Diagnostic Performance of Contrast-Enhanced Ultrasound for Ovarian Cancer: A Meta-Analysis

Abstract

This meta-analysis is the first study aimed at assessing the overall diagnostic performance of contrast-enhanced ultrasound for ovarian cancer. PubMed, Embase and Medline databases were systematically searched for relevant articles published up to June 2014. Data were pooled to yield summary sensitivity, specificity, diagnostic odds ratio and receiver operating characteristic curves using Meta-Disc Version 1.4 software. Ten independent studies with 579 ovarian tumors were enrolled in this meta-analysis. The pooled sensitivity, specificity and diagnostic odds ratio statistics were 0.89 (0.83–0.94), 0.91 (0.88–0.93) and 91.70 (41.41–203.05), respectively, and the area under the summary receiver operating characteristic curve was 0.9619 (standard error: 0.0125), all indicating that contrast-enhanced ultrasound has high diagnostic accuracy in differentiation of malignant from benign ovarian tumors.

Introduction

Ovarian cancer is the most lethal gynecologic malignancy (Siegel et al. 2012), partly because it is detected late, with greater than 70% of patients presenting at an advanced stage (Cohen et al. 2014). Five-year survival for all stages is 47%, and for advanced stages, <30% (Carter and Downs 2011). Early detection is one of the most important strategies for improving patient prognosis (Suh et al. 2012). In the early stage, ovarian cancer is usually asymptomatic, and moreover, only a small number of the relatively common ovarian masses detected by imaging techniques are malignant, which makes it crucial to differentiate benign from malignant ovarian masses. Transvaginal sonography usually is the initial diagnostic modality of choice for assessment of most adnexal masses; however, some masses, especially early-stage ovarian cancer, remain difficult to classify by conventional transvaginal sonography, even in experienced hands (Veyer et al., 2010, Wang et al., 2011, Xiang et al., 2013). Solid tumors and their metastases persist and grow through angiogenesis, which is characterized by a neovascular network with irregularly branching vessels derived from pre-existing normal venules that contain numerous arteriolar–venous malformations without an intact basement membrane (Feldmann et al. 1999). Thus, imaging of vessels in tumors may help to assess the risk of ovarian cancer. However, radiologic assessment of tumor vascularity is not yet well established. The lack of screening tests for diagnosis of early-stage ovarian cancer is an important determinant of the mortality rate of this disease (Enakpene et al. 2009).

Contrast-enhanced ultrasound (CEUS), with the use of contrast agents consisting of gas microbubbles that are administered intravenously and remain intravascular, has been used to evaluate many tumors in the liver, kidneys, pancreas, breasts and other organs (Jakobsen et al. 2005), improving the characterization of tumor angiogenesis and perfusion. In addition, the kinetics of contrast agents in tumors can be evaluated objectively by quantifying time–intensity curve (TIC) parameters. A few studies have investigated the use of contrast-enhanced sonography in the differential diagnosis of malignant versus benign ovarian masses. Diagnostic accuracy in these published studies varied widely, with the sensitivity ranging from 74% to 100% and the specificity ranging from 42% to 98% (D'Arcy et al., 2004, Fleischer et al., 2009, Huchon et al., 2012, Kupesic and Kurjak, 2000, Marret et al., 2004, Orden et al., 2003, Testa et al., 2007, Testa et al., 2009, Veyer et al., 2010, Xiang et al., 2013), probably as a result of advances in technology, the heterogeneity of patient populations and so on. The overall accuracy of CEUS in the diagnosis of ovarian cancer has never been systematically assessed. The purpose of this study was to perform a systematic review and meta-analysis of published information to assess the overall diagnostic performance of CEUS in ovarian cancer.