Ann Surg Oncol DOI 10.1245/s10434-014-3723-5
ORIGINAL ARTICLE – TRANSLATIONAL RESEARCH AND BIOMARKERS
Significance of Allelic Percentage of BRAF c.1799T [ A (V600E) Mutation in Papillary Thyroid Carcinoma Shih-Ping Cheng, MD, PhD1,2,6, Yi-Chiung Hsu, PhD5, Chien-Liang Liu, MD1,2, Tsang-Pai Liu, MD1,2, Ming-Nan Chien, MD1,3, Tao-Yeuan Wang, MD1,4, and Jie-Jen Lee, MD, PhD1,2,6 Mackay Junior College of Medicine, Nursing, and Management, Taipei, Taiwan; 2Department of Surgery, Mackay Medical College and Mackay Memorial Hospital, Taipei, Taiwan; 3Division of Endocrinology and Metabolism, Department of Medicine, Mackay Medical College and Mackay Memorial Hospital, Taipei, Taiwan; 4Department of Pathology, Mackay Medical College and Mackay Memorial Hospital, Taipei, Taiwan; 5Institute of Statistical Science, Academia Sinica, Taipei, Taiwan; 6Graduate Institute of Medical Sciences and Department of Pharmacology, Taipei Medical University, Taipei, Taiwan 1
ABSTRACT Background. Somatic BRAF mutation is frequently observed in papillary thyroid carcinoma (PTC). Recent evidence suggests that PTCs are heterogeneous tumors containing a subclonal or oligoclonal occurrence of BRAF mutation. Conflicting results have been reported concerning the prognostic significance of the mutant allele frequency. Our present aim was to investigate the association between the percentage of BRAF c.1799T [ A (p.Val600Glu) alleles and clinicopathological parameters in PTC. Methods. Genomic DNA was extracted from fresh-frozen specimens obtained from 50 PTC patients undergoing total thyroidectomy. The BRAF mutation status was determined by Sanger sequencing. The percentage of mutant BRAF alleles was quantified by mass spectrometric genotyping, pyrosequencing, and competitive allele-specific TaqMan PCR (castPCR). Results. Positive rate of BRAF mutation was 72 % by Sanger sequencing, 82 % by mass spectrometric genotying, and 84 % by pyrosequencing or castPCR. The average percentage of mutant BRAF alleles was 22.5, 31, and 30.7 %, respectively.
Electronic supplementary material The online version of this article (doi:10.1245/s10434-014-3723-5) contains supplementary material, which is available to authorized users. Ó Society of Surgical Oncology 2014 First Received: 31 October 2013 J.-J. Lee, MD, PhD e-mail: [email protected]
There was a good correlation among three quantification methods (Spearman’s rho = 0.87–0.97; p \ 0.0001). The mutant allele frequency was significantly correlated with tumor size (rho = 0.47–0.52; p \ 0.01) and extrathyroidal invasion. The frequency showed no difference in pathological lymph node metastasis. Conclusions. The percentage of mutant BRAF alleles is positively associated with tumor burden and extrathyroidal invasion in PTC. Relatively good correlations exist among mass spectrometric genotyping, pyrosequencing, and castPCR in quantification of mutant BRAF allele frequency.
Papillary thyroid carcinoma (PTC) is the most common malignant tumor of the thyroid gland. Thyroid cancer incidence has increased substantially over the last decade, and this change is primarily attributed to an increase in PTC.1,2 Although most PTCs are curable or controllable by the combination of surgery, radioactive iodine ablation, and thyrotropin suppression, a considerable number of patients die of persistent or recurrent disease that is refractory to conventional therapy.3 With a growing understanding of molecular oncology, therapeutic agents that target activating genetic alterations in thyroid cancer have been rapidly developed.4 The v-raf murine sarcoma viral oncogene homolog B (BRAF) mutation occurs frequently in PTC.5 The most common BRAF mutation comprises a single base transversion at codon 600 in exon 15, identified as c.1799T [ A (p.Val600Glu) and generally referred to as V600E. The prevalence of BRAF mutation was found to be increased significantly over a 15-year period, and it was proposed
S.-P. Cheng et al.
that this change may contribute to the increasing incidence of thyroid cancer.6 BRAF c.1799T [ A mutation greatly increases the kinase activity and is associated with elevated transcriptional output of the mitogen-activated protein kinase (MAPK) pathway.7,8 Data from meta-analyses consistently indicate that the presence of BRAF mutation in thyroid cancer is associated with extrathyroidal invasion, lymph node metastasis, advanced TNM stage, and disease recurrence.5,9–11 If thyroid cancer cells harboring BRAF mutation are biologically more aggressive, tumors containing more BRAF mutationpositive cells will theoretically have a worse outcome. In this regard, Guerra and colleagues used pyrosequencing to quantify the percentage of mutant BRAF alleles, demonstrating that the mutant allele frequency correlated with age and tumor volume, and the odds ratio of recurrent disease was 5.31-fold higher in PTCs harboring a high ([30 %) percentage of mutant BRAF alleles.12 On the contrary, Gandolfi et al. 13 found no association between the percentage of mutant alleles and metastatic behavior. The reason for this discrepancy remains to be elucidated. Methods to detect BRAF mutation have not been standardized. Conventional DNA sequencing is binary (present or absent) and takes no account of the ratio of mutated DNA to wild type. Contemporary technological developments in genotyping are advantageous in terms of cost, throughput, scalability, and sensitivity.14 The primary aim of this study was to perform a multiplatform analysis to characterize the biological significance of the percentage of BRAF c.1799T [ A alleles in PTC. The secondary aim was to compare different methods for quantification of mutant BRAF allele frequency.
PATIENTS AND METHODS Patients and Tissue Samples This study was approved by the Institutional Review Board of Mackay Memorial Hospital (12MMHIS175). All patients provided written informed consent before the procurement of tissue specimens. A total of 50 patients who underwent total thyroidectomy for PTC were consecutively selected. Five patients who underwent lobectomy for follicular adenoma were also included as negative controls. All patients were euthyroid at the time of surgery and were not taking any medications that could affect serum thyrotropin levels. PTC tissues from the center of the lesions and corresponding normal thyroid tissues from the contralateral lobes of the same patient were obtained. All PTC tissue samples were carefully dissected to exclude surrounding normal tissue. Tissues were immediately snap-frozen in
liquid nitrogen and stored at -80 °C. The tissue diagnosis was confirmed by frozen section. Genomic DNA Extraction Genomic DNA was extracted using the QIAamp DNA mini kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. Serial dilutions of DNA from B-CPAP thyroid cancer cells (homozygous BRAF mutation ?/?) were used as positive controls.15 Sanger Sequencing Exon 15 of the BRAF gene was amplified using a specific custom-made oligonucleotide primer pair (electronic supplementary Table 1). The PCR products were subjected to sequencing reaction using the forward primer and BigDye terminator V3.1 cycle sequencing reagents (Applied Biosystems, Life Technologies, Carlsbad, CA, USA). DNA sequence was read on an ABI PRISM 3730xL DNA analyzer. Mass Spectrometric Genotyping The iPLEX assay is a single-base primer extension assay.16 Locus-specific PCR primers and allele-specific detection primers (electronic supplementary Table 1) were designed using MassARRAY Assay Design software (Sequenom Inc., San Diego, CA, USA). Allele detection was performed using matrix-assisted laser desorption/ ionization time-of-flight mass spectrometer (MALDI-TOF MS). The mass spectrograms were analyzed by the MassARRAY TYPER software 4.0. All analyses showed a very high call rate ([95 %, the median of call rate [97 %), and a very high consistency rate between blind duplicates ([97 %). Pyrosequencing Analysis Primer design (electronic supplementary Table 1) was carried out using the PyroMark Assay Design software 2.0 (Qiagen). The biotinylated PCR products were attached to Streptavidin-Sepharose beads (Amersham, GE Healthcare Bio-Sciences, Piscataway, NJ, USA) and processed to obtain single-stranded DNA using the PSQ 96 Sample Preparation Kit. The sequencing-by-synthesis reaction of the complementary strand was automatically performed using the PyroMark Gold Q24 reagents (Qiagen) according to kit specifications. The percentage of mutant alleles was then calculated by PyroMark Q24 version 2.0.6 software using the allele quantification mode.
