INTRODUCTION

Thyroid cancer is the most common endocrine related malignancy in the Philippines. Data showed Filipino patients are at higher risk to develop thyroid malignancy with an increasing incidence annually. Currently, the initial screening test utilized to evaluate thyroid nodules is ultrasonography with studies showing promising results in detecting and evaluating thyroid carcinoma employing the use of the Thyroid Imaging, Reporting and Data System (TIRADS). TIRADS is a standardized classification system to evaluate and characterize thyroid nodules. However, there are studies stating that TIRADS is of limited clinical value for risk stratification of indeterminate cytological results.

OBJECTIVES

The primary objective of this study is to determine the correlation of the results of TIRADS, Fine Needle Aspiration Cytology (FNAC), and histopathology of thyroid nodules among patients who underwent thyroidectomy at Makati Medical Center from January 2016 to March 2020.

METHODS

This is a retrospective, analytical, observational, cross-sectional study wherein medical records of patients who were diagnosed with thyroid nodules goiter who underwent thyroid ultrasound with TIRADS scoring, Fine needle Aspiration Biopsy (FNAB) and ultimately thyroidectomy were reviewed. The primary endpoint included diagnostic performance of TIRADS classification and the possible factors that may contribute to discordance to FNAB and Histopathology.

RESULTS

One hundred twenty-five patients who underwent thyroidectomy were reviewed. These patients underwent thyroidectomy on the basis of their fine needle aspiration biopsy results. With FNAB as a reference standard, TIRADS had good sensitivity of 100% and low specificity of 27.7% in detecting thyroid malignancy. Patients who had FNAB positive or suspicious for malignancy are 1.37 times more likely to yield a positive TIRADS compared to patients who are FNAB negative (LR+), and 94% less likely to yield a negative TIRADS result (LR-). When TIRADS is positive, the positive predictive value was 31.3% and when TIRADS is negative, the negative predictive value (NPV) was nearly 100%. Overall, the accuracy of TIRADS in thyroid malignancy is 45.6% with ROC area at 0.638, indicating fair discriminative power of TIRADS to differentiate between benign vs malignant thyroid nodules. With histopathology as a reference standard, TIRADS had good sensitivity of 96.3% and low specificity of 33.8%. Patients who are histopathology positive are 1.45 times more likely to yield a positive TIRADS compared to patients who are histopathology negative (LR+), and 89% less likely to yield a negative TIRADS result (LR-). When TIRADS is positive, the positive predictive value was 52.5% and when TIRADS is negative, the negative predictive value (NPV) was 92.3%. Overall, the accuracy of TIRADS in thyroid malignancy is 60.8% with ROC area at 0.65, indicating fair discriminative power to differentiate between benign versus malignant thyroid nodules.

CONCLUSION

TIRADS classification provides high sensitivity value in detecting thyroid malignancies but has fair discriminative power to differentiate between benign versus malignant thyroid nodules. Factors that are associated with discordant classification between TIRADS and FNAB were seen in those who underwent total thyroidectomy with lymph node dissection, and solid composition. There is insufficient evidence to determine whether any of the patient or nodule characteristics were associated with discordance between TIRADS and histopathology.

Human ; Thyroid Nodule