BRAF mutation detection in indeterminate thyroid cytology specimens

Underlying cytologic, molecular, and pathologic characteristics of papillary thyroid carcinoma


  • We thank Marcia Kurs-Lasky for assistance with statistical analysis.



BRAF mutations are highly specific for papillary thyroid carcinoma (PTC) and many cytology specimens with BRAF mutations are expected to demonstrate cytologic features typical of PTC. However, indeterminate thyroid cytology cases are inevitable and understanding the significance of the BRAF mutation within the context of the Bethesda System for Reporting Thyroid Cytopathology would be valuable.


Thyroid cytology cases submitted for conventional cytomorphologic evaluation and BRAF mutational analyses were selected from the authors' cytopathology files from April 2007 to October 2011. From this group, the diagnostic usefulness of BRAF mutations in indeterminate and malignant cases was assessed and analyses of cytologic and histopathologic features associated with the mutations in this gene were performed.


A total of 131 cases with a BRAF mutation were identified. Of these, 119 underwent surgical pathology resection follow-up and demonstrated PTC. Approximately 75% of the cases were cytologically diagnosed as being positive for malignancy and these cases were associated with both the classic and tall cell variants of PTC at the time of resection, a greater likelihood of extrathyroidal extension, and the V600E type of BRAF mutation. In contrast, BRAF-mutated cases with diagnoses of atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS) and follicular neoplasm/suspicious for follicular neoplasm were found to be more strongly associated with the follicular variant of PTC, a K601E BRAF mutation, and a lower likelihood of extrathyroidal extension. However, a subset of AUS/FLUS cases with the V600E BRAF mutation appeared to represent sampling variability of the classic or tall cell variants of PTC.


Bethesda thyroid diagnoses in the setting of a BRAF mutation reflect differences in PTC subtypes, the nature of cytology specimens, and molecular characteristics. Cancer (Cancer Cytopathol) 2013;121:197–205. © 2012 American Cancer Society.


The well-differentiated nature of most thyroid cancers and their morphologic overlap with benign nodules results in the placement of 10% to 26% of thyroid cytology cases in 1 of the 3 indeterminate diagnostic categories of the Bethesda System for Reporting Thyroid Cytopathology (BSRTC).1-6 The indeterminate categories, atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS), follicular neoplasm/suspicious for follicular neoplasm (FN/SFN), and suspicious for malignancy (SMC), represent diagnoses with an increasing risk of cancer. With the rising incidence of thyroid cancer and improved detection of nodules, more thyroid nodules are evaluated by fine-needle aspiration (FNA) and there is a need for ancillary tests that may be used as an adjunct to thyroid cytology so that the additional information would help treat patients more effectively. To this end, several molecular markers have been studied for their applicability to the practice of thyroid FNA. One such marker, BRAF, is a serine-threonine kinase belonging to the RAF family of proteins.7 The BRAF mutation has been well studied and is well known for its association with the classic (CL) and tall cell (TCV) variants of papillary thyroid carcinoma (PTC), extrathyroidal extension (ETE), and a higher stage of disease at the time of presentation.8 Furthermore, the BRAF mutation is highly specific for PTC and is potentially a very useful ancillary marker for FNA practice.9 To the best of our knowledge, studies published to date have focused on the V600E mutation, which is the most common type of BRAF mutation. This mutation is associated frequently with CL and TCV PTC.9 Given the robust morphologic features of PTC demonstrated in CL and TCV PTC, one may question the diagnostic value of BRAF testing for FNA cytology specimens. Conversely, sampling variability is a common problem in FNA cytology and a sparsely sampled PTC, even if it is CL or TCV, may result in one of the indeterminate diagnoses. We hypothesized that a highly specific ancillary test such as BRAF mutational analysis could potentially add information to the indeterminate diagnoses and provide a better understanding of these FNA specimens. Specifically, in the current study, we examined the distribution of indeterminate thyroid FNA diagnoses among BRAF-mutated PTC cases, determined whether these indeterminate diagnoses were because of suboptimal sampling or differences in PTC characteristics, and investigated the significance of the BRAF mutation within the context of the BSRTC.


