| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology and Institute of Radiation Medicine, Seoul National
University Hospital, 101 Daehangno, Jongno-gu, Seoul 110-744, Korea.
2 Department of Radiology, Seoul National University College of Medicine, Seoul,
Korea.
3 Department of Diagnostic Radiology, Research Institute of Radiological
Science, Yonsei University College of Medicine, Seoul, Korea.
Received October 21, 2007;
accepted after revision March 16, 2008.
Addres correspondence to J. M. Lee
(leejm{at}radcom.snu.ac.kr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. MDCT scans obtained during the portal venous phase in 78 patients with gallbladder wall thickening caused by various pathologic conditions were retrospectively reviewed by two blinded observers. The CT features of benign and malignant gallbladder wall thickening were compared by means of univariate and multivariate analyses. The study cases were then divided into five patterns according to enhancement pattern. Using these five patterns, two radiologists reviewed the MDCT images and recorded their diagnostic confidence for differentiating benign versus malignant cause on a 5-point scale. The diagnostic performance of CT was evaluated by each observer using a receiver operating characteristic curve analysis.
RESULTS. The thicknesses of the inner and outer layers
("thick" enhancing inner layer 
2.6 mm, "thin"
outer layer 
3.4 mm), strong enhancement of the inner wall, and irregular
contour of the affected wall were significant predictors for a malignant cause
of gallbladder wall thickening (p < 0.001). The two-layer pattern
with a strongly enhancing thick inner layer and weakly enhancing or
nonenhancing outer layer and the one-layer pattern with a heterogeneously
enhancing thick layer were patterns that were significantly associated with
gallbladder cancer (p < 0.05). When we consider those two
enhancing patterns as a sign of malignancy, the diagnostic accuracy of MDCT
was 89.1% and 87.6% for the two observers, respectively.
CONCLUSION. Analyzing the enhancement pattern of a thickened gallbladder wall on MDCT is helpful in differentiating gallbladder cancer from benign inflammatory diseases.
Keywords: biliary tree • cholecystitis • flat gallbladder wall thickening • gallbladder cancer • gallbladder wall thickening • MDCT
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Misinterpretation of diffuse gallbladder wall thickening on imaging studies can lead to unnecessary cholecystectomy in patients without intrinsic gallbladder disease; conversely, misdiagnosis in patients who do require cholecystectomy may result in delayed treatment with increased morbidity [1]. Large series have shown that the overall 5-year survival rate of patients with gallbladder cancer after gallbladder resection is 40%, but the 5-year survival rate of all patients with gallbladder cancer is less than 5% in most series, primarily because most patients have unresectable disease when they are first seen [4, 5]. Furthermore, considering that survival after simple cholecystectomy for T1 disease is reported to be near 100% [6–8], cross-sectional imaging has an important role in distinguishing between gallbladder wall thickening resulting from a malignant cause and gallbladder wall thickening resulting from benign causes and in rendering a diagnosis of gallbladder cancer at an early stage.
Researchers have vigorously attempted to differentiate malignant from benign diffuse gallbladder wall thickening using various imaging techniques including sonography, CT, and MRI. Although sonography has been the screening method of choice in the diagnosis of gallbladder disease [9], making an accurate diagnosis may be difficult using sonography because thickening of the gallbladder wall is nonspecific [10, 11]. Furthermore, because of recent developments and spatial refinements, MDCT has become the imaging technique that is more widely used to detect and characterize gallbladder wall thickening.
Recently, Jung et al. [12] reported four characteristic MRI findings of gallbladder wall thickening according to each gallbladder disease entity. However, considering that MDCT, with its high spatial and temporal resolution, is now more widely used than MRI to evaluate acute abdominal diseases, the schematic analysis of MDCT findings for gallbladder wall thickening is timely and warranted. Furthermore, until now the diagnostic accuracy of CT for the detection of gallbladder cancer that presents with flat wall thickening has never been described to our knowledge. Therefore, the purpose of this study was to evaluate the MDCT findings of the three most common gallbladder diseases that present with diffuse wall thickening—that is, the flat-wall-thickening type of gallbladder cancer, chronic cholecystitis, and acute cholecystitis—with special emphasis on enhancement patterns of the gallbladder wall.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3 mm based on previous studies
[1,
3,
13–15].
