The association of the BRAFV600E mutation with prognostic factors and poor clinical outcome in papillary thyroid cancer

A meta-analysis

Authors


Abstract

BACKGROUND:

The effects of the BRAFV600E mutation on prognostic factors and poor clinical outcomes in papillary thyroid cancer (PTC) have not been fully quantified. The authors performed comprehensive meta-analysis to assess the strength of associations between these conditions and the BRAFV600E mutation.

METHODS:

The authors identified the clinical studies that examined the association of the BRAFV600E mutation in surgical specimens with clinicopathologic outcomes between January 2003 and October 2010 using the Medline database. One hundred thirty-one relevant studies were hand-searched. The authors selected 27 studies that included 5655 PTC patients. They calculated the pooled odds ratios (ORs) or risk ratios with 95% confidence intervals (CIs) for each study using a random effect model.

RESULTS:

The average prevalence rate of the BRAFV600E mutation was 49.4%. In 26 studies, compared with the patients who had the wild-type BRAF genes, the PTC patients with the BRAFV600E mutation had increased ORs of an extrathyroidal invasion (OR, 2.14; 95% CI, 1.68-2.73), a lymph node metastasis (OR, 1.54; 95% CI, 1.21-1.97), and an advanced TNM stage (OR, 2.00; 95% CI, 1.61-2.49). In 8 studies, patients with the mutation had 2.14-fold increased risk of recurrent and persistent disease (95% CI, 1.67-2.74). The associations were generally consistent across the different study populations.

CONCLUSIONS:

This meta-analysis demonstrates that the BRAFV600E mutation is closely related to the high-risk clinicopathological factors and poorer outcome of PTC. The results obtained here suggest that the BRAFV600E mutation should be considered as a poor prognostic marker in PTC and may lead to better management for individual patients. Cancer 2012;. © 2011 American Cancer Society.

INTRODUCTION

Differentiated thyroid cancer is the most common endocrine malignancy, and its prevalence is increasing worldwide.1 A recent study using the National Cancer Institute's Surveillance Epidemiology and End Results database found increased incidence across all tumor sizes during the time period between 1988 and 2005 in both sexes and that this increase is not only a result of increased diagnostic activity.2

The majority of the incidences of thyroid cancer are papillary thyroid cancer (PTC), which accounts for approximately 85%.3 Although most PTC patients have excellent overall survival, it is important to note that there is a 15% recurrence rate for the 10-year period after the initial treatment even in TNM stage I,4 and that mortality is possible in cases of asymptomatic small PTC.5, 6 With the rising trend in the worldwide incidence of this type of cancer, the need to identify those tumors that pose the greatest risk to patients is great.

The B-type raf proto-oncogene (BRAF) V600E mutation is a representative oncogenic mutation that occurs in approximately 45% of PTC cases. This mutation has emerged as a promising prognostic variable of PTC and can be obtained in aspiration specimens preoperatively.7 Since its initial description in PTC, various authors have reported associations of the BRAFV600E mutation with aggressive clinical course and poor clinicopathologic features, such as extrathyroidal invasion, lymph node metastasis, and advanced TNM stage.8, 9 However, the clinical significance of this mutation in terms of a prognostic marker in PTC has been controversial,3 because several authors reported that this mutation was not correlated with poor clinical outcome or with pathologic aggressiveness.10-12

In the present study, we performed a comprehensive meta-analysis to quantify the associations of the BRAFV600E mutation with high-risk clinicopathological factors and prognostic outcomes. In addition, we conducted subgroup analyses to assess the effects of the factors that might modify this association.

MATERIALS AND METHODS

Selection Criteria and Search Strategy

We extensively searched for studies that examined the association of the BRAFV600E mutation with high-risk clinicopathological factors and prognostic outcome. The investigations that were included in this study are described below. The BRAFV600E mutation data from the primary PTC tissues were included. The presence of an extrathyroidal invasion, a lymph node metastasis in the surgical pathology, an advanced TNM stage,13 any cancer recurrence, persistent disease at last follow-up, and cancer-specific death were chosen as the measures of the outcome.

The studies with no clinical-pathologic data were excluded. The studies that included the BRAFV600E analysis in patients with preoperative fine needle aspiration biopsies were excluded to avoid false-negative results for the cytologic specimens. The studies confined to low-risk groups or tumor size <2 cm were excluded to avoid a disproportional inclusion of the low-risk cases. Abstracts and single case reports were excluded.

