Patient Selection for Prostate Biopsy: Risk-based or PSA-based?

Abstract

Aim: The AUA Best Practice Statement on PSA advocates a risk-based approach to selecting men for biopsy, based not on a PSA threshold, but on the risk of finding cancer on biopsy. We have therefore studied the association between candidate risk factors and prostate biopsy results. Method: All patients who underwent a prostate biopsy within the Epsom and St Helier NHS Trust between July 2005 and July 2007 were identified. Those with a serum PSA value of <15.0ng/ml were selected for inclusion. Univariate and multivariate logistic regression analysis was done to analyse risk factors for the presence of cancer, and of significant cancer, on biopsy. Significant prostate cancer on biopsy was arbitrarily defined as either a Gleason score of 7 or more, or the presence of cancer in 50% or more of the cores. Results: Of 400 cases, 153 (38%) were found to have prostate cancer on biopsy, of which 93 were classed as significant. On multivariate analysis, smaller prostate volume, abnormal DRE, PSA level and absence of previous negative biopsy were independent predictors of significant prostate cancer. The multivariate logistic regression results were used to create a predictive nomogram for significant cancer. Conclusion: The PSA level is merely one of several factors that predict biopsy results. A risk-based, rather than a PSA-based, approach to selecting men for prostate biopsy has the potential to both reduce the number of men undergoing biopsy, and increase the detection of significant cancers.