[How reliable is cancer diagnosis with needle biopsy of the prostate?].

Aim of the study was to investigate the reproducibility and the reliability of the morphological diagnosis, of different morphological parameters and representativity of needle biopsy material compared with transurethral resection material. 416 needle biopsies of carcinomas of the prostate were studied by 5 pathologists. 130 biopsies were evaluated twice by the same investigators and 325 at least by one other investigator. In 65 cases, tissue obtained by transurethral resection was evaluated by the same investigator who had seen the biopsy. The diagnosis of carcinoma of the prostate was highly reproducible, only in 2% of the cases the original diagnosis was not confirmed; 5 out of 27 systematically evaluated parameters proved to be very reliable and reproducible, i.e. histological tumor classification, cribriform glands, invasion of perineural spaces, nucleoles, nuclear polymorphism, 8 parameters were sufficiently reproducible and 14 were unreliable among which cytologic differentiation was one. The diagnostic reproducibility between biopsy material and transurethral resection material was only 48.8%. This very low figure can be explained by the marked histologic heterogeneity of the tumors of which only 13% were classified as uniform carcinomas. The conclusions of the investigation are: 1. The diagnosis carcinoma is highly reliable and reproducible in needle biopsies. 2. Needle biopsy tissue is not representative for the entire tumor. 3. Grading of prostate carcinoma is highly subjective and therefore no sound basis for therapy and prognosis in the individual patient.