Accuracy of Thyroid Imaging and Reporting Data Systems in Risk Stratification of Thyroid Nodules- A Retrospective Observational Study

Abstract

Introduction: Thyroid cancer is the most common endocrine malignancy worldwide. Thyroid ultrasound reporting based on Thyroid Imaging and Reporting Data Systems (TIRADS) classification has enormously changed the clinical management of patients presenting with thyroid nodules. Aim: To study the thyroid nodules using high-resolution ultrasound and to correlate the TIRADS and Bethesda system for thyroid cytopathology. Methods and Materials: The present study was a retrospective observational study which was conducted in the Department of Radiology, Father Muller Medical College, Mangalore, Karnataka, India. Total 226 patients who came for thyroid ultrasound for evaluation of thyroid swelling were included. The thyroid nodules were evaluated based on features like shape, margin, echogenicity, internal composition, presence and type of calcification. Thyroid lesions were categorised using TIRADS classification. Diagnostic accuracy of TIRADS system was evaluated by comparing with the Fine Needle Aspiration Cytology (FNAC). Cytopathology report was considered to be the gold standard. The p-value and odds ratio were determined so that it was known how significantly the presence or absence of a specific ultrasound feature was seen associated with benign lesions and malignant lesions respectively in the study population. For each TIRADS category, the risk of malignancy was determined. Cervical lymph nodes were evaluated based on the size of lymph node, margin, necrosis, presence or absence of central echoegenic hilum and microcalcification. The percentage of cases that were accurately determined by TIRADS, (to avoid unnecessary FNAC) were determined. Data was statistically analysed using Statistical Package for the Social Sciences (SPSS Inc, Chicago, Illinois, USA), version 20.c2. Results: The risk of malignancy in TIRADS categories 1 and 2 was found to be 0%, 0.5% in category 3, 3.85% in category 4A, 65.32% in category 4B, 84.22% in category 4C, and 100% in category 5. Out of the five sonological features assessed, shape (taller than wider) showed the highest Positive Predictive Value (PPV) (96.15%) for malignancy. Also, of the cervical lymph nodes studied, sonological features such as loss of central echogenic hilum, presence of an irregular margin, microcalcification and necrosis were found to have sensitivity of 100%, 67.15%, 43.56%, and 76.56%, respectively and specificity of 97.1%, 96.5%, 11.8%, and 35.3%, respectively to differentiate between benign and metastatic nodes. The estimated decrease in unnecessary FNACs was found to be 41.15%-84.30%. Conclusion: The widespread use of imaging ultrasound techniques has generated an overwhelming increase in the recognition of thyroid nodules. This study will help to improve the ultrasound characterisation of thyroid nodules and establish risk groups to decide which patient should be evaluated for FNAC.

Keywords

• hypoechogenecity
• malignant nodule
• microcalcification
• thyroid cancer