Original Article
Morphological Parameters Able to Predict BRAFV600EMutated Malignancies on Thyroid Fine-Needle Aspiration Cytology: Our Institutional Experience Esther Diana Rossi, MD, PhD1; Tommaso Bizzarro, BD1; Maurizio Martini, MD, PhD; Sara Capodimonti, BD, PhD; Guido Fadda, MD1; Luigi Maria Larocca, MD1; and Fernando Schmitt, MD, PhD2,3,4
BACKGROUND: The BRAFV600E mutation is the most common diagnostic/prognostic marker in papillary thyroid carcinoma (PTC). Its evaluation is typically performed with DNA-based techniques; nonetheless, a few articles have recently proposed the morphological prediction of BRAFV600E in histological PTCs. We investigated this morphological parameter in our cytological series. METHODS: We re-analyzed all 72 cytohistological samples diagnosed as positive for malignancy (favoring PTC) on fine-needle aspiration cytology from January 2012 to December 2013. We included 22 male patients and 50 female patients. The cytological cases were processed with liquid-based cytology. We performed molecular analysis and immunocytochemistry for the VE1 BRAFV600E antibody. RESULTS: We reported 47 mutated cases and 25 wildtype (WT) cases with 100% cytohistological concordance. The cytological evaluations revealed plump cells (abundant eosinophilic cytoplasm and PTC nuclei) in all 47 mutated cases, with only 6 having a focal plump cell component (20% cells). Furthermore, 5 of the 25 WT cases showed focal plump cells. A distinctive sickle nuclear shape was found only in the mutated cases. VE1 yielded 100% positivity for all 24 mutated cases that were tested, including 3 cases with focal plump cells. CONCLUSIONS: We demonstrated that the BRAFV600E mutation might be predicted in cytological samples on the basis of some specific morphological features. Although the detection of plump cells (mainly focal) was also observed in WT cases, the detection of sickle-shaped nuclei provided the highest specificity and sensitivity as a predictive mutational parameter. These morphological features might be a valid tool for selecting cases for molecular analysis. C 2014 American Cancer Society. Cancer (Cancer Cytopathol) 2014;122:883-91. V
KEY WORDS: BRAF mutation; liquid-based cytology; thyroid carcinoma; thyroid lesions; plump cells.
INTRODUCTION Activating somatic mutations in the BRAF oncogene are common findings in thyroid carcinoma, as demonstrated by the 45% to 70% prevalence (mainly in the classic variant of papillary thyroid carcinoma ([PTC]).1-4 More than 95% of BRAF mutations involve the hotspot identified on exon 15 of the B isoform of the RAF kinase gene with a consequent valine–to–glutamic acid substitution.1-8 Furthermore, this mutational assessment has provided significant insights into a prognostically unfavorable outcome.2 In fact, it is frequently related to more aggressive behavior involving multifocality, extrathyroid infiltration, and lymph nodes and/or distant metastases.3,9-14 Corresponding author: Esther Diana Rossi, MD, PhD, Division of Anatomic Pathology and Histology of the Catholic University of the Sacred Heart, Agostino Gemelli School of Medicine, Largo Francesco Vito 1, 00168 Rome, Italy; Fax: (011) 3906 3015 7008; [email protected] 1
Division of Anatomic Pathology and Histology of the Catholic University of the Sacred Heart, Agostino Gemelli School of Medicine, Rome, Italy; Institute of Molecular Pathology and Immunology, University of Porto, Porto, Portugal; 3Department of Laboratory Medicine and Pathobiology, Faculty of Medicine, University of Toronto, Toronto, Canada; 4Department of Pathology, University Health Network, Toronto, Canada. 2
The last 2 authors share senior authorship. Received: June 18, 2014; Revised: July 25, 2014; Accepted: July 30, 2014 Published online August 25, 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cncy.21475, wileyonlinelibrary.com
Cancer Cytopathology
December 2014
883
Original Article
The diagnostic role of the BRAFV600E mutation in PTC has been extensively considered in an ample literature and has been mainly linked to specific histotypes, including the classic variant of PTC and the tall cell variant (TCV).1-4 According to the literature, the gold standard for BRAF analysis is an assessment with DNA-based methods; nonetheless, this may not represent the most cost-effective option. Despite their invaluable advantages, the difficulties encountered with the worldwide dissemination of these techniques have enhanced the enthusiasm for the recent introduction of the monoclonal V600E antibody (clone VE1) as well as the possible morphological prediction of specific features in thyroid BRAFV600Emutated cases.15-17 Few articles have underlined the distinctive morphological features of histological samples of BRAFV600Emutated PTC,15,16 including some architectural parameters (tumor-associated stromal reactions and infiltrative tumor borders) and cell-specific parameters (nuclear features and polygonal eosinophilic cells [defined as plump cells]). This alternative morphological perspective has gained popularity as a possibly foreseeable predictive aid.15,16 Hall et al.17 hypothesized that these morphological changes are the result of the involvement of the Warburg effect in enabling cancer cells to convert glucose into lactic acid. This effect causes the upregulation of oncogenes involved in RAS-BRAF-MEK-ERK pathways, which induce glycolysis even in the presence of abundant glucose.17 Consequently, an awareness of these histological findings raised our curiosity and led us to evaluate our cytological series of thyroid lesions diagnosed as positive for malignancy (favoring PTC). Because of the lack of infiltrative growth evidence on fine-needle aspiration cytology (FNAC), we focused on some specific cellular details of BRAFV600E-mutated follicular cells, mainly in the research of plump cells, which were defined as large polygonal tumor cells with a cell height less than twice the width that had features of squamoid metaplasia with homogeneous, eosinophilic, and moderate to abundant cytoplasm and shared nuclear features of PTC.15,16 We retrospectively investigated and quantified this morphological feature in 72 cytologically positive-formalignancy thyroid lesions with surgical pathology outcomes. Moreover, we discuss the immunocytochemistry with the VE1 antibody in 30 of the 72 cases. 884
MATERIALS AND METHODS Our series included all 72 retrospective positive-formalignancy (favoring PTC) samples from January 2012 to December 2013. All the cases were recorded at the Division of Anatomic Pathology and Histology of the Catholic University of the Sacred Heart (Agostino Gemelli Hospital of Rome). All FNAC was performed under sonographic guidance (ultrasound), mostly by surgeons and endocrinologists, and was processed with the ThinPrep 5000 liquid-based cytology (LBC) method (Hologic Co., Marlborough, Mass). Our patients were studied with ultrasound during a thyroid checkup performed at our hospital’s center for thyroid diseases. The series included 22 male patients and 50 female patients with a median age of 27 years (range 5 19-73 years). All aspirations (usually 2 passes for each lesion) were performed with 25- to 27-gauge needles; no rapid onsite assessment of the adequacy of the material was performed. All patients were appropriately informed about the use of LBC for processing their samples, and a written informed consent form was signed. Our study followed the tenets of the Declaration of Helsinki, and we received internal ethics approval for the study. The technical steps for liquid-based preparations are clearly described in previous articles from our group.9,10 Each resulting slide was fixed in 95% ethanol and stained with the Papanicolaou technique, whereas the remaining material was stored in PreservCyt solution (Hologic) to be possibly used for the preparation of additional slides for further investigations (including both immunocytochemistry and molecular analysis). Our cytological cases were classified as positive for malignancy (favoring PTC) according to the morphological criteria adopted by the Societa Italiana di Anatomia Patologica e Citologia Diagnostica–IAP Italian Working Group classification, which is similar to the Bethesda thyroid system.18,19 The lower limit of adequacy for each sample was established according to the British Royal College of Pathologists classification with 6 groups of thyroid follicular epithelial cells within the submitted slides, each having at least 10 well-visualized epithelial cells.20 Additional slides for immunocytochemistry and molecular analysis were obtained from the material stored in PreservCyt solution, so that analyses could be performed with even 2 mL of the material remaining eluted in 5 mL of PreservCyt solution. Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
