Objectives: To describe the changes in the Gleason grading system over time and evaluate how a shift in Gleason grading has affected the overall predictive accuracy of the system in predicting biochemical disease-free survival after radical prostatectomy.
Methods: The Columbia University Urologic Oncology Database was reviewed, and 1515 patients who met the inclusion criteria were identified who had undergone radical prostatectomy from 1988 to 2004. The patients were divided into two time cohorts (1988 to 1997 and 1998 to 2004). To determine whether a shift in the Gleason sum distribution has occurred, a chi-square test was performed. Survival curves and log-rank tests were used to compare the biochemical disease-free survival between cohorts stratified by the Gleason sum. To estimate the predictive ability of the Gleason system over time, concordance indexes were calculated.
Results: A shift toward greater Gleason sums over time was confirmed using the chi-square test (P <0.001). A significant difference was observed in biochemical disease-free survival between the two time cohorts for those with Gleason sum 6 cancer (P <0.01). The concordance indexes corresponding to Gleason sum alone for each time cohort were 0.71 and 0.87, demonstrating that the Gleason sum's predictive ability improved significantly over time. After adjusting for other variables, the Gleason sum continued to demonstrate a significantly improved predictive ability in the more recent time cohort.
Conclusions: We found a trend toward the assignment of increasing Gleason sums over time in our data set. This shift in Gleason sum distribution between the two time cohorts has resulted in a significant improvement in the predictive ability of the Gleason system.