Utilization and Outcomes of Serial Prostate MRI for Prostate Cancer Active Surveillance
Andrew Gusev, BA, Jeffrey Twum-Ampofo, MD, Alberto Pieretti, MD, Carl Ceraolo, BA, Florian Rumpf, BA, Alice Yu, MD, Keyan Salari, MD PhD, Douglas Dahl, MD, Adam S. Feldman, MD MPH.
Massachusetts General Hospital, Boston, MA.
BACKGROUND:
Multiparametric Magnetic Resonance Imaging (mpMRI) is increasingly utilized in the diagnosis and risk stratification of men with prostate cancer (PCa). During Active Surveillance (AS) men often receive multiple MRIs, yet it is not well described how results of subsequent MRIs affect their outcomes. We investigated our AS cohort to assess the outcomes of men who underwent at least two prostate mpMRIs during their time on surveillance.
METHODS:
We retrospectively reviewed our institutional database of 1268 men enrolled in AS for localized PCa from 1996-2016. We identified men who underwent at least two prostate mpMRIs during the course of surveillance. MRIs without a reported Prostate Imaging Reporting and Data System (PIRADS) version 2.0 score were excluded. Radiologic progression was defined as new lesion (PIRADS ≥ 3) on a previously negative MRI or increase in PIRADS score of a previous lesion. Lesions with PIRADS score > 3 were defined as high-risk. Cox proportional hazards regression was used to identify factors associated with freedom from treatment. Survival analyses were conducted using Kaplan-Meier method and log rank test.
RESULTS:
The study cohort included 201 men. The majority of patients had GG 1 (97%) and clinical stage T1c (93%) disease. Median values were as follows: number of MRIs meeting criteria: 2 (range 2-3); follow up time: 5.1 years (IQR 3.7-7.7); time between MRIs: 2 years (IQR 1.5-2.8). On MRI 1, 10% of patients had PIRADS 3 lesions, 19% PIRADS 4, 6% PIRADS 5, while 65% were negative. Figure 1 demonstrates the distribution of PIRADS scores from MRI 2, arranged by initial lesion. Overall on repeat MRI, 62% of PIRADS scores did not change, while 27% increased and 11% decreased. Compared to lesions that remained the same, PIRADS increase was predictive of earlier progression to treatment (HR 3.9 [95% CI 2.2-6.9] p<0.001), while PIRADS decrease did not have a significant effect. In 131 men with a negative initial scan, MRI 2 remained negative 63% of the time. This cohort of 83 men with two negative MRIs experienced prolonged freedom from treatment, 88% and 83% at 5 and 10 years, compared to 70% and 47% for those whom had at least one PIRADS ≥ 3 lesion [Figure 2]. Compared to a second negative MRI, a new development of a high-risk lesion, which occurred in 28% of initial negatives, was significantly predictive of subsequent progression to treatment (HR 5.7 [95% CI 2.8-11.6] p<0.001).
CONCLUSIONS:
While a majority of men do not experience PIRADS score change on serial prostate mpMRI, those who have radiologic progression progress to treatment sooner. Men with multiple negative MRIs have significantly prolonged freedom from treatment and may be candidates for an increased interval between surveillance MRI and biopsy. 
Back to 2020 Abstracts