High-Risk Ultrasound Findings Are Superior to Afirma Molecular Profiling for Prediction of Malignancy in Resected Thyroid Nodules: An Institutional Experience

INTRODUCTION: Afirma testing of thyroid nodules is used to stratify cancer risk for patients when fine needle aspiration (FNA) is indeterminate, and to decrease unnecessary operation. However, Afirma is associated with high cost, relatively low specificity, and positive predictive value (PPV). We hypothesized that high-risk ultrasound findings (HRUF) provide useful information for the management of indeterminate thyroid nodule in our population. METHODS: Data were retrospectively collected from all patients who underwent ultrasound, FNA, Afirma testing, and subsequent resection from 2015 to 2021. After inclusion criteria were met, 133 patients were divided based on postoperative pathological findings, and statistical analysis was conducted. RESULTS: In 133 patients, HRUF was associated with malignant surgical pathology (p < 0.001). Afirma did not significantly correlate with malignancy (p = 0.372) with >63% of suspicious cases being benign on postoperative pathology. HRUF had a PPV of 44.8% compared with Afirma’s 34.3%. Also, negative predictive value of HRUF was higher than Afirma at 88.8% vs 74.3%, respectively. Last, HRUF has higher sensitivity/specificity than Afirma with 86.6%/50.0% compared with 72.7%/36%, respectively. CONCLUSION: In our study, HRUF was more predictive of malignancy than Afirma in resected nodules. Afirma was previously shown to predict benign pathology in cytologically indeterminate nodules, decreasing unnecessary operation; however, some findings demonstrate a lower specificity and PPV than HRUF in certain populations. In our study, HRUF was superior to Afirma in predicting postoperative pathology. This suggests that simple, accessible, reproducible tests, such as ultrasound, may be a practical and cost-effective alternative to Afirma, or may be used in conjunction with Afirma.