Diagnostic value of ultrasound elastography in thyroid cancer – Hoang Dinh Anh

Journal of military pharmaco-medicine n o 1-2019 158 DIAGNOSTIC VALUE OF ULTRASOUND ELASTOGRAPHY IN THYROID CANCER Hoang Dinh Anh1; Pham Thi Dieu Huong1; Nguyen Thi Hoang Oanh1 Nguyen Minh Hai1; Nguyen Thi Ha1; Nguyen Thi Ly1; Nguyen Vu Thang1 SUMMARY Objectives: To describe some clinical features of thyroid nodules, B-mode ultrasound, Doppler and strain elastography in order to differentiate benign from malignant tumors. Subjects and methods: 72 patients with thyroid nodules were performed clinical examination, 2D ultrasound, Doppler and strain elastography for qualitative assessment of the stiffness of thyroid tumors according to Asteria scores (2008) from ES1 - ES4, semi-quantitative calculating elasticity of the tumors (E2) and normal parenchyma (E1), strain ratio: SR = E2/E1, comparing with fine needle aspiration results. Results: The study group had mean age of 47.31 ± 12.91, most of them were female. B-mode ultrasound: suspicious features of malignance: ill-defined margin (50%); taller- than-wide (11.1%); micro-calcification (13.9%). Ultrasound elastography: 45.8% of patients were classified as ES3 (hard) and ES4 (very hard); strain ratio SR: 2.51 ± 1.18. A comparison of these results with post-operative biopsy showed that the highest sensitivity and specificity were obtained with SR ≥ 2.5. Conclusion: Qualitative assessment of the elasticity of lesions using color map in combination with calculating mean strain ratio of the whole lesion (SR ≥ 2.5) would help to evaluate more exactly the stiffness of lesions on elastography to assess thyroid cancer. * Keywords: Thyroid tumor; Ultrasound elastography; Diagnosis. INTRODUCTION Thyroid nodule is the second most common endocrine disease, after diabetes. The prevalence of the disease diagnosed by physical examination only accounts for 4 - 7%. However, with ultrasound technology, the rate of thyroid nodule detection may reach to 19 - 67%. More than 90% of thyroid tumors are benign adenomas and cysts, only 5 - 10% are malignant nodules. Therefore, early detection and treatment for patients with malignant nodules as well as recognizing patients with benign disease without the need of fine needle aspiration (FNA) are required. So far, ultrasound is still the main method of diagnistic image of thyroid disease. However, with a diagnosed thyroid tumor, the indication of FNA is quite excessive. This increases the number of unecessary FNA performance, which is financial burden and especially makes a stress to the patients. For routine follow-up without FNA, some researchers have recently introduced a new technique - elastography. 1. 103 Military Hospital Correspoding author: Pham Thi Dieu Huong (huongdr106@gmail.com) Date received: 20/10/2018 Date accepted: 03/12/2018 Journal of military pharmaco-medicine n o 1-2019 159 This technique allows to evaluate the biological properties of body tissues based on the principle that when a force is applied to the body tissue, it causes the tissue to deform or move. Normal (soft) tissue is more distorted but less translocated, while pathological tissue (harder), is less distorted and more translocated. The hardness of the tissue is calculated from the corresponding deformation and displacement. Studies in evaluating the effect of ultrasound elastography on focal lesions in general and thyroid nodules in particular in our country are not much. To contribute to the studies and assertion of the value of ultrasound elastography, proceeding to apply this modality daily in the hospital, we conducted this study with the aims to: Evaluate some clinical features, 2D ultrasound and elastography of thyroid nodules, value of ultrasound elastography in diagnosis of malignant nodules. SUBJECTS AND METHODS 1. Subjects. 72 patients were diagnosed with thyroid nodule from 10 - 2015 to 10 - 2016 at Department of Functional Exploration, 103 Military Hospital. 2. Methods. * Study design: Cross-sectional study. - 2D and color Doppler ultrasound were performed to evaluate the size and structure of thyroid parenchyma as well as the position, morphology and vascularity of thyroid nodules. Kwak's TIRADS classification relies on signs of malignancy as following: - Solid nodule. - Hypoechogenicity or marked hypoechogenicity. - Microlobulated or irregular margins. - Microcalcifications. - Taller-than-wide shape. - Elastography: Score ES3 - ES4, Strain ratio ≥ 2.5. Tables 1: TI-RADS Description Risk of malignancy 1 Normal Normal thyroid gland 0% 2 Benign Cyst, spongiform nodule Macrocalcifications Hyperechogenicity 0% 3 Probably benign Hyper or iso-echoic, regular shape and borders, peripheral vascularization < 5% Journal of military pharmaco-medicine n o 1-2019 160 4A Low suspicion for