How reliable is contact endoscopy of the nasopharynx in patients with nasopharyngeal cancer?

Abstract

To evaluate diagnostic reliability of contact rhinoscopy in patients with nasopharyngeal carcinoma. A cross-sectional randomized single-blinded study was carried out to evaluate the reliability of the findings of contact rhinoscopy in patients with nasopharyngeal carcinoma assessed by different observers. Random images of contact endoscopy of 157 subjects including normal patients and patients with nasopharyngeal carcinoma, both prior to external beam radiation therapy (RT), the non-RT group, and after external beam radiation therapy, the post-RT group, were available for analysis. Three independent observers of varying clinical experience and histopathologic knowledge viewed the images and made a diagnosis for each image. The diagnosis of each image was correlated with the histology of the biopsy from the corresponding patient. In the non-RT group, Kappa values for inter- and intraobserver reliability to differentiate normal epithelium from primary nasopharyngeal carcinoma were 0.894, 1.000, 0.794 and 0.694, 1.000, 0.776 for the house officer, radiologist, and pathologist, respectively. There were no significant differences in interobserver and intraobserver reliabilities between the assessors. In the post-RT group, Kappa values for the inter- and intraobserver reliability to diagnose recurrent carcinoma or atypia, squamous metaplasia, and radiation change for the three observers were 0.820, 0.718, 0.775 and 0.731, 0.622, 0.734, respectively. There were no significant differences in interobserver and intraobserver reliabilities between the assessors. The clinical diagnosis of nasopharyngeal carcinoma by contact endoscopy is highly reliable and is not dependent on the clinical experience or knowledge of histopathology of the observers.