Objectives: To evaluate clinical predictors of disease reclassification or progression (DR/P) in prostate cancer patients elected to active surveillance (AS).
Material and methods: Patients were assessed on the basis of DR/P criteria. Predictors of DR/P were evaluated by multivariate logistic regression and Cox proportional hazards.
Results: The median DR/P free time was 16.5 months. DR/P was detected in 30 out of 84 cases (35.7%). In DR/P cases, the median prostate volume (PV) was significantly lower (34.7 vs. 42.7 ml) and the percentage of cases with 2 or 3 vs. 1 initial biopsy positive cores (BPC) was significantly higher (36.7 vs. 7.4%). The multivariate logistic regression model showed that PV (OR 0.9; p = 0.021) and initial n >1 BPC (OR 9.8; p = 0.001) were independent predictors of DR/P. By Cox multivariate proportional hazards, only n >1 BPC predicted early DR/P (hazard ratio 3.1; p = 0.003).
Conclusions: In a contemporary cohort of patients elected to AS, independent factors stratifying the risk of DR/P were PV and initial BPC, which also predicted early DR/P. In patients elected to AS, the identification of risk factors of DR/P require early re-biopsy. Confirmatory studies are required.
© 2016 S. Karger AG, Basel.