Ultrasound in Med. & Biol., Vol. -, No. -, pp. 1–11, 2014 Copyright Ó 2014 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/$ - see front matter

http://dx.doi.org/10.1016/j.ultrasmedbio.2014.06.015

d

Original Contribution DIFFERENTIAL DIAGNOSIS OF GALLBLADDER WALL THICKENING: THE USEFULNESS OF CONTRAST-ENHANCED ULTRASOUND JUN-MEI XU,* LE-HANG GUO,* HUI-XIONG XU,*y SHU-GUANG ZHENG,*y LIN-NA LIU,*y LI-PING SUN,* MING-DE LU,y XIAO-YAN XIE,y WEN-PING WANG,z BING HU,x KUN YAN,{ HONG DING,z SHAO-SHAN TANG,k LIN-XUE QIAN,# and BAO-MING LUO** * Department of Medical Ultrasound, Shanghai Tenth People’s Hospital, Tenth People’s Hospital of Tongji University, Shanghai, China; y Department of Medical Ultrasonics, First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China; z Department of Ultrasound, Zhongshan Hospital of Fudan University, Shanghai, China; x Department of Ultrasound in Medicine, Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, Shanghai, China; { Department of Ultrasound, Peking University School of Oncology, Beijing Cancer Institute, Beijing, China; k Department of Ultrasound, Shengjing Hospital of China Medical University, Shenyang, China; # Department of Ultrasound, Beijing Friendship Hospital, Capital Medical University, Beijing, China; and ** Department of Ultrasound, Second Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China (Received 8 May 2014; revised 23 June 2014; in final form 26 June 2014)

Abstract—The purpose of this study was to evaluate the usefulness of contrast-enhanced ultrasound (CEUS) in the differential diagnosis of gallbladder wall (GBW) thickening and determine the predictors of malignant GBW thickening. One hundred fifty-nine patients with GBW thickening, including 76 men and 83 women, from eight institutions were enrolled. CEUS was performed after injection of a sulfur hexafluoride microbubble-based ultrasound contrast agent. Multiple logistic regression analysis was used to reveal independent predictors associated with malignant GBW thickening. The final diagnoses were 48 gallbladder carcinomas and 111 benign gallbladder diseases. Maximal thicknesses of the GBW in malignant and benign GBW thickening were 17.3 ± 5.2 (6–30) mm and 8.6 ± 5.1 (4–26) mm respectively (p , 0.001). CEUS revealed significant differences in intralesional vessels, enhancement homogeneity, time to hypo-enhancement, inner layer discontinuity, outer layer discontinuity and adjacent liver involvement (all p-values , 0.05) between malignant and benign GBW thickening. Patient age . 46.5 y, focal GBW thickening, inner layer discontinuity and outer layer discontinuity were found to be associated with malignancy by multiple logistic regression analysis (all p-values , 0.05). Receiver operating characteristic curve analysis revealed Az values for patient age, focal GBW thickening, inner wall discontinuity and outer wall discontinuity of 0.709 (95% confidence interval [CI]: 0.627–0.790), 0.714 (95% CI: 0.630–0.798), 0.860 (95% CI: 0.791–0.928) and 0.858 (95% CI: 0.783–0.933), respectively. CEUS is useful in the differential diagnosis between malignant and benign GBW thickening. Focal GBW thickening, inner wall discontinuity and outer wall discontinuity observed on CEUS are diagnostic clues for malignant GBW thickening. (E-mail: [email protected]) Ó 2014 World Federation for Ultrasound in Medicine & Biology. Key Words: Gallbladder, Wall thickening, Contrast-enhanced ultrasound, Differential diagnosis, Chronic cholecystitis, Gallbladder carcinoma.

INTRODUCTION

GBW thickening such as chronic cholecystitis, acute cholecystitis, adenomyomatosis and GB carcinoma is difficult with conventional US (Joo et al. 2013; Kim et al. 2012; Wibbenmeyer et al. 1995; Yoshimitsu et al. 2001). The survival rates for patients with GB cancer remain dismal; thus, diagnosis at an early stage has important implications in selecting the appropriate surgical method and improving the long-term prognosis (Kapoor et al. 2011; Wibbenmeyer et al. 1995). In the early stage, 30% to 40% of GB carcinomas present as wall thickening. Unfortunately, GB carcinoma is sometimes diagnosed at an advanced stage, after tumors

Gallbladder wall (GBW) thickening is a frequently detected finding on ultrasound (US) in the daily clinical practice of abdominal imaging examination (Joo et al. 2013; Kim et al. 2012; Yoshimitsu et al. 2001). Differentiation between commonly observed causes of

Address correspondence to: Hui-Xiong Xu, Department of Medical Ultrasound, Shanghai Tenth People’s Hospital, Tenth People’s Hospital of Tongji University, No. 301, Yanchangzhong Road, Shanghai 200072, China. E-mail: [email protected] 1

2

Volume -, Number -, 2014

Ultrasound in Medicine and Biology

Table 1. Patient characteristics and conventional US features of the 159 patients with GBW thickening

