Aims: Magnetic resonance colonography (MRC) is emerging as a potential complementary investigation for the diagnosis of colorectal cancer (CRC) and also for benign pathology such as diverticular disease. A meta-analysis reporting the use of MRC is yet to be performed. The aim of this study was to evaluate the diagnostic accuracy of MRC compared with the gold-standard investigation, conventional colonoscopy (CC).
Methods: A literature search was carried out to identify studies containing comparative data between MRC findings and CC findings. Quantitative meta-analysis for diagnostic tests was performed, which included the calculation of independent sensitivities, specificities, diagnostic odds ratios, the construction of summary receiver operating characteristic (SROC) curves, pooled analysis and sensitivity analysis. The study heterogeneity was evaluated by the Q-test using a random-effect model to accommodate the cluster of outcomes between individual studies.
Results: In all, 8 comparative studies were identified, involving 563 patients. The calculated pooled sensitivity for all lesions was 75% (95% CI: 47% to 91%), the specificity was 96% (95% CI: 86% to 98%) and the area under the ROC curve was 90% (weighted). On sensitivity analysis, MRC had a better diagnostic accuracy for CRC than for polyps, with a sensitivity of 91% (95% CI: 97% to 91%), a specificity of 98% (95% CI: 66% to 99%) and an area under the ROC curve of 92%. There was no significant heterogeneity between the studies with regard to the diagnostic accuracy of MRC for CRC.
Conclusion: This meta-analysis suggests that MRC is an imaging technique with high discrimination for cases presenting with colorectal cancer. The exact diagnostic role of MRC needs to be clarified (e.g. suitable for an elderly person with suspected CRC). Further evaluation is necessary to refine its applicability and diagnostic accuracy in comparison with other imaging methods such as computed tomography colonography.