Magnetic resonance imaging (MRI) for local staging before salvage radical prostatectomy: a meta-analysis

PURPOSE: The reliability of magnetic resonance imaging (MRI) as a local and nodal staging tool in radio-recurrent prostate cancer (PCa) is still unclear. The present study aims at evaluating the predictive value of MRI in the detection of extracapsular extension (ECE), seminal vesical invasion (SVI) and nodal involvement (LNI) in patients after primary radio (EBRT) and/or brachytherapy (BT) before salvage radical prostatectomy (SRP). METHODS: This systematic review and meta-analysis were performed in line with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. Pubmed, Scopus, and Web of Science databases were systemically reviewed to extract the data on diagnostic performance of MRI in radio-recurrent PCa. RESULTS: Four studies comprising 94 radio-recurrent PCa patients were included. The pooled prevalence of ECE, SVI, and LNI was 61%, 41%, and 20%, respectively. The pooled sensitivity for ECE, SVI and LNI detection was 53% (CI 95% 19.8–83.6%), 53% (CI 95% 37.2–68%) and 33% (CI 95% 4.7–83.1%) respectively, whereas specificity was 75% (CI 95% 40.6–92.6%), 88% (CI 95% 71.7–95.9%) and 92% (CI 95% 79.6–96.8%). The sensitivity analysis revealed that a single outlying study using only T2-weighted imaging instead of multiparametric MRI reported significantly higher sensitivity with significantly lower specificity. CONCLUSIONS: This is the first meta-analysis reporting reliability of staging MRI in a radio-recurrent setting. MRI provides poor sensitivity while maintaining high specificity for local and nodal staging before SRP. However, current evidence is limited to the low number of heterogenous studies at meaningful risk of bias. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s00345-023-04383-2.