Timo Soeterik

106 CHAPTER 6 METHODS Study population and setting The validation cohort was derived from the population of 625 prostate cancer patients that underwent RARP in 2014 to 2017 at the Canisius Wilhelmina Hospital in Nijmegen, The Netherlands. Data from the prospectively collected local RARP database including preoperative characteristics such as age, initial serum PSA level, highest ISUP grade on biopsy, and postoperative characteristics such as pathologic tumour stage, ISUP grade, and margin status were used. Additional variables specifically needed for the validation (e.g. side of EPE on MRI, side of EPE during pathological evaluation) were collected retrospectively. Radiological reporting was done according to the European Society of Uro-Radiology (ESUR) guidelines and from 2015 PI-RADS v2. 11,12 Presence of EPE on MRI was defined as loss of prostate capsule and irregularity of the capsule. Presence of reported “uncertain EPE”; e.g. presence of solely capsule abutment or broad capsular contact, were regarded as EPE-negative. These definitions were consistent with those used for model development . 10 The cohort consisted of patients diagnosed locally and referred patients (tertiary centre and other regional [teaching] hospitals), reflecting a real-world population. Patients with complete cases were included for analysis. Derivation of the study sample from the overall population was further clarified using a patient inclusion flow chart (Figure 1). Assessment of present EPE All radical prostatectomy specimens were formaldehyde fixed and completely imbedded in paraffin. All specimens were standardized inked in 3 ink colors for optimal orientation and cut in a standardized way. From the apical region and bladder neck region a 5 mm thick section was taken and sectioned perpendicular to the inked surface and totally embedded. EPE was defined as tumour that bulges beyond the prostatic contour, tumour that is admixed with periprostatic fat tissue or in the posterolateral area, as tumour within connective tissue or between nerve bundles of the neurovascular bundle. In the bladder neck region, the presence of tumour between thick smooth muscle bundles in the absence of benign prostatic glands was considered as bladder neck invasion. 13 In case information on EPE was missing, slides were revised by a single uropathologist (H.K.V.).