Contrast-enhanced ultrasound for differential diagnosis of benign and malignant ovarian tumors: systematic review and meta-analysis.

X Ma,Y Zhao,B Zhang,W Ling,H Zhuo,H Jia,P Li

DOI: https://doi.org/10.1002/uog.14800

2015-01-01

Abstract:Objective To assess the performance of contrast-enhanced ultrasound (CE-US) in the differential diagnosis of malignant and benign ovarian tumors. Methods We conducted a comprehensive literature search of PubMed and EMBASE to identify published articles evaluating the diagnostic potential of CE-US for the differentiation of benign and malignant ovarian tumors. Inclusion criteria were: (1) the study assessed the accuracy (or sensitivity and specificity) of CE-US for diagnosis of benign and malignant ovarian tumors; (2) it used surgery and histopathology as the reference standard for distinguishing between benign and malignant tumors; (3) it included data allowing construction of a 2x2 contingency table for true-and false-positives and negatives. We present summary sensitivity, specificity, diagnostic odds ratio (OR) and areas under the summary receiver-operating characteristics curves (AUCs). Results Preliminary screening identified 103 papers, of which 11 fulfilled our predefined inclusion criteria and underwent final analysis. The pooled sensitivity and specificity of CE-US for diagnosis of benign and malignant ovarian tumors were 93% (95% CI, 89-96%) and 95% (95% CI, 92-96%), respectively. The pooled diagnostic OR was 171.2 (95% CI, 65.9-444.6) and the AUC was 0.98. I-2 values of sensitivity, specificity and diagnostic OR were 38.3%, 31.7% and 48.4%, respectively, all indicating moderate heterogeneity. Conclusions The evidence from available studies suggests CE-US is useful for discriminating between benign and malignant ovarian tumors; however, further studies are needed to examine whether CE-US has improved diagnostic test accuracy compared with that of standard two-dimensional Doppler sonography. Copyright (C) 2015 ISUOG. Published by John Wiley & Sons Ltd.