Interobserver Agreement in Interpreting Thyroid FNA with a Diagnosis of Atypia/Follicular Lesion of Undetermined Significance (FLUS/AUS)
Constanino Theoharis, Mitchell Wachtel, Zubair Baloch, Marilyn Bui, Isam Eltoum, Oscar Lin, Aylin Simsir, Ed Stelow, Adebowale Adeniran, Guoping Cai, Malini Harigopal, Angelique Levi, Berrin Ustun, Meiklejohn McKenzie, Diane Kowalski, David Chhieng. Yale University, New Haven, CT; University of Virginia, Charlottesville, VA; University of Pennsylvania, Philadelphia, PA; Memorial Sloan-Kettering Cancer Center, New York, NY; University of Alabama, Birmingham, AL; Moffit Cancer Center, Miami, FL; Texas Tech University, Lubbock, TX; New York University, New York, NY
Background: The Bethesda System for Reporting Thyroid Cytopathology (BSRTC) was developed to refine definitions and improve clinical communication and management. However, the diagnostic category of atypia/follicular lesion of undetermined significance (FLUS/AUS) remains heterogenous in terms of usage and clinical outcome. Because of the “gray zone” that exists in the interpretation of thyroid FNA demonstrated minor architectural and/or cytologic atypia, this study was undertaken to evaluate the degree of interobserver agreement in the evaluation of thyroid FNAs originally interpreted as FLUS/AUS.
Design: Twenty-three thyroid FNAs including 18 cases originally diagnosed as FLUS/AUS, 2 as negative for malignant cells, 2 as positive for malignancy, and 2 as follicular neoplasms were selected. Two representatives slides from each case were circulated to 13 board certified or eligible cytopathologists; all were from academic institutions with 6 were from the same institution. Each reviewer was asked to evaluate each cases using the BSRTC. The kappa statistics was calculated.
Results: Only 2 cases (22%) were in complete agreement: one originally interpreted as negative and one as positive for malignancy; both cases were confirmed on histology. There was a majority agreement (among 10 or more) in 6 cases: one originally diagnosed as positive, one negative, and 4 FLUS/AUS. Both positive and negative cases were confirmed on histology; among the 4 FLUS/AUS, 2 were found to be negative and 2 follicular adenoma on histology. For the entire group of reviewers, the mean kappa statistic was 0.34±0.13. The mean kappa statistic was 0.42±0.7 and 0.29±0.14 among reviewers of the same institution and among reviewers from different institutions. The difference was statistically significant.
Conclusions: The interobserver agreement for thyroid FNA cases originally classified as FLUS/AUS was fair among academic cytopathologists. It appeared that the interobserver agreement was better among cytopathologists who were from the same institution.
Wednesday, March 6, 2013 1:00 PM
Poster Session VI # 43, Wednesday Afternoon