Magnetic resonance imaging for long-term active surveillance biopsy decision-making.

Abstract

To evaluate whether long-term magnetic resonance imaging (MRI) findings, alone or in combination with prostate-specific antigen (PSA) density or the result of prior surveillance biopsies, could be used to omit biopsies occurring after patients have been on active surveillance (AS) for ≥5 years. Analysis of a single-institution AS cohort of patients with Grade Group (GG) 1 prostate cancer who underwent a scheduled 6-year MRI and biopsy. Multivariable logistic regression was used to test the association between the Prostate Imaging-Reporting and Data System (PI-RADS) score, PSA density, status of prior AS biopsy, and reclassification to GG ≥2. We then analysed how many GG ≥2 cancers would be missed if the 6-year biopsy was avoided based on significant predictors. In all, 49 of 221 men (22%) were reclassified to GG ≥2. PSA density and PI-RADS scores 4-5 were significant predictors; a prior negative AS biopsy was inversely associated with GG ≥2. The risk of missing GG ≥2 if the 6-year biopsy is omitted in men with an unsuspicious MRI (PI-RADS 1-2) was 12% (95% confidence interval [CI] 5.3-19%). After testing combinations of MRI findings, PSA density, and the status of the prior biopsy, we found the lowest risk of GG ≥2 in patients with a prior negative biopsy and PSA density of <0.10 ng/mL/cc (5.3%, 95% CI 1.5-10%). An unsuspicious MRI at 6 years is not enough alone to omit the 6-year AS biopsy under current guidelines. Combining MRI findings, PSA density, and prior negative AS biopsies is a promising strategy to tailor surveillance intensity. Our findings should be replicated in larger cohorts prior to implementation in clinical practice.