malignancy One suspicious feature 5 - 10% 4B Moderate suspicion for malignancy Two suspicious features 10 - 80% 4 4C Moderate concern but not classic for malignancy Three to four suspicious features > 80% 5 Biopsy proven malignancy All five suspicious features or metastatic lymph nodes * Elastography criteria: Patients with multi-nodular in one or two thyroid lobes, ultrasonography was performed with one that has malignant signs on the 2D ultrasound. - Qualitative hardness of nodules were evaluated based on prominent color of the nodules according to the classification proposed by Asteria et al (2008). Color scale of the hardness: ES1 - ES4. - Semi-quantitative assessment of the hardness: Drawing the circumference of nodules, calculating the average elasticity index (E2). The region of interest (ROI) at the same depth was calculated as the elasticity index of the benign parenchyma (E1). Then calculate the SR (Strain ratio) according to the formula: SR = E2/E1. Picture 1: Color scale in respond to different hardness of thyroid nodules (from ES1 to ES4) and calculation of SR on elastography in one patient. * Data analysis: Data were analyzed by SPSS 16.0, percentage, average value, sensitivity level, specificity, the accuracy of the method compared with the pathological results. Journal of military pharmaco-medicine n o 1-2019 161 RESULTS AND DISCUSSION 1. General characteristics of the study group. In our study, 72 patients were diagnosed with thyroid nodules by clinical examination and 2D ultrasound. They were selected randomly from patients who admited to 103 Military Hospital for ultrasound examination or for thyroid surgery. All patients were performed FNA or histologically diagnosed after surgery. We examined 2D ultrasound and strain elastography of thyroid tumor before having histological results. The results of the study group: Table 2: Some clinical characteristics of the study group. n = 72 % Age (mean) 47.31 ± 12.91 Male 4 5.6 Gender Female 68 94.4 One 41 56.9 Quantity Many 31 43.1 Right lobe 28 38.9 Left lobe 16 22.2 Physical examination of thyroid tumor Location Two lobe 28 38.9 Soft 42 58.3 Tumor density Hard/firmness, less movement 30 41.7 Yes 6 8.3 Lymph nodes in the neck No 66 91.7 The study group was mainly middle-aged with an average age of 47.31 ± 12.91 years old; most of them were female (94.4%). Clinical examination: It is difficult to detect cancer markers, only suspicious features of malignance are the feature of firmness, less movement. We found this symptom in 30/72 patients (41.7%). In terms of age, sex and clinical characteristics suspect thyroid cancer, our results were similar to those of other authors’s studies such as Tran Ngoc Dung [1] in 1,181 patients with thyroid carcinoma, showed an average age of 41.3 ± 8.9 years, the most common age was 30 - 49 years (45.97%), Nguyen Huu Thinh (2012) [2] in 44 patients with thyroid tumor, mean age: 45 years old, female accounted for 81.8%, clinical diagnosis suspected malignant in 40.9%. Journal of military pharmaco-medicine n o 1-2019 162 2. Features of 2D ultrasound, Doppler and TIRADS classification. Table 3: 2D ultrasound features of thyroid tumor. Ultrasound characteristics n = 72 % Solid 50 69.4 Cystic 3 4.2 Composition Mixed 19 26.4 Hyperechoic 0 0 Hypoechoic 21 29.2 Isoechoic 26 36.1 Echogen Mixechoic 25 34.7 Defined 36 50.0 Margin Ill-defined 36 50.0 < 10 mm 19 26.4 Size > 10 mm 53 73.6 Tall > wide 8 11.1 Shape Wide > tall 64 88.9 Micro-calcification 10 13.9 Coarse-calcification 11 15.3 Calcification No calcification 51 70.8 The diagnosis of thyroid tumors mainly relied on 2D ultrasound and Doppler. Nguyen Huu Thinh [2] studied 44 patients with thyroid nodules showed suspicious features of malignance such as hypoechoic structure (47.7%); calcification (31.8%); ill-defined margin (25%). Our results were similar with suspicious features of malignance such as hypoechoic structure (29.2%); ill-defined margin (50%); taller than wide (11.1%); micro-calcification (13.9%). If only every suspected malignant sign on 2D ultrasound, the diagnosis is more difficult. Then, we categorized TIRADS of thyroid tumor: Figure 1: TIRADS classification. Journal of military pharmaco-medicine n o 1-2019 163 Kwak (2011) [6] based on 5 malignant signs on 2D ultrasound for grading of TIRADS. Our results: 60 patients (83.3%) of TIRADS 2 and TIRADS 3. There was no image suspected malignant. 12 patients (16.6%) had 1 to 4 signs of malignancy with TIRADS 4. This group of patients should be further investigated by FNA for diagnosis. 