190 IMPROVED CONCORDANCE OF BIOPSY AND PATHOLOGICAL GLEASON SCORE USING QUANTITATIVE GLEASON GRADING OF PROSTATE BIOPSY SPECIMENS

Abstract

INTRODUCTION AND OBJECTIVES: In men with prostate cancer, biopsy Gleason grade is typically assigned based on the most aggressive Gleason pattern of any single biopsy core, not the most prevalent Gleason pattern of all positive biopsies. This practice may partially explain the poor concordance between biopsy and pathologic Gleason grade in men treated with radical prostatectomy (RP). We hypothesized that quantitative Gleason grading, based on the percent of high-grade tumor aggregated across all positive biopsy cores, is a better predictor of true pathologic Gleason grade than traditional biopsy grading practices. METHODS: We performed an analysis of our institutional prostate cancer database, including all men with biopsy Gleason grade of 7 treated with RP since 2005, when our pathologists began routinely including percent of high-grade cancer on biopsy reports. Quantitative Gleason grade was determined by calculating the percent of highgrade tumor (Gleason 4 or 5) from all positive biopsy cores. The t-test, ROC curve analysis, and decision curve analysis were used to compare the efficacy of quantitative vs. traditional grading of biopsy specimens in predicting pathologic Gleason score. RESULTS: 225 patients met the study inclusion criteria. Using traditional Gleason grading, 188 men (84%) were assigned biopsy Gleason grade 3 4 and 37 (16%) were grade 4 3. The distribution of quantitative Gleason grade was as follows: 125 men (55% of cohort) 0–25% high-grade tumor on biopsy, 61 men (27%) 26–50% high-grade tumor, 22 men (10%) 51–75% high-grade tumor, and 17 men (8%) 76-100% high-grade tumor. The mean percent of high-grade biopsy tissue was greater in those men with pathological 4 3 compared to 3 4 tumors (41.2% vs. 17.5%, p 0.01). ROC and decision curve analyses of quantitative vs. traditional Gleason grading in predicting pathological Gleason grade are shown in the figure. Quantitative grading resulted in a greater AUC on ROC analysis (0.77 vs. 0.69) and a greater net-benefit on decision curve analysis. CONCLUSIONS: Quantitative Gleason grading is better associated with pathological Gleason grade than traditional biopsy grading practices. Quantitative Gleason grading may reduce rates of both upand down-grading and result in more precise risk stratification for men undergoing RP for prostate cancer.