Low Prevalence of High-Grade Lesions Detected With Autofluorescence Bronchoscopy in the Setting of Lung Cancer Screening in the Pan-Canadian Lung Cancer Screening Study

Background Lung cancer screening with low-dose CT (LDCT) scan has been demonstrated to reduce lung cancer mortality. Preliminary reports suggested that up to 20% of lung cancers may be CT scan occult but detectable by autofluorescence bronchoscopy (AFB). We evaluated the prevalence of CT scan occult, invasive, and high-grade preinvasive lesions in high-risk participants undergoing screening for lung cancer. Methods The first 1,300 participants from seven centers in the Pan-Canadian Early Detection of Lung Cancer Study who had ≥ 2% lung cancer risk over 5 years were invited to have an AFB in addition to a LDCT scan. We determined the prevalence of CT scan and AFB abnormalities and analyzed the association between selected predictor variables and preinvasive lesions plus invasive cancer. Results A total of 776 endobronchial biopsies were performed in 333 of 1,300 (25.6%) participants. Dysplastic or higher-grade lesions were detected in 5.3% of the participants (n = 68; mild dysplasia: n = 36, moderate dysplasia: n = 25, severe dysplasia: n = 3, carcinoma in situ [CIS]: n = 1, and carcinoma: n = 4). Only one typical carcinoid tumor and one CIS lesion were detected by AFB alone, for a rate of CT scan occult cancer of 0.15% (95% CI, 0.0%-0.6%). Fifty-six prevalence lung cancers were detected by LDCT scan (4.3%). The only independent risk factors for finding of dysplasia or CIS on AFB were smoking duration (OR, 1.05; 95% CI, 1.02-1.07) and FEV₁ percent predicted (OR, 0.99; 95% CI, 0.98-0.99). Conclusions The addition of AFB to LDCT scan in a high lung cancer risk cohort detected too few CT occult cancers (0.15%) to justify its incorporation into a lung cancer screening program. Trial Registry ClinicalTrials.gov; No.: NCT00751660; URL: www.clinicaltrials.gov