Patient characteristics Age (y) Gender (male/female) GBW thickness (mm) 4–10 mm 11–20 mm .20 mm GBW thickening morphology Diffuse thickening Focal thickening GBW thickening location Total Bottom Body Neck GBW echogenicity Hyper-echogenic Iso-echogenic Hypo-echogenic Mixed Intramural cystic spaces (yes/no) Intramural echogenic foci (yes/no) Intralesional color Doppler signals None Scarce Abundant Involvement of adjacent liver (present/absent) Coexisting gallstones (present/absent)

Malignant (n 5 48)

Benign (n 5 111)

p-value

63.0 6 9.6 (28–78) 22/26 17.3 6 5.2 (6–30) 5 30 13

52.4 6 14.6 (24–78) 54/57 8.6 6 5.1 (4–26) 77 32 2

,0.001* 0.746 ,0.001* ,0.001*

8 40

66 45

,0.001*

9 12 16 11

65 18 23 5

,0.001*

4 26 18 0 0/48 0/48

46 40 15 10 12/99 10/101

,0.001*

2 28 18 28/20 13/35

39 57 15 23/88 37/74

,0.001*

0.018* 0.032

,0.001* 0.439

GBW 5 gallbladder wall. * Statistically significant difference.

have invaded adjacent organs. Approximately 15%–30% of patients manifest no pre-operative or intra-operative evidence of malignancy and are not diagnosed until the typical histologic appearance is identified on postoperative microscopic evaluation (Contini et al. 1999). Although transabdominal US and computed tomography (CT) are widely used, these imaging modalities may not yield a precise pre-operative assessment of the nature of GBW thickening (Chun et al. 1997; Yoshimitsu et al. 2001). Therefore, new US techniques such as 3-D US, high-frequency US, endoscopic US and contrast-enhanced US (CEUS), have been employed with the aim of improving diagnosis (Liu et al. 2012; Xie et al. 2010; Xu et al. 2003; Zheng et al. 2013). Among them, increasing attention has been paid to CEUS, which can reveal the macro- and microcirculation of GB lesions, thus providing additional diagnostic information in comparison with conventional US, and may in turn facilitate the diagnosis of GB diseases (Adamietz et al. 2007; Inoue et al. 2007; Liu et al. 2012; Meacock et al. 2010; Piscaglia et al. 2012; Sparchez and Radu 2012; Tsuji et al. 2012; Xie et al. 2010; Zheng et al. 2013). However, until now, no reports have been published to evaluate the usefulness of CEUS in the differentiation diagnosis between malignant and benign GBW thickening. The aim of this

study was to evaluate the diagnostic performance of CEUS in diagnosing GBW thickening prospectively. METHODS Patients This study received approvals from the institutional ethics committees of the eight university hospitals. All patients gave written informed consent after the procedure had been carefully explained them. The clinical investigation was conducted according to the principles expressed in the Declaration of Helsinki. Between January 2007 and December 2013, we prospectively examined, with CEUS, patients referred to eight university hospitals with GBW thickening identified on conventional US and scheduled for surgery by the clinicians. The inclusion criteria were as follows: (i) GBW thickness $4 mm at conventional US; (ii) willingness to participate in the study and undergo CEUS examination; (iii) confirmation of diagnoses by pathologic examination after surgical excision or typical clinical and imaging results. Exclusion criteria included recent coronary syndrome or unstable angina, pregnancy or lactation, age ,18 y or .80 y and inability to retrieve imaging data. Finally, a total of 201 patients met the inclusion criteria. Among the 201 there were 22 patients with recent unstable

CEUS in diagnosis of gallbladder wall thickening d J.-M. XU et al.

3

Fig. 1. Seventy-seven-year-old woman with chronic cholecystitis with gallstone. (a) Baseline ultrasound reveals that the maximal thickness of the diffusely thickened gallbladder wall (thin arrows) is 7 mm and a gallstone (thick arrow) with posterior acoustic shadowing is present. (b) Color Doppler ultrasound reveals no flow signal in the gallbladder wall (thin arrows). (c) Twenty-four seconds after contrast administration, contrast-enhanced ultrasound reveals triple-layer enhancement of the gallbladder wall (thin arrows) with a hyper-enhancing inner layer, hypo-enhancing middle layer and hyper-enhancing outer layer. No discontinuity of the inner or outer layer is seen. (d) Sixty-one seconds after contrast administration, the gallbladder wall (thin arrows) is hypo-enhanced compared with adjacent liver.

angina, 2 pregnant women, 1 lactating woman, 4 patients younger than 18 y, 6 patients older than 80 y and 7 patients with missing imaging data, who were excluded from the study. A total of 159 patients with GBW thickening were finally included in this study. There were 76 (47.8%) men and 83 (52.2%) women. Age ranged from 24 to 78 y (mean 6 standard deviation [SD]: 55.6 6 14.1 y). The basic patient characteristics are summarized in Table 1. Reference diagnoses for all patients were based on histopathologic results obtained from surgery (n 5 157) or typical clinical and imaging results (n 5 2). Conventional and contrast-enhanced ultrasound The following US scanners were used depending on availability in different institutions: LOGIQ E9 (GE Healthcare, Milwaukee, WI, USA), LOGIQ 9 (GE Healthcare), Acuson Sequoia 512 (Siemens Medical Solutions, Mountain View, CA, USA), Aplio XV (Toshiba Medical System, Tokyo, Japan), IU 22 (Philips Medical Systems, Bothell, WA, USA); Technos DU8 (Esaote, Genoa, Italy). Machine settings were standardized and optimized before implementation with the help of engi-

