The Clinical Value Of Quantitative Parameters Of Ultrasonic Elastography In Diagnosis Of Thyroid Nodules

BackgroundThyroid nodules are thyroid masses with the abnormal hardness and structure, as the common thyroid disease. It’s reported that more women have suffered thyroid nodules than men, and the incident rate in elders is higher. Most thyroid nodules are benign, and about2.7percent to17percent ones are malignant. The main therapeutic method for malignant nodules is surgery. The chief purpose of clinical diagnosis is subjected to determine the nature of nodules, for making the suitable therapeutic method to induce the unnecessary surgery and the operative complication, and then to improve the patients’life quality. However, the clinical characteristic, imaging and cytology feature of benign and malignant lesions sometimes overlap in them. Therefore, it’s difficult for differentiation before surgery.Nowadays, the differentiating methods between benign and malignant thyroid nodules concentrate to imaging and fine needle aspiration. As a non-invasive method, imaging examinations includes CT, MRI, radioisotope scanning, PET, and ultrasound. Those strongly radioactive methods are not fit for pregnant women, containing CT, radioisotope scanning and PET; MRI and PET with a high price could not be applied widely. Among all the imaging examinations, ultrasound with the most sensitive about100%to thyroid nodules, could detect lesions about diameter1-2mm.The diagnosis of thyroid nodules ever came from conventional ultrasound examinations, including traditional gray scale ultrasound for appearance and color Doppler ultrasound for vascularity. Gray scale ultrasound could detect the number, size, shape, location, cover, inner echo (cystic or solid or calcification), and relation with the closing tissue and so on. Color Doppler ultrasound could provide the inner or rounding vascularity of nodules, including the distribution, volume, velocity. With the increasing resolution of ultrasound, the detection of tiny lesion improves greatly. But, the multifocal pathological structure of thyroid nodules brings the complication of image, so that the overlap existence of the benign and malignant nodules influences the diagnosis.Conventional ultrasound could provide appearance about the benign and malignant nodules except hardness. Thus, it is confined in the differentiation of thyroid nodules to some extent. Ultrasound elastography could reflect the nodule hardness to some extent. It is a fine noninvasive and convenient supplement to traditional ultrasound. Ultrasound elastography has showed its superiority in several fields. It has a high clinical value in the breast tumors, and then has been applied for thyroid nodules.According to the clinical evaluation of thyroid nodules, the hardness of nodules has a close relation with the pathology. And the malignant nodules are much harder than benign ones. Ultrasound elastography could reflect directly the nodule’s elasticity. Ultrasound elastography contains three types (strain imaging, transient elastography, vibration elastography). Thyroid is superficial organ, so strain imaging is more suitable for examining than others. The elasticity of tissue could been got via the probe compressing it. In the current times, most of researchers put emphasize on elastography score for thyroid nodules at home and abroad, however, more studies are needed to carry out for establishing the value of elastography quantitative indexes in thyroid nodules.It is reported that strain ratio is beneficial for differentiating thyroid nodules, but there is not a unified the best cut point. At the same times, studies for Region of interest A (ROI A) value were barely showed.Our study for differentiating thyroid nodules were divided into two parts, and took post-operative pathology as the golden standard.1) Receiver operator characteristic curve(ROC curve) was used to realized strain ratio in order to evaluate the thyroid nodules. And the best cutoff point is determined.2) Receiver operator characteristic curve(ROC curve) was used to realized ROI A value in order to evaluate the thyroid nodules. And the best cutoff point is determined. Strain ratio, elatography score, ROI A value were divided into four groups, including strain ratio, ROI A value, elastography score combined with strain ratio, and elastography score combined with ROI A value. Comparing the four groups was aimed at a better method of differentiating thyroid nodules.Objectives1) Explore the clinical value of ultrasound elastography quantitative parameters in differentiating thyroid nodules;2) Evaluate the application of four groups (strain ratio, ROI A value, elastography score combined with strain ratio, and elastography score combined with ROI A value) for thyroid nodules.Materials and Methods1) Research objects2)Two hundred twenty-four patients (61men and163women) with233thyroid nodules were included in the study from February2010to July2011in our hospital. Patient age ranged from4to77years. All nodules were confirmed by operation.3) The cases were selected according to the following criterion:lesions without autoimmune thyroid disease; at least twice as much normal thyroid tissue surrounding the target lesion.1) Examination and methods The Hitachi HI Vision900and Hitachi Vision Preirus that was used to evaluate the patients provides real-time elastography, a program for measuring strain ratio, and a linear6.0to13.0MHz probe. First, patients underwent a B-mode ultrasound examination and color Doppler flow image. They then underwent real-time ultrasonic elastography, and the observer produced slight and steady compression of the target lesion, finally stored the fine elastogram, which reflected each layer tissue clearly.Strain ratio and ROI A value were calculated. Region of interest A (ROI A) encompassed the maximum target tumor area, and the second ROI (ROI B) colored with green included the lateral normal thyroid tissue with ROI A. All strain ratios and ROI A value were compared with post-operative pathology, then sensitivity and specitivity were calculated. The best cutoff value was acquired via ROC curve.The four-score system distinguishes five types of nodules:0=cystic lesions (blue-green-red; BGR);1=soft solid lesions with stiffness similar with the surrounding normal thyroid tissue (homogeneous green);2=lesions with stiffness slightly harder than the surrounding tissue (ratio of blue and green less than50%);3=harder lesions with stiffness greater than the surrounding tissue (ratio of blue and green more than50%);4=the hardest lesions (whole lesion, even the surrounding tissue is colored blue). Strain ratio, elatography score, ROI A value were divided into four groups, including strain ratio, ROI A value, elastography score combined with strain ratio, and elastography score combined with ROI A value. It was comparing the sensitivity, specificity, accuracy, positive predictive value, negative predictive value,and odd product of four groups that was aimed at a better method of differentiating thyroid nodules.2) Statistical analysis The statistical software SPSS13.0was used to analyze the data. The difference between benign and malignant nodules was analyzed using a t-test and chi-square test was used to evaluate the strain ratio and ROI A value. The best cutoff value for strain ratio and ROI A value was calculated by receiver-operating characteristic (ROC) analysis. According to the best cutoff value, thyroid nodules include benign and malignant group, and t-test and chi-square test was used to evaluate the two combinations; k-independent samples nonparametric test was used to evaluate the difference of four groups. Sensitivity, specificity, accuracy, positive predictive value, negative predictive value, and odd product of four groups were assessed compared to the histologic results.Results1) Pathology of all nodules:There were105malignant nodules:100papillary carcinoma,1medullar carcinoma,1follicular carcinoma. There were128nodules in benign group, including:29thyroid adenoma,96nodular goiter, and others3(calcification, interstitium fibro, focus lympathyroiditis).2) Part one:strain ratioThe mean strain ratio of benign lesions was2.18±1.45; the mean strain ratio of nodular goiter was2.23±1.53; the mean strain ratio of adenoma was1.87±0.91. The mean strain ratio of malignant lesions was4.36±2.51; the mean strain ratio of papillary carcinoma was4.36±2.51. According to t test and chi-square test, strain ratio was significantly different between benign and malignant lesions (p<0.001). Strain ratio was not significantly different between nodular goiter and adenoma (P=0.224). The ROC analysis produced a cutoff value (2.49) for strain ratio. The AUC was0.826. The higher is strain ratio, the possibility of malignant nodule is higher.3) Part two:ROI A value, and comparing of four groupThe mean ROI A value of benign lesions was0.0027±0.0014; the mean ROI A value of nodular goiter was0.0027±0.0014; the mean ROI A value of adenoma was0.0029±0.0013. The mean ROI A value of malignant lesions was0.0012±0.0007; the mean ROI A value of papillary carcinoma was0.0012±0.0007. According to t test and chi-square test, ROI A value was significantly different between benign and malignant lesions (p<0.001). ROI A value was not significantly different between nodular goiter and adenoma (P=0.627). The ROC analysis produced a cutoff value (0.00175) for ROI A value. The AUC was0.826. The lower is ROI A value, the possibility of malignant nodule is higher. According to chi-square test, the two combinations was significantly different between benign and malignant lesions (p<0.001). Comparison of sensitivity, specificity, accuracy, positive predictive value, negative predictive value, and odd product of the four groups (strain ratio, ROI A value, elastography score combinated with strain ratio, and elastography score combinated with ROI A value).(Tab.1) Tab.1Comparison of the four groups a:elastography score combinated with strain ratio b:elastography score combinated with ROIA value PV+:positive predict value PV_.:negative predict value sensitivity specificity accuracy PV+PV_OPConclusions1.Strain ratio demonstrates the fine clinical value of differentiating thyroid nodules.2.ROI A value demonstrates the fine clinical value of differentiating thyroid nodules. The diagnosis capability of four groups was median, and the group (elastography score combined with ROI A value) showed the highest comprehensive diagnostic capability.