTC were identified (representing 128 patients—101 female, 27 male). All patients had undergone either lobectomy with completion thyroidectomy or total thyroidectomy. The cytologic diagnosis of “positive for malignancy” accounted for only 27 % of the final histologic diagnosis of FVPTC, while suspicious for carcinoma, follicular neoplasm, follicular lesion of undetermined significance, and benign accounted for 11, 23, 23, and 16 % of the final diagnosis of FVPTC, respectively. Only 18 % of the 55 cases tested were positive for BRAF mutation. The subtle nuclear features of FVPTC pose challenges for an accurate diagnosis. Therefore, a better approach is to triage these cases for surgical intervention and/ or further evaluation of the particular nodule. Our triage efficacy for FVPTC was 84 %; however, the diagnostic accuracy of PTC was 38 %. A negative diagnosis on FNA has diagnostic and management implications for up to 16 % of cases because they may have no further immediate diagnostic studies or treatment. BRAF mutation analysis provides minimal effect on diagnostic accuracy. Keywords Follicular variant of papillary thyroid carcinoma . Fine needle aspiration . Diagnostic accuracy . Triage efficiency
Introduction Palpable thyroid nodules are detected in about 4 to 7 % of the adult population. Only a small portion of these detected nodules are malignant [1]. The very high prevalence of the thyroid nodules in the general population and low rate of malignancy underscore the need for a reliable preoperative evaluation of any given nodule. To this end, the procedure of choice for triaging patients with palpable thyroid nodules is fine needle aspiration (FNA) biopsy [2]. Papillary thyroid carcinoma (PTC) is the most common type of malignancy of the thyroid gland, accounting for 70 to
Endocr Pathol
80 % of all thyroid cancers [3]. Numerous histological variants of PTC have been described in the literature, and some are known to have prognostic significance [4]. The follicular variant of PTC (FVPTC) is the most common variant after classic PTC, making up 10 to 15 % of all PTC [5, 6]. FNA biopsy is very sensitive for the diagnosis of the conventional PTC. On the contrary, reported sensitivity of FNA in FVPTC is significantly low [7–11]. Given this low sensitivity, FVPTC poses a significant diagnostic challenge to pathologists, endocrinologists, and surgeons [12]. The diagnosis of FVPTC is made by finding the typical nuclear features, which are seen in classic PTC. In theory, the rate and ease of diagnosis of FVPTC should be similar to classic PTC since the diagnostic criteria are identical. However cytological diagnosis of FVPTC is challenging and poses a diagnostic problem for the following reasons: (1) nuclear features of PTC are less obvious and focal, (2) absence of papillary groups or large syncytial sheets, and (3) predominance of follicular architecture with rare nuclear alteration and presence of variable colloid component [13–15]. Another important reason for the difficulty on cytomorphology is the overlap in histologic appearance between FVPTC and other lesions in the follicular neoplasm category, namely, follicular adenoma, follicular carcinoma, and benign nonneoplastic follicular lesions [16–18]. For these reasons, it is not uncommon for the diagnosis of FVPTC to be missed on FNA. In the literature, the studies looking into clinical behaviors and long-term survival similarities and differences between FVPTC and classic PTC reported different results. A recent study with long-term follow-up showed that FVPTC generally has favorable clinicopathologic features, which translates into less nodal metastasis and extrathyroidal extension. However, the mean survival for both groups of patients was not significantly different [19–22]. This current study was designed to determine the triage efficiency and diagnostic accuracy of cytology in the diagnosis of FVPTC, based on the modified Bethesda classification system, and to determine the implications for patient management based on the current recommendation for intervention. Given that the diagnosis of FVPTC is challenging, we propose that the goal of FNA diagnosis should be to improve the triage efficiency of patients with FVPTC for surgery or repeat FNA. This will ensure that patients with equivocal diagnosis have appropriate follow-up.
