Meta-analysis of gadoxetic acid disodium (Gd-EOB-DTPA)-enhanced magnetic resonance imaging for the detection of liver metastases

Abstract

Objective: To determine the accuracy of MR imaging with Gd-EOB-DTPA for the detection of liver metastases.

Materials and methods: PUBMED, EMBASE, the Web of Science, and the Cochrane Library were searched for original articles published prior to February 2012. The criteria for the inclusion of articles were as follows: reported in the English language; MR imaging with Gd-EOB-DTPA was performed to detect liver metastases; histopathologic analysis (surgery, biopsy), intraoperative observation (manual palpatation, intraoperative ultrasonography), and/or follow-up US was the reference standard; and data were sufficient for the calculation of true-positive or false-negative values. The methodological quality was assessed by using the quality assessment of diagnostic studies instrument. The data were extracted to calculate sensitivity, specificity, predictive value, diagnostic odds ratio, and areas under hierarchical summary receiver operating characteristic (HSROC) curve to perform heterogeneity test and threshold effect test, as well as publication bias analysis and subgroup analyses.

Results: From 229 citations, 13 were included in the meta-analysis with a total of 1900 lesions. We detected heterogeneity between studies and evidence of publication bias. The methodological quality was moderate. The pooled weighted sensitivity with a corresponding 95% confidence interval (CI) was 0.93 (95% CI: 0.90, 0. 95), the specificity was 0.95 (95% CI: 0.91, 0.97), the positive likelihood ratio was 18.07 (95% CI: 10.52, 31.04), the negative likelihood ratio was 0.07 (95% CI: 0.05, 0.10), and the diagnostic odds ratio was 249.81 (95% CI: 125.12, 498.74). The area under the receiver operator characteristic curve was 0.98 (95% CI: 0.96, 0.99).

Conclusion: MR imaging with Gd-EOB-DTPA is a reliable, non-invasive, and no-radiation-exposure imaging modality with a high sensitivity and specificity for detection of liver metastases. Nonetheless, it should be applied cautiously, and large scale, well-designed trials are necessary to assess its clinical value.

Figure 1. Flowchart illustrating the selection of studies.

Figure 2. Methodological quality of the 13 included studies on a per-lesion basis.

Figure 3. Forest plots of the SEN, SPE, PLR and NLR with corresponding 95%CIs for MR imaging with Gd-EOB-DTPA for the detection of liver metastases.

Figure 4. Summary receiver operating characteristic (SROC) curves from the bivariate model of MR imaging with Gd-EOB-DTPA in detection of liver metastases.

Figure 5. The funnel plot of publication bias.

Linear regression of the inverse root of effective sample sizes (ESS) on a log dOR is performed as a test for funnel plot asymmetry.