Usefulness of Diagnostic Qualifiers for Thyroid Fine-Needle Aspirations With Atypia of Undetermined Significance

Transcription

1 Anatomic Pathology / AUS Qualifiers in Thyroid FNAs Usefulness of Diagnostic Qualifiers for Thyroid Fine-Needle Aspirations With Atypia of Undetermined Significance Paul A. VanderLaan, MD, PhD, 1 Ellen Marqusee, MD, 2 and Jeffrey F. Krane, MD, PhD 1 Key Words: Thyroid; Fine-needle aspiration; Cytology; Atypia of undetermined significance Abstract The diagnostic category of atypia of undetermined significance (AUS) in the Bethesda System for reporting the results of thyroid fine-needle aspirations (FNAs) is intended to encompass findings associated with a low risk of malignancy. It is unclear if there are patterns within this evolving, heterogeneous category associated with differing risk of malignancy that might warrant alternative classification or clinical management. Therefore, a retrospective review of 512 AUS FNAs from January 2005 to May 2009 was done. Most malignancies associated with AUS were papillary carcinoma (86/96 [90%]), of which 85% (73/86) were follicular variants. Atypia qualifiers were correlated with the follow-up rate of malignancy. The risk of malignancy for architectural atypia alone was approximately half that observed for cytologic, both cytologic and architectural, or unspecified atypia. Architectural atypia alone was less likely to be papillary carcinoma and more likely to be follicular adenoma. The lower risk of malignancy associated with isolated architectural atypia compared with other patterns of AUS should be considered in clinical decision making and in future management guidelines. Thyroid nodule fine-needle aspiration (FNA) has proven to be a reliable, relatively noninvasive method for determining the risk of malignancy of a thyroid nodule, thereby identifying patients most appropriately referred for surgery. The recently proposed reporting schema for thyroid FNAs, born out of the 2007 National Cancer Institute (NCI) Thyroid Fine Needle Aspiration State of the Science Conference 1,2 with the subsequent Bethesda System for Reporting Thyroid Cytopathology, 3 provides a uniform 6-tiered diagnostic framework for classifying thyroid nodule FNAs based on malignancy risk from multiple studies. The category known as atypia of undetermined significance (AUS), or follicular lesion of undetermined significance, is composed of a heterogeneous mix of diagnostic scenarios that are indeterminate but anticipated to have a low risk of malignancy. A number of recent studies have investigated the biologic behavior of thyroid nodules with an AUS diagnosis based on clinical and surgical follow-up data, 4-13 with an observed risk of malignancy ranging from 5% to 48%. This wide range of risk may be partly attributable to the heterogeneous nature of the AUS category, and it remains to be seen if there are particular specimen characteristics that are more strongly associated with malignancy. To this end, we report a retrospective analysis of thyroid FNAs with a diagnosis of AUS with their associated outcome data to determine if any salient diagnostic features exist that could better stratify the subsequent risk of malignancy. Materials and Methods After receiving approval from the institutional review board, a retrospective analysis was conducted of the 4, Am J Clin Pathol 2011;136: Downloaded 572 from

2 Anatomic Pathology / Original Article thyroid FNAs performed at Brigham and Women s Hospital, Boston, MA, from January 2005 to May All FNAs were performed by staff endocrinologists using a 25-gauge needle with ultrasound guidance (typically 3-4 passes). The specimen was collected immediately in CytoLyt (Hologic, Marlborough, MA), and Papanicolaou-stained ThinPrep slides were prepared using the ThinPrep 2000 (Hologic). Rapid on-site evaluation with direct smear preparation was not performed routinely. All cases were reported by a staff cytopathologist using a 6-tiered diagnostic system with diagnostic criteria essentially identical to those of the 2007 National Cancer Institute Thyroid FNA State of the Science Conference guidelines 1,2 and the Bethesda System for Reporting Thyroid Cytopathology. 