| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Radiology, University of California, San Francisco, 505 Parnassus Ave., Box 0628, Room M372, San Francisco, CA 94143-0628.
Received June 30, 2006;
accepted after revision February 5, 2007.
Address correspondence to F. V. Coakley
(fergus.coakley{at}radiology.ucsf.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We retrospectively identified the cases of 36 patients with pathologically proven adenomyomatosis (n = 22) or gallbladder cancer (n = 14) who had undergone preoperative abdominal CT. Two reviewers independently evaluated the presence and nature of morphologic gallbladder abnormalities, including the presence of intramural diverticula (i.e., small cystic-appearing spaces within the gallbladder wall). The reviewers used a five-point scale (1, definitely absent; 5, definitely present) to rate the overall likelihood of the presence of adenomyomatosis and gallbladder cancer. Ratings were dichotomized such that a diagnosis was considered present at a rating of 4 or 5 and considered absent at lower ratings.
RESULTS. Reviewer 1 detected a morphologic gallbladder abnormality in 17 patients and correctly characterized the abnormality in 14 (82%) of the patients (eight with adenomyomatosis and six with gallbladder cancer). Reviewer 2 detected an abnormality in 18 patients and was correct for 13 (72%) of the patients (eight with adenomyomatosis and five with gallbladder cancer). In particular, reviewer 1 detected intramural diverticula in eight patients, and all had the pathologic diagnosis of adenomyomatosis, whereas reviewer 2 detected intramural diverticula in 11 patients, and eight (73%) had the pathologic diagnosis of adenomyomatosis.
CONCLUSION. CT is limited in the detection and differentiation of adenomyomatosis and gallbladder cancer, but the diagnosis of adenomyomatosis can be made with reasonable accuracy when thickening of the gallbladder wall is seen to contain small cystic-appearing spaces.
Keywords: adenomyomatosis • CT • gallbladder • gastrointestinal radiology
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The radiologic evidence of both adenomyomatosis and gallbladder cancer is focal or diffuse thickening of the gallbladder wall. Studies have addressed the imaging findings of these diseases individually with varying techniques, and the finding of cystlike spaces in a thickened gallbladder wall is said to indicate adenomyomatosis [7–9]. We are unaware, however, of any study of the utility of CT in differentiating the two conditions. The importance of differentiation is increasing because gallbladder wall thickening has become a common and often incidental finding on MDCT, which has greater spatial resolution than previous techniques. In our clinical experience, gallbladder wall thickening can be a dilemma for interpreting radiologists. We undertook this study to determine the accuracy of CT in differentiating adenomyomatosis from gallbladder cancer.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT Technique
Helical CT scans were obtained with an MDCT (LightSpeed, GE Healthcare)
(n = 26) or a single-detector (n = 8) scanner (HiSpeed, GE
Healthcare) at slice thicknesses of 3 mm (n =2), 5 mm (n
=26), and 7 mm (n = 8). Two of the CT examinations were performed at
outside hospitals, and the type of scanner was not documented. All patients
received 150 mL of IV iohexol (Omnipaque 350, GE Healthcare) 70–80
seconds before CT. Thirty-five of 36 patients received 800 mL of oral
diatrizoate meglumine (Hypaque, GE Healthcare).
Image Interpretation
Two attending abdominal imaging radiologists who had undergone fellowship
training in body imaging and had 10 and 4 years of faculty experience
independently reviewed all CT scans without knowledge of the clinical or
histopathologic findings. The reviewers were aware of the study design and
that patients had a diagnosis of either adenomyomatosis or gallbladder cancer.
Images were reviewed on a PACS workstation (Impax, Agfa). The reviewers
recorded the presence or absence of focal or diffuse gallbladder wall
thickening as absent, equivocal, or present. Given the inherent inaccuracy of
measuring small distances on CT and given that gallbladder wall thickening is
partially a subjective finding that incorporates the degree of gallbladder
distention, we did not use quantitative criteria for gallbladder wall
thickening but relied on the expert judgment of the reviewers. In general,
however, a gallbladder wall thicker than 3–5 mm was considered abnormal.
When recorded as present, gallbladder wall thickening was characterized as
solid, diverticular (defined as small cystic-appearing spaces within the
thickened wall that were visually of fluid or near-fluid density) or as being
of other or indeterminate appearance. Finally, on the basis of their overall
evaluation, the reviewers rated the overall likelihood of the presence of
adenomyomatosis and gallbladder cancer on the following five-point scale: 1,
definitely absent; 2, probably absent; 3, indeterminate; 4, probably present;
5, definitely present. Assessment of these findings and the use of the
five-point scale were not based on quantitative criteria but were left to the
expert judgment of the reviewers.
