Nomograms for predicting the risk of biochemical recurrence in patients with prostate cancer after surgery

Abstract

Background. Prostate cancer (PCa) patients often develop recurrent disease after radical surgery. A tool that can accurately predict the risk of disease progression in the population of Russian patients will be very helpful to choose an optimal treatment strategy and prevent possible recurrence.Objective: to analyze preoperative and postoperative prognostic factors for PCa progression and identify the most significant of them.Materials and methods. This study included 2,255 patients with localized and locally advanced PCa who underwent radical surgery. We constructed nomograms for predicting the risk of disease progression after surgery using mathematical models.Results. We created nomograms for predicting the risk of biochemical recurrence and probability of relapse-free survival by the level of prostate specific antigen (PSA) in patients with no lymph node metastases (pN0) according to the results of morphological examination and in patients with lymph node metastases (pN1). The accuracy of nomograms reached 71 % (area under the ROC curve (AUC) 0.7119) and 76 % (AUC 0.7617), respectively.Conclusion. The nomograms demonstrated high accuracy of prognosis and can be used in the population of Russian patients.