Elastography by acoustic radiation force impulse technology for differentiation of benign and malignant breast lesions: a meta-analysis

To perform a meta-analysis assessing the ability of elastography by acoustic radiation force impulse (ARFI) technology to differentiate benign and malignant breast lesions. PubMed, the Cochrane Library, and the Web of Knowledge before September 24, 2014 were searched. Published studies that evaluated the diagnostic performance of ARFI for characterization of focal breast lesions were included. A total of fifteen studies, including 1720 patients with 1873 breast lesions (743 cancers, 1130 benign lesions), was analyzed. Among the included studies, virtual touch tissue imaging (VTI) was used in six studies, virtual touch tissue quantification (VTQ) in eight, combined VTI and VTQ in four, and virtual touch tissue imaging quantification (VTIQ) in three. Summary sensitivity and summary specificity for distinguishing malignant from benign breast lesions were 0.913 [95 % confidence interval (CI), 0.779-0.969] and 0.871 (95 % CI 0.773-0.930) for VTI, 0.849 (95 % CI 0.805-0.884) and 0.889 (95 % CI 0.771-0.950) for VTQ, and 0.935 (95 % CI 0.892-0.961) and 0.881 (95 % CI 0.818-0.924) for combined VTI and VTQ, respectively. The area under summary receiver operating characteristic (sROC) curve of VTI, VTQ, and combined VTI and VTQ were 0.95, 0.88, and 0.96, respectively. Significant publication bias was found only in the VTQ assessment (p = 0.025). The obtained sensitivity of VTIQ ranged from 80.4 to 90.3 %, while the specificity ranged from 73.0 to 93.0 %. The summary diagnostic value of VTIQ could not be evaluated due to insufficient data. Elastography by ARFI technology could be used as a good identification tool for differentiating benign and malignant breast lesions.