Nomogram for predicting Gleason grouping upgrading(GGU)in a cohort receiving radical prostatectomy based on 2014 ISUP grouping system: development and internal validation

Abstract

Objective: To analyze the predictive factors of GGU between biopsy and radical prostatectomy pathology based on 2014 ISUP grouping system, then establish and evaluate nomogram. Methods: Patients undergoing radical prostatectomy in Shanghai Ruijin Hospital from March 2012 to March 2019 were reviewed, and the clinical and pathological information were collected. Age(68.1±7.2), body mass indes(BMI) (24.2±3.2)kg/m2, prostate specific antigen(PSA) 11.5(6.7-20.4)ng/ml, prostate specific antigen destiny(PSAD) 0.35(0.20-0.66). Before March 2017, the number of biopsy cores were 6 to 8; After then, all patients toke 12 cores systemic biopsy. Based on 2014 ISUP grouping system, the differences between biopsy and radical prostatectomy grades were counted. The independent predictors of GGU were analyzed by univariate and multivariate logistic regression analysis, then the nomogram for predicting GGU were established and evaluated. Results: 429 patients were enrolled. There were 161 (37.5%) patients in GGU group and 268 (62.5%) patients in non-GGU group. After multivariate logistic regression analysis, body mass index (BMI)>28 kg/m2(OR=2.54, P=0.021), prostate specific antigen density (PSAD)(OR=1.65, P=0.018)and 2014 ISUP grouping sysyem (OR=0.53, P<0.001) of biopsy specimen were independent impact factors of GGU. The predicting model was established according to BMI, PSAD and 2014 ISUP grouping system. The area under the ROC cure of the model was 0.735 (95%CI 0.681-0.789). The nomogram model was well calibrated, with the mean absolute error of 6.7%, which means the prediction of GGU is fairly consistent with the actual situation. Conclusions: Based on the 2014 ISUP grouping system, BMI>28 kg/m2, PSAD and 2014 ISUP grouping of biopsy specimen were independent predictors of GGU. The nomogram model for predicting GGU has a good statistical significance.