DUAL DIGITAL VIDEO-AUTOFLUORESCENCE IMAGING FOR DETECTION OF PRE-NEOPLASTIC LESIONS

Abstract

Summary Aim: The incorporation of autofluorescence (AF) to white light bronchoscopy has led to improved sensitivity for the detection of pre-neoplastic lesions in the airways. However, AF has difficulty distinguishing benign epithelial changes such as bronchitis, previous biopsy, and airway fibrosis from pre-invasive lesions, which necessitates extensive biopsy. This frequently results in longer procedural time and need for additional sedation that may compromise patient safety, increase the risk of bronchospasm, and bleeding from multiple endobronchial biopsies. We postulate that dual imaging with simultaneous video and AF bronchoscopy of the tracheobronchial tree could improve the low specificity observed with AF in the detection of pre-invasive lesions, leading to targeted biopsy, good correlation with pathological diagnosis and shorter procedural time. Methods: Forty-eight patients with known or suspected of lung cancer underwent video and AF bronchoscopy, which were provided as real-time dual images with SAFE 3000 (Pentax, Tokyo) between March and August 2006. Biopsy specimens were taken from all suspicious areas with two random specimens from normal areas. Values were expressed as median and range, and p < 0.05 was considered statistically significant. Results: Twenty-five suspicious sites were detected by dual imaging bronchoscopy, and 126 endobronchial biopsies were evaluated, of which 22 (17.5%) were graded as moderate dysplasia and worse. Sensitivity and specificity of dual imaging for the detection of high-grade dysplasia were 86% and 94%, respectively, with good correlation between bronchoscopic assessment and pathology (r = 0.77, p < 0.0001). However, there were three random biopsy specimens obtained from normal or abnormal sites that showed severe dysplasia in two and moderate dysplasia in one. Median time taken for airway examination was 4 min (range, 4—4.8), and 5 min (range, 4—5) for biopsy, giving a total procedural time of 9 min (range, 8—10). There were no procedurerelated complications noted.