The accuracy of diagnostic biopsy specimens in predicting tumor grades by Gleason's classification of radical prostatectomy specimens.

We reviewed the tissue histology of 115 patients with clinically localized carcinoma to determine the correlation between tumor grades in the biopsy and the prostatectomy specimen. Gleason's primary and secondary pattern score systems were used, and each specimen was graded on a scale of 2 to 10 by a referee pathologist in a blind fashion. If the difference in the summed primary and secondary grades in the 2 specimens was no more than 1 grade, the discrepancy was regarded as insignificant. In all but 32 cases initial diagnostic biopsy specimens predicted accurately the final prostatectomy specimen score. The discrepancy was 3 grades in 7 cases and 2 grades in 25 cases. As a result, the lesions in 19 cases were changed from a well differentiated (2 to 4), a moderately differentiated (5 to 7) or a poorly differentiated (8 to 10) lesion to another of these categories. The lesions were upgraded from a well differentiated to a moderately differentiated category in 9 cases and from a moderately to a poorly differentiated category in 4 cases. The lesions were downgraded from a moderately differentiated to a well differentiated category in 6 cases. In 13 other cases the discrepancy was 2 but the lesion remained within the moderately differentiated group. Although in 19 cases the cancerous tissue occupied less than 10 per cent of the biopsy specimen, accurate prediction could be made in 16. The results indicate that diagnostic biopsy specimens will predict the grade of the primary tumor in a majority (72 per cent) but not all of the cases.