The detection of thyroid lesions is a frequently encountered especially in the adult population. Data from literature emphasize that they are found in more than 65 % of individuals. Most of these lesions are benign (90-92 %), even though the incidence of malignancy has been increasing due to frequent ultrasonographic head and neck evaluation, which can now identify small subcentimeter suspicious nodules. However, a 20 % of them, falling into the category of indeterminate lesions can lead to some pitfalls and tricky evaluations. Globally Fine needle aspiration (FNA) has been established as a safe, useful, first-line diagnostic tool, with a high positive predictive value for identifying malignancy. The development of classification system originated in order to obtain a practical classification system, able to combine each entity with a category and then with a specific risk of malignancy (ROM) and management. It is well-known that, among the others, The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) represent the most worldwide used system. The 3r edition of TBSRTC, published in July 2023, subclassified indeterminate lesions into: a) atypia of undetermined significance (AUS) with nuclear atypia or other atypia, b) follicular neoplasm (FN) and c) suspicious for malignancy (SFM). However, despite the high positive predictive value (97 %-99 %), sensitivity (65 %-99 %) and specificity (72 %-100 %) of thyroid FNAC, diagnostic pitfalls exist that can lead to false positive (FP) and/or false negative (FN) results. This inconvenience is mostly due to the overlapping of morphological features in terms of cells and even background. This review discusses the most important practical issue also related to the application of TBSRTC and the evaluation of morphological challenges that can lead to pitfalls and diagnostic errors.
Keywords: Classification systems; Fine needle aspiration cytology; Immunocytochemistry; Personalized medicine; Thyroid malignancies.
Copyright © 2025 The Authors. Published by Elsevier Inc. All rights reserved.