The percentage of disease-specific cells for molecular analysis was at least 50% in all LBC samples, whereas the percentage was 30% for the immunocytochemical VE1 evaluation. Molecular Analysis
DNA was extracted from LBC samples stored in PreservCyt solution and from paraffin-embedded tissues. Each LBC sample was centrifuged, the supernatant was discarded, and the cellular pellet was processed. The pellet was incubated at 56 C for 3 hours in 180 lL of an ATL lysis buffer and 20 lL of proteinase K (20 mg/mL) from the QIAamp DNA mini kit (Qiagen, Hilden, Germany). For histological samples, 10 lm of slide tissue was deparaffinized and, after an ethanol treatment, was incubated at 56 C overnight in 180 lL of an ATL lysis buffer and 20 lL of proteinase K (20 mg/mL) from the QIAamp DNA mini kit. DNA was extracted according to the manufacturer’s protocol, and we assessed the quantity and quality of the DNA spectrophotometrically (absorbance at 260 nm and ratio of the absorbance at 260 nm to the absorbance at 280 nm; spectrum 5 220-320 nm; Biochrom, Cambridge, United Kingdom) and by separation on an Agilent 2100 bioanalyzer (Agilent, Palo Alto, Calif). Low-purity or insufficient DNA samples were extracted a second time. After the first amplification on a Rotor-Gene Q (Qiagen), the mutational analysis of BRAF was achieved with an anti-EGFR MoAb response (BRAF status) kit (reference UP033, Diatech Pharmacogenetics) by pyrosequencing via the PyroMark Q96 ID system (Biotage AB, Uppsala, Sweden). The sensitivity of this method was 5% in our laboratory. The following sequence was analyzed: ACAGT/AGAAA (exon 15).21 Immunocytochemistry and Immunohistochemistry
Because we retrospectively analyzed all 72 cytohistological series, the 30 cases with an analysis of the VE1 antibody on LBC were enrolled from our previous study.22 Immunohistochemistry for the VE1 antibody was run on LBC specimens according to the standard protocol reported in our previous article22 and was evaluated by 2 pathologists (E.D.R. and G.F.) blinded to the BRAF molecular status. The interpretation and validation of this VE1 application with cytology and LBC were supported by our previous experience.22 Briefly, the cytoplasm positivity of the samples was graded from 0 to 31 and was determined on both cytoCancer Cytopathology
December 2014
logical and histological samples. The intensity was defined as follows: 0, negative; 11, positivity in less than 30% of cells; 21, positivity in 30% or more than of cells and less than 80% of cells; and 31, positivity in more than 80% of cells. For our convenience, we report this arbitrary distribution with 4 categories of staining intensity (0, 11, 21, and 31), which are also based on the percentage of stained cells. The results for each category were homogeneous, regardless of the number of tumor cells stained. We did not find any discrepancy in the intensity or percentage of VE1-positive cells between cytological and histological samples. Histology
All surgical specimens were fixed in 10% buffered formaldehyde and embedded in paraffin, and the 5-lm-thick microtomic sections were stained with hematoxylin-eosin. The BRAF mutational analysis was also performed on DNA extracted from surgical specimens that were at least 70% tumor. All the perithyroid adipose tissue was embedded and examined for lymph node research. The diagnosis of PTC was based on the presence of true papillary structures and distinctive nuclear features, whereas the diagnosis of the follicular variant of papillary thyroid carcinoma (FVPC) relied on the detection of the nuclear features of PTC in multiple foci within the tumor.21 The diagnosis of TCV was characterized by a predominance of neoplastic cells whose heights were at least 3 times their widths and that had classic PTC nuclear features. The oxyphilic variant of PTC was defined by distinctive neoplastic cells showing marked cellular atypia with an oxyphilic basal zone, midplaced nuclei, and a slightly clearer apical region. All the cases were classified according to the seventh edition of the tumor-node-metastasis–based staging system recommended by the American Joint Commission on Cancer.23 Statistical Analysis
Statistical analysis was performed with a commercially available statistical software package (SPSS 10.0 for Windows, Chicago, Ill). Comparisons of categorical variables were performed with the chi-square statistic; Fisher’s exact test was used when that was appropriate. A P value less than .05 was considered significant.
RESULTS We re-analyzed all 72 cytological samples diagnosed as positive for malignancy between January 2012 and 885
Original Article
December 2013 with surgical follow-up. Table 1 shows the distribution of the clinical features in 47 BRAFV600Emutated cases (65.3%) and 25 wild-type (WT) BRAFV600E cases (34.7%; Fig. 1A). Regardless of the BRAFV600E mutation, 57 of the 72 nodules were equal to or smaller than 2 cm. There were no significant differences in sex or tumor size with respect to the 2-cm cutoff. All cases were histologically confirmed, and they were diagnosed as a classic variant of PTC (n 5 52) or FVPC (n 5 20). We excluded all cases showing TCV and oxyTABLE 1. Clinical Features of 72 Cytohistological Series
Clinical Feature Sex, n Male Female Age (years), range (median) Size, n 2 cm >2 cm Histology, n PTC FVPC T stage, n TI-TII TIII-TIV N stage, n pN0 pN1
BRAFV600E Mutation (47 Cases)
Wild-Type BRAF (25 Cases)
12 35 19-70 (25.5)
10 15 20-73 (26.5)
40 7
17 8
37 10
15 10
33 (25 TI 1 8 TII) 14 (12 TIII 1 2 TIV)
19 (13 TI 16 TII) 6 (5 TIII 11 TIV)
22 25
15 10
P Value NS (.28)
NS NS (.13)
NS (.11)
NS (.78)
NS (.33)
Abbreviations: FVPC, follicular variant of papillary thyroid carcinoma; N stage, lymph node stage; NS, not significant; PTC, papillary thyroid carcinoma; T stage, tumor stage.
philic PTC on histology. Moreover, the T stage and the N stage did not reach statistical significance in the 2 groups. As for the T stage, we had a higher number of T3 patients in the mutated group versus the WT group (25.5% vs 20%). For the statistical evaluation, we subdivided the cases into TI-TII and TIII-TIV groups. Lymph node metastases were found in 25 of the 47 BRAFV600Emutated cases (54%) and in 10 of the 25 WT cases (40%). As underlined in Table 1, there were no statistical significant differences for any of the parameters. In accordance with Virk et al15 and Finkelstein et al,16 we have underlined the presence of some cellular features predictive of the BRAFV600E mutation in Table 2. Regardless of the BRAFV600E mutation, all our samples showed typical nuclear features of PTC (grooves, pleomorphic membranes, and nuclear pseudo-inclusions). In addition, we found plump cells showing a polygonal shape and characterized by abundant eosinophilic cytoplasm. Noticeably, in our LBC preparations, the specific color of these plump cells seemed to be more intense than the pink eosinophilic shade on histology (Figs. 1B, 2, and 3). Moreover, for our convenience, we qualified these plump cell component as (1) diffuse (up to >20%) or (2) focal (20%). As documented in Table 2, these plump cells were significantly associated with the BRAFV600E mutation (P 5 .0002, odds ratio 5 70.23, 95% confidence interval 5 3.46-1427), with only 6 of 47 mutated cases having a focal plump cell component. Moreover, we found that 5 of the 25 WT cases had focal plump cells (Table 2). Our data revealed a significant correlation
Figure 1. (A) Details of the morphological features of wild-type cells obtained from a positive-for-malignancy papillary thyroid carcinoma case on liquid-based cytology (340). (B) Details of the morphological features of plump cells obtained from a positive-for-malignancy papillary thyroid carcinoma case on liquid-based cytology (340).
886
Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
TABLE 2. Evaluation of the Cytological Parameters of Positive-for-Malignancy Cases With and Without the BRAFV600E Mutation Cytological Feature
Mutated BRAF (47 Cases)
Wild-Type BRAF (25 Cases)
6 41 47 47 10 15
5b 0 25 0 12 6
Plump cellsa 20% >20% Classic PTC nuclear featuresc Sickle-shaped nuclei Psammoma bodies Cystic changes
Sensitivity
Specificity
PPV
NPV
Diagnostic Accuracy
90.4%
100%
100%
80%
93%
65.3% 100% 45.5% 71.4%
0% 100% 26% 37.3%
100% 100% 21.3% 31.9%
0% 100% 52% 76%
65.3% 100% 31.9% 47.2%
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; PTC, papillary thyroid carcinoma. a P 5.0002 (odds ratio 5 70.23, 95% confidence interval 5 3.46-1427). b The remaining 20 cases did not show any evidence of plump cell components. c These nuclear features included nuclear pseudo-inclusions, irregular membranes, and nuclear grooves.
