Enhanced Detection of Sinonasal and Nasopharyngeal Pathologies by Narrow-Band Imaging- a Prospective Observational Study

Narrow-band imaging (NBI) is an advanced endoscopic technique that uses filtered light to enhance the visualization of microvascular patterns, which can aid in identifying preneoplastic and cancerous lesions in the upper aerodigestive tract. While conventional fiber optic endoscopy under white light (WL) is essential for examining the sinonasal region, NBI offers a unique advantage by focusing on the vascular patterns rather than the lesions themselves. Literature on NBI's diagnostic applications, especially for distinguishing normal from pathological nasal mucosa, remains sparse. This study aims to classify sinonasal lesions based on the Bruno et al. classification system and evaluate the microvascular features of various pathologies under NBI, comparing these findings with histopathological results. A prospective observational study was conducted in a tertiary care setting enrolling 60 patients presenting with endoscopically identified sinonasal and nasopharyngeal lesions requiring biopsy. All patients were subjected to narrow-band imaging, followed by biopsy. Microvascular changes in the mucosa were classified using the Bruno et al. system, and were then compared against blinded histopathology results. Among the 60 patients, 50 exhibited NBI patterns classified as SN 1, 2, 3, or 4 (considered unsuspicious), while 10 displayed the SN 5 pattern (suspicious). Of the lesions categorized as suspicious, 9 were confirmed malignant through biopsy, while 1 was negative for malignancy. Using HPE as the gold standard, NBI demonstrated 88.8% sensitivity and 96% specificity in identifying malignancy with a PPV of 80%, NPV of 98% and diagnostic accuracy of 95%. Narrow-band imaging (NBI) proves to be an effective tool for distinguishing between benign and malignant sinonasal pathologies. Our study contributes valuable insights to the limited literature on the application of NBI technology in nasal pathologies.