ABSTRACT
RATIONALE AND OBJECTIVES: Only a few studies have systematically evaluated risk factors for pneumothorax and pulmonary hemorrhage in computed tomographically (CT)-guided transthoracic lung biopsy (TLB). We evaluated the diagnostic yield of CT-guided TLB and determined risk factors for pneumothorax and hemorrhage. METHODS: One hundred seventy-two CT-guided TLBs were performed on 159 patients (mean age 66 +/- 11 years; 72% male) using a 16-gauge core biopsy needle. Lesion and patient characteristics, lung function analysis, CT signs of emphysema, histopathologic diagnoses, and complications were recorded. Statistical analysis was performed with multivariate regression analysis. RESULTS: Histopathologic diagnosis was established in 153 cases (89%). Although lesion size was higher (47 +/- 29 vs. 43 +/- 35 mm, P = .191) and depth was lower (22 +/- 23 vs. 6 +/- 23 mm, P = .350) in procedures with histopathologic diagnosis, no parameter showed significant impact on diagnostic yield. Sensitivity and specificity for detection of malignancy were 93% and 100%, respectively, whereas positive and negative predictive values were 100% and 88%. Overall accuracy was 95%. Pneumothorax occurred in 45 procedures (26%). Hemorrhage was recorded in 17 procedures (10%). There was higher frequency of pneumothorax in smaller lesions (35 +/- 23 vs. 50 +/- 31 mm, P = .003; odds ratio = .96) and greater depth (29 +/- 29 vs. 20 +/- 19 mm, P = .05; odds ratio = 1.03). CT signs of emphysema revealed higher incidence of hemorrhage (35% vs. 23%; P = .04; odds ratio=41.03). Other parameters were nonsignificant. CONCLUSIONS: The high diagnostic yield of CT-guided TLB was not affected by lesion characteristics or emphysema. Pneumothorax rate was influenced by lesion size and depth. Hemorrhage was associated with CT signs of emphysema.
