Gouveia C, Can NT, Bostrom A, Grenert JP, van Zante A, Orloff LA. Lack of Association of BRAF Mutation With Negative Prognostic Indicators in Papillary Thyroid CarcinomaThe University of California, San Francisco, Experience. JAMA Otolaryngol Head Neck Surg. 2013;139(11):1164-1170. doi:10.1001/jamaoto.2013.4501
Copyright 2013 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Papillary thyroid carcinoma (PTC) is the most common endocrine neoplasm. B-type raf kinase (BRAF) V600E mutation has been proposed as a negative prognostic indicator in PTC, and patients harboring it should receive more aggressive initial therapy.
To assess the significance of BRAF V600E mutation in PTC in the largest US sample to date.
We identified patients from our institution’s pathology archives diagnosed as having PTC and meeting criteria for BRAF mutation testing. Medical records were analyzed for BRAF status (positive or negative) and a list of standardized clinicopathologic features.
A total of 429 patients with PTC at an academic medical center.
Main Outcomes and Measures
Clinicopathologic features in patients with PTC with and without BRAF mutation.
Of 429 cases with PTC, 314 (73.2%) were positive for the BRAF mutation and 115 (26.8%) tested negative. BRAF mutation was significantly associated with tumor margin positivity (P = .03) and lymph node metastasis (P = .002) on univariate analysis but not on multivariate study. BRAF mutation was a predictor of male sex (odds ratio [OR], 3.2; 95% CI, 1.4-7.2), total thyroidectomy (OR, 2.6; 95% CI, 1.1-6.2), and a negative predictor of follicular variant PTC (OR, 0.1; 95% CI, 0.1-0.4). There was no significant association between BRAF positivity and tumor multicentricity, lymphovascular invasion, extranodal extension, central neck involvement, advanced stage (stage III or IV), and distant metastasis.
Conclusions and Relevance
BRAF V600E mutation has been extensively studied in relation to negative prognostic indicators in PTC, with no consistent relationship emerging. Two recent meta-analyses showed an overall association between BRAF status and aggressive disease features and called for tailoring treatment plans in patients accordingly. In this, the largest US study to date, BRAF status was not significantly associated with most clinicopathologic features suggestive of more aggressive disease.
Papillary thyroid carcinoma (PTC) is the most common endocrine neoplasm, accounting for 88% of all thyroid malignant neoplasms.1 Although treatment is often curative, there is a 15% recurrence rate over the course of 10 years, and approximately 10% of patients die as a result of disease progression.2,3 Furthermore, the incidence of PTC is increasing across all demographics.4
The association of B-type raf (BRAF) kinase V600E mutation to PTC has been extensively studied. Multiple individual publications have shown correlation of BRAF mutation with known negative prognostic clinicopathologic features, including advanced age at diagnosis, male sex, larger tumor size, extrathyroidal extension, tumor multifocality, advanced TNM stage, lymph node metastasis, and recurrence.5- 11 In addition, BRAF mutation has been reported to be associated with overall increased mortality.12 However, the results across studies have been inconsistent, and several similar large retrospective trials have failed to corroborate these results.13- 17 The overall significance of BRAF mutation in PTC therefore remains debatable.18
Two recent meta-analyses have shown an overall correlation of BRAF with extrathyroidal extension, lymph node metastasis, advanced TNM stage, and recurrent disease.2,19 There remains no definitive correlation between BRAF V600E mutation and the clinicopathologic features of PTC, but these studies add support to the consensus that BRAF mutation is an overall negative prognostic indicator and call for consideration of tailoring initial treatment accordingly. The usual limitations of meta-analysis, along with wide variation in BRAF mutation rates and heterogeneity in geographic subgroups among the included studies, restrict the translation of these results. There is a continuing need for large, nation-wide studies of the relationship between BRAF mutation and PTC.
The aim of this study was to summarize our institution’s experience with BRAF mutation in PTC to provide the largest US study to date. To accomplish this, we analyzed a spectrum of clinical and pathologic features in PTC tumors with and without BRAF V600E mutation.
This retrospective study was approved by our institutional review board. The Department of Pathology at the University of California, San Francisco, began routine BRAF V600E testing of all PTC tumor specimens 0.4 cm or larger in April 2009. We identified 439 patients as having had BRAF testing of their thyroid specimens. We included all patients with PTC who underwent partial or total thyroidectomy with or without lymph node sampling or neck dissection between April 2009 and April 2012. We excluded 10 patients for the following reasons: 4 for incomplete medical records, 3 for initial pathological results that were performed at an outside hospital, 2 for unknown primary tumor, and 1 for recurrent disease.