The current study was approved by the University of Pittsburgh Institutional Review Board. Prospective determination of BRAF mutation status on thyroid cytology specimens began at the study institution in 2007.10 Thyroid FNA cytology cases demonstrating a BRAF mutation were searched for in the pathology files of 6517 thyroid FNA cases (including all diagnostic categories) from April 2007 through October 2011. From these thyroid FNA cases, those with surgical pathology correlates demonstrating PTC were selected for further clinicopathologic investigation. The clinical parameters evaluated included the age and gender of each patient. The procedures for the collection, evaluation, and reporting were standardized and followed for thyroid FNA cases. Thyroid FNAs were performed under ultrasound guidance by a radiologist or endocrinologist using a 23-gauge, 25-gauge, or 27-gauge needle. Three to 4 passes were obtained in the majority of cases. For each thyroid nodule, a minimum of 4 direct smears (2 alcohol-fixed Papanicolaou-stained and 2 air-dried Diff-Quik–stained smears) and a monolayered ThinPrep (Hologic Inc, Marlborough, Mass) slide were made. Residual material from the passes was placed in 400 μL of nucleic acid preservative solution (Roche Molecular Biochemicals, Mannheim, Germany) for molecular analysis. The specimen vial with the nucleic acid preservative solution was stored frozen at −20°C before molecular analysis.

Cytologic interpretation and classification were based on BSRTC.11 At the study institution, transition to BSRTC took place in September 2008. Cytology diagnoses that were made before that date were reviewed and translated into the BSRTC by one of the authors (N.P.O.). Briefly, the 6 BSRTC categories were comprised of the following: unsatisfactory/nondiagnostic, benign, AUS/FLUS, FN/SFN, SMC, and positive for malignancy (PMC). As stated earlier, 3 of these diagnoses (AUS/FLUS, FN/SFN, and SMC) were considered to be indeterminate diagnoses. For the AUS/FLUS diagnoses, further evaluation was made to characterize the underlying cytologic features of the lesional cells. Each AUS/FLUS case was examined to determine whether it demonstrated architectural atypia, cytologic atypia, or both. This determination was made by one of the authors (N.P.O.) through review of the cytopathology reports and/or slide review of individual cases when the report was ambiguous. Molecular testing was performed on all cytology cases with a diagnoses of AUS/FLUS, FN/SFN, SMC, and PMC. BRAF mutational analysis was not performed on benign and unsatisfactory/nondiagnostic cases.

For those cases in which molecular testing was ordered, nucleic acid from the FNA material was isolated using the MagNA Pure Compact instrument (Roche Applied Science, Indianapolis, Ind) and Nucleic Acid Isolation Kit I (Roche Applied Science) according to the manufacturer's instructions. BRAF point mutations were detected using real-time LightCycler polymerase chain reaction (PCR) (Roche Applied Science) and fluorescence melting curve analysis (FMCA) as previously described.10, 12 Details of the molecular assays were described in the supplemental data of a previous reports.10 Cases that demonstrated BRAF point mutations by FMCA were confirmed by Sanger sequencing. The adequacy of molecular samples was determined by the quantity and quality of the isolated nucleic acids and the percentage of epithelial cells in the sample. The quantity and quality of the isolated nucleic acids were assessed by PCR amplification of the BRAF gene and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) cDNA, respectively. The percentage of epithelial cells was measured by comparing the expression of the cytokeratin gene KRT7 with that of the housekeeping gene GAPDH, which is expressed in all cells. The threshold for molecular adequacy was set at the difference in amplification between KRT7 and GAPDH of > 3.5 cycles. This difference corresponded to a minimum of 10% thyroid epithelial cells in the sample. Molecular testing and analyses were performed by trained technologists in the molecular anatomic pathology laboratory who were not aware of the cytology diagnoses. Our previous studies that reported on various clinical, cytologic, and molecular aspects of PTC have included some of the BRAF-mutated cases in the current cohort. Because the aims of these studies were different, the number of overlapping cases was relatively small and were comprised as follows: 17 cases from Nikiforov et al,10 31 cases from Yip et al,13 and 3 cases from Ohori et al.14

Surgical pathology thyroid resection cases corresponding to the cytology cases demonstrating a BRAF mutation were identified and reviewed. Cases that did not clearly state the subtype of papillary carcinoma were reexamined by 3 of the authors (R.S., N.P.O., and Y.N.) to provide a precise subtype. For each resection case, the nodule corresponding to the FNA target was identified and the size of the PTC, the presence or absence of ETE, and pathologic stage were recorded. The cytology diagnoses (AUS/FLUS, FN/SFN, SMC, and PMC) were correlated with the surgical pathology findings. Comparisons were made between the groups, with a particular focus on the contrast from the PMC group. Statistical analysis was performed using the Student t test, Fisher exact test, and chi-square test.


