| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 From the Department of Radiology, Samsung Medical Center, College of Medicine, Sungkyunkwan University, 50 Ilwon-dong, Kangnam-ku, Seoul 135-710, Korea (H.K.L., W.J.L., S.H.K.), and the Departments of Radiology (Y.H.K., J.K.H., B.I.C.) and Pathology (Y.I.K.), Seoul National University Hospital, Korea. From the 1997 RSNA scientific assembly. Received April 20, 1998; revision requested July 2; revision received August 19; accepted November 20. Address reprint requests to H.K.L.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: A retrospective review of upper GI examinations was performed in 25 patients with proved low-grade gastric MALT lymphomas. Upper GI examinations were reviewed for common findings and most probable diagnosis, and these findings were correlated with pathologic findings in resected specimens in 15 patients.
RESULTS: The common findings at upper GI examination included mucosal nodularity (n = 13), ulcer (n = 12), rugal thickening (n = 6), mass (n = 4), and enlarged areae gastricae (n = 2). The most probable diagnoses were early gastric carcinoma (n = 7), advanced gastric carcinoma (n = 6), gastritis (n = 9), and lymphoma (n = 3). Of 17 lesions found on resected specimens, six ulcers and two masses were not depicted at barium study. Disorganized convergent rugae projecting to multiple points and vague ulcer margins were present in four and seven lesions, respectively. Multiple ulcers were seen in two patients.
CONCLUSION: Although the common radiographic and pathologic findings observed in low-grade gastric MALT lymphomas were similar to those of gastric carcinomas or gastritis, disorganized convergent rugae, vague ulcer margins, and multiplicity of lesions may be helpful in differentiating them from gastric carcinomas or gastritis.
Index terms: Gastritis, 72.202, 72.291 • Lymphoma, 72.34 • Lymphoma, diagnosis, 72.123, 72.1261 • Peptic ulcer, 72.25 • Stomach, biopsy, 72.123, 72.1261
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
There have been some reports (14–16) describing the radiographic findings of pseudolymphoma and gastric MALT lymphoma (17–20). Suekane et al (18) stated that friable mucosa and ulcers simulating early gastric carcinoma were frequent findings in low-grade gastric MALT lymphoma. Yoo et al (20) reported that round nodularity was the most common finding at double-contrast barium study. To our knowledge, however, there has been no prior report describing the radiologic-pathologic correlation of this entity in the English-language radiology literature.
The purpose of this study was to evaluate the findings of double-contrast upper GI examinations in 25 patients with histopathologically proved low-grade gastric MALT lymphoma and to correlate them with pathologic findings.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
All patients underwent double-contrast upper GI examinations that were adequate for the analysis of the lesions. All upper GI examinations were performed within 2 weeks (mean, 5 days) of endoscopy. The upper GI examination was performed to evaluate the extent of the lesions in patients (n = 20) who were scheduled to undergo surgery. The surgeons at our institutions wanted the barium study before surgery to help them decide what type of surgery to perform. The barium study in the remaining five patients treated with antibiotics or chemotherapy was requested as a baseline study for follow-up. We used intramuscular injection of 20 mg of scopolamine butylbromide (Buscopan; Boehringer Ingelheim Korea, Seoul, Korea) as a hypotonic agent and oral administration of 200–250 mL of 140% wt/vol barium suspension (Solotop-HD; Taejoon Pharm, Kyungkido, Korea). Oral administration of an effervescent agent was used for adequate gastric distention. Under fluoroscopic observation, multiple spot radiographs were obtained after adequate mucosal coating.
The findings at upper GI examination were reviewed independently at each institution by four radiologists (Y.H.K., H.K.L., J.K.H., W.J.L.). This retrospective review was performed without knowledge of the endoscopic and pathologic findings but with knowledge of the final diagnosis of low-grade gastric MALT lymphoma. Final interpretations were made by consensus. The radiologic signs evaluated were the pattern of areae gastricae, presence of ulcer, mass formation, mucosal nodularity, and rugal thickening. The extent of the lesions was categorized as being in the antrum, body, or fundus.
