Active surveillance in prostate cancer patients: Modeling upgrading and upsizing at 1 year rebiopsy

90 Background: Prostate biopsy findings at diagnosis and follow-up are essential criteria in Active Surveillance (AS). In a previous work (Nicolai Eur Urol Suppl 2013) upgrading (UPG) and upsizing (UPS) at 1-yr rebiopsy resulted to be independent outcomes with different predictors. Aim of the present work is to validate these models on an independent population and to evaluate the probability of not-developing UPG, thus the chance of continuing AS Methods: Pts enrolled in 2011-2017 were considered for validation, while 318 AS pts enrolled in 2005-2011 were considered for model development. UPG model included: age (risk), PSA density (risk), prostate volume (>60 cc, protective); UPS model: age (protective), % core length containing cancer (>5%, risk), number of positive cores (>1, risk) Performance on the independent population was evaluated through AUC and calibration. Logistic model for not-developing UPG at 1 yr was fitted using all available AS pts Results: 433 pts were included in the validation set. UPG and UPS were registered in 43 and 29 pts, respectively; 38 pts had UPG+UPS. Predictors for UPG and UPS were mainly confirmed in the validation cohort (Table), with Odds Ratios (OR) very similar to the development model. Discrimination was confirmed (AUCs) and calibration was excellent (slope~1, R2>0.90). Probability of not-developing UPG (evaluated on 751 pts) was associated to age, prostate volume and PSA density (Table). Conclusions: UPG and UPS in AS pts should be considered as independent events and their management implying different strategies. Specifically, age, volume and PSA density play a key role in the chance to continue AS, i.e. in not developing UPG. A nomogram estimating the likelihood of continuing AS has been developed and will be presented. [Table: see text]