A total of 131 thyroid FNA cases with a BRAF mutation (tested on AUS/FLUS, FN/SFN, SMC, and PMC cases) were identified. Of these, 120 cases had surgical pathology resection follow-up available. A total of 119 cases represented PTC and 1 case represented a diffuse large B-cell lymphoma, which was excluded from the current study. There was no false-positive (nonmalignant) result from FNA cytology cases demonstrating a BRAF mutation. These 119 indeterminate (AUS/FLUS, FN/SFN, and SMC) and PMC cases became the core data set of the current study (Table 1). The presurgical cytology diagnoses for these cases were as follows: AUS/FLUS (11 cases), FN/SFN (4 cases), SMC (14 cases), and PMC (90 cases). Therefore, approximately one-quarter (29 of 119 cases; 24.4%) of the cytology specimens demonstrating BRAF mutations were given indeterminate diagnoses.

Table 1. BRAF-Mutated Thyroid Cytology Cases With Clinicopathologic and Molecular Correlation
FeaturesAUS/FLUSa n = 11FN/SFNb n = 4SMCc n = 14PMCd n = 90Statistical Comparisons (P)
  1. Abbreviations: AA, architectural atypia; A+C, architectural and cytologic atypia; AUS/FLUS, atypia of undetermined significance/follicular lesion of undetermined significance; CA, cytologic atypia; CL, classic; FN/SFN, follicular neoplasm/suspicious for follicular neoplasm; FV, follicular variant; ONC, oncocytic; PMC, positive for malignancy; PTC, papillary thyroid carcinoma; SD, solid; SMC, suspicious for malignancy; TCV, tall cell variant; WL, Warthin-like.

Mean age, y47.733.843.047.9a+b vs c: .84
a+b vs d: .31
c vs d: .21
GenderMan: 0Man: 1Man: 0Man: 24a+b vs c: >.99
Woman: 11Woman: 3Woman: 14Woman: 66a+b vs d: .11
c vs d: .04
Mean size, cm1. vs c: .44
a+b vs d: .12
c vs d: .86
PTC size ≤1.0 cm5/11 (45.5%)2/4 (50%)5/14 (35.7%)27/90 (30%)a+b vs c: .82
a+b vs d: .24
c vs d: .76
Key cytologic featuresAA = 3Cellular microfollicular patternMost but not all malignant criteria presentNuclear enlargement, elongation, grooves, and pseudoinclusions 
CA = 6
A+C = 2
PTC subtypeFV = 5 (45.4%)FV = 4 (100%)FV = 3 (21.4%)FV = 5 (5.6%)FV subtype
CL = 4 (36.4%)CL = 10 (71.4%)CL = 68 (75.5%)a+b vs c: .08
TCV = 1 (9.1%)WL = 1 (7.1%)TCV = 15 (16.7%)a+b vs d: <.001
SD = 1 (9.1%)ONC = 1 (1.1%)c vs d: .07
Other = 1 (1.1%) 
Extrathyroidal extension4/11 (36.4%)0/4 (0%)8/14 (57.1%)52/90 (57.8%)a+b vs c: .20
a+b vs d: .05
c vs d: >.99
Lymph node metastasis3/5 (60%)0/1 (0%)6/14 (42.9%)41/82 (50%)a+b vs c: >.99
a+b vs d: >.99
c vs d: .84
BRAF mutational statusV600E: 6V600E: 1V600E: 14V600E: 89K601E mutation
K601E: 5K601E: 3K601E: 0K601E: 0a+b vs c: .002
Other: 0Other: 0Other: 0Other: 1a+b vs d: <.001
c vs d: —

