Background: Although conventional CT scan-guided needle biopsy is an established diagnostic method for pulmonary lesions, few large studies have been conducted on the diagnostic outcomes of CT fluoroscopy-guided lung biopsy. We have conducted a retrospective analysis to evaluate the diagnostic outcomes of 1,000 CT fluoroscopy-guided lung biopsies performed with 20-gauge coaxial cutting needles.
Methods: We determined the diagnostic yield of CT fluoroscopy-guided lung biopsies performed with 20-gauge coaxial cutting needles for 1,000 lesions in 901 patients. Independent risk factors for diagnostic failure (ie, nondiagnostic, false-positive, and false-negative results) were determined with multivariate logistic regression analysis.
Results: The biopsy results were nondiagnostic in 0.6% of the lesions (6 of 1,000 lesions). The sensitivity and specificity for the diagnosis of malignancy was 94.2% (741 of 787 lesions) and 99.1% (211 of 213 lesions), respectively; diagnostic accuracy was 95.2% (952 of 1,000 lesions). For lesions measuring <or= 1.0 cm, the diagnostic accuracy was 92.7% (140 of 151 lesions). The significant independent risk factors for diagnostic failure were as follows: the acquisition of two or fewer specimens (odds ratio [OR], 2.43; p = 0.007), lesions in the lower lobe (OR, 2.50; p = 0.003), malignant lesions (OR, 7.16; p = 0.007), and lesions measuring <or= 1.0 cm (OR, 3.85; p = 0.016) and >or= 3.1 cm (OR, 4.32; p = 0.007).
Conclusions: CT fluoroscopy-guided lung biopsy performed with 20-gauge coaxial cutting needles resulted in a high diagnostic yield, even in the case of small lesions. Factors such as the acquisition of two or fewer specimens, lesions in the lower lobe, malignant lesions, and lesions measuring <or= 1.0 cm or >or= 3.1 cm significantly increased the rate of diagnostic failure.