Reducing unnecessary biopsies while detecting clinically significant prostate cancer including cribriform growth with the ERSPC Rotterdam risk calculator and 4Kscore

Abstract

Introduction: The use of risk calculators predicting clinically significant prostate cancer (csCaP) on biopsy reduces unnecessary biopsies and overdiagnosis of indolent disease compared to a Prostate Specific Antigen (PSA) strategy. Updating these tools using more specific outcome measures and contemporary predictors could potentially lead to further reductions. Our objective was to assess clinical impact of the 4 kallikrein (4K) score, the Rotterdam Prostate Cancer Risk Calculator (RPCRC), and the combination of both for predicting csCaP based on the latest International Society of Urological Pathology grading system and cribriform growth pattern.

Materials and methods: Our prospective cohort consisted of 2,872 men from the first screening round in the European Randomized Study of Screening for Prostate Cancer Rotterdam; biopsy indication PSA ≥ 3.0. The predictive performance of the 4Kscore, RPCRC, and the combination of RPCRC with 4Kscore were assessed with area under the receiver operator characteristic curve (AUC) and calibration plots. Decision curve analysis was used to evaluate the reduction of unnecessary biopsy and indolent CaP.

Results: The csCaP was present in 242 (8%) men, and indolent CaP in 578 (20%). The 4Kscore and RPCRC had similar high AUCs (0.88 vs. 0.87; P = 0.41). The 4Kscore-RPCRC combination improved AUC to 0.89 compared to 4Kscore (P < 0.01) and RPCRC (P < 0.01). The RPCRC and 4Kscore reduced the number of Bx with 42 and 44, respectively, per 100 men at risk compared to a ≥PSA 3.0 strategy without increasing missed csCaP. The RPCRC-4Kscore combination resulted in a slight additional net reduction of 3.3 biopsies per 100 men.

Conclusions: The RPCRC and 4Kscore had similar reductions of unnecessary biopsies and overdiagnosis of indolent disease. Combination of both models slightly reduced unnecessary biopsies further. Gain in net benefit must, however, be weighed against additional costs and availability of tests.

Keywords: Clinical prediction model; Cribriform growth; European Randomized Study of Screening for Prostate Cancer; International Society of Urological Pathology grading system; Kallikrein panel; Prostate cancer.