Because there were only 4 cases of FN/SFN, they were grouped together with the AUS/FLUS cases for statistical comparison with the SMC and PMC cases: AUS/FLUS + FN/SFN versus SMC and AUS/FLUS + FN/SFN versus PMC. In addition, statistical comparisons between the SMC and PMC cases were made. The mean age of the patients ranged from 33.8 years (FN/SFN) to 47.9 years (PMC), but there was no statistical difference noted between the groups. Comparisons of gender distribution between the AUS/FLUS + FN/SFN versus SMC cases and AUS/FLUS + FN/SFN versus PMC cases did not yield significant differences. However, we found significantly more females in SMC group compared with the PMC group (P = .04). There were no significant differences noted with regard to the size of the PTC between the groups. Cytologically, the 11 AUS/FLUS cases demonstrated architectural atypia in 3 cases, cytologic atypia in 6 cases, and both architectural and cytologic atypia in 2 cases (Figs. 1-3). For the 5 cases demonstrating follicular variant (FV) PTC at the time of resection, 2 showed architectural atypia, 2 had both architectural and cytologic atypia, and 1 demonstrated cytologic atypia alone in the cytology specimen. All 5 cases showing a CL or TCV PTC finding demonstrated cytologic atypia in the cytology specimen. The 1 case with a solid PTC finding demonstrated architectural atypia in the cytology specimen. FN/SFN cases showed the typical cellular microfollicular pattern. The SMC cases demonstrated most features of PTC but lacked 1 or 2 critical criteria, usually intranuclear pseudoinclusions. The PMC cases were robust in their demonstration of the cytologic features (nuclear enlargement, nuclear elongation, grooves, and pseudoinclusions) of PTC (Fig. 4). When comparing the subtypes of PTC in the resection specimens, we found that the AUS/FLUS + FN/SFN group had significantly more FV PTCs compared with the PMC group (P < .001) but not with the SMC group (P = .08). There was no difference noted between the SMC and PMC groups (P = .07). With regard to ETE, we found that the AUS/FLUS + FN/SFN group had fewer cases of ETE, which nearly reached statistical significance when compared with that of the PMC group (P = .05). There were no differences noted in the incidence of lymph node metastasis (based on lymph node dissection) between the groups. However, relatively fewer cases in the AUS/FLUS + FN/SFN group had undergone lymph node dissection (6 of 15 cases; 40%) when compared with the incidence of lymph node dissection in the SMC and PMC groups (96 of 104 cases; 92%). Therefore, the evaluation of surgically sampled lymph nodes may not have resulted in a fair comparison.

Figure 1.

A case of atypia of undetermined significance/follicular lesion of undetermined significance with architectural atypia demonstrated microfollicular structures in an otherwise hypocelluar specimen (Papanicolaou stain, × 600). Molecular testing revealed a K601E BRAF mutation and the determination was an encapsulated follicular variant papillary thyroid carcinoma.

Figure 2.

A case of atypia of undetermined significance/follicular lesion of undetermined significance with cytologic atypia demonstrated nuclear enlargement and grooves in occasional cells (Papanicolaou stain, × 600). Molecular testing revealed a V600E BRAF mutation and the determination was a classic papillary thyroid carcinoma with extrathyroidal extension.

Figure 3.

A case of atypia of undetermined significance/follicular lesion of undetermined significance with architectural and cytologic atypia demonstrated occasional microfollicular structures with nuclear enlargement and irregular nuclear membranes (Papanicolaou stain, × 600). Molecular testing revealed a K601E BRAF mutation and the determination was an encapsulated follicular variant papillary thyroid carcinoma.

Figure 4.

A positive for malignancy case demonstrated the typical features of papillary carcinoma with nuclear enlargement, irregular nuclear membranes, and multiple intranuclear pseudoinclusions (Papanicolaou stain, × 600). Molecular testing revealed a V600E BRAF mutation and the determination was a classic papillary thyroid carcinoma with extrathyroidal extension.

Finally, the vast majority of BRAF mutations were of the V600E type (110 of 119 cases; 92.4%), and a small number of variants (9 of 119 cases; 7.6%) were identified. Eight of the variants were of the K601E type and 1 was a complex BRAF mutation, p.T599_600insT. The case of the complex BRAF mutation, p.T599_600insT, was previously reported as a case report.15 The K601E variant BRAF mutation was found in 5 of 11 (45.5%) AUS/FLUS cases and 3 of 4 (75%) FN/SFN cases. For the K601E-mutated AUS/FLUS cases, a review of the surgical resection specimens revealed 4 FV PTC cases and 1 solid PTC. One of the V600E-mutated AUS/FLUS cases was found to be a FV PTC at the time of resection. All 3 K601E-mutated FN/SFN cases were found to be FV PTCs at resection. The 1 case of a V600E-mutated FN/SFN was found to be a FV PTC at the time of resection. The V600E-mutated PMC cases were as follows: CL, 77 cases; TCV, 15 cases; Warthin-like, 1 case; oncocytic, 1 case; and other (infarcted), 1 case. In contrast, none of the 104 SMC and PMC cases was found to demonstrate the K601E variant (P < .001). The 1 case of the p.T599_600insT variant mutation was found in a case cytologically diagnosed as PMC and later shown to represent a CL PTC in the resected specimen.


