Clinically Significant ISUP Upgrading in the Multiparametric MRI Era: Biopsy Tumor Burden Outperforms Complex Machine...

Area of Science:

• Urology
• Oncology
• Radiology

Background:

• Multiparametric MRI (mpMRI)-guided biopsy is used for prostate cancer (PCa) diagnosis.
• Clinically significant upgrading (CSU) of ISUP Grade Group (GG) from biopsy to radical prostatectomy (RP) occurs frequently.
• Predictors of CSU require identification to improve preoperative risk stratification.

Purpose of the Study:

• To identify preoperative predictors of CSU (biopsy GG ≤ 2 to RP GG ≥ 3) in PCa patients.
• To compare a parsimonious logistic model against machine learning (ML) classifiers for predicting CSU.
• To evaluate the performance of predictive models using discrimination, calibration, and decision curve analysis.

Main Methods:

• Single-center exploratory analysis of 96 PCa patients undergoing mpMRI, biopsy, and RP.
• Predictive modeling focused on patients with biopsy GG 1-2 (n=64).
• LASSO-guided feature selection and Firth-penalized logistic regression for a reference model, benchmarked against ML classifiers.

Main Results:

• CSU occurred in 15.6% of eligible patients (10/64).
• Positive core ratio was the dominant independent predictor of CSU (adjusted OR 1.54 per 10% increase).
• A two-variable logistic model (positive core ratio, PSA density) outperformed ML classifiers in discrimination and calibration (AUC ≈ 0.75-0.79).

Conclusions:

• Biopsy tumor burden, particularly positive core ratio, is a primary driver of CSU in mpMRI-informed PCa diagnosis.
• A simple logistic model incorporating biopsy tumor burden metrics shows promise for preoperative risk stratification.
• External validation is necessary to confirm the stability and generalizability of these findings.