Of 20 patients who underwent surgical resection, the pathologic specimens were available in 15 patients. The pathologic specimens were reviewed by one experienced pathologist (Y.I.K.) for the extent and morphologic features of the lesions. The depth of invasion and involvement of lymph nodes were also determined at histopathologic examination. Although detailed histopathologic mapping of the whole resected specimen was not performed, the pathologist tried to evaluate the histopathologic features of all visible lesions. We then correlated the findings of double-contrast upper GI examinations with those of pathologic examinations, and we reviewed the clinical indications for endoscopy and the clinical staging of the disease.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
On the basis of records of clinical, surgical, and histopathologic examinations, the Ann Arbor staging scheme was applied in all 25 patients. Of these, 23 patients had stage IE, and two had stage IIE.
Of 25 patients, five underwent administration of antibiotics for eradication of H pylori. Complete remission was defined as resolution of clinical evidence of disease and of endoscopic lesions, with negative biopsy specimen findings and absence of extragastric localization. Partial remission was defined as improvement of clinical symptoms and endoscopic lesions with a decrease in tumoral infiltration. Complete remission was attained in one patient, and partial remission was attained in four. Three of four patients with partial remission underwent surgical resection, and the remaining one underwent chemotherapy. Seventeen patients underwent curative surgical resection without preoperative medical therapy. Three patients were lost to follow-up after endoscopy and upper GI examination. Two patients received adjunctive radiation therapy, and one patient received adjunctive chemotherapy owing to the involvement of regional lymph nodes (n = 2) and resection margins (n = 1).
There were no surgical deaths in the current study, but two patients had early perioperative complications: leakage from gastrojejunostomy site (n = 1) and wound infection (n = 1). During follow-up, one patient experienced postoperative adhesive ileus, which was resolved with conservative treatment. The remaining 17 patients had no peri- or postoperative complications. All 25 patients are alive with disease in complete remission after follow-up (range, 6–70 months; mean, 38.1 months).
Findings at Upper GI Examination
All 25 patients with low-grade MALT lymphoma showed abnormal findings at upper GI examination. Of these patients, nine had lesions in the antrum, seven had lesions in the body, seven had lesions in the antrum and body, one had lesions in the fundus, and the remaining one had lesions in the body and fundus. Thus, 16 patients (64%) had involvement of the gastric antrum.
The findings observed at upper GI examination are summarized in Table 1. Mucosal nodularity, found in 13 patients (52%), was the most common finding (Fig 1). These nodules were distinct in margin and round or oval in shape. The nodules were variable in size (less than 1 cm in greatest diameter) and were located in the antrum (n = 6), antrum and body (n = 5), fundus (n = 1), and nearly the whole stomach (n = 1). Thus, 12 patients (92%) had nodularity in the gastric antrum.
|
|
|
|
|
|
|
|
|
|
|
|
Pathologic Findings and Radiographic-Pathologic Correlation
In the pathologic specimens of 15 patients, 17 discrete abnormalities were found. Three patients had two discrete lesions. One patient had no definite focal abnormality on the specimen. Nine lesions involved the antrum, six involved the body, and two involved the gastric angle. The correlation of the pathologic and radiographic findings in 15 patients is summarized in Table 2.
|
At histopathologic examination, all lesions invaded only the submucosal layer. One patient with no definite focal abnormality at pathologic examination demonstrated disseminated tumor cell infiltration along the mucosa and submucosa from antrum to fundus. Regional lymph node dissection was performed in all patients who underwent surgical resection. Two of 20 patients had involvement of regional lymph nodes. The resection margin was involved by tumor cells in two patients. All ulcers and focal masses demonstrated at double-contrast upper GI examinations and in pathologic specimens had tumor cells.