Materials and Methods A retrospective review of the files from Yale-New Haven Hospital’s Department of Pathology for the period from January 2008 to December 2010 identified a total of 152 FNA biopsy cases in which the patients underwent thyroid resection and the histological diagnosis showed FVPTC. Of the 152,
eight were excluded from this study due to nondiagnostic samples. Of the 144 thyroid FNA specimens, 74 specimens were processed in-house at Yale-New Haven Hospital. The inhouse thyroid FNAs were performed by a radiologist or endocrinologist using a 25-gauge needle. Two to four passes were obtained in most of the cases. For each thyroid nodule, direct smears (air-dried Diff-Quik-stained slides [Dade Behring, Deerfield, IL, USA] and alcohol-fixed Papanicolaoustained slides) were obtained. The needle was then rinsed in Cytorich Red solution (Thermo Fisher Scientific) for ThinPrep liquid-based preparation (Hologic, Marlborough, MA, USA). Seventy FNA specimens were prepared at outside institutions, and they were submitted with one or more of Papanicolaou-stained slides, Diff-Quik stained-slides, and ThinPrep liquid-based preparations. Depending upon submitting institutions, occasional cell-block slides were received. Cytologic interpretation and classification was based on the modified Bethesda System for Reporting Thyroid Cytopathology (BSRTC). At our institution, transition to the BSRTC took place in January 2008. The BSRTC categories were composed of the following: nondiagnostic/unsatisfactory, benign, indeterminate (AUS/FLUS), follicular neoplasm, suspicious for malignancy, and positive for malignancy. As a result, after the exclusion of the nondiagnostic cases, a total of 144 thyroid FNA cases from 128 patients were included in this retrospective study. The final cytologic and histologic diagnoses were identified from the pathology report. The cytologic diagnoses were rendered by seven board-certified cytopathologists with a range of years of experience from 3 to 10+ years. The triage efficiency in this study considered the FNA diagnoses that would potentially trigger surgical intervention or repeat FNA biopsy, and this included the following: positive for PTC, suspicious for PTC, follicular neoplasm, and indeterminate (AUS/FLUS). Diagnostic accuracy was considered as making the accurate diagnosis of PTC and that included positive for PTC and suspicious for PTC.
Results A total of 144 thyroid FNA samples from 128 patients were identified and reviewed. Sixteen patients had a repeat FNA biopsy. The clinical, pathologic, and molecular characteristics are highlighted in Table 1. The patient population was predominantly female (101 female and 27 male) and ranged in age from 14 to 80 years (median age, 51 years). Forty-three patients (representing 34 %) were aged 45 years and below, while the majority of the patients (85; 66 %) were older than 45 years. This is clinically important, given the significance that age has on the staging of these tumors. All patients had confirmed FVPTC on histology. Of the 128 patients, 28 had
Endocr Pathol Table 1 Clinical, pathologic, and molecular characteristics of FVPTC
Table 2 Distribution of cytologic diagnosis
Characteristics
Patients, number (%)
Cytologic diagnosis
Subdiagnosis
Number of cases
51 43 (34) 85 (66)
Positive Suspicious Follicular neoplasm
PTC PTC
39 (27 %) 16 (11 %) 33 (23 %) 10
Age Median (years) ≤45 >45 Sex Female Male Tumor size (cm) Median ≤4 >4 Vascular invasion Present Absent Extrathyroidal extension Present Absent Central nodal metastases Present Absent CBA Thyroid surgery Lobectomy Total thyroidectomy ± CLND BRAF V600E mutation Present Absent CBA
Hurthle cell neoplasm Microfollicular-patterned neoplasm FVPTC
101 (79) 27 (21) Indeterminate 1.3 122 (95) 6 (5) 6 (5) 122 (95) 5 (4) 123 (96) 22 (17) 67 (52) 39 (31) 28 (22) 100 (78) 10 (8) 45 (35) 73 (57)
CBA cannot be assessed
undergone lobectomy with completion thyroidectomy, 48 had undergone total thyroidectomy, 38 had undergone total thyroidectomy with lymph node dissection, and 14 had undergone total thyroidectomy with intraoperative frozen section. On histologic examination, vascular invasion was identified in 6 cases (5 %), while extrathyroidal extension and central lymph node metastases were present in 5 (4 %) and 22 (17 %) cases, respectively. Overall, final pathologic staging on surgical resections were as follows: pT1 (104/144, 72 %), pT2 (24/144, 17 %), and pT3 (16/144, 11 %). All the patients with lymph node metastasis had metastasis to central neck lymph nodes (N1 disease). The overall mean size of the resected thyroid nodules was 1.3 cm (range 0.1 to 11.3 cm). By using the modified 2007 National Cancer Institute (NCI)/ Bethesda guidelines, FNA diagnosis resulted in the following distribution of cytological diagnosis: positive for PTC (39/ 144, 27 %), suspicious for PTC (16/144, 11 %), indeterminate (33/144, 23 %), follicular neoplasm (33/144, 23 %), and negative (23/144, 16 %) (Table 2).
