Related Experiment Video
Updated: Mar 31, 2026
Author Spotlight: Integrating Ultrasound Imaging with Biochemical Markers for Thyroid Disease Diagnosis
Published on: February 9, 2024
Decreased Interobserver Variability in Nuclear Atypia of Undetermined Significance After Consensus Review of Thyroid
Heather Chen-Yost1, Richard Cantley1, Xiaobing Jin1
1Department of Pathology, Michigan Medicine, University of Michigan, Ann Arbor, Michigan, USA.
Introduction:
The 2023 Bethesda System for Reporting Thyroid Cytopathology (TBS) subcategorizes Atypia of Undetermined Significance (AUS) into two tiers: nuclear atypia and other. Despite this standardized reporting system, limitations still exist, specifically the low reproducibility of AUS. We sought to determine if group review can improve interobserver variability with the two-tiered AUS before implementation at our institution.
Methods:
A total of 155 liquid-based preparations (ThinPrep) from 2020 were reviewed independently by 5 board-certified cytopathologists (CP) and classified using the 6 categories of the third edition of TBS. Classification was concordant if four of five CPs agreed. Discordant cases were subsequently reviewed together for consensus. Clinical outcomes were recorded after completion of the slide review.
Results:
Discordance decreased from 36.1% for individual review to 7.8% after group review, most commonly amongst TBS I, II, and III. Although concordance was achieved during group review for most cases, 21.4% remained discordant, the majority of which were also amongst these categories (7/12; 58.3%). For TBS III-other and TBS III-nuclear, there were no discordances after it was emphasized at group review that nuclear atypia should only be used for cases of "cannot exclude papillary thyroid carcinoma" with conservative use in oncocytes. TBS III-nuclear was more predictive of malignancy than TBS III-other, which had no malignant outcomes.
Conclusion:
Interobserver variability remains for TBS I, II, and III-other despite group review. The implementation of the two-tiered AUS will simplify our clinical practice and is expected to decrease interobserver variability in the use of AUS-nuclear atypia, which has a higher risk of malignancy.