All patient medical records were reviewed for BRAF status, age, sex, type of surgery, PTC subtype, multicentricity, extrathyroidal extension, lymphovascular invasion, tumor margin status, lymph node status (including central neck involvement), extranodal extension, distant metastasis, and final TNM stage.
All specimens were formalin fixed and paraffin embedded. Areas of PTC with at least 50% tumor cells were identified and circled on hematoxylin-eosin–stained slides by a pathologist. These areas were scraped from corresponding unstained slides for DNA extraction using either the QIAamp DNA Mini Kit or EZ1 automated extraction system (both, Qiagen). A portion of BRAF exon 15 encompassing codon 600 was amplified by polymerase chain reaction (PCR) with specific primers, and codon 600 was analyzed with fluorescence-labeled hybridization probes in a real-time LightCycler 480 PCR (Roche Applied Science) melting curve assay following the protocol of Rowe et al.20 A melting temperature of approximately 65°C corresponds with the wild-type sequence, while melting at approximately 60°C indicates the T to A transversion at nucleotide 1799 that results in the V600E mutation. This assay was validated to have sensitivity for V600E mutation detection down to a minimum of at least 25% tumor cells in the specimen.
Associations of demographic and clinical variables with BRAF were analyzed using Fisher exact test for nominal variables and a Mann-Whitney nonparametric test for ordinal and continuous variables. All P values were 2-tailed, and P < .05 was considered statistically significant. Variables associated with BRAF V600E mutation at P < .10 were used in a multivariate logistic regression model for BRAF mutation positivity. All computations were performed using SAS statistical software (version 9.2).
A total of 429 patients with PTC from our institution met inclusion criteria. A total of 314 (73.2%) were positive for BRAF, and 115 (26.8%) tested negative. Results of demographic features (Table 1), univariate analysis (Table 2), and multivariate analysis (Table 3) are shown.
There was no significant association of BRAF mutation with patient age (P = .91). A higher proportion of male patients was found to be BRAF positive compared with female patients (80.4% vs 70.9%), but this was not significant (P = .07). On multivariate analysis, male sex was found to be a predictor of positive BRAF status (odds ratio [OR], 3.2; 95% CI, 1.4-7.2). Patients who underwent total thyroidectomy were more likely to test positive for BRAF than those who received hemithyroidectomy (77.1% v. 50.0%; P < .001), and total thyroidectomy was a significant predictor of BRAF mutation (OR, 2.6; 95% CI, 1.1-6.2).
Classic-type PTC was significantly associated with BRAF mutation (P < .001) vs follicular variant and other, less common subtypes. Follicular variant was a predictor of negative status (OR, 0.1; 95% CI, 0.1-0.4). When examining patients with classic-type PTC in our population, smaller tumor size was the only clinicopathologic feature significantly associated with BRAF V600E mutation (Table 4). There was no correlation between BRAF mutation and tumor size (P = .15).
The BRAF mutation-positive group had a significant association with tumor margin positivity (P = .03) and a trend toward increased extrathyroidal extension (P = .06). In multivariate models, these were not significantly associated with BRAF status. There was no significant association between tumor multicentricity (P = .59) or lymphovascular invasion with BRAF V600E mutation (P = .27 and P = .64, respectively).
BRAF status had no association with central neck dissection (P = .72) or with positive central compartment lymph nodes (P = .63). Lymph node metastasis was significantly associated with BRAF mutation (P = .002). This association did not remain significant in multivariate analysis (OR, 1.4; 95% CI, 0.7-2.6). There was no significant relationship with BRAF status and extranodal extension (P = .32). There was no significant association between distant metastasis or advanced-stage disease (stage III and IV) and BRAF positivity.
The significance of BRAF mutation in the management of PTC remains unclear. Several articles have shown a significant association between poor prognostic variables and BRAF status,5- 12 but the results are inconsistent with some studies showing a lack thereof.13- 17 Despite this, 2 recent meta-analyses2,19 that showed an overall relationship between BRAF mutation and high-risk features concluded that steps should be taken to modify initial treatments in BRAF-positive patients with PTC.