Microfollicular patterned (INa) Nuclear atypia (INb) Negative Goiter Lymphocytic thyroiditis All cases
11 12 33 (23 %) 2 31 23 (16 %) 19 4 144 (100 %)
Of the 23 negative cases, the cytologic diagnosis was predominantly rendered as hyperplastic nodule (goiter) (19/ 23, 83 %) while lymphocytic thyroiditis accounted for the remaining 4 cases. Of these 23 nodules with a negative cytologic diagnosis, 16 (70 %) measured less than 1 cm in the greatest dimension on surgical resection. The mean size of the resected nodules in the negative diagnostic category was 0.7 cm (range, 0.2 to 2.5 cm). Sixteen of these 23 cases were in-house cases, and a review of the cytology slides revealed that 12 of the cases (80 %) displayed no atypical nuclear features and 4 (20 %) displayed subtle nuclear clearing and rare nuclear enlargement and nuclear groves in small subsets of follicular cells which were unrecognized by the pathologists at the time the negative cytological interpretation was rendered (Figs. 1 and 2). These cases were stratified into encapsulated and nonencapsulated FVPTC based on the presence or absence of a capsule on histological evaluation of the
Fig. 1 Follicular variant of papillary thyroid carcinoma. Cellular aspirate showing crowded groups. Most of the nuclei are round and uniform, but there is a small subpopulation with nuclear grooves. (Papanicolaou stain, original magnification ×200)
Endocr Pathol Table 3 Clinical, pathologic, and molecular characteristics of FVPTC cases diagnosed as negative on fine needle aspiration based on stratification into EFVPTC vs. NEFVPTC Characteristics
Fig. 2 Follicular variant of papillary thyroid carcinoma. Crowded groups showing nuclear enlargement and rare nuclear grooves. Scant sticky colloid is present. (Diff-Quik stain, original magnification ×400)
tumors (Table 3). The two histologic subtypes are identical in terms of median age and number of patients aged 45 years and below. There were no significant differences in terms of vascular invasion, extrathyroidal extension, and central lymph node metastases between the two groups. The sonographic characteristics of these 23 cases were also evaluated to determine if there were unique features that could help in better triage of these negative cases (Table 4). The nodules were predominantly solid (91 %) and hypoechoic (65 %). Other sonographic findings included absence of microcalcifications (65 %) and presence of well-defined, regular margins (61 %). The shape of the nodule was insignificant as almost half were oval while the other half had nodules which were taller than wide. At our institution, the term “indeterminate” corresponds to the NCI 2007 guidelines category “Follicular cells of undetermined significance/Atypia of undetermined significance (FLUS/AUS).” This category is further subclassified into two diagnostic subtypes: (1) FNA specimens showing scant cellularity with a predominantly microfollicular architecture and scant colloid, which raises the concern for follicular neoplasm (INa) and (2) FNA specimens showing subtle atypical nuclear features, which raises the concern for papillary thyroid carcinoma (INb). The INa category accounted for 2 of the 33 indeterminate cases (6 %), while the INb category accounted for 31 (94 %). Follicular neoplasm category comprises three diagnostic subtypes as follows: (1) cellular FNA sample showing predominantly microfollicular architecture (MF), (2) some features suggestive of but not diagnostic for papillary carcinoma follicular variant (FL), and (3) cellular FNA specimen showing predominantly Hurthle cells with scant colloid (HCN). The distribution of cytologic diagnosis of cases in the follicular neoplasm category is as follows: MF (11/33, 33 %), FL (12/33, 37 %), and HCN (10/33, 30 %). Sixteen of the 144 cases (11 %) had a cytologic diagnosis of “suspicious for PTC,” while 39 (27 %) had a cytologic
Age Median (years) ≤45 >45 Sex Female Male Tumor size (cm) Median ≤4 >4 Vascular invasion Present Absent Extrathyroidal extension Present Absent Central nodal metastases Present Absent CBA Thyroid surgery Lobectomy Total thyroidectomy ± CLND BRAF V600E mutation Present Absent CBA
