ベセスダIII甲状腺結節のTBSRTC亜分類とBRAF V600E検査の統合による精密管理：WHO第5版に基づくサブタイプ特異的な診断格差を強調する研究

To investigate the malignant risk of ultrasound-guided fine-needle aspiration (UG-FNA) cytology-diagnosed Bethesda III thyroid nodules and the significance of combined BRAF V600E mutation testing, and to establish a clinically actionable algorithm integrating cytomorphologic and molecular features.

A retrospective analysis was conducted on UG-FNA results from 10,839 patients (12,528 nodules) at Deyang People's Hospital between December 2021 and September 2024. Liquid-based cytology (LBC) combined with cell block technology identified 732 Bethesda III nodules (detection rate: 5.8%). Among these, 180 cases with postoperative histological follow-up were analyzed, including 101 cases with preoperative BRAF V600E testing using cell blocks. Using histology (5th edition WHO classification) as the gold standard, the influence of gender, age, nodule location, diameter, and cytological subcategory (with nuclear atypia vs. with other features) on malignant risk was analyzed using univariate and multivariate logistic regression. The diagnostic value of BRAF V600E testing was evaluated.

Among 180 Bethesda III nodules, malignancies accounted for 62.2% (112/180), predominantly papillary thyroid carcinoma (PTC; 110/112, 98.2%), while low-risk neoplasms and benign lesions constituted 15.0% (27/180) and 22.8% (41/180), respectively. Multivariate analysis identified nodule size <1 cm (P < 0.001) and the "nuclear atypia" subcategory as independent predictors, with the latter showing extreme predictive value for malignancy (OR = 121.854, P < 0.001; capturing 98.2% malignancies) and modest association with low-risk neoplasms (OR = 7.014, P = 0.001). BRAF V600E testing (n=101) demonstrated 82.0% sensitivity (100% specificity) for PTC but exhibited striking subtype-dependent performance (classical PTC: 90.2% [46/51] vs. FVPTC: 40.0% [4/10], P = 0.001). ROC analysis revealed that combined cytology-BRAF testing achieved AUCs of 0.873 (95%CI:0.797-0.948) for malignancy and 0.892 (95%CI:0.822-0.962) for PTC, though accuracy was significantly lower for FVPTC (AUC = 0.481) versus classical PTC (AUC = 0.911, P < 0.001).

In Bethesda III nodules, a diameter <1 cm or the "with nuclear atypia" subcategory indicates a high malignant risk (OR = 121.854), warranting active management. Preoperative BRAF V600E testing provides excellent detection for classical PTC (sensitivity 90.2%, specificity 100%) but exhibits limited sensitivity for FVPTC (40%). Based on the 3rd edition TBSRTC and 5th edition WHO classification, our risk-stratified algorithm could reduce unnecessary surgeries by 25-30% in indeterminate nodules while maintaining perfect specificity for classical PTC, achieving optimal clinical decision-making.