Mucosal nodularity, enlarged areae gastricae, and thickened rugae had involvement of MALT lymphoma in some areas and H pylori gastritis in other areas. It was almost impossible to distinguish MALT lymphoma from H pylori gastritis in each of these lesions demonstrated at double-contrast upper GI examination. H pylori gastritis of varying degrees coexisted with low-grade MALT lymphoma in all cases.
Of 17 lesions found at pathologic examination, only nine lesions in eight patients were detected at upper GI examination. Six shallow ulcers and two masses were not depicted at double-contrast upper GI examination. However, six shallow ulcers, two deep ulcers, and one focal mass seen at upper GI examination correlated well with the lesions seen in the resected specimens in terms of morphology.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The paradox of lymphomas arising in the stomach has been explained by the observation of MALT in the stomach in response to infections by H pylori and by the presence of this organism in more than 90% of gastric MALT lymphomas (22–25). In some cases, low-grade MALT lymphomas have regressed with eradication of H pylori (26). In our study, 23 (92%) of 25 patients had histopathologic evidence of H pylori infection. This finding correlates well with findings of previous reports (24–26) and supports the suggestion that H pylori infection acts as a predisposing factor in the development of gastric MALT lymphoma.
In our study, the mean age of the patients was 32 years, and a female predominance of 16:9 was noted. These results did not match with the findings of previous reports (5,6,8,9,13,17,18,27,28) of MALT lymphoma or pseudolymphomas. In previous reports (5,6,8,9,13,17,18,27,28), the mean age was 50-60 years, and an equal sex ratio or slight male predominance was noted. Although it is difficult to explain the exact cause of the younger age of patients and of female predominance our study, the small number of patients included or selection bias may contribute to these differences. Another possible explanation for the younger age would be in practice pattern. In our country, gastric carcinoma is a common malignancy, and gastroscopy is a routine procedure in patients with GI symptoms. Widespread and early use of endoscopy, even in young patients, may enable early diagnosis.
Low-grade MALT lymphoma manifests at an early stage and has a good prognosis (9,27–29). All 25 patients had Ann Arbor stage IE (n = 23) or IIE (n = 2) and remained alive during a mean follow-up of 38.1 months in our study. The stage of primary gastric lymphoma was one of the most important prognostic factors. Other prognostic factors are tumor location, tumor size, depth of invasion, grade of tumor, and involvement of regional lymph nodes (29–33). Manifestation at an early stage, the superficial nature of the lesions (confined to submucosa), and the infrequent involvement of regional lymph nodes indicate a good prognosis for the patients in our study.
The strategy for treatment of low-grade MALT lymphoma remains somewhat controversial. Some investigators (29,32,33) reported that adequate staging and long-term survival were attained with radical gastrectomy, with or without adjuvant treatment, in patients with low-grade MALT lymphoma. However, other nonsurgical treatments, such as chemotherapy or eradication of H pylori, also were recommended (26,34). In the patients in our study, radical gastrectomy, with or without adjuvant therapy, was performed in 20 patients. All these patients have been alive with disease in complete remission until now. Although direct comparison would be difficult because of the small number of patients who underwent nonsurgical treatment, surgical resection can be an appropriate treatment of low-grade MALT lymphoma, especially in the case of failure of antibiotics or in institutes that have skilled surgeons. The effects of nonsurgical treatment, especially eradication of H pylori, should be evaluated in further studies.
Gastric lymphomas are demonstrated on barium studies by a spectrum of findings, including thickened rugae, submucosal masses, centrally ulcerated (bull's-eye or target) lesions, polypoid lesions, nodules, and ulcers. There have been some reports in which radiographic findings of gastric MALT lymphomas have been described in the Japanese-language (17,18) and English-language literature (19,20). According to one Japanese report (18) in which 20 patients with low-grade gastric MALT lymphoma were analyzed, the largest group of patients with this disease, to our knowledge, friable mucosa and shallow ulcers mimicking early gastric carcinoma were frequent findings at upper GI examination. At endoscopy, the lesions affected the body and the antrum and consisted of an irregular and large ulcer with raised edges and erosions, small nodules, and localized or diffuse rugal thickening (28). Our study results are similar to those of the study by Suekane et al (18).