Encapsulated (n=8)
Nonencapsulated (n=15)
45 5 3
49 4 11
7 0
11 3
0.8 8 0
0.4 15 0
0 8
0 15
0 8
1 14
1 5 2
3 7 5
1 7
3 12
1 3 4
1 2 12
diagnosis of “positive for PTC.” The calculated triage efficiency (potential for triggering surgical intervention or repeat FNA) is 84 % (121/144). The diagnostic accuracy is 38 % (54/ 144) (Table 5). The in-house cytologic specimens with the diagnoses of indeterminate, suspicious for PTC, and positive for PTC were triaged for diagnostic and/or prognostic BRAF mutation analysis. Ten (18 %) of the 55 cases tested harbored the BRAF V600E mutation. (Table 1)
Discussion Classic PTC and its variants represent the most common malignancy, with definitive preoperative diagnosis of classic PTC being around 94 % [2, 14]. The definitive preoperative cytological diagnosis of FVPTC is generally low. Most of the studies have reported sensitivities ranging from 9 to 37 % [5,
Endocr Pathol Table 4 Sonographic findings in FVPTC
Characteristics
Number (%)
Consistency Solid Cystic Mixed Echogenicity
21 (91) 0 (0) 2 (9)
Hypoechoic Isoechoic Microcalcification Present Absent Margin Regular Irregular Shape Oval Taller than wide
15 (65) 8 (35) 8 (35) 15 (65) 14 (61) 9 (39) 12 (52) 11 (48)
7–11, 17]. Tielens et al. [19] reported a sensitivity of 75 %; however, the results of the study are misleading as patients with suspicious FNA were included as true-positive results, leading to an apparent increase in calculated sensitivity. The diagnostic dilemma in the preoperative detection of FVPTC is not only in enhancing the specificity of preoperative diagnosis but also to simultaneously increase the sensitivity of detection so that patients with malignancy are discovered soon enough to allow effective therapy. In our present study, the definitive preoperative diagnosis was 27 % (39/144). But our instructional diagnostic accuracy was 38 % (54/144), which included both diagnostic categories of positive for PTC as well as suspicious for PTC. One might argue about the inclusion of the suspicious category in the diagnostic accuracy. We defined diagnostic accuracy as the ability to identify nodules requiring immediate attention. This includes the categories of positive and suspicious for malignancy. More so, the category of suspicious is added because in most cases when a diagnosis of suspicious for malignancy is made, there is a reasonably high inclination towards Table 5 Triage efficiency Cytologic diagnosis
Number of cases (%)
Current intervention
Positive
39 (27)
Suspicious Follicular neoplasm Indeterminate Negative All cases (ex-nondiagnostic)
16 (11) 33 (23) 33 (23) 23 (16) 144 (100)
Total thyroidectomy Lobectomy Lobectomy Repeat FNA No intervention
malignancy but there is a holdback usually due to scant evidence. Many studies have shown that precise recognition of FVPTC is low due to the rarity of the nuclear features, which are necessary to making a definitive diagnosis. Separate studies by Jain et al. and Logani et al. concluded that majority of the cytology specimens from patients with FVPTC were interpreted as suspicious for malignant lesion or follicular neoplasm [23, 24]. In this present study, 23 % (33/144) resulted in an indeterminate result. The indeterminate category in our institution comprises the lesions in the FLUS/AUS category of the Bethesda classification. Majority of the cases were called indeterminate due to nuclear atypia (31/33, 94 %), and rare cases (2/33, 6 %) were called indeterminate due to architectural atypia with the low cellularity and predominantly microfollicular pattern. Only one of those cases with an indeterminate call had a repeat FNA biopsy, which resulted in subsequent indeterminate cytologic diagnosis. In this particular study, the majority of the cases in the indeterminate category did not have a repeat FNA as recommended by current Bethesda guideline but rather had surgical intervention because of other concomitant risk factors. If they did have a repeat FNA, this would upgrade or possibly downgrade the original cytological diagnosis of indeterminate and consequently would impact the triage efficiency and diagnostic accuracy. Twenty-three percent (33/144) of our cases were in the category of follicular neoplasm, which should trigger immediate surgical intervention (lobectomy) according to the thyroid cancer guidelines [25]. For follicular neoplasms, most investigators believe that intraoperative evaluation does not provide accurate definitive diagnosis and a unilateral thyroid lobectomy is the standard initial surgical intervention [5]. The NCI Thyroid FNA State of the Science Conference held in 2007 proposed six basic diagnostic classifications for the reporting of thyroid FNA specimens [26]. At our institution [27], we use the modified Bethesda classification system for classifying thyroid FNA specimens. We further subclassify the follicular neoplasm category into three subcategories: (1) microfollicular-patterned neoplasm, (2) follicular lesion with some features suggestive of but not diagnostic of FVPTC, and (3) Hurthle cell neoplasm. Of the cytological diagnosis of follicular neoplasm, 37 % (12/33) exhibited enough atypical nuclear features which are beyond the indeterminate (AUS/ FLUS) diagnostic category but not enough to place those cases in the suspicious for PTC category. In our experience, there are a number of cases in which some nuclear features of PTC (most commonly nuclear clearing and rare grooves) are present. The frequency of these nuclear features varies from case to case but generally not enough nuclear changes to place those cases in the suspicious for PTC category. At our institution, our experience has been that the subcategory of follicular lesion with some features suggestive of but
Endocr Pathol
not diagnostic for FVPTC (FL) allows the pathologist to characterize follicular lesion one step further and gives more information to the clinician as to the level of concern that the lesions in this subcategory elicit. It is to ensure that patients with lesions in this subcategory are ultimately referred for surgical consultation. Two of the 12 cases diagnosed as FL on cytology had frozen sections performed at the time of surgery and were upgraded to positive for PTC and had total thyroidectomies based on the intraoperative consultation. The rest of the follicular neoplasm diagnoses were simply called either Hurthle cell neoplasm (10/33, 30 %) or microfollicularpatterned neoplasm (11/33, 33 %), which in turn puts these cases in the gray zone which will include adenomatous goiter, follicular adenoma, follicular carcinoma, Hurthle cell neoplasm, and, less likely, FVPTC [28, 29]. The reported rate of malignancy with the cytological diagnosis of follicular neoplasm varies but generally ranges from 8 to 30 % [30, 31]. As a result, even though the Bethesda guidelines recommend lobectomy for a cytological diagnosis of follicular neoplasm, there are still a good number of institutions where the Bethesda classification is not implemented [32]. Terms such as follicular lesion are used; hence, it may not be a routine practice in these institutions to perform immediate lobectomy when the term “follicular lesion” is used in the place of “follicular neoplasm.” In our experience, any cellular follicular neoplasm with subtle nuclear atypia should be assumed to be FVPTC until proven otherwise and careful search for the subtle nuclear features of PTC should be performed so that FVPTC can be distinguished from other lesions in the follicular neoplasm category (Fig. 3). This is the goal of subclassification of follicular neoplasm, and in this way, the pathologist will deliver more useful information to the clinician. We evaluated common sonographic findings of FVPTC in all cases with a negative cytologic diagnosis in our present study. The small sample size is admittedly a limitation, but it
Fig. 3 Follicular variant of papillary thyroid carcinoma. Morphologic appearance of follicular architecture and subtle nuclear features such as clearing, enlargement, and grooves (H & E stain, original magnification ×400)
