RÉSUMÉ
BACKGROUND: Controversy exists concerning the utility of fine needle aspiration cytology (FNAC) in diagnosing salivary masses. Thus, the goal of this study was to evaluate the efficacy of preoperative fine needle aspiration cytology compared with those of frozen section (FS) and final histopathologies in terms of sensitivity, specificity, accuracy and pitfalls. METHODS: A total of 66 aspirates were from the salivary gland, 50 (75.8%) from the parotid gland and 16 (24.2%) from the submandibular gland, from Jan. 1990 to Jul. 1998. The results from 49 of 66 aspirates were compared with the results from frozen-section pathology. RESULTS: The 66 cases included 59 cases of benign lesions (49 cases of pleomorphic adenomas, 4 cases of basal cell adenomas, 3 cases of Warthin's tumors, and 3 cases of other benign tumors) and 7 cases of malignant lesions (3 cases of mucoepidermoid carcinomas and 4 cases of other malignant tumors). The accuracy in distinguishing malignant from benign lesions was 89.4% for FNAC. The sensitivity, the specificity, and the type-specific accuracy of FNAC were 89.7%, 96.3%, and 70.0%, respectively. A comparison of the cytopathologic diagnosis and the FS diagnosis was made. The accuracy of the FS in distinguishing malignant lesions was 93.8%. The sensitivity, the specificity, and the type-specific accuracy of FS were 80.0%, 95.4%, and 87.7%, respectively. The salivary gland tumors mainly misdiagnosed in FNAC were basal-cell adenomas and Warthin's tumor in benign disease and mucoepidermoid carcinomas in malignant tumors. CONCLUSION: Though FNAC still has some problems and pitfalls, the comparative accuracies of FNAC and FS are similar. A review of the value of FS diagnosis in the surgical treatment of salivary gland lesions suggests that FS diagnosis alone should not determine the surgical management of salivary gland tumors. FNAC is a sensitive and specific diagnostic tool and is complementary to a frozen section biopsy of a salivary gland mass.