Clinical utility of percutaneous biopsies of musculoskeletal tumors

Percutaneous biopsies are frequently used for musculoskeletal lesions. We suspect published accuracy rates (80-97%) overestimate clinical utility. We retrospectively reviewed 120 consecutive percutaneous biopsies performed by interventional radiologists at our institution. Patients underwent core biopsy, fine-needle aspiration (FNA), or both. The biopsy interpretations were considered clinically useful if they allowed proper treatment to proceed and not useful if they were nondiagnostic or if surgical specimens disagreed with percutaneous specimen. Patients were categorized by biopsy type, tissue type, and tumor type. Ninety of 120 percutaneous biopsies were clinically useful, 27 were nondiagnostic, and three were incorrect; in these latter 30 patients we proceeded to open biopsy. Patients with both biopsies had clinically useful results (80.6%) more often than FNA or core alone(68.0% and 66.7%, respectively). Biopsies of bone lesions were clinically useful more often than those of soft tissue. Myxoid histology was associated with decreased clinical accuracy. Clinical utility was independent of tumor type. No single characteristic predicted increased probability of open biopsy. The clinical utility rate was acceptable, but below published accuracy rates. The combination of both biopsies was better than FNA alone. Myxoid findings rarely helped to guide definitive treatment. Treatment decision making requires balancing biopsy results with clinical data.