Utility of thin-layer preparations in thyroid fine-needle aspiration: Diagnostic accuracy, cytomorphology, and optimal sample preparation

BACKGROUND. The efficacy of preparing thyroid fine-needle aspirations (FNAs) as thin-layer slides has not been evaluated extensively. METHODS. To evaluate the efficacy of thin-layer cytology, the authors examined thyroid FNAs from 135 unselected palpable lesions that were aspirated using uniform procedures and prepared as air-dried, Diff-Quik(R)-stained direct smears (DS) and Papanicolaou-stained thin-layer slides (TL). Diagnoses rendered independently on masked slides from each thyroid nodule were compared with the reported final cytologic or histologic diagnoses based on all available pathologic material. Slides were analyzed for diagnostically important cytologic features and the optimal number of TL slides per sample was assessed. RESULTS. TL diagnoses agreed with final diagnoses in 85% of cases compared with 96% for DS. All 11 neoplasms were recognized in both preparations. A final diagnosis of chronic lymphocytic thyroiditis (CLT) was missed in 10 of 26 TL (39%) compared with 2 of 26 DS (8%) (P = 0.013, McNemar's test). Cytologic features appreciated less often in TL than DS (McNemar's test) included diffuse colloid, spherules, tissue fragments, and lymphocytes. Multinucleated giant cells were identified more often in TL than DS. Two TL slides were sufficient for diagnosis in 83% of cases. CONCLUSIONS. The diagnostic accuracy of TL was 85% compared with 96% for DS. CLT was diagnosed accurately in 62% of TL compared with 92% of DS. Cytologic features in TL and DS may differ. Preparation of only two TL slides is adequate for definitive diagnosis in most cases.