Perez and Dorfman (14) reported radiologic findings of pseudolymphoma, including tumor mass with ulceration, infiltrating constrictive lesions associated with ulceration, solitary well-defined ulcer craters, and enlarged gastric rugae. Other authors (15,16) described radiographic and pathologic findings of pseudolymphoma. They described radiologic findings such as ulcerating masses, submucosal masses, enlarged gastric rugae, and benign-appearing ulcers. At pathologic examination, ulcers were shown in all cases, and about half demonstrated multiple lesions. Considering that the majority of pseudolymphomas originated from MALT, our findings of low-grade MALT lymphoma are comparable to those of pseudolymphoma. However, multiple lesions were less frequent in our study (20%) than were pseudolymphomas.
Early gastric lymphoma was defined as lymphoma invasion limited to the mucosa or the submucosa of the stomach wall, regardless of the presence of lymph node metastases (35). These tumors usually manifested as localized smooth enlargement of the rugae, poorly defined shallow ulcers, or well-defined deep ulcers. The ulcers usually were associated with mucosal convergence, enlargement of the rugae, and smooth marginal elevation. Multiple lesions were noted in 40%–50% of cases (30,35). Kitamura et al (30) reported that five of 10 early gastric lymphomas had pathologic features of MALT lymphoma.
In our study, the depth of invasion of all tumors was limited to the submucosa, and only two patients had regional lymph node metastases. On the basis of depth of invasion, the MALT lymphomas in our study can be classified as early gastric lymphoma. The findings obtained from our analysis of the upper GI examinations and pathologic examinations are similar to those of early gastric lymphoma, which frequently exhibits enlarged rugae, either shallow or deep ulcers, and multiplicity of lesions.
In our study, low-grade MALT lymphomas were confused with gastric carcinoma (especially early gastric carcinoma), gastritis, and lymphoma at double-contrast upper GI examination and pathologic examination. Compared with the findings of previous reports of MALT lymphoma (28), pseudolymphoma (14–16), and early gastric lymphoma (35), our results are not surprising. Chiles and Platz (15) reported that pseudolymphomas and early gastric lymphomas were usually diagnosed as gastric carcinomas or benign ulcers. Differentiation from gastric carcinoma was the most frequent diagnostic problem (30,35). Carcinoma that originates in the mucosa produces mucosal destruction and desmoplastic changes in the submucosa. In gastric carcinoma of the ulcerative type, these pathologic changes usually manifest as well-defined, irregular, shallow ulcers with abrupt disruption of the normal mucosa and often manifest as a striking convergence of the rugae. In contrast, early lymphoma that arises from the lymphatic tissue in the deeper layer mainly produces submucosal tumor growth, and desmoplastic reaction is rare. Therefore, smooth enlarged rugae with slight convergence indicating the submucosal nature of early lymphoma and poorly defined ulceration indicating poor desmoplastic reaction were findings suggestive of early lymphoma (35).
Kitamura et al (30) reported a higher prevalence of multifocal lesions in early gastric lymphoma than in gastric carcinoma. In our study, four (29%) of 14 ulcers had disorganized convergent rugae projecting to multiple points, and seven (50%) had vague ulcer margins. Two patients (14%) had multiple ulcers. Our results and the findings of previous reports (30,35) suggest that disorganized convergent rugae, vague ulcer margins, and multiple lesions may be useful in differentiating low-grade gastric MALT lymphoma from gastric carcinoma, especially the ulcerative type.