Allelic Percentage of BRAF Mutation
Competitive Allele-Specific TaqMan PCR The BRAF mutation status was also determined with a commercially available TaqMan Mutation Detection Assay (Applied Biosystems). Competitive allele-specific TaqMan PCR (castPCR) using the BRAF_476_mu probe for the detection of BRAF c.1799T [ A mutation were run in triplicates on an Applied Biosystems 7900HT Fast RealTime PCR System. The mutant allele frequency was determined by comparing the cycle threshold (Ct) values of the wild-type and mutant allele assays (DCt) in reference to the control samples using Mutation Detector software 2.0. The percentage of mutant BRAF alleles was calculated as % = 1/(2DCt) 9 100.17
TABLE 1 Clinicopathological parameters and BRAF c.1799T [ A mutation status determined by Sanger sequencing BRAF mutation ? (n = 36)
BRAF mutation - (n = 14)
Female sex
29 (81)
11 (79)
1.000
Age at diagnosis (years)a
52 (37–58)
37 (32–51)
0.125
Body mass index (kg/m2)a
24.5 (21.6–28.9) 23.4 (19.4–25.8) 0.266
Body weight (kg)a
63 (52–76)
58 (56–65)
0.496
Tumor size (cm)a
2.2 (1.5–3.3)
1.4 (1.0–2.2)
0.015
Thyroiditis Extrathyroidal invasionb None
Statistical Analysis Statistical analyses were performed using STATA 11.0 (Stata Corp., College Station, TX, USA) and GraphPad Prism 6.02 (GraphPad Software, La Jolla, CA, USA). Continuous variables were compared using the Mann– Whitney test for two groups or Kruskal–Wallis test and Dunn’s procedure for more than two groups. Chi square, Fisher’s exact test, or Cochran-Armitage trend test were used to compare categorical variables. For correlation studies, Spearman’s rank-correlation test was used. Throughout the analysis, p-values \ 0.05 (two-sided hypotheses) were considered to be statistically significant.
p Value
4 (11)
4 (29)
9 (25)
10 (71)
Microscopic
12 (33)
3 (21)
Macroscopic
15 (42)
1 (7)
0.197 0.002
Multifocality
13 (36)
4 (29)
0.746
Lymphovascular invasion
12 (33)
2 (14)
0.295
Clinical lymph node metastasis
13 (36)
3 (21)
0.501
Pathological lymph node metastasis
21 (58)
8 (57)
1.000
I
15 (42)
11 (79)
II
0 (0)
2 (14)
III
9 (25)
1 (7)
IV
12 (33)
0 (0)
TNM stageb
RESULTS
0.003
Data are expressed as number (%) or median (interquartile range)
Patients Characteristics The study cohort consisted of 40 (80 %) female and 10 (20 %) male patients. Mean age at diagnosis was 47 years. A total of 16 (32 %) patients with clinically apparent nodal metastasis (cN1) had therapeutic central compartment neck dissection with or without lateral neck dissection. The remaining 34 (68 %) patients without preoperative or intraoperative evidence of central neck metastases underwent prophylactic central compartment dissection. Fortyseven tumors were classical PTC (including five nonincidental microcarcinomas), two were follicular variant PTC, and one was solid variant PTC. Mean tumor size was 2.3 cm. Eight patients showed histological evidence of chronic lymphocytic thyroiditis in the corresponding normal thyroid tissue. The numbers of tumor-infiltrating lymphocytes were not specifically determined in the study. Of 34 patients undergoing prophylactic central compartment dissection, 13 (38 %) showed microscopic nodal metastasis. Therefore, a total of 29 (58 %) had pathological nodal metastasis (pN1). The final TNM stage was determined on the basis of pathological data. One patient had
a
Mann–Whitney test
b
Cochran-Armitage trend test
evidence of lung metastasis at the time of diagnosis, and she was classified as TNM stage 2 because of her young age. Tumor size was associated with clinical and pathological lymph node metastasis (p = 0.016 and 0.025, respectively). Mutational Analysis by Sanger Sequencing Tumor BRAF mutation was found in 36 (72 %) of 50 PTCs. No mutation was found in the contralateral normal tissues or five follicular adenomas, indicating that this mutation was somatically acquired and tumor-specific. As shown in Table 1, the tumors with BRAF mutation were associated with larger tumor size (p = 0.015), higher probabilities of extrathyroidal invasion (p = 0.002), and more advanced TNM stage (p = 0.003). Two (40 %) of five microcarcinomas were positive for BRAF mutation. Tumor BRAF mutation status was not associated with
S.-P. Cheng et al.
BRAF c.1799T>A (% alleles)
40
extrathyroidal invasion and TNM stage between the second and third tertile (p = 0.347 and 0.113, respectively). A seemingly higher disease stage in the second tertile may reflect the relatively older age in this subgroup. As shown in Fig. 1b, there was a strong positive correlation between tumor size and mutant allele frequency (Spearman’s rho = 0.47; p = 0.0005). Analyzed from another perspective, PTC in association with clinical nodal metastasis showed a mutant allele percentage significantly higher than cN0 PTC (p = 0.037; Fig. 2a). There was no difference in the mutant allele frequency between PTC with or without pathological nodal metastasis (p = 0.984; Fig. 2b).
(a)
30
20
10
Mutant allele frequency (%)
0
40
(b)
30
Mutational Analysis by Pyrosequencing 20 10 0 0
1
2
3
4
5
6
Tumor size (cm)
FIG. 1 a The percentage of mutant BRAF c.1799T [ A (p.Val600Glu) alleles detected by mass spectrometric genotyping in 50 papillary thyroid carcinomas. b Scatter plot showing the correlation between tumor size and the percentage of mutant BRAF alleles. Lines represent simple linear regression ±95 % confidence interval. Spearman’s rho = 0.47, p = 0.0005; simple regression adjusted r2 = 0.21, p = 0.001
clinical or pathological lymph node metastasis (p = 0.501 and 1.000, respectively). Mutational Analysis by Mass Spectrometry The percentage of mutant BRAF alleles in genomic DNA from 50 PTCs was determined by mass spectrometric genotyping (Fig. 1a). There was no mutant allele in nine PTCs and five follicular adenomas. For 41 (82 %) PTCs carrying BRAF mutation, the range of mutant allele frequency was 3.2–35.2 %, with a mean and median of 22.5 and 25.2 %, respectively. Overall, five samples failed to be assigned to positive BRAF mutation by Sanger sequencing (90 % concordance). These discordant samples showed a mutant allele frequency from 3.2 to 13.5 % (mean 6.6 %). Tumor size varied from 1.3 to 4.0 cm (mean 2.4 cm). There was no significant difference in characteristics between the cases with concordant results and those with discordant results, except for a lower frequency of mutant allele in the discordant samples. Patients were categorized on the basis of mutant allele frequency into tertiles (Table 2). Tumor BRAF mutation was associated with body weight (p = 0.023), tumor size (p = 0.004), extrathyroidal invasion (p = 0.024), and TNM stage (p = 0.029). There was no difference in
The percentage of mutant BRAF alleles from five follicular adenomas varied from 3 to 5 %. Therefore, the cutoff was set at 5 %, consistent with previous reports.12,13 For 42 PTCs positive for BRAF mutation, the range of mutant allele frequency was 6–46 %, with a mean and median of 31 and 36 %, respectively. The mutant allele frequency was highly consistent between mass spectrometric genotying and pyrosequencing (Fig. 3a). One patient with a 7 % BRAF mutation allele based on pyrosequencing analysis had no mutant BRAF allele detectable by mass spectrometry (98 % concordance). The percentage of mutant BRAF alleles was strongly correlated with tumor size (Spearman’s rho = 0.52; p = 0.0001), but not age (p = 0.782) or body weight (p = 0.100). Comparing medians of the mutant allele frequency, significant differences were observed in extrathyroidal invasion (p = 0.034) and clinical nodal metastasis (p = 0.041), but not pathological nodal metastasis (p = 0.937). Tumor BRAF mutation determined by pyrosequencing was associated with higher probabilities of stage III/IV disease (p = 0.018). Mutational Analysis by castPCR Five follicular adenomas were in the range of 0.03– 1.7 % mutant allele. A conservative 2.5 % cutoff was chosen to avoid unreliable estimates due to stochastic fluctuations in low copy number situations.18 In concordance with the results obtained by pyrosequencing, 42 PTCs were positive for BRAF mutation based on castPCR assay. The range of mutant allele frequency was 3.0– 61.6 %, with a mean and median of 30.7 and 29.1 %, respectively. The mutant allele frequency was moderately correlated between mass spectrometric genotying and castPCR assay (Fig. 3b). A similar correlation coefficient was observed between pyrosequencing and castPCR (Spearman’s rho = 0.87; p \ 0.0001).