One hundred two patients were excluded from the analysis for one of the
following reasons: an MDCT study was not performed (n = 57); a
polypoid or nodular gallbladder lesion was detected on CT (n = 24); a
gallbladder-replacing tumor was seen on CT review (n = 6); or
gallbladder wall thickening was < 3 mm (n = 15). The remaining 78
patients were included in our study. The mean age of the study population was 60.4 years (age range, 27–94 years). The study population included 53 men (age range, 27–94 years; mean age, 60.3 years) and 25 women (age range, 40–85 years; mean age, 60.4 years). Among the 78 study patients, MDCT was used as the initial imaging examination in 50 patients and as the secondary imaging technique in 28; the latter group of patients underwent MDCT because of nondiagnostic or abnormal findings on the initial imaging examination, which included sonography (n = 16) or helical CT (n = 9) with inadequate diagnostic quality or MRI (n = 3). The mean interval from MDCT to surgery was 16.4 days (range, 0–96 days).
Of the 78 cases of gallbladder wall thickening, 29 resulted from a malignant cause (infiltrative adenocarcinoma, n = 29) and 49 resulted from benign disease (chronic cholecystitis, n = 20; adenomyomatosis, n = 4; acute cholecystitis, n = 25). The diagnoses of all gallbladder wall thickening cases were confirmed by reviewing patients' surgical and pathology records. Patients with malignant gallbladder wall thickening underwent extended radical cholecystectomy (n = 11) or open cholecystectomy (n = 18). Patients with benign gallbladder wall thickening underwent extended cholecystectomy (n = 2), open cholecystectomy (n = 39), or laparoscopic cholecys tectomy (n = 8).
CT Scan Acquisition
The 78 patients included in our study under went CT at our institution on
an MDCT scanner: a Sensation 16 (Siemens Medical Solutions) (n = 20),
a LightSpeed Ultra (GE Healthcare) (n = 33), or an MX8000 4-MDCT
(Philips Healthcare) (n = 25) scanner. Each patient received
90–150 mL (1.5 mL/kg of body weight) of a nonionic contrast material
(iopromide [Ultravist 370, Schering Korea]) through an 18-gauge angio graphic
catheter inserted into a forearm vein. The contrast material was injected at a
rate of 3–5 mL/s for 30 seconds using an automatic power injector.
CT scans were routinely obtained during full inspiration with the patient in a supine position. Scanning routinely began at the level of the lung bases and proceeded in a craniocaudal direction up to the lower pole of the kidney. CT was performed using the following parameters: detector collimation, 0.75–2.5 mm; table pitch, 1:1.3–1.7; matrix, 512 x 512; and reconstruction intervals, 2.5–3 mm. MDCT scans were obtained 65–75 seconds after initiation of the contrast injection (i.e., during the portal venous phase).
CT Scan Analysis
The CT scans acquired in the 78 patients were independently reviewed on a
PACS workstation (m-view, Marotech) by two attending abdominal radiologists,
one with 17 years of clinical experience and another with 15 years of clinical
experience, who had no knowledge of the final radiologic or pathologic
findings. Reviewers were free to change the window width and level and to use
magnification at their discretion. During analysis of the CT features, cases
of malignant and of benign diffuse gallbladder wall thickening were randomly
intermixed. The review process was divided into two sessions over a 4-week
interval.
The first interpretation session—The reviewers independently analyzed the CT images, which were in random order regarding the CT findings, and differences in their assessments were resolved by consensus. First, to measure the involved wall thickness, reviewers measured the thickness of the inner layer and of the outer layer, respectively, when a two-layer pattern was identified, but they measured the entire thickness when layering of the gallbladder wall could not be identified. Second, reviewers noted the character and extent of the involved gallbladder wall—that is, irregular or smooth, focal or diffuse, respectively. Third, reviewers documented the degree of enhancement of the involved gallbladder wall during the portal venous phase. The degree of enhancement of the thickened gallbladder wall was compared with the attenuation of normal liver parenchyma. Ancillary findings—such as subserosal edema, transient hepatic attenuation difference (THAD) in adjacent liver parenchyma, pericholecystic fluid or air, increased pericholecystic fat stranding, enlarged pericholedochal lymph nodes, and gallbladder stones—were also evaluated.
On the basis of CT data from the first interpretation session, statistical
analysis was performed for differentiating benign from malignant causes of
wall thickening using the chi-square test. Depending on the statistical
analysis results, when two layers of thickened gallbladder wall could be
differentiated, a "thick" inner layer was defined as 2.6 mm
and a "thin" outer layer, as 3.4 mm. When gallbladder wall
layering could not be identified, we considered the wall to be one thick
layer. For each patient, the enhancement pattern of the thickened wall was
determined by the presence of striation, thick nesses of the inner and outer
layers, and the degree of enhancement of each layer compared with that of
normal hepatic parenchyma.