Over the last decade, significant advances have been made in the understanding of the molecular pathogenesis of thyroid neoplasia.8 Common thyroid neoplasms involve mutations in genes associated with the mitogen-activated protein kinase (MAPK) pathway (eg, RAS, RAF) and these mutations are mutually exclusive. Many ancillary tests used in conjunction with thyroid specimens have focused on markers in this pathway. Given the potential applicability to FNA cytology, molecular testing has been anticipated to impact the practice of thyroid FNA cytopathology significantly. However, any new technology must be tested against the standard procedures. In parallel to the development of molecular technologies, we have witnessed changes in the practice of thyroid FNA cytology. The pivotal event marking this recent change was the National Cancer Institute Thyroid Fine-Needle Aspiration State of the Science Conference (held in October 2007 in Bethesda, Maryland) of thyroid pathologists, cytopathologists, radiologists, endocrinologists, and surgeons to standardize the various aspects of thyroid cytology. After the meeting, the proceedings of the BSRTC were published in June 2008.16 Many institutions in the United States and other countries have adopted or are in the process of adopting this 6-tiered classification system. Overall, the BSRTC has made significant contributions in bringing pathologists, radiologists, and clinicians closer together in understanding and communicating thyroid cytology diagnoses. The main differences between BSRTC and other previous systems have been in the number and definition of categories stratifying the indeterminate diagnoses. The BSRTC has 3 indeterminate diagnoses (AUS/FLUS, FN/SFN, and SMC) with increasing risk of malignant outcome. These 3 categories comprise 10% to 26% of thyroid FNA diagnoses.1-6 Ali and Cibas estimated the risk of malignancy as follows: 5% to 15% for AUS/FLUS, 15% to 30% for FN/SFN, and 60% to 75% for SMC.11 Considering the outcome estimates of these indeterminate diagnoses, a reliable ancillary test that improves the assessment of risk would be potentially helpful in guiding clinical management.

At the time of the initial Bethesda publications in 2008, the recommendation for the widespread use of molecular testing was believed to be provisional because of the limited number of validation studies that had been performed up to that point.17 The gradual acceptance of molecular testing has been reflected in subsequent statements made in a variety of society publications. In 2009, the American Thyroid Association stated “Recent large prospective studies have confirmed the ability of genetic markers (BRAF, RAS, RET/PTC) and protein markers (Galectin-3) to improve preoperative diagnostic accuracy for patients with indeterminate thyroid nodules.”18 To the best of our knowledge, of the markers in the MAPK pathway, the BRAF mutation has been studied most extensively. Greater than 40 BRAF mutations have been detected in human cancer, and of these, V600E is the most common type in PTC.19 It is found in approximately 45% of PTCs, although there are demographic differences, with regions in East Asia (in particular Korea) reporting a higher incidence of approximately 75%.20-22 In the thyroid gland, the BRAF mutation is recognized as a very specific marker for PTC and the overall value of this marker in cytopathology practice currently is being elucidated.