3 Since adopting a 6-tiered diagnostic system, pathologists in our laboratory have been encouraged, but not required, to use brief descriptive language to characterize the nature of the atypia when reporting an AUS diagnosis. Descriptors of cytologic atypia, architectural atypia, or both cytologic and architectural atypia have been used to broadly describe the atypia, but without providing details (such as rare intranuclear pseudoinclusions) that could be misconstrued clinically as representing a more severe diagnosis (such as suggestive of papillary carcinoma). Representative images of these qualifiers of AUS are shown in Image 1. AUS with architectural atypia encompasses focal or mild features similar to but not sufficient for a diagnosis of A B C D Image 1 A, Architectural atypia: rare microfollicular groups present in an otherwise hypocellular specimen (Papanicolaou, 400). B, Cytologic atypia: occasional group of follicular cells with mild cytologic changes including few nuclear grooves, powdery chromatin, and the presence of small distinct nucleoli (Papanicolaou, 1,000). C, Cytologic and architectural atypia: rare clusters of follicular cells with a combination of mild cytologic changes and microfollicular architecture (Papanicolaou, 1,000). D, Atypia unspecified: follicular cells with mild nuclear crowding and enlargement with slightly pale chromatin (Papanicolaou, 600). Downloaded from Am J Clin Pathol 2011;136:

3 VanderLaan et al / AUS Qualifiers in Thyroid FNAs suspicious for a follicular neoplasm. This would include a predominance of microfollicles in a sparsely cellular specimen or focal crowding/disorder of follicular cells (not attributable to preparation artifact) in the absence of associated nuclear changes. AUS with cytologic atypia is broadly regarded as a specimen with features similar to but not sufficient for a diagnosis of suspicious for papillary thyroid carcinoma or malignant. Here, the cytologic/nuclear features of papillary thyroid carcinoma (nuclear grooves, powdery chromatin, nuclear membrane thickening/irregularity, nuclear molding/ crowding, and rarely nuclear pseudoinclusions) would be present in a small number of cells in a paucicellular or normocellular specimen, or, alternatively, the changes could be mild but more widespread. The diagnosis of AUS with both cytologic and architectural atypia applies to cases with a combination of characteristics from the cytologic and architectural subcategories. A diagnosis of AUS without a cytologic or architectural atypia qualifier could be given for cases including an atypical lymphocytic infiltrate, Hürthle cell change deemed insufficient for a suspicious diagnosis or in a clinical context in which a neoplasm is less probable (such as a goiter or Hashimoto thyroiditis), or atypia complicated by obscuring blood or other technical artifacts. Unqualified AUS could also represent the failure of the responsible cytopathologist to attempt to classify the atypia into the architectural or cytologic category. Additional diagnostic descriptors provided in the original report were also evaluated. These descriptors included a qualitative assessment of the severity of atypia present (eg, mild atypia); the overall cellularity of the specimen (eg, scantly cellular); the qualitative number of atypical cells present (eg, rare atypical cells); and the presence of colloid, histiocytes, hemosiderin-laden macrophages, cyst lining cells, or Hürthle cell changes. The AUS cases compiled were not rereviewed or reclassified retrospectively by us, reflecting the actual use of these atypia qualifiers and additional descriptors in everyday practice. Clinical outcome for the aspirated thyroid nodule was categorized as benign or malignant. A benign outcome was defined as a benign cytologic diagnosis made on followup FNA or a tissue diagnosis of multinodular hyperplasia, follicular adenoma, adenomatous nodule, Hashimoto thyroiditis, or incidental papillary thyroid microcarcinoma (not associated with the targeted FNA nodule) on histologic evaluation. Outcomes were classified as malignant when a histologic diagnosis of malignancy was made on resection of the aspirated nodule. When evaluating the surgical resection specimen, only the target nodule was considered and correlated with the preceding FNA findings. All cases had at least 1 year of clinical follow-up after the initial AUS thyroid FNA. Data processing and statistics were performed by using Microsoft Excel (Microsoft, Redmond, WA) and GraphPad software (GraphPad Software, La Jolla, CA). Categorical analysis was performed by using a 2-tailed χ 2 test or a 2-tailed Fisher exact test when appropriate, with a predetermined level of significance set at a P value of.05. Results Overall outcome data for this cohort have been published elsewhere. 