Background

Thyroid cancer is the most common malignant endocrine tumor. Its incidence has been considerably increasing over the last 4 decades [1, 2]. Thyroid cancer usually arises from follicular epithelial cell or parafollicular cells [3]. The latter is the cell of origin of medullary thyroid cancer [3]. Follicular cell-derived thyroid cancer is by far the most common type (95 %) and is classified to differentiated thyroid cancer (DTC), poorly differentiated (PDTC) and undifferentiated (anaplastic) subtypes [3]. DTC is further classified to papillary thyroid cancer (PTC) and follicular thyroid cancer (FTC) [28]. In Saudi Arabia, DTC is the fourth most common type of cancer in general and the second most common cancer in women [4]. Over the last decade, there has been a significant progress in our understanding of DTC pathogenesis [5]. Genetic alterations in genes of key cellular signaling pathways, particularly mitogen activated protein kinase (MAPK) and phosphatidylinositol-3-kinase (PI3K) signaling pathways are the main genetic mechanisms of thyroid carcinogenesis [5, 6].

PTC is the most common subtype of thyroid cancer accounting for about 90–95 % of cases. BRAF V600E mutation is the most prevalent genetic alteration in PTC (~45 %) [5]. This transversion mutation results in thymine-to-adenine (T > A) change at nucleotide position 1799 (T1799A) substituting valine by glutamic acid at amino acid position 600 (V600E). When present, this mutation results in constitutive activation of the BRAF protein with consequent activation of the downstream mediators [5, 7, 8].

A large number of studies from different ethnic backgrounds reported BRAF V600E mutation in 29–83 % of PTC [7]. These studies were mostly undertaken in patients with DTC of various ages. To our knowledge, no previous study reported BRAF V600E mutation in a pure adult population. The relationship of BRAF V600E mutation and age of the patient at diagnosis has been inconsistent. Some studies showed significant correlation [9, 10] while others didn’t demonstrate such a correlation [11]. We and others have found a much lower rate of BRAF V600E mutation in DTC from pediatric and adolescent patients [12, 13]. Furthermore, we also found a much lower rate of the recently described TERT mutations in pediatric compared to adult DTC [13]. We speculated that the adult and pediatric DTCs are likely to have different molecular signatures as suggested by these findings and also the significantly different clinical and pathological phenotype [14–16]. Saudi population is highly inbred with high rate of consanguinity [17, 18]. Although BRAF V600E is somatic in nature, one may wonder whether this high rate of consanguinity may have an impact on the prevalence of BRAF V600E mutation and prevalence of other non-spot BRAF mutations. Although BRAF V600E mutation has been previously reported from Saudi Arabian population of various ages [19, 20], it has not been investigated in purely adult population. In the course of this study, we also found a relatively high rate of non-hot spot BRAF mutations. Our aim in this study is to determine the prevalence of the BRAF V600E mutation and other non-conventional exon 15 BRAF mutations in pure adult patients (>18 years) with DTC in this highly inbred population.

Methods

Tumor samples and DNA extraction

We studied a sample consisting of 204 sporadic DTCs. The clinical and pathological data of these patients are summarized in Table 1. The only exclusion criterion was age at diagnosis of ≤18 years. DTC samples included 114 (55.9 %) conventional papillary thyroid cancer (CPTC), 55 (27 %) follicular variant papillary thyroid cancer (FVPTC), 29 (14.2 %) tall cell papillary thyroid cancer (TC-PTC), 3 (1.5 %) hurthle cell carcinomas (HCC), and 3 (1.5 %) diffuse sclerosing type papillary thyroid cancer (DSC) collected from the pathology Department of the King Faisal Specialist Hospital and Research Centre (KFSH&RC), Riyadh, Saudi Arabia (Table 1). The patients did not have family history of thyroid cancer. Histological diagnosis was confirmed by an experienced endocrine pathologist (H.A) who also carefully selected the tumor samples from paraffin blocks to ensure that tumor tissue was dissected for DNA extraction. Slices of 10-micron thickness of tumor samples were dissected from formalin fixed and paraffin embedded tissue (FFPT). Genomic DNA was extracted from the tumor tissues as previously described [21] using the Gentra Puregene DNA extraction kit (Qiagen, Valencia, CA).

Table 1 Clinical and histopathological characteristic features of 204 cases of adult DTC
Full size table

Ethical approval

This study was approved by the Ethical Committee and Institutional Review Board of King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia (RAC-2130015).

