Development of prediction models based on risk scores for clinically significant prostate cancer on MRI/TRUS fusion biopsy.

The implementation of population screening for prostate cancer has increased the number of patients with biochemical suspicion. Prediction models may reduce the number of unnecessary biopsies by identifying patients who benefit the most from them. Our aim is to develop a prediction model that is easily applicable in patients with suspicion of prostate cancer in the urology clinic setting to avoid unnecessary biopsies.

We developed prediction models based on risk scores for the detection of prostate cancer and clinically significant prostate cancer using the TRIPOD guidelines. For this, we conducted an observational and retrospective review of computerised medical records of 204 patients undergoing prostate fusion biopsy between 2018 and 2021. We also reviewed other prediction models for prostate cancer including radiological parameters and targeted sampling of suspicious lesions.

A total of 204 patients underwent a biopsy, 138 were diagnosed of prostate cancer, and from them, 60 of clinically significant prostate cancer. Multivariate regression and random forest analysis were performed. Age, PSA density, diameter of the index lesions and PIRADS score on MRI were identified as predictors with an Area Under the Curve ranging between 0.71 and 0.80 and acceptable calibration results. Risk scores may avoid between 21.7% and 48.1% of biopsies.

Our prediction models are characterised by ease of use and may reduce unnecessary biopsies with satisfactory discrimination and calibration results while bringing benefits to the healthcare system and patients.

Urologic oncology. 2024 Sep 02 [Epub ahead of print]

Hector Ayerra Perez, Javier Fermin Barba Abad, Julene Argaluza Escudero, Javier Extramiana Cameno, Egoitz Tolosa Eizaguirre

Department of Urology, Araba University Hospital, OSI Araba Osakidetza, Vitoria-Gasteiz, Spain; Urologic Cancer Group, Bioaraba Health Research Institute, Vitoria-Gasteiz, Spain. Electronic address: ., Department of Urology, Araba University Hospital, OSI Araba Osakidetza, Vitoria-Gasteiz, Spain., Epidemiology and Public Health Group, Bioaraba Health Research Institute, Vitoria-Gasteiz, Spain., Department of Urology, Araba University Hospital, OSI Araba Osakidetza, Vitoria-Gasteiz, Spain; Urologic Cancer Group, Bioaraba Health Research Institute, Vitoria-Gasteiz, Spain.