Background The diagnosis of Bethesda Ⅲ/Ⅳ/Ⅴ thyroid nodules has always been a challenge for clinicians. While BRAF (V600E) gene detection is commonly used for preoperative assessment, it exhibits high specificity but low sensitivity. The American college of radiology and the Chinese medical association ultrasound branch have respectively proposed the American college of radiology thyroid imaging reporting and data system (ACR-TIRADS) and the Chinese thyroid imaging reporting and data system (C-TIRADS). The combination of TIRADS with BRAF (V600E) gene detection aims to improve the diagnostic accuracy of Bethesda Ⅲ/Ⅳ/Ⅴ thyroids nodules and reduce unnecessary surgical procedures and fine needle aspiration biopsies.

Objective To compare the diagnostic value of American college of radiology thyroid imaging reporting and data system (ACR-TIRADS), Chinese thyroid imaging reporting and data system (C-TIRADS) in Bethesda Ⅲ/Ⅳ/Ⅴ thyroid nodules, and explore the additional value of BRAF (V600E) gene detection.

Methods This study included 484 thyroid nodules from 484 patients who underwent fine needle aspiration biopsies and BRAF (V600E) gene detection in the Chinese PLA General Hospital from January 2020 to July 2023. Taking histopathological diagnosis as the gold standard, the diagnostic value of ACR-TIRADS, C-TIRADS, and their combination with BRAF (V600E) gene detection in Bethesda Ⅲ/Ⅳ/Ⅴ thyroid nodules was compared.

Results In both ultrasound stratification systems, the malignant risk of thyroid nodules gradually increased with higher grades (all P for trend＜0.001). The optimal cut-off values for ACR-TIRADS and C-TIRADS were 5 and 4c, respectively. BRAF (V600E) gene detection significantly improved the diagnostic performance of ACR-TIRADS (0.809 vs 0.778, P ＜0.001) and C-TIRADS (0.815 vs 0.783, P＜0.001) for Bethesda Ⅲ/Ⅳ/Ⅴ thyroid nodules without reducing diagnostic specificity. There was no statistically significant difference in the area under the curve (AUC) between ACR-TIRADS and C-TIRADS, whether assessed individually or in combination (0.783 vs 0.778, P=0.755; 0.815 vs 0.809, P=0.675). However, C-TIRADS exhibited higher sensitivity (88.83% vs 83.90%, P＜0.001; 95.33% vs 90.13%, P＜0.001) and accuracy (84.50% vs 81.41%, P ＜0.001; 89.67% vs 86.36%, P＜0.001) compared to ACR-TIRADS, either alone or in combination. Notably, the specificity of ACR-TIRADS, either alone or in combination was equal, and higher than that of C-TIRADS (71.72% vs 67.68%, P＜0.001).

Conclusion BRAF (V600E) gene detection significantly enhances the diagnostic performance of ACR-TIRADS and C-TIRADS for Bethesda Ⅲ/Ⅳ/Ⅴ thyroid nodules without compromising specificity. Whether used alone or in combination with BRAF (V600E) gene detection, C-TIRADS exhibits the best overall diagnostic performance, demonstrating promising clinical utility.