Image-guided pleural biopsy: diagnostic yield and complications.

ABSTRACT

BACKGROUND: Pleural biopsy and cytology are standard procedures for the investigation of pleural disease. Recent medical literature has suggested that image-guided pleural biopsy shows improved sensitivity for the diagnosis of pleural malignancy, when compared with the more commonly performed reverse bevel needle biopsy such as Abrams' needle. In our centre there has been an increasing trend towards performing image-guided pleural biopsies, and to our knowledge there is no large published series documenting the complication rate and diagnostic yield. METHODS: The radiology and pathology databases were searched for all image-guided [computed tomography (CT) and ultrasound (US)] pleural biopsies from January 2001 to December 2004. All imaging and histology were reviewed, and final diagnostic information about patients was obtained from the respiratory multidisciplinary team database and patient notes. A record was made of complications following biopsy, presence of pleura in the biopsy, and adequacy of tissue for histological diagnosis. RESULTS: A total of 82 patients underwent 85 image-guided pleural biopsies over a 4-year period. 80 cases were performed under CT and five under US guidance. The rate of new pneumothorax detected by chest radiography was 4.7%. No patient required a chest drain or blood transfusion to treat complications. In 10 (12%) cases, there was inadequate tissue to reach a confident histological diagnosis and in eight (9%) of these, no pleura was present. Assuming all suspicious and inadequate biopsies are treated as benign, which is the worst case scenario, image-guided pleural biopsy has a sensitivity and specificity of 76% and 100%, respectively, for the diagnosis of malignant disease. CONCLUSIONS: Image-guided pleural biopsy is a safe procedure with few associated complications and has a higher sensitivity than previously published series for reverse cutting needle biopsy in the diagnosis of malignant pleural disease.