Allelic Percentage of BRAF Mutation TABLE 2 Clinicopathological parameters and BRAF c.1799T [ A mutation status determined by mass spectrometric genotyping Mutant allele 0–10 % (n = 16) Female sex Age at diagnosis (years)
Mutant allele 10–27.5 % (n = 16)
13 (81) a
Body mass index (kg/m2)a
14 (88)
37 (32–58) 23.4 (20.1–26.8)
56 (52–59) 24.0 (20.5–26.0)
Mutant allele [ 27.5 % (n = 18) 13 (72) 41 (33–50) 26.6 (23.0–29.4)
p-Value
0.615 0.053 0.191
Body weight (kg)a,b
58 (54–69)
55 (47–65)
74 (58–80)
0.023
Tumor size (cm)a,c
1.3 (1.1–2.0)
2.3 (1.5–3.2)
2.5 (2.0–3.5)
0.004
Thyroiditis
5 (31)
2 (13)
1 (6)
10 (63)
4 (25)
5 (28)
Microscopic
5 (31)
3 (19)
7 (39)
Macroscopic
1 (6)
9 (56)
6 (33)
Multifocality
4 (25)
9 (56)
4 (22)
0.111
Lymphovascular invasion
1 (6)
6 (38)
7 (39)
0.062
Clinical lymph node metastasis
2 (13)
5 (31)
9 (50)
0.070
Pathological lymph node metastasis
9 (56)
9 (56)
11 (61)
1.000
I
11 (69)
4 (25)
11 (61)
II
2 (13)
0 (0)
0 (0)
III IV
2 (13) 1 (6)
5 (31) 7 (44)
3 (17) 4 (22)
Extrathyroidal invasion None
0.140 0.024
TNM stage
0.029
Data are expressed as number (%) or median (interquartile range) a
Kruskal–Wallis test
b
Dunn’s post hoc tests: tertile 1 vs. tertile 2, p = 1.000; tertile 2 vs. tertile 3, p = 0.021; tertile 1 vs. tertile 3, p = 0.256
c
Dunn’s post hoc tests: tertile 1 vs. tertile 2, p = 0.098; tertile 2 vs. tertile 3, p = 0.855; tertile 1 vs tertile 3, p = 0.003
The percentage of mutant BRAF alleles was positively correlated with tumor size (Spearman’s rho = 0.47; p = 0.006). In addition, the mutant allele frequency was associated with extrathyroidal invasion (p = 0.010) and stage III/IV disease (p = 0.009), but not clinical or pathological nodal metastasis (p = 0.075 and 0.602, respectively). DISCUSSION The occurrence of BRAF mutation in PTC has been extensively investigated. It appears that the development of BRAF mutation in the thyroid gland is associated with environmental factors and iodine intake.19,20 The prevalence of BRAF mutation in PTC varies according to the detection methods, ranging from 27 to 90 % in the literature.9,10,21 Among the mutation detection methods, Sanger sequencing is least sensitive and capable of detecting samples having 20–30 % mutant alleles.22 Additionally, the mutation detection may be influenced by DNA extraction methods. Kim et al. 23 showed that direct DNA sequencing and pyrosequencing to detect BRAF mutation were more accurate in fresh-frozen than formalin-fixed and paraffin-embedded tissue sections. In this study, we
confirmed that Sanger sequencing may fail to detect BRAF mutation in a small portion of frozen PTC specimens which contain a low allelic fraction. BRAF mutation can be found in a substantial number of microcarcinomas.23,24 In transgenic mouse models, mice with targeted expression of BRAFV600E in thyroid cells developed PTC characteristics of the human disease.25,26 These findings indicate that BRAF mutation represents an early, initiating event in tumor development. However, discordant patterns of BRAF mutation were found in about 40 % of the multifocal PTCs.27,28 By analyzing the relationship between genome-wide allelic imbalances and BRAF mutation, Jovanovic and collaborators reported that BRAF mutation did not represent the earliest transforming event.29 Furthermore, de novo BRAF mutation was observed in metastatic lymph nodes from mutation-negative primary tumors.30 These results imply that BRAF mutation may be a late genetic change heterogeneously occurring in the tumor bulk. As such, the role of BRAF mutation in pathogenesis of thyroid cancer appears complicated and conflicting, and is being actively debated.31 Resolving these contradicting views is important to understanding the oncologic and therapeutic significance of
S.-P. Cheng et al. P = 0.037
(a)
50
(a)
Pyrosequencing (%)
Mutant BRAF (% alleles)
40
30
20
10
40 30 20 10 0 0
(b)
20
30
40
1
Mass genotyping (%)
(b)
P = 0.984 60
40
50
castPCR (%)
Mutant BRAF (% alleles)
10
cN
cN
0
0
30
20
40 30 20 10
10
0 0
1 pN
pN
0
0
FIG. 2 Vertical scatter plots of the percentage of mutant BRAF alleles grouped by a clinical and b pathological lymph node metastasis. Lines represent median values. Mann–Whitney tests were performed to evaluate the difference between groups
BRAF mutation. In the hierarchical model of clonal evolution, early somatic mutations tend to be propagated in many or all clones, whereas later events occur only in some clones.32 A 50 % percentage of mutant BRAF alleles would be the expected finding, if all PTC cells carry one mutant and one wild-type BRAF allele. To test this concept, Guerra and colleagues first demonstrated that most PTCs consist of a mixture of tumor cells carrying wild-type and mutant BRAF.33 Gandolfi et al.13 also found an average mutated allele percentage of 27.4 % with a range between 7.5 and 49.8 %. In line with their observations, our data showed that the mutant allele frequency was less than 25 % in 18 (44 %) of 41 PTCs positive for mutation. To further examine the biological relevance of clonality of BRAF mutation, we studied the correlation between the mutant allele frequency and clinicopathological parameters. Our results indicate that larger tumors and tumors with extrathyroidal invasion had a significantly higher mutant allele frequency. These findings are consistent with recent metaanalyses.5,9–11 Of note is the fact that a strong correlation existed between tumor size and mutant allele frequency. In agreement with the present observations, Guerra and
10
20
30
40
Mass genotyping (%)
FIG. 3 a Scatter plot showing the correlation of the mutant BRAF allele frequency determined by mass spectrometric genotying and pyrosequencing analysis. Spearman’s rho = 0.97, p \ 0.0001; simple regression adjusted r2 = 0.98, p \ 0.001. b Scatter plot showing the correlation of the mutant BRAF allele frequency determined by mass spectrometric genotying and competitive allele-specific TaqMan PCR (castPCR). Spearman’s rho = 0.87, p \ 0.0001; simple regression adjusted r2 = 0.85, p \ 0.001
colleagues demonstrated that tumor volume directly correlated with the mutant allele frequency.12 A reasonable interpretation of these results is that BRAF mutation provides cancer cells a survival and growth advantage, resulting in selection of BRAF mutant alleles during tumor progression. These data also refute an alternative hypothesis which proposed that BRAF mutation occurs as a primary genetic event but is later removed by the DNA repair machinery.31 In contrast to the results of previous meta-analyses,5,9–11 our data indicate that BRAF mutation was not associated with lymph node metastasis. It is worth noting that in most studies, neck dissections were performed for suspicious lymphadenopathy.9 We observed a small and inconsistent difference in the mutant allele frequency between clinically node-positive and node-negative tumors. When prophylactic central compartment dissection was performed, the mutant allele frequency was not associated with pathological nodal status. In accordance, Gandolfi et al.13 found that the average mutated allele
Allelic Percentage of BRAF Mutation
percentage was significantly lower in corresponding metastatic nodes compared with the primary PTCs (18.6 vs. 28.6 %). Therefore, it is likely that cancer cells harboring BRAF mutation do not really have enhanced metastatic capacity. More frequent lymph node metastasis observed in BRAF mutation-positive tumors may result from a larger tumor burden and more extensive extrathyroidal invasion. It was recently showed that BRAF mutation was the only independent predictor of central compartment lymph node metastasis in PTC;34 however, the design of the study was biased by including BRAF status as a criterion for prophylactic nodal dissection. A plethora of genetic and phenotypic heterogeneity comes from ongoing genetic instability within a cancer. Most of the spontaneous mutations are evolutionarily neutral or deleterious. Rare mutations that confer competitive advantage can be selected for by Darwinian forces and will eventually become predominant. Nonetheless, achieving clonal dominance can take a long time, partly because of spatial constraints.35 Our findings have important biological implications in that they point out the possibility that activating BRAF mutation may not necessarily be an initiating prerequisite in PTC tumorigenesis, but this secondary genetic alteration provides a survival advantage to cancer cells, and accumulates. The fact that the mutant allele frequency was associated with extrathyroidal invasion also makes it unlikely that expansion emerges from fixation of neutral mutations over time (genetic drift). Our results seemingly suggest that quantitative assessment of mutation frequency provides no additional information on clinical significance of BRAF mutation (Tables 1 and 2). However, even a small portion of mutation-negative tumor cells may hamper efficacy of BRAF-targeted therapy. In RAS/RAF wild-type tumors, RAF kinase inhibitors may paradoxically activate the MAPK pathway in an RAS-dependent manner, thus enhancing tumor growth in xenograft models.36 Further studies may be undertaken to determine the effects of clonality measures on clinical response to BRAF-targeted therapy. A limitation of our study is the inability to exclude stromal contamination. Tumor samples always have some degree of contamination by normal stromal, endothelial, and inflammatory cells. The mutant allele frequency in unselected cells would be lower than that in laser-captured PTC cells.33 Nonetheless, the average percentage of mutant BRAF alleles analyzed by different methods in the present study (22.5 to 31 %) was compatible with those of recent reports (21.9 to 27.4 %).12,13 Using next-generation sequencing, Nikiforova and colleagues also demonstrated that the mutant allele frequency was 18–44 % (mean 34 %) in PTCs, corresponding to 36–88 % of cells with heterozygous mutation.37
CONCLUSIONS The percentage of mutant BRAF alleles in PTC is significantly associated with tumor burden and extrathyroidal invasion. A positive correlation between the mutant allele frequency and tumor size implies that BRAF mutation may play a role in tumor progression. ACKNOWLEDGMENT The authors would like to thank the Translational Core Laboratories of National Translational Medicine and Clinical Trial Resource Center for technical assistance in MALDI-TOF MS experiments. This work was supported by the National Science Council of Taiwan (100-2314-B-195-001-MY3) and Mackay Memorial Hospital (MMH-10206 and MMH-E-102-10). DISCLOSURES Shih-Ping Cheng, Yi-Chiung Hsu, Chien-Liang Liu, Tsang-Pai Liu, Ming-Nan Chien, Tao-Yeuan Wang, and Jie-Jen Lee have nothing to disclose.
