W. Castillo1, S. Jain1, E. Idzikowski1, H. Bava1, M. Kotdawala1, A. Manzella1, S. Z. Trooskin1, A. M. Laird1, 2, T. Beninato1, 2 1Robert Wood Johnson – UMDNJ, New Brunswick, NJ, USA 2Cancer Institute Of New Jersey, New Brunswick, NJ, USA
Introduction: Thyroid nodules are common, with the prevalence estimated to be as high as 49-57%. With increased use of imaging, a large number of asymptomatic, incidental thyroid nodules are detected. To determine malignancy, a definitive diagnosis can only be made by invasive techniques including fine needle aspiration (FNA) or surgery, potentially resulting in unnecessary biopsies of benign nodules. To better assess at-risk lesions and reduce the morbidity of unnecessary procedures, the American College of Radiology developed the Thyroid Imaging Reporting and Data System (ACR TI-RADS) to stratify malignancy risk based on ultrasound findings. However, the applicability of this scoring system to tumors other than the most common thyroid malignancy, papillary thyroid cancer, is not clear. This study aims to elucidate the efficacy of ACR TI-RADS and appropriate recommendation for biopsy in patients with follicular thyroid cancer (FTC).
Methods: We performed a retrospective review of 143 patients who underwent surgery for FTC. Sixty-five patients met the inclusion criteria which included patients with a discrete thyroid nodule and adequate information on pre-operative ultrasound imaging reports to calculate a TI-RADS score. A corresponding TI-RADS score was calculated if not provided in the report. Data were analyzed to determine whether biopsy or follow-up would have been recommended based on ACR TI-RADS score alone. Pearson’s chi-square was then performed to identify any association between TI-RADS scores and disease specific characteristics, and indications for surgery.
Results: Of 65 included patients, the most common indications for surgery were a Bethesda III (30.8%) or IV (27.7%) FNA, followed by multinodular goiter (21.5%). Nearly half of patients (n=30,46%) with FTC had a calculated TI-RADS score of I-II. The remainder were primarily TI-RADS III and IV (III 20.0%, IV 26.2%, V 7.7%). Nodules recommended for biopsy were slightly larger however the difference was not significant (41.2 vs 34.1mm, p=0.171). Utilizing the TI-RADS score algorithm and accounting for nodule size, 30 (46.2%) of the 65 patients would have undergone biopsy. Of those 35 not recommended for biopsy, only 4 (11.4%) would have been recommended to have ultrasound follow up. When further stratified into those recommended for biopsy and those not recommended for biopsy, there was no difference in pathologic characteristics, such as nodule size, presence of vascular or capsular invasion, and preoperative FNA Bethesda classification.
Conclusion: Utilization of the ACR TI-RADS classification system failed to recommend biopsy of more than half of all FTCs, suggesting a shortcoming in broad applicability to thyroid malignancies. As such, the clinician should take particular care with the use of TI-RADS when FTC is suspected. Additional investigations are needed to improve image-based risk stratification of follicular thyroid cancers.