3. Result of strain elastography in thyroid tumor. Figure 2: Ultrasound elastography window with satisfactory force compression, full blue column, ES4 color pattern (patient B.T.H.Y 26 years old). After performing 2D ultrasound, classified nodules according to TIRADS, all patients were performed strain elastography. Patients with multiple nodules in one or two thyroid lobes, strain elastography at the highest TIRADS category nodule was selected. Strain elastography was performed by continuous compression, with moderate force, reaching the standardized display of full blue bars on the screen. Thereby, hardness assessment of the tumor by the color on the elasticity map. Table 4: Elasticity map of thyroid nodule. Strain elastography (ES) Patients (n = 72) Percent (%) ES 1 (soft) 4 5.6 ES 2 (soft - hard) 35 48.6 ES 3 (hard) 28 38.9 ES 4 (very hard) 5 6.9 The results showed that 33/72 patients (45.8%) had solid and hardened tumors (ES3 - ES4), which were considered to be malignant. However, this is a qualitative sign, the diagnostic value is not high. We continued to evaluate the elastic parameters, which are more objective than the color spectrum. Journal of military pharmaco-medicine n o 1-2019 164 Table 5: Strain elastography parameters. Mean Min Max E1 1.19 ± 0.54 0.6 3.3 E2 2.6 ± 0.98 1.0 4.6 SR (E2/E1) 2.51 ± 1.18 0.6 5.7 ≥ 2.5 23 patients (31.9%) SR value < 2.5 49 patients (68.1%) For qualitative assessment of thyroid tumor by strain elastography: We performed continuous pressure for 6 seconds, measuring elasticity index by drawing the perimeter of the entire tumor (E2), compared with that of benign parenchyma at the same depth (E1). Value of E2 (2.6 ± 0.98) was significantly higher than E1 (1.19 ± 0.54). Strain ratio = 2.51 ± 1.18. Compared with the results of the study by Nguyen Huu Thinh [2] (SR: 2.76 ± 0.41), our results were lower, because our study group was mainly benign and SR is valuable for hardness assessment of malignant nodules. 4. The sensitivity, specificity of strain elastography ultrasound compared with FNA and histology results. Table 6: The sensitivity, specificity of strain elastography. Histopathology Elasticity pattern Malignant Benign Total ES3 - ES4 10 13 23 ES1 - ES2 2 47 49 Total 12 60 72 We found that tumors with ES3 and ES4 types were high risk malignant tumors, compared with results of histology after surgery. We suggested that SR value > 2.5 is significant in the diagnosis of cancer with sensitivity 83.3%; specificity 78.3%; accuracy 79.1%. Our results were similar to the results of Sahin’s study [5]: SR > 2.45 is a sign to diagnose thyroid cancer on ultrasound elastography with the sensitivity of 73.9% and the specificity of 73%. Strain elastography combined with 2D ultrasound, TIRADS classification increased sensitivity, specificity and accuracy to 91.7%, 98.3%; and 98.3%, respectively. According to Gilles Russ (2013) [7]: 2D ultrasound, TIRADS classification combined with ultrasound elastography will increase the diagnostic sensitivity of thyroid cancer to 98.5%. In some recent studies, the authors used strain elastography to assess the malignancy of thyroid tumors. However, it is rare for the authors to calculate the strain ratio through the mean elasticity index of the whole tumor. In this study, if an ultrasound of TIRADS 4B or above, combined with ultrasound elastography with ES3 or ES4 color pattern, and strain ratio SR ≥ 2.5, the possibility of malignancy is very high. CONCLUSION Strain elastography is a simple, easy to perform, effective, non-invasive method. Qualitative assessment of the elasticity of lesions using color map in combination with calculating mean strain ratio of the whole lesion (SR ≥ 2.5) would help to evaluate more exactly the stiffness of lesions on elastography to assess thyroid cancer. Journal of military pharmaco-medicine n o 1-2019 165 REFERENCE 1. Trần Ngọc Dũng. Nghiên cứu mô bệnh học lâm sàng 1,118 trường hợp ung thư biểu mô tuyến giáp. Chuyên đề Bệnh viện Quân y 103. Học viện Quân y. 2010, tr.172-175. 2. Nguyễn Hữu Thịnh. Nghiên cứu đặc điểm siêu âm đàn hồi mô ở bệnh nhân u tuyến giáp. Luận văn Thạc sỹ Y học. Đại học Y Huế. 2012. 3. Asteria C, Giovanardi A, Pizzocaro A et al. US-elastography in the differential diagnosis of benign and malignant thyroid nodules. Thyroid. 2008, 18, pp.523-531. 4. Gilles Russ et al. Prospective evaluation of thyroid imaging reporting and data system on 4,550 nodules with and without elastography. Eur J Endocrinol. 2013, 168, pp.649-655. 5. Kwak J.Y, Han K.H, Yoon J.H et al. Thyroid imaging reporting and data system for US features of nodules: a step in establishing better stratification of cancer risk. Radiology. 2011, 260, pp.892-899. 6. Rago T, Santini F, Scutari M, Pinchera A, Vitti P. Elastography: New developments in ultrasound for predicting malignancy in thyroid nodules. J Clin Endocrinol Metab. 2007, 92, pp.2917-2922. 7. Sahin M, Cakal E, Ozbek M et al. Elastography in the differential diagnosis of thyroid nodules in Hashimoto thyroiditis. Med Oncol. 2014, 31, p.97.