Abstract
OBJECTIVES: To determine the performance of MRI for detecting clinically significant prostate cancer (csPCa) recurrence following focal cryo-ablation over five years of oncological surveillance. SUBJECTS AND METHODS: A total of 305 men with intermediate-risk prostate cancer undergoing focal cryo-ablation (FCA) at our referral centre between 03/2017 and 03/2024 were prospectively enrolled in an outcomes registry. Selection criteria, treatment planning and oncological surveillance were standardized. The initial surveillance protocol included MRI at six months, two, 3.5, five years, and surveillance prostate biopsy at six months, two and five years. The surveillance protocol biopsy at six months and two years was abandoned for a negative MRI in May 2019 and November 2020. Performance statistics of MRI for detecting csPCa recurrence were assessed among mandatory MRI and biopsy dyads at six months and two years. Clinically significant prostate cancer detection rates for suspicious (positive) and negative MRI included all subjects. RESULTS: Sensitivity, specificity, positive predictive value and negative predictive value of MRI to predict csPCa is 25%, 93%, 29% and 91%. The csPCa detection rate for positive and negative MRI was 40% and 7.0%, respectively. The area under the receiver operating characteristic curve for MRI as a predictor of csPCa was 0.60, suggesting limitations of MRI for predicting csPCa. A limitation is that a validated instrument for positive MRI was not available during the study. CONCLUSIONS: Six-month MRI is not recommended owing to its very low csPCa detection rate. The increasing csPCa detection rates at two and five years suggest these are reasonable time points to perform MRI. Prostate biopsy should be performed on all cases with positive MRI, and can safely be deferred in most cases with negative MRI.