Diagnostic accuracy of virtual non-contrast enhanced dual-energy CT for diagnosis of adrenal adenoma: A systematic review and meta-analysis.

Objective: To compare the diagnostic accuracy of dual-energy (DE) virtual non-contrast computed tomography (vNCT) to non-contrast CT (NCT) for the diagnosis of adrenal adenomas.

Methods: Search of multiple databases and grey literature was performed. Two reviewers independently applied inclusion criteria and extracted data. Risk of bias was assessed using QUADAS-2. Summary estimates of diagnostic accuracy were generated and sources of heterogeneity were assessed.

Results: Five studies (170 patients; 192 adrenal masses) were included for diagnostic accuracy assessment; all used dual-source dual-energy CT. Pooled sensitivity for adrenal adenoma on vNCT was 54% (95% CI: 47-62%). Pooled sensitivity for NCT was 57% (95% CI: 45-69%). Pooling of specificity was not performed since no false positives were reported. There was a trend for overestimation of HU density on vNCT as compared to NCT which appeared related to contrast timing. Potential sources of bias were seen regarding the index test and reference standard for the included studies. Potential sources of heterogeneity between studies were seen in adenoma prevalence and intravenous contrast timing.

Conclusions: vNCT images generated from dual-energy CT demonstrated comparable sensitivity to NCT for the diagnosis of adenomas; however the included studies are heterogeneous and at high risk for some types of bias.

Key Points: • Similar sensitivity of vNCT to NCT for diagnosis of adenoma • Heterogeneity could be related to vNCT from early (<=60 sec) CECT studies • Could not pool specificity as there were no false positives • Small number of heterogeneous studies at high risk of bias.