Computed tomography-guided bronchoscopy in the diagnosis of small peripheral pulmonary lesions: A retrospective study of 240 examinations in a single academic center.

Abstract

Factors that affect the diagnostic yield in computed tomography (CT)-guided bronchoscopy have not yet been fully evaluated. To improve the diagnostic yield of peripheral pulmonary lesions (PPLs) by CT-guided bronchoscopy, we quantitatively analyzed factors affecting the diagnostic yield. The data were collected for 240 PPLs in 237 patients examined by using CT-guided bronchoscopy between October 2003 and November 2011 in our respiratory center. The association of diagnostic yield with the CT bronchus sign (CT-BS), lesion size, location, number of tissue specimens, and type of bronchoscope was retrospectively assessed. The diagnostic yield of PPLs with negative CT-BS was significantly lower (2.9%) than that for PPLs with positive CT-BS (52.2%; p<0.01). Among the PPLs with positive CT-BS, the yield was significantly higher in those in the left S(3) than for lesions in other bronchial segments (83.3% vs. 50.3%; p<0.05). Lesion size was not significantly associated with diagnostic yield. The yield was significantly lower in PPLs without lung tissue specimens than in lesions with biopsy specimens (p<0.01). Moreover, a thin bronchoscope produced a higher yield in comparison with other bronchoscope types (66.0% vs. 47.6%; p<0.05). Multivariate analysis revealed that the number of biopsy specimens was an independent factor affecting diagnostic yield. CT-guided bronchoscopy is valuable in the diagnosis of PPLs with positive CT-BS regardless of lesion size; however, PPLs with negative CT-BS are not good candidates for CT-guided bronchoscopy. Obtaining tissue specimens by biopsy is a critical factor in diagnosing PPLs.