Statistical Analysis
Ratings for the overall likelihood of adenomyomatosis and gallbladder
cancer were dichotomized in such a way that a diagnosis was considered present
at a rating of 4 or 5 and absent at lower ratings. Histopathologic findings
were used as the standard of reference, although to elucidate any potential
systematic causes of misinterpretation, the lead investigator closely reviewed
CT findings and histopathologic reports in cases of incorrect reviewer
diagnosis. Descriptive statistics for the diagnosis of adenomyomatosis and
gallbladder cancer were calculated for each reviewer on the basis of this
dichotomization. The influence of tumor size on CT detection of gallbladder
cancer was examined by comparing the mean size of detected tumors with that of
undetected tumors. Diagnostic accuracy was calculated as the percentage of
correct diagnoses by each reviewer when a convincing morphologic abnormality
was seen in the gallbladder wall (i.e., the overall likelihood of either
adenomyomatosis or gallbladder cancer was rated 4 or 5). Interobserver
agreement was assessed by kappa value (calculated on the basis of the
dichotomized ratings) and was classified as follows: poor, 0–0.20; fair,
0.21–0.40; moderate, 0.41–0.60; good, 0.61–0.80; and
excellent, 0.81–1.00
[10].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
|
With respect to specific morphologic findings, reviewer 1 detected
intramural diverticula in eight patients, and all had the pathologic finding
of adenomyomatosis. Reviewer 2 detected intramural diverticula in 11 patients,
and eight (73%) had the pathologic finding of adenomyomatosis. Review of the
images of three patients with intramural diverticula seen by reviewer 2 in
whom histopathologic examination showed only gallbladder cancer showed that
reviewer 2 correctly classified one of the patients as having gallbladder
cancer on the basis of overall impression. The other two patients had
undetected cancer (Figs. 5 and
7). Reviewers had good
agreement for the diagnosis of adenomyomatosis (
= 0.73) and moderate
agreement for the diagnosis of gallbladder cancer ( =0.42).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our results suggest that if gallbladder wall thickening is seen, CT is reasonably accurate in the differentiation of adenomyomatosis from gallbladder cancer. Reviewers 1 and 2 correctly characterized such abnormalities in 82% and 72% of cases, respectively. In particular, the finding of intramural diverticula on CT seems indicative of adenomyomatosis, even in a population that includes patients with gallbladder cancer. Eight of eight patients with intramural diverticula seen by reviewer 1 had adenomyomatosis, and eight of 11 patients with intramural diverticula seen by reviewer 2 had adenomyomatosis. Review of the cases of three patients with intramural diverticula seen by reviewer 2 but in whom the histopathologic finding was gallbladder cancer showed one correct classification of cancer on the basis of overall impression. The other two patients had undetected cancer. The utility of the finding of intramural diverticula on CT in the diagnosis adenomyomatosis can be likened to the socalled pearl necklace sign on MRI [9]. Our study dealt only with CT, but results of a study [7] involving 20 patients with proven adenomyomatosis suggested MRI was more accurate in the diagnosis. This conclusion seems intuitively reasonable given the excellent depiction of small fluid-filled structures, such as intramural diverticula, on images obtained with T2-weighted sequences.
The two cases of undetected cancer are particularly instructive with respect to lessons learned from our study. In the patient with undetected cancer shown in Figure 5, the finding of intramural diverticula is highly questionable. Maintaining a high threshold for making this finding should help prevent the critical error of mistaking a malignant tumor for a benign entity. In the patient shown in Figure 7, there appeared to be unquestionable cystic thickening in the gallbladder fundus, yet the pathologic report mentioned only a 5-mm malignant tumor in a region of the gallbladder remote from the radiologic finding. Our supposition is that this patient probably had adenomyomatosis but that this finding was not recorded at histopathologic review given the overriding importance of the associated malignant lesion. This scenario highlights the importance of judicious correlation of histopathologic and radiologic results, particularly for findings such as adenomyomatosis that histopathologists may not believe merit formal reporting. The situation also highlights a pitfall of retrospective studies. Irrespective of the presumptive explanation for the apparent misdiagnosis by reviewer 2 in this case, our data indicate that the presence of intramural diverticula favors a diagnosis of adenomyomatosis. However, neither the finding of intramural diverticula nor that of a normal gallbladder excludes the presence of gallbladder cancer, particularly small or infiltrative malignant tumors.
From a practical viewpoint, it seems reasonable to report the presence of intramural diverticula on CT as favoring the diagnosis of adenomyomatosis. It also may be reasonable and prudent to suggest follow-up sonography when gallbladder wall thickening is seen on CT, even if intramural diverticula are present, if for no other reason than that any subsequent diagnosis of gallbladder cancer can lead to a potential malpractice claim of misdiagnosis. Another lesson we learned from this study is that the lesions of adenomyomatosis can appear solid and resemble malignant disease (Figs. 3 and 6). This misinterpretation is clearly less dangerous to the patient than missing a cancer diagnosis.
