Molecular testing of the presence of pathogenic genomic variants in a tumor without quantifying the variant allele fraction (VAF) does not differentiate the variation extent among tumors, often resulting in an inconclusive diagnosis because of interpatient variability. To examine the association between the quantification of VAFs of BRAF V600E and TERT promoter variants and a definitive cancer diagnosis of thyroid tumors. This diagnostic study analyzed a cohort of 378 surgically resected thyroid tumors with a maximum dimension of 1 cm or larger between March 15, 2016, and March 16, 2020, and a separate cohort of 217 residual thyroid fine-needle aspiration (FNA) biopsy specimens obtained from January 22, 2020, to March 2, 2021, at Mount Sinai Hospital, Toronto, Ontario, Canada. Data analysis was conducted between February 1, 2021, and February 1, 2023. Quantitative VAF assays of BRAF V600E and TERT promoter variants (C228T and C250T) were performed by digital polymerase chain reaction molecular assays. The VAFs of BRAF V600E and TERT promoter variants were correlated with tumor histologic diagnoses and histopathologic features to delineate the association of VAF assays with tumor malignancy. The receiver operating characteristic curve analysis, sensitivity, specificity, positive predictive value, negative predictive value, and logistic regression analysis based on follow-up histopathologic types were used to determine the diagnostic utility of the quantitative molecular assays. A total of 595 specimens, including 378 surgically resected thyroid tumors and 217 thyroid nodule FNA biopsy specimens, were collected from 580 patients (436 [75.2%] female with a mean [SD] age of 50 [16] years and 144 [24.8%] male with a mean [SD] age of 55 [14] years). Sensitive VAF assays of 378 thyroid tumors revealed the presence of the BRAF V600E variant in 162 tumors (42.9%), with 26 (16.0%) at a low VAF of 1% or less and 136 (84.0%) at a high VAF of greater than 1%, and the presence of TERT promoter variants in 49 tumors (13.0%), including 45 C228T variants (91.8%), 15 (33.3%) of which were quantified as having a low VAF (≤1%) and 30 (66.7%) as having a high VAF (>1%), and 4 C250T variants (8.2%) with VAFs between 40.0% and 47.0%. All tumors detected with BRAF V600E and/or TERT promoter variants, whether at low or high VAFs, received a definitive cancer diagnosis. Further analysis delineated a significant association between high VAFs of either variant individually or different VAF levels for both variants in coexistence and aggressive histopathologic features of tumors. Excluding low VAFs assisted in identifying patients at an intermediate-to-high risk of recurrence (odds ratio, 5.3; 95% CI, 1.9-14.6; P = .001). The VAF assays on the residual FNA biopsy specimens showed a high agreement to those on surgical tissues (κ = 0.793, P < .001) and stratified malignancy in 40 of 183 indeterminate FNA cases (21.9%), with a sensitivity of 93.8% (95% CI, 67.7%-99.7%), specificity of 90.0% (95% CI, 75.4%-96.7%), positive predictive value of 78.9% (95% CI, 53.9%-93.0%), and negative predictive value of 97.3% (95% CI, 84.2%-99.9%). This diagnostic study suggests that sensitive quantitative VAF assays of BRAF V600E and TERT promoter variants can elucidate the interpatient variability in tumors and facilitate a definitive cancer diagnosis of thyroid nodules by differentiating the variation extent of genomic variants, even at low VAFs.
Read full abstract