does indicate that solid lesions, hypoechogenicity, absence of microcalcifications, and presence of well-defined regular margins are regular features seen in FVPTC. Unfortunately, these individual features are also seen in benign thyroid lesions to variable degrees; hence, the presence of a single individual sonographic characteristic is insufficient to suspect FVPTC. All the sonographic characteristics must be evaluated in conjunction with the clinical impression, and the intervention is based on the overall assessment. One major limitation in this study was our inability to describe the pattern of vascularity in these lesions. Even though this is not a peculiar feature of malignancy, marked intrinsic hypervascularity with increased flow in the central portion of the nodule relative to surrounding thyroid parenchyma is the most commonly described vascularity pattern in thyroid malignancies [33]. We stratified the 16 cases of FVPTC with benign cytologic diagnosis into two—encapsulated and nonencapsulated—to determine if there was a difference in the biologic behavior of the tumors in both categories. Even though it does not appear that there is a difference in biologic potential, the small sample size is a major limitation as one cannot reasonably draw a meaningful conclusion based on this. However, studies have shown that the nonencapsulated variant behaves more like classic PTC, with a significantly higher rate of BRAF V600E mutations and nodal metastases [34]. BRAF is mutated in a small percentage of FVPTC. Most of these mutated cases are usually recognized by the pathologist either as suspicious or positive for PTC on FNA. As a result of this, BRAF analysis has a limited value in preoperative diagnosis of FVPTC [35]. In this study, 18 % of the 55 cases harbored the BRAF mutation. In turn, BRAF molecular test neither improves triage efficacy nor diagnostic accuracy. There are several studies that have utilized ancillary studies to improve the morphologic evaluation of indeterminate lesions, enabling a better selection of patients for surgery. The expression of HBME-1, cytokeratin 19, and galectin-3 has been used to classify cytologically indeterminate lesions into high-risk (those that need immediate surgery) and low-risk (those that can be followed up) groups [36, 37]. The addition of RET proto-oncogene to HBME-1 and galectin-3 has been found to increase the overall diagnostic accuracy of cytology in follicular proliferations to 97 % [38]. These immunohistochemical stains can be performed on liquid-based preparations. It should be noted, however, that many of these markers are also focally expressed in benign nodular goiters, so they are used in conjunction with one another and with careful cytologic and morphologic evaluation. In summary, the diagnosis of FVPTC on FNA biopsy can be very challenging for the pathologist. The definitive diagnosis of FVPTC is not always possible. However, FNA biopsy plays an important role in identifying these patients and triaging them appropriately. While evaluating any follicular lesion, consideration should be given to FVPTC and strict
Endocr Pathol
cytologic criteria should be applied to identify these patients. Pathologists should develop a very low threshold for detection of FVPTC, and careful search for the presence of subtle nuclear atypia is critical to achieving this. These features can be seen focally and admixed with colloid and benign follicular epithelium. Our study suggests that emphasizing triage efficiency (i.e., FNA classification that triggers surgical intervention or repeat FNA), rather than diagnostic accuracy, might improve the clinical management of patients with FVPTC. The ultimate goal should be to minimize the negative diagnosis so that patients would not be lost to follow-up. The triage efficiency in our study was 84 %, an indication that up to one in six cases of FVPTC could still be missed and these cases could potentially have no intervention. In the 16 % of FVPTC where a benign diagnosis was rendered on FNA, most were microcarcinomas and no specific clinical or radiologic characteristics distinguished these cases from the other FVPTC cases that were appropriately triaged by FNA.
Funding This study was supported by funding from the Department of Pathology, Yale University School of Medicine, New Haven, CT. Financial Disclosures The authors have no financial disclosures.
