BRAFV600E Mutation in Papillary Thyroid Carcinoma:Significant Association with Node Metastasesand Extra Thyroidal Invasion

Avik Chakraborty & Archana Narkar &

Rita Mukhopadhyaya & Shubhada Kane & Anil D’Cruz &

M. G. R. Rajan

Published online: 22 November 2011# Springer Science+Business Media, LLC 2011

Abstract B-Raf (BRAF) is the strongest activator in thedownstream of MAP kinase signaling. The somatic pointmutation of BRAF gene (V600E) is the most common andspecific event in papillary thyroid carcinoma (PTC).However, its prevalence is variable among differentstudies and its association with clinico-pathologicalfeatures is controversial. This study tests the prevalenceof BRAFV600E mutation in thyroid cancer patients inIndian subcontinental population. We analyzed 140 thyroidtumor specimens for BRAF gene mutation at codon 600using mutant-allele-specific amplification, single-strandconformation polymorphism, Mutector assay, and DNAsequencing of the PCR-amplified exon 15. BRAFmutation at codon 600 was detected in 46 of 86 PTCpatients (53.4%) from Indian subcontinental cohort.Frequency of mutation varied across the subtypes ofPTCs. BRAFV600E mutation was more common in theconventional PTC (38 out of 62; 61%) than in thefollicular variant of PTC (2 out of 17; 11.7%). None of

the 8 follicular thyroid adenomas, 14 follicular thyroidcarcinomas, 16 medullary thyroid carcinomas, and 16benign hyperplasia patients showed any exon 15 mutation.We found significant correlation between BRAF mutationstatus and extra-thyroidal invasion, lymph node metastasis,and tumor stage. However no correlation was observedwith gender, age, and tumor size of the patients. Thusour findings suggest that BRAFV600E is a prevalentgenetic alteration in adult sporadic PTCs in Indian cohortand it may be responsible for the progression of classicvariant of PTC to metastatic and poorly differentiatedsubtype and likely to have significant impact on itsdiagnostic and prognostic management.

Keywords Thyroid carcinoma . BRAF mutation .

Lymph node metastases . Extra-thyroidal invasion

Introduction

Thyroid cancer has a wide range of biological and clinicalbehavior ranging from very indolent well-differentiatedcarcinoma to highly aggressive poorly differentiated andanaplastic carcinoma [1]. Genetic alteration(s) is a drivingforce for thyroid tumorigenesis and progression and havebeen associated with specific types of thyroid cancer. Ofall the histological types of thyroid cancer, the mostprevalent is papillary thyroid cancer (PTC) whichaccounts for 80–90% [2, 3]. The genetic alterationsimplicated in PTCs are RET/PTC rearrangement [4, 5],RAS mutations [6], and BRAF mutations [7]. The aberrantactivation of classic signal pathway receptor tyrosinekinase-RAS-RAF-MEK-ERK due to either of thesealterations leads to PTC. RAF kinase which is a

A. Chakraborty :A. Narkar (*) :M. G. R. RajanRadiation Medicine Centre, BARC,C/o Tata Memorial Hospital Annexe,Parel, Mumbai 400012, Indiae-mail: narkar.aa@gmail.com

R. MukhopadhyayaMolecular Biology Division, BARC,Trombay, Mumbai 400085, India

S. KaneDepartment of Pathology, Tata Memorial Hospital,Parel, Mumbai 400012, India

A. D’CruzHead and Neck Services, Tata Memorial Hospital,Parel, Mumbai 400012, India

Endocr Pathol (2012) 23:83–93DOI 10.1007/s12022-011-9184-5

component of this pathway connects aberrant activatingsignals to cell cycle machinery and plays a key role intransformation. It exhibits various isoforms and, amongthem, BRAF is the strongest activator of the downstreamMAP kinase signaling. The somatic point mutation ofBRAF which occurs commonly in a broad range of humancancers, most notably in melanoma, is due to the thymineto adenine transversion at nucleotide position 1799 ofBRAF gene, resulting in a valine to glutamic acidsubstitution at residue 600 (BRAFV600E) [8].

This sporadic point mutation in the BRAF gene hasbeen identified as the most common and specific geneticevent in PTC with a prevalence ranging from 29% to 83%[9, 10]. The BRAF mutation found in thyroid cancer isalmost exclusively T1799A transversion mutation in exon15. It can now be detected pre-operatively on FNABspecimens and has been proposed to have potential as adiagnostic adjunctive tool in evaluation of thyroidnodules with indeterminate cytological findings [11–13]as well as a prognostic marker to improve risk evaluationand recurrence prediction [14–16]. Several studies havebeen addressed to establish a correlation betweenBRAFV600E mutation and clinico-pathological features ofPTC. However, the results have been controversial [17–19].The discrepancies in the results may be due to the differentsample sizes including different PTC variants and differentgeographic areas.

It is known that uncertain prevalence might bias thepotential usefulness of the detection of BRAF mutation as adiagnostic and prognostic molecular marker for PTC. [8].Hence, it is mandatory to build up the knowledge ofprevalence of BRAF mutation and it's relation with theclinical behavior of PTC in different cohorts all over theworld. This will be certainly useful in developing BRAFmutation as a molecular marker for the management of PTCin the clinics.