A number of authors [7–9, 11] have found limitations of using CT to diagnose adenomyomatosis. CT findings suggesting gallbladder carcinoma include a focal soft-tissue mass replacing the gallbladder lumen, biliary obstruction at the level of the porta hepatis with ductal dilatation, lymphadenopathy, and direct hepatic invasion. Irregular or smooth thickening of the gallbladder wall, intraluminal masses, and gallstones can be seen on CT in both adenomyomatosis and carcinoma of the gallbladder [12], but our findings support the notion that identification of intramural diverticula can help to differentiate the two. Although not specifically addressed in our study, other causes of focal and diffuse gallbladder wall thickening should be remembered when this finding is visualized in practice. These causes include polyps of varying types, cholecystitis, hepatitis, cirrhosis, and ascites [13, 14].
Our study had several limitations. We assessed only cases of histopathologically proven adenomyomatosis documented in the pathology report. Adenomyomatosis was likely underreported in the pathology reports because this condition is considered a benign finding. Our study therefore might have been biased toward more severe cases. In addition, we did not assess whether our recorded CT findings were of value in differentiating adenomyomatosis and gallbladder carcinoma from other pathologic conditions of the gallbladder, such as cholecystitis. However, benign and malignant diseases of the gallbladder commonly are differentiated at daily CT readouts for patients without suspected acute gallbladder inflammation, and our study was aimed at this common clinical scenario.
We performed only limited feature analysis of findings we considered likely to be useful in differentiating adenomyomatosis from gallbladder cancer. For example, we did not specifically study the enhancement characteristics of both entities for potential differences, although our impression is that the solid-tissue elements of both conditions become enhanced in relatively similar ways (Figs. 1 and 2). The gallbladder was not routinely subdivided for purposes of imaging and pathologic interpretation, and we generally assumed that an imaging abnormality corresponded to the reported pathologic diagnosis. Despite that assumption, in at least one case an imaging finding suggestive of adenomyomatosis in the gallbladder fundus did not match the pathologic report of a small malignant tumor in the gallbladder neck. We believe the CT finding represented unreported adenomyomatosis (Fig. 7). We cannot be certain that such errors did not occur in other patients, and it is difficult to confirm or refute this possibility in a retrospective study.
Our study included only patients with adenomyomatosis and those with gallbladder cancer, and the two reviewers were aware of this study design. As such, the reviewers were sensitized to considering these diagnoses and may have excluded other diagnoses from consideration. Our results therefore may be overestimates of the accuracy of reviewers in daily practice. The interval between imaging and surgery was variable, and it is possible that there might have been changes in gallbladder pathologic findings in this period. Adenomyomatosis, however, is a chronic condition unlikely to change with great rapidity over time, and among the gallbladder cancer patients, the interval during which changes might well have occurred was generally short (mean, 20 days; range, 1–84 days).
Our study was performed over an 8-year period. Accordingly, the CT technique was heterogeneous and included both single-detector and MDCT scanners. We did not have enough patients to analyze whether continuing improvements in CT image quality over the study period might have improved diagnostic accuracy. The relatively low sensitivity we found for gallbladder cancer also could have been a function of older technology at the start of the study, although selection bias also is likely a factor, many of the malignant tumors being in the lower stages. Our sample size was small. Further assessment with larger case series may be helpful in determining ability to generalize our findings to a broader population.
Finally, reviewers used their expert judgment to evaluate specific criteria such as gallbladder wall thickening and to determine the overall likelihood of the presence of gallbladder cancer or adenomyomatosis. Gallbladder wall thickening is arguably a subjective observation that can be influenced by the degree of gallbladder distention. Although we did not record it, the degree of distention was presumably one of the factors used by reviewers to decide whether wall thickening was present. Likewise, with respect to the overall diagnostic impression, the specific factors that influenced each reviewer in arriving at a given diagnosis are ultimately unknown. The high kappa values we documented for interobserver agreement suggest the two reviewers were using similar standards.
We conclude that CT is limited in the detection and differentiation of adenomyomatosis and gallbladder cancer but that the diagnosis of adenomyomatosis can be made with reasonable accuracy when thickening of the gallbladder wall is seen to contain small cystic-appearing spaces.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. J. Kim, J. M. Lee, J. Y. Lee, S. H. Kim, J. K. Han, B. I. Choi, and J. Y. Choi Analysis of Enhancement Pattern of Flat Gallbladder Wall Thickening on MDCT to Differentiate Gallbladder Cancer from Cholecystitis Am. J. Roentgenol., September 1, 2008; 191(3): 765 - 771. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