Enhancement was classified as one of five patterns: type 1 was a one-layer pattern, and types 2–5 were two-layer patterns. The type 1 pattern was a heterogeneously enhancing one-layer gallbladder wall or indistinguishable layering of the gallbladder wall (Fig. 1A); type 2, strongly enhancing thick inner layer and weakly enhancing or nonenhancing outer layer (Fig. 1B); type 3, borderline enhancement and thickness of the inner layer with small cystic spaces and nonenhancing outer layer (Fig. 1C); type 4, weakly enhancing thin inner layer and nonenhancing thin outer layer (Fig. 1D); and type 5, weakly enhancing thin inner layer and nonenhancing thick outer layer (Fig. 1E).
|
|
|
|
|
The second interpretation session—According to our experience reviewing MDCT and correlating MDCT findings with pathologic results, the prevalent enhancing patterns of gallbladder cancer were the type 1 and 2 patterns (Figs. 2A and 2B). Therefore, those two patterns (types 1 and 2) were considered to be malignant gallbladder wall thickening, although some cases of chronic granulomatous cholecystitis, such as xanthogranulomatous cholecystitis, showed a type 1 enhancement pattern. Type 3 was considered to be a borderline enhancement pattern including adenomyomatosis (Fig. 2C); type 4, chronic cholecystitis (Fig. 2D); and type 5, acute cholecystitis (Fig. 2E).
|
|
|
|
|
Statistical Analysis
Statistical differences in the CT features of malignant and benign causes
of gallbladder wall thickening were analyzed using the chi-square test. A
multivariate stepwise logistic regression model was used to determine the best
predictors of a differential diagnosis between benign and malignant
gallbladder wall thickening using an SPSS software package (version 10.0,
SPSS). Receiver operating characteristic (ROC) curve analysis was performed to
evaluate the diagnostic performance of pattern analysis for differentiating
benign from malignant flat gallbladder wall thickening. Binormal ROC curves
were fitted using a MedCalc program (version 6.15.000, MedCalc Software). The
diagnostic capability of MDCT using enhancement pattern analysis was
determined by calculating the area under the ROC curve
(Az) for each reader. Interobserver agreement between the
two readers for pattern analysis was quantified using weighted kappa
statistics. A kappa value of less than 0.20 was considered to indicate poor
agreement; 0.20–0.39, fair agreement; 0.40–0.59, moderate
agreement; 0.60–0.79, substantial agreement; and 0.80 or greater,
excellent agreement.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2.6 mm thick and
hyperenhancement of the inner wall showed the two highest values for
suggesting a malignant cause of wall thickening. In addition, an outer wall
3.4 mm thick, irregular wall thickening, focal involvement, and weak
enhancement or nonenhancement of the outer wall showed a high chi-square
value. These six results involving gallbladder wall thickening were
statistically significant (p 0.01).
|
Besides being an ancillary finding, subserosal edema showed statistical
significance (p = 0.038). The incidence of gallbladder stone, THAD in
the gallbladder bed, peri cholecystic fluid or air, pericholedochal lymph node
enlargement, and pericholecystic fat infiltration did not differ significantly
between the gallbladder carcinoma group and the acute or chronic cholecystitis
group. Using stepwise logistic regression analysis, three CT findings were
identified for malignant gallbladder wall thickening—that is, enhancing
inner wall 2.6 mm thick (p < 0.0001), outer wall 3.4 mm
thick (p = 0.0262), and irregular wall thickening (p =
0.0300).
Second Interpretation Session: Diagnostic Performance of the Enhancement Pattern Approach
The diagnostic performance of enhancement pattern analysis and the final
diagnoses of the 78 cases of gallbladder wall thickening are summarized in
Table 2. When type 1 and 2
enhancement patterns were regarded as suggestive of gallbladder cancer, the
Az values of each observer to predict a malignant cause of
flat gallbladder wall thickening were 0.891 and 0.876, respectively. Using the
five-pattern analysis, the sensitivity for detecting the flat-wall-thickening
type of gallbladder cancer was 82.8% and 75.9% and the positive predictive
value was 80% and 84.6%, respectively, for the two observers. Interobserver
agreement between the two readers for pattern analysis showed substantial
agreement with a kappa value of 0.785.
