Over-diagnosis and under-diagnosis of screen- vs non-screen-detected prostate cancers with in men with prostate-specific antigen levels of 2.0-10.0 ng/mL.

Abstract

To evaluate possible over- and under-diagnosis of prostate cancer in a screened vs a referral population in the same range of prostate-specific antigen (PSA). In all, 1445 patients undergoing radical prostatectomy and with a PSA level of <10 ng/mL were evaluated; 237 were from outside Tyrol (Austria) and represented the unscreened group, and 1208 were Tyrolean screening volunteers. Over-diagnosis was defined as a pathological stage of pT2a and a Gleason score of <7 with no positive surgical margins. Under-diagnosis was defined as a pathological stage of >or=pT3a or positive surgical margins. The chi-square test was used to assess the differences, with P < 0.05 considered to indicate statistical significance. There were no significant differences in patient age or PSA levels between the study groups. There was over-diagnosis in the screening and referral groups in 17.4% and 8.9%, respectively, and under-diagnosis in 18.6% and 42.2%, respectively. This study suggests that patients with prostate cancer participating in a screening programme are less likely to be under-diagnosed or have extracapsular disease than their counterparts in a referral population, even in the same PSA range, after radical prostatectomy. Furthermore, there was more under-diagnosis in the referral group than over-diagnosis in the screened group.