Qualitative and Quantitative Reporting of a Unique Biparametric MRI: Towards Biparametric MRI-Based Nomograms for Prediction of Prostate Biopsy Outcome in Men With a Clinical Suspicion of Prostate Cancer (IMPROD and MULTI-IMPROD Trials)

Abstract

Background: Multiparametric MRI of the prostate has been shown to improve the risk stratification of men with an elevated prostate-specific antigen (PSA). However, long acquisition time, high cost, and inter-center/reader variability of a routine prostate multiparametric MRI limit its wider adoption.

Purpose: To develop and validate nomograms based on unique rapid biparametric MRI (bpMRI) qualitative and quantitative derived variables for prediction of clinically significant cancer (SPCa).

Study type: Retrospective analyses of single (IMPROD, NCT01864135) and multiinstitution trials (MULTI-IMPROD, NCT02241122).

Population: 161 and 338 prospectively enrolled men who completed the IMPROD and MULTI-IMPROD trials, respectively.

Field strength/sequence: IMPROD bpMRI: 3T/1.5T, T₂ -weighted imaging, three separate diffusion-weighted imaging (DWI) acquisitions: 1) b-values 0, 100, 200, 300, 500 s/mm² ; 2) b values 0, 1500 s/mm² ; 3) values 0, 2000 s/mm² .

Assessment: The primary endpoint of the combined trial analysis was the diagnostic accuracy of the combination of IMPROD bpMRI and clinical variables for detection of SPCa.

Statistical tests: Logistic regression models were developed using IMPROD trial data and validated using MULTI-IMPROD trial data. The model's performance was expressed as the area under the curve (AUC) values for the detection of SPCa, defined as ISUP Gleason Grade Group ≥2.

Results: A model incorporating clinical variables had an AUC (95% confidence interval) of 0.83 (0.77-0.89) and 0.80 (0.75-0.85) in the development and validation cohorts, respectively. The corresponding values for a model using IMPROD bpMRI findings were 0.93 (0.89-0.97), and 0.88 (0.84-0.92), respectively. Further addition of the quantitative DWI-based score did not improve AUC values (P < 0.05).

Data conclusion: A prediction model using qualitative IMPROD bpMRI findings demonstrated high accuracy for predicting SPCa in men with an elevated PSA. Online risk calculator: http://petiv.utu.fi/multiimprod/ Level of Evidence: 1 Technical Efficacy Stage: 2 J. Magn. Reson. Imaging 2020;51:1556-1567.

Keywords: PSA; biparametric MRI; diffusion weighted imaging; multi-institutional trial; prostate cancer; prostate cancer screening.