The overall incidence of thyroid cancer in India is 3–4%with 80% of it being PTC [20]. However, the type andprevalence of the BRAF mutations in PTC patients remainsso far unidentified in Indian population. It is known thatdifferent geographic areas and genetic background amongraces may contribute to difference in frequency and typeof mutations as well as their role in the molecularpathogenesis of thyroid cancer [5, 8].

The aim of the present study was to establish theprevalence of this genetic alteration and to define associ-ation of BRAFV600E mutation with clinico-pathologicalcharacteristics of PTC in Indian subcontinental population.For this purpose, 140 sporadic thyroid carcinoma patientswith different histological variants were included in thestudy. The mutational analysis of the BRAF gene wasperformed by an assay based on mutant-allele-specificamplification (MASA), Mutector assay, single-strand

conformation polymorphism (SSCP) or by polymerasechain reaction (PCR) and direct sequencing.

Materials and Methods

Tumor Tissue Specimens

Frozen thyroid tissues from thyroid cancer patients who hadattended the clinic between 2002 and 2006 were retrievedfrom the Tissue Repository of Tata Memorial Hospital,Mumbai, India for analysis of BRAF gene mutation atcodon 600. A total of 140 specimens with adequate clinicaland pathological information were studied. Tissue speci-mens included 86 PTC, 14 follicular thyroid carcinomas(FTC), 8 follicular thyroid adenomas (FTA), 16 medullarythyroid carcinomas (MTC), and 16 benign hyperplasias.Acquisition of the tissue specimens was approved by theScientific Review Committee of Tata Memorial Hospital.Experiments were performed in accordance with regulationsof the hospital's Human Ethics Committee.

All the patients were selected randomly. Their clinicalinformation was retrieved retrospectively from patients'medical records, pathology reports, and subsequentclinical courses. Histological diagnosis was reconfirmedby an experienced pathologist after hematoxylin andeosin staining of sections. Patients were staged using thetumor-node-metastases (TNM) system and classifiedaccording to the presence of extra-thyroidal extension,cervical nodes, and distant metastases.

Cell Lines

Five thyroid tumor cell lines were used in this study(Table 1). Follicular carcinoma (WRO), papillary carcinoma(NPA), and anaplastic carcinoma (ARO and FRO) celllines were procured from Dr. Giorgio Stassi's laboratory(Universita Degli Studi Di Palermo, Palermo, Italy) whichhad been originally developed by Dr. Guy J. F. Juillard atthe University of California at Los Angeles. Medullarythyroid carcinoma cell line (TT) was received fromNational Centre for Cell Sciences, Pune, India. All thesethyroid carcinoma cell lines were grown in Iscov's

Table 1 BRAFV600E mutation in thyroid carcinoma cell lines

Cell line Type Metastasis BRAF V600E

WRO FTC Yes −NPA PTC Unknown +

ARO ATC Yes +

FRO ATC Yes +

TT MTC Unknown −

84 Endocr Pathol (2012) 23:83–93

Modified Dulbecco's Medium containing 10% fetal calfserum at 37°C and 5% CO2 [21].

Molecular Analysis

Genomic DNA Extraction

Nucleic acid extraction from frozen tissues, cell lines, andnormal peripheral blood mononuclear cells (PBMNC) wascarried out using QIAampR DNA Mini Kit (Qiagen,Germany) with some modifications. All samples weresubjected to digestion in digestion buffer (Tris-Cl100 mM, EDTA 1 mM, 1% SDS) containing 1 mg/ml ofproteinase-K at 60°C for 12 h. Concentration of DNA wasquantified by measurement of OD at 260 nm in NanodropReader (Nanodrop Technologies, ND1000, USA).Onehundred nanogram of genomic DNA was used as templatefor detection of mutation.

PCR Amplification

Genomic DNA (100 ng) was used as template to amplify215-bp fragment of BRAF gene exon 15. Sequences ofthe forward 5′-TCATAATGCTTGCTCTGATAGGA-3′andreverse 5′-GGCCAAAATTTAATCAGTGGA-3′ primerswere based on flanking introns [21]. PCR reaction wasset with an initial denaturation of 3 min at 94°C followedby 35 cycles of denaturation for 30 s at 94°C, annealingfor 45 s at 55°C, extension for 45 s at 72°C and a finalextension step for 10 min at 72°C. The NPA and ARO celllines were used as positive controls for BRAF T1799Amutation, and WRO cell line and PBMNC were used ascontrols for wild-type BRAF gene.

Sequencing

BRAF gene exon 15 amplicons generated from tumortissue, cell lines, and PBMNCs were subjected to automatedDNA sequencing in an ABI PRISM 377–18 DNASequencer (ABI, USA). Each amplicon was sequencedin both directions using Big Dye Terminator Version.3.1Cycle Sequencing kit (ABI, USA) and above-mentionedforward and reverse primers, respectively. All ampliconswere cleaned by QIAquick gel extraction kit (Qiagen,Germany) before being used as template. Cycle sequencingreaction and PCR program was followed as per manu-facturer's protocol. Sequences were compared by theBLAST program: www.ncbi.nlm.nih.gov/BLAST.

