Published in:
01-09-2019 | Thyroid Cancer
A Systematic Review and Meta-Analysis of the Diagnostic Performance of BRAF V600E Immunohistochemistry in Thyroid Histopathology
Authors:
Ranjit Singarayer, Ozgur Mete, Laure Perrier, Lehana Thabane, Sylvia L. Asa, Stan Van Uum, Shereen Ezzat, David P. Goldstein, Anna M. Sawka
Published in:
Endocrine Pathology
|
Issue 3/2019
Login to get access
Abstract
Immunohistochemistry (IHC) in evaluating thyroid surgical specimens may facilitate diagnostic and prognostic evaluation, with potential therapeutic implications. We performed a systematic review and meta-analysis examining the analytic validity of IHC in detecting BRAFV600E mutations in thyroid cancer (primary or metastatic). We screened citations from three electronic databases (until December 20, 2018), supplemented by a hand search of authors’ files and cross-references of reviews. Citations and full-text papers were independently reviewed in duplicate, and consensus was achieved on inclusion of papers. Two reviewers independently critically appraised and abstracted data from included papers. Random-effect meta-analyses were conducted for sensitivity and specificity estimates. We reviewed 1499 unique citations and 93 full-text articles. We included 1 systematic review and 30 original articles. The published review (from 2015) needed to be updated as there were multiple subsequent original studies. The pooled sensitivity of IHC in detecting a BRAFV600E mutation was 96.8% (95% confidence interval [CI] at 94.1%, 98.3%) (29 studies, including 2659 BRAFV600E mutant tumors). The IHC pooled specificity was 86.3% (95% CI 80.7%, 90.4%) (28 studies, including 1107 BRAFV600E wild-type specimens). These meta-analyses were subject to statistically significant heterogeneity, partly explained by antibody type (sensitivity and specificity) and tissue/tumor type (specificity). In conclusion, BRAF IHC is highly sensitive and reasonably specific in detecting the BRAFV600E mutation; however, there is some variability in analytic performance.