2D Shear Wave Elastography Increases Diagnostic Accuracy of the TIRADS-ACR and ATA Classification System in Thyroid Nodule Selection: Cyto- and Histological Correlation.

Abstract

Purpose: The aim was to evaluate whether elastography changes the accuracy of thyroid nodule malignancy risk classification by TI-RADS ACR and ATA systems.

Materials and methods: This was a prospective study with 191 nodules (180 patients). Nodule assessments by B-mode ultrasonography (US) and two-dimensional shear wave elastography (2D-SWE) were compared with a) fine-needle aspiration biopsy cytological (Bethesda II) and post-resection histology results (Bethesda III to VI). Nodules were divided into benign and malignant. B-mode US evaluated: echogenicity, composition, dimensions, contours, limits, presence of halo and echogenic foci. Elastography classified nodules from I - completely softened to IV - completely hard; mean nodule deformation value (assessed in m/s and kPa); deformation ratio between nodule and thyroid parenchyma (TDR) and deformation ratio between nodule and prethyroid musculature (MDR).

Results: Significant univariate parameters for B-mode: hypoechogenicity, halo, microcalcifications, irregular contours and ill-defined limits; for elastography, all parameters. MDR (in kPa) was the best elastographic parameter: nodules with MDRs> 1.53 exhibited a higher chance of malignancy (AUC-ROC=0.831). B-mode ACR-TIRADS performance had AUC: 0.678; 95% CI: 0.596-0.760 and ATA, AUC: 0.680; 95%: 0.597-0.763. Multivariable analysis indicated that combination of prognostic models with any elastographic parameter increased performance. ATA classification combined with elastogram pattern and MDR (in kPa), increased the AUC to 0.892; 95% IC: 0.845-0.939.

Conclusion: 2D-SWE can increase accuracy of the most widespread B-mode prognostic models - TI-RADS ACR and ATA.