Presentation Authors: Cosimo De Nunzio*, Giuseppe Simone, Riccardo Lombardo, rome, Italy, Marcello Scarcia, Torino, Italy, Mariaconsiglia Ferrero, Rocco Papalia, Rome, Italy, Giuseppe Mario Ludovico, Torino, Italy, Angela Sica, Rome, Italy, Alessandro Giacobbe, Marco Ordera, Torino, Italy, Giovanni Muto, Turin, Italy, Michele Gallucci, Andrea Tubaro, Ospedale F. Miulli, Rome, Italy
Introduction: Recent studies have evaluated the possible role of MRI in first set biopsy, however only few nomograms include fusion biopsies to predict prostate cancer. The aim of our study is to analyze the role of fusion biopsy in improving prostate cancer and high-grade prostate cancer detection developing a clinical nomogram.
Methods: From 2016 onwards, we consecutively enrolled, in five centers in Italy, men undergoing first 12 core trans-rectal ultrasound-guided prostate needle biopsy (SB) and first fusion biopsy (FB). Indications for prostate biopsy were a positive digital rectal examination (DRE), PSAâ‰¥4ng/ml and/or positive MRI. Demographic, clinical and histopathological data were collected. Logistic regression analysis was used to evaluate predictors of cancer and high-grade cancer. Based on the multivariable logistic regression a nomogram to predict cancer and a nomogram to predict high grade (Gleason >3+4) were generated. LROC, calibration plots and decision curve analysis were used to assess discrimination, calibration and net benefit of the model respectively.
Results: Overall 2123 patients were enrolled patients with a median age of 68(61/73) years, median PSA of 6.6(4.7/9.9) ng/ml and a median prostate volume of 48(36/68) cc. Overall 937/2123 (44%) presented prostate cancer (SB: 669/1653: 40% vs FB:268/470:57%) and out of them 392/937 (42%)(SB: 273/669: 41% and FB: 119/268: 44%) presented high grade prostate cancer. Overall 470/2123 (22%) performed a FB, no differences in terms of age, PSA and PV were found when compared to SB. The accuracy of the cancer model was 0.75, calibration was excellent, and a net benefit of adding a fusion biopsy on the base model was recorded in the range of probabilities between 30-80%. The accuracy of the high-grade cancer model was 0.78, calibration was excellent, and no-net benefit was observed when adding a fusion biopsy to the high-grade model.
Conclusions: The present nomograms have excellent properties in the prediction of prostate cancer and high-grade prostate cancer. Fusion biopsy increased the model only for prostate cancer detection overall and particularly in the 30-80% probability. Our nomograms if validated could be used to identify the best patients candidate for a fusion biopsy.