Mutant-Allele-Specific Amplification

MASA was carried out with two forward primers onecomplimentary to wild-type sequence, another having

mutant-allele-specific nucleotide base substitutions at the3′end (5′-TAG GTG ATT TTG GTC TAG CTA CAG T-3′and 5′-GGT GAT TTT GGT CTA GCT ACA AA- 3′) andthe common reverse primer used earlier. Each primerwas designed to amplify the wild-type allele of BRAFT1799A and transversion mutation, respectively [22].Amplification reactions were performed in a PTC-1148 MJThermal Cycler (MJ-Bio-Rad, USA), following an initialdenaturation of 2 min at 94°C and subsequent denaturationfor 30 s at 94°C, annealing for 45 s at 52°C andextension for 45 s at 72°C for 35 cycles (40 cycles fornegative samples for verification). Amplicons wereseparated on 2% agarose gel and visualized by ethidiumbromide staining. All samples were examined twice forthe conformation of BRAF mutation.

Single-Strand Conformation Polymorphism

Exon 15 amplicons of BRAF gene were diluted 1:1 indenaturing solution (0.25% Bromophenol blue and 95%formamide and 0.1 M NaOH) before boiling for 3 min inboiling water bath and rapid chilling on ice to preventreannealing of single-strand products. Electrophoresiswas carried out in the Macrokin-S Electrophoresis Unit(Techno Source, India) with Fotodyne DNA Seq. System,Model - 4200, USA, at 9°C, 120 V, 30 mA for 6 h. Gelswere stained with silver nitrate [23].

Mutector Assay

Mutector assay was also used to detect BRAF mutationat nucleotide position 1799. The Mutector assay isdesigned to detect point mutations of known DNAsequence variation. For this, a detection primer (providedin kit) allowing primer extension in cases of mutantamplicons and preventing primer extension from wild-type template was used. If primer extension does notoccur, labeled nucleotides are not incorporated, and acolor reaction is not observed. When the target base ismutated (e.g., T→A point mutation at BRAF T1799),primer extension continues and a strong color reaction isobserved. The Mutector assay is highly sensitive and candetect as little as 1% of mutant DNA from a mixedsample. The assay was performed using 10 μL of PCRproducts (215-bp fragment) according to the manufac-turer's instructions (TrimGen, Sparks, MD) [24]. Positiveand negative controls were used as mentioned for PCRanalysis.

Statistical Analysis

Patients' clinico-pathological parameters and mutationstatus were resumed using mean and frequency. Correlation

Endocr Pathol (2012) 23:83–93 85

between BRAF mutation and clinico-pathological param-eters of the patients was determined by χ2 test andnonparametric two-tailed Fisher test and multivariatelogistic regression analysis. A univariate as well asmultivariate analysis using logistic regression wasperformed to assess the effect of clinico-pathologicalvariables on lymph node metastasis. All analyses wereperformed using Sigma Stat 3.5. P<0.05 was consideredas statistically significant and tests were two-tailed [25].

Results

BRAF Mutation in Thyroid Cancer Cell Lines

Direct sequencing of exon 15 amplicons from fivehuman thyroid cancer cell line DNA (Fig. 1) wasconfirmatory for their respective genotypes. As seen inthe electropherogram, NPA was homozygous for BRAFT1799A (V600E) mutation, while ARO and FRO celllines were heterozygous. This missense mutation wasabsent in WRO and TT cell lines which are homozygousfor wild-type allele. All these cell lines (Table 1) wereused as controls for further experiments.

BRAF Mutation in Thyroid Tumor Samples

BRAFV600E mutation was determined primarily by MASAassay, Mutector assay, DNA sequencing, and SSCP in 25samples. Results obtained by all three techniques werecomparable. Mutation detected in all the positive samplesby MASA (Fig. 2a) involved a T>A transversion atnucleotide 1799 (V600E) and was heterozygous as shownby DNA sequencing (Fig. 3b). Mutector assay (Fig. 3a)showed high sensitivity for the mutation allowing quanti-tative detection of percent mutant template DNA in samples(data not shown). It is however not cost effective forscreening large number of samples by this method. Thus,MASA-PCR which is specific, reproducible, and alsoeconomic was found to be most suitable for screening all

of the 140 tumors along with adjacent normal tissue. Fiftypercent of all the samples were randomly selected for DNAsequencing. SSCP analysis (Fig. 3) corroborated with theabove findings but was found to be inconsistent amongsamples.

In summary, BRAF T1799A (V600E) mutation wasfound in 46 out of 86 PTC patients (53.4%) of Indiansubcontinental cohort. Percent incidence of this mutationvaried among pathological subtypes of PTC (Table 2).BRAF gene mutation was much more common in theconventional PTC (38 out of 62; 61.3%) than in thefollicular variants of PTC (2 out of 17; 11.7%). None ofthe 54 non-PTC thyroid tumors including 8 FTA, 14FTCs, 16 MTCs, and 16 benign samples showed anymutation at this nucleotide position in exon 15. Also,normal thyroid tissue samples from the surroundings ofBRAF mutation-positive malignant thyroid areas showedno mutations of BRAF thereby suggesting that themutations were somatically acquired.