|
There were 11 false-positive diagnoses of gallbladder cancer, 11 for observer 1 and eight for observer 2 (Table 3), including 10 cases of chronic cholecystitis (Fig. 3A, 3B) and one case of acute cholecystitis. Among the false-positive cases, seven cases of chronic cholecystitis, including three cases of xanthogranulomatous cholecystitis, were seen by both observers. The three cases of xanthogranulomatous cholecystitis showed the type 1 pattern, whereas the six cases of chronic cholecystitis mimicked the type 2 pattern. To the contrary, there were four false-negative cases of gallbladder cancer, two for observer 1 and four for observer 2 (Table 3). Among these cases, both observers interpreted two cases as type 4 or 5 (Fig. 4A, 4B). The other two cases were interpreted as the type 3 pattern by observer 2 and as the type 2 pattern by observer 1. In fact, two gallbladder cancer cases that observer 2 interpreted as showing the type 3 pattern were revealed at pathology to have cancer cells infiltrating along the dilated Rokitansky-Aschoff sinuses within the thickened mucosal layer.
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Based on the results of this study, the enhancement patterns with a high prevalence of gallbladder cancer are the type 2 and 1 patterns—that is, two-layer wall thickening showing a strongly enhancing thick inner layer and a weakly enhancing or nonenhancing outer layer (55.2% in observer 1, 51.7% in observer 2) and one-layer wall thickening showing heterogeneous enhancement (37.9% in observer 1, 34.5% in observer 2) (Table 3), respectively. From these results, we observe that considerable portions of the flat-wall-thickening type of infiltrating gallbladder cancer manifest as gallbladder wall thickening with well-preserved layering. This is a new finding identified in our study and should be added as an imaging feature of gallbladder cancer to the existing literature that indicates a thickened gallbladder wall with disruption or obliteration of the normal layered pattern suggests carcinoma [12, 14]. Considering that recently we encountered patients with early gallbladder cancer presenting with focal or flat thickening of the gallbladder wall, the widespread use of MDCT and laparoscopic cholecystectomy, and that diagnosing a thickened gallbladder wall can be a problematic dilemma for the radiologist [15], our results of enhancement pattern analysis of the gallbladder wall are valuable in selecting treatment options and determining which surgical procedures are required [3, 16].
In our study, we evaluated the enhancement features of flat gallbladder wall thickening, which has often been considered benign inflammation. To our knowledge, at the time of writing this article, differentiation of gallbladder wall thickening focusing on layered enhancement patterns using MDCT images had not yet been described in the published literature. Our results regarding the enhancement patterns of thickened gallbladder wall are similar in some aspects to a previous report by Jung et al. [12] regarding the layered patterns of gallbladder wall on MRI. In that study, Jung and colleagues classified thickened gallbladder wall into four patterns on T2-weighted images on the basis of the presence of striation, the features of the inner and outer layers, and the signal intensity of each layer. In their results, acute and chronic cholecystitis showed two layers with an ill-defined or a discrete margin, whereas gallbladder cancer showed diffuse nodular wall thickening without layering. In addition, adenomyomatosis showed multiple cystic spaces of high signal intensity in the thickened wall. However, that study is different from ours in that the inclusion criteria consisted of thickened gallbladder wall of more than 3 mm irrespective of the extent of involvement or the gross type of gallbladder cancer [12]. Furthermore, given that MDCT and sonography are regarded in many clinical institutes as the first-line imaging techniques for the evaluation of abdominal pain and that MRI is a second-line imaging technique, we believe that our approach using analysis of MDCT enhancement patterns is more valuable in many clinical scenarios.
The results of our study also revealed several imaging features with a high
odds ratio for the diagnosis of gallbladder cancer. According to our results
using this odds ratio, the important findings suggesting a malignant cause of
gallbladder wall thickening are the following: first, enhancing thick inner
layer of the gallbladder wall 2.6 mm; second, stronger enhancement of the
inner layer than of liver parenchyma; third, thin outer layer 3.4 mm;
fourth, irregular contour of the affected wall; and, fifth, focal involvement,
in that order of highest odds ratio. These results agree with those of a
previous study regarding the enhancing features of gallbladder cancer on
biphasic helical CT [17]. Even
though the authors of that study included all morphologic types of gallbladder
cancer (i.e., mass-forming, wall thickening, and polypoid types), the most
common enhancement patterns of gallbladder cancer were a highly enhanced thick
inner wall layer seen during the arterial phase, but the most common pattern
of chronic cholecystitis was isoattenuation of the thin inner layer.