Given this controversy, the purpose of our study was to provide a large, single-institution sampling to evaluate the association of BRAF mutation and poor prognostic variables in PTC. To our knowledge, this is the largest US study examining this relationship. Our results showed a lack of association between BRAF mutation status and most clinicopathologic risk factors studied. Only male sex, increased rate of total thyroidectomy, and classic-type PTC were significantly associated with positive BRAF status on multivariate analysis.
Male sex is an established poor prognostic factor in PTC19 and was found to be a predictor of positive BRAF status on multivariate analysis in our study. This association is inconsistent with findings in prior studies, with many articles unable to establish a relationship between patient sex and the presence of BRAF mutation.6,10,12,14,15,21- 23 There was a significant association between BRAF V600E positivity and receiving a total vs subtotal thyroidectomy in our patient population. The cause for this association is unclear but may be due to BRAF mutation being associated with higher-risk fine-needle aspiration features or neck ultrasonographic findings that would sway a surgeon toward definitive surgery; we did not examine these factors in our study. We also found a significant association between BRAF mutation and classic-type PTC, whereas follicular variant PTC was found to be a negative predictor of BRAF mutation status on multivariate analysis. This association between BRAF mutation and PTC variants is well documented and highlights the importance of classifying subtype in the study and management of PTC.7,10,13- 15,22,23 We conducted a separate analysis of patients with classic-type PTC only and found no significant differences from our overall study population. Our study included 59 patients with nonclassic PTC. While this could contribute to our study’s overall findings, it is a small percentage of our sample and is accounted for in the multivariate analysis.
There are several possible explanations for the difference between our findings and those of prior studies showing significant association of BRAF mutation and negative prognostic factors in PTC. We analyzed data for 429 patients, the largest study in the United States and one of the largest internationally, which could show differences vs smaller individual studies with less power. BRAF-positive patients accounted for 74% of our study population. While this is within the range of prior studies, it is on the higher end of the spectrum and likely represents a heterogeneous population. Our sample is from 1 region of the United States, in contrast with prior studies, most of which are international. However, the patient population in our study is ethnically diverse. Genetic, dietary, and other environmental differences have been hypothesized as accounting for differences in incidence of BRAF mutation, although we did not analyze specific ethnic factors in this study. In the meta-analysis by Kim et al,2 subgroup analysis by country and BRAF mutation rates showed the association between BRAF mutation and negative prognostic variables remained true overall but could be significantly affected by each, strengthening our study’s findings.
Our study did not evaluate disease recurrence and mortality association. Papillary thyroid carcinoma is a slowly progressive, often curable disease, so obtaining long-term follow-up data is an area in need of further study. The role and value of central neck dissection in PTC is still being investigated, and its relationship to BRAF mutation has been inconsistent.11,24,25 Our study found no significant association between central neck dissection rates or positive central neck lymph nodes with BRAF mutation status.
In conclusion, the results of the present study show little correlation between BRAF V600E mutation and negative prognostic indicators in PTC. If anything, the absence of BRAF mutation may prove to be a favorable prognostic indicator and a basis for deescalation of therapy, such as more selective use of radioactive iodine or lymph node dissection, in patients with PTC. Such data have not yet been adequately studied. Certainly, efforts to recommend more aggressive initial therapies to BRAF-positive patients with PTC should be tempered until more long-term data related to outcome are available.
Submitted for Publication: February 27, 2013; final revision received May 16, 2013; accepted June 21, 2013.
Corresponding Author: Lisa A. Orloff, MD, Department of Otolaryngology–Head and Neck Surgery, University of California, San Francisco, 2380 Sutter St, Second Floor, San Francisco, CA 94115 (email@example.com).
Published Online: September 12, 2013. doi:10.1001/jamaoto.2013.4501.
Author Contributions: Dr Orloff had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Gouveia, Can, van Zante, Orloff.
Acquisition of data: Gouveia, Can, Grenert, van Zante, Orloff.
Analysis and interpretation of data: Gouveia, Can, Bostrom, Grenert, Orloff.
Drafting of the manuscript: Gouveia, Can, Bostrom.
Critical revision of the manuscript for important intellectual content: Gouveia, Grenert, van Zante, Orloff.
Statistical analysis: Gouveia, Bostrom.
Administrative, technical, or material support: Grenert, Orloff.
Study supervision: van Zante, Orloff.
Conflict of Interest Disclosures: None reported.
Previous Presentation: This study was presented at the American Head and Neck Society 2013 Annual Meeting; April 11, 2013; Orlando, Florida.