Background: Prior case series have shown promising diagnostic sensitivity for CT scan-guided bronchoscopy.
Methods: This was a prospective randomized trial comparing CT scan-guided bronchoscopy vs conventional bronchoscopy for the diagnosis of lung cancer in peripheral lesions and mediastinal lymph nodes. All procedures were performed using a protocolized number of passes for forceps, transbronchial needles, and brushes. Cytologists and pathologists were blinded as to bronchoscopy type. Patients with negative results underwent open surgical biopsy (for nodules or lymph nodes) or were observed for > or = 2 years if they had a nodule < 1 cm in size.
Results: Fifty patients were enrolled into the study (CT scan-guided bronchoscopy, 26 patients; conventional bronchoscopy, 24 patients). Two patients, one from each arm, dropped out of the study. Ultimately, 36 patients were proven to have cancer, and 27 of these patients (75%) had their diagnosis made by bronchoscopy. The sensitivity for malignancy of CT scan-guided bronchoscopy vs conventional bronchoscopy for peripheral lesions was similar (71% vs 76%, respectively; p = 1.0). The sensitivity for malignancy of CT guided bronchoscopy vs conventional bronchoscopy for mediastinal lymph nodes was higher (100% vs 67%, respectively) but did not reach statistical significance (p = 0.26). On a per-lymph-node basis, there was a trend toward higher diagnostic accuracy with CT scan guidance (p = 0.09). The diagnostic yield was higher in larger lesions (p = 0.004) and when CT scanning confirmed target entry (p = 0.001).
Conclusion: We failed to demonstrate a significant difference between CT scan-guided bronchoscopy and conventional bronchoscopy for the diagnosis of lung cancer in peripheral lesions and mediastinal lymph nodes. Further study of improved steering methods combined with CT scan guidance for the diagnosis of lung cancer in peripheral lesions is warranted.