REFERENCES 1. Cramer JD, Fu P, Harth KC, Margevicius S, Wilhelm SM. Analysis of the rising incidence of thyroid cancer using the surveillance, epidemiology and end results national cancer data registry. Surgery. 2010;148:1147–52. 2. Husson O, Haak HR, van Steenbergen LN, et al. Rising incidence, no change in survival and decreasing mortality from thyroid cancer in The Netherlands since 1989. Endocr Relat Cancer. 2013;20:263–71. 3. Eustatia-Rutten CF, Corssmit EP, Biermasz NR, Pereira AM, Romijn JA, Smit JW. Survival and death causes in differentiated thyroid carcinoma. J Clin Endocrinol Metab. 2006;91:313–9. 4. Kapiteijn E, Schneider TC, Morreau H, Gelderblom H, Nortier JW, Smit JW. New treatment modalities in advanced thyroid cancer. Ann Oncol. 2012;23:10–8. 5. Xing M. BRAF mutation in papillary thyroid cancer: pathogenic role, molecular bases, and clinical implications. Endocr Rev. 2007;28:742–62. 6. Mathur A, Moses W, Rahbari R, Khanafshar E, Duh QY, Clark O, et al. Higher rate of BRAF mutation in papillary thyroid cancer over time: a single-institution study. Cancer. 2011;117:4390–5. 7. Wan PT, Garnett MJ, Roe SM, et al; Cancer Genome Project. Mechanism of activation of the RAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell. 2004;116:855–67. 8. Pratilas CA, Taylor BS, Ye Q, Viale A, Sander C, Solit DB, et al. (V600E)BRAF is associated with disabled feedback inhibition of RAF-MEK signaling and elevated transcriptional output of the pathway. Proc Natl Acad Sci USA. 2009;106:4519–24. 9. Li C, Lee KC, Schneider EB, Zeiger MA. BRAF V600E mutation and its association with clinicopathological features of papillary thyroid cancer: a meta-analysis. J Clin Endocrinol Metab. 2012;97:4559–70. 10. Kim TH, Park YJ, Lim JA, Ahn HY, Lee EK, Lee YJ, et al. The association of the BRAF(V600E) mutation with prognostic factors and poor clinical outcome in papillary thyroid cancer: a meta-analysis. Cancer. 2012;118:1764–73. 11. Tufano RP, Teixeira GV, Bishop J, Carson KA, Xing M. BRAF mutation in papillary thyroid cancer and its value in tailoring initial treatment: a systematic review and meta-analysis. Medicine (Baltimore). 2012;91:274–86. 12. Guerra A, Fugazzola L, Marotta V, Cirillo M, Rossi S, Cirello V, et al. A high percentage of BRAFV600E alleles in papillary
S.-P. Cheng et al.
13.
14. 15.
16.
17. 18.
19.
20.
21.
22.
23.
24.
25.
thyroid carcinoma predicts a poorer outcome. J Clin Endocrinol Metab. 2012;97:2333–40. Gandolfi G, Sancisi V, Torricelli F, Ragazzi M, Frasoldati A, Piana S, et al. Allele percentage of the BRAF V600E mutation in papillary thyroid carcinomas and corresponding lymph node metastases: no evidence for a role in tumor progression. J Clin Endocrinol Metab. 2013;98:E934–42. MacConaill LE. Existing and emerging technologies for tumor genomic profiling. J Clin Oncol. 2013;31:1815–24. Cheng SP, Liu CL, Hsu YC, Chang YC, Huang SY, Lee JJ. Expression and biologic significance of adiponectin receptors in papillary thyroid carcinoma. Cell Biochem Biophys. 2013;65: 203–10. Thomas RK, Baker AC, Debiasi RM, Winckler W, LaFramboise T, Lin WM, et al. High-throughput oncogene mutation profiling in human cancer. Nat Genet. 2007;39:347–51. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001;29:e45. Kloos RT, Reynolds JD, Walsh PS, Wilde JI, Tom EY, Pagan M, et al. Does addition of BRAF V600E mutation testing modify sensitivity or specificity of the Afirma Gene Expression Classifier in cytologically indeterminate thyroid nodules? J Clin Endocrinol Metab. 2013;98:E761–8. Frasca F, Nucera C, Pellegriti G, Gangemi P, Attard M, Stella M, et al. BRAF(V600E) mutation and the biology of papillary thyroid cancer. Endocr Relat Cancer. 2008;15:191–205. Guan H, Ji M, Bao R, Yu H, Wang Y, Hou P, et al. Association of high iodine intake with the T1799A BRAF mutation in papillary thyroid cancer. J Clin Endocrinol Metab. 2009;94:1612–7. Jeong D, Jeong Y, Park JH, Han SW, Kim SY, Kim YJ, et al. BRAF (V600E) mutation analysis in papillary thyroid carcinomas by peptide nucleic acid clamp real-time PCR. Ann Surg Oncol. 2013;20:759–66. Lade-Keller J, Romer KM, Guldberg P, Riber-Hansen R, Hansen LL, Steiniche T, et al. Evaluation of BRAF mutation testing methodologies in formalin-fixed, paraffin-embedded cutaneous melanomas. J Mol Diagn. 2013;15:70–80. Kim HS, Kim JO, Lee DH, Lee HC, Kim HJ, Kim JH, et al. Factors influencing the detection of the BRAF T1799A mutation in papillary thyroid carcinoma. Oncol Rep. 2011;25:1639–44. Virk RK, Van Dyke AL, Finkelstein A, Prasad A, Gibson J, Hui P, et al. BRAFV600E mutation in papillary thyroid microcarcinoma: a genotype-phenotype correlation. Mod Pathol. 2013;26:62–70. Knauf JA, Ma X, Smith EP, Zhang L, Mitsutake N, Liao XH, et al. Targeted expression of BRAFV600E in thyroid cells of
26.
27.
28.
29.
30.
31. 32. 33.
34.
35. 36.
37.
transgenic mice results in papillary thyroid cancers that undergo dedifferentiation. Cancer Res. 2005;65:4238–45. Charles RP, Iezza G, Amendola E, Dankort D, McMahon M. Mutationally activated BRAF(V600E) elicits papillary thyroid cancer in the adult mouse. Cancer Res. 2011;71:3863–71. Park SY, Park YJ, Lee YJ, Lee HS, Choi SH, Choe G, et al. Analysis of differential BRAF(V600E) mutational status in multifocal papillary thyroid carcinoma: evidence of independent clonal origin in distinct tumor foci. Cancer. 2006;107:1831–8. Giannini R, Ugolini C, Lupi C, Proietti A, Elisei R, Salvatore G, et al. The heterogeneous distribution of BRAF mutation supports the independent clonal origin of distinct tumor foci in multifocal papillary thyroid carcinoma. J Clin Endocrinol Metab. 2007;92:3511–6. Jovanovic L, Delahunt B, McIver B, Eberhardt NL, Grebe SK. Most multifocal papillary thyroid carcinomas acquire genetic and morphotype diversity through subclonal evolution following the intra-glandular spread of the initial neoplastic clone. J Pathol. 2008;215:145–54. Vasko V, Hu S, Wu G, Xing JC, Larin A, Savchenko V, et al. High prevalence and possible de novo formation of BRAF mutation in metastasized papillary thyroid cancer in lymph nodes. J Clin Endocrinol Metab. 2005;90:5265–9. Xing M. BRAFV600E mutation and papillary thyroid cancer: chicken or egg? J Clin Endocrinol Metab. 2012;97:2295–8. Aparicio S, Caldas C. The implications of clonal genome evolution for cancer medicine. N Engl J Med. 2013;368:842-51. Guerra A, Sapio MR, Marotta V, Campanile E, Rossi S, Forno I, et al. The primary occurrence of BRAF(V600E) is a rare clonal event in papillary thyroid carcinoma. J Clin Endocrinol Metab. 2012;97:517–24. Howell GM, Nikiforova MN, Carty SE, Armstrong MJ, Hodak SP, Stang M T, et al. BRAF V600E mutation independently predicts central compartment lymph node metastasis in patients with papillary thyroid cancer. Ann Surg Oncol. 2013;20:47–52. Marusyk A, Polyak K. Tumor heterogeneity: causes and consequences. Biochim Biophys Acta. 2010;1805:105–17. Hatzivassiliou G, Song K, Yen I, Brandhuber BJ, Anderson DJ, Alvarado R, et al. RAF inhibitors prime wild-type RAF to activate the MAPK pathway and enhance growth. Nature. 2010;464:431–5. Nikiforova MN, Wald AI, Roy S, Durso MB, Nikiforov YE. Targeted next-generation sequencing panel (ThyroSeq) for detection of mutations in thyroid cancer. J Clin Endocrinol Metab. 2013;98:E1852–60.