PCR amplification and sequencing

Exon 15 of the BRAF gene was PCR amplified using genomic DNA from the tumor samples. The forward and reverse primers and the PCR conditions were as described previously [22]. We limited our search to exon 15 as the previously described mutations were mostly described in this exon. The PCR amplicons were confirmed on 2 % agarose gel and directly sequenced using the Big Dye terminator v3.1 cycle sequencing ready reaction kit (Applied Biosystems) and by ABI PRISM 3730X1, genetic analyzer (Applied Biosystems). Identified mutations were confirmed in both forward and reverse directions by an independent PCR amplification and sequencing reaction. For the confirmed mutations, matched normal tissue samples were further analyzed to check whether the mutations are somatic or germline in nature. The sequencing results were read against the BRAF gene (GeneBank No.: NM_004333.4).

Statistical analysis

Statistical analyses were performed using SPSS (The Statistical Package for Social Sciences) version 20 software (IBM Corp., NY USA). Data were expressed as median and range for numerical values and numbers and percentages for categorical data. The T test was used to analyze continuous variables and Fisher exact and χ2 tests were used for categorical data. In all analyses, a two-tailed P < 0.05 was considered statistically significant.

Availability of data and supporting materials

Paraffin embedded tumor tissues, genomic DNAs and data of tumor molecular genetics. The SPSS (The Statistical Package for Social Sciences) version 20 software was obtained from IBM Corp., NY USA. The protein molecules can be downloaded from http://www.rcsb.org/pdb/home/home.do using the PDB ID and can be viewed using SWISS PDB and Jmol viewer downloaded freely from http://www.spdbv.vital-it.ch/ and http://www.jmol.sourceforge.net/, respectively.

Results

Prevalence of BRAF mutation and identification of rare, non-hotspot BRAF mutation in adult DTCs

Of 204 DTCs, 99 samples (48.5 %) harbored a BRAF mutation (Table 2). The hotspot BRAF V600E mutation was found in 95 cases (46.5 %). As shown in Table 3, four samples pertaining to 4 unrelated patients harbored other non-hotspot BRAF mutations. One sample harbored C > T transition mutation at nucleotide position 1796 resulting in a T599I amino acid change. The tumor was a follicular variant papillary thyroid cancer (FV-PTC) from a female patient. The other sample harbored a very rare 3-base insertion mutation at nucleotide position 1798 resulting in duplication of threonine 599 (T599dup) and this mutation was identified in a tumor from a 44-year old lady with a 3-cm FV-PTC without extrathyroidal invasion, lymph node or distant metastases. Two samples, one in a 30-year old lady with CPTC without extrathyroidal invasion, lymph node or distant metastases and the other was in a 30-year old lady with 1 cm FV-PTC without extrathyroidal extension, lymph node or distant metastases, harbored an A > G point mutation at nucleotide position 1801 resulting in a K601E amino acid change (Fig. 1). These mutations were confirmed to be somatic in nature as they were absent in matched normal tissue samples. These three non-hot spot mutations have already been recorded in the COSMIC (Catalogue of Somatic Mutations in Cancer) database, UK.

Table 2 Summary of exon 15 BRAF mutations identified in 204 cases of adult DTC
Full size table
Table 3 Various non-hotspot BRAF mutations identified in four out of 204 cases of adult DTC
Full size table
Fig. 1
figure 1

Identification of BRAF mutations in adult differentiated thyroid cancer. a The chromatogram shows a rare, somatic, non-hotspot, insertion mutation of BRAF gene in exon 15 at nucleotide position 1798 [1798insTAC (T599dup)]. b Shows a rare, somatic, non-hotspot, point mutation of BRAF gene in exon 15 at nucleotide position 1796 [C1796T (T599I)]. c A representative sequencing results shows a rare, somatic, non-hot spot, point mutation of BRAF gene in exon 15 at nucleotide position 1801 [A1801G (K601E)] from two independent tumor samples. In all the above cases, sequencing results of the matched normal tissue of each sample are shown in the left side of the panel. All the samples were repeated by independent PCR reactions with forward and reversed sequencing

Full size image

Association of BRAF V600E mutation with clinicopathological features and outcome of DTC