We conducted an electronic search of the papers in the Medline database from January 2003 to August 2010. The search was restricted to English language publications. Search term combinations were “BRAF,” “V600E,” “T1799A,” “papillary thyroid cancer,” and “PTC.” All of the reference lists from the main reports and relevant reviews were inspected for additional eligible studies. The relevant unpublished data that were presented at international meetings were also included. One clinician investigator (T.H.K.) reviewed each full-text report for eligibility, extracted data from included studies, and performed meta-analysis. Duplications of data were carefully avoided by examining the names of all authors and medical centers that were involved in each publication. If there were >1 report relating to the same centers, the report with the information most relevant to our analysis was included. We contacted the authors for additional tabular data when necessary. All of the procedures conformed to the guidelines for the meta-analysis of observational studies in epidemiology.14

Data Analyses and Statistical Methods

We used RevMan (version 5) to calculate the summary odds ratios (ORs) or risk ratios (RRs) with 95% confidence intervals (CIs), using a random effects model for all of the analyses. We assessed the heterogeneity of the studies using the chi-square test of heterogeneity and the I2 measure of inconsistency. Significant heterogeneity was defined as a chi-square test P value of <.10 or as an I2 measure >50%.15 The extent to which the combined metarisk or heterogeneity were affected by individual studies was assessed further by successively excluding every study from the meta-analysis. We investigated potential sources of the identified heterogeneity among the studies using a stratification process according to the country in which the research was conducted, the mean patient age, the number of patients, and the prevalence rate of the BRAFV600E mutation. In these analyses, we used 99% CIs to reduce the potential of chance differences arising from multiple testing. The potential for publication bias was assessed using a funnel plot analysis.

RESULTS

Results of the Search and Study Characteristics

Figure 1 shows the study selection process. A total of 462 abstracts and titles were obtained through electronic searches. Of these abstracts and papers, 124 full-text papers were deemed to be relevant and were examined in detail. An additional 6 reports and 1 unpublished study from our institution presented at the Ninth Asia and Oceania Thyroid Association Congress were also included. After this review, 27 studies met the inclusion/exclusion criteria,11, 16-40 and these studies contributed 5655 patients with PTC to the meta-analysis. The main features of the 26 eligible studies that investigated prognostic factors are summarized in Table 1. Of these studies, 18 studies evaluated extrathyroidal invasions. Data regarding lymph node metastasis and TNM stages were reported in 25 and 22 studies, respectively. The 8 studies that investigated recurrent and persistent disease are summarized in Table 2.11, 20, 26, 32, 36-39 The funnel plots for each outcome did not suggest the presence of a publication bias (data not shown).

Figure 1.

The study selection process is shown. PTC, papillary thyroid cancer.

Table 1. Summary of the 26 Included Studies That Compared the BRAFV600E Mutation With Clinical and Pathologic Risk Factors of PTC
StudyCountryPTC Patients, No.% FemaleAge, Mean y (SD)BRAFV600E Mutation Rate, %
  • Abbreviations: N/A, not available; PTC, papillary thyroid cancer; SD, standard deviation.

  • a

    As the raw data were not directly reported in the original publication, we used the tabulated data of this study from the recent review of Dr. Xing9 (Tables 1, 3).

  • b

    Unpublished information from our institution (presented at the Ninth Asia and Oceania Thyroid Association Congress).

Xu 200340USA5675N/A37.5
Namba 200334Japan17076N/A28.8
Fugazzola 200411Italy5677N/A32.1
Liu 200530Taiwan10570<45, 58%46.7
Xing 200539USA, Italy, Ukraine21974N/A48.9
Kim 200528Korea7986<45, 54%81.0
Jo 200625Korea161834563.4
Adeniran 200618USA966739 (18)41.7
Kim 200627USA103795333.0
Riesco-Eizaguirre 200636Spain677543 (14)41.8
Lee 200629Korea100904858.0
Kebebew 200726aUSA209734651.0
Durante 200719Italy93715060.2
Lupi 200731Italy5007545 (14)43.8
Abrosimov 200716Russia40854757.5
Nakayama 200733Japan40735265.0
Frasca 200821Italy323834638.6
Goutas 200822Greece558244 (13)27.3
Abubaker 200817Saudi Arabia29675>45, 39%51.7
Oler & Cerutti 200935Brazil1208445 (14)48.3
Guan 200923China1032844461.9
Ito 200924Japan631895138.4
Kim 2009bKorea4098446 (13)68.2
Rivera 201037USA6175>45, 64%70.5
Musholt 201032Germany2906648 (18)42.1
Sykorova 201038Czech Republic242814733.5
Table 2. Summary of the 8 Included Studies That Compared the BRAFV600E Mutation With Recurrent and Persistent Disease
StudyCountryNo. of Follow-up CasesBRAFV600E mutation Rate, %Median Follow-up, yStage of DiseaseInitial TreatmentPoorer OutcomeConfirmation Method
  • Abbreviations: N/A, not available; PET, positron emission tomography; RAI, radioactive iodine; Tg, thyroglobulin; TT, total thyroidectomy; USG, ultrasonography.