13 Briefly, a diagnosis of AUS was given in 512 cases (10.9%): 99 males and 413 females, mean age, 53 years (range, 9-88 years). From this cohort, 63 (12.3%) were repeated AUS diagnoses for the same thyroid nodule and 118 (23.0%) had incomplete or no further follow-up on record. A total of 199 (38.9%) cases underwent surgical resection. Taken together with cases having a benign repeated FNA, a total of 331 thyroid nodules with an initial FNA diagnosis of AUS had clinical follow-up data available. Based on overall outcome data, 96 nodules were malignant (29.0%), whereas 235 nodules were benign (71.0%). Outcome data were initially stratified by atypia qualifier including benign cytologic follow-up and histologic outcome Table 1. Architectural atypia alone had a malignant outcome (11/70 [16%]) nearly half as frequently as the other qualifier Table 1 Cytologic and Histologic Outcome Data for Atypia of Undetermined Significance in Thyroid Fine-Needle Aspirations Based on Atypia Qualifier * Atypia Qualifier Benign Malignant Total Architectural 84 (59/70) 16 (11/70) 21.1 (70/331) Cytologic 71 (42/59) 29 (17/59) 17.8 (59/331) Cytologic and architectural 72 (60/83) 28 (23/83) 25.1 (83/331) Unspecified 62.2 (74/119) 37.8 (45/119) 36.0 (119/331) All nonarchitectural atypia 67.4 (176/261) 32.6 (85/261) 78.9 (261/331) Total 71.0 (235/331) 29.0 (96/331) 100 (331/331) * Data are given as percentage (number/total) of cases. Architectural atypia vs atypia unspecified, P <.002. Architectural atypia vs all nonarchitectural atypia, P = Am J Clin Pathol 2011;136: Downloaded 574 from

4 Anatomic Pathology / Original Article Table 2 Surgical Outcome Data for Atypia of Undetermined Significance in Thyroid Fine-Needle Aspirations Based on Atypia Qualifier * Atypia Qualifier Benign Malignant Total Architectural 76 (34/45) 24 (11/45) 22.6 (45/199) Cytologic 50 (17/34) 50 (17/34) 17.1 (34/199) Cytologic and architectural 54 (27/50) 46 (23/50) 25.1 (50/199) Unspecified 36 (25/70) 64 (45/70) 35.2 (70/199) Total 51.6 (103/199) 48.2 (96/199) (199/199) * Data are given as percentage (number/total) of cases. Architectural atypia vs cytologic atypia, P <.04. Architectural atypia vs cytologic and architectural atypia, P <.04. Architectural atypia vs atypia unspecified, P = categories: cytologic atypia alone (17/59 [29%]), architectural and cytologic atypia (23/83 [28%]), and atypia unspecified (45/119 [37.8%]). Statistical significance was achieved when comparing architectural atypia alone with the total of the rest of the groups (P =.005) and architectural atypia compared with atypia unspecified (P <.002). No other differences between groups achieved statistical significance. Table 3 Specific Histologic Diagnoses Based on Atypia Qualifier Atypia Qualifier/Diagnosis No. (%) Architectural Benign 34 (100) Adenoma 23 (68) MNH 9 (26) HASH 2 (6) Malignant 11 (100) PTC 11 (100) Cytologic * Benign 17 (100) Adenoma 9 (53) MNH 8 (47) Malignant 17 (100) PTC 17 (100) Cytologic and architectural Benign 27 (100) Adenoma 23 (85) MNH 4 (15) Malignant 23 (100) PTC 20 (87) FTC 3 (13) Unclassified Benign 25 (100) Adenoma 14 (56) MNH 10 (40) HASH 1 (4) Malignant 45 (100) PTC 38 (84) FTC 5 (11) ANA 1 (2) DLBCL 1 (2) ANA, anaplastic (undifferentiated) thyroid carcinoma; DLBCL, diffuse large B-cell lymphoma; FTC, follicular thyroid carcinoma; HASH, chronic lymphocytic thyroiditis (Hashimoto thyroiditis); MNH, multinodular hyperplasia; PTC, papillary thyroid carcinoma. * Architectural atypia vs cytologic atypia: PTC vs all other diagnoses, P <.04; adenoma vs all other diagnoses, P <.04; benign follicular nodule (adenoma + MNH) vs all other diagnoses, P <.07, not significant. Architectural atypia vs unclassified atypia: PTC vs all other diagnoses, P =.002; adenoma vs all other diagnoses, P =.0009; benign follicular nodule (adenoma + MNH) vs all other diagnoses, P = Recent studies suggest that benign cytologic follow-up alone may not have significance equivalent to that of benign histologic outcome. 