Trends in Prostate MRI Interpretation and Performance Over Time
Alexander Homer, B.A.1, Rebecca M. Ortiz, B.A.2, Borivoj Golijanin, B.S.2, Sai Allu, B.S.1, Natalie Passarelli, B.A.1, Vikas Bhatt, M.D.3, Gyan Pareek, M.D.3, Elias S. Hyams, M.D.3.
1Warren Alpert Medical School of Brown University, Providence, RI, USA, 2Minimally Invasive Urology Institute, The Miriam Hospital, Providence, RI, USA, 3Department of Urology, The Warren Alpert Medical School of Brown University; Minimally Invasive Urology Institute, The Miriam Hospital, Providence, RI, USA.
BACKGROUND: Multiparametric MRI (mpMRI) is a standard of care approach for evaluation of clinically significant prostate cancer (csPCa) in men with elevated PSA or other risk factors. Criticisms have included radiologist learning curve and user-dependent interpretation. We evaluated trends in the interpretation of MRI and diagnostic accuracy of MRI fusion biopsy through a quality initiative. METHODS: Patients with no history of prostate cancer who underwent a 3T multiparametric MRI and TRUS fusion biopsy at a single center from 2017-2020 were retrospectively analyzed. csPCa was defined as Grade Group 2 (GG2) or greater, cancer volume >0.5 mL, or presence of extra-prostatic extension. Temporal trends in the probability of csPCa, GG2+ disease, positive biopsy on >50% of cores, perineural invasion, PIRADS 3, 4, or 5, csPCa among PIRADS 3 and 4/5 patients, as well as the association between PSA density and high grade (HG) disease in PIRADS 4/5 patients were evaluated with logistic regression. Trends in cancer volume were analyzed using ordinary least squares regression. RESULTS: During the study period, 311 subjects met inclusion criteria. The percentage of PIRADS 3, 4, and 5 findings were 18.3%, 29.9%, and 14.5%, respectively. There were increased odds of PIRADS 4/5 (OR 1.53, 95% CI: 1.27-1.84, p<0.001), csPCa (OR 1.40, 95% CI: 1.17 - 1.68, p<0.001), >50% positive cores (OR 1.80, 95% CI: 1.44-2.27, p<0.001), and perineural invasion (OR 1.81, 95% CI: 1.30-2.64, p<0.001) per year. Cancer volume increased by an average of 0.028 mL (SE: 0.0077, p<0.001) per year. csPCa was found in 17.5%, 46.2%, and 60% of PIRADS 3, 4 and 5 exams, respectively. Per year, the odds of finding csPCa in PIRADS 4/5 lesions (OR 1.27, 95% CI: 1.006-1.611, p=0.047) and PIRADS 3 (OR 1.52, 95% CI: 1.26-1.86, p=<0.001) lesions significantly increased. Odds of finding HG cancer in PIRADS 4/5 lesions did not significantly change, but increased in PIRADS 3 lesions (OR 1.42, 95% CI : 1.17-1.74, p<0.001). PSA density >0.15 (OR: 2.31 CI:1.12-4.84, p=0.024) and log PSA density (OR 1.76, CI: 1.085-3.03, p=0.03) were associated with higher PIRADS 4/5 specificity. CONCLUSIONS: Over time, odds of PIRADS 4/5 results and csPCa among PIRADS 3-5 lesions for men undergoing mpMRI for prostate cancer risk have increased, suggesting radiologists are becoming more adept and may reflect trends toward more aggressive cancer. Significant disease for PIRADS 5 was lower than expected. A continued challenge of mpMRI is assessment of negative predictive value, as those with normal imaging are often monitored rather than biopsied. Continued assessment of MRI accuracy and trends is essential to ensure high quality risk assessment.
Back to 2023 Abstracts