The diagnostic accuracy of high b-value diffusion- and $ T_{2} $-weighted imaging for the detection of prostate cancer: a meta-analysis

Purpose This study aims to investigate the role of diffusion-weighted imaging (DWI) and $ T_{2} $-weighted imaging ($ T_{2} $WI) in combination for the detection of prostate cancer, specifically assessing the role of high b-values (> 1000 s/$ mm^{2} $), with a systematic review and meta-analysis...
Ausführliche Beschreibung

Purpose This study aims to investigate the role of diffusion-weighted imaging (DWI) and $ T_{2} $-weighted imaging ($ T_{2} $WI) in combination for the detection of prostate cancer, specifically assessing the role of high b-values (> 1000 s/$ mm^{2} $), with a systematic review and meta-analysis of the existing published data. Methods The electronic databases MEDLINE, EMBASE, and OpenSIGLE were searched between inception and September 1, 2017. Eligible studies were those that reported the sensitivity and specificity of DWI and $ T_{2} $WI for the diagnosis of prostate cancer by visual assessment using a histopathologic reference standard. The QUADAS-2 critical appraisal tool was used to assess the quality of included studies. A meta-analysis with pooling of sensitivity, specificity, likelihood, and diagnostic odds ratios was undertaken, and a summary receiver-operating characteristics (sROC) curve was constructed. Predetermined subgroup analysis was also performed. Results Thirty-three studies were included in the final analysis, evaluating 2949 patients. The pooled sensitivity and specificity were 0.69 (95% CI 0.68–0.69) and 0.84 (95% CI 0.83–0.85), respectively, and the sROC AUC was 0.84 (95% CI 0.81–0.87). Subgroup analysis showed significantly better sensitivity with high b-values (> 1000 s/$ mm^{2} $). There was high statistical heterogeneity between studies. Conclusion The diagnostic accuracy of combined DWI and $ T_{2} $WI is good with high b-values (> 1000 s/$ mm^{2} $) seeming to improve overall sensitivity while maintaining specificity. However, further large-scale studies specifically looking at b-value choice are required before a categorical recommendation can be made. Ausführliche Beschreibung