Correlation with Clinico-Pathological Parameters

The correlation between BRAF mutation and variousclinico-pathological parameters were assessed in 86patients with PTC (Table 3). By univariate analysis,BRAF mutation status showed significant association withconventional type of PTC (P<0.011), TNM stage (P<0.005),extra-thyroidal extension (P<0.005) and Lymph nodemetastasis (P<0.005). There was no significant correlationbetween BRAF mutation and sex, age, or tumor size atdiagnosis. A multivariate analysis was performed with thesame clinico-pathological parameters. Only extra-thyroidalextension (odds ratio=9.1; 95% confidence interval, 2.75–30.4; P<0.005) and lymph node metastasis (odds ratio=3.5;95% confidence interval, 1.12–11; P<0.031) were correlatedwith BRAF mutation with and without stepwise multiplelogistic regression method.

All the tumors were divided into morphological sub-types, wherein conventional (classical) variant of PTC wasthe largest group. Correlation of BRAF mutation with the

Fig. 1 BRAFV600E mutationin thyroid cancer cell linesas demonstrated by DNAsequencing. The sequencesshow the wild-type BRAFgene in WRO and TT,homozygous BRAF mutationT1799A (V600E) in NPAand heterozygous BRAFmutation T1799A in AROand FRO

86 Endocr Pathol (2012) 23:83–93

same parameters was sought within this group. But nosignificant correlation was found either by univariate or bymultivariate analysis.

To understand the prognostic significance of clinico-pathologic factors in determining lymph node metastasis,linear, and multiple regression analyses were performed. Inunivariate analysis, tumor size, tumor stage, extra-thyroidalextension (ETE), and BRAF mutation were significantpredictive factors for lymph node metastasis (Table 4). Butin multivariate analysis tumor stage and BRAF mutationwere the only significant prognostic factor for lymph nodemetastasis (odds ratio=12.2; 95% confidence interval, 4.1–35.6; P<0.001; and odds ratio=4.9; 95% confidenceinterval, 1.5–15.5; P<0.006; respectively; Table 4).

In summary, our results predict that BRAF gene mutationof PTC is correlated with lymph node metastasis andETE by both univariate and multivariate analysis; but itis not an independent risk indicator within the classicalsubtype, as observed by other investigators. Tumor stageand BRAF gene mutation were the only significant

prognostic factors for lymph node metastasis in PTC.However, BRAF mutation status is not independent fromPTC subtypes in predicting extra-thyroid extension andlymph node metastasis [25].

Pathological Findings

Among the 86 heterogeneous entities of PTC, conventionalsubtype was predominating over other types. In general BRAFmutation, status was associated with certain clinical featureslike older patients, larger tumors, solid tumors, higher rate ofnode metastasis, and ETE. The presence of BRAF mutationalso correlated well with pathologic features like papillarypatterns, characteristic nuclear features, intra-thyroidal inva-sion, sclerotic stroma, etc. It was seen more commonly inconventional types and all cases of tall-cell variant (3/3).BRAF negativity was associated more frequently with youngerpatients, smaller tumors, cystic tumors, lower rate of lymphnode metastasis, ETE, non-papillary patterns like follicularand solid pattern circumscribed border, subtle nuclear features,

Fig. 2 BRAF mutation inthyroid tumors. (a) MASAanalysis of BRAF mutation.Genomic DNA extracted fromtumor tissue was analyzed forBRAF mutation by MASA-PCRas described in Materials andMethods. PCR products wereanalyzed in a 2% agarose gel.Genomic DNA extracted fromWRO and ARO thyroidcancer cell lines were used asnegative and positive control,respectively.(b) SSCP patternin thyroid tumors showingWRO (BRAFT1799WT),ARO (BRAF T1799Atransversion mutation) andDNA samples from thyroidcancer patients. SamplesP12 and P46 show wild-typeBRAF whereas samples P21and P71 exhibit mutated BRAF.Notice the abnormalsingle-strand and hetero-duplexpattern in the mutated samples(arrows)

Endocr Pathol (2012) 23:83–93 87

and lack of sclerotic stroma. This group included mainlyfollicular variant of PTC (FVPTC) and other tumor types.

Both cases of poorly differentiated thyroid carcinoma(PDTC) were of large cell type. Both patients were male.On histological examination, the tumor showed areas ofPTC with vacant nuclei and nuclear grooves (Fig. 4). BRAF

mutation was present in both patients, indicating theprogression of PTC towards PDTC. Three cases werethoroughly examined for distant metastasis. All the caseswere having lymph node metastasis and ETE and one ofthem was PDTC. Two out of three distant metastasispatients were BRAF mutated [26].

Fig. 3 BRAF mutation in thyroid tumors. (a) Results of Mutectorassay by BRAF codon 600 mutations detection kit; TrimGen, Sparks,MD. Wells with color reaction (green) represent samples positive forT1799A BRAF mutation. NTC is no-template control, NPA thePapillary thyroid cancer cell line served as the positive control, andWRO the follicular thyroid cancer cell line served as the negative

control. The representative positive and negative thyroid tumorsamples were P#21, P#71, and P#46, respectively. (b) Thecorresponding sequence chromatographs of tumor P#21, tumor P#71and NPA, with BRAF T1799A mutation and normal tissue P#21,P#71, and WRO with wild-type BRAF T1799

Table 2 Prevalence ofBRAFV600E mutation in 140thyroid neoplasia patients

Histology Total BRAFV600E

Positive Negative

Number Percentage, % Number Percentage, %

Papillary carcinoma 86 46 53.4 40 46.6

Classical variant 62 38 61.3 24 38.7

Poorly differentiated 2 2 100 0 0

Follicular variant of PTC 17 2 11.7 15 88.2

Tall cell variant 3 3 100 0 –

Hürthle cell carcinoma 1 1 – 0 –

Diffuse sclerosing variant 1 0 – 1 –

Follicular carcinoma 14 0 – 14 –

Follicular adenoma 8 0 – 8 –

Medullary carcinoma 16 0 – 16 –

Benign hyperplasia 16 0 – 16 –

88 Endocr Pathol (2012) 23:83–93

However, in summarizing statistical analysis andmorphological findings, our results indicate that BRAFmutational status, though associated strongly with distinctclinical, pathologic, and biologic variables, it is not anindependent prognostic indicator “per se” in PTC.Morphologic typing appears to be mandatory prognosticindicator in PTC.

