Objective: Fine needle aspiration (FNA) cytological analysis fails to confirm the benignity or malignancy of Bethesda III, IV and V thyroid nodules. Molecular tests performed on FNA samples have demonstrated interesting results in improving the diagnosis of these nodules. The aim of this study was to assess the performance of a large next-generation sequencing (NGS) panel in thyroid nodules with indeterminate cytology (Bethesda III, IV, V).
Methods: Retrospective, monocentric study including 121 patients with cytologically indeterminate thyroid nodules (Bethesda III, IV and V), which underwent a routine FNA procedure for molecular testing, using the AmpliSeq general cancer NGS panel, with available final histological diagnosis. The main objective was to estimate the negative predictive value (NPV) of malignancy of the AmpliSeq panel in Bethesda III, and IV thyroid nodules. Performance assessment (sensitivity, specificity, positive predictive value (PPV) and NPV) was carried out in the grouped categories III and IV, in the overall cohort, and in each Bethesda category. The final histological diagnosis was used as the designated gold standard.
Results: Histologically, 86 nodules were benign and 35 malignant. Molecular analysis yielded a positive result in 40 nodules. Panel performances assessed in the grouped categories Bethesda III and IV, demonstrated a 55.0% (CI95% [31.5;76.9]) sensitivity, a 76.9% (CI95% [66.0;85.7]) specificity, a 37.9% (CI95% [25.7;51.9]) PPV and an 87.0% (CI95% [80.2;91.7]) NPV, considering a 20 % prevalence of malignancy.
Conclusions: The performances of the AmpliSeq panel are promising, however, the NPV is not sufficient to avoid diagnostic surgery in cytologically indeterminate thyroid nodules.