Patients included in this study were all adults and the mean ± SD age was 43.5 ± 15 years in patients with BRAF V600E mutation and 36.2 ± 13.3 years in patients with wild type BRAF (P 0.0001). Extrathyroidal invasion was more frequently observed in tumors with BRAF V600E mutation occurring in 54 tumors (56.8 %) compared to only 41 (39 %) tumors with wild type BRAF (P = 0.017). Lymph node metastasis was more commonly found in 51 tumors (53.7 %) with BRAF V600E mutation compared to 40 tumors (38.1 %) with wild type BRAF (P = 0.038). Similarly, BRAF V600E mutation was more common in higher stage tumors (stage III/IV) occurring in 31 patients (32.6 %) compared with only 13 (12.4 %) in the wild type BRAF (P = 0.001). BRAF V600E mutation was not associated with gender, tumor size, tumor multifocality, distant-metastasis, and persistent/recurrent disease at 6–12 months after initial management and at the last follow up (Table 4).

Table 4 Comparison of clinical and histopathological characteristics between patients with positive and negative BRAF V600E mutation in adult DTCs
Full size table

Discussion

In this study, we analyzed exclusively a sample of adult DTC patients for BRAF mutations as it has been reported that adult DTCs harbor higher frequencies of BRAF mutations than pediatric DTC [13, 23, 24] and previous data showed significant differences in the clinical, pathological and molecular basis of pediatric compared with adult DTCs [15, 16, 25–28]. We report here 48.5 % (99/204) of BRAF mutations in adult DTC from Saudi Arabia. As expected, the vast majority of the identified mutations were the well described mutation, BRAF V600E (95/204, 46.5 %). In addition, we also found three rare, non-hotspot BRAF mutations (T599I, T599dup and K601E) in four cases (4/204, 2 %). Although these non-hot spot mutations were previously described, their rate in this study seems higher than expected. Consanguinity is common in Saudi Arabia [17, 18] and although these mutations are somatic in nature, it is possible that there might be a genetic predisposition to their occurrence as inheritance of certain SNPs in the DNA repairing machinery genes may theoretically predispose to the development of BRAF and other driver mutations in PTC.

Detailed analyses of the previously reported BRAF mutations in PTCs showed that in addition to the most common BRAF V600E mutation, many unique and rare non-hotspot mutations have been detected in the exon 15 of the BRAF gene and these mutations have recently been compiled in a report [29]. In our study, we found three rare, non-hotspot BRAF mutations in four cases. In a tumor from a 47-year old female with FV-PTC, we found the rare non-hotspot missense T599I mutation. This mutation has also been previously reported in a follicular variant [30] and in a solid variant [31] PTC. In both cases, the T599I mutation was originally identified as a concomitant mutation with a complex BRAF mutation V600delinsAL (replacement of valine with an alanine and a leucine) in the same allele, and VKSRdel (deletion of valine 600, lysine 601, serine 602, and arginine 603), respectively [30, 31]. However, we did not find any co-existing mutation with the T599I mutation in our patient’s tumor sample. Previously reported functional analysis of this mutation revealed that the T599I mutant had a moderately increased kinase activity but with dramatically increased phosphorylation and activation of ERK [8]. This phenomenon might be related to the effect of dimerization which is likely achieved via complexing with wild-type BRAF and RAF1 [32]. As illustrated in Fig. 2, this mutation is located in the core region of the kinase domain and more specifically in the activation loop that is encoded by the nucleotides of the BRAF exon 15. Moreover, the rare non-hotspot mutation T599I resides in one of the two critical phosphorylation sites (T599 and S602) in the activation loop. As BRAF is a serine/threonine kinase, it has been hypothesized that phosphorylation of Threonine 599 and Serine 602 residues disturb the hydrophobic interaction of P-loop (phosphate binding loop) with the A-loop (activation loop) [8]. Therefore, it is likely that phosphorylating either the S/T residues would substantially result in destabilization of inactive BRAF protein and constitutively switching to active conformation. In a previous report from Saudi Arabia, deletion of threonine at this position (T599del) was described in 1 out of 69 benign follicular adenomas, and 1 of 115 PTC harbored an insertion of additional threonine at this position (T599dup) [20]. This latter mutation was also detected in another patient from our series. This mutation was previously described in a pilocytic astrocytoma [33] and in anaplastic thyroid cancer (ATC) samples with tall cell like phenotypes [34, 35]. It has been demonstrated that the T599dup displays an increased kinase activity and enhanced MEK phosphorylation followed by ERK activation and potentially comparable to those of the hotspot BRAF V600E mutation [33]. In addition, this insertion mutation has also been shown to transform the NIH3T3 and MCF-10A cells in culture. It has been speculated that insertion of an additional amino acid at this T599 residue of the protein backbone destabilizes the inactive conformation of the kinase domain which may result in its conversion to an active mutant BRAF [33].

