Development of a Molecular Assay for Detection and Quantification of the BRAF Variation in Residual Tissue From Thyroid Nodule Fine-Needle Aspiration Biopsy Specimens

IMPORTANCE: Thyroid cancer, predominantly papillary thyroid carcinoma (PTC), is common, but an estimated 30% of ultrasonography-guided fine-needle aspiration (FNA) biopsies of thyroid nodules are indeterminate. BRAF variation, associated with poor clinicopathological characteristics, is a useful molecular marker for diagnostics. OBJECTIVE: To develop a sensitive molecular assay for BRAF V600E detection in remaining tissue of thyroid FNA biopsies to identify patients with cancer carrying a BRAF variation. DESIGN, SETTING, AND PARTICIPANTS: This diagnostic study used tumor tissue from surgical formalin-fixed, paraffin-embedded (FFPE) specimens and residual tissue from thyroid FNA biopsies for genomic DNA extraction. FFPE specimens served as the validation set, and residual tissue from FNA biopsies served as the test set. A molecular assay was developed for accurate detection of BRAF V600E variation using locked nucleic acid (LNA) probe–based droplet digital polymerase chain reaction (dPCR), and the assay was validated by BRAF V600E immunohistochemical staining (IHC). The study was conducted between February 2019 and May 2021. RESULTS: A total of 271 specimens, including 77 FFPE specimens (with a follow-up of 48 matched surgical specimens) and 146 residual FNA samples, were collected from 223 patients (mean [SD] age, 53.8 [15.3] years; 174 [78.0%] women; 49 [22.0%] men). The molecular assay by dPCR was first established to specifically and accurately detect and quantify wild-type BRAF and variant BRAF in DNA from human follicular thyroid carcinoma–derived FTC-133 and papillary thyroid carcinoma–derived BCPAP cells. The linearity of quantification of BRAF V600E was calculated (y = 0.7339x; R(2) = 0.9996) with sensitivity at 0.02 copies/μL and reproducibility in detecting variant DNA at various dilutions(coefficient of variance in 0.3% DNA, 9.63%; coefficient of variance in 1.0% DNA, 7.41%). In validation testing, the dPCR assay and IHC staining exhibited 100% specificity in concordantly identifying BRAF V600E in PTCs (κ = 0.873; P < .001) and sensitivity of 32.0% (95% CI, 19.1% to 44.9%) in dPCR and 26.0% (95% CI, 13.1% to 38.9%) in IHC staining, with an improvement by 23.08% in dPCR compared with the IHC staining. The dPCR assay further detected BRAF V600E in 39 of 146 residual FNA specimens (26.7%). At short-term follow-up, 48 patients, including 14 of 39 patients with BRAF variation and 34 of 107 patients without BRAF variation on residual FNA specimens, underwent resection. The dPCR assay of BRAF status in the matched surgical specimens showed BRAF V600E variations in 12 patients and wild-type BRAF in 36 patients, with a high agreement to that in residual tissue of FNA specimens (κ = 0.789; P < .001). Among 14 patients with BRAF variations on residual FNA, 13 were diagnosed with PTC and 1 was diagnosed with anaplastic thyroid cancer at the thyroidectomy. CONCLUSIONS AND RELEVANCE: This diagnostic study developed a sensitive molecular assay for detection and quantification of BRAF V600E variation in residual tissue from thyroid FNA biopsies to identify patients with cancer harboring BRAF V600E in a cost-effective manner, highlighting the clinical value of molecular assay of the remaining FNA tissue in the management of thyroid nodules.