Optimal Prostate Cancer Diagnostic Pathways for Men With Prostatomegaly in the MRI Era

Abstract

To evaluate utilities of multiparametric MRI and targeted biopsy to detect clinically significant prostate cancer in men with prostatomegaly. We conducted a retrospective review of multiparametric MRI obtained for elevated PSA between 2017 and 2020. We selected patients with prostates ≥80g who had undergone biopsy. Clinically significant prostate cancer was defined as grade group ≥2. Predictive and logistic regression analyses quantified impacts of diagnostic components. A total of 338 patients met inclusion criteria: 89 (26.3%) had clinically significant prostate cancer. On MRI, positive predictive value for clinically significant prostate cancer was 26.5% for PIRADS 4% and 73.5% for PIRADS 5; negative predictive value for MRI without suspicious lesions was 98.8%. Applying PSA density to MRI yielded a negative predictive value of 78.9% for PIRADS 4 lesions at PSA density <0.05 and a positive predictive value of 90.5% for PIRADS 5 lesions at PSA density ≥0.15. Targeted (versus standard) biopsy reduced likelihood of missing clinically significant prostate cancer by >50% (12.2% vs 28.3%). MRI in-bore biopsies trended towards better accuracy versus MRI-transrectal ultrasound fusion biopsies (75% versus 52%). On logistic regression analyses, MRI improved predictive accuracy (area under the curve 0.91), and PIRADS score demonstrated the strongest association with clinically significant prostate cancer (odds ratio 6.42, P<.001). For large prostates, MRI is less predictive of clinically significant prostate cancer but effectively rules out malignancy. PSA density better informs biopsy decisions for PIRADS 4 and 5 lesions. There may be a pronounced role for targeted biopsy, specifically in-bore, in prostatomegaly.