  • a

    As the raw data were not directly reported in the original publication, we used the tabulated data of this study from the recent review of Dr. Xing9 (Tables 1, 3).

Fugazzola 200411Italy4738.3N/AI/II, 66%N/ARecurrenceN/A
Xing 200539USA, Italy, Ukraine18848.91.3I/II, 78%TT or near TT and/or RAI ablationRecurrenceRadioiodine scan, Tg, and/or pathology
Riesco-Eizaguirre 200636Spain6741.83.0I/II, 42%TT and RAI ablation and T4 suppressionRecurrencePathology, Tg, radioiodine scan, and/or abnormal imaging studies
Kebebew 200726aUSA20951.06.0I/II, 83%Thyroidectomy and/or RAI ablationRecurrence and persistent diseaseN/A
Elisei 200820Italy10237.315.0I/II, 80%TT and RAI ablationPersistent disease including cancer-specific deathUSG, radioiodine scan, and Tg
Rivera 201037USA5770.5N/AN/AThyroidectomy and/or RAI ablationPersistent diseaseTg, radioiodine, or PET scan
Sykorova 201038Czech Republic20033.5N/AI/II, 77%N/ARecurrenceN/A
Musholt 201032Germany25142.15.5 (mean)T1, 15%; N0, 33%; M0, 88%N/APersistent diseaseTg, radioiodine, or PET scan

Meta-Analysis of the BRAFV600E Mutation Effects on Prognostic Factors

In each category of prognostic factors, the pooled ORs were higher in the patients with the BRAFV600E mutation than in those with the wild-type gene. For extrathyroidal invasions (Fig. 2), the average OR from 18 studies was 2.14 (95% CI, 1.68-2.73). The heterogeneity of the data was significant (P = .002), and the I2 estimate of the variance between the studies was 56%.

Figure 2.

Random effect model of the odds ratios (ORs) with 95% confidence intervals (CIs) of the presence of an extrathyroidal invasion associated with the BRAFV600E mutation is shown. The x-axis is a log scale. Each solid square represents an OR. The horizontal lines indicate the 95% CIs. IV, inverse variance; wt, wild type.

For the cases of lymph node metastasis (Fig. 3), 25 studies were included in the meta-analysis. Despite the wide range of variance among the studies, the overall OR was 1.54 (95% CI, 1.21-1.97). The heterogeneity of the data was significant (P < .00001), and the I2 estimate of the variance between the studies was 68%. For the presence of an advanced TNM stage (Fig. 4), the overall OR from 22 studies was 2.00 (95% CI, 1.61-2.49). The chi-square test and I2 test of heterogeneity revealed a considerable level of heterogeneity in the risk estimates (P = .005, I2 = 49%).

Figure 3.

Random effect model of the odds ratios (ORs) with 95% confidence intervals (CIs) of the presence of lymph node metastasis associated with the BRAFV600E mutation is shown.

Figure 4.

Random effect model of the odds ratios (ORs) with 95% confidence intervals (CIs) of the advanced TNM stages (III and IV) associated with the BRAFV600E mutation is shown.

Subgroup Analyses of the BRAFV600E Mutation Effects on the Prognostic Factors

To investigate the potential sources of heterogeneity that might modify effects of the BRAFV600E mutation on prognostic factors, we performed subgroup analyses according to the patient's country, the mean patient age, the number of cases, and the prevalence rate of the BRAFV600E mutation (Table 3). The effect estimates were broadly consistent in the subgroups that were analyzed. There was a significant level of heterogeneity in the effect estimates regarding the presence of extrathyroidal invasion and lymph node metastasis when the studies were grouped according to the BRAFV600E mutation rate. The group that had a high prevalence of the BRAFV600E mutation tended to have a smaller metarisk of extrathyroidal invasion and lymph node metastasis than the group with a lower prevalence of this mutation. When the studies that were included in the present study were stratified by country, there was a significant level of heterogeneity in the effect estimates of lymph node metastasis and advanced TNM stages.