Discussion

Since the initial discovery of BRAF mutation in humancancers [8], there have been more than 40 mutationsidentified in the BRAF gene. Among these mutations, theT1799A point mutation is the most common and accounts

for more than 90% of all the mutations found in the BRAFgene [27].

BRAFV600E mutation was reported in thyroid carcinomainitially by Kimura et al. 2003, and since then, significantprogress has been made in understanding its tumorigenicrole and clinical importance in this disease [17, 18]. It hasbeen shown that BRAF plays critical role as intermediate ofERK phosphorylation in MAPK pathway. This pathwayacts in parallel with TSH receptor-cAMP-PKA-CREBpathway for subsequent cellular proliferation and differen-tiation [28]. There is a crosstalk between these twopathways, either agonistic or antagonistic, depending onthe cell type [29].

This activating mutation has been specifically reportedin PTC with a frequency ranging from 29% to 83% in

Table 3 Correlation between BRAFV600E mutation and clinico-pathological parameters in papillary thyroid carcinoma

BRAFV600E χ2 test Multivariate analysis

Positive Negative P valueb P value Odds ratio (95% CI)

Numbera Percentage, % Number Percentage, %

Tumor stage (PTC) <0.005 0.381 1.4 (0.61–3.5)

T I 2/18 11.1 16 88.9

T II 10/22 45.4 12 54.6

T III–IV 34/46 73.9 12 26.1

Gender (PTC) <0.65 0.073 0.31 (0.08–1.1)

Male 16/32 50.0 16 50.0

Female 30/54 55.5 24 44.5

Age (PTC)c <0.085 0.11 2.6 (0.8–8.5)

≥45 years 24/38 63.1 14 36.8

<45 years 21/48 43.7 27 56.2

Node metastasis (PTC) <0.005 0.031 3.5 (1.1–11.4)

Yes 35/47 74.4 12 25.6

No 11/39 28.2 28 71.8

Tumor size (mm)c <0.068 0.94 0.96 (0.35–2.6)

Range, mean±SD

<10 mm 3/12 25.0 9 75.0

10–40 mm 32/52 61.5 20 38.5

>40 mm 11/22 50.0 11 50.5

Distant metastasis 2/3 66.6 1 33.4

Extra-thyroidal extension <0.005 <0.005 9.1 (2.7–30.4)

Yes 33/41 80.5 8 19.5

No 13/45 28.9 32 72.1

Tumor type 0.011 0.098 3.1 (0.82–12.2)

Conventional 40/65 61.5 25 38.4

Non-conventional 6/21 28.5 15 7.1

a Data are presented as number/total number of patients, with percentages in parenthesesb Clinico-pathological parameters are compared by χ2 testc Age and tumor size were assessed as continuous variables

Endocr Pathol (2012) 23:83–93 89

several thyroid tumor cohorts depending on the epidemio-logical factors, heterogeneity of the histological variants, orthe age group analyzed [9, 10]. Recently Mathur et al. 2011have reported much higher rate (88%) of BRAF mutation inPTC which increased significantly over a 15-year period atthe authors' institution [26]. The findings suggested that ahigher rate of BRAF mutation in papillary thyroid cancermay contribute to the increasing incidence of thyroidcancer. Xing et al. 2005 have shown that this mutation isassociated with a poorer clinico-pathological outcome andis a novel independent molecular prognostic marker in therisk evaluation of thyroid cancer [16].

In India 3–4% of cancer patients suffer from thyroidneoplasia [20]. To the best of our knowledge, there is nopublished data on the prevalence of BRAF mutation inIndian subcontinental population. In this article, we reportthe frequent occurrence of BRAF mutation in sporadicPTCs, mainly in the classic variant of PTC, for the firsttime in the concerned population and confirm that thismutation is strongly associated with the papillary growthpattern.

In the present study, 140 patients of thyroid carcinomawere included from different parts of the country.BRAFV600E mutation was found in 46 out of 86 PTCpatients (53.4%), in comparison with observations fromother Asian countries, Japan 36%, Taiwan 46.6%, andUkraine 22.9% as quoted by Fugazzola et al. [30]. Wedid not find the concerned mutation in the adjacent normalthyroid tissues of those patients with BRAFV600E-positivetumor samples, which is suggestive of its somatic origin[8, 14].

