Fine-needle aspiration cytology in tumors and tumor-like conditions of the oral and maxillofacial region: diagnostic reliability and limitations

Abstract

Background: Fine-needle aspiration cytology (FNAC) is a well established diagnostic technique. The purpose of the current study was to evaluate its diagnostic accuracy, sensitivity, and specificity in tumors and tumor-like conditions in the oral and maxillofacial region.

Methods: The results of 1022 cases of FNAC performed over a 7-year period (1990-1996) were retrospectively reviewed and analyzed. The material was collected from 228 major salivary glands, 67 perioral soft tissues and facial bones, 139 tumors of the oral cavity, 18 paranasal tumors, 41 oropharyngeal lesions, and 21 nasopharyngeal lesions, whereas 508 cases represented cervical lymph nodes.

Results: In 804 cases FNAC established or documented the initial diagnosis whereas in 198 cases it confirmed suspected recurrences. Definitive histologic diagnosis was established in 809 of the 1022 cases after surgical treatment and the histologic diagnoses were compared with preoperative FNAC results. In 20 cases the cytologic material was considered unsatisfactory and therefore the cytologic diagnosis was impossible to establish. In 12 of these cases diagnosis was made based on histologic examination after surgery. Malignancy was found in three of the latter cases. In the total number of 1022 cases there were 18 false-negative and 2 false-positive cytologic results. In another 19 cases a difference in the histologic type of the lesion was found. The diagnostic accuracy of the method was 98.2% with a sensitivity of 96.2% and a specificity of 99.4%.

Conclusions: Four major causes of error were identified: material adequacy criteria, sampling technique, cytologic interpretation, and limitations of the procedure. FNAC proved to be a safe, accurate, inexpensive, and patient-friendly in the effort to establish preoperative diagnosis in tumors and tumor-like conditions of the oral and maxillofacial region. The authors have proposed modification in the mathematic computation of diagnostic accuracy, adding the parameter of nonconcordance with histologic typing of FNAC.