%0 Journal Article
%A Püllen, Lukas
%A Radtke, Jan P
%A Wiesenfarth, Manuel
%A Roobol, Monique J
%A Verbeek, Jan F M
%A Wetter, Axel
%A Guberina, Nika
%A Pandey, Abhishek
%A Hüttenbrink, Clemens
%A Tschirdewahn, Stephan
%A Pahernik, Sascha
%A Hadaschik, Boris A
%A Distler, Florian A
%T External validation of novel MRI-based models for prostate cancer prediction.
%J BJU international
%V 125
%N 3
%@ 1464-4096
%C Oxford
%I Wiley-Blackwell39962
%M DKFZ-2019-02710
%P 407-416
%D 2020
%Z 2020 Mar;125(3):407-416
%X To validate three novel risk models (RM) combining mpMRI and clinical parameters to predict significant prostate cancer (sPC) through an external cohort, including the recently updated European Randomised Study of Screening for PC (ERSPC) risk-calculator.We retrospectively analyzed 307 men who underwent mpMRI prior to transperineal ultrasound fusion biopsy between 10/2015 and 07/2018 at two German centers. mpMRI was rated by PI-RADSv2.0 and sPC was defined as ISUP Gleason grade group ≥2.The prediction performance of the three models (MRI-ERSPC-3/4, RM by Radtke et al. and RM by Distler et al.) were compared using ROC-curve-analyses with area under the curve (AUC), calibration curve analyses and decision curves to assess net-benefit.ROC-AUCs of the three novel models (MRI-ERSPC-3/4; Radtke's RM and Distler's RM were 0.82; 0.85 and 0.83, respectively). Calibration curve analyses showed the best intercept for MRI-ERSPC-3 and -4 of 0.35 and 0.76. Net benefit analyses indicated clear benefit of MRI-ERSPC-3/4-RM compared to the other two validated models. The MRI-ERSPC-RC-3/4 demonstrated a discrimination benefit for a risk threshold of up to 15
%F PUB:(DE-HGF)16
%9 Journal Article
%$ pmid:31758738
%R 10.1111/bju.14958
%U https://inrepo02.dkfz.de/record/147733