ORIGINAL ARTICLE – TRANSLATIONAL RESEARCH AND BIOMARKERS
Significance of Allelic Percentage of BRAF c.1799T [ A (V600E) Mutation in Papillary Thyroid Carcinoma Shih-Ping Cheng, MD, PhD1,2,6, Yi-Chiung Hsu, PhD5, Chien-Liang Liu, MD1,2, Tsang-Pai Liu, MD1,2, Ming-Nan Chien, MD1,3, Tao-Yeuan Wang, MD1,4, and Jie-Jen Lee, MD, PhD1,2,6 Mackay Junior College of Medicine, Nursing, and Management, Taipei, Taiwan; 2Department of Surgery, Mackay Medical College and Mackay Memorial Hospital, Taipei, Taiwan; 3Division of Endocrinology and Metabolism, Department of Medicine, Mackay Medical College and Mackay Memorial Hospital, Taipei, Taiwan; 4Department of Pathology, Mackay Medical College and Mackay Memorial Hospital, Taipei, Taiwan; 5Institute of Statistical Science, Academia Sinica, Taipei, Taiwan; 6Graduate Institute of Medical Sciences and Department of Pharmacology, Taipei Medical University, Taipei, Taiwan 1
ABSTRACT Background. Somatic BRAF mutation is frequently observed in papillary thyroid carcinoma (PTC). Recent evidence suggests that PTCs are heterogeneous tumors containing a subclonal or oligoclonal occurrence of BRAF mutation. Conflicting results have been reported concerning the prognostic significance of the mutant allele frequency. Our present aim was to investigate the association between the percentage of BRAF c.1799T [ A (p.Val600Glu) alleles and clinicopathological parameters in PTC. Methods. Genomic DNA was extracted from fresh-frozen specimens obtained from 50 PTC patients undergoing total thyroidectomy. The BRAF mutation status was determined by Sanger sequencing. The percentage of mutant BRAF alleles was quantified by mass spectrometric genotyping, pyrosequencing, and competitive allele-specific TaqMan PCR (castPCR). Results. Positive rate of BRAF mutation was 72 % by Sanger sequencing, 82 % by mass spectrometric genotying, and 84 % by pyrosequencing or castPCR. The average percentage of mutant BRAF alleles was 22.5, 31, and 30.7 %, respectively.
Electronic supplementary material The online version of this article (doi:10.1245/s10434-014-3723-5) contains supplementary material, which is available to authorized users. Ó Society of Surgical Oncology 2014 First Received: 31 October 2013 J.-J. Lee, MD, PhD e-mail: [email protected]
There was a good correlation among three quantification methods (Spearman’s rho = 0.87–0.97; p \ 0.0001). The mutant allele frequency was significantly correlated with tumor size (rho = 0.47–0.52; p \ 0.01) and extrathyroidal invasion. The frequency showed no difference in pathological lymph node metastasis. Conclusions. The percentage of mutant BRAF alleles is positively associated with tumor burden and extrathyroidal invasion in PTC. Relatively good correlations exist among mass spectrometric genotyping, pyrosequencing, and castPCR in quantification of mutant BRAF allele frequency.
Papillary thyroid carcinoma (PTC) is the most common malignant tumor of the thyroid gland. Thyroid cancer incidence has increased substantially over the last decade, and this change is primarily attributed to an increase in PTC.1,2 Although most PTCs are curable or controllable by the combination of surgery, radioactive iodine ablation, and thyrotropin suppression, a considerable number of patients die of persistent or recurrent disease that is refractory to conventional therapy.3 With a growing understanding of molecular oncology, therapeutic agents that target activating genetic alterations in thyroid cancer have been rapidly developed.4 The v-raf murine sarcoma viral oncogene homolog B (BRAF) mutation occurs frequently in PTC.5 The most common BRAF mutation comprises a single base transversion at codon 600 in exon 15, identified as c.1799T [ A (p.Val600Glu) and generally referred to as V600E. The prevalence of BRAF mutation was found to be increased significantly over a 15-year period, and it was proposed
S.-P. Cheng et al.
that this change may contribute to the increasing incidence of thyroid cancer.6 BRAF c.1799T [ A mutation greatly increases the kinase activity and is associated with elevated transcriptional output of the mitogen-activated protein kinase (MAPK) pathway.7,8 Data from meta-analyses consistently indicate that the presence of BRAF mutation in thyroid cancer is associated with extrathyroidal invasion, lymph node metastasis, advanced TNM stage, and disease recurrence.5,9–11 If thyroid cancer cells harboring BRAF mutation are biologically more aggressive, tumors containing more BRAF mutationpositive cells will theoretically have a worse outcome. In this regard, Guerra and colleagues used pyrosequencing to quantify the percentage of mutant BRAF alleles, demonstrating that the mutant allele frequency correlated with age and tumor volume, and the odds ratio of recurrent disease was 5.31-fold higher in PTCs harboring a high ([30 %) percentage of mutant BRAF alleles.12 On the contrary, Gandolfi et al. 13 found no association between the percentage of mutant alleles and metastatic behavior. The reason for this discrepancy remains to be elucidated. Methods to detect BRAF mutation have not been standardized. Conventional DNA sequencing is binary (present or absent) and takes no account of the ratio of mutated DNA to wild type. Contemporary technological developments in genotyping are advantageous in terms of cost, throughput, scalability, and sensitivity.14 The primary aim of this study was to perform a multiplatform analysis to characterize the biological significance of the percentage of BRAF c.1799T [ A alleles in PTC. The secondary aim was to compare different methods for quantification of mutant BRAF allele frequency.
PATIENTS AND METHODS Patients and Tissue Samples This study was approved by the Institutional Review Board of Mackay Memorial Hospital (12MMHIS175). All patients provided written informed consent before the procurement of tissue specimens. A total of 50 patients who underwent total thyroidectomy for PTC were consecutively selected. Five patients who underwent lobectomy for follicular adenoma were also included as negative controls. All patients were euthyroid at the time of surgery and were not taking any medications that could affect serum thyrotropin levels. PTC tissues from the center of the lesions and corresponding normal thyroid tissues from the contralateral lobes of the same patient were obtained. All PTC tissue samples were carefully dissected to exclude surrounding normal tissue. Tissues were immediately snap-frozen in
liquid nitrogen and stored at -80 °C. The tissue diagnosis was confirmed by frozen section. Genomic DNA Extraction Genomic DNA was extracted using the QIAamp DNA mini kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. Serial dilutions of DNA from B-CPAP thyroid cancer cells (homozygous BRAF mutation ?/?) were used as positive controls.15 Sanger Sequencing Exon 15 of the BRAF gene was amplified using a specific custom-made oligonucleotide primer pair (electronic supplementary Table 1). The PCR products were subjected to sequencing reaction using the forward primer and BigDye terminator V3.1 cycle sequencing reagents (Applied Biosystems, Life Technologies, Carlsbad, CA, USA). DNA sequence was read on an ABI PRISM 3730xL DNA analyzer. Mass Spectrometric Genotyping The iPLEX assay is a single-base primer extension assay.16 Locus-specific PCR primers and allele-specific detection primers (electronic supplementary Table 1) were designed using MassARRAY Assay Design software (Sequenom Inc., San Diego, CA, USA). Allele detection was performed using matrix-assisted laser desorption/ ionization time-of-flight mass spectrometer (MALDI-TOF MS). The mass spectrograms were analyzed by the MassARRAY TYPER software 4.0. All analyses showed a very high call rate ([95 %, the median of call rate [97 %), and a very high consistency rate between blind duplicates ([97 %). Pyrosequencing Analysis Primer design (electronic supplementary Table 1) was carried out using the PyroMark Assay Design software 2.0 (Qiagen). The biotinylated PCR products were attached to Streptavidin-Sepharose beads (Amersham, GE Healthcare Bio-Sciences, Piscataway, NJ, USA) and processed to obtain single-stranded DNA using the PSQ 96 Sample Preparation Kit. The sequencing-by-synthesis reaction of the complementary strand was automatically performed using the PyroMark Gold Q24 reagents (Qiagen) according to kit specifications. The percentage of mutant alleles was then calculated by PyroMark Q24 version 2.0.6 software using the allele quantification mode.
