Purpose: To estimate sensitivity and specificity of CESM for breast cancer diagnosis.
Methods: Systematic review and meta-analysis of the accuracy of CESM in finding breast cancer in highly selected women. We estimated summary receiver operating characteristic curves, sensitivity and specificity according to quality criteria with QUADAS-2.
Results: Six hundred four studies were retrieved, 8 of these reporting on 920 patients with 994 lesions, were eligible for inclusion. Estimated sensitivity from all studies was: 0.98 (95% CI: 0.96-1.00). Specificity was estimated from six studies reporting raw data: 0.58 (95% CI: 0.38-0.77). The majority of studies were scored as at high risk of bias due to the very selected populations.
Conclusion: CESM has a high sensitivity but very low specificity. The source studies were based on highly selected case series and prone to selection bias. High-quality studies are required to assess the accuracy of CESM in unselected cases.
Keywords: Accuracy; Contrast-enhanced spectral mammography; Meta-analysis; Sensitivity; Specificity; Systematic review.
Copyright © 2016 Elsevier Ltd. All rights reserved.