In some reports (36–38), H pylori gastritis manifested as thickened, nodular rugae in the gastric antrum and enlarged areae gastricae. Mucosal nodularity was the most frequent finding at upper GI examination in our study, so low-grade MALT lymphoma can mimic H pylori gastritis. Although all low-grade gastric MALT lymphomas did not have ulcers, the presence of ulcers may play a useful role in the differentiation of the two diseases. However, low-grade gastric MALT lymphoma manifesting as nodularity or thickened rugae alone is difficult to differentiate from H pylori gastritis on the basis of findings at upper GI examination.
Although we could not perform detailed histopathologic mapping of the entire gastrectomy specimen, all abnormal areas demonstrated at upper GI examination had involvement of MALT lymphoma with a background of diffuse gastritis. Hoshida et al (29) reported that multiple superficial spreading (slightly elevated lesions or thickened rugae), restricted invasion within the submucosa, and manifestation at an early stage were characteristic features of low-grade MALT lymphoma. In other reports (9,30) of MALT lymphoma and early gastric lymphoma, superficial spreading lesions were noted as frequent findings. Considering the superficial spreading nature of low-grade MALT lymphoma, the mucosal nodularity, enlarged areae gastricae, and thickened rugae seen at upper GI examination in our study could be represented by infiltration of MALT lymphoma.
When upper GI examination findings were correlated with pathologic examination findings, six shallow ulcers and two focal elevated lesions were missed at upper GI examination. Of the missed shallow ulcers, four were interpreted as normal, and two were interpreted as mucosal nodularity. Two cases of missed focal elevated lesions were interpreted as rugal thickening and mucosal nodularity. The inadequate upper GI examination, shallow nature of the lesions, and nonspecific findings, such as mucosal nodularity and rugal thickening that are commonly seen in H pylori gastritis, may contribute to the low detection rate of main lesions at upper GI examination.
In conclusion, low-grade gastric MALT lymphomas manifested as mucosal nodularity, either shallow or deep ulcers, and rugal thickening at upper GI examination and showed similar findings to those of gastric carcinoma and H pylori gastritis. Disorganized convergent rugae, vague ulcer margins, and multiple lesions may be helpful in differentiating low-grade gastric MALT lymphomas from gastric carcinomas.
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, H.K.L., Y.H.K.; study concepts and design, H.K.L., J.K.H.; definition of intellectual content, H.K.L.; literature research, Y.H.K.; clinical studies, W.J.L., H.K.L., J.K.H., B.I.C.; data acquisition, S.H.K., Y.H.K.; data analysis, Y.H.K., Y.I.K.; manuscript preparation, Y.H.K.; manuscript editing, H.K.L.; manuscript review, B.I.C., H.K.L.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M.-S. Park, K. W. Kim, J.-S. Yu, C. Park, J. K. Kim, S.-W. Yoon, K.-H. Lee, Y. H. Ryu, H. Kim, M.-J. Kim, et al. Radiographic Findings of Primary B-Cell Lymphoma of the Stomach: Low-Grade Versus High-Grade Malignancy in Relation to the Mucosa-Associated Lymphoid Tissue Concept Am. J. Roentgenol., November 1, 2002; 179(5): 1297 - 1304. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Choi, H. K. Lim, S. J. Lee, J. H. Lim, S. H. Kim, W. J. Lee, J. H. Lee, Y.-H. Kim, P.-L. Rhee, J. J. Kim, et al. Gastric Mucosa-Associated Lymphoid Tissue Lymphoma: Helical CT Findings and Pathologic Correlation Am. J. Roentgenol., May 1, 2002; 178(5): 1117 - 1122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. An, J. K. Han, Y. H. Kim, A. Y. Kim, B. I. Choi, Y. A Kim, and C. W. Kim Gastric Mucosa-associated Lymphoid Tissue Lymphoma: Spectrum of Findings at Double-Contrast Gastrointestinal Examination with Pathologic Correlation RadioGraphics, November 1, 2001; 21(6): 1491 - 1502. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Levine Invited Commentary RadioGraphics, November 1, 2001; 21(6): 1502 - 1504. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