Allelic Percentage of BRAF Mutation
Competitive Allele-Specific TaqMan PCR The BRAF mutation status was also determined with a commercially available TaqMan Mutation Detection Assay (Applied Biosystems). Competitive allele-specific TaqMan PCR (castPCR) using the BRAF_476_mu probe for the detection of BRAF c.1799T [ A mutation were run in triplicates on an Applied Biosystems 7900HT Fast RealTime PCR System. The mutant allele frequency was determined by comparing the cycle threshold (Ct) values of the wild-type and mutant allele assays (DCt) in reference to the control samples using Mutation Detector software 2.0. The percentage of mutant BRAF alleles was calculated as % = 1/(2DCt) 9 100.17
TABLE 1 Clinicopathological parameters and BRAF c.1799T [ A mutation status determined by Sanger sequencing BRAF mutation ? (n = 36)
BRAF mutation - (n = 14)
Female sex
29 (81)
11 (79)
1.000
Age at diagnosis (years)a
52 (37–58)
37 (32–51)
0.125
Body mass index (kg/m2)a
24.5 (21.6–28.9) 23.4 (19.4–25.8) 0.266
Body weight (kg)a
63 (52–76)
58 (56–65)
0.496
Tumor size (cm)a
2.2 (1.5–3.3)
1.4 (1.0–2.2)
0.015
Thyroiditis Extrathyroidal invasionb None
Statistical Analysis Statistical analyses were performed using STATA 11.0 (Stata Corp., College Station, TX, USA) and GraphPad Prism 6.02 (GraphPad Software, La Jolla, CA, USA). Continuous variables were compared using the Mann– Whitney test for two groups or Kruskal–Wallis test and Dunn’s procedure for more than two groups. Chi square, Fisher’s exact test, or Cochran-Armitage trend test were used to compare categorical variables. For correlation studies, Spearman’s rank-correlation test was used. Throughout the analysis, p-values \ 0.05 (two-sided hypotheses) were considered to be statistically significant.
p Value
4 (11)
4 (29)
9 (25)
10 (71)
Microscopic
12 (33)
3 (21)
Macroscopic
15 (42)
1 (7)
0.197 0.002
Multifocality
13 (36)
4 (29)
0.746
Lymphovascular invasion
12 (33)
2 (14)
0.295
Clinical lymph node metastasis
13 (36)
3 (21)
0.501
Pathological lymph node metastasis
21 (58)
8 (57)
1.000
I
15 (42)
11 (79)
II
0 (0)
2 (14)
III
9 (25)
1 (7)
IV
12 (33)
0 (0)
TNM stageb
RESULTS
0.003
Data are expressed as number (%) or median (interquartile range)
Patients Characteristics The study cohort consisted of 40 (80 %) female and 10 (20 %) male patients. Mean age at diagnosis was 47 years. A total of 16 (32 %) patients with clinically apparent nodal metastasis (cN1) had therapeutic central compartment neck dissection with or without lateral neck dissection. The remaining 34 (68 %) patients without preoperative or intraoperative evidence of central neck metastases underwent prophylactic central compartment dissection. Fortyseven tumors were classical PTC (including five nonincidental microcarcinomas), two were follicular variant PTC, and one was solid variant PTC. Mean tumor size was 2.3 cm. Eight patients showed histological evidence of chronic lymphocytic thyroiditis in the corresponding normal thyroid tissue. The numbers of tumor-infiltrating lymphocytes were not specifically determined in the study. Of 34 patients undergoing prophylactic central compartment dissection, 13 (38 %) showed microscopic nodal metastasis. Therefore, a total of 29 (58 %) had pathological nodal metastasis (pN1). The final TNM stage was determined on the basis of pathological data. One patient had
a
Mann–Whitney test
b
Cochran-Armitage trend test
evidence of lung metastasis at the time of diagnosis, and she was classified as TNM stage 2 because of her young age. Tumor size was associated with clinical and pathological lymph node metastasis (p = 0.016 and 0.025, respectively). Mutational Analysis by Sanger Sequencing Tumor BRAF mutation was found in 36 (72 %) of 50 PTCs. No mutation was found in the contralateral normal tissues or five follicular adenomas, indicating that this mutation was somatically acquired and tumor-specific. As shown in Table 1, the tumors with BRAF mutation were associated with larger tumor size (p = 0.015), higher probabilities of extrathyroidal invasion (p = 0.002), and more advanced TNM stage (p = 0.003). Two (40 %) of five microcarcinomas were positive for BRAF mutation. Tumor BRAF mutation status was not associated with
S.-P. Cheng et al.
BRAF c.1799T>A (% alleles)
40
extrathyroidal invasion and TNM stage between the second and third tertile (p = 0.347 and 0.113, respectively). A seemingly higher disease stage in the second tertile may reflect the relatively older age in this subgroup. As shown in Fig. 1b, there was a strong positive correlation between tumor size and mutant allele frequency (Spearman’s rho = 0.47; p = 0.0005). Analyzed from another perspective, PTC in association with clinical nodal metastasis showed a mutant allele percentage significantly higher than cN0 PTC (p = 0.037; Fig. 2a). There was no difference in the mutant allele frequency between PTC with or without pathological nodal metastasis (p = 0.984; Fig. 2b).
(a)
30
20
10
Mutant allele frequency (%)
0
40
(b)
30
Mutational Analysis by Pyrosequencing 20 10 0 0
1
2
3
4
5
6
Tumor size (cm)
FIG. 1 a The percentage of mutant BRAF c.1799T [ A (p.Val600Glu) alleles detected by mass spectrometric genotyping in 50 papillary thyroid carcinomas. b Scatter plot showing the correlation between tumor size and the percentage of mutant BRAF alleles. Lines represent simple linear regression ±95 % confidence interval. Spearman’s rho = 0.47, p = 0.0005; simple regression adjusted r2 = 0.21, p = 0.001
clinical or pathological lymph node metastasis (p = 0.501 and 1.000, respectively). Mutational Analysis by Mass Spectrometry The percentage of mutant BRAF alleles in genomic DNA from 50 PTCs was determined by mass spectrometric genotyping (Fig. 1a). There was no mutant allele in nine PTCs and five follicular adenomas. For 41 (82 %) PTCs carrying BRAF mutation, the range of mutant allele frequency was 3.2–35.2 %, with a mean and median of 22.5 and 25.2 %, respectively. Overall, five samples failed to be assigned to positive BRAF mutation by Sanger sequencing (90 % concordance). These discordant samples showed a mutant allele frequency from 3.2 to 13.5 % (mean 6.6 %). Tumor size varied from 1.3 to 4.0 cm (mean 2.4 cm). There was no significant difference in characteristics between the cases with concordant results and those with discordant results, except for a lower frequency of mutant allele in the discordant samples. Patients were categorized on the basis of mutant allele frequency into tertiles (Table 2). Tumor BRAF mutation was associated with body weight (p = 0.023), tumor size (p = 0.004), extrathyroidal invasion (p = 0.024), and TNM stage (p = 0.029). There was no difference in
The percentage of mutant BRAF alleles from five follicular adenomas varied from 3 to 5 %. Therefore, the cutoff was set at 5 %, consistent with previous reports.12,13 For 42 PTCs positive for BRAF mutation, the range of mutant allele frequency was 6–46 %, with a mean and median of 31 and 36 %, respectively. The mutant allele frequency was highly consistent between mass spectrometric genotying and pyrosequencing (Fig. 3a). One patient with a 7 % BRAF mutation allele based on pyrosequencing analysis had no mutant BRAF allele detectable by mass spectrometry (98 % concordance). The percentage of mutant BRAF alleles was strongly correlated with tumor size (Spearman’s rho = 0.52; p = 0.0001), but not age (p = 0.782) or body weight (p = 0.100). Comparing medians of the mutant allele frequency, significant differences were observed in extrathyroidal invasion (p = 0.034) and clinical nodal metastasis (p = 0.041), but not pathological nodal metastasis (p = 0.937). Tumor BRAF mutation determined by pyrosequencing was associated with higher probabilities of stage III/IV disease (p = 0.018). Mutational Analysis by castPCR Five follicular adenomas were in the range of 0.03– 1.7 % mutant allele. A conservative 2.5 % cutoff was chosen to avoid unreliable estimates due to stochastic fluctuations in low copy number situations.18 In concordance with the results obtained by pyrosequencing, 42 PTCs were positive for BRAF mutation based on castPCR assay. The range of mutant allele frequency was 3.0– 61.6 %, with a mean and median of 30.7 and 29.1 %, respectively. The mutant allele frequency was moderately correlated between mass spectrometric genotying and castPCR assay (Fig. 3b). A similar correlation coefficient was observed between pyrosequencing and castPCR (Spearman’s rho = 0.87; p \ 0.0001).