Table 3. Subgroup Analyses Grouped by Country, Number of Patients, Patient Age, and BRAFV600E Mutation Rate in Terms of the Presence of the Following: Extrathyroidal Invasion, Lymph Node Metastasis, and Advanced TNM Stage
SubgroupTotal No. of PTC PatientsOR (99% CI)I2, %Chi-Square Test of Heterogeneity [P]
  1. Abbreviations: CI, confidence interval; N/A, the heterogeneity in this case was not applicable because 1 study was analyzed; OR, odds ratio; PTC, papillary thyroid cancer.

Extrathyroidal invasion
 Country  236.46 on 5 df [.26]
  Europe, 5 studies11902.48 (1.76-3.50)7013.32 on 4 df [.010]
  China, Taiwan, 2 studies9001.84 (1.06-3.20)00.91 on 1 df [.34]
  Korea, 3 studies6652.14 (1.39-3.30)00.40 on 2 df [.82]
  Japan, 2 studies1662.16 (0.86-5.45)622.61 on 1 df [.11]
  USA, 3 studies3671.92 (0.99-3.76)727.03 on 2 df [.03]
  Others, 2 studies4101.37 (0.81-2.31)805.08 on 1df [.02]
 No. of PTC patients  01.24 on 2 df [.54]
  <100, 5 studies3572.69 (1.42-5.11)6812.51 on 4 df [.01]
  100-200, 5 studies6021.93 (1.23-3.02)02.75 on 4 df [.60]
  >200, 8 studies29582.10 (1.66-2.65)6922.31 on 7 df [.002]
 Mean patient age  473.77 on 2 df [.15]
  <45 years, 5 studies12382.48 (1.63-3.76)03.93 on 4 df [.42]
  ≥45 years, 8 studies18762.22 (1.69-2.91)6218.62 on 7 df [.009]
  Not recorded, 5 studies8031.65 (1.11-2.45)6812.49 on 4 df [.01]
 BRAFV600E mutation rate  899.14 on 1 df [.003]
  <50%, 9 studies17552.68 (2.02-3.55)2911.34 on 8 df [.18]
  ≥50%, 9 studies21621.68 (1.28-2.22)5618.33 on 8 df [.02]
Lymph node metastasis
 Country  6514.16 on 5 df [.01]
  Europe, 9 studies15801.73 (1.25-2.40)5216.61 on 8 df [.03]
  China, Taiwan, 2 studies9001.41 (0.96-2.07)693.26 on 1 df [.07]
  Korea, 4 studies7471.02 (0.67-1.56)00.53 on 3 df [.91]
  Japan, 3 studies7971.15 (0.76-1.74)594.87 on 2 df [.09]
  USA, 4 studies4652.34 (1.39-3.95)8317.65 on 3 df [.0005]
  Others, 2 studies3921.53 (0.88-2.65)9010.01 on 1 df [.002]
 No. of PTC patients  162.38 on 2 df [.30]
  <100, 9 studies5651.15 (0.70-1.90)2510.60 on 8 df [.23]
  100-200, 6 studies7111.60 (1.04-2.45)8635.49 on 5 df [<.00001]
  >200, 10 studies38241.58 (1.30-1.92)7131.22 on 9 df [.0003]
 Mean patient age  00.59 on 2 df [.75]
  <45 years, 6 studies12991.56 (1.13-2.14)509.99 on 5 df [.08]
  ≥45 years, 12 studies29091.57 (1.25-1.97)7645.39 on 11 df [<.00001]
  Not recorded, 7 studies8921.38 (0.95-2.02)7523.73 on 6 df [.0006]
 BRAFV600E mutation rate  888.00 on 1 df [.005]
  <50%, 15 studies29011.81 (1.44-2.28)7657.84 on 14 df [<.00001]
  ≥50%, 10 studies21991.26 (0.99-1.60)3513.85 on 9 df [.13]
Advanced TNM staging    
 Country  6815.57 on 5 df [.008]
  Europe, 8 studies13192.34 (1.67-3.30)3510.76 on 7 df [.15]
  China, Taiwan, 2 studies9002.06 (1.32-3.21)00.76 on 1 df [.38]
  Korea, 3 studies6561.49 (0.94-2.39)01.16 on 2 df [.56]
  Japan, 3 studies7971.35 (0.92-1.99)778.56 on 2 df [.01]
  USA, 4 studies4653.23 (1.67-6.25)253.98 on 3 df [.26]
  Others, 1 study2761.34 (0.68-2.63)N/AN/A
 No. of PTC patients  01.67 on 2 df [.43]
  <100, 8 studies4772.47 (1.38-4.40)5314.89 on 7 df [.04]
  100-200, 5 studies5911.76 (1.04-2.99)01.75 on 4 df [.78]
  >200, 9 studies35641.83 (1.49-2.25)6523.11 on 8 df [.003]
 Mean patient age  112.24 on 2 df [.33]
  <45 years, 5 studies11792.24 (1.51-3.34)6110.29 on 4 df [.04]
  ≥45 years, 11 studies26281.85 (1.47-2.34)6225.98 on 10 df [.004]
  Not recorded, 6 studies8251.62 (1.07-2.44)02.91 on 5 df [.71]
 BRAFV600E mutation rate  00.00 on 1 df [.97]
  <50%, 13 studies25371.87 (1.47-2.38)5627.18 on 12 df [.007]
  ≥50%, 9 studies20951.88 (1.43-2.48)4414.23 on 8 df [.08]

