Gleason Grading of Prostatic Carcinoma

Reproducibility

A major study of interobserver reproducibility of Gleason grading was recently published. In this study, 46 needle biopsies containing prostatic carcinoma were assigned Gleason scores by l0 urologic pathologists. The scores were separated into 4 groups: scores 2-4, 5-6, 7, and 8-10. In 38 (82.6%) of the cases 7 or more of the urologic pathologists agreed on the Gleason score group. These cases were designated "consensus" cases. There was consensus in 12 of 15 cases originally designated Gleason score 5-6, 3 with unanimity, in 9 of 13 cases originally designated Gleason score 7, 1 with unanimity, and in all 16 cases originally designated Gleason score 8-10, 5 with unanimity. Interobserver reproducibility for these cases was substantial for actual Gleason score and moderate to substantial for the Gleason score groups.

The same slides were distributed to 41 general pathologists in the state of Georgia for Gleason grading. There was consistent undergrading of consensus Gleason score groups 5-6 (47%). 7 (47%), and 8-10 (25%). Overall interobserver reproducibility for the "consensus" cases was in the low moderate range.

A questionnaire containing a number of variables possibly associated with good interobserver reproducibility was given to each of the 41 general pathologists. The strongest variable associated with significantly better interobserver reproducibility was having learned Gleason grading at a meeting or course. This finding led to three studies of tutorial efforts to improve Gleason grading, including 2 web-sites (See "Links"), and all 3 have shown significant improvement of Gleason grading following these tutorials. These studies are the rationale for the present expanded website tutorial.