Allelic Percentage of BRAF Mutation TABLE 2 Clinicopathological parameters and BRAF c.1799T [ A mutation status determined by mass spectrometric genotyping Mutant allele 0–10 % (n = 16) Female sex Age at diagnosis (years)
Mutant allele 10–27.5 % (n = 16)
13 (81) a
Body mass index (kg/m2)a
14 (88)
37 (32–58) 23.4 (20.1–26.8)
56 (52–59) 24.0 (20.5–26.0)
Mutant allele [ 27.5 % (n = 18) 13 (72) 41 (33–50) 26.6 (23.0–29.4)
p-Value
0.615 0.053 0.191
Body weight (kg)a,b
58 (54–69)
55 (47–65)
74 (58–80)
0.023
Tumor size (cm)a,c
1.3 (1.1–2.0)
2.3 (1.5–3.2)
2.5 (2.0–3.5)
0.004
Thyroiditis
5 (31)
2 (13)
1 (6)
10 (63)
4 (25)
5 (28)
Microscopic
5 (31)
3 (19)
7 (39)
Macroscopic
1 (6)
9 (56)
6 (33)
Multifocality
4 (25)
9 (56)
4 (22)
0.111
Lymphovascular invasion
1 (6)
6 (38)
7 (39)
0.062
Clinical lymph node metastasis
2 (13)
5 (31)
9 (50)
0.070
Pathological lymph node metastasis
9 (56)
9 (56)
11 (61)
1.000
I
11 (69)
4 (25)
11 (61)
II
2 (13)
0 (0)
0 (0)
III IV
2 (13) 1 (6)
5 (31) 7 (44)
3 (17) 4 (22)
Extrathyroidal invasion None
0.140 0.024
TNM stage
0.029
Data are expressed as number (%) or median (interquartile range) a
Kruskal–Wallis test
b
Dunn’s post hoc tests: tertile 1 vs. tertile 2, p = 1.000; tertile 2 vs. tertile 3, p = 0.021; tertile 1 vs. tertile 3, p = 0.256
c
Dunn’s post hoc tests: tertile 1 vs. tertile 2, p = 0.098; tertile 2 vs. tertile 3, p = 0.855; tertile 1 vs tertile 3, p = 0.003
The percentage of mutant BRAF alleles was positively correlated with tumor size (Spearman’s rho = 0.47; p = 0.006). In addition, the mutant allele frequency was associated with extrathyroidal invasion (p = 0.010) and stage III/IV disease (p = 0.009), but not clinical or pathological nodal metastasis (p = 0.075 and 0.602, respectively). DISCUSSION The occurrence of BRAF mutation in PTC has been extensively investigated. It appears that the development of BRAF mutation in the thyroid gland is associated with environmental factors and iodine intake.19,20 The prevalence of BRAF mutation in PTC varies according to the detection methods, ranging from 27 to 90 % in the literature.9,10,21 Among the mutation detection methods, Sanger sequencing is least sensitive and capable of detecting samples having 20–30 % mutant alleles.22 Additionally, the mutation detection may be influenced by DNA extraction methods. Kim et al. 23 showed that direct DNA sequencing and pyrosequencing to detect BRAF mutation were more accurate in fresh-frozen than formalin-fixed and paraffin-embedded tissue sections. In this study, we
confirmed that Sanger sequencing may fail to detect BRAF mutation in a small portion of frozen PTC specimens which contain a low allelic fraction. BRAF mutation can be found in a substantial number of microcarcinomas.23,24 In transgenic mouse models, mice with targeted expression of BRAFV600E in thyroid cells developed PTC characteristics of the human disease.25,26 These findings indicate that BRAF mutation represents an early, initiating event in tumor development. However, discordant patterns of BRAF mutation were found in about 40 % of the multifocal PTCs.27,28 By analyzing the relationship between genome-wide allelic imbalances and BRAF mutation, Jovanovic and collaborators reported that BRAF mutation did not represent the earliest transforming event.29 Furthermore, de novo BRAF mutation was observed in metastatic lymph nodes from mutation-negative primary tumors.30 These results imply that BRAF mutation may be a late genetic change heterogeneously occurring in the tumor bulk. As such, the role of BRAF mutation in pathogenesis of thyroid cancer appears complicated and conflicting, and is being actively debated.31 Resolving these contradicting views is important to understanding the oncologic and therapeutic significance of
S.-P. Cheng et al. P = 0.037
(a)
50
(a)
Pyrosequencing (%)
Mutant BRAF (% alleles)
40
30
20
10
40 30 20 10 0 0
(b)
20
30
40
1
Mass genotyping (%)
(b)
P = 0.984 60
40
50
castPCR (%)
Mutant BRAF (% alleles)
10
cN
cN
0
0
30
20
40 30 20 10
10
0 0
1 pN
pN
0
0
FIG. 2 Vertical scatter plots of the percentage of mutant BRAF alleles grouped by a clinical and b pathological lymph node metastasis. Lines represent median values. Mann–Whitney tests were performed to evaluate the difference between groups
BRAF mutation. In the hierarchical model of clonal evolution, early somatic mutations tend to be propagated in many or all clones, whereas later events occur only in some clones.32 A 50 % percentage of mutant BRAF alleles would be the expected finding, if all PTC cells carry one mutant and one wild-type BRAF allele. To test this concept, Guerra and colleagues first demonstrated that most PTCs consist of a mixture of tumor cells carrying wild-type and mutant BRAF.33 Gandolfi et al.13 also found an average mutated allele percentage of 27.4 % with a range between 7.5 and 49.8 %. In line with their observations, our data showed that the mutant allele frequency was less than 25 % in 18 (44 %) of 41 PTCs positive for mutation. To further examine the biological relevance of clonality of BRAF mutation, we studied the correlation between the mutant allele frequency and clinicopathological parameters. Our results indicate that larger tumors and tumors with extrathyroidal invasion had a significantly higher mutant allele frequency. These findings are consistent with recent metaanalyses.5,9–11 Of note is the fact that a strong correlation existed between tumor size and mutant allele frequency. In agreement with the present observations, Guerra and
10
20
30
40
Mass genotyping (%)
FIG. 3 a Scatter plot showing the correlation of the mutant BRAF allele frequency determined by mass spectrometric genotying and pyrosequencing analysis. Spearman’s rho = 0.97, p \ 0.0001; simple regression adjusted r2 = 0.98, p \ 0.001. b Scatter plot showing the correlation of the mutant BRAF allele frequency determined by mass spectrometric genotying and competitive allele-specific TaqMan PCR (castPCR). Spearman’s rho = 0.87, p \ 0.0001; simple regression adjusted r2 = 0.85, p \ 0.001
colleagues demonstrated that tumor volume directly correlated with the mutant allele frequency.12 A reasonable interpretation of these results is that BRAF mutation provides cancer cells a survival and growth advantage, resulting in selection of BRAF mutant alleles during tumor progression. These data also refute an alternative hypothesis which proposed that BRAF mutation occurs as a primary genetic event but is later removed by the DNA repair machinery.31 In contrast to the results of previous meta-analyses,5,9–11 our data indicate that BRAF mutation was not associated with lymph node metastasis. It is worth noting that in most studies, neck dissections were performed for suspicious lymphadenopathy.9 We observed a small and inconsistent difference in the mutant allele frequency between clinically node-positive and node-negative tumors. When prophylactic central compartment dissection was performed, the mutant allele frequency was not associated with pathological nodal status. In accordance, Gandolfi et al.13 found that the average mutated allele
Allelic Percentage of BRAF Mutation
percentage was significantly lower in corresponding metastatic nodes compared with the primary PTCs (18.6 vs. 28.6 %). Therefore, it is likely that cancer cells harboring BRAF mutation do not really have enhanced metastatic capacity. More frequent lymph node metastasis observed in BRAF mutation-positive tumors may result from a larger tumor burden and more extensive extrathyroidal invasion. It was recently showed that BRAF mutation was the only independent predictor of central compartment lymph node metastasis in PTC;34 however, the design of the study was biased by including BRAF status as a criterion for prophylactic nodal dissection. A plethora of genetic and phenotypic heterogeneity comes from ongoing genetic instability within a cancer. Most of the spontaneous mutations are evolutionarily neutral or deleterious. Rare mutations that confer competitive advantage can be selected for by Darwinian forces and will eventually become predominant. Nonetheless, achieving clonal dominance can take a long time, partly because of spatial constraints.35 Our findings have important biological implications in that they point out the possibility that activating BRAF mutation may not necessarily be an initiating prerequisite in PTC tumorigenesis, but this secondary genetic alteration provides a survival advantage to cancer cells, and accumulates. The fact that the mutant allele frequency was associated with extrathyroidal invasion also makes it unlikely that expansion emerges from fixation of neutral mutations over time (genetic drift). Our results seemingly suggest that quantitative assessment of mutation frequency provides no additional information on clinical significance of BRAF mutation (Tables 1 and 2). However, even a small portion of mutation-negative tumor cells may hamper efficacy of BRAF-targeted therapy. In RAS/RAF wild-type tumors, RAF kinase inhibitors may paradoxically activate the MAPK pathway in an RAS-dependent manner, thus enhancing tumor growth in xenograft models.36 Further studies may be undertaken to determine the effects of clonality measures on clinical response to BRAF-targeted therapy. A limitation of our study is the inability to exclude stromal contamination. Tumor samples always have some degree of contamination by normal stromal, endothelial, and inflammatory cells. The mutant allele frequency in unselected cells would be lower than that in laser-captured PTC cells.33 Nonetheless, the average percentage of mutant BRAF alleles analyzed by different methods in the present study (22.5 to 31 %) was compatible with those of recent reports (21.9 to 27.4 %).12,13 Using next-generation sequencing, Nikiforova and colleagues also demonstrated that the mutant allele frequency was 18–44 % (mean 34 %) in PTCs, corresponding to 36–88 % of cells with heterozygous mutation.37
CONCLUSIONS The percentage of mutant BRAF alleles in PTC is significantly associated with tumor burden and extrathyroidal invasion. A positive correlation between the mutant allele frequency and tumor size implies that BRAF mutation may play a role in tumor progression. ACKNOWLEDGMENT The authors would like to thank the Translational Core Laboratories of National Translational Medicine and Clinical Trial Resource Center for technical assistance in MALDI-TOF MS experiments. This work was supported by the National Science Council of Taiwan (100-2314-B-195-001-MY3) and Mackay Memorial Hospital (MMH-10206 and MMH-E-102-10). DISCLOSURES Shih-Ping Cheng, Yi-Chiung Hsu, Chien-Liang Liu, Tsang-Pai Liu, Ming-Nan Chien, Tao-Yeuan Wang, and Jie-Jen Lee have nothing to disclose.
