Correlation between thyroid ultrasound and histology in patients with indeterminate cytology results : a local experience

Abstract

BACKGROUND : Thyroid nodules are common. Most patients with indeterminate fine needle aspiration cytology (FNAC) results are subjected to thyroidectomy for fear of malignancy. However, only 20–30% of these cases are found to be malignant. The aim was to determine the value of thyroid ultrasound in diagnosing malignancy in patients with indeterminate fine needle aspiration cytology results in our practice. Fine needle aspiration was performed after ultrasound, thereby preventing architectural distortion of the nodule and to ensure that the most suspicious nodule was biopsied. METHODS : A retrospective review of records of patients who presented to two University of Pretoria academic hospitals, in South Africa, from 2001 to 2015 with nodular thyroid goitre was undertaken. Patients had a thyroid ultrasound scan, FNAC and had undergone thyroid surgery. RESULTS : Records of 104 patients were retrospectively evaluated. Patients were predominantly female (93.3%). Histology report was available for 101 of the 104 patient records. Malignancy was identified in 23 (22.8%). The overall sensitivity and specificity of thyroid image reporting and data system (TIRADS) score in this study were 69.5% and 61.5% respectively. The TIRADS classification had high sensitivity amongst the 26 patients with indeterminate cytology, with sensitivity and specificity of 85.7% and 52.6% respectively, however it was not statistically significant (p = 0.28). CONCLUSION : There was poor identification of malignancy with the use of ultrasound TIRADS classification in cases of indeterminate FNAC results, Bethesda III or IV, in our study. This is probably related to the diversity of ultrasonographers in our practice. It is recommended that there should be a dedicated thyroid ultrasonographer for a better and consistent TIRADS classification that surgeons can rely on for guiding surgical intervention.