Figure 2. Details of the morphological features of plump cells organized into small groups with the height of the cells less than twice the width (liquid-based cytology, 340).
between BRAF mutation analysis and a peculiar nuclear shape as an additional cellular parameter. This feature was present in 100% of our BRAFV600E-mutated cases and was absent in WT cases (Table 2). We defined this feature as a sickle-shaped nucleus (Fig. 3). These sickle-shaped nuclei tended to be smaller and eccentrically located within pleomorphic and irregular membranes, so the detection of nuclear pseudo-inclusions might be more difficult. On balance, the WT cases were characterized by PTC nuclei with a bigger nuclear size and less abundant cytoplasm in comparison with mutated cases. We did find 10 and 12 psammoma bodies in the mutated and WT groups, respectively, as detailed in Table 2. Some cases showed cystic changes (15 mutated cases vs 6 WT cases) without any statistical significance. We found a 100% cytohistological correlation for these parameters (Fig. 4). Cancer Cytopathology
December 2014
Figure 3. Details of the morphological features of sickleshaped malignant nuclei obtained from a positive-formalignancy papillary thyroid carcinoma case on liquid-based cytology (340).
In fact, the same peculiar parameters were studied and evaluated for the corresponding histological sections, and this resulted in 100% cytohistological concordance for the quantitative evaluation of plump cells and sickleshaped nuclei (Fig. 4A,B). Also in Table 2, a specific evaluation of the sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy is presented for each analyzed feature. Table 3 reports the results for our 30 cases with VE1 expression, including 24 BRAFV600E-mutated cases and 6 WT cases. The results for each category of immunocytochemistry were homogeneous, regardless of the number of stained cells in each of the categories. As for the distribution of VE1, our mutated series had 13 cases with 21 positivity and 11 cases with 31 positivity (Fig. 5), whereas 887
Original Article
Figure 4. (A) Specific histological details of polygonal cells with homogeneous eosinophilic cytoplasm (plump cells) at the bottom of the picture obtained from a follicular variant of papillary thyroid carcinoma (hematoxylin-eosin, 360). (B) Specific histological details of plump cells with details on the sickle-shaped nuclei (hematoxylin-eosin, 360).
TABLE 3. Evaluation of 30 of 72 Cases With VE1 Expression22
Cytological Feature Mutated BRAFV600E Wild-type BRAF
Plump Cells
VE1-Negative (0 and 11 Intensity)
VE1-Positive (21 and 31 Intensity)
20% >20% Absent 20% >20%
0 0 5b 1c 0
3a 21 0 0 0
a
All 3 focal plump cell cases were 21 VE1-positive. These 5 cases included all 5 wild-type cases without plump cells. c Only 1 case belonged to the group with 20% plump cells (expressed as 11 VE1). b
our 6 negative cases were distributed as 4 cases with 0 positivity (these 4 cases did not present with any plump cell components) and 2 cases with 11 positivity (one belonged to the group without plump cells, and the second belonged to the group with 20% plump cells). Four cases with focal plump cells belonged to the VE 1 series (3 were mutated cases with 21 VE1, whereas 1 was a WT case with 11 VE1).
DISCUSSION The aims of this study were to evaluate the specific morphological features of the BRAFV600E mutation on cytological thyroid malignant lesions (namely PTC and its variants) and to provide a possible means of predictive molecular detection. In the field of thyroid oncology, the BRAFV600E mutation is synonymous with thyroid carcinoma, mainly 888
Figure 5. Immunocytochemistry cytoplasm staining results for a mutated case with plump cells on liquid-based cytology: 31 VE1 positivity in tumor cells (avidin-biotin complex, 3400).
in the classic variant of PTC.1-6 The high specificity of this somatic mutation underlines its early driving role in thyroid tumorigenesis and its association with a more aggressive phenotype. Several series have documented mutational frequencies of 40% to 69% in PTC and 100% in TCV, as reported by Nikiforova and others.1-8 The increasing prevalence of the BRAFV600E mutation in the classic variant of PTC has raised the awareness of the role of its detection in everyday practice, which is mostly performed via molecular assays, including real-time polymerase chain reaction, allele-specific polymerase chain reaction, high-resolution melting analysis, Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
pyrosequencing, and Sanger sequencing.3,7,11,15,16,24,25 Despite the excellent accuracy of these methods, they have some limitations (ie, more laborious and time-consuming investigations, more expensive equipment, need for dedicated technical personnel, and contamination by nonneoplastic components), which may result in an unpredictable rate of nondiagnostic molecular yields. To overcome these drawbacks, alternative approaches, including recently introduced antibodies against mutated proteins and the identification of specific morphological features of mutated thyroid carcinomas, have been encouraged.15,16,23,26-29 To the best of our knowledge, very few detailed studies have found that PTCs with the BRAFV600E mutation show morphologically distinctive features.15,16 The articles by Virk et al15 and Finkelstein et al16 underline a series of easily reproducible cellular and architectural parameters that may help in predicting the BRAFV600E mutation. They reported the highest diagnostic accuracy (82%) for infiltrative borders and tumor-associated fibrosis as well as plump cells. In fact, Virk et al identified these plump cells as large polygonal tumor cells with a cell height less than twice the width that had features of squamoid metaplasia with homogeneous, eosinophilic, and moderate to abundant cytoplasm lining slit-like spaces and shared nuclear features of PTC. In their analysis, the authors reported that 29 of 50 PTCs were characterized by these plump cells with a 72% correlation with the BRAFV600E mutation and with a 28% correlation with WT plump cells. Virk et al’s findings are in keeping with the data from Finkelstein et al, who found these identical morphological features in 65% of BRAFV600E-mutated PTCs. To the best of our knowledge, our study is the first cytological evaluation of these predictive parameters; this is underlined by studies in the recent literature performed only with histological samples without any reference to their application to cytological thyroid samples. Our analysis was based on the immediate and glaring evidence of a nonuniform amount of these plump cells, and we aimed at (1) their cytological identification and (2) a tentative quantification of this pattern into diffuse (>20% plump cells) and limited categories (20% plump cells). Furthermore, the definition itself of these plump cells may induce a misdiagnoses of their features, so we intentionally excluded from our revision all cytologically malignant cases with a histological diagnosis of oxyphilic Cancer Cytopathology
December 2014
PTC and TCV PTC because these latter entities showed some distinctive and peculiar features, including marked cellular atypia, midplaced nuclei, and a slightly clearer apical region (in oxyphilic PTC) with a height exceeding twice the width (in TCV).23 Our study is partially in keeping with the data provided by Virk et al.15 We have reported 47 BRAFV600Emutated cases (65.3%) among 72 cytologically malignant cases. Although Virk et al described these plump cells in 72% of their BRAFV600E cases, we found them in all our 47 mutated cases; nevertheless, this was affected by a slightly intense pink cytoplasm, which may have been due to our liquid-based method (Figs. 1 and 3). The quantitative interpretation of these cells seems to be a critical parameter. Virk et al15 and Finkelstein et al16 deemed that even focal plump cells were linked with the BRAFV600E mutation, and accordingly, we found 6 of 47 BRAFV600E-mutated cases with focal plump cells. However, in our series, we reported 5 of 25 WT cases with focal plump cells. Eventually, what significantly emerged in the interpretation of focal plump cells was the problem of the semiquantitative evaluation of the percentage score, which might not predict a BRAFV600E mutation or WT assessment. To overcome the issues and contradictory interpretations of the focal plump cell component, we discussed the possible role of the VE1 antibody.23 Not only did a VE1 analysis confirm positivity in all 24 mutated cases and negativity in all 6 WT cases, but it also was likely to identify mutated cases with focal plump cells as well as sickle-shaped nuclei. On the basis of this compelling evidence, we encourage a collegial consensus on the morphology of focal plump cells followed by a VE1 analysis and finally a molecular throughput platform, including pyrosequencing or other high-sensitivity methods. Even with the additional aid of immunocytochemistry and the statistical significance of plump cells, we are conscious that this parameter was not univocally linked with the BRAFV600E mutation when psammoma bodies and cystic changes were evaluated (31.9% and 47.2%, respectively). Nevertheless, an accurate cytological analysis highlighted a peculiar nuclear shape in all the BRAFV600E cases. In fact, in our mutated samples, nuclei tended to be smaller and eccentrically located with a particular sickle shape that was absent in the WT BRAF counterpart. The sickle nuclear shape showed very irregular nuclear membranes associated with the classic features of PTC nuclei. We 889
Original Article
found that the BRAF activating mutation may have a different effect on the nucleus/cytoplasm ratio in favor of abundant cytoplasm, whereas the WT group exhibited typical PTC nuclei with a bigger nuclear size and less abundant cytoplasm in comparison with the mutated group. This morphological evidence suggests that cancer cells display a strikingly different metabolism generated through the active signaling pathway by mutated oncogenes.15 This different metabolism of BRAF-mutated cells, through the upregulation of the gene involved in glycolysis, may hesitate with increased mitochondrial respiration and justify the abundant eosinophilic cytoplasm of the cells.15 The correlation between thyroid variants of carcinoma and predictive mutational features is unlikely to be underlined. Although Virk et al15 selected BRAFV600Emutated PCT with the exclusion of variants such as FVPC, Finkelstein et al16 introduced a histotype distinction even though they did not detect any mutation in the FVPC variant. The lack of a specific and exclusive correlation between the mutation and the PTC histotype was suggested by our 10 BRAFV600E FVPC cases characterized by both plump cells and sickle nuclei. In conclusion, we have shown for the first time that specific cell features are highly associated with the BRAF mutation and may be used for a possible cytological prediction of this mutation30. Nonetheless, our results show that sickle nuclei may be a strict foreseeable parameter of the BRAFV600E mutation with the highest specificity (100%) and specificity (100%). Eventually, the detection of more than 20% plump cells and even more of these sickle nuclei may pave the way for triaging cases for BRAFV600E molecular analysis. FUNDING SUPPORT
4.
