Is physical tumor banking too costly and too inefficient to effectively meet the needs of cancer research?