|
|
||||||||
Original Report |
1 Department of Radiologic Pathology, Armed Forces Institute of Pathology, 6825
16th St. NW, Washington, DC 20306-6000.
2 Present address: Department of Radiology, Duke University Medical Center, Box
3808, Durham, NC 27710.
3 Department of Gastrointestinal Pathology, Armed Forces Institute of Pathology,
Washington, DC 20306-6000.
4 Department of Radiology and Nuclear Medicine, Uniformed Services University of
the Health Sciences, 4301 Jones Bridge Rd., Bethesda, MD 20814.
Received December 23, 2002;
revised April 4, 2003;
The opinions and assertions contained herein are the private views of the
authors and are not to be construed as official or as reflecting the views of
the Department of the Army or Department of Defense.
Abstract
|
|
|---|
CONCLUSION. Of the 13 patients who had upper gastrointestinal examinations, seven had findings of smooth submucosal masses with ulcerations or depressions. These findings overlap with those of a gastrointestinal stroma tumor and lymphoma. CT findings were specific for the diagnosis of lipoma in eight of nine patients. CT should be used to evaluate large submucosal masses in the stomach to establish a preoperative diagnosis.
|
|
|---|
Classically, imaging findings that suggest the diagnosis of a gastric lipoma are a smooth, sharply marginated, oval or spherical mass that is compressible during fluoroscopic examination and may exhibit significantly decreased attenuation on barium studies [2, 7]. At endoscopy, some of the lesions are even translucent enough to allow the correct diagnosis. In other lesions, ulceration produces a bull's eye appearance, making these lesions indistinguishable from other submucosal tumors. We believe that this ulcerated appearance has not been emphasized in the radiology literature. Most lipomas are solitary, but cases of multiple lipomas have been reported [8]. CT has proven to be of considerable value in the diagnosis of gastrointestinal lipomas [912]. The lesions appear as well-circumscribed areas of uniform, fatty density with an attenuation ranging from 70 to 120 H. Thus, a gastric lipoma can be definitively diagnosed using CT, which obviates endoscopy or even surgery if the patient is asymptomatic [11, 12].
To our knowledge, no large series focusing on gastric lipomas has yet been reported. We sought to review the clinical, pathologic, and imaging characteristics of 16 gastric lipomas. We paid particular attention to comparing the imaging characteristics that we observed with those that have previously been reported and believe that our study represents the largest group of gastric lipomas reported in the radiology literature.
|
|
|---|
Upper gastrointestinal examinations were available for review for 13 patients, and CT scans were available for review for nine patients. For six patients, both gastrointestinal examinations and CT scans were available. Because patients at the Armed Forces Institute of Pathology are referred from many institutions, the images were obtained using a variety of equipment and different protocols, so the acquisition techniques were not standardized. In addition, only a limited number of images from the upper gastrointestinal examinations were available.
Two radiologists reviewed the upper gastrointestinal examinations to evaluate the size, margins, and density of each lesion; presence of transparency; and evidence of ulceration. CT scans were also reviewed to evaluate tumor margins and to determine whether fatty density, increased density within the mass, or ulceration were present. For three patients, specific Hounsfield values of the lesions were available. The specific location of each lipoma was determined from the upper gastrointestinal examinations; CT scans; and, when necessary, endoscopic correlations.
|
|
|---|
Pathologic Findings
In all 16 patients, gastric lipomas were confirmed histologically, 15 at
the Armed Forces Institute of Pathology and one at an outside hospital. Gross
pathologic specimens were available for only 10 of the 16 patients with proven
lipomas, but a gross pathologic description was available for all 16. Eleven
lesions were located in the antrum, two in the fundus, two in the pylorus, and
one in the body of the stomach. The average size of the lipomas was 6.5 cm
(range, 3.59.0 cm) measured at the greatest dimension. All the lesions
had smooth margins, although some had central ulcerations. The tumors in eight
patients were ulcerated at the time of resection, and the tumors in two
patients had areas of depressions that had undergone necrosis but had not yet
developed into ulcers. These depressions were indistinguishable from ulcers
seen on the upper gastrointestinal examinations.
Imaging Findings
All 13 lesions available for review on the upper gastrointestinal
examinations were detected. Upper gastrointestinal examinations were performed
in six of the patients with ulcers and in the two with central depressions.
These examinations revealed five (83%) of the 13 ulcers (Fig.
1A,
1B,
1C) and both of the depressions
(Fig. 2). In one patient, the
ulcer was not detected; two patients who had ulcers did not have upper
gastrointestinal examinations. Thus, the upper gastrointestinal examinations
revealed seven (88%) of the eight ulcers.
|
|
|
|
One 8.5-cm lesion in the fundus of the stomach was hypodense on the upper gastrointestinal examination, a finding suggestive of a lipoma (Fig. 3). All the lesions had smooth margins (Figs. 1A, 1B, 1C, 2, 3, 4A, 4B, 5), but only this lesion exhibited the significantly decreased density that is consistent with a fat-containing tumor (Fig. 3). The other 12 lesions, particularly the eight with either ulceration or central depressions, mimicked nonfatty submucosal tumors such as a gastrointestinal stroma tumor or a lymphoma (Figs. 1A, 1B, 1C, 2, 3, 4A). Two patients with abdominal pain had pyloric lesions that had prolapsed into the duodenal bulb. Obviously, compressibility of the lesions could not be evaluated.