13,14 Therefore, we also evaluated clinical outcome data using surgical pathology diagnoses alone as the gold standard Table 2. Once again, architectural atypia alone had a malignant outcome (11/45 [24%]) roughly half that of all other qualifiers with a statistically significant difference observed from cytologic atypia alone (17/34 [50%]; P <.04), cytologic and architectural atypia (23/50 [46%]; P <.04), and unspecified atypia (45/70 [64%]; P =.0001). To further characterize outcome differences, we compared the histologic diagnoses among AUS qualifiers Table 3. The category of architectural atypia alone exhibited statistically significant differences from the cytologic atypia alone and unqualified atypia categories. The architectural atypia alone qualifier was less likely to be papillary carcinoma and more likely to be an adenoma. The 96 malignant diagnoses included 86 papillary carcinomas (90%), 8 follicular carcinomas (9%), and 1 case each of undifferentiated carcinoma and diffuse large B-cell lymphoma. The small number of follicular carcinomas identified belonged to the cytologic and architectural atypia and unqualified atypia categories and not to the architectural atypia alone group. The follicular variant comprised the majority of the papillary carcinomas (73/86 [85%]) with no statistically significant difference in subtype of papillary carcinoma among the AUS qualifiers (data not shown). To determine if any other specimen descriptors present in the report could predict malignancy, we evaluated outcome with respect to the severity of atypia present; the specimen cellularity; the qualitative number of atypical cells; and the presence of colloid, histiocytes, hemosiderin-laden macrophages, cyst lining cells, or Hürthle cell changes. No statistically significant difference in outcome was noted for any of these descriptors (data not shown). Discussion As defined by the Bethesda System, the atypia of undetermined significance category is heterogeneous with respect to inclusion criteria, intended to result in a Downloaded from Am J Clin Pathol 2011;136:

5 VanderLaan et al / AUS Qualifiers in Thyroid FNAs risk of malignancy intermediate between the benign and suspicious categories. This study represents an attempt to identify the diagnostic characteristics of cases within the AUS category that may predict a higher or lower risk of malignancy. Our results indicate that cases deemed atypical based solely on architectural atypia have nearly half the risk of malignancy as cases classified as cytologic atypia alone, as cases classified as cytologic and architectural atypia, and in cases in which the atypia is unclassified. Furthermore, on histologic follow-up, the AUS pattern of architectural atypia alone is more likely than the nonarchitectural atypia based patterns of AUS (cytologic atypia alone and unspecified atypia) to represent a follicular adenoma. Although no other diagnostic qualifiers examined were found to confer significantly different predictive risk, it is important to acknowledge that the present study was based on a retrospective review of primary pathology reports without performing a slide review. The subclassification of the atypia type and the inclusion of any other diagnostic features in the final report were at the discretion of the original cytopathologist, reflecting a real-world working experience with this diagnostic category. As such, the varying reporting practices of individual cytopathologists within our laboratory may have adversely affected the power of our study to detect the importance of the various qualifiers in AUS. Although others have reported some prognostic usefulness for associated findings in atypical thyroid FNA samples, 11 the lack of association seen in the present study may mean that these common findings (presence or absence of colloid, histiocytes, hemosiderin-laden macrophages, cyst lining cells, or Hürthle cell change) are not robustly associated with significant risk stratification or, alternatively, that any association that exists could not be readily identified within this study design. Despite these limitations, our findings are in keeping with limited previous reports suggesting that different subtypes of AUS specimens indeed vary with respect to the risk of malignancy. As shown herein, a thyroid nodule with a diagnosis of AUS with architectural atypia alone has a risk of malignancy of approximately 16%, which falls between that of the benign (2%-3%) and suspicious for a follicular neoplasm (20%-30%) categories. 