In our study, both the poorly differentiated carcinoma(PDTC) samples with large cell component were positivefor BRAFV600E. The tumors contained areas of PTC withvacant nuclei and nuclear grooves. Similar finding werereported by Nikiforova et.al 2003 and Soares et al. 2004[32, 33]. According to our observation two out of threedistant metastasis cases were harboring BRAF mutationwith one of them being PDTC. In a study, reported byFugazzola et.al, 2006 none of the ten poorly differentiatedtumors were positive for BRAF mutation, indicating theabsence of BRAF mutation in poorly differentiated thyroidcarcinoma when not associated with a well-differentiatedcomponent; which is in contrary to our observation [26].The possibility that BRAF mutation reduces the risk oftumor progression is in conflict [32, 33] and reason forthese discrepancies is not clear. As suggested by Puxeddu etal. 2004, it might be related to differences in cancerselection or to the low number of cases studied in eachtumor cohort, due to unavailability of samples [34].

As far as the correlation of BRAF mutation and clinicalfeatures of PTC patients are concerned, we did not find

Table 4 Prognostic factors forlymph node metastasis

aAge and tumor size wereassessed as continuous variablesfor both univariate and multi-variate analysis

LN metastasis Univariate analysis Multivariate analysis

Yes No P value Odds ratio P value Odds ratio

Agea 0.81 1.1 (0.4–2.67) 0.22 0.97 (0.94–1.0)

Gender 0.82 1.1 (0.5–2.7) 0.72 1.2 (0.4–3.6)

Female 29 25

Male 18 14

Tumor sizea 0.037 1.3 (1.0–1.7) 0.50 1.3 (1.0–1.9)

TNM stage 0.001 12.2 (4.1–35.6) 0.001 12.2 (4.1–35.6)

I 0 18

II 1 21

III and IV 46 0

ETE 0.001 5.6 (2.1–14.3) 0.052 3.1 (0.99–9.8)

Yes 31 10

No 16 29

BRAF V600E 0.001 7.4 (2.8–19.3) 0.006 4.9 (1.5–15.5)

Positive 35 12

Negative 11 28

90 Endocr Pathol (2012) 23:83–93

The distribution of BRAF mutation in PTC showed aclear subtype-related pattern. We found the significantprevalence of BRAF mutation in classical PTC (64%).Whereas in FVPTC, only 11% harbored BRAF mutation.This percentage of prevalence correlated well with theaverage prevalence of 60% and 12% incidence of BRAF inclassical PTC and FVPTC as observed from the ninereports summarized in the review by Xing 2005 [7]. Also,all the tall-cell variants, which were more aggressive innature, harbored BRAF mutation as expected [7, 31].

significant correlation between BRAF mutation andgender, age, tumor size at diagnosis. Wherein we foundsignificant correlation of BRAF mutation with nodalmetastasis (P<0.005), extra thyroidal invasion (P<0.005),and tumor stage (P<0.005) which is in agreement with

the observation reported by Frasca et al. 2008 in a series of323 PTC patients indicating that the presence of BRAFmutation in PTC is associated with aggressive tumorbehavior [19]. In multivariate analysis, BRAF mutationcorrelated only with ETE and lymph node metastasis. In

Fig. 4 Pathological variants ofPTC. (a) Conventional PTCwith characteristic nuclearfeature; (b) FVPTC withcircumscribed border (capsular);(c) tall-cell variant of PTCwith compactly arrangedpapillae and dense eosinophiliccytoplasm; (d) Poorlydifferentiated PTC with vacantnucleus and nuclear grooves;(e) Hürthle cell carcinoma;(f) Diffuse sclerosing variant ofPTC; (g) Lymph node metastasisin PTC; (h) ETE with cysticchanges in PTC

Endocr Pathol (2012) 23:83–93 91

another study on 500 cases of PTC in a homogenousItalian cohort from a single institution, Lupi et al. 2007demonstrated a strong association of BRAF mutation withextra-thyroidal invasion, lymph node metastasis, andadvanced tumor stages [35]. However, we could find nocorrelation of BRAF with any of these clinico-pathologicalparameters, when performed within the classical orfollicular subtype of PTC. In our consecutive series of86 patients, tumor stage was another prognostic factor,other than BRAF mutation, for lymph node metastasis.Similar to our finding, one large-scale study on 410 PTCpatients reported by Cheng S et al. 2011 indicated thatBRAF mutation cannot be considered as independent riskdeterminant within a particular subtypes PTC [25]. Thisrelationship is still controversial as it is observed in somestudies, but not in other studies. Different factors such asthe number of study subjects in a cohort, differentrecruitment and follow up procedures, and tumor classifi-cation may have been contributing factors making many ofthese studies less consistent and reliable [18]. Our resultspredict that BRAF gene mutation of PTC is correlated withlymph node metastasis and ETE by both univariate andmultivariate analysis; but it is not an independent riskindicator within the classical subtype.

We have not included studies on other genetic alterationssuch as RET/PTC and RAS in the present report. Howeverit has been reported that BRAFV600E occurs mutuallyexclusively and its activation may be responsible forprogression to classic PTC [21, 31, 35, 36]. According toour study, all the PDTC, Hürthle cell, tall-cell variant, andmost of the distant metastatic patients were BRAF positive,which suggests that this mutation may be involved inthyroid cancer progression to poorly differentiated andaggressive phenotypes [37]. The results from the transgenicmouse studies also revealed that BRAF mutation initiateddevelopment of PTC and its transition to anaplastic thyroidcancer [38]. Consistent with other observations, we did notfind any patient with FTC, FTA, MTC, or benignhyperplasia positive for BRAF mutation [26, 32].