|
|
|
|
All nine CT scans showed a fatty density. The density of the lesions in three patients in whom density was measured was 70, 100, and 117 H, respectively (Figs. 1B, 4B, 5, and 6). One lesion had stranded soft-tissue attenuation coursing through it (Fig. 6) and a significant amount of low-density fat; an ulceration was present in the gross pathology specimen. Thus, CT findings were absolutely diagnostic in eight of nine patients (Figs. 1A, 1B, 1C and 4A, 4B, 5, 6) and highly suggestive of the correct diagnosis in the ninth patient (Fig. 7A, 7B). We found it interesting that CT revealed ulceration in one (17%) of six patients who had an ulcerated lipoma (Fig. 1C). The ulceration was also seen on the patient's upper gastrointestinal examination (Fig. 1A).
|
|
|
We observed considerable overlap in the size of the lipomas with ulceration and those without ulceration. The average size of the lipomas with ulceration was 7.6 cm and ranged from 4 to 9 cm; the average size of the lesions without ulceration was 6.1 cm and ranged from 3.5 to 9 cm.
|
|
|---|
The central depressions in the lipomas in two patients were the only unexpected findings noted in the pathology evaluation. The depressions were thought to be underlying necrosis of the lipomas with no apparent ulcerations. Ulceration might have eventually occurred if the lesions were not resected. These lesions were indistinguishable from ulceration on upper gastrointestinal examinations (Figs. 1A, 1B, 1C and 2).
Despite the relativity large size of the lipomas, only one in the fundus exhibited significantly decreased attenuation on the upper gastrointestinal examination that was suggestive of a lipoma (Fig. 3). All lipomas had a smooth surface, even the eight with ulcerations and the two with central depressions. On upper gastrointestinal examination, most of the lesions displayed characteristics of a nonfatty mural mass, suggesting that the lipomas arose from the gastric submucosa or muscular layers, which is where one might expect to find a gastrointestinal stroma tumor, lymphoma, carcinoid, or metastasis. The upper gastrointestinal examination was highly sensitive (88%) for the detection of ulcerations and depressions, which are typical findings for a carcinoid, lymphoma, metastasis, and occasionally gastrointestinal stroma tumor [2, 7]. Lipomas should also be included in this differential diagnosis.
As we discussed earlier, two of our patients had lesions that had prolapsed into the duodenum [2]. We also encountered two lesions that were exophytic, one of which was observed on an upper gastrointestinal examination to be causing significant external compression, a finding suggestive of a serosal rather than a submucosal lesion. CT confirmed the diagnosis of a lipoma in both patients.
CT is the imaging examination of choice for obtaining a specific diagnosis of lipoma. The smallest lesion revealed on CT was 4 cm and had an ulcer that was visible on both upper gastrointestinal examination and CT (Fig. 1A, 1B, 1C). A homogeneous gastric mass with a density of between 70 and 120 H has been previously reported as pathognomonic for the diagnosis of gastric lipoma [912]. On CT, gastric lipomas have also been reported to display linear strands of soft-tissue attenuation at the base as well as ulceration of the mucosa that correlated with prominent fibrovascular septa [2]. This finding is similar to that seen in our patient in Figure 7A, 7B. Taylor et al. [2] believed that the presence of these strands visualized on CT in an otherwise uniform, fatty tumor should be taken as a sign of benignity and cautioned against mistaking it for a liposarcoma, which is extremely rare in the alimentary tract. A gastric lipoma can usually be definitively diagnosed using CT, thereby obviating endoscopy or surgery in an asymptomatic patient [2, 11, 12].
If a large (> 2 cm) submucosal mass is detected on an endoscopic or upper gastrointestinal examination, a CT scan should be obtained because CT findings can allow one to make a specific diagnosis of lipoma. If CT does show the characteristic features of a lipoma, a biopsy is not needed [3, 6, 11, 12]. Biopsy is indicated in patients with lesions that are not totally fatty; surgery is indicated in symptomatic patients [1].
Use of MRI has been limited in diagnosing gastric lipomas [13], but MRI is extremely sensitive to fat and could be used instead of CT in certain patient populations, especially in children and perhaps in patients allergic to iodinated contrast agents.
Our retrospective study has limitations. It covered more than 30 years, and the cases studied were collected from multiple institutions. We had the basic clinical data and good copies of the upper gastrointestinal examinations and CT scans available for review. Microscopic slides for 15 of the 16 patients were available for review as were gross specimen photographs for 10 of the 16 patients with gastric lipomas. Therefore, we believe that we have good clinical, radiologic, and pathologic material to document our findings.
In conclusion, our review of data for 16 patients with proven gastric lipomas revealed two significant findings. First, on upper gastrointestinal examinations, lipomas have the appearance of any submucosal tumor; they are indistinguishable from gastrointestinal stroma tumors or lymphoma. Second, CT findings are specific for the diagnosis of gastric lipoma. CT should be used to characterize large submucosal masses before endoscopic biopsy is performed.
|
|
|---|
This article has been cited by other articles:
![]() |
W. M. Thompson Imaging and Findings of Lipomas of the Gastrointestinal Tract Am. J. Roentgenol., April 1, 2005; 184(4): 1163 - 1171. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |