Close Surveillance with Long-Term Follow-up of Subjects with Preinvasive Endobronchial Lesions.

RATIONALE: Autofluorescence bronchoscopy (AFB) and computed tomography (CT) enable lung cancer (LC) detection at the early (pre-)invasive stage. However, LC risk in patients with preinvasive endobronchial lesions is unclear. OBJECTIVES: To assess LC incidence and identify potential risk determinants in patients with preinvasive lesions. METHODS: In our tertiary care referral center, 164 subjects with preinvasive lesions were monitored up to 12.5 years by repeated AFB and CT. Occurrence of LC was monitored. Clinical management depended on histological grade, with cancer patients receiving standard care. Potential risk determinants (smoking status, baseline histology, cancer history, and chronic obstructive pulmonary disease [COPD] status) were evaluated in relation to cancer occurrence, event-free survival (EFS), and overall survival (OS). MEASUREMENTS AND MAIN RESULTS: During surveillance (median of 30 mo, range 4-152) of 164 subjects with preinvasive lesions (80 high grade and 84 low grade at inclusion), 61 LCs were detected in 55 subjects (median time to event 16.5 mo). Twenty-three LCs (38%) were detected by CT, and 38 (62%) were detected by AFB. More cancers (36 of 61; 59%) developed from separate, rather than initial lesional sites. Subjects with high-grade lesions were more likely to be diagnosed with LC at the same or another site in the lungs than those with low-grade lesions (P = 0.03). Independent risk determinants for OS were previous curatively treated cancer and COPD (P ≤ 0.05). CONCLUSIONS: Presence of preinvasive lesions, especially high-grade lesions, may serve as LC risk markers. LCs occur both at preinvasive lesion sites and elsewhere in the bronchial epithelium or lung parenchyma. Prospective validation of biomarkers and randomized intervention studies are needed to determine optimal management strategies.