In conclusion, this study was undertaken to find outprevalence of BRAFV600E in Indian population and itsassociation with clinico-pathological parameters. We found53% of PTCs harbored BRAF mutation which is wellwithin the range (29–83%) reported in the literature. Thisgenetic event is predominantly associated with the classicvariant of PTC also in Indian subjects and correlatedsignificantly with aggressive features among all theclinico-pathological parameters that we studied. However,statistical analysis and morphological findings, in com-bination, suggests that BRAF is not an independentprognostic indicator “per se” in PTC.

Thus our findings along with those reported in theliterature suggest that morphologic typing in combina-

tion with BRAF mutation status appears to be betterprognostic indicator in PTC. Relevantly, considering thehigher prevalence in Indian cohort, BRAF gene can be apromising target for small molecular inhibitors for betterprognosis of radio-iodine refractory thyroid carcinomapatients [39].

Acknowledgment The authors would like to thank Dr. GiorgioStassi (Universita Degli Studi Di Palermo, Palermo, Italy) for kindlyproviding the thyroid cancer cell lines and Dr. K. M. Mohandas (DeanAcademics and HOD DDCN, Tata Memorial Hospital, India) for kindsuggestions in stratification of tissue samples for analysis.

References

1. Gimm O, Thyroid Cancer. Cancer Letter 163:143 – 156, 20012. Hundahl SA, Fleming ID, Fremgen AM, Menck HR. A National

Cancer Data Base report on 53,856 cases of thyroid carcinomatreated in the USA, 1985–1995. Cancer 83:2638–2648, 1998

3. Paterson IC, Greenlee R, Adams Jones D. Thyroid cancer inWales, 1985–1996: a cancer registry-based study. Clin Oncol (RColl Radiol) 11:245–251, 1999

4. Nikiforov YE. RET/PTC rearrangement in thyroid tumors.Endocrine Pathology 13:3–16, 2002

5. Santoro M, Melillo RM, Carlomagno F, Fusco A & VecchioG. Molecular mechanisms of RET activation in human cancer.Annals of the New York Academy of Sciences 963: 116–121,2002

6. Giordano TJ, Kuick R, Thomas DG, Misek DE, Vinco M, SandersD, Zhu Z, Ciampi R, Roh M, Shedden K et al. Molecularclassification of papillary thyroid carcinoma: distinct BRAF,RAS, and RET/PTC mutation specific gene expression profilesdiscovered by DNA microarray analysis. Oncogene 24: 6646–6656, 2005

7. Xing M, BRAF Mutation in Thyroid Cancer. Endocr Rela Cancer12:245–262, 2005

8. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S,Teague J, Woffendin H, Garnett MJ, Bottomley W et al. Mutationsof the BRAF gene in human cancer. Nature (Lond) 417: 949–954,2002

9. Kim KH, Kang DW, Kim SH, Seong I O & Kang DY. Mutationsof the BRAF gene in papillary thyroid carcinoma in a Koreanpopulation. Yonsei Medical Journal 45:818–821, 2004

10. Namba H, Nakashima M, Hayashi T, Hayashida N, Maeda S,Rogounovitch TI, Ohtsuru A, Saenko VA, Kanematsu T &Yamashita S. Clinical Implication of hot spot BRAF mutation,V599E, in papillary thyroid cancers. J Clin Endo and Metab88:4393–4397, 2003

11. Xing M, Tufano RP, Tufaro AP, et al. Detection of BRAF mutationon fine needle aspiration biopsy specimens: a new diagnostictool for papillary thyroid cancer. J Clin Endo and Metab489:2867–2872, 2004

12. Jin L, Sebo TJ, Nakamura N, Qian X, Oliveira A, Majerus JA,Johnson MR, Lloyd RV. BRAF mutation analysis in fine needleaspiration (FNA) cytology of the thyroid. Diagn Mol Pathol15:136–143, 2006

13. Chung KW, Yang SK, Lee GK, Kim EY, Kwon S, Lee SH, Parkdo J, Lee HS, Cho BY, Lee ES, Kim SW. Detection ofBRAFV600E mutation on fine needle aspiration specimens ofthyroid nodule refines cyto-pathology diagnosis, especially inBRAF600E mutation-prevalent area. Clin Endocrinol (Oxf)65:660–666, 2006

92 Endocr Pathol (2012) 23:83–93

14. Kebebew E, Weng J, Bauer J, Ranvier G, Clark OH, Duh QY,Shibru D, Bastian B, Griffin A. The prevalence and prognosticvalue of BRAF mutation in thyroid cancer. Ann Surg 246:466–471, 2007

15. Kim TY, Kim WB, Rhee YS, Song JY, Kim JM, Gong G, Lee S,Kim SY, Kim SC, Hong SJ, ShongYK. The BRAF mutation isuseful for prediction of clinical recurrence in low-risk patientswith conventional papillary thyroid carcinoma. Clin Endocrinol(Oxf) 65:364–368, 2006

16. Xing M, Westra WH, Tufano RP, Cohen Y, Rosenbaum E,Rhoden KJ et al. BRAF mutation predicts a poorer clinicalprognosis for papillary thyroid cancer. J Clin Endocrinol Metab90:6373–6379, 2005