5. 6. 7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURES The authors made no disclosures.
REFERENCES 1.
2.
3.
890
Jung CK, Little MP, Lubin JH, et al. The increase in thyroid cancer incidence during the last four decades is accompanied by a high frequency of BRAF mutations and a sharp increase in RAS mutations. J Clin Endocrinol Metab. 2014;99:E276-E285. Mathur A, Moses W, Rahbari R, et al. Higher rate of BRAF mutation in papillary thyroid cancer over time: a single institution study. Cancer. 2011;117:4390-4395. Xing M. BRAF mutation in papillary thyroid cancer: pathogenic role, molecular bases, and clinical implications. Endocr Rev. 2007; 28:742-762.
19. 20.
21.
22.
23.
Kim TH, Park YJ, Lim JA, et al. The association of the BRAF(V600E) mutation with prognostic factors and poor clinical outcome in papillary thyroid cancer. Cancer. 2012;118:1764-1773. Nikiforova MN, Nikiforov Y. Molecular diagnostics and predictors in thyroid cancer. Thyroid. 2009;19:1351-1361. Nikiforov YE. Molecular diagnostics of thyroid tumors. Arch Pathol Lab Med. 2011;135:569-577. Oler G, Camacho CP, Hojaij FC, Michaluart P Jr, Riggins GJ, Cerutti JM. Gene expression profiling of papillary thyroid carcinoma identifies transcripts correlated with BRAF mutational status and lymph node metastasis. Clin Cancer Res. 2008;14:4735-4742. Trovisco V, Vieira de Castro I, Soares P, et al. BRAF mutations are associated with some histological types of papillary thyroid carcinoma. J Pathol. 2004;202:247-251. Rossi ED, Martini M, Capodimonti S, et al. Diagnostic and prognostic value of immunocytochemistry and BRAF mutation analysis on liquid-based biopsies of thyroid neoplasms suspicious for carcinoma. Eur J Endocrinol. 2013;168:853-859. Rossi ED, Martini M, Capodimonti S, et al. BRAF (V600E) mutation analysis on liquid-based cytology-processed aspiration biopsies predicts bilaterality and lymph node involvement in papillary thyroid microcarcinoma. Cancer Cytopathol. 2013;121:291-297. Rivera M, Ricarte-Filho J, Tuttle RM, et al. Molecular, morphologic, and outcome analysis of thyroid carcinomas according to degree of extrathyroid extension. Thyroid. 2010;20:1085-1093. Yip L, Nikiforova MN, Carty SE, et al. Optimizing surgical treatment of papillary thyroid carcinoma associated with BRAF mutation. Surgery. 2009;146:1215-1223. Guerra A, Fugazzola L, Marotta V, et al. A high percentage of BRAFV600E alleles in papillary thyroid carcinoma predicts a poorer out come. J Clin Endocrinol Metab. 2012;97:2333-2340. Gandolfi G, Sancisi V, Piana S, Ciarrocchi A. Time to re-consider the meaning of BRAF V600E mutation in papillary thyroid carcinoma. Int J Cancer. 2014 may 15 DOI: 10.1002/ijc28976 (epub ahead of print). Virk RK, Theoharis CGA, Prasad A, Chhieng D, Prasad ML. Morphology predicts BRAF (V600E) mutation in papillary thyroid carcinoma: an interobserver reproducibility study. Virchows Arch. 2014;464:435-442. Finkelstein A, Levy GH, Hui P, et al. Papillary thyroid carcinomas with and without BRAF V600E mutations are morphologically distinct. Histopathology. 2012;60:1052-1059. Hall A, Meyle KD, Lange MK, et al. Dysfunctional oxidative phosphorylation makes malignant melanoma cells addicted to glycolysis driven by the V600EBRAF oncogene. Oncotarget. 2013;4: 584-599. Fadda G, Basolo F, Bondi A, et al. Cytological classification of thyroid nodules. Proposal of the SIAPEC-IAP Italian Consensus Working Group. Pathologica. 2010;102:405-408. Cibas ES, Ali SZ. The Bethesda system for reporting thyroid cytopathology. Thyroid. 2009;19:1159-1165. British Thyroid Association. Guidelines for the Management of Thyroid Cancer. 2nd ed. London, United Kingdom: Royal College of Physicians of London; 2007. Martini M, Teofili L, Cenci T, et al. A novel heterozygous HIF2AM535I mutation reinforces the role of oxygen sensing pathway disturbances in the pathogenesis of familial erythrocytosis. Haematologica. 2008;93:1068-1071. Rossi ED, Martini M, Capodimonti S, et al. Analysis of immunocytochemistry and molecular BRAF expression in thyroid carcinoma: a cytohistologic institutional experience. Cancer Cytopathol. 2014;122:527-535. Compton CC, Byrd DR, Garcia-Aguilar J, Kurtzman SH, Olawaiye A, Washington MK, eds. AJCC Cancer Staging Atlas. 2nd ed. New York, NY: Springer; 2013.
Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
24. Kim SK, Kim DL, Han HS, et al. Pyrosequencing analysis for detection of a BRAFV600E mutation in an FNAB specimen of thyroid nodules. Diagn Mol Pathol. 2008;17:118-125. 25. Yeo MK, Liang ZL, Oh T, et al. Pyrosequencing cut-off value identifying BRAFV600E mutation in fine needle aspiration samples of thyroid nodules. Clin Endocrinol (Oxf). 2011;75:555-560. 26. Ghossein RA, Katabi N, Fagin JA. Immunohistochemical detection of mutated BRAF V600E supports the clonal origin of BRAFinduced thyroid cancers along the spectrum of disease progression. J Clin Endocrinol Metab. 2013;98:E1414-E1421. 27. Routhier CA, Mochel MC, Lynch K, et al. Comparison of 2 monoclonal antibodies for immunohistochemical detection of BRAF V600E mutation in malignant melanoma, pulmonary carci-
Cancer Cytopathology
December 2014
noma, gastrointestinal carcinoma, thyroid carcinoma, and gliomas. Hum Pathol. 2013;44:2563-2570. 28. Koperek O, Kornauth C, Capper D, et al. Immunohistochemical detection of the BRAF V600E-mutated protein in papillary thyroid carcinoma. Am J Surg Pathol. 2012;36:844-850. 29. Zimmermann AK, Camerisch U, Rechsteiner MP, BodeLesniewska B, Rossle M. Value of immunohistochemistry in detection of BRAF(V600E) mutations in fine needle aspiration biopsies of papillary thyroid carcinoma. Cancer Cytopathol. 2014;122:48-58. 30. Rossi ED, Bizzarro T, Fadda G, Larocca LM, Schmitt F. Is morphology alone able to predict BRAF-mutated malignancies on thyroid FNAC? Virchows Arch. 2014;465:247-248.
