Purpose: Preoperative prostate specific antigen is widely used to predict unfavorable pathological features and biochemical relapse after radical prostatectomy. Recent reports that hemodilution may be responsible for lower prostate specific antigen in obese men led to concerns that prostate specific antigen may be less effective for prognosticating in men with increased body mass index. We determined whether the clinical usefulness of prostate specific antigen is negatively impacted by obesity by examining its operating characteristics and predictive accuracy as a function of body mass index.
Materials and methods: We performed a multicenter retrospective analysis of the records of 11,705 men who underwent radical prostatectomy from 1988 to 2007 from Veterans Affairs hospitals of the Shared Equal Access Regional Cancer Hospital Database, the Duke Prostate Center and Johns Hopkins Hospital. ROC curve analysis, the concordance index and the test for interaction were used to compare the ability of prostate specific antigen to predict unfavorable tumor characteristics and biochemical recurrence across body mass index categories.
Results: There were no significant differences in the area under ROC curves across increasing body mass index categories for prostate specific antigen to predict pathological Gleason sum (7 or greater, 7 [4 + 3] or greater, or 8 or greater), positive surgical margins, extracapsular extension or seminal vesicle invasion in all 3 cohorts. There was no significant difference in prostate specific antigen accuracy to predict biochemical failure across increasing body mass index categories.
Conclusions: In 3 cohorts of men treated with radical prostatectomy the ability of preoperative prostate specific antigen to predict adverse pathological features and posttreatment biochemical recurrence is not significantly affected by obesity. However, adjusting for obesity related hemodilution may still be required to properly interpret prostate specific antigen results in men with increased body mass index.