Risk of malignancy and neoplasia predicted by three molecular testing platforms in indeterminate thyroid nodules on fine-needle aspiration.

ABSTRACT

BACKGROUND: The management of thyroid nodules with indeterminate cytology is challenging. Recently, molecular testing on fine-needle aspirates (FNAs) has been advocated to determine whether clinical follow-up or surgery is warranted for patients. Three different testing platforms were performed on aspirates from our institution (Afirma Thyroid FNA Analysis, RosettaGX Reveal, and Interpace ThyGenX/ThyraMIR). This study compares their diagnostic efficacy. METHODS: We conducted a retrospective analysis of indeterminate thyroid FNAs with correlating molecular testing over 4 years (2015-2018). The aspirates included diagnoses of follicular lesion of undetermined significance, follicular neoplasm, or suspicious for malignancy (SM). Based on cases that underwent surgical resection (Afirma, n = 37; Rosetta, n = 19; Interpace, n = 14), we calculated sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for risk of malignancy and neoplasia. RESULTS: The three tests performed similarly when predicting risk of malignancy. They showed high sensitivity (80-100%) and NPV (90-100%) but lower specificity (10-64%) and PPV (21-44%). When assessing their value to predict neoplasia, each test had a high PPV (76-89%) but low NPV (20-33%). The sensitivity for neoplasm was intermediate to high (50-93%), and the specificity remained extremely variable (11-67%). CONCLUSION: Overall, these molecular platforms performed similarly, displaying high NPV but low to intermediate PPV for malignancy and low NPV but high PPV for neoplasm. The risk of neoplasm is a good index for surgery, and we argue that many of the neoplasms are low-risk tumors. We endorse conservative treatment with lobectomy for cases that are indeterminate at FNA but suspicious by molecular testing.