References 1. Ramacciotti CE, Pretorius HT, Chu EW, et al. Diagnostic accuracy and use of aspiration biopsy in the management of thyroid nodules. Arch Intern Med 144:1169–1173, 1984. 2. Wang HH Reporting thyroid fine needle aspiration: Literature review and a proposal. Diagn Cytopathol 34:67–76, 2006. 3. Adeniran AJ, Zhu Z, Gandhi M, et al. Correlation between genetic alterations and microscopic features, clinical manifestations, and prognostic characteristics of thyroid papillary carcinomas. Am J Surg Pathol 30:216–222, 2006. 4. Lam AK, Lo CY, Lam KS. Papillary carcinoma of thyroid: A 30-yr clinicopathological review of the histological variants. Endocrine Pathol 16:323–30,2005. 5. Kesmodel SB, Terhune KP, Canter RJ, et al. The diagnostic dilemma of follicular variant of papillary thyroid carcinoma. Surgery 134: 1005–1012, 2003. 6. Shih SR, Shun CT, Su DH, et al. Follicular variant of papillary thyroid carcinoma: Diagnostic limitations of fine needle aspiration cytology. Acta Cytol 49:383–386, 2005. 7. Lin HS, Komisar A, Opher E, Blaugrund SM. Follicular variant of papillary carcinoma: the diagnostic limitations of preoperative fineneedle aspiration and intraoperative frozen section evaluation. Laryngoscope 110:1431–1436, 2000. 8. Caraway NP, Sneige N, Samaan NA. Diagnostic pitfalls in thyroid fine-needle aspiration: a review of 394 cases. Diagn Cytopathol 9: 345–50, 1993. 9. Powari M, Dey P, Saikia UN. Fine needle aspiration cytology of follicular variant of papillary carcinoma of thyroid. Cytopathology 14:212–215, 2003.
10. Yang J, Schnadig V, Logrono R, Wasserman PG. Fine-needle aspiration of thyroid nodules: A study of 4703 patients with histologic and clinical correlations. Cancer 111:306–315, 2007. 11. Miller JM, Hamburger JI, Kini SR. The needle biopsy diagnosis of papillary thyroid carcinoma. Cancer 48:989 –993, 1981. 12. Salajegheh A, Petcu EB, Smith RA, Lam AK. Follicular variant of papillary thyroid carcinoma: a diagnostic challenge for clinicians and pathologists. Postgrad Med J 84:78–82, 2008. 13. El Hag IA, Kollur SM. Benign follicular thyroid lesions versus follicular variant of papillary carcinoma: Differentiation by architectural pattern. Cytopathology 15:200–205, 2004. 14. Zacks JF, de las Morenas A, Beazley RM, O’Brien MJ. Fine-needle aspiration cytology diagnosis of colloid nodule versus follicular variant of papillary carcinoma of the thyroid. Diagn Cytopathol 18: 87–90, 1998. 15. Wu HH, Jones JN, Grzybicki DM, Elsheikh TM. Sensitive cytologic criteria for the identification of follicular variant of papillary thyroid carcinoma in fine-needle aspiration biopsy. Diagn Cytopathol 29: 262–266, 2003. 16. Williams MD, Suliburk JW, Staerkel GA, et al. Clinical significance of distinguishing between follicular lesion and follicular neoplasm in thyroid fine-needle aspiration biopsy. Ann Surg Oncol 16:3146– 3153, 2009. 17. Jogai S, Adesina AO, Temmim L, et al. Follicular variant of papillary thyroid carcinoma—A cytological study. Cytopathology 15:212– 216, 2004. 18. Dobrinja C, Trevisan G, Liguori G, et al. Sensitivity evaluation of fine-needle aspiration cytology in thyroid lesions. Diagn Cytopathol 37:230–235, 2009. 19. Tielens ET, Sherman SI, Hruban RH, Ladenson PW. Follicular variant of papillary thyroid carcinoma. A clinicopathologic study. Cancer 73:424–31, 1994. 20. Lin HW, Bhattacharyya N. Clinical behavior of follicular variant of papillary thyroid carcinoma: presentation and survival. Laryngoscope 120:712–716, 2010. 21. Chang HY, Lin JD, Chou SC, et al. Clinical presentations and outcomes of surgical treatment of follicular variant of the papillary thyroid carcinomas. Jpn J Clin Oncol 36:688–693, 2006. 