REFERENCES 1. Cramer JD, Fu P, Harth KC, Margevicius S, Wilhelm SM. Analysis of the rising incidence of thyroid cancer using the surveillance, epidemiology and end results national cancer data registry. Surgery. 2010;148:1147–52. 2. Husson O, Haak HR, van Steenbergen LN, et al. Rising incidence, no change in survival and decreasing mortality from thyroid cancer in The Netherlands since 1989. Endocr Relat Cancer. 2013;20:263–71. 3. Eustatia-Rutten CF, Corssmit EP, Biermasz NR, Pereira AM, Romijn JA, Smit JW. Survival and death causes in differentiated thyroid carcinoma. J Clin Endocrinol Metab. 2006;91:313–9. 4. Kapiteijn E, Schneider TC, Morreau H, Gelderblom H, Nortier JW, Smit JW. New treatment modalities in advanced thyroid cancer. Ann Oncol. 2012;23:10–8. 5. Xing M. BRAF mutation in papillary thyroid cancer: pathogenic role, molecular bases, and clinical implications. Endocr Rev. 2007;28:742–62. 6. Mathur A, Moses W, Rahbari R, Khanafshar E, Duh QY, Clark O, et al. Higher rate of BRAF mutation in papillary thyroid cancer over time: a single-institution study. Cancer. 2011;117:4390–5. 7. Wan PT, Garnett MJ, Roe SM, et al; Cancer Genome Project. Mechanism of activation of the RAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell. 2004;116:855–67. 8. Pratilas CA, Taylor BS, Ye Q, Viale A, Sander C, Solit DB, et al. (V600E)BRAF is associated with disabled feedback inhibition of RAF-MEK signaling and elevated transcriptional output of the pathway. Proc Natl Acad Sci USA. 2009;106:4519–24. 9. Li C, Lee KC, Schneider EB, Zeiger MA. BRAF V600E mutation and its association with clinicopathological features of papillary thyroid cancer: a meta-analysis. J Clin Endocrinol Metab. 2012;97:4559–70. 10. Kim TH, Park YJ, Lim JA, Ahn HY, Lee EK, Lee YJ, et al. The association of the BRAF(V600E) mutation with prognostic factors and poor clinical outcome in papillary thyroid cancer: a meta-analysis. Cancer. 2012;118:1764–73. 11. Tufano RP, Teixeira GV, Bishop J, Carson KA, Xing M. BRAF mutation in papillary thyroid cancer and its value in tailoring initial treatment: a systematic review and meta-analysis. Medicine (Baltimore). 2012;91:274–86. 12. Guerra A, Fugazzola L, Marotta V, Cirillo M, Rossi S, Cirello V, et al. A high percentage of BRAFV600E alleles in papillary
S.-P. Cheng et al.
13.
14. 15.
16.
17. 18.
19.
20.
21.
22.
23.
24.
25.
thyroid carcinoma predicts a poorer outcome. J Clin Endocrinol Metab. 2012;97:2333–40. Gandolfi G, Sancisi V, Torricelli F, Ragazzi M, Frasoldati A, Piana S, et al. Allele percentage of the BRAF V600E mutation in papillary thyroid carcinomas and corresponding lymph node metastases: no evidence for a role in tumor progression. J Clin Endocrinol Metab. 2013;98:E934–42. MacConaill LE. Existing and emerging technologies for tumor genomic profiling. J Clin Oncol. 2013;31:1815–24. Cheng SP, Liu CL, Hsu YC, Chang YC, Huang SY, Lee JJ. Expression and biologic significance of adiponectin receptors in papillary thyroid carcinoma. Cell Biochem Biophys. 2013;65: 203–10. Thomas RK, Baker AC, Debiasi RM, Winckler W, LaFramboise T, Lin WM, et al. High-throughput oncogene mutation profiling in human cancer. Nat Genet. 2007;39:347–51. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001;29:e45. Kloos RT, Reynolds JD, Walsh PS, Wilde JI, Tom EY, Pagan M, et al. Does addition of BRAF V600E mutation testing modify sensitivity or specificity of the Afirma Gene Expression Classifier in cytologically indeterminate thyroid nodules? J Clin Endocrinol Metab. 2013;98:E761–8. Frasca F, Nucera C, Pellegriti G, Gangemi P, Attard M, Stella M, et al. BRAF(V600E) mutation and the biology of papillary thyroid cancer. Endocr Relat Cancer. 2008;15:191–205. Guan H, Ji M, Bao R, Yu H, Wang Y, Hou P, et al. Association of high iodine intake with the T1799A BRAF mutation in papillary thyroid cancer. J Clin Endocrinol Metab. 2009;94:1612–7. Jeong D, Jeong Y, Park JH, Han SW, Kim SY, Kim YJ, et al. BRAF (V600E) mutation analysis in papillary thyroid carcinomas by peptide nucleic acid clamp real-time PCR. Ann Surg Oncol. 2013;20:759–66. Lade-Keller J, Romer KM, Guldberg P, Riber-Hansen R, Hansen LL, Steiniche T, et al. Evaluation of BRAF mutation testing methodologies in formalin-fixed, paraffin-embedded cutaneous melanomas. J Mol Diagn. 2013;15:70–80. Kim HS, Kim JO, Lee DH, Lee HC, Kim HJ, Kim JH, et al. Factors influencing the detection of the BRAF T1799A mutation in papillary thyroid carcinoma. Oncol Rep. 2011;25:1639–44. Virk RK, Van Dyke AL, Finkelstein A, Prasad A, Gibson J, Hui P, et al. BRAFV600E mutation in papillary thyroid microcarcinoma: a genotype-phenotype correlation. Mod Pathol. 2013;26:62–70. Knauf JA, Ma X, Smith EP, Zhang L, Mitsutake N, Liao XH, et al. Targeted expression of BRAFV600E in thyroid cells of
26.
27.
28.
29.
30.
31. 32. 33.
34.
35. 36.
37.
transgenic mice results in papillary thyroid cancers that undergo dedifferentiation. Cancer Res. 2005;65:4238–45. Charles RP, Iezza G, Amendola E, Dankort D, McMahon M. Mutationally activated BRAF(V600E) elicits papillary thyroid cancer in the adult mouse. Cancer Res. 2011;71:3863–71. Park SY, Park YJ, Lee YJ, Lee HS, Choi SH, Choe G, et al. Analysis of differential BRAF(V600E) mutational status in multifocal papillary thyroid carcinoma: evidence of independent clonal origin in distinct tumor foci. Cancer. 2006;107:1831–8. Giannini R, Ugolini C, Lupi C, Proietti A, Elisei R, Salvatore G, et al. The heterogeneous distribution of BRAF mutation supports the independent clonal origin of distinct tumor foci in multifocal papillary thyroid carcinoma. J Clin Endocrinol Metab. 2007;92:3511–6. Jovanovic L, Delahunt B, McIver B, Eberhardt NL, Grebe SK. Most multifocal papillary thyroid carcinomas acquire genetic and morphotype diversity through subclonal evolution following the intra-glandular spread of the initial neoplastic clone. J Pathol. 2008;215:145–54. Vasko V, Hu S, Wu G, Xing JC, Larin A, Savchenko V, et al. High prevalence and possible de novo formation of BRAF mutation in metastasized papillary thyroid cancer in lymph nodes. J Clin Endocrinol Metab. 2005;90:5265–9. Xing M. BRAFV600E mutation and papillary thyroid cancer: chicken or egg? J Clin Endocrinol Metab. 2012;97:2295–8. Aparicio S, Caldas C. The implications of clonal genome evolution for cancer medicine. N Engl J Med. 2013;368:842-51. Guerra A, Sapio MR, Marotta V, Campanile E, Rossi S, Forno I, et al. The primary occurrence of BRAF(V600E) is a rare clonal event in papillary thyroid carcinoma. J Clin Endocrinol Metab. 2012;97:517–24. Howell GM, Nikiforova MN, Carty SE, Armstrong MJ, Hodak SP, Stang M T, et al. BRAF V600E mutation independently predicts central compartment lymph node metastasis in patients with papillary thyroid cancer. Ann Surg Oncol. 2013;20:47–52. Marusyk A, Polyak K. Tumor heterogeneity: causes and consequences. Biochim Biophys Acta. 2010;1805:105–17. Hatzivassiliou G, Song K, Yen I, Brandhuber BJ, Anderson DJ, Alvarado R, et al. RAF inhibitors prime wild-type RAF to activate the MAPK pathway and enhance growth. Nature. 2010;464:431–5. Nikiforova MN, Wald AI, Roy S, Durso MB, Nikiforov YE. Targeted next-generation sequencing panel (ThyroSeq) for detection of mutations in thyroid cancer. J Clin Endocrinol Metab. 2013;98:E1852–60.
