Evaluation of the efficacy of us-elastography and TI-RADS scoring system for the differentiation of benign-malignant thyroid nodules

PURPOSE: In this study, we aimed to evaluate and compare the sensitivity and specificity of the TI-RADS scoring system and US-Elastography for differentiation of benign-malignant thyroid nodules and evaluate their efficacy on their own, and in combination. MATERIALS and METHOD: 50 thyroid nodules in 49 patients between January 2015-April 2015 were included in the study. All cases were evaluated by gray scale, power Doppler and elastography findings with an Esaote MyLab 70XVG system. Marked hypoechogenicity, microcalcifications, irregular borders, solid nature and taller-than-wide appearance were considered suspicious findings for purpose of TI-RADS scoring. In Power Doppler examination, vascularity of the nodules were categorized into peripheral, intranodular and anarchic vascularity. Using strain elastography method, color maps were acquired in transverse and longitudinal planes, which were then evaluated qualitatively according to the Rago scale. Semi-quantitative evaluation was performed via elasticity indices, which were calculated as the ratio of the strain of the nodule to the strain of the normal thyroid parenchyma at the same level and to the softest parenchyma at the same level. ROC analysis was performed. Statistical significance was evaluated with chi-square and Fisher's exact test for categorical variables; and Student's t test and Mann-Whitney U test for continuous variables. The level of statistical significance was set at p<0.05. FINDINGS: 45 nodules were benign and 5 were malignant. For the purpose of detecting malignancy, irregular borders was 80,0% sensitive and 93,3% specific; the presence of a halo 80,0% sensitive and 40,0% specific; being solid 100% sensitive and 53,3% specific; microcalcifications 40,0% sensitive and 93,3% specific; being hypoechoic 80,0% sensitive and 53,3% specific; being markedly hypoechoic 60,0% sensitive and 97,8% specific. There was no significant difference between benign and malignant groups with respect to vascularity (chi-square test, p=0,37). Being TI-RADS 4b and 4c was 80% sensitive and 86,7% specific in detecting malignancy, whereas being Rago category 4 or 5 was 40,0% sensitive and 100% specific. Elasticity indices of thyroid/nodule and softest parenchyma/nodule were measured three times in both transverse and longitudinal planes. When the average of these three measurements were considered the index, cut-off values as calculated in ROC analysis were 1,59, 1,18, 2,38 and 2,50 respectively; whereas when only the maximum index was considered they were 1,79, 1,59, 2,53 and 2,73 respectively. The areas under the curve varied between 0,729-0,893. When nodules that were categorized as malignant either by TI-RADS or Rago scores were considered malignant, a sensitivity of 80% and specificity of 86,7% was calculated. There was moderate correlation between TI-RADS and Rago scoring systems (r=0,505). CONCLUSION: TI-RADS scoring system, which evaluates gray scale US findings collectively, is a practical and efficient method for evaluation of thyroid nodules. US-Elastography, alone, is 40% sensitive and 100% specific for detection of malignancy; however, these values increase to 80% and 87% when combined with gray scale findings. Semi-quantitative measurements, compared to qualitative analysis, decrease specificity (from 100% to 71,1%) but increase the sensitivity significantly (from 40% to 100%). More studies with larger patient series are required to evaluate the efficiacy of TI-RADS and US-Elastography when used in combination. Key words: TI-RADS, elastography, thyroid, nodule, FNAB.