Even though we did not evaluate arterial phase images, based on a previous report regarding differentiating malignant from benign common bile duct strictures [18], we assumed that hyperenhancement of the involved gallbladder wall during the portal venous phase is the main factor distinguishing malignant from benign causes of gallbladder wall thickening. In addition, although we included only the flat-wall-thickening type of gallbladder cancer in our study, according to articles in the literature, associated mural irregularity or marked asymmetry is helpful for distinguishing complicated cholecystitis from gallbladder carcinoma [3, 19, 20].
The reported CT features of acute cholecystitis include gallstones, thickening of the gallbladder wall, pericholecystic fluid, stranding of pericholecystic fat, high-attenuation bile, blurring of the interface between the gallbladder and the liver, and transient increases in attenuation of the portion of the liver adjacent to the gallbladder [16, 21, 22]. However, the flat-wall-thickening type of gallbladder cancer is often misdiagnosed as inflammatory gallbladder disease when it is accompanied by acute cholecystitis [23]. In our study, we observed that acute cholecystitis frequently showed the type 5 enhancement pattern, which consists of a smooth weakly enhancing thin inner layer and a nonenhancing thick outer layer (Figs. 1E and 2E). The weakly enhancing thin inner layer is consistent with inflamed or sloughed mucosa, and the nonenhancing thick outer layer is compatible with an edematous loose connective tissue layer.
Adenomyomatosis usually presents as focal wall thickening (fundal or segmental type or both types) or as diffuse wall thickening of the gallbladder (diffuse type); therefore, it is important to differentiate adeno myomatosis from gallbladder carcinoma [24]. As the investigators of several studies have reported, cystlike spaces indicating Rokitansky-Aschoff sinuses in a thickened gallbladder wall on sonography or MRI are thought to indicate adenomyomatosis [24–27]. According to the results of a previous study by Ching et al. [15], although CT is limited in diagnosing adenomyomatosis, this diagnosis can be made with reasonable accuracy when the thickened gallbladder wall contains small cystic-appearing spaces. In our study, the degree of enhancement and the thickness of the enhancing wall were borderline between gallbladder cancer and cholecystitis, and small intramural cystic lesions with water density were observed in two of four cases of adenomyomatosis.
In addition to the intrinsic limits of any retrospective study, this study has several limitations. First, because our patient population was a subset of all the patients at our institution who underwent surgery for gallbladder wall thickening, the possibility of a selection bias must be considered. Second, we did not evaluate arterial phase images because in many of our study patients CT images were obtained during the portal venous phase. However, considering that the routine CT protocol for evaluation of acute or chronic right upper quadrant abdominal pain in many institutions is portal venous phase imaging, evaluation of portal venous phase images is clinically relevant. Third, although state-of-the-art MDCT scanners can produce coronal or sagittal reformatted images as well as thin axial images, the CT data used in the present study were acquired at section thicknesses ranging from 2.5 to 3.0 mm. Considering that the gallbladder is not exclusively perpendicular to the transverse plane, combined interpretation of axial and reformatted images would improve the diagnostic performance of CT for the evaluation of the extent of gallbladder wall involvement. Fourth, the variability in CT techniques of our CT scanners, including imaging delay from injection, may have influenced the degree of enhancement of the gallbladder wall and liver parenchyma or the measurement of the gallbladder wall layer. However, we attempted to use the same contrast dose based on patient body weight and fixed the injection duration (30 seconds) to minimize the influence of contrast injection regarding gallbladder wall enhancement. Finally, because our study population was composed of patients with diffuse gallbladder wall thickening who had undergone surgical resection of the gallbladder, there could have been a selection bias.
In conclusion, our study results show that certain MDCT findings—that
is, hyperenhancing thick inner wall 2.6 mm during the portal venous
phase, weakly enhancing or nonenhancing thin outer wall 3.4 mm, and
irregular and focal wall thickening—indicate a malignant cause of flat
gallbladder wall thickening rather than benign disease. In particular, the
results of this study show that the two-layer pattern of a hyperenhancing
thick inner layer with a weakly enhancing or nonenhancing thin outer layer is
the prevalent pattern of the flat-wall-thickening type of gallbladder
cancer.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. C. Kim, D. M. Yang, W. Jin, S. H. Joo, B. S. Kim, K. R. Joo, and S. J. Lim Large Fibrous Polyps of the Gallbladder Simulating Gallbladder Carcinoma J. Ultrasound Med., April 1, 2009; 28(4): 537 - 540. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
