A single-center multidisciplinary study analyzing thyroid nodule risk stratification by comparing the thyroid imaging...

The thyroid imaging reporting and data system (TI-RADS) and 2015 American Thyroid Association (ATA) guidelines are two well-known risk stratification systems for classifying thyroid nodules based on cancer risk. This study aims to evaluate the diagnostic efficacy of these two systems in predicting malignancy in patients undergoing thyroid surgery.

We studied data on 120 individuals who were scheduled to undergo surgery for benign or malignant nodular diseases of the thyroid gland between October 2017 and October 2019. The TI-RADS category and ultrasound pattern based on ATA guidelines were assigned to dominant thyroid nodule categories by two experienced radiologists blinded to patients' previous thyroid ultrasonography and fine-needle aspiration biopsy results. A pathologist with experience in thyroid diseases blinded to patients' sonographic and clinical data reviewed the thyroidectomy specimens.

A total of 120 patients, 88 women and 32 men, were included in our study. Final histopathological results were as follows: 50% (n=60) papillary thyroid carcinoma, 36.6% (n=44) benign nodular thyroid diseases, 4.1% (n=5) follicular adenoma, 2.5% (n=3) hurtle cell adenoma, 1.7% (n=2) follicular thyroid carcinoma, 1.7% (n=2) medullary thyroid carcinoma, 1.7% (n=2) hurtle cell carcinoma, and 1.7% (n=2) follicular tumor of uncertain malignancy potential. The sensitivity, specificity, positive predictive value, and negative predictive value for TI-RADS were 80%, 56%, 72%, and 67%, respectively, and that for ATA were 80%, 64%, 76%, and 69%, respectively.

The TI-RADS and ATA showed similar rates of sensitivity, specificity, NPV, and PPV. Our observed risk of malignancy was higher than expected for the ACR TI-RADS 3-5 categories and the very low, low, and intermediate suspicion risk strata in the ATA guidelines. We found no difference between observed and expected malignancy risk for the ACR TI-RADS 2's and ATA's high suspicion categories.