Introduction

To reduce overdiagnosis and overtreatment on Prostate Cancer (PCa), a biopsy (Bx) should only be offered to men with an increased risk of having a potentially life-threatening PCa. The Digital Rectal Examination (DRE) version of the Rotterdam PCa Risk Calculator (RPCRC), was developed to include information on prostate volume but to circumvent the need for imaging studies, enabling easier implementation into daily practice of urologists and general practitioners (GPs). Our objective was to assess the level of agreement between DRE findings (irregularities and estimation of volume) of two urologists in men suspicious for PCa and, subsequently, the potential effect on risk prediction using the DRE-based RPCRC.

Methods

A prospective cohort of asymptomatic and unscreened men with PSA <=50.0 ng/mL and TRUS (transrectal ultrasonography) volume <=110 mL who underwent 16-core TRUS-guided Bx were evaluated. _x000D_ Both urologists’ DRE findings were graded normal or abnormal (i.e. nodularity and/or induration), and volume classified as 25mL, 40mL or 60mL, according to the RPCRC algorithm. Inter-rater agreement analysis using Cohens’s kappa (?) statistic was performed to determine consistency of DRE outcome and volume assessment. Receiver operating characteristic (ROC) curve analysis and calibration plots were constructed per urologist to determine the effect of inter-rater differences. Decision curve analysis (DCA) was applied to evaluate the clinical usefulness of the DRE based model.

Results

Of the 241 men included in the study, 41% (n = 98) had a positive Bx (81 PCa were clinically significant). There was substantial agreement in the DRE examination (abnormal/normal) (? = 0.78; P < 0.001), and volume estimation (? = 0.79; P < 0.001). _x000D_ ROC analyses showed good discrimination (0.75–0.78) and were highly comparable for both urologists. In our high risk cohort, at a probability threshold of 25%, the DRE-based RPCRC reduces the Bx rate by 9%, without missing cancers.

Conclusions

This is the first study to validate the DRE-version of the RPCRC. Most crucial in this validation is the effect of the inter observer variability of a subjective predictor like DRE and DRE assessed volume. The slight differences in DRE findings between the two urologists had very little impact on the performance of the RPCRC. The DRE-based RPCRC can be considered a useful Bx outcome prediction tool.

Funding

None.