Coaxial percutaneous biopsy technique with automated biopsy devices: value in improving accuracy and negative predictive value.

Abstract

Three hundred percutaneous biopsies were performed in 267 consecutive patients by means of a coaxial technique with use of 18- and 20-gauge automated cutting biopsy devices and 22-gauge aspiration needles. Thoracic, hepatic, renal, pancreatic, adrenal, splenic, retroperitoneal, and musculoskeletal soft-tissue masses were sampled. For malignant masses (229 cases), the sensitivity was 79% for cytologic analysis, 88% for histologic analysis, and 92% for both combined. In benign disease (71 cases), a correct specific diagnosis was made with cytologic analysis in 38%, with histologic analysis in 97%, and with both combined in 97%. The negative predictive value was 60% for cytologic analysis, 72% for histologic analysis, and 80% for both combined. When only cancer-negative results in which a specific benign diagnosis was made were considered, the negative predictive value was 100% for cytologic analysis, 97% for histologic analysis, and 97% for both combined. The positive predictive value was 100% for both cytologic and histologic analysis. Bleeding complications occurred in 3% of biopsies, including in one patient who died.