neers from the manufacturers after review of the study protocol. The imaging settings were used consistently for the same type of US scanner. Abdominal convex transducers were used, and transducer frequency ranged from 1.0 to 6.0 MHz. Contrast-specific imaging (CSI) modes were used for CEUS in all US systems at a low mechanical index (,0.2), which enables effective tissue cancellation to generate almost pure microbubble images and avoids destruction of microbubbles in the circulation. All patients fasted at least 8 h before US examination. Patients were scanned in the supine and left lateral decubitus positions. Baseline US and CEUS were performed by radiologists with more than 5 y of experience in GB US and at least 2 y of experience in GB CEUS. The US examination was performed according to a standardized protocol arrived at in a consensus meeting before the study. Each patient underwent a complete US scanning of the GB and the adjacent liver before CEUS. The most thickened portion of the GBW was measured on US. A second-generation blood pool US contrast agent, BR1 (SonoVue, Bracco Imaging, Milan, Italy), consisting of phospholipid-stabilized shell microbubbles filled with sulfur hexafluoride gas, was used in this study. In each

4

Ultrasound in Medicine and Biology

Volume -, Number -, 2014

Fig. 2. Sixty-three-year-old man with acute gangrenous cholecystitis with gallstone. (a) Baseline ultrasound reveals that the maximal thickness of the diffusely thickened hypo-echoic gallbladder wall (thin arrows) is 19 mm and a debris-like gallstone (thick arrow) is present. (b) Color Doppler ultrasound reveals scarce flow signals in the gallbladder wall (thin arrows). (c) Twenty-three seconds after contrast administration, contrast-enhanced ultrasound reveals non-enhancement of the gallbladder wall (thin arrows) with discontinuity (asterisks) of the outer layer, indicating destruction of the gallbladder wall. The gallstones in the cavity, which appear as hyper-echoic masses on US, are non-enhanced. (d) Thirtysix seconds after contrast administration, hybrid-mode contrast-enhanced ultrasound with the contrast image overlapping the baseline image reveals non-enhancement of the gallbladder wall and discontinuity of the outer layer more clearly.

case, a dose of 1.5 to 2.4 mL contrast agent was administered through a 20-gauge cannula embedded in the antecubital vein in bolus fashion (within 1–2 s), followed by a flush of 5 mL of 0.9% normal saline. The dose of contrast agent was recommended by the manufacturers, and generally the same dose was used for the same machine. Contrast-enhanced ultrasound was performed after baseline US. The thickest portion of the GBW was placed in the center of the screen and the transducer was kept in a stable position. The imaging mode was shifted to CSI CEUS mode, and the mechanical index settings were adjusted to provide sufficient tissue cancellation with the maintenance of adequate depth penetration. Focus was positioned just below the bottom of the lesion and maintained at the same position throughout the examination. CEUS images were recorded continuously for a period of 120 s immediately after injection of contrast, without any change in machine settings. The transducer was then moved to scan the adjacent liver tissue 120 s after contrast administration to exclude liver infiltration or liver metastasis. The CEUS vascular phases were defined as follows: arterial phase (,30 s) and venous phase (31– 120 s) (Liu et al. 2012; Xie et al. 2010; Zheng et al. 2013).

Baseline US images and CEUS cine clips were stored digitally on hard disks in the imaging systems. Image reading and analysis Baseline US images and CEUS cine clips were evaluated by two experienced investigators (i.e., more than 6 y of experience in GB US and 4 y of experience in GB CEUS) independently. Disagreement was solved by another senior investigator who had more than 15 y experience in GB US and 9 y of experience in GB CEUS. Before image analysis, several tentative cases were allocated to the two investigators for evaluation and discussion unless high inter-observer agreement was achieved. All images were reviewed using Digital Imaging and Communications in Medicine viewing software (Showcase, Trillium Technology, Ann Arbor, MI, USA). The Baseline US images and CEUS clips were randomly selected for presentation so that GB diseases were not grouped by diagnosis, and any identifying information (i.e., site number, case sequence number, name, gender and age of each patient) was concealed. Investigators were blind to the final diagnoses, relevant laboratory data and other imaging results. Baseline US and CEUS

CEUS in diagnosis of gallbladder wall thickening d J.-M. XU et al.

5

Fig. 3. Sixty-seven-year-old man with moderately differentiated gallbladder adenocarcinoma. (a) Baseline ultrasound reveals an enlarged gallbladder with focal septum thickening (thin arrows) in the gallbladder body. The maximal thickness of the gallbladder is 15 mm, and debris (thick arrow) is seen in the cavity. (b) Color Doppler ultrasound reveals scarce flow signals in the thickened gallbladder wall (thin arrows). (c) Eighteen seconds after contrast administration, contrastenhanced ultrasound reveals hyper-enhancement of the thickened gallbladder wall (thin arrows). The debris in the cavity is non-enhanced. (d) Twenty-six seconds after contrast administration, contrast-enhanced ultrasound still reveals hyperenhancement of the thickened gallbladder wall.