891
Morphological Parameters Able to Predict BRAFV600EMutated Malignancies on Thyroid Fine-Needle Aspiration Cytology: Our Institutional Experience Esther Diana Rossi, MD, PhD1; Tommaso Bizzarro, BD1; Maurizio Martini, MD, PhD; Sara Capodimonti, BD, PhD; Guido Fadda, MD1; Luigi Maria Larocca, MD1; and Fernando Schmitt, MD, PhD2,3,4
BACKGROUND: The BRAFV600E mutation is the most common diagnostic/prognostic marker in papillary thyroid carcinoma (PTC). Its evaluation is typically performed with DNA-based techniques; nonetheless, a few articles have recently proposed the morphological prediction of BRAFV600E in histological PTCs. We investigated this morphological parameter in our cytological series. METHODS: We re-analyzed all 72 cytohistological samples diagnosed as positive for malignancy (favoring PTC) on fine-needle aspiration cytology from January 2012 to December 2013. We included 22 male patients and 50 female patients. The cytological cases were processed with liquid-based cytology. We performed molecular analysis and immunocytochemistry for the VE1 BRAFV600E antibody. RESULTS: We reported 47 mutated cases and 25 wildtype (WT) cases with 100% cytohistological concordance. The cytological evaluations revealed plump cells (abundant eosinophilic cytoplasm and PTC nuclei) in all 47 mutated cases, with only 6 having a focal plump cell component (20% cells). Furthermore, 5 of the 25 WT cases showed focal plump cells. A distinctive sickle nuclear shape was found only in the mutated cases. VE1 yielded 100% positivity for all 24 mutated cases that were tested, including 3 cases with focal plump cells. CONCLUSIONS: We demonstrated that the BRAFV600E mutation might be predicted in cytological samples on the basis of some specific morphological features. Although the detection of plump cells (mainly focal) was also observed in WT cases, the detection of sickle-shaped nuclei provided the highest specificity and sensitivity as a predictive mutational parameter. These morphological features might be a valid tool for selecting cases for molecular analysis. C 2014 American Cancer Society. Cancer (Cancer Cytopathol) 2014;122:883-91. V
KEY WORDS: BRAF mutation; liquid-based cytology; thyroid carcinoma; thyroid lesions; plump cells.
INTRODUCTION Activating somatic mutations in the BRAF oncogene are common findings in thyroid carcinoma, as demonstrated by the 45% to 70% prevalence (mainly in the classic variant of papillary thyroid carcinoma ([PTC]).1-4 More than 95% of BRAF mutations involve the hotspot identified on exon 15 of the B isoform of the RAF kinase gene with a consequent valine–to–glutamic acid substitution.1-8 Furthermore, this mutational assessment has provided significant insights into a prognostically unfavorable outcome.2 In fact, it is frequently related to more aggressive behavior involving multifocality, extrathyroid infiltration, and lymph nodes and/or distant metastases.3,9-14 Corresponding author: Esther Diana Rossi, MD, PhD, Division of Anatomic Pathology and Histology of the Catholic University of the Sacred Heart, Agostino Gemelli School of Medicine, Largo Francesco Vito 1, 00168 Rome, Italy; Fax: (011) 3906 3015 7008; [email protected] 1
Division of Anatomic Pathology and Histology of the Catholic University of the Sacred Heart, Agostino Gemelli School of Medicine, Rome, Italy; Institute of Molecular Pathology and Immunology, University of Porto, Porto, Portugal; 3Department of Laboratory Medicine and Pathobiology, Faculty of Medicine, University of Toronto, Toronto, Canada; 4Department of Pathology, University Health Network, Toronto, Canada. 2
The last 2 authors share senior authorship. Received: June 18, 2014; Revised: July 25, 2014; Accepted: July 30, 2014 Published online August 25, 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cncy.21475, wileyonlinelibrary.com
Cancer Cytopathology
December 2014
883
Original Article
The diagnostic role of the BRAFV600E mutation in PTC has been extensively considered in an ample literature and has been mainly linked to specific histotypes, including the classic variant of PTC and the tall cell variant (TCV).1-4 According to the literature, the gold standard for BRAF analysis is an assessment with DNA-based methods; nonetheless, this may not represent the most cost-effective option. Despite their invaluable advantages, the difficulties encountered with the worldwide dissemination of these techniques have enhanced the enthusiasm for the recent introduction of the monoclonal V600E antibody (clone VE1) as well as the possible morphological prediction of specific features in thyroid BRAFV600Emutated cases.15-17 Few articles have underlined the distinctive morphological features of histological samples of BRAFV600Emutated PTC,15,16 including some architectural parameters (tumor-associated stromal reactions and infiltrative tumor borders) and cell-specific parameters (nuclear features and polygonal eosinophilic cells [defined as plump cells]). This alternative morphological perspective has gained popularity as a possibly foreseeable predictive aid.15,16 Hall et al.17 hypothesized that these morphological changes are the result of the involvement of the Warburg effect in enabling cancer cells to convert glucose into lactic acid. This effect causes the upregulation of oncogenes involved in RAS-BRAF-MEK-ERK pathways, which induce glycolysis even in the presence of abundant glucose.17 Consequently, an awareness of these histological findings raised our curiosity and led us to evaluate our cytological series of thyroid lesions diagnosed as positive for malignancy (favoring PTC). Because of the lack of infiltrative growth evidence on fine-needle aspiration cytology (FNAC), we focused on some specific cellular details of BRAFV600E-mutated follicular cells, mainly in the research of plump cells, which were defined as large polygonal tumor cells with a cell height less than twice the width that had features of squamoid metaplasia with homogeneous, eosinophilic, and moderate to abundant cytoplasm and shared nuclear features of PTC.15,16 We retrospectively investigated and quantified this morphological feature in 72 cytologically positive-formalignancy thyroid lesions with surgical pathology outcomes. Moreover, we discuss the immunocytochemistry with the VE1 antibody in 30 of the 72 cases. 884
MATERIALS AND METHODS Our series included all 72 retrospective positive-formalignancy (favoring PTC) samples from January 2012 to December 2013. All the cases were recorded at the Division of Anatomic Pathology and Histology of the Catholic University of the Sacred Heart (Agostino Gemelli Hospital of Rome). All FNAC was performed under sonographic guidance (ultrasound), mostly by surgeons and endocrinologists, and was processed with the ThinPrep 5000 liquid-based cytology (LBC) method (Hologic Co., Marlborough, Mass). Our patients were studied with ultrasound during a thyroid checkup performed at our hospital’s center for thyroid diseases. The series included 22 male patients and 50 female patients with a median age of 27 years (range 5 19-73 years). All aspirations (usually 2 passes for each lesion) were performed with 25- to 27-gauge needles; no rapid onsite assessment of the adequacy of the material was performed. All patients were appropriately informed about the use of LBC for processing their samples, and a written informed consent form was signed. Our study followed the tenets of the Declaration of Helsinki, and we received internal ethics approval for the study. The technical steps for liquid-based preparations are clearly described in previous articles from our group.9,10 Each resulting slide was fixed in 95% ethanol and stained with the Papanicolaou technique, whereas the remaining material was stored in PreservCyt solution (Hologic) to be possibly used for the preparation of additional slides for further investigations (including both immunocytochemistry and molecular analysis). Our cytological cases were classified as positive for malignancy (favoring PTC) according to the morphological criteria adopted by the Societa Italiana di Anatomia Patologica e Citologia Diagnostica–IAP Italian Working Group classification, which is similar to the Bethesda thyroid system.18,19 The lower limit of adequacy for each sample was established according to the British Royal College of Pathologists classification with 6 groups of thyroid follicular epithelial cells within the submitted slides, each having at least 10 well-visualized epithelial cells.20 Additional slides for immunocytochemistry and molecular analysis were obtained from the material stored in PreservCyt solution, so that analyses could be performed with even 2 mL of the material remaining eluted in 5 mL of PreservCyt solution. Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
