A study on the value of narrow-band imaging (NBI) for the general investigation of a high-risk population of nasopharyngeal carcinoma (NPC).

Background: This study aims to explore the feasibility of narrow-band imaging (NBI) applied for the diagnostic screening of a high-risk population of nasopharyngeal carcinoma (NPC) and increase the accuracy rate of nasopharyngeal biopsy and the diagnosis rate of early-stage patients.

Methods: The positive high-risk population of NPC to EB virus antibody was followed up. At the same time, serological screening and pharyngorhinoscopy were carried out. The specific methods were as follows: (1) all subjects received nasopharyngeal examinations through both the HD endoscopic white light mode (WL) and NBI mode, (2) nasopharyngeal biopsy was conducted on positive subjects with microscopic examination, and, finally, (3) a comparative analysis was conducted between the biopsy pathology results and microscopy results. In addition, the following comparative indicators were recorded under different modes: sensitivity, specificity, accuracy, positive likelihood ratio, and negative likelihood ratio. Then, the area under the ROC curve and the kappa coefficient were calculated.

Results: A total of 115 subjects were detected to be positive by microscopic examination under the WL mode. Among these subjects, 19 subjects were diagnosed with NPC. In addition, 24 subjects were detected to be positive by microscopic examination under the NBI mode. Among these subjects, 23 subjects were diagnosed with NPC. Under the WL mode, the specific values of the comparative indicators were as follows: sensitivity, 82.61%; specificity, 0%; and area under the ROC curve, 0.413. Furthermore, the WL mode in the diagnosis on the high-risk population of NPC exhibited poor consistency with the biopsy pathology results (kappa coefficient = - 0.069). Under the NBI mode, the specific values of the comparative indicators were as follows: sensitivity, 100%; specificity, 98.96%; and area under the ROC curve, 0.995. Furthermore, the NBI mode in the diagnosis on the high-risk population of NPC exhibited relatively satisfactory consistency with the biopsy pathology results (kappa coefficient = 0.973). Therefore, the NBI mode is significantly superior to the WL mode.

Conclusion: NBI endoscopic examinations should be conducted on a routine basis for high-risk populations of NPC. This can decrease the frequency of biopsies and enhance diagnostic effects.