Interestingly one female patient whose diagnosis could not be ruled

Interestingly, one female patient whose diagnosis could not be ruled out as malignant disease by many methods, including EUS, was finally diagnosed with inflammatory pseudo-tumor after explorative laparotomy. This case exemplifies the difficulty in clinically distinguishing malignant from benign disease. A case of immunoglobulin G-related pseudo-tumor in the rora mimicking pancreatic cancer was reported by Kim et al. (Park et al. 2008). We think this case may be an example of immunoglobulin G-related pseudo-tumor involving the stomach, which is undistinguishable from benign and malignant diseases. Thus, there are subgroups that are undistinguishable from benign and malignant diseases on EUS. Further studies of these subgroups are needed, and additional workup is needed to avoid unnecessary laparotomy.
Our study has some potential limitations. First, we could not use standard EUS protocols because our study was retrospective. Second, although our study contained more diverse disease categories compared with previous studies, we could not include all diseases in which there are large gastric folds. However, this is difficult limitation to overcome because patients with large gastric folds are relatively rare. Third, we included patients without definite histologic diagnoses and used clinical follow-up examinations to make our diagnoses; however, previous studies have also used this method (Gines et al. 2006).



Tumor angiogenesis, required for invasive tumor growth and metastasis, is an important point in the control of cancer progression (Folkman 2002; Stacker and Achen 2013). Ischemic tumors rarely grow larger than 3 mm3, but once they become vascularized, their growth rate may significantly accelerate (Folkman and Shing 1992). Microvessel density (MVD) has been the most common index of tumor vasculature since 1972 (Brem et al. 1972). Several reports suggest that higher tumor MVD correlates with tumor aggressiveness rora and poor survival in several malignancies, including liver, breast, gastric and pancreatic tumors and neuroblastoma (Couvelard et al. 2005; Fernandez-Aguilar et al. 2006; Poon et al. 2002; Ribatti and Ponzoni 2005; Zhao et al. 2005). Interestingly, studies have failed to demonstrate any significant correlation between MVD and tumor progression or patient prognosis (Hillen et al. 2006; Marioni et al. 2005; Qian et al. 2009). Such a discrepancy could be explained by the technical drawbacks in the evaluation of MVD. The diversity of tumor vasculature could also be a contributor.
Hepatocellular carcinoma (HCC) is a common malignancy with poor prognosis in China and other developing countries (Wei et al. 2014). The vasculature of HCC is suitable for morphologic classification (Chen et al. 2011; Ding et al. 2011). HCC vasculature can be classified into two major types of vessels based on morphology: (i) capillary-like microvessels, which possess a complete basement membrane and are surrounded by pericytes, and (ii) sinusoid-like vessels, which have a long, wide lumen structure and insufficient pericytes (Chen et al. 2011; Kin et al. 1994; Tanigawa et al. 1997; Yamamoto et al. 2001). Recent studies have also reported that HCC patients with sinusoid-like vasculature have a shorter survival time, although the intra-tumor MVD of a tumor with sinusoid-like vasculature is significantly less than that of a tumor with capillary-like microvessels (Chen et al. 2011; Ding et al. 2011). However, there is currently no non-invasive method for the evaluation of HCC microvasculature.
The ability to visualize angiogenesis through non-invasive methods may provide new opportunities to stratify patients for anti-angiogenic treatment and monitor treatment efficacy (Iagaru and Gambhir 2013). Contrast-enhanced ultrasonography (CEUS) of the tumor vasculature in animal models appears to be effective in evaluating tumor vasculature in a non-invasive manner (Forsberg et al. 2011; Zhou et al. 2011; Zhuo et al. 2014). CEUS has been reported to have an effect on the evaluation of microvascular morphology and angiogenesis in HCC (Hatanaka et al. 2010; Schacherer et al. 2010; Xiao et al. 2010; Zheng et al. 2013). The purpose of this study was to identify the parameters of pre-surgical contrast-enhanced ultrasonography for the classification of HCC, based primarily on the different tumor vasculature morphologies related to the prediction of post-surgical patient prognosis.