The percentage of disease-specific cells for molecular analysis was at least 50% in all LBC samples, whereas the percentage was 30% for the immunocytochemical VE1 evaluation. Molecular Analysis
DNA was extracted from LBC samples stored in PreservCyt solution and from paraffin-embedded tissues. Each LBC sample was centrifuged, the supernatant was discarded, and the cellular pellet was processed. The pellet was incubated at 56 C for 3 hours in 180 lL of an ATL lysis buffer and 20 lL of proteinase K (20 mg/mL) from the QIAamp DNA mini kit (Qiagen, Hilden, Germany). For histological samples, 10 lm of slide tissue was deparaffinized and, after an ethanol treatment, was incubated at 56 C overnight in 180 lL of an ATL lysis buffer and 20 lL of proteinase K (20 mg/mL) from the QIAamp DNA mini kit. DNA was extracted according to the manufacturer’s protocol, and we assessed the quantity and quality of the DNA spectrophotometrically (absorbance at 260 nm and ratio of the absorbance at 260 nm to the absorbance at 280 nm; spectrum 5 220-320 nm; Biochrom, Cambridge, United Kingdom) and by separation on an Agilent 2100 bioanalyzer (Agilent, Palo Alto, Calif). Low-purity or insufficient DNA samples were extracted a second time. After the first amplification on a Rotor-Gene Q (Qiagen), the mutational analysis of BRAF was achieved with an anti-EGFR MoAb response (BRAF status) kit (reference UP033, Diatech Pharmacogenetics) by pyrosequencing via the PyroMark Q96 ID system (Biotage AB, Uppsala, Sweden). The sensitivity of this method was 5% in our laboratory. The following sequence was analyzed: ACAGT/AGAAA (exon 15).21 Immunocytochemistry and Immunohistochemistry
Because we retrospectively analyzed all 72 cytohistological series, the 30 cases with an analysis of the VE1 antibody on LBC were enrolled from our previous study.22 Immunohistochemistry for the VE1 antibody was run on LBC specimens according to the standard protocol reported in our previous article22 and was evaluated by 2 pathologists (E.D.R. and G.F.) blinded to the BRAF molecular status. The interpretation and validation of this VE1 application with cytology and LBC were supported by our previous experience.22 Briefly, the cytoplasm positivity of the samples was graded from 0 to 31 and was determined on both cytoCancer Cytopathology
December 2014
logical and histological samples. The intensity was defined as follows: 0, negative; 11, positivity in less than 30% of cells; 21, positivity in 30% or more than of cells and less than 80% of cells; and 31, positivity in more than 80% of cells. For our convenience, we report this arbitrary distribution with 4 categories of staining intensity (0, 11, 21, and 31), which are also based on the percentage of stained cells. The results for each category were homogeneous, regardless of the number of tumor cells stained. We did not find any discrepancy in the intensity or percentage of VE1-positive cells between cytological and histological samples. Histology
All surgical specimens were fixed in 10% buffered formaldehyde and embedded in paraffin, and the 5-lm-thick microtomic sections were stained with hematoxylin-eosin. The BRAF mutational analysis was also performed on DNA extracted from surgical specimens that were at least 70% tumor. All the perithyroid adipose tissue was embedded and examined for lymph node research. The diagnosis of PTC was based on the presence of true papillary structures and distinctive nuclear features, whereas the diagnosis of the follicular variant of papillary thyroid carcinoma (FVPC) relied on the detection of the nuclear features of PTC in multiple foci within the tumor.21 The diagnosis of TCV was characterized by a predominance of neoplastic cells whose heights were at least 3 times their widths and that had classic PTC nuclear features. The oxyphilic variant of PTC was defined by distinctive neoplastic cells showing marked cellular atypia with an oxyphilic basal zone, midplaced nuclei, and a slightly clearer apical region. All the cases were classified according to the seventh edition of the tumor-node-metastasis–based staging system recommended by the American Joint Commission on Cancer.23 Statistical Analysis
Statistical analysis was performed with a commercially available statistical software package (SPSS 10.0 for Windows, Chicago, Ill). Comparisons of categorical variables were performed with the chi-square statistic; Fisher’s exact test was used when that was appropriate. A P value less than .05 was considered significant.
RESULTS We re-analyzed all 72 cytological samples diagnosed as positive for malignancy between January 2012 and 885
Original Article
December 2013 with surgical follow-up. Table 1 shows the distribution of the clinical features in 47 BRAFV600Emutated cases (65.3%) and 25 wild-type (WT) BRAFV600E cases (34.7%; Fig. 1A). Regardless of the BRAFV600E mutation, 57 of the 72 nodules were equal to or smaller than 2 cm. There were no significant differences in sex or tumor size with respect to the 2-cm cutoff. All cases were histologically confirmed, and they were diagnosed as a classic variant of PTC (n 5 52) or FVPC (n 5 20). We excluded all cases showing TCV and oxyTABLE 1. Clinical Features of 72 Cytohistological Series
Clinical Feature Sex, n Male Female Age (years), range (median) Size, n 2 cm >2 cm Histology, n PTC FVPC T stage, n TI-TII TIII-TIV N stage, n pN0 pN1
BRAFV600E Mutation (47 Cases)
Wild-Type BRAF (25 Cases)
12 35 19-70 (25.5)
10 15 20-73 (26.5)
40 7
17 8
37 10
15 10
33 (25 TI 1 8 TII) 14 (12 TIII 1 2 TIV)
19 (13 TI 16 TII) 6 (5 TIII 11 TIV)
22 25
15 10
P Value NS (.28)
NS NS (.13)
NS (.11)
NS (.78)
NS (.33)
Abbreviations: FVPC, follicular variant of papillary thyroid carcinoma; N stage, lymph node stage; NS, not significant; PTC, papillary thyroid carcinoma; T stage, tumor stage.
philic PTC on histology. Moreover, the T stage and the N stage did not reach statistical significance in the 2 groups. As for the T stage, we had a higher number of T3 patients in the mutated group versus the WT group (25.5% vs 20%). For the statistical evaluation, we subdivided the cases into TI-TII and TIII-TIV groups. Lymph node metastases were found in 25 of the 47 BRAFV600Emutated cases (54%) and in 10 of the 25 WT cases (40%). As underlined in Table 1, there were no statistical significant differences for any of the parameters. In accordance with Virk et al15 and Finkelstein et al,16 we have underlined the presence of some cellular features predictive of the BRAFV600E mutation in Table 2. Regardless of the BRAFV600E mutation, all our samples showed typical nuclear features of PTC (grooves, pleomorphic membranes, and nuclear pseudo-inclusions). In addition, we found plump cells showing a polygonal shape and characterized by abundant eosinophilic cytoplasm. Noticeably, in our LBC preparations, the specific color of these plump cells seemed to be more intense than the pink eosinophilic shade on histology (Figs. 1B, 2, and 3). Moreover, for our convenience, we qualified these plump cell component as (1) diffuse (up to >20%) or (2) focal (20%). As documented in Table 2, these plump cells were significantly associated with the BRAFV600E mutation (P 5 .0002, odds ratio 5 70.23, 95% confidence interval 5 3.46-1427), with only 6 of 47 mutated cases having a focal plump cell component. Moreover, we found that 5 of the 25 WT cases had focal plump cells (Table 2). Our data revealed a significant correlation
Figure 1. (A) Details of the morphological features of wild-type cells obtained from a positive-for-malignancy papillary thyroid carcinoma case on liquid-based cytology (340). (B) Details of the morphological features of plump cells obtained from a positive-for-malignancy papillary thyroid carcinoma case on liquid-based cytology (340).
886
Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
TABLE 2. Evaluation of the Cytological Parameters of Positive-for-Malignancy Cases With and Without the BRAFV600E Mutation Cytological Feature
Mutated BRAF (47 Cases)
Wild-Type BRAF (25 Cases)
6 41 47 47 10 15
5b 0 25 0 12 6
Plump cellsa 20% >20% Classic PTC nuclear featuresc Sickle-shaped nuclei Psammoma bodies Cystic changes
Sensitivity
Specificity
PPV
NPV
Diagnostic Accuracy
90.4%
100%
100%
80%
93%
65.3% 100% 45.5% 71.4%
0% 100% 26% 37.3%
100% 100% 21.3% 31.9%
0% 100% 52% 76%
65.3% 100% 31.9% 47.2%
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; PTC, papillary thyroid carcinoma. a P 5.0002 (odds ratio 5 70.23, 95% confidence interval 5 3.46-1427). b The remaining 20 cases did not show any evidence of plump cell components. c These nuclear features included nuclear pseudo-inclusions, irregular membranes, and nuclear grooves.