17. Lee JH, Lee ES & Kim YS. Clinicopathologic significance ofBRAF V600E mutation in papillary carcinomas of the thyroid: ameta-analysis. Cancer 110:38–46, 2007

18. Xing M. BRAF mutation in papillary thyroid cancer: pathogenicrole, molecular bases, and clinical implications. EndocrineReviews 28 742–762, 2007

19. Frasca F, Nucera C, Pellegriti G et al. BRAF (V600E) mutationand the biology of papillary thyroid cancer. Endo Rel Can15:191–205, 2008

20. Yeole BB. Descriptive epidemiology of thyroid cancer in greaterBombay. Indian J Cancer 35(2):57–64, 1998

21. Kimura ET, Nikiforova MN, Zhu Z et al. High Prevalence of BRAFmutations in thyroid cancer: genetic evidence for constitutiveactivation of the RET/PTC-RAS-BRAF signaling pathway inpapillary thyroid carcinoma. Can Res 63:1454–1457, 2003

22. Sapio MR, Posca D, Troncone G et al. Detection of BRAFmutation in thyroid papillary carcinomas by mutant allele-specificPCR amplification (MASA). Eur J of Endocrinol 154:341–348,2006

23. Saldanha G, Purnell D, Fletcher A et al. High BRAF mutationfrequency does not characterize all melanocytic tumor types. Int JCancer 111:705–710, 2004

24. Cohen NG, Cohen Y, Rosenbaum E, et al. T1799A BRAFMutations in Conjunctival Melanocytic Lesions. InvestigativeOphthalmology & Visual Science 46: 3027–3032, 2005

25. Cheng S, Serra S, Mercado M, Ezzat S, Asa SL. A high-throughput proteomic approach provides distinct signatures forthyroid cancer behavior. Clin Cancer Res 17(8):2385–94, 2011

26. Mathur A, Moses W, Rahbari R, Khanafshar E, Duh QY, Clark O,Kebebew E. Higher rate of BRAF mutation in papillary thyroidcancer over time: A single-institution study. Cancer 117(19):4390–5, 2011

27. Garnett MJ, Marais R. Guilty as charged: B-RAF is a humanoncogene. Cancer Cell 6:313–319, 2004

28. Kondo T, Ezzat S and Asa SL. Pathogenetic mechanisms inthyroid follicular-cell neoplasia, Nature Reviews 6:292–306,2006

29. Vaillancourt RR, Gardner AM AND Johnson GL. B-Raf-Dependent Regulation of the MEK-1/Mitogen-Activated ProteinKinase Pathway in PC12 Cells and Regulation by Cyclic AMP.Mol and Cell Bio 14: 6522–6530, 1994

30. Fugazzola L, Puxeddu E, Avenia N et al. Correlation between B-RAFV600E mutation and clinico–pathologic parameters in papillarythyroid carcinoma: data from a multicentric Italian study and reviewof the literature. Endo Rel Can 13:455–464, 2006

31. Frattini M, Ferrario C, Bressan P, Balestra D, De Cecco L,Mondellini P, Bongarzone I, Collini P, Gariboldi M, Pilotti S et al.Alternative mutations of BRAF, RET and NTRK1 are associatedwith similar but distinct gene expression patterns in papillarythyroid cancer. Oncogene 23:7436–7740, 2004

32. Nikiforova MN, Kimura ET, Gandhi M, et al. BRAF mutations inthyroid tumors are restricted to papillary carcinomas andanaplastic or poorly differentiated carcinomas arising frompapillary carcinomas. J Clin Endo & Metab 88 (11): 5399–5404,2003

33. Soares P, Trovisco V, Rocha AS, Feijao T, Rebocho AP et al.BRAF mutations typical of papillary thyroid carcinoma aremore frequently detected in undifferentiated than in insular andinsular-like poorly differentiated carcinomas. Virchows Archiv444 572–576, 2004

34. Puxeddu E, Moretti S, Elisei R, et al. BRAFV599E mutation isthe leading genetic event in adult sporadic papillary thyroidcarcinomas. J Clin Endo & Metab 89(5):2414–2420, 2004

35. Soares P, Trovisco V, Rocha AS, Lima J, Castro P et al. BRAFmutations and RET/PTC rearrangements are alternative events inthe etiopathogenesis of PTC. Oncogene 22:4578–4580, 2003

36. Nikiforova MN, Ciampi R, Salvatore G, Santoro M, Gandhi M, etal. Low prevalence of BRAF mutations in radiation-inducedthyroid tumors in contrast to sporadic papillary carcinomas.Cancer Letters 209:1–6, 2004

37. Michels JJ, Jacques M, Amar MH, Bardet S. Prevalence andprognostic significance of tall cell variant of papillary thyroidcarcinoma. Human Pathology 38, 212– 219, 2007

38. Knauf JA, Ma X, Smith EP, Zhang L, Mitsutake N et al. Targetedexpression of BRAFV600E in thyroid cells of transgenic miceresults in papillary thyroid cancers that undergo dedifferentiation.Cancer Res 65(10):4238–4245, 2005

39. King A J, Patrick D R, Batorsky R S et al. Demonstration of agenetic therapeutic index for tumors expressing oncogenicBRAF by the kinase inhibitor SB-590885. Cancer Res 66(23):11100–11105, 2006

Endocr Pathol (2012) 23:83–93 93