Impact of MRI on outcomes in active surveillance (AS) for localized prostate cancer in a hospital registry.

Abstract

280 Background: MRI is increasingly used in follow-up of men undergoing AS for localized prostate cancer. Large cohorts predating the MRI era have shown that the risk of disease progression or stopping an AS protocol is ~20-30% at 5 years, but the impact of MRI on outcomes in AS is unclear. Methods: We studied men who initiated AS at Dana-Farber Cancer Institute between 2001-2017 for Gleason 6 prostate cancer diagnosed on transrectal ultrasound-guided biopsy, and who underwent a prostate MRI as part of their AS protocol. Progression on AS was defined as finding Gleason ≥7 cancer on repeat biopsy or at radical prostatectomy. Early MRI was defined as one performed within 1 year of diagnosis. Results: A total of 148 men were identified from a prospective IRB-approved database. Median PSA at diagnosis was 4.8 (range 0.7-14.3), median number of positive biopsy cores was 1 (1-7), median maximal core involvement by cancer was 10% (1-65), and most men (87%) had T1c disease. Overall, 54 (36%) progressed on AS over a median follow-up of 5.6 years (95% CI 5.1-6.0), with the majority (n=44, 81%) progressing at repeat biopsy; the 5-year progression rate was 32% (25-41). There were no major differences in baseline clinicopathologic characteristics between men undergoing early (n=103, 70%) or delayed (n=45, 30%) MRI, but men who had an early MRI underwent fewer subsequent biopsies (median 2 vs. 3, p=0.010) and more targeted biopsies (47% vs. 36, p=0.212). Men who underwent early MRI had a higher rate of progression compared to those who had a delayed MRI (5-year rate 43% [33-55] vs. 12% [5-26], log-rank p=0.001). However, when immortal time bias was accounted for by considering MRI receipt as a time-dependent covariate, early MRI was not associated with progression (HR=0.74 [0.50-1.09], p=0.129). Conclusions: The 5-year progression rate on AS was ~30% in the MRI era, a rate similar to that reported in series that predate the adoption of MRI. A higher rate of progression was seen in men undergoing MRI within 1 year of diagnosis, but MRI timing did not influence risk of progression when accounting for the time interval before an MRI was performed. Longer follow-up is needed to determine the impact of MRI on cancer-specific survival in men undergoing AS.