Meta-Analysis of the BRAFV600E Mutation Effect on Recurrent and Persistent Disease

For recurrent and persistent disease, 9 studies were initially included in the meta-analysis. Heterogeneity was present (P = .001, I2 = 48%), mainly from the study by Ito el al,24 which had important confounding factors effecting the evaluation of the effect of BRAFV600E mutation on recurrence. A significant portion (41.5%) of the included cases underwent partial thyroidectomy rather than total or near-total thyroidectomy, and only 11 patients of 631 received radioiodine ablation as initial treatment.41 On exclusion of this study (Fig. 5), the heterogeneity was diminished (P = .43, I2 = 0%). The overall RR of developing recurrent and persistent disease from the 8 remaining studies including 1121 follow-up cases was 2.14 (95% CI, 1.67-2.74).

Figure 5.

Random effect model of the risk ratios with 95% confidence intervals (CIs) of the recurrent and persistent disease associated with the BRAFV600E mutation is shown.

DISCUSSION

In the present study, we confirmed that the BRAFV600E mutation of PTC was related to the high-risk clinicopathological factors of PTC and poor clinical outcome. To assess the strength of association of the BRAFV600E mutation with adverse clinical and pathological outcomes, we performed a meta-analysis of 27 studies that evaluated 5655 patients. Overall, our results showed that PTC patients with the BRAFV600E mutation exhibited a 1.5- to 2.1-fold increase in the risk of recurrent and persistent disease as well as extrathyroidal extension, lymph node metastasis, and higher TNM stages compared with patients with the wild-type form of the BRAF gene. The present study's results are notable because of the inclusion of a very large number of subjects comprising a diverse population and the use of a careful selection process that allowed us to avoid overlapping cases.

There are various clinicopathological risk factors in PTC related to recurrence and cancer death. Tumor factors include certain histologic variants, such as tall cell variant, large tumor size, the presence of lymph node metastasis, extrathyroidal extension, and distant metastasis.5, 42, 43 Patient variables include old age at initial presentation, presence in children aged ≤10 years, and male sex.5 Among these clinicopathological risk factors and staging systems to prognosticate outcomes, we chose extrathyroidal invasion, lymph node metastasis, and advanced TNM stage, because they are considered to be reliable predictors for poor prognosis.5, 6, 9

The prognostic value of the BRAFV600E mutation has been tested in many studies, and variable results have been obtained related to this mutation's association with traditional risk factors for recurrent disease or higher mortality rate of PTC patients. Recent, careful analyses have reported significant correlations of this mutation with the presence of an extrathyroidal invasion, nodal metastasis, advanced TNM stage, and higher risk of recurrent and persistent disease.26, 31 Summary data in a review article, which was published in 2007, showed significant associations of this mutation with aggressive tumor phenotypes and recurrent PTC.9 By comparison with the previous review covering a wide range of topics, the current meta-analysis provides new insights into the prognostic importance of the BRAFV600E mutation in PTC and includes studies published after 2007. In a short meta-analysis that was also published in 2007, Lee et al reported that this mutation was associated with a 2-fold increase in the risk of extrathyroidal invasions and higher tumor stages. Contrary to our study, the BRAFV600E mutation did not show statistically significant association with lymph node metastasis. This might be explained by the difference in size of included cases between the Lee et al study and the current study (the number of analyzed cases regarding lymph node metastasis was 1159 vs 4986, respectively), although the effect size was similar (OR, 1.50; 95% CI, 0.99-2.27; P = .055 vs OR, 1.54; 95% CI, 1.21-1.97; P < .00001, respectively).44