images were analyzed by the same investigator, with the baseline US images evaluated in advance. The following US features of GBW thickening were evaluated: morphologic type (diffuse or local thickening); location (fundus, body, neck or entire GB); echogenicity (hyper-echoic, iso-echoic, hypo-echoic or mixed); intra-lesional color Doppler US signals (abundant, scarce or none); involvement of adjacent liver (presence or absence); coexisting GB stone (presence or absence). On CEUS, intralesional vascularity during the arterial phase was divided into branched, linear or dotted vessels. The extent of enhancement of the thickened GBW during the arterial and venous phases was divided into hyper-, iso-, hypoor non-enhancement, with reference to the adjacent normal liver tissue. Enhancement homogeneity was classified as homogeneous or inhomogeneous enhancement. Time to arrival of contrast at the thickening GBW, time to arrival of contrast at adjacent liver parenchyma, time to peak enhancement, time to iso-enhancement and time to hypo-enhancement (washout time) were evaluated by reviewing the cine loops visually frame by frame. GBW layer enhancement was divided into single, double or triple layer. In general, single-layer enhancement indi-

cated the thickened GBW enhanced at the same level. Double-layer enhancement indicated the GBW enhanced at two different levels and there was clear distinction between them. Triple-layer enhancement indicated hyperenhanced inner and outer layers with a hypo-enhanced middle layer. Discontinuity of the inner layer and outer layer of the GBW was observed. Involvement of adjacent liver tissue was also documented; the involved liver often exhibited hypo-enhancement in the late phase. Statistical analysis Statistical analyses were performed using the SPSS 13.0 software package (SPSS, Chicago, IL, USA). Continuous data were expressed as means 6 SD. Comparisons between qualitative data were performed using the c2-test or Fisher exact probability test. Comparisons between quantitative data were evaluated using the independent t-test if a normal distribution was achieved and the non-parametric Mann–Whitney U-test otherwise. Multiple logistic regression analysis was applied to reveal independent predictors associated with malignant GBW thickening, and subsequently receiver operating characteristic (ROC) curve analysis was carried out to evaluate

6

Ultrasound in Medicine and Biology

Volume -, Number -, 2014

Fig. 4. Sixty-nine-year-old woman with moderately differentiated gallbladder adenocarcinoma. (a) Baseline ultrasound reveals hypo-echoic focal gallbladder wall thickening (thin arrows). The maximal thickness of the gallbladder is 14 mm. The outer layer of the gallbladder wall appears continuous. (b) Color Doppler ultrasound reveals abundant flow signals in the thickened gallbladder wall (thin arrows). (c) Sixteen seconds after contrast administration, contrast-enhanced ultrasound reveals irregular hyper-enhancement of the thickened gallbladder wall (thin arrows). The inner and outer layers of the gallbladder wall are discontinuous (asterisks), and the adjacent liver is involved by the hyper-enhanced lesion. (d) Fifty-two seconds after contrast administration, contrast-enhanced ultrasound reveals hypo-enhancement of the thickened gallbladder wall, and a triple-layer pattern is observed.

their diagnostic performance using the area under the ROC curve (Az). The diagnostic value is regarded as low (Az, 0.5–0.7), moderate (Az, 0.7–0.9) or high (Az . 0.9), respectively. The corresponding cutoff value and associated sensitivity and specificity were obtained from the ROC analysis. Youden’s index was calculated as sensitivity 1 (specificity – 1). A two-tailed p , 0.05 indicated a statistically significant difference.

(n 5 2), schwannoma (n 5 1), adenoma (n 5 6), xanthogranulomatous cholecystitis (n 5 1), acute cholecystitis (n 5 1), acute gangrenous cholecystitis (n 5 1), chronic cholecystitis (n 5 65), episode of chronic cholecystitis (n 5 3), chronic granulomatous inflammation with acute infection (n 5 1) and adenomyomatosis (n 5 22, including diffuse adenomyomatosis in 4, segmental adenomyomatosis in 11 and mass-like adenomyomatosis in 7) (Table 1).

RESULTS Final diagnoses One hundred fifty-seven patients were histologically confirmed with specimens after surgery and the remaining 2 cases of cholecystitis were confirmed by followup with clinical and imaging data. The final diagnoses of GBW thickening were 48 GB carcinomas, including adenocarcinoma (n 5 44), squamous cell carcinoma (n 5 1), adenosquamous carcinoma (n 5 2) and undifferentiated carcinoma (n 5 1); and 111 benign lesions including cholesterol polyp (n 5 5), inflammatory polyp (n 5 1), adenomatous polyp (n 5 1), inflammatory myofibroblastic tumor (n 5 1), chronic GBW abscess

Basic and conventional US characteristics Between those with malignant and those with benign GBW thickening, there were significant differences in age, GBW thickness, GBW thickening morphologic type, GBW thickening location, GBW echogenicity, intramural cystic spaces in GBW, intramural echogenic foci, intra-lesional color Doppler signals and involvement of adjacent liver (all p-values ,0.05). Malignant GBW thickening was more often encountered in elderly patients and patients with GBW thickness .10 mm, focal GBW thickening, iso- or hypo-echoic GBW, GBW with abundant color Doppler signals and involvement of

CEUS in diagnosis of gallbladder wall thickening d J.-M. XU et al.