1,2 Conversely, a diagnosis of AUS with a component of cytologic atypia (characterized as cytologic atypia alone or cytologic atypia accompanied by architectural atypia) has a malignancy risk closer to 30%. Therefore, one could conclude that there is a higher risk of malignancy when incomplete features of papillary carcinoma are present as opposed to features more likely to be associated with a follicular neoplasm. A recent study by Renshaw 9 reports a similar dichotomy for atypical thyroid aspirates with histologic follow-up. In cases diagnosed as atypical, rule out papillary carcinoma, a malignant outcome was found in 38% of cases, whereas in cases diagnosed as atypical, rule out follicular neoplasm, a malignant outcome was found in 22% of cases, approximately half as frequently. Another study has also shown that focal features of papillary carcinoma in thyroid aspirates categorized as atypical were strongly associated with a malignancy at subsequent surgical resection. 15 Thus, it seems clear that the identification of rare or incomplete nuclear/cytologic features of papillary carcinoma has a risk of malignancy higher than the identification of architectural atypia alone. In our study population, a large majority of the malignancies identified in association with AUS were papillary carcinoma (86/96 [90%]), and most of these were follicular variants (73/86 [85%]). These findings suggest that the lesser degrees of atypia identified in an AUS FNA sample may preferentially correlate with detection of the controversial 16 and commonly less aggressive follicular variant of papillary carcinoma when compared with a suspicious or positive diagnosis. It is not surprising that architectural atypia alone would be less predictive of identifying papillary carcinoma. Perhaps more surprising is that although architectural atypia alone was associated with the presence of a benign follicularpatterned nodule, none of the few follicular carcinomas in our population were associated with the pattern of architectural atypia alone. The lack of association between follicular carcinoma and AUS with architectural atypia may reflect a fundamental difference in this pattern relative to the suspicious for follicular neoplasm category, but could also be because of the low overall prevalence of follicular carcinoma in this study. The management of a given thyroid nodule following an AUS diagnosis is dictated by the anticipated risk of malignancy, and there seems to be mounting evidence from multiple laboratories that AUS has a risk of malignancy higher than originally predicted This finding may translate into a more aggressive management algorithm by clinicians as familiarity with this diagnostic category builds. If, however, most of the carcinomas detected with AUS are follicular variants of papillary carcinoma, as was the case in our study, an argument could be made that aggressive management may not be warranted for this typically indolent entity. The wide range of malignancy risk reported by authors from different laboratories in association with AUS speaks to the heterogeneity of lesions for which this diagnosis is used on FNA, but may also reflect variation in the diagnostic threshold for the follicular variant of papillary carcinoma in surgical pathology. With the appropriate management for AUS unresolved, identification of lesions with differing risks of malignancy may be valuable in guiding clinical decision making. Our findings support the continued use of AUS for cases in which isolated architectural atypia is present because these 576 Am J Clin Pathol 2011;136: Downloaded 576 from