A large number of studies have addressed the application of BRAF mutation testing to thyroid cytology specimens. However, only a subset of these studies correlated the BRAF mutation results with indeterminate diagnoses that were stratified into ≥ 2 subcategories and only 6 studies used the BSRTC or a comparable system (Table 2).12, 23-34 In addition, the vast majority of the studies concerning thyroid cytology specimens specifically examined the V600E mutation only. To the best of our knowledge, the study by Proietti et al is the only one that identified the K601E variant in a series of thyroid FNA cases that categorized the diagnoses according to the BSRTC.27 Overall, these studies demonstrated that the majority of cases with a BRAF mutation were cytologically diagnosed as PMC. However, a review of the indeterminate diagnoses revealed a great variability in the distribution of the BRAF-mutated cases. In several studies, only rare BRAF-mutated cases (< 3%) were placed in the combined group of AUS/FLUS and FN/SFN diagnoses.23-25, 29, 30, 32, 34 Conversely, other studies demonstrated a greater percentage (> 10%) of BRAF-mutated cases being diagnosed in the combined group of AUS/FLUS and FN/SFN diagnoses.11, 26-28, 31, 33 The most likely explanation for this difference is the variability in the application of the cytologic criteria of the BSRTC or equivalent systems (eg, British Thyroid Association diagnostic categories Thy 1-Thy 5). Specifically, the demarcation between AUS/FLUS and SMC and between FN/SFN and SMC may vary from individual to individual and from institution to institution. For example, the threshold for the degree of atypia and the number of cells demonstrating atypia required may not be clear when distinguishing between AUS/FLUS and SMC.35 Likewise, cases with a microfollicular pattern and cytologic atypia may be interpreted as FN/SFN or SMC depending on the interpretation of the cytopathologist. In addition to the interpretative aspects, sampling methodologies and specimen processing may contribute to variabilities in diagnostic reporting. It is interesting to note that this type of variability was observed in geographic areas with higher incidences of BRAF-mutated PTCs (eg, Korea) as well as in areas with lower incidences (eg, the United States). Given these differences, the authors of studies from the United States and abroad have reached different conclusions regarding the contribution of BRAF testing in thyroid cytology specimens. Some may argue that BRAF mutational testing is not helpful in the majority of indeterminate cases because most BRAF-mutated cases are cytologically diagnosed as PMC in a subset of studies.23-25, 29, 30, 32, 34 These studies primarily focused on the BRAF V600E mutation and found very few cases (< 3%) in the AUS/FLUS and FN/SFN categories but substantially more cases (12.5% to 50.8%) were found in the SMC category. In this regard, the value of BRAF mutational testing should be viewed within the context of the risk of malignancy of the SMC diagnosis. Individual cytology laboratories vary with regard to their threshold for placing a case in the SMC category. If the risk at a particular laboratory is very high (eg, > 90%), there is most likely little value in BRAF mutational testing because many of these patients may be treated similarly to those with a malignant diagnosis. However, if the risk is at the lower end of the range for an SMC diagnosis (ie, approximately 55%), the results of BRAF mutational testing may or may not impact patient management, depending on the institution's management strategy (eg, lobectomy for a BRAF-negative result or total thyroidectomy in the case of a BRAF-positive finding). It is interesting to note that we found significantly more women in SMC group compared with the PMC group (P = .04), but could not find an explanation for this difference.

Table 2. Distribution of BRAF-Positive Cases Among Indeterminate and Malignant Diagnosesa
  • Abbreviations: AUS/FLUS, atypia of undetermined significance/follicular lesion of undetermined significance; BSRTC, Bethesda System for Reporting Thyroid Cytopathology; FN/SFN, follicular neoplasm/suspicious for follicular neoplasm; PMC, positive for malignancy; SMC, suspicious for malignancy.

  • Bold type indicates studies that used the BSRTC or equivalent.

  • a

    aStudies that combined all indeterminate cases into 1 category were not included.

Marchetti 200923No0/59 (0%)18/59 (30.5%)41/59 (69.5%)V600E
Nikiforov 200912Yes1/19 (5.2%)4/19 (21.1%)4/19 (21.1%)10/19 (52.6%)V600E
Zatelli 200924Yes0/42 (0%)1/42 (2.4%)10/42 (23.8%)31/42 (73.8%)V600E
Kim 201025No0/212 (0%)50/212 (23.6%)162/212 (76.4%)V600E
Moses 201026No3/23 (13.0%)10/23 (43.5%)10/23 (43.5%)V600E
Proietti 201027Yes4/26 (15.4%)8/26 (30.8%)8/26 (30.8%)6/26 (23.0%)V600E/K601E
Adeniran 201128Yes12/64 (18.8%)0/64 (0%)4/64 (6.2%)48/64 (75.0%)V600E
Colanta 201129No0/24 (0%)3/24(12.5%)21/24 (87.5%)V600E
Kim 201131Yes45/221 (20.4%)1/221 (0.4%)46/221 (20.8%)129/221 (58.4%)V600E
Pelizzo 201132No2/95 (2.1%)28/95 (29.5%)65/95 (68.4%)V600E
Chang 201233No16/92 (17.4%)19/92 (20.6%)57/92 (62.0%)V600E
Kang 201230Yes0/276 (0%)0/276 (0%)76/276 (27.5%)200/276 (72.5%)V600E
Marchetti 201234No0/63 (0%)32/63 (50.8%)31/63 (49.2%)V600E
Current studyYes11/119 (9.2%)4/119 (3.4%)14/119 (11.8%)90/119 (75.6%)V600E/K601E/other