7

Fig. 5. Sixty-four-year-old man with moderately differentiated gallbladder adenocarcinoma. (a) Baseline ultrasound reveals hypo-echoic diffuse gallbladder wall thickening (thin arrows) with gallstones stuffed in the cavity. The maximal thickness of the gallbladder is 13 mm. The outer layer of the gallbladder wall appears continuous, and the inner layer is blurred. (b) Color Doppler ultrasound reveals no flow signals in the thickened gallbladder wall (thin arrows). (c) Nineteen seconds after contrast administration, contrast-enhanced ultrasound reveals irregular hyperenhancement (asterisks) of the inner layer which projects into the cavity. Triple-layer enhancement is observed; linear vessels are observed in the middle layer, and the continuous line of the outer layer is seen. (d) Thirty-nine seconds after contrast administration, contrast-enhanced ultrasound reveals hypo-enhancement of the thickened gallbladder wall (thin arrows).

adjacent liver parenchyma (Table 1) (Figs. 1–5). Hepatic invasion was seen in 28 (58.3%) of 48 patients with malignant GBW thickening and 23 (20.7%) of 111 with benign GBW thickening (p , 0.001). CEUS characteristics On CEUS, such features as intra-lesional vascularity during the arterial phase, enhancement homogeneity, time to hypo-enhancement, inner layer discontinuity, outer layer discontinuity and adjacent liver involvement significantly differed between malignant and benign GBW thickening (all p-values ,0.05). Malignant GBW thickening was more often associated with GBWs with branched or linear intra-lesional vessels, inhomogeneous GBW enhancement, quicker time to hypo-enhancement, discontinuous inner or outer layer and involvement of adjacent liver (Table 2, Figs. 1–5). Multiple logistic regression analysis On multiple logistic regression analysis, age, GBW thickening morphologic type, inner layer discontinuity and outer layer discontinuity were found to be

associated with malignant GBW thickening (all p-values ,0.05). Although liver involvement had a high odds ratio, the associated p-value was 0.997 (Table 3, Figs. 1–5). ROC analyses On ROC analyses, focal GBW thickening, inner wall discontinuity and outer wall discontinuity yielded Az values of 0.714, 0.860 and 0.858, respectively, whereas the Az values for patient age and liver involvement were 0.709 and 0.769, respectively (Table 4, Fig 6). The associated sensitivity ranged from 52.1% to 100%, and the specificity, from 37.1% to 100%. Youden’s index for inner layer discontinuity was 0.719, which was the highest for all significant independent variables. Wall thickness, although not a significant predictor, had an Az value of 0.879 (95% confidence interval [CI]: 0.826– 0.932) (p , 0.001). The cutoff value for wall thickness was 9.5 mm, and the associated sensitivity and the specificity were 97.9% and 61.8%, respectively (Figs. 1–5). The combination of GBW thickness $10 mm and inner layer discontinuity yielded an Az value of 0.857 (95% CI: 0.785–0.929) (Table 4).

8

Ultrasound in Medicine and Biology

Volume -, Number -, 2014

Table 2. Contrast-enhanced ultrasound features of the 159 patients with GBW thickening

GBW enhancement layer Single layer Double layer Triple layer Intra-lesional vessels Branched Linear Dotted Enhancement pattern (homogeneous/inhomogeneous) Contrast time Arrival time to liver (s) Arrival time to GBW (s) Time to peak (s) Time to iso-enhancement (s) Time to hypo-enhancement (s) Enhancement extent Arterial phase Hyper-enhancement Iso-enhancement Hypo-enhancement No enhancement Venous phase Hyper-enhancement Iso-enhancement Hypo-enhancement Inner layer discontinuity (presence/absence) Outer layer discontinuity (presence/absence) Liver infiltration (presence/absence)

Malignant (n 5 48)

Benign (n 5 111)

p-value

23 4 21

68 5 38

0.166

8 38 2 19/29

0 40 71 84/27

,0.001*

18.9 6 4.4 (11–34) 15.8 6 3.4 (9–24) 20.6 6 4.0 (12–32) 26.6 6 6.4 (13–40) 42.9 6 20.4 (18–120)

18.0 6 4.2 (10–29) 15.6 6 4.0 (7–27) 20.1 6 4.4 (11–31) 30.4 6 13.5 (13–120) 69.2 6 32.3 (18–176)

0.247 0.763 0.521 0.060 ,0.001*

40 6 2 0

103 7 1 0

0 2 46 41/7 37/11 25/23

0 11 100 15/96 6/105 0/111

,0.001*

0.052

0.228

,0.001* ,0.001* ,0.001*

GBW 5 gallbladder wall. * Statistically significant difference.

DISCUSSION Abdominal US is the initial imaging modality of choice for GB disease. US allows an accurate diagnosis in 70%–80% of advanced GB carcinoma cases, but in only 20%–30% of early cases (Gore et al. 2002; Tsuchiya 1991). GB polypoid carcinoma, usually characterized by an irregular margin and abundant intralesional blood flow, can be well visualized on US (Corwin et al. 2011). In contrast, superficial or infiltrating GB carcinoma, which always appears as GBW thickening, is hard to visualize on US. However, approximately 68% of GB carcinomas are infiltrating tumors, and the remainder (32%) exhibit intraluminal polypoid growth (Sons et al. 1985). In addition, destruction of the GBW Table 3. Multivariate logistic regression analysis in prediction of malignant GBW thickening Odds ratio Age Focal gallbladder wall thickening Inner layer discontinuity Outer layer discontinuity Liver involvement

95% Confidence interval

1.157 4269.885

1.019–1.314 16.867–1080920

281.234 769.958 4E1010

12.920–6121.753 7.595–78054.706 0.000

p-value 0.025 0.003 ,0.001 0.005 0.997

beneath the lesion and infiltration into the adjacent liver tissue, two features highly suspicious of malignancy, are hard to visualize on conventional US (Liu et al. 2012). The difficulty in US diagnosis of GB malignancy with wall thickening may be ascribed to several factors. Coexistence of gallstones in GB carcinoma is common (26%–79% of all cases), and the gallstones might occupy the entire GB and obscure a superficial or infiltrating lesion, making an accurate diagnosis difficult. Acute inflammation with GBW edema or thickening, as well as purulent bile or sludge, might hinder identification of concurrent GB malignancy. It has been reported that approximately one-fifth of patients with GB carcinoma present with acute cholecystitis, and the diagnostic rates of preoperative imaging range from 0 to 25.4% for GB manifesting as acute cholecystitis (Liang et al. 2009). The aforementioned factors limit the application of conventional US in the diagnosis of GB carcinoma manifesting as GBW thickening. In this series, the thickness of the GBW was significantly greater in malignant GBW thickening than in benign GBW thickening. However, owing to overlap in GBW thicknesses, exclusive consideration of wall thickness measured on US may be of limited value. Multiple logistic regression analysis also revealed that GBW thickness failed to be an independent predictor of GB

CEUS in diagnosis of gallbladder wall thickening d J.-M. XU et al.

9

Table 4. Receiver operating characteristic analyses of the independent variables in differentiation of malignant from benign GBW thickening Az

95% CI

Age Focal GBW thickening Inner layer discontinuity Outer layer discontinuity Liver involvement Age 1 GBW thickness 1 inner layer discontinuity

0.709 0.714 0.860 0.858 0.769

0.627–0.790 0.630–0.798 0.791–0.928 0.783–0.933 0.666–0.855

0.620

0.518–0.723

GBW thickness 1 inner layer discontinuity

0.857

0.785–0.929

Cutoff value

Sens

Spec

YI

.46.5 y Focal wall thickening Discontinuous Discontinuous Involved

100% 83.3% 85.4% 77.1% 52.1%

37.1% 59.5% 86.5% 94.6% 100%

0.371 0.428 0.719 0.717 0.521

Age $ 70 y 1 GBW $ 10 mm 1 inner layer discontinuity GBW $ 10 mm 1 inner layer discontinuity

25.0%

99.1%

0.241

81.3%

90.1%

0.714

GBW 5 gallbladder wall; Sens 5 sensitivity; Spec 5 specificity; YI 5 Youden’s index.

malignancy, although ROC analysis indicated that the Az value for GBW was as high as 0.879. On the other hand, focal wall thickening was found to be an independent predictor of GB malignancy by multiple logistic regression analysis. However, substantial overlap still existed between malignant and benign GBW thickening, and sometimes focal wall thickening was not able to be clearly visualized on conventional US, as illustrated in Figures 4 and 5. Instead, focal wall thickening was able to be visualized on CEUS, because most of the malignant focal wall thickening was hyper-enhanced and thus, there was a high contrast between the lesion and surrounding structures. A previous study found that disruption of the normal layers of the GBW is also associated with neoplastic GBW thickening, which can be easily detected by endoscopic US. However, endoscopic US is invasive and is difficult to use as a routine examination (Kim et al. 2012). High-frequency US has been reported to be

Fig. 6. Receiver operating characteristic (ROC) curve of the significant independent factors for gallbladder malign ancy. Morphology 5 focal wall thickening; Inner 5 inner layer discontinuity; Outer 5 outer layer discontinuity; Involvement 5 liver involvement.

capable of visualizing the GBW layer and, thus, increasing diagnostic ability. Irregular thickening of the outer wall, focal inner layer discontinuity, inner layer irregularity, inner layer thickening .1 mm and loss of the multilayer pattern in the GBW have been reported to be significantly associated with cancer (Joo et al. 2013). However, the penetration of high-frequency US is limited, and deeply located lesions may be invisible. In addition, high-frequency US still suffers from the difficulties encountered by conventional US: the lesion may be obscured by infected bile or bile sludge. As indicated in Tables 1 and 2, larger numbers of significant p-values were obtained with baseline US compared with CEUS. Multiple logistic regression analysis revealed only that age, focal GBW thickening, inner layer discontinuity and outer layer discontinuity were independent predictors of malignant GBW thickening. In addition, ROC analysis indicated that inner layer and outer layer discontinuity on CEUS achieved the highest diagnostic performance, as denoted by the area under the curve (Az). The inner layer might represent the mucosa, and the outer layer, the serosa. Disruption of these structures is a direct indicator of the extent of the lesion. Both are difficult to visualize on conventional US, whereas these are easily visualized with CEUS, as illustrated in Figures 4 and 5. The results are in agreement with previous reports that GBW destruction revealed on CEUS may be an important diagnostic clue (Liu et al. 2012; Tsuji et al. 2012; Xie et al. 2010; Zheng et al. 2013). Therefore, the use of CEUS in detecting GB malignancy should be justified. In addition, given the fact that US contrast agents are relatively harmless with a low incidence of side effects and are not nephrotoxic, CEUS can be used in patients with iodine allergies, impaired renal funct ion and urinary tract obstruction, which may be contraindications to contrast-enhanced computed tomography or magnetic resonance imaging. However, it should be taken into consideration that GBW discontinuity is also seen in some benign cases such as acute gangrenous cholecystitis and serious chronic cholecystitis (Tang et al.

10

Ultrasound in Medicine and Biology

2013; Zechner et al. 2012). The combination of baseline US and CEUS features might be helpful in that a GBW thickness $10 mm plus inner wall discontinuity also yielded reasonable diagnostic performance (Az 5 0.857). Our study has some limitations. First, most of the patients included in this study underwent surgery; thus, a majority of patients who did not undergo surgery, for example, for GBW thickening caused by liver cirrhosis, kidney function failure and heart failure, were excluded. Most of the patients with acute or chronic cholecystitis who did not require surgery were also excluded. This could lead to a selection bias; thus, the results should be carefully applied to a specific clinical scenario. Second, there was a discrepancy in patient numbers between malignant and benign GBW thickening; however, GB carcinoma is a rare tumor, and thus, this limitation is hard to overcome. Third, only one patient with xanthogranulomatous cholecystitis was included in this study. According to previous reports, the differential diagnosis between xanthogranulomatous cholecystitis and GB carcinoma is difficult with imaging study (Maker and Maker 2012; Martins et al. 2012); thus, future studies including more cases of xanthogranulomatous cholecystitis are mandatory. Fourth, time to adjacent liver parenchyma, time to peak enhancement, time to iso-enhancement and time to hypo-enhancement (washout time) should be measured quantitatively using specific software instead of the visual observation used in the present study; otherwise the measure might be inaccurate. However, different equipment and contrast dosages limited the use of quantification software. In addition, US equipment with markedly different performance was used in this study. Although it may not influence the outcome of the study, because the machine settings were optimized and low acoustic power contrast-specific modes were used, the potential bias should not be neglected. On the other hand, in this study different amounts of contrast agent was used for different machines. The various dosages were used according to the recommendations of the manufacturers. The manufacturers think the dose recommended could depict the microcirculation the best. However, for the same machine, only one dosage was used. This could minimize this confounding factor. Finally, given that so many features on baseline US and CEUS had statistical significance, the overall diagnostic performance of US, and the comparison with CEUS, is important. A retrospective blind read of all cases to compare the accuracy of conventional US versus combined conventional US and CEUS is mandatory. CONCLUSIONS Contrast-enhanced ultrasound provides another useful tool in differential diagnosis between malignant

Volume -, Number -, 2014

and benign GBW thickening. Focal GBW thickening, inner wall discontinuity and outer wall discontinuity visualized on CEUS are diagnostic clues to malignant GBW thickening. Acknowledgments—This work was supported in part by Grants 20114003 and 2013SY066 from the Shanghai Health and Family Planning Commission; Grant 81301299 from the National Natural Scientific Foundation of China; and Grant 2012045 of the Shanghai Talent Development Project from the Shanghai Human Resource and Social Security Bureau.

REFERENCES Adamietz B, Wenkel E, Uder M, Meyer T, Schneider I, Dimmler A, Bautz W, Janka R. Contrast enhanced sonography of the gallbladder: A tool in the diagnosis of cholecystitis? Eur J Radiol 2007;61: 262–266. Chun KA, Ha HK, Yu ES, Shinn KS, Kim KW, Lee DH, Kang SW, Auh YH. Xanthogranulomatous cholecystitis: CT features with emphasis on differentiation from gallbladder carcinoma. Radiology 1997;203:93–97. Contini S, Dalla Valle R, Zinicola R. Unexpected gallbladder cancer after laparoscopic cholecystectomy an emerging problem? Reflections on four cases. Surg Endosc 1999;13:264–267. Corwin MT, Siewert B, Sheiman RG, Kane RA. Incidentally detected gallbladder polyps: Is follow-up necessary? Long-term clinical and US analysis of 346 patients. Radiology 2011;258:277–282. Gore RM, Yaghmai V, Newmark GM, Berlin JW, Miller FH. Imaging of benign and malignant disease of the gallbladder. Radiol Clin North Am 2002;40:1307–1323. Inoue T, Kitano M, Kudo M, Sakamoto H, Kawasaki T, Yasuda C, Maekawa K. Diagnosis of gallbladder diseases by contrastenhanced phase-inversion harmonic ultrasonography. Ultrasound Med Biol 2007;33:353–361. Joo I, Lee JY, Kim JH, Kim SJ, Kim MA, Han JK, Choi BI. Differentiation of adenomyomatosis of the gallbladder from early-stage, wall-thickening-type gallbladder cancer using high-resolution ultrasound. Eur Radiol 2013;23:730–738. Kapoor A, Kapoor A, Mahajan G. Differentiating malignant from benign thickening of the gallbladder wall by the use of acoustic radiation force impulse elastography. J Ultrasound Med 2011;30: 1499–1507. Kim HJ, Park JH, Park DI, Cho YK, Sohn CI, Jeon WK, Kim BI, Choi SH. Clinical usefulness of endoscopic ultrasonography in the differential diagnosis of gallbladder wall thickening. Dig Dis Sci 2012;57:508–515. Liang JL, Chen MC, Huang HY, Ng SH, Sheen-Chen SM, Liu PP, Kung CT, Ko SF. Gallbladder carcinoma manifesting as acute cholecystitis: Clinical and computed tomographic features. Surgery 2009; 146:861–868. Liu LN, Xu HX, Lu MD, Xie XY, Wang WP, Hu B, Yan K, Ding H, Tang SS, Qian LX, Luo BM, Wen YL. Contrast-enhanced ultrasound in the diagnosis of gallbladder diseases: A multi-center experience. PLoS One 2012;7:e48371. Maker AV, Maker VK. Malignant masquerade of xanthogranulomatous cholecystitis on intra-operative ultrasound of the liver. J Surg Oncol 2012;106:525–526. Martins PN, Sheiner P, Facciuto M. Xanthogranulomatous cholecystitis mimicking gallbladder cancer and causing obstructive cholestasis. Hepatobiliary Pancreat Dis Int 2012;11:549–552. Meacock LM, Sellars ME, Sidhu PS. Evaluation of gallbladder and biliary duct disease using microbubble contrast-enhanced ultrasound. Br J Radiol 2010;83:615–627. Piscaglia F, Nolsøe C, Dietrich CF, Cosgrove DO, Gilja OH, Bachmann Nielsen M, Albrecht T, Barozzi L, Bertolotto M, Catalano O, Claudon M, Clevert DA, Correas JM, D’Onofrio M, Drudi FM, Eyding J, Giovannini M, Hocke M, Ignee A, Jung EM, Klauser AS, Lassau N, Leen E, Mathis G, Saftoiu A, Seidel G, Sidhu PS, ter Haar G, Timmerman D, Weskott HP. The EFSUMB

CEUS in diagnosis of gallbladder wall thickening d J.-M. XU et al. guidelines and recommendations on the clinical practice of contrast enhanced ultrasound (CEUS): Update 2011 on non-hepatic applications. Ultraschall Med 2012;33:33–59. Sons HU, Borchard F, Joel BS. Carcinoma of the gallbladder: Autopsy findings in 287 cases and review of the literature. J Surg Oncol 1985; 28:199–206. Sparchez Z, Radu P. Role of CEUS in the diagnosis of gallbladder disease. Med Ultrason 2012;14:326–330. Tang S, Wang Y, Wang Y. Contrast-enhanced ultrasonography to diagnose gallbladder perforation. Am J Emerg Med 2013;31:1240–1243. Tsuchiya Y. Early carcinoma of the gallbladder: Macroscopic features and US findings. Radiology 1991;179:171–175. Tsuji S, Sofuni A, Moriyasu F, Itokawa F, Ishii K, Kurihara T, Tsuchiya T, Ikeuchi N, Umeda J, Tanaka R, Itoi T. Contrastenhanced ultrasonography in the diagnosis of gallbladder disease. Hepatogastroenterology 2012;59:336–340. Wibbenmeyer LA, Sharafuddin MJ, Wolverson MK, Heiberg EV, Wade TP, Shields JB. Sonographic diagnosis of unsuspected gallbladder cancer: Imaging findings in comparison with benign gallbladder conditions. AJR Am J Roentgenol 1995;165: 1169–1174.

11

Xie XH, Xu HX, Xie XY, Lu MD, Kuang M, Xu ZF, Liu GJ, Wang Z, Liang JY, Chen LD, Lin MX. Differential diagnosis between benign and malignant gallbladder diseases with real-time contrast-enhanced ultrasound. Eur Radiol 2010;20:239–248. Xu HX, Yin XY, Lu MD, Liu L, Yue DC, Liu GJ. Comparison of three- and two-dimensional sonography in diagnosis of gallbladder diseases: Preliminary experience. J Ultrasound Med 2003;22: 181–191. Yoshimitsu K, Honda H, Aibe H, Shinozaki K, Kuroiwa T, Ine H, Asayama Y, Masuda K. Radiologic diagnosis of adenomyomatosis of the gallbladder: Comparative study among MRI, helical CT, and transabdominal US. J Comput Assist Tomogr 2001;25:843–850. Zechner PM, Rienm€uller S, Dorr K, Genger C, Wurzer H. Contrastenhanced ultrasound detects gallbladder perforation in a patient with acute abdominal pain. Am J Emerg Med 2012;30: 516.e5–516.e6. Zheng SG, Xu HX, Liu LN, Lu MD, Xie XY, Wang WP, Hu B, Yan K, Ding H, Tang SS, Qian LX, Luo BM. Contrast-enhanced ultrasound versus conventional ultrasound in the diagnosis of polypoid lesion of gallbladder: A multi-center study of dynamic micro-vascularization. Clin Hemorheol Microcirc 2013;55:359–374.

Differential diagnosis of gallbladder wall thickening: the usefulness of contrast-enhanced ultrasound.

he purpose of this study was to evaluate the usefulness of contrast-enhanced ultrasound (CEUS) in the differential diagnosis of gallbladder wall (GBW)...
2MB Sizes 2 Downloads 8 Views