Development and validation of a model for predicting the risk of prostate cancer.

Background: Abnormal hematologic parameters before patients undergoing prostate biopsy play a pivotal role in guiding the surgical management of prostate cancer (PCa) incidence. This study aims to establish the first nomogram for predicting PCa risk for better surgical management.

Methods: We retrospectively reviewed and analyzed the data including basic information, preoperative hematologic parameters, and imaging examination of 540 consecutive patients who underwent transrectal ultrasound (TRUS)-guided prostate biopsy for elevated prostate-specific antigen (PSA) in our medical center between 2017 and 2021. Logistic regression analysis was used to determine the risk factors for PCa occurrence, and the nomogram was constructed to predict PCa occurrence. Finally, the data including 121 consecutive patients in 2022 were prospectively collected to further verify the results.

Results: In retrospective analyses, univariate and multivariate logistic analyses identified that three variables including age, diabetes, and De Ritis ratio (aspartate transaminase/alanine transaminase, AST/ALT) were determined to be significantly associated with PCa occurrence. A nomogram was constructed based on these variables for predicting the risk of PCa, and a satisfied predictive accuracy of the model was determined with a C-index of 0.765, supported by a prospective validation group with a C-index of 0.736. The Decision curve analysis showed promising clinical application. In addition, our results also showed that the De Ritis ratio was significantly correlated with the clinical stage of PCa patients, including T, N, and M stages, but insignificantly related to the Gleason score.

Conclusions: The increased De Ritis ratio was significantly associated with the risk and clinical stage of PCa and this nomogram with good discrimination could effectively improve individualized surgical management for patient underdoing prostate biopsy.