Value of contrast-enhanced ultrasound in diagnosis and differential diagnosis of polypoid lesions of gallbladder ≥ 1 cm

180 patients diagnosed with PLGs ≥ 1 cm in Shengjing Hospital of China Medical University by conventional US were selected, including 63 (35%) male and 117 (65%) female patients with a mean age of 46.8 years old (SD ± 13.1 years old). Then, all patients underwent CEUS after signing the written informed consent. The patient cannot meet one of the exclusive criteria (severe cardiovascular or cardiopulmonary disease; pregnant or lactating females; with a history of allergies, especially those allergic to contrast agents) to ensure the safety of CEUS examination. Except for five cases confirmed by magnetic resonance cholangiopancreatography (MRCP) or contrast-enhanced computer tomography (CECT), the remaining 175 cases were all confirmed by pathological examination after surgery which was performed within one week after the CEUS.

Two kinds of US systems were used: Toshiba Aplio 500 and GE LogiQ 9. The central frequency of the broadband convex array probe was 3.5 MHz and the mechanical index (MI) ranged from 0.08 to 0.12. The second-generation blood pool contrast agent, SonoVue (Bracco, Milan, Italy), consisting of phospholipid-stabilized shell microbubbles filled with sulfur hexafluoride gas, was used. In each case, a 2.4 ml bolus injection of SonoVue was administered via the antecubital vein, immediately followed by a 5 ml bolus of saline solution. If the CEUS result was unsatisfactory, the procedure was repeated with a time interval of 10 min. The images were stored in DICOM files.

Abdominal US and CEUS were performed by experienced sonographers, who worked for > 5 years with CEUS and were not involved in later data analysis. Each patient fasted for at least 8 h before the examination.

Before CEUS was performed, each patient underwent conventional US to thoroughly check the GB and liver tissue. The lesion’s location, size, shape, echo characteristics, boundary, width at the base, and blood flow within the lesion were recorded. The continuity of the GB wall and echo changes of the surrounding liver were also observed.

After the conventional US, CEUS was then performed. Each patient was asked to hold a deep breath for the entire procedure, or hold the breath for the first 30 s, to obtain the desired views. The enhancement process was divided into arterial phase (0–30 s after contrast injection) and venous phase (31–180 s after contrast injection). The extent of enhancement was classified into hypo-, iso-, hyper-, and non-enhancement, with reference to the surrounding normal GB wall. The contrast arrival time (AT) of the lesions and wash-out time were recorded by observing the entrance and exit of contrast medium in the CEUS image. Because in the process of CEUS, affected by the patient's breathing, the lesions move greatly, which is not suitable for the time-contrast curve. After 120 s, the transducer was moved to examine the entire hepatic parenchyma, especially the adjacent tissue, to exclude liver infiltration or metastases. For those patients with multiple lesions, the largest lesion was generally selected for the study. The whole process of CEUS was recorded and stored in the hard disk of the US machine to facilitate later analysis. Intraoperative and pathological findings were collected and compared with those of conventional US and CEUS images.

All statistical analyses were carried out using the SPSS 17.0 software (SPSS, Chicago, IL, USA). Quantitative data were expressed as mean ± SD, and qualitative data were expressed as a percentage. Comparisons among qualitative data were tested using the Chi-squared test. The independent t test was used for comparisons between groups. Any P values < 0.05 were considered statistically significant.

Histological diagnosis revealed 29 malignant cases and 146 benign cases. The malignant cases included 19 cases of adenocarcinoma, three cases of adenosquamous carcinoma, and seven cases developed from benign lesions (one case of adenomyomatasis and six cases of adenoma). The benign cases included 100 cases of cholesterol polyps, 29 cases of adenoma, ten cases of adenomyomatosis, three cases of chronic cholecystitis, two cases of cholecystolithiasis, one case of metaplastic polyp, and one case of hyperplastic polyp. The other five cases were confirmed malignant lesions by MRCP or CECT.

Among the 146 benign lesions, 77 cases (52.7%) were single and 69 cases (47.3%) were multiple; 129 cases (88.4%) had narrow bases or a thin pedicle. The mean diameter of the benign lesions was 1.33 ± 0.53 cm (range 1.0–5.8 cm), and the blood flow signal of only 14.4% of those lesions could have been detected.

Among the 34 malignant lesions, 28 cases (82.4%) were single with wide bases. The mean diameter of the malignant lesions was 3.29 ± 1.70 cm (range 1.1–7.2 cm). Blood flow was evident in 27 cases (79.4%) on CDFI. Significant differences between malignant and benign lesions were observed in diameter, number, width of the base, and blood flow signals within the lesion (all P < 0.05). The results are presented in Table 1.

The enhancement patterns of PLGs on CEUS images are shown in Table 2.

The mean AT of PLGs of different pathological types was computed, as shown in Table 3. No significant differences were observed in AT between the malignant and benign lesions. Different pathological types of PLGs enhancement patterns are shown in Table 2. The benign lesions (121/146, 82.9%) were mainly iso-enhanced in the arterial phase, whereas the malignant lesions (25/34, 73.5%) were mainly hyper-enhanced. The difference between the two groups was statistically significant (P = 0.000). However, no significant differences were observed in enhancement intensity among benign lesions during the arterial phase. In the venous phase, the contrast agent was frequently shown to be synchronously washed out in most benign lesions (125/146, 85.6%) in comparison to the GB wall. The majority of malignant lesions (31/34, 91.2%) typically exhibited wash-out in advance of the GB wall by almost 40 s. The difference in wash-out pattern between malignant and benign lesions indicated statistical significance (P = 0.000). Among the benign lesions, the proportion of adenomas showing hypo-enhancement during the venous phase was higher than that of cholesterol polyps, and this difference between the two groups was statistically significant (P = 0.048).

The accuracy of CEUS in the diagnosis of PLGs ≥ 1 cm was 92.22% (166/180), and in the diagnosis of benign and malignant PLGs was 92.47% (135/146) and 91.17% (31/34), respectively. The accuracy of different types of PLGs is presented in Table 4. No statistically significant differences were observed between the two groups. When we considered the discontinuity of the GB wall on CEUS images as a standard for malignant lesions that had infiltrated the GB wall or the adjacent hepatic parenchyma, the sensitivity, specificity, and accuracy were 93.33% (14/15), 92.12% (152/165), and 92.22% (166/180), respectively. If we considered the wash-out time ≤ 40 s as the diagnostic standard for malignant lesions, the sensitivity, specificity, and accuracy were 88.24% (30/34), 85.62% (125/146), and 86.11% (155/180), respectively.