Answers are grounded in this paper and cite their sources.
BACKGROUND: Atypia of undetermined significance (AUS; Bethesda category III) is a heterogeneous thyroid fine-needle aspiration (FNA) category with variable malignancy risk. Molecular testing guides AUS management in North America but is unavailable or cost-prohibitive in much of the world. This study evaluated repeat FNA as primary AUS triage and the contribution of cytomorphologic subclassification in a population-based cohort without molecular testing. METHODS: The authors identified all AUS thyroid FNAs in the Fraser Health Authority, British Columbia (2014-2019), with follow-up through 2022. Index cytology was subclassified by architectural, morphologic/nuclear, and oncocytic atypia. Repeat-FNA outcomes, surgical pathology, and clinical/radiologic data were correlated; noninvasive follicular thyroid neoplasm with papillary-like nuclear features was classified as nonmalignant. RESULTS: Among 611 AUS nodules, 271 (44.4%) underwent repeat FNA and 237 (38.8%) resection. Resection-based risk of malignancy (ROM) was 32.9% (78 of 237); minimum observed malignancy was 12.8% (78 of 611). The second FNA was benign in 116 (42.8%), persistent AUS in 78 (28.8%), nondiagnostic in 58 (21.4%), and upgraded in 19 (7.0%). Minimum observed malignancy was 2.6% (3 of 116) after a benign versus 47.4% (9 of 19) after an upgraded second FNA. A benign repeat FNA obviated surgery in 103 of 116 (88.8%). AUS-nuclear carried a higher ROM than AUS-other (38.6% vs. 21.5%; p = .008; adjusted odds ratio, 2.33; 95% CI, 1.24-4.38) but did not predict repeat-FNA outcome (p = .420). CONCLUSIONS: In a molecular-testing-free, population-based cohort, repeat FNA effectively triaged AUS: benign resolution was frequent and marked very low malignancy risk, whereas persistent or upgraded cytology selected nodules for surgery. Cytomorphologic subclassification provided complementary risk stratification, supporting repeat FNA as cost-effective first-line management where molecular testing is unavailable.
No takes yet. Share an insight, caveat, or question.
Khaitan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: