Affiliation:
1. Allegheny Health Network Pittsburgh Pennsylvania USA
2. Diane Vido Thompson LLC Murrysville Pennsylvania USA
3. Interpace Diagnostics, Inc Parsipanny New Jersey USA
Abstract
AbstractBackgroundMolecular analysis of fine‐needle aspiration biopsies (FNAB) improves the diagnostic accuracy of cytologically indeterminate thyroid nodules (ITNs). Recently, the use of MPTXv2 has been shown to further improve the accuracy of risk stratification of ITNs.MethodsA total of 338 patient samples with atypia of undetermined significance (n = 258) or follicular neoplasm (n = 80) cytology diagnosis and corresponding surgical outcomes or clinical follow‐up, collected between 2016 and 2020 were included [Correction added on 19 June 2024, after first online publication: In the preceding sentence, the n values 260 and 78 have been changed to 258 and 80, respectively.]. All samples underwent multiplatform testing (MPTXv1), which includes an oncogene panel (ThyGeNEXT®) plus a microRNA risk classifier (ThyraMIR®). A blinded, secondary analysis was performed to assess the added utility of MPTXv2 (ThyraMIR®v2). The average length of follow‐up for the surveillance group (n = 248) was 30 months.ResultsSensitivity at moderate threshold was 96% and specificity at positive threshold was 99% for MPTXv2. At 14% disease prevalence, the negative predictive value at the moderate threshold was 99% and the positive predictive value at the positive threshold was 89% for MPTXv2. MPTXv2 had fewer patients classified into the moderate‐risk group than MPTXv1, which was statistically significant (p < .001). Using surgical resection, the gold standard for outcomes, MPTXv2 showed a statistically greater area under the curve (p = .028) than MPTXv1, demonstrating greater accuracy for MPTXv2.ConclusionBoth test versions demonstrated robust performance with low false‐positive molecular results. Data suggest that incorporation of MPTXv1, and more recently MPTXv2, into clinical practice within our healthcare network resulted in improved accuracy of ITN risk stratification.