In contrast to the experience of those who found that majority of BRAF-mutated cases occurred in the PMC and SMC categories, we and others found a greater percentage of BRAF-mutated (V600E, K601E, and others) cases in the AUS/FLUS and FN/SFN categories; therefore, BRAF mutational testing was found to be a useful predictor of PTC diagnosis in these indeterminate cases.11, 26-28, 31, 33 Furthermore, we observed that BRAF mutational status interpreted within the context of the BSRTC diagnosis provided greater insight into the neoplasm and the characteristics of the FNA sample. BRAF-positive mutational status in cases of AUS/FLUS and FN/SFN indicated a high likelihood of a diagnosis of FV PTC, a lower likelihood of ETE by PTC, and the presence of the K601E mutation when compared with those cases with a PMC diagnosis. In particular, all 4 of the positive FN/SFN cases in the current study resulted in a determination of FV PTC and 3 of these 4 cases were found to have the K601E variant mutation. This information gives insight into the lesional characteristics and may or may not impact patient management. For institutions following the standard BSRTC guidelines, the typical clinical practice is to surgically evaluate patients with a diagnosis of FN/SFN using a diagnostic lobectomy. If malignancy is found, a completion lobectomy is performed. In such a management paradigm, the molecular test result is not likely to impact patient management. However, at an institution such as ours, which has started to include the molecular information into the clinical management scheme, a positive BRAF result does influence the clinical decision-making and those patients with a BRAF-positive FN/SFN diagnosis may undergo a total thyroidectomy because malignancy is found in the majority of these cases.10 The cost-effectiveness of this approach has been demonstrated in a previous study by Yip et al.36 Approximately one-half (5 of 11 cases) of the AUS/FLUS cases resulted in a FV PTC outcome and 4 of these 5 cases demonstrated the K601E variant mutation (Table 3). For these FN/SFN and AUS/FLUS cases with the K601E mutation variant, the cytomorphology of the PTC (not sampling) was most likely the limiting factor precluding a more definitive diagnosis. Conversely, 5 of the remaining 6 AUS/FLUS cases that resulted in a CL, TCV, or solid diagnosis demonstrated the V600E mutation, and suggested that these 6 cases of AUS/FLUS most likely represented insufficient sampling of the diagnostic cytologic features of PTC. Although we found no difference in the incidence of lymph node metastasis, we noted that relatively fewer cases of AUS/FLUS and FN/SFN had had lymph node dissection performed (6 of 15 cases; 40%) when compared with the SMC and PMC cases (96 of 104 cases; 92%).

Table 3. Specific Details Regarding the Outcome of AUS/FLUS Cases
Cytologic FeaturesSurgical Pathology Resection–PTC SubtypeBRAF Mutation Type
  1. Abbreviations: AA, architectural atypia; A+C, architectural and cytologic atypia; AUS/FLUS, atypia of undetermined significance/follicular lesion of undetermined significance; CA, cytologic atypia; CL, classic; FV, follicular variant; PTC, papillary thyroid carcinoma; SD, solid; TCV, tall cell variant.


In summary, BRAF analyses in conjunction with the BSRTC diagnoses provide greater insight into the characteristics of PTCs. The majority of cases with a BRAF mutation demonstrate overt cytologic features of PTC and are diagnosed as PMC. At the time of resection, they usually result in a finding of CL or TCV PTC. Conversely, approximately one-quarter of cases with a BRAF mutation are given one of the indeterminate diagnoses. The clinicopathologic significance of the BRAF mutation appears to be related to the type of indeterminate diagnostic category. BRAF-mutated cases with the cytology diagnosis of SMC result in a CL PTC determination in the majority of resections. For AUS/FLUS and FN/SFN cases, a FV PTC outcome and K601E mutation are much more common and ETE is noted less frequently. However, AUS/FLUS cases with a V600E BRAF mutation most likely represent CL or TCV PTC and for these cases, better sampling may have yielded a SMC or PMC diagnosis. BSRTC diagnostic categories of thyroid FNA specimens showing BRAF mutations reflect differences in the underlying pathologic characteristics of PTC.


No specific funding was disclosed.


The authors made no disclosures.