To Evaluate Transoral Flexible Endoscope With Narrow Band Imaging in Nasopharyngeal Cancer

Nasopharyngeal carcinoma (NPC) is endemic in our region and is the 9th most common cancer in Hong Kong. Traditionally diagnosis has been through a nasoendoscopic examination of the nasopharynx with white light followed by a biopsy of suspicious lesions for a confirmatory diagnosis. However, given the geometry of the anatomy of the nasopharynx, with its inherent crevices and varying amounts of lymphoid tissues, lesions are not always easily identified leading to potential missed lesions. The non-specific aspect of white light also leads to excessive biopsies that are not without risk and of some discomfort to patients. Recent advances in liquid biopsies have also allowed for the detection of earlier and smaller lesions that are not always easily identified on nasoendoscopy but rather are seen on Magnetic Resonance Imaging (MRI)1 . An alternative imaging technique is the use of Narrow Band Imaging (NBI) to view the vasculature of the mucosa to identify suspicious lesions for pre-malignancy and malignancy that has been popularized in the gastrointestinal tract. In NPC, NBI with the flexible nasoendoscopes has been used in the diagnosis of NPC with varying success2-5 . Our own group's research has found that in NPC NBI has limitations arising from a lack of consensus on vascular findings on NBI that constitute malignancy, lack of magnification and long focal length of current nasoendoscopes5-8 . Flexible endoscopy using conventional esophago-gastroscopy endoscope (OGD) with NBI and magnification power up to 80x overcome the limitations of current nasoendoscopes, however their size precludes the passing of these endoscopes through the nasal cavity. Here in this pilot study we will seek to use an OGD with NBI passed transorally and retroflexed into the nasopharynx to view the nasopharynx with increased magnification and clarity to evaluate the feasibility of this study in the diagnosis of nasopharyngeal carcinoma. Study questions: 1. Is it feasible to use an OGD with