6 Anatomic Pathology / Original Article aspirates have the low risk of malignancy envisioned for the AUS category when the Bethesda System was implemented. Conversely, evidence continues to accumulate indicating that AUS with focal features of papillary carcinoma or cytologic atypia has a malignancy risk that, although lower than for a diagnosis of suspicious for papillary carcinoma, is higher than had been anticipated for the AUS category in the initial implementation of the Bethesda System. Further study is warranted to determine how or if focal features of papillary carcinoma can be reproducibly and meaningfully separated from the suspicious for papillary carcinoma category. The AUS diagnostic category as it is currently defined is heterogeneous with respect to inclusion criteria and the estimated risk of malignancy for different lesions. Our study and the studies of others support the notion that architectural atypia alone has a much lower risk of malignancy than focal atypia related to traditional cytologic features of papillary carcinoma. Further experience is needed to establish the risk of malignancy associated with other less common patterns of AUS. Although the practice of using diagnostic qualifiers of AUS is not dictated by the Bethesda System, we favor this practice to facilitate further refinement of this diagnostic category with the aim of providing better risk stratification and clinical management for AUS aspirates. From the Departments of 1 Pathology and 2 Medicine, Brigham and Women s Hospital and Harvard Medical School, Boston, MA. Address reprint requests to Dr Krane: Cytopathology Division, Head and Neck Pathology Service, Dept of Pathology, Brigham and Women s Hospital, 75 Francis St, Boston, MA References 1. Baloch ZW, Cibas ES, Clark DP, et al. The National Cancer Institute Thyroid Fine Needle Aspiration State of the Science Conference: a summation. Cytojournal. 2008;5:6. 2. Cibas ES, Ali SZ. The Bethesda System for Reporting Thyroid Cytopathology. Am J Clin Pathol. 2009;132: Ali SZ, Cibas ES. The Bethesda System for Reporting Thyroid Cytopathology. New York, NY: Springer; Nayar R, Ivanovic M. The indeterminate thyroid fineneedle aspiration: experience from an academic center using terminology similar to that proposed in the 2007 National Cancer Institute Thyroid Fine Needle Aspiration State of the Science Conference. Cancer. 2009;117: Layfield LJ, Morton MJ, Cramer HM, et al. Implications of the proposed thyroid fine-needle aspiration category of follicular lesion of undetermined significance : a five-year multi-institutional analysis. Diagn Cytopathol. 2009;37: Theoharis CG, Schofield KM, Hammers L, et al. The Bethesda Thyroid Fine-Needle Aspiration Classification System: year 1 at an academic institution. Thyroid. 2009;19: Shi Y, Ding X, Klein M, et al. Thyroid fine-needle aspiration with atypia of undetermined significance: a necessary or optional category? Cancer. 2009;117: Faquin WC, Baloch ZW. Fine-needle aspiration of follicular patterned lesions of the thyroid: diagnosis, management, and follow-up according to National Cancer Institute (NCI) recommendations. Diagn Cytopathol. 2010;38: Renshaw AA. Should atypical follicular cells in thyroid fine-needle aspirates be subclassified [published correction appears in Cancer Cytopathol. 2010;118:303]? Cancer Cytopathol. 2010;118: Jo VY, Stelow EB, Dustin SM, et al. Malignancy risk for fineneedle aspiration of thyroid lesions according to the Bethesda System for Reporting Thyroid Cytopathology. Am J Clin Pathol. 2010;134: Somma J, Schlecht NF, Fink D, et al. Thyroid fine needle aspiration cytology: follicular lesions and the gray zone. Acta Cytol. 2010;54: Marchevsky AM, Walts AE, Bose S, et al. Evidencebased evaluation of the risks of malignancy predicted by thyroid fine-needle aspiration biopsies. Diagn Cytopathol. 2010;38: VanderLaan PA, Marqusee E, Krane JF. Clinical outcome for atypia of undetermined significance in thyroid fineneedle aspirations: should repeat FNA be the preferred initial approach? Am J Clin Pathol. 2011;135: Renshaw AA. Does a repeated benign aspirate change the risk of malignancy after an initial atypical thyroid fine-needle aspiration? Am J Clin Pathol. 2010;134: Weber D, Brainard J, Chen L. Atypical epithelial cells, cannot exclude papillary carcinoma, in fine needle aspiration of the thyroid. Acta Cytol. 2008;52: Elsheikh TM, Asa SL, Chan JK, et al. Interobserver and intraobserver variation among experts in the diagnosis of thyroid follicular lesions with borderline nuclear features of papillary carcinoma. Am J Clin Pathol. 2008;130: Downloaded from Am J Clin Pathol 2011;136: