| Objectives:1.To examine and compare detailed contrast-enhanced ultrasonography (CEUS)images of Serous cystadenomas (SCAs) and Mucinous cystadenomas ( MCAs), and to determine whether there are significant findings that can contribute to the discrimination between these two diseases.2. To compare the different diagnostic performance between conventional ultrasonography (US) and CEUS together with endoscopic ultrasonography (EUS)and contrast-enhanced EUS (CH-EUS) in the common pancreatic cystic neoplasms.3. To compare the diagnostic performance of CEUS, EUS and contrast-enhanced magnetic resonance imaging (CE-MRI) in the differentiation of benign from pre-malignant/malignant pancreatic cystic lesions.Materials and Methods:1. From April 2015 to February 2017, 83 patients (47 patients with SCAs and 36 patients with MCAs) were enrolled in this study. Sixty-seven cases were confirmed by surgical pathology and 16 by comprehensive clinical diagnoses. All of the CEUS characteristics of these lesions were recorded: size, location, shape, wall characteristics, septa characteristics, and the presence of a honeycomb pattern or nodules. CEUS examinations were performed by two ultrasound physicians.2. From April 2015 to February 2017, we performed CEUS and CH-EUS orn 105 pancreatic cysts. Among them, 42 were SCAs, 34 were MCAs, 18 were intraductal papillary mucinous neoplasms (IPMNs),and 11 were solid pseudopapillary neoplasms(SPNs).85 were confirmed by surgical pathology and 20 by comprehensive clinical diagnoses. CEUS examinations were performed by two ultrasound physicians, andEUS examinations by two endoscopic ultrasound physicians. A uniform diagnosis was made between two ultrasound physicians and two endoscopic ultrasound physicians respectively after a discussion referred to a diagnostic criterion.3. From April 2015 to February 2017, we performed CEUS , EUS and CE-MRI on 110 pancreatic cysts which included 38 SCAs, 31 MCAs, 16 IPMNs, 8SPNs, 8 pancreatic pseudocysts, 3 neuroendocrine neoplasms, 2 cysts, 2 cystadenocarcinomas,an acinar cell carcinoma and a lymphangioma. All of the lesions were divided into benign or pre-malignant/malignant groups. A uniform diagnosis was made between two physicians respectively after a discussion referred to a diagnostic criterion. Then six groups were divided according to their location and size: pancreatic head(uncinated process)<3cm, pancreatic head (uncinated process)≥3cm, pancreatic body(neck)<3cm, pancreatic body (neck)≥3cm, pancreatic tail<3cm, pancreatic tail>3cm,and the diagnostic performance of different imaging were analysed between various groups.Results:1. Location (p=0.026), shape (p=0.001),thickness of the wall (p=0.024), the number of septa (p=0.003), and the honeycomb pattern (p=0.002) were statistically significantly different. A head-neck location, a lobulated shape, an inner regular honeycomb pattern, and a thin wall (less than 3 mm thick) were significant in diagnosing patients with SCAs. When two of these four findings were combined, we could achieve a sensitivity of 74.5% and a specificity of 77.8% to diagnose SCA;when three of these four findings were combined, the specificity was 91.6%. A body-tail location,a round/oval shape,0-2 septa, and a thick wall (3 mm or more thick) were most often detected in patients with MCAs. When two of these four findings were combined, we could achieve a sensitivity of 86.1% and a specificity of 63.8% to diagnose MCA; when three of these four findings were combined, the area under the curve (Az) was highest at 0.796, with a sensitivity of 80.5% and a specificity of 80.8%.2. For different diseases of SCAs, MCAs, IPMNs and SPNs, the diagnostic accuracy of US were 54.8%, 73.5%, 55.6%, 45.5% respectively; and 76.2%, 79.4%, 61.1%,63.6% for CEUS; 81.0%, 85.3%, 77.8%, 63.6% for EUS ; 81.0%, 91.2%, 77.8%,72.7% for CH-EUS. There was significant statistical significance between CEUS and US when diagnose SCAs (P=0.012), and for the collective diseases (P=0.013);Interobserver agreement was excellent (k=1.000) between EUS and CH-EUS when diagnose SCAs and IPMNs. There was no significant statistical significance between CEUS and EUS in the diagnostic performance.3. The diagnostic sensitivity of CEUS, EUS and CE-MRI in the differentiation of pre-malignant/malignant from benign pancreatic cystic lesions were 73.8%, 86.9%,80.3% respectively, the specificity were 69.4%, 81.6%, 83.7% respectively, and accuracy were 71.8%, 84.5%, 81.8% respectively. There was no significant statistical significance among these imaging modalities. However, when the lesion located in the pancreatic tail and the maximum diameter was smaller than 3cm, there was significant statistical significance between CEUS and EUS (P=0.039).Conclusions:1. The characteristics of tumour location, shape, thickness of the wall, the number of septa, and the honeycomb pattern by CEUS play an important role in the diagnosis of SCAs and MCAs. A combination of these findings can provide better diagnostic performance in the discrimination of SCAs from MCAs.2. CEUS has better diagnostic performance than EUS in the common pancreatic cystic neoplasms, especially in diagnosing SCAs. The diagnostic performance is similar between CH-EUS and EUS, CEUS and EUS in the common pancreatic cystic neoplasms.3. The diagnostic performance was similar between CH-EUS, EUS and MRI in the differentiation of benign from pre-malignant/malignant pancreatic cystic lesions.However, EUS has better diagnostic performance than CEUS when the lesion locates in the pancreatic tail and the maximum diameter is smaller than 3cm. |
PDF Full Text Request