Fig. 2
figure 2

Schematic diagram of the BRAF. a Schematic diagram of the BRAF shows the rare, somatic, non-hot spot, mutations of BRAF gene in exon 15 at various nucleotide positions, numbers in the boxes indicate exons and boxes indicate various domains: RBD-ras binding domain, CRD-cysteine rich domain, KD-kinase domain, L–G loop, a conserved glycine motif. b A diagram of native modeled structure of BRAF homodimer. Each monomer molecule is indicated in different color (red and green). The structure of BRAF protein in complex PDB ID is 4E26. Amino acid residues of rare, non-hotspot, somatic mutation identified in adult differentiated thyroid cancer are plotted in BRAF native protein structure using SWISS PDB viewer and the mutated residues are shown in sphere shape. c Shows the monomer cartoon diagram of native modeled structure of BRAF kinase domain. The structure of BRAF protein kinase domain (monomer) PDB ID is 4WO5. Amino acid residues of rare, non-hotspot, somatic mutation identified in adult differentiated thyroid cancer are plotted in BRAF native protein structure and the molecule is visualized by JSmol

Full size image

We also identified another rare non-hotspot mutation (K601E) in two tumor samples of adult DTC, one from a CPTC and the other from a FV-PTC. This mutation has been well documented in DTC and is considered the second most common BRAF mutation in DTC after BRAF V600E. Furthermore, it also has been found to be associated with tumor cells with histological features of FVPTC [10], more specifically in the encapsulated follicular variant of PTC. Recently, it has been shown that this K601E mutation may also be found rarely in other sub-types of thyroid cancer including follicular thyroid carcinoma (FTC) [36]. In the study by Schulten et al. this mutation was found in three cases, an FTC, a PTC and a FV-PTC [20]. Transient transfection mediated functional characterization of this mutation in HEK 293T cells showed enhanced kinase activity resulting in subsequent strong MEK phosphorylation and activation of ERK1/2 kinases in the absence of TSH [30] suggesting that this is a typical gain-of-function mutation and likely to drive cell proliferation and thyroid carcinogenesis in a kinase dependent fashion. We found that all of these three rare non-hotspot mutations were in cases of FV-PTC and the K601E was also found in one case of CPTC. Interestingly, the histopathological features and the outcome of patients with those mutations were favorable. In all cases, the tumor size was relatively small (1–3 cm) and there was no evidence of extrathyroidal invasion, lymph node or distant metastases. All patients were in remission after the initial management.

Apart from these rare non-hot spot BRAF mutations, BRAF V600E was common in this series of adult DTCs. Similar to previous studies, we found a statistically significant association between the BRAF V600E mutation and age (P < 0.0001), extrathyroidal invasion (P = 0.017), lymph node metastasis (P = 0.038) and TNM stage III/IV (P = 0.001) but not with other clinical and histopathological features such as gender, tumor size, tumor multifocality, distant metastasis, persistent and persistent/recurrent disease. This might be due to the relatively small sample size. To the best of our knowledge, this study is the first to determine the frequency of BRAF mutation exclusively in a pure sample of adult DTC patients. The previously reported studies were a mixture of adult and pediatric cases [19, 20]. BRAF V600E mutation has been detected with a frequency ranging between ~30–85 % in classical PTC. On the other hand, in case of FVPTC, the frequency ranges between ~0 and 35 % [7, 11, 37]. The frequency of the BRAF mutations identified in our study is comparable to that of the other studies in the Middle East region [19, 20, 38].

Conclusions

We have reported the frequency of exon 15 BRAF mutations (48.5 %) in a pure sample of adult DTC patients from an ethnically different population with a high rate of consanguinity and high rate of thyroid cancer. In addition to the common occurrence of BRAF V600E mutation, we describe a relatively high rate of non-hotspot BRAF mutations (T599I, T599dup and K601E). These mutations were detected in 4 out of 204 cases (2 %) of DTC in our study. Our results show a high rate and a strong prognostic role of the classical BRAF V600E mutation and also suggest a common occurrence of non-hot spot mutations in adult DTC from this highly inbred population.