Abstract
BACKGROUND: Risk models to predict prostate cancer on biopsy, whether they include only prostate-specific antigen (PSA) or other markers, are intended for use in all men of screening age. However, the association between PSA and cancer probably depends on a man's recent screening history. METHODS: The authors examined the effect of prior screening on the ability to predict the risk of prostate cancer by using a previously reported, 4-kallikrein panel that included total PSA, free PSA, intact PSA, and human kallikrein-related peptidase 2 (hK2). The study cohort comprised 1241 men in Gothenburg, Sweden who underwent biopsy for elevated PSA during their second or later visit for the European Randomized Study of Screening for Prostate Cancer. The predictive accuracy of the 4-kallikrein panel was calculated. RESULTS: Total PSA was not predictive of prostate cancer. The previously published 4-kallikrein model increased predictive accuracy compared with total PSA and age alone (area under the curve [AUC], 0.66 vs 0.51; P < .001) but was poorly calibrated and missed many cancers. A model that was developed with recently screened men provided important improvements in discrimination (AUC, 0.67 vs 0.56; P < .001). Using this model reduced the number of biopsies by 413 per 1000 men with elevated PSA, missed 60 of 216 low-grade cancers (Gleason score < or =6), but missed only 1 of 43 high-grade cancers. CONCLUSIONS: Previous participation in PSA screening dramatically changed the performance of statistical models that were designed to predict biopsy outcome. A 4-kallikrein panel was able to predict prostate cancer in men who had a recent screening history and provided independent confirmation that multiple kallikrein forms contribute important diagnostic information for men with elevated PSA. Cancer