ObjectiveTo determine the diagnostic test accuracy of multiparametric magnetic resonance imaging (mpMRI) in detecting seminal vesicle invasion (SVI).MethodsThe Medical Literature Analysis and Retrieval System Online (MEDLINE), PubMed, the Excerpta Medica dataBASE (EMBASE) and Cochrane databases were search up to May 2023. We included studies that investigated the accuracy of mpMRI in detecting SVI when compared to radical prostatectomy specimens as the reference standard. Data extraction was performed by two independent reviewers to construct 2 × 2 tables, as well as patient and study characteristics. The methodological quality of the included studies was assessed with the Quality of Assessment of Diagnostic Accuracy Studies‐2 tool. Sensitivity and specificity were pooled and presented graphically with summary receiver operator characteristic (SROC) plots.ResultsA total of 27 articles with 4862 patients were included for analysis. The summary sensitivity and specificity were 0.57 (95% confidence interval [CI] 0.45–0.68) and 0.95 (95% CI 0.92–0.99), respectively. Meta‐regression indicated that there was no evidence that coil strength (P = 0.079), coil type (P = 0.589), year of publication (P = 0.503) or use of the Prostate Imaging‐Reporting and Data System (P = 0.873) significantly influenced these results. The summary diagnostic odds ratio was 28.3 (95% CI 15.0–48.8) and the area under the curve for the SROC curve was 0.87. The I2 statistic was a modest 11.9%. In general, methodological quality was good.ConclusionThe use of mpMRI in detecting SVI has excellent specificity but poor sensitivity. Both endorectal coils and magnetic field strength do not significantly impact the accuracy of MRI. These findings suggest that mpMRI cannot reliably rule out SVI in patients with prostate cancer.

中文翻译：

---

MRI 检测前列腺癌精囊浸润的准确性：系统评价和荟萃分析


目的确定多参数磁共振成像 （mpMRI） 检测精囊浸润 （SVI） 的诊断试验准确性。方法检索截至 2023 年 5 月的医学文献分析和检索系统在线 （MEDLINE）、PubMed、Excerpta Medica 数据库 （EMBASE） 和 Cochrane 数据库。我们纳入了调查 mpMRI 检测 SVI 准确性的研究，与根治性前列腺切除术标本相比作为参考标准。由两名独立评价员进行资料提取，构建 2 × 2 个表格，以及患者和研究特征。使用 Quality of Assessment of Diagnostic Accuracy Studies-2 工具评估纳入研究的方法学质量。将敏感性和特异性合并，并以图形方式呈现汇总受试者操作员特征 （SROC） 图。结果共纳入 27 篇文献，4862 例患者进行分析。总敏感性和特异性分别为 0.57 （95% 置信区间 [CI] 0.45-0.68） 和 0.95 （95% CI 0.92-0.99）。Meta 回归分析表明，没有证据表明弹簧圈强度 （P = 0.079）、弹簧圈类型 （P = 0.589）、出版年份 （P = 0.503） 或使用前列腺成像和数据系统 （P = 0.873） 显着影响这些结果。总诊断比值比为 28.3 （95% CI 15.0–48.8），SROC 曲线曲线下面积为 0.87。I2 统计数据为 11.9%。总的来说，方法学质量良好。结论mpMRI 检测 SVI 特异性好，但敏感性差。直肠内线圈和磁场强度都不会显着影响 MRI 的准确性。 这些发现表明 mpMRI 不能可靠地排除前列腺癌患者的 SVI。