Thyroid imaging reporting and data system score combined with Bethesda system for malignancy risk stratification in thyroid nodules with indeterminate results on cytology.

Background: The thyroid imaging reporting and data system (TI-RADS) was designed to better select patients who had undergone fine-needle aspiration biopsies (FNABs) with high sensitivity and accuracy. However, the combination of TI-RADS scores and Bethesda system categories in indeterminate thyroid nodules has not been examined extensively.

Objective: This study aimed to stratify indeterminate thyroid nodules (Bethesda categories III, IV and V) according to risk of malignancy as determined by combining TI-RADS score with Bethesda system classification.

Methods: Retrospective study. Histopathological, cytological and ultrasound (US) data were available for 242 cases after surgery, including 136 indeterminate nodules. Methods: All thyroid cytopathological slides and US reports were reviewed and classified according to Bethesda system and TI-RADS categories. The malignancy rate was determined for each Bethesda category, TI-RADS score and both methods combined of indeterminate nodules.

Results: The malignancy rates were 8·7%, 51·3% and 67·5% for Bethesda categories III, IV and V, respectively. Based on histopathological comparison, the accuracy was 66·7% for TI-RADS greyscale. TI-RADS 3 and 4A scores were observed in 80% of Bethesda III cases, which led to 80% sensitivity and 90% of negative predictive value (NPV). In contrast, for nodules scored as TI-RADS 4B and 5, the combined cytological results of Bethesda IV and V resulted in a higher risk of malignancy (75% and 76·9%, respectively, P < 0·001).

Conclusions: In view of the high NPV of TI-RADS 3/4A only in Bethesda III category, a surgical approach could be considered for lesions defined as Bethesda III, IV and V when TI-RADS 4B and 5 were concomitant.