Several mechanisms are involved in the aggressive phenotype of PTC that is promoted by the BRAFV600E mutation.45 The activating mutation is located in exon 15 of the B isoform of the Raf kinase gene, which results in a constitutive activated state of kinase activity and promotes tumorigenesis through the mitogen-activated protein kinase (MAPK) pathway.46 Tumor initiation by the oncogenic BRAFV600E mutation renders thyroid cells susceptible to transforming growth factor beta-induced epithelial-mesenchymal transition, through a MAPK-dependent process.47 The BRAFV600E mutation may also increase the production of cancer-promoting molecules, such as extracellular matrix proteins, thrombospondin-1,48 and platelet-derived growth factor.49 This increase in the number of these molecules may be involved in the BRAFV600E mutation-mediated progression of PTC and in the aggressiveness of PTC. Furthermore, it has been demonstrated that selective BRAFV600E inhibitor (PLX4720) inhibits migration and invasion of BRAFV600E thyroid cancer cells and tumor aggressiveness in preclinical models of human thyroid cancer.50

Recurrent and persistent disease demands additional therapy and can certainly affect the PTC patient's quality of life; current staging systems are incapable of capturing this risk.3 As outlined in this study, the information of the BRAFV600E mutation can predict poorer outcomes and may lead to better management for individual PTC patients. Because the BRAFV600E mutational status can be measured in preoperative fine needle aspiration biopsy specimens, the ability to identify which patients have aggressive PTC could help surgeons to tailor the operative approach at initial surgery, such as total thyroidectomy rather than lobectomy, and also could potentially provide a long-lasting cure. The implications of the present study correspond with a recent clinical study reporting that preoperative knowledge of the BRAFV600E mutation may productively alter initial PTC surgical management in 24% of cases and may prevent the increased rate of morbidity that has been associated with reoperative neck exploration.51

We observed heterogeneity in the overall effect calculation on each prognostic factor and conducted subgroup analyses to explore possible causes of this heterogeneity. There was no statistically significant difference between subgroups when the studies were divided according to the mean patient age or the number of included cases. The country in which the research was conducted and the frequency of the BRAFV600E mutation contributed to this heterogeneity, suggesting that ethnic differences or the geographical area where the research was conducted may have affected this association.

When we divided the studies into 2 groups according to the prevalence of the BRAFV600E mutation (≥50% and <50%), the prognostic factors were statistically correlated with BRAFV600E-positive PTC compared with wild-type BRAF PTC in both strata. It is noteworthy that the pooled effect sizes of extrathyroidal extension and lymph node metastasis were approximately 30% smaller in the BRAFV600E-prevalent group than in the less prevalent group, although it was unclear whether the different nodal status and extrathyroidal invasion depended on the BRAFV600E mutation rate.

It is important to note that the present study has several limitations. Because we were unable to perform stratified analyses of the summary data from the reported studies, we could not identify the diverse sources of the heterogeneity of the effect size. Multiple outcome variables require cautious interpretation, because the outcomes might be inter-related. For example, as the patients with extrathyroidal invasion tend to have more advanced TNM stages than those with no invasion, the extrathyroidal invasion act as a confounder between the BRAFV600E mutation and higher stages. However, the existence of confounding factors as well as the sources of heterogeneity could be efficiently assessed using an individual level meta-analysis with an existing dataset.52 The meta-analyses of the reported data may also be affected by a publication bias; however, this limitation is unlikely to have had an important impact on the present study. There was no evidence of funnel plot asymmetry upon visual inspection of all the studied outcomes. The inclusion of the unpublished data did not affect the conclusions of the meta-analysis when we reassessed by omitting the results.

In this meta-analysis, the association of the BRAFV600E mutation with cancer-specific mortality was not evaluated. To date, Elisei et al reported increased mortality in those with the BRAFV600E mutation of 102 PTC patients with adequate long-term follow-up (median, 15 years).20 Sufficient amounts of long-term data have not yet been accumulated to accurately evaluate the effect on cancer-specific mortality, because PTC is a slowly progressing cancer with a distinctively low mortality rate.

In conclusion, the meta-analysis confirmed that the BRAFV600E mutation in PTC was correlated with high-risk clinicopathological factors and poor clinical outcome. The results obtained here suggest that the BRAFV600E mutation should be considered a poor prognostic marker in PTC, and that BRAFV600E mutational analysis may lead to better management for individual PTC patients.

FUNDING SOURCES

No specific funding was disclosed.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary