Narrow band imaging in oral cancer did not improve visualisation of the tumour borders: a prospective cohort study.

Background: In oral cancers, tumour borders are typically defined by white light (WL). Narrow-band imaging (NBI) is an optical endoscopic technique commonly used for the larynx and for cancers of unknown primary. However, evidence for using NBI in oral cancers is insufficient.

Aims/objectives: We investigated whether NBI is a better predictor of the true mucosal tumour borders than WL in oral cancers. Additionally, we examined the agreement between NBI-based Takano intrapapillary capillary loop classifications and pathology reports.

Materials and methods: In this prospective study, the tumour borders were assessed by both NBI and WL at the time of surgical resection and then compared. Pathology reports of the examined areas were used as gold standard.

Results: Forty-nine participants were included. After exclusion of 15 patients due to missing data, 34 were included in analyses. In 26.5% of the assessments, the tumour borders defined by NBI were outside the borders defined by WL. However, 55.5% of these were false-positives.

Conclusions and significance: The delineation of mucosal tumour borders in oral cancers by NBI was not better than that by WL in this study. Several methodological challenges may have influenced the findings of this study, similar to the limitations reported in previous studies.