PD48-02 DECISION-ANALYTIC MODELING STUDY OF THE PRECISION TRIAL: DOES PRE-BIOPSY MRI DO MORE GOOD THAN HARM?

Abstract

You have accessJournal of UrologyProstate Cancer: Detection & Screening V (PD48)1 Apr 2020PD48-02 DECISION-ANALYTIC MODELING STUDY OF THE PRECISION TRIAL: DOES PRE-BIOPSY MRI DO MORE GOOD THAN HARM? Andrew Vickers* and Sigrid Carlsson Andrew Vickers*Andrew Vickers* More articles by this author and Sigrid CarlssonSigrid Carlsson More articles by this author View All Author Informationhttps://doi.org/10.1097/JU.0000000000000942.02AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Several recent studies, including the randomized PRECISION trial, have investigated whether men with elevated prostate-specific antigen (PSA) levels should undergo magnetic resonance imaging (MRI) before prostate biopsy, with targeted biopsy performed if MRI is positive and no biopsy for MRI negative patients. The finding that MRI reduces the number of low-grade and increases the number of high-grade cancers has been interpreted as supporting MRI. However, a technique that overgraded tumors would report similar findings. Study findings at the group level need to be complemented by modeling studies of individual patient counterfactuals. For instance, a patient who would have been grade Grade Group (GG) 1 by regular transrectal ultrasound (TRUS) biopsy but GG2 by MRI would only benefit if confirmatory biopsy after TRUS would show GG2 (hence MRI would avoid an additional biopsy) or if delayed treatment affected chance of cure. If neither are true, the patient would be harmed by curative treatment resulting from the increased grade. METHODS: We used the PRECISION trial results to create different scenarios for individual patient counterfactuals, varying the proportion of MRI-detected high-grade cancers that were reclassified from low-grade vs. benign TRUS biopsy. We assumed that only patients diagnosed with GG2+ would be treated curatively with active surveillance for GG1. We considered treatment to constitute high harm, active surveillance to be moderate harm and additional biopsy to constitute minor harm. As there is evidence that long-term rates of prostate cancer death are very low for both patients with negative TRUS biopsy and those with GG1 followed on active surveillance, delayed treatment was considered to be a moderate harm. We tested a range of scenarios for degree of harm. RESULTS: In most scenarios, MRI was associated with very low levels of net benefit that were highly dependent on assuming that the higher rates of GG2+ disease on MRI resulted from reclassification from benign disease on TRUS. In the base case, the highest net benefit for MRI across scenarios was the equivalent of 50 MRIs to prevent one overdiagnosis of GG1. MRI had was harmful in many scenarios. MRI had important benefit only under the scenario where MRI and TRUS tumors were, grade for grade, of similar oncologic risk, a position not supported by best evidence. CONCLUSIONS: Decision-analytic modelling of how MRI impacts the treatment of individual patients does not support the routine use of MRI before diagnostic biopsy, with biopsy only targeted to suspicious lesions and no biopsy for negative MRI. Source of Funding: This work was supported in part by the National Institutes of Health/National Cancer Institute (NIH/NCI) with a Cancer Center Support Grant to Memorial Sloan Kettering Cancer Center [P30 CA008748], and a SPORE grant in Prostate Cancer [P50-CA92629]. © 2020 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 203Issue Supplement 4April 2020Page: e993-e994 Advertisement Copyright & Permissions© 2020 by American Urological Association Education and Research, Inc.MetricsAuthor Information Andrew Vickers* More articles by this author Sigrid Carlsson More articles by this author Expand All Advertisement PDF downloadLoading ...