False negatives in thyroid cytology: impact of large nodule size and follicular variant of papillary carcinoma.

ABSTRACT

OBJECTIVES/HYPOTHESIS: Fine-needle aspiration (FNA) cytology is well established in the diagnosis of thyroid nodules. However, false-negative rates for malignancy of 3% to 10% are reported. The purpose of the present study was to investigate the impact of nodule size and follicular variant of papillary carcinoma (FVPTC) on false-negative FNA rates in thyroid nodules and on malignancy rates in nodules with indeterminate cytology. STUDY DESIGN: Retrospective study. METHODS: A total of 765 consecutive ultrasound-guided FNAs were reviewed. Histological correlation was available in 262 cases. RESULTS: The overall sensitivity of FNA for malignancy was 84%, and the false-negative rate 9.1%. Nodules ≥ 3 cm were significantly more likely to ultimately be diagnosed as cancer by histology than nodules <3 cm (14% vs. 6.8%, P = .006); however, they were also significantly more likely to undergo surgery than smaller nodules (P < .0001). Among the surgical series, the false-negative rate was 10.9% in nodules ≥ 3 cm and 6.1% in nodules <3 cm (P = .71). Most false negatives were due to FVPTC. FVPTC was significantly more likely to be missed by preoperative cytology than conventional or other variants of papillary carcinoma (P < .001). Among cases with indeterminate cytology, nodule size and Thy-3f versus Thy-3a subclassification did not have any significant impact on likelihood of malignancy. CONCLUSIONS: The sensitivity of FNA for detection of FVPTC is reduced compared to conventional papillary carcinoma. The impact of nodule size is not significant. LEVEL OF EVIDENCE 4.