22. Lang BH, Lo CY, Chan WF, et al. Classical and follicular variant of papillary thyroid carcinoma: a comparative study on clinicopathologic features and long-term outcome. World J Surg 30:752–758, 2006. 23. Jain M, Khan A, Patwardhan N, et al. Follicular variant of papillary thyroid carcinoma: A comparative study of histopathologic features and cytology results in 141 patients. Endocr Pract 7:79–84, 2001. 24. Logani S, Gupta PK, LiVolsi VA, et al. Thyroid nodules with FNA cytology suspicious for follicular variant of papillary thyroid carcinoma: follow-up and management. Diagn Cytopathol 23:380–385, 2000 25. Cooper DS, Doherty GM, Haugen BR, et al. Management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid 16:109–142, 2006 26. Baloch ZW, Cibas ES, Clark DP, et al. The National Cancer Institute Thyroid fine needle aspiration state of the science conference: A summation. Cytojournal 5:6, 2008 27. Theoharis CG, Schofield KM, Hammers L, et al. The Bethesda thyroid fine-needle aspiration classification system: year 1 at an academic institution. Thyroid 19:1215–1223, 2009. 28. Baloch ZW, Fleisher S, LiVolsi VA, Gupta PK. Diagnosis of “follicular neoplasm”: A gray zone in thyroid fine-needle aspiration cytology. Diagn Cytopathol 26:41–44, 2002. 29. Yang GC, Goldberg JD, Ye PX. Risk of malignancy in follicular neoplasms without nuclear atypia statistical analysis of 397 thyroidectomies. Endocr Pract 9:510–516, 2003. 30. Raber W, Kaserer K, Niederle B, Vierhapper H. Risk Factors for Malignancy of Thyroid Nodules Initially Identified as Follicular
Endocr Pathol
31.
32.
33.
34.
Neoplasia by Fine-Needle Aspiration: Results of a Prospective Study of One Hundred Twenty Patients. Thyroid 10:709–712, 2000. Marhefka GD, McDivitt JD, Shakir KM, Drake AJ 3rd Diagnosis of follicular neoplasm in thyroid nodules by fine needle aspiration cytology: does the result, benign vs. suspicious for a malignant process, in these nodules make a difference? Acta Cytol. 53:517–523, 2009. Fu L, Chhieng D, Theoharis C, et al. Impact of Second Review of Thyroid FNA on the Implementation of Bethesda Classification: An Analysis of Cases on a Consult Service. Modern Pathol. 26(S2):89A, 2013. Yoon JH, Kim E, Hong SW, et al. Sonographic features of the follicular variant of papillary thyroid carcinoma. J Ultrasound Med 27:1431–1437, 2008. Gupta S, Ajise O, Dultz L, et al. Follicular variant of papillary thyroid cancer, encapsulated, non-encapsulated and diffuse: Distinct biologic and clinical entities. Arch Otolaryngol Head Neck Surg. 138:227– 233, 2012.
35. Proietti A, Giannini R, Ugolini C, et al. BRAF status of follicular variant of papillary thyroid carcinoma and its rela\tionship to its clinical and cytological features. Thyroid 20:1263– 1270, 2010. 36. Fadda G, Rossi ED, Raffaelli M, et al. Follicular thyroid neoplasms can be classified as low- and high-risk according to HBME-1 and Galectin-3 expression on liquid-based fine-needle cytology. Eur J Endocrinol. 165:447–453, 2011. 37. Cochand-Priollet B, Dahan H, Laloi-Michelin M, et al. Immunocytochemistry with cytokeratin 19 and anti-human mesothelial cell antibody (HBME1) increases the diagnostic accuracy of thyroid fine-needle aspirations: preliminary report of 150 liquidbased fine-needle aspirations with histological control. Thyroid. 21:1067–1073, 2011. 38. Rossi ED, Raffaelli M, Minimo C, et al. Immunocytochemical evaluation of thyroid neoplasms on thin-layer smears from fine-needle aspiration biopsies. Cancer. 105:87–95, 2005
