20 Molecular and Cytomorphologic Comparison of Bethesda System Versions 1.0 and 2.0: A Study of Thyroid FNA Categories III and IV in a Safety-Net Hospital

Abstract Introduction/Objective Thyroid fine needle aspiration (FNA) specimens are currently classified using the Bethesda System for Reporting Thyroid Cytopathology (TBSRTC). The updated 2023 edition of TBSRTC removed the terms Follicular Lesion of Undetermined Significance (FLUS) from Category III and Suspicious for Follicular Neoplasm (SFN) from Category IV. Additionally, Category III (Atypia of Undetermined Significance, AUS) has been further subclassified into AUS—nuclear atypia and AUS—other, based on predominant cytomorphologic features. In this study, we retrospectively analyzed thyroid FNAs classified as Bethesda Categories III and IV at our institution, with a focus on molecular findings and implications for classification under the updated system. Methods/Case Report We performed a retrospective review of all thyroid FNAs diagnosed as Bethesda Categories III and IV between 2020 and 2022, as retrieved from our laboratory information system. Category III cases were subclassified as AUS, FLUS, AUS/FLUS, or Hürthle cell lesion of undetermined significance. Category IV cases were subclassified as follicular neoplasm (FN), suspicious for follicular neoplasm (SFN), or FN/SFN. When available, molecular testing results were reviewed for mutation type, associated malignancy risk, and correlation with cytomorphologic features. Results A total of 427 thyroid FNA cases were reviewed. Category III comprised 383 cases: 274 (71.5%) FLUS, 94 (24.5%) AUS, 14 (3.6%) AUS/FLUS, and 1 (0.3%) Hürthle cell lesion of undetermined significance. Molecular testing revealed mutations in 46 cases (12%): 37 FLUS and 9 AUS. No mutations were detected in AUS/FLUS or Hürthle cell lesion cases. The most frequently identified alterations included BRAF (n = 6, including BRAF-TERT fusion), HRAS (n = 6), NRAS (n = 3), RET/PTC (n = 2), and single cases of KRAS, TSHR, DICER1, and P53. Category IV included 44 cases: 12 (27.3%) FN, 25 (56.8%) SFN, and 7 (15.9%) FN/SFN. Mutations were detected in 11 cases (25%): 2 FN, 7 SFN, and 2 FN/SFN. Detected alterations included HRAS (n = 6), NRAS (n = 4), KRAS (n = 1), BRAF (n = 1), and a TERT promoter mutation (n = 1). Cytopathologists at our institution tended to classify cases with architectural atypia or combined nuclear and architectural atypia as FLUS, while AUS was more commonly assigned to cases with isolated nuclear atypia. The distinction between FN and SFN appeared subjective, with notable interobserver variability. Conclusion Our findings suggest an overutilization of Category III, particularly in cases that may better align with Categories II or IV under the updated TBSRTC. With the reclassification of FLUS into AUS—other and the removal of the SFN subcategory, there is a risk of increased reliance on Category III, especially when cytopathologists are reluctant to assign a definitive follicular neoplasm diagnosis. Continued evaluation is warranted to assess how the revised TBSRTC impacts diagnostic categorization and patient management.