Figure 2. Details of the morphological features of plump cells organized into small groups with the height of the cells less than twice the width (liquid-based cytology, 340).
between BRAF mutation analysis and a peculiar nuclear shape as an additional cellular parameter. This feature was present in 100% of our BRAFV600E-mutated cases and was absent in WT cases (Table 2). We defined this feature as a sickle-shaped nucleus (Fig. 3). These sickle-shaped nuclei tended to be smaller and eccentrically located within pleomorphic and irregular membranes, so the detection of nuclear pseudo-inclusions might be more difficult. On balance, the WT cases were characterized by PTC nuclei with a bigger nuclear size and less abundant cytoplasm in comparison with mutated cases. We did find 10 and 12 psammoma bodies in the mutated and WT groups, respectively, as detailed in Table 2. Some cases showed cystic changes (15 mutated cases vs 6 WT cases) without any statistical significance. We found a 100% cytohistological correlation for these parameters (Fig. 4). Cancer Cytopathology
December 2014
Figure 3. Details of the morphological features of sickleshaped malignant nuclei obtained from a positive-formalignancy papillary thyroid carcinoma case on liquid-based cytology (340).
In fact, the same peculiar parameters were studied and evaluated for the corresponding histological sections, and this resulted in 100% cytohistological concordance for the quantitative evaluation of plump cells and sickleshaped nuclei (Fig. 4A,B). Also in Table 2, a specific evaluation of the sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy is presented for each analyzed feature. Table 3 reports the results for our 30 cases with VE1 expression, including 24 BRAFV600E-mutated cases and 6 WT cases. The results for each category of immunocytochemistry were homogeneous, regardless of the number of stained cells in each of the categories. As for the distribution of VE1, our mutated series had 13 cases with 21 positivity and 11 cases with 31 positivity (Fig. 5), whereas 887
Original Article
Figure 4. (A) Specific histological details of polygonal cells with homogeneous eosinophilic cytoplasm (plump cells) at the bottom of the picture obtained from a follicular variant of papillary thyroid carcinoma (hematoxylin-eosin, 360). (B) Specific histological details of plump cells with details on the sickle-shaped nuclei (hematoxylin-eosin, 360).
TABLE 3. Evaluation of 30 of 72 Cases With VE1 Expression22
Cytological Feature Mutated BRAFV600E Wild-type BRAF
Plump Cells
VE1-Negative (0 and 11 Intensity)
VE1-Positive (21 and 31 Intensity)
20% >20% Absent 20% >20%
0 0 5b 1c 0
3a 21 0 0 0
a
All 3 focal plump cell cases were 21 VE1-positive. These 5 cases included all 5 wild-type cases without plump cells. c Only 1 case belonged to the group with 20% plump cells (expressed as 11 VE1). b
our 6 negative cases were distributed as 4 cases with 0 positivity (these 4 cases did not present with any plump cell components) and 2 cases with 11 positivity (one belonged to the group without plump cells, and the second belonged to the group with 20% plump cells). Four cases with focal plump cells belonged to the VE 1 series (3 were mutated cases with 21 VE1, whereas 1 was a WT case with 11 VE1).
DISCUSSION The aims of this study were to evaluate the specific morphological features of the BRAFV600E mutation on cytological thyroid malignant lesions (namely PTC and its variants) and to provide a possible means of predictive molecular detection. In the field of thyroid oncology, the BRAFV600E mutation is synonymous with thyroid carcinoma, mainly 888
Figure 5. Immunocytochemistry cytoplasm staining results for a mutated case with plump cells on liquid-based cytology: 31 VE1 positivity in tumor cells (avidin-biotin complex, 3400).
in the classic variant of PTC.1-6 The high specificity of this somatic mutation underlines its early driving role in thyroid tumorigenesis and its association with a more aggressive phenotype. Several series have documented mutational frequencies of 40% to 69% in PTC and 100% in TCV, as reported by Nikiforova and others.1-8 The increasing prevalence of the BRAFV600E mutation in the classic variant of PTC has raised the awareness of the role of its detection in everyday practice, which is mostly performed via molecular assays, including real-time polymerase chain reaction, allele-specific polymerase chain reaction, high-resolution melting analysis, Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
pyrosequencing, and Sanger sequencing.3,7,11,15,16,24,25 Despite the excellent accuracy of these methods, they have some limitations (ie, more laborious and time-consuming investigations, more expensive equipment, need for dedicated technical personnel, and contamination by nonneoplastic components), which may result in an unpredictable rate of nondiagnostic molecular yields. To overcome these drawbacks, alternative approaches, including recently introduced antibodies against mutated proteins and the identification of specific morphological features of mutated thyroid carcinomas, have been encouraged.15,16,23,26-29 To the best of our knowledge, very few detailed studies have found that PTCs with the BRAFV600E mutation show morphologically distinctive features.15,16 The articles by Virk et al15 and Finkelstein et al16 underline a series of easily reproducible cellular and architectural parameters that may help in predicting the BRAFV600E mutation. They reported the highest diagnostic accuracy (82%) for infiltrative borders and tumor-associated fibrosis as well as plump cells. In fact, Virk et al identified these plump cells as large polygonal tumor cells with a cell height less than twice the width that had features of squamoid metaplasia with homogeneous, eosinophilic, and moderate to abundant cytoplasm lining slit-like spaces and shared nuclear features of PTC. In their analysis, the authors reported that 29 of 50 PTCs were characterized by these plump cells with a 72% correlation with the BRAFV600E mutation and with a 28% correlation with WT plump cells. Virk et al’s findings are in keeping with the data from Finkelstein et al, who found these identical morphological features in 65% of BRAFV600E-mutated PTCs. To the best of our knowledge, our study is the first cytological evaluation of these predictive parameters; this is underlined by studies in the recent literature performed only with histological samples without any reference to their application to cytological thyroid samples. Our analysis was based on the immediate and glaring evidence of a nonuniform amount of these plump cells, and we aimed at (1) their cytological identification and (2) a tentative quantification of this pattern into diffuse (>20% plump cells) and limited categories (20% plump cells). Furthermore, the definition itself of these plump cells may induce a misdiagnoses of their features, so we intentionally excluded from our revision all cytologically malignant cases with a histological diagnosis of oxyphilic Cancer Cytopathology
December 2014
PTC and TCV PTC because these latter entities showed some distinctive and peculiar features, including marked cellular atypia, midplaced nuclei, and a slightly clearer apical region (in oxyphilic PTC) with a height exceeding twice the width (in TCV).23 Our study is partially in keeping with the data provided by Virk et al.15 We have reported 47 BRAFV600Emutated cases (65.3%) among 72 cytologically malignant cases. Although Virk et al described these plump cells in 72% of their BRAFV600E cases, we found them in all our 47 mutated cases; nevertheless, this was affected by a slightly intense pink cytoplasm, which may have been due to our liquid-based method (Figs. 1 and 3). The quantitative interpretation of these cells seems to be a critical parameter. Virk et al15 and Finkelstein et al16 deemed that even focal plump cells were linked with the BRAFV600E mutation, and accordingly, we found 6 of 47 BRAFV600E-mutated cases with focal plump cells. However, in our series, we reported 5 of 25 WT cases with focal plump cells. Eventually, what significantly emerged in the interpretation of focal plump cells was the problem of the semiquantitative evaluation of the percentage score, which might not predict a BRAFV600E mutation or WT assessment. To overcome the issues and contradictory interpretations of the focal plump cell component, we discussed the possible role of the VE1 antibody.23 Not only did a VE1 analysis confirm positivity in all 24 mutated cases and negativity in all 6 WT cases, but it also was likely to identify mutated cases with focal plump cells as well as sickle-shaped nuclei. On the basis of this compelling evidence, we encourage a collegial consensus on the morphology of focal plump cells followed by a VE1 analysis and finally a molecular throughput platform, including pyrosequencing or other high-sensitivity methods. Even with the additional aid of immunocytochemistry and the statistical significance of plump cells, we are conscious that this parameter was not univocally linked with the BRAFV600E mutation when psammoma bodies and cystic changes were evaluated (31.9% and 47.2%, respectively). Nevertheless, an accurate cytological analysis highlighted a peculiar nuclear shape in all the BRAFV600E cases. In fact, in our mutated samples, nuclei tended to be smaller and eccentrically located with a particular sickle shape that was absent in the WT BRAF counterpart. The sickle nuclear shape showed very irregular nuclear membranes associated with the classic features of PTC nuclei. We 889
Original Article
found that the BRAF activating mutation may have a different effect on the nucleus/cytoplasm ratio in favor of abundant cytoplasm, whereas the WT group exhibited typical PTC nuclei with a bigger nuclear size and less abundant cytoplasm in comparison with the mutated group. This morphological evidence suggests that cancer cells display a strikingly different metabolism generated through the active signaling pathway by mutated oncogenes.15 This different metabolism of BRAF-mutated cells, through the upregulation of the gene involved in glycolysis, may hesitate with increased mitochondrial respiration and justify the abundant eosinophilic cytoplasm of the cells.15 The correlation between thyroid variants of carcinoma and predictive mutational features is unlikely to be underlined. Although Virk et al15 selected BRAFV600Emutated PCT with the exclusion of variants such as FVPC, Finkelstein et al16 introduced a histotype distinction even though they did not detect any mutation in the FVPC variant. The lack of a specific and exclusive correlation between the mutation and the PTC histotype was suggested by our 10 BRAFV600E FVPC cases characterized by both plump cells and sickle nuclei. In conclusion, we have shown for the first time that specific cell features are highly associated with the BRAF mutation and may be used for a possible cytological prediction of this mutation30. Nonetheless, our results show that sickle nuclei may be a strict foreseeable parameter of the BRAFV600E mutation with the highest specificity (100%) and specificity (100%). Eventually, the detection of more than 20% plump cells and even more of these sickle nuclei may pave the way for triaging cases for BRAFV600E molecular analysis. FUNDING SUPPORT
4.
