Development and validation of nomogram to improve the specificity of multiparametric MRI for clinically significant prostate cancer.

Abstract

To develop and validate a nomogram to improve the specificity of prostate imaging reporting and data system (PI-RADS) on multiparametric magnetic resonance imaging (MRI) for clinically significant prostate cancer on targeted fusion biopsy. A retrospective review of patients who underwent fusion biopsy for PI-RADS 3-5 lesions using UroNav and Artemis systems between 2016 and 2022 was performed. Patients were divided into those with CS disease on fusion biopsy (Gleason grade group ≥2) versus those without. Multivariable analysis was used to identify variables associated with CS disease. A 100-point nomogram was constructed, and ROC curve was generated. 1485 lesions (1032 patients) were identified, 510 (34%) were PI-RADS 3, 586 (40%) were PI-RADS 4, and 389 (26%) were PI-RADS 5. Of these, 11% of PI-RADS 3, 39% of PI-RADS 4, and 61% of PI-RADS 5 showed CS disease. CS disease was associated with older age (OR 1.04, 95% CI 1.02-1.06, p < 0.01), previous negative biopsy (OR 0.52, 95% CI 0.36-0.74, p < 0.01), presence of multiple PI-RADS 3-5 lesions (OR 0.61, 95% CI 0.45-0.83, p < 0.01), peripheral zone location (OR 1.88, 95% CI 1.30-2.70, p < 0.01), PSA density (OR 1.48 per 0.1 unit, 95% CI 1.33-1.64, p < 0.01), PI-RADS score 4 (OR 3.28, 95% CI 2.21-4.87, p < 0.01), and PI-RADS score 5 (OR 7.65, 95% CI 4.93-11.85, p < 0.01). Area under ROC curve was 82% for nomogram compared to 75% for PI-RADS score alone. We report a nomogram that combines PI-RADS score with other clinical parameters. The nomogram outperforms PI-RADS score for the detection of CS prostate cancer.