CPC Risk Calculator

コペンハーゲン前立腺がんセンターリスク計算

The Copenhagen Prostate Cancer Center (CPC) Risk Calculator can estimate the individual risk of biochemical recurrence (defined as first ≥PSA 0.2 ng/ml) after radical prostatectomy for localised prostate cancer. The calculator predicts the absolute risk of biochemical recurrence for the following one- or five years in one-year intervals after surgery. The CPC Risk Calculator has been validated in two large institutional cohorts from Denmark and USA.

The dataset is based on 3.746 men who were followed for 20 years after surgery for prostate cancer without receiving any adjuvant treatment at Copenhagen Prostate Cancer Center, Denmark and Stanford Urology, CA, USA. No patients were lost to follow-up. Initial analyses were performed using time between date of surgery and whatever came first biochemical recurrence (outcome of interest) or death without biochemical recurrence (competing risk). Patients without biochemical recurrence were right-censored at latest follow-up. Two multiple cause-specific Cox proportional hazard regression models, one for biochemical recurrence and one for death without recurrence were combined into estimates of the absolute risks of biochemical recurrence. The analyses include preoperative PSA, pT-category, surgical margin status, and radical prostatectomy Gleason score. All analyses were repeated in yearly intervals after surgery using data from patients who survived without biochemical recurrence. The data were internally validated according statistical standards. External validation showed c-index >85% and Brier scores < 5%.

The Copenhagen Prostate Cancer Center is an international research unit affiliated with The Department of Urology at Rigshospitalet, Copenhagen University Hospital, Denmark. Stanford Urology is affiliated with Stanford Medical Center, Stanford, CA, USA.

The statistical model was developed by Prof. Thomas Gerds at The Department of Biostatistics, Faculty of Health and Medical Sciences, University of Copenhagen. The App was developed in collaboration between Andreas Røder ([email protected]) and Daman.dk.