5. 6. 7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURES The authors made no disclosures.
REFERENCES 1.
2.
3.
890
Jung CK, Little MP, Lubin JH, et al. The increase in thyroid cancer incidence during the last four decades is accompanied by a high frequency of BRAF mutations and a sharp increase in RAS mutations. J Clin Endocrinol Metab. 2014;99:E276-E285. Mathur A, Moses W, Rahbari R, et al. Higher rate of BRAF mutation in papillary thyroid cancer over time: a single institution study. Cancer. 2011;117:4390-4395. Xing M. BRAF mutation in papillary thyroid cancer: pathogenic role, molecular bases, and clinical implications. Endocr Rev. 2007; 28:742-762.
19. 20.
21.
22.
23.
Kim TH, Park YJ, Lim JA, et al. The association of the BRAF(V600E) mutation with prognostic factors and poor clinical outcome in papillary thyroid cancer. Cancer. 2012;118:1764-1773. Nikiforova MN, Nikiforov Y. Molecular diagnostics and predictors in thyroid cancer. Thyroid. 2009;19:1351-1361. Nikiforov YE. Molecular diagnostics of thyroid tumors. Arch Pathol Lab Med. 2011;135:569-577. Oler G, Camacho CP, Hojaij FC, Michaluart P Jr, Riggins GJ, Cerutti JM. Gene expression profiling of papillary thyroid carcinoma identifies transcripts correlated with BRAF mutational status and lymph node metastasis. Clin Cancer Res. 2008;14:4735-4742. Trovisco V, Vieira de Castro I, Soares P, et al. BRAF mutations are associated with some histological types of papillary thyroid carcinoma. J Pathol. 2004;202:247-251. Rossi ED, Martini M, Capodimonti S, et al. Diagnostic and prognostic value of immunocytochemistry and BRAF mutation analysis on liquid-based biopsies of thyroid neoplasms suspicious for carcinoma. Eur J Endocrinol. 2013;168:853-859. Rossi ED, Martini M, Capodimonti S, et al. BRAF (V600E) mutation analysis on liquid-based cytology-processed aspiration biopsies predicts bilaterality and lymph node involvement in papillary thyroid microcarcinoma. Cancer Cytopathol. 2013;121:291-297. Rivera M, Ricarte-Filho J, Tuttle RM, et al. Molecular, morphologic, and outcome analysis of thyroid carcinomas according to degree of extrathyroid extension. Thyroid. 2010;20:1085-1093. Yip L, Nikiforova MN, Carty SE, et al. Optimizing surgical treatment of papillary thyroid carcinoma associated with BRAF mutation. Surgery. 2009;146:1215-1223. Guerra A, Fugazzola L, Marotta V, et al. A high percentage of BRAFV600E alleles in papillary thyroid carcinoma predicts a poorer out come. J Clin Endocrinol Metab. 2012;97:2333-2340. Gandolfi G, Sancisi V, Piana S, Ciarrocchi A. Time to re-consider the meaning of BRAF V600E mutation in papillary thyroid carcinoma. Int J Cancer. 2014 may 15 DOI: 10.1002/ijc28976 (epub ahead of print). Virk RK, Theoharis CGA, Prasad A, Chhieng D, Prasad ML. Morphology predicts BRAF (V600E) mutation in papillary thyroid carcinoma: an interobserver reproducibility study. Virchows Arch. 2014;464:435-442. Finkelstein A, Levy GH, Hui P, et al. Papillary thyroid carcinomas with and without BRAF V600E mutations are morphologically distinct. Histopathology. 2012;60:1052-1059. Hall A, Meyle KD, Lange MK, et al. Dysfunctional oxidative phosphorylation makes malignant melanoma cells addicted to glycolysis driven by the V600EBRAF oncogene. Oncotarget. 2013;4: 584-599. Fadda G, Basolo F, Bondi A, et al. Cytological classification of thyroid nodules. Proposal of the SIAPEC-IAP Italian Consensus Working Group. Pathologica. 2010;102:405-408. Cibas ES, Ali SZ. The Bethesda system for reporting thyroid cytopathology. Thyroid. 2009;19:1159-1165. British Thyroid Association. Guidelines for the Management of Thyroid Cancer. 2nd ed. London, United Kingdom: Royal College of Physicians of London; 2007. Martini M, Teofili L, Cenci T, et al. A novel heterozygous HIF2AM535I mutation reinforces the role of oxygen sensing pathway disturbances in the pathogenesis of familial erythrocytosis. Haematologica. 2008;93:1068-1071. Rossi ED, Martini M, Capodimonti S, et al. Analysis of immunocytochemistry and molecular BRAF expression in thyroid carcinoma: a cytohistologic institutional experience. Cancer Cytopathol. 2014;122:527-535. Compton CC, Byrd DR, Garcia-Aguilar J, Kurtzman SH, Olawaiye A, Washington MK, eds. AJCC Cancer Staging Atlas. 2nd ed. New York, NY: Springer; 2013.
Cancer Cytopathology
December 2014
Morphology of BRAF-Mutated Thyroid/Rossi et al
24. Kim SK, Kim DL, Han HS, et al. Pyrosequencing analysis for detection of a BRAFV600E mutation in an FNAB specimen of thyroid nodules. Diagn Mol Pathol. 2008;17:118-125. 25. Yeo MK, Liang ZL, Oh T, et al. Pyrosequencing cut-off value identifying BRAFV600E mutation in fine needle aspiration samples of thyroid nodules. Clin Endocrinol (Oxf). 2011;75:555-560. 26. Ghossein RA, Katabi N, Fagin JA. Immunohistochemical detection of mutated BRAF V600E supports the clonal origin of BRAFinduced thyroid cancers along the spectrum of disease progression. J Clin Endocrinol Metab. 2013;98:E1414-E1421. 27. Routhier CA, Mochel MC, Lynch K, et al. Comparison of 2 monoclonal antibodies for immunohistochemical detection of BRAF V600E mutation in malignant melanoma, pulmonary carci-
Cancer Cytopathology
December 2014
noma, gastrointestinal carcinoma, thyroid carcinoma, and gliomas. Hum Pathol. 2013;44:2563-2570. 28. Koperek O, Kornauth C, Capper D, et al. Immunohistochemical detection of the BRAF V600E-mutated protein in papillary thyroid carcinoma. Am J Surg Pathol. 2012;36:844-850. 29. Zimmermann AK, Camerisch U, Rechsteiner MP, BodeLesniewska B, Rossle M. Value of immunohistochemistry in detection of BRAF(V600E) mutations in fine needle aspiration biopsies of papillary thyroid carcinoma. Cancer Cytopathol. 2014;122:48-58. 30. Rossi ED, Bizzarro T, Fadda G, Larocca LM, Schmitt F. Is morphology alone able to predict BRAF-mutated malignancies on thyroid FNAC? Virchows Arch. 2014;465:247-248.
891
