Integration of Multiparameter MRI into Conventional Pretreatment Risk Factors to Predict Adverse Features after Radical Prostatectomy.

Abstract

The purpose of this study was to Integration of Multiparameter MRI into conventional pretreatment risk factors to develop nomogram for the prediction of adverse features (i.e., positive margin (PM) or extra-prostatic extension (EPE)) indicated for postoperative radiotherapy in patients with prostate cancer. We analyzed 203 histologically proven prostate cancer patients underwent preoperative 3T multiparametric MRI including high b value (0, 1500 s/mm2) diffusion-weighted imaging between 2019 and 2022 at our hospital. Patients with a history of neoadjuvant hormonal therapy, or patients whose surgeon's experience was ≤3 years were excluded. Age, surgical technique, serum PSA level, PSA density, clinical T stage, and biopsy Gleason score were clinical predictors. MRI parameters include prostate volume, tumor size, ECE score, seminal vesicle invasion score, tumor location (apex, peripheral region, or bladder neck), apparent diffusion coefficient (ADC), tumor contact length (TCL), PI-RADS score. Predictors were used in nomograms developed from multivariable logistic regression analysis to estimate the probability of positive margin (PM) or extra-prostatic extension (EPE) after RP. The nomogram's predictive accuracy and discriminative ability were determined by the concordance index with calibration and receiver operating characteristic (ROC) curves. (1) Patient characters: Table 1 lists the MRI characteristics of the patients. Median PSA level is 17.6ng/ml. All patients received extra-fascial resection. 44% of the patients underwent robotic surgery. 24% of the patients had positive margins. In 226 patients who was clinically confined to the prostate, 100 (44%) had a postoperative pathological upgrade of pT3a or above. 46 (29%) of the 158 patients with biopsy grade group (GG)1 was confirmed GG3-5 after RP. Using mpMRI parameters and clinicopathological information, we produced nomograms that may accurately predict adverse Features that are indications for postoperative radiotherapy after RP, which may help individualize treatment decision-making.