Endocrine Abstracts

Background: The majority of thyroid nodules are clinically insignificant, however considerable overlap in radiological characteristics exists for benign and malignant lesions. Recently, validated radiological criteria using thyroid imaging reporting and data system scoring (TIRADS) have shown correlation between cumulative suspicious features and risk of malignancy¹.

Objective: To assess sensitivity, specificity and reliability of ultrasound scan (USS) in evaluation and differentiation of benign from malignant thyroid lesions.

Methods: Patients presenting with a thyroid nodule between 2007 and 2012 to the thyroid service at our institution and in whom surgical histology data was subsequently available were studied. Pre-operative thyroid ultrasound images were anonymised and retrospectively reviewed by two independent, blinded, radiologists. Reports were generated for individual and overall TIRADS score. Only static images were available for analysis. TIRADS scores ranged from 1 to 5. Suspicious nodules scored between 4a–c and 5. TIRADS scores of 1–3 were considered benign. Final histology was used as the definitive outcome variable. TIRADS sensitivity and specificity for malignancy was determined by dichotomising scores between ‘benign’ and ‘malignant’ and shifting the level at which nodules were considered malignant.

Results: 58 patients (45F and 13M), with 71 nodules, 49 benign, and 22 malignant nodules were studied. Eighteen patients had a malignancy, of which three were multifocal.

A TIRADS score of 4a or higher yielded sensitivity for malignancy of 68.2–81.8%; 5.63–9.86% of malignancies were missed by this algorithm (range represents reporting discrepancies between radiologists).

Conclusions: In our centre the TIRADS diagnostic accuracy on static ultrasound appeared inferior to current established diagnostic techniques.

Prospective use of TIRADS using dynamic images alongside current clinical diagnostic techniques requires further study.

Reference: 1. Kwak JY 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(3) 892–899.