Timo Soeterik

63 Impact of MRI on Risk Classification BACKGROUND In 1998, D’Amico and colleagues proposed the D’Amico classification, stratifying patients with prostate cancer into low-, intermediate-, or high-risk disease to predict disease progression following radical treatment based on pretreatment clinical T-stage, Gleason score and serum prostate-specific antigen (PSA) level. 1 The classification forms the basis for both the National Comprehensive Cancer Network and the European Association of Urology staging systems and is therefore commonly used by urologists to assist patient counselling. 2,3 Multiparametric magnetic resonance imaging (mpMRI) of the prostate is one of the most influential emerging diagnostic modalities within the field of prostate cancer of the last 2 decades, establishing an important role in the detection of clinically significant disease 4,5 and local tumour staging. 6,7 With regard to local staging of prostate cancer, mpMRI potentially enables more accurate estimation of the local tumour size and extent compared with digital rectal examination (DRE), 8 and thus may improve risk stratification and treatment selection. However, it is recommended that imaging alone should not replace the DRE for defining clinical tumour stage due to lack of interobserver reproducibility, issues with patient selection and contradictory results. 9,10 Previous studies have shown that use of mpMRI information for risk group classification leads to substantial upstaging of patients to a higher risk group. 11,12 In addition, these studies have demonstrated that mpMRI use may lead to alterations in treatment advice. However, these studies regarded both relatively small, single-centre populations, including solely patients that underwent radical prostatectomy. Therefore, results may not be generalisable to the overall population of newly diagnosed prostate cancer patients. In addition, the possible stage migration due to mpMRI could impact the prognostic characteristics of the original D’Amico classification system, due tomigration of patients with more favorable characteristics to higher risk groups. For instance, it remains unclear whether mpMRI stage ≥ T3 is prognostically equivalent to clinical stage ≥ T3 assessed by DRE. 12 In this study, we aim to establish the impact of mpMRI use for local staging on prostate cancer risk stratification and treatment assignment in a large real-world multi- centre cohort of newly diagnosed prostate cancer patients. In addition, we will evaluate the accuracy of both DRE and mpMRI for the detection of non-organ-confined disease in a subset of patients undergoing robot-assisted radical prostatectomy (RARP). 4