| Japanese Journal of Clinical Oncology | Pages |
Primary Gastric T-cell Lymphomas: Report of Two Cases and a Review of the Literature
Introduction
Materials And Methods
Case Report
Literature Review
Case Reports
Clinical Summary
Pathological Findings
Review of the Cases
Discussion
Acknowledgments
References
Primary Gastric T-cell Lymphomas: Report of Two Cases and a Review of the Literature
INTRODUCTION
The digestive tract is one of the organs that is most frequently affected by extranodal lymphomas (1). Determination of the cellular origin of the clonally expanding population in the lesion is now possible by immunological and molecular techniques that characterize surface antigens and detect rearrangements of immunoglobulin genes or those of T-cell receptors, respectively. In Japan, about half of the lymphoproliferative disorders have been shown to be T-cell in origin (2). In contrast, most primary lymphomas of the digestive tract are of B-cell origin (5-8) and those of the T-cell phenotype consist of only a few per cent and only a limited number of clinical reports are available. Previously, most of the primary T-cell lymphomas of the digestive tract (PGITL) were thought to have originated in the small intestine and primary gastric T-cell lymphoma (PGTL) was considered an extremely rare disease (2-4). Consequently, their clinicopathological features have not been well characterized.
Adult T-cell leukemia (ATL) is caused by HTLV-1 infection and is characterized by unique clinical features and a very poor prognosis (9,10). It frequently involves the gastrointestinal tract including the stomach and small intestine (11). Case reports of PGTL and primary small intestine T-cell lymphoma (PSITL) have included those cases positive for serum HTLV-1 antibody or having evidence of proliferation of HTLV-1-infected T-cells. The latter should be diagnosed as lymphoma type ATL based on its definition (12). In addition, the majority of reported cases of PGTL are from Japan, where HTLV-1 is endemic and about 700 cases of ATL are diagnosed every year (13). Therefore, it is very important to elucidate the clinicopathological characteristics of gastrointestinal T-cell lymphomas, including PGTL and PSITL, discriminating them from ATL affecting the gastrointestinal tract.
MATERIALS AND METHODS
Case Report
Primary gastrointestinal lymphomas were defined according to the definition by Dawson et al. (14). T-cell origin of the lymphoma cells was determined by immunohistochemical analysis of formalin-fixed or frozen sections of the tumors that were obtained by surgical resection. Two patients were included in this case report who were admitted to the hospital of the University of Tokyo (case 1) and to the hospital of the Institute of Medical Science, University of Tokyo (case 2). Specimens obtained by gastrectomy were fixed in 10% formaldehyde and embedded in paraffin or quickly frozen in dry ice-acetone. Sections 2-4 µm thick were prepared and used for hematoxylin-eosin (HE) staining or immunohistochemistry. The sections were stained using the avidin-biotin-peroxidase complex (ABC) method. Antibodies used for immunohistochemical studies of frozen sections were: Leu-1 (CD5), Leu-2a (CD8), Leu-3a (CD4), Leu-4 (CD3), Leu-7 (CD57), Leu-14 (CD22), IL-2R (CD25), HLA-DR (Becton Dickinson, Mountain View, CA), IgM, IgD, Ig[kappa], Ig[lambda] (DAKO, Glostrup, Denmark) for case 1 and Leu-2a, Leu-3a, Leu-4 and L26 (CD20) (DAKO) for case 2. Immunostaining of the paraffin sections using CD3 (DAKO) was also performed for both cases. Molecular analyses of these cases were not carried out because of the poor quality of DNA samples obtained.
Literature Review
We surveyed the literature on PGTL, diagnosed by immunophenotyping and/or molecular techniques, that provided clinical information supporting the diagnosis of primary gastrointestinal lymphoma, permitting evaluation of the prognosis. Consequently, some case reports were excluded from our summary of the information (described in the Review of the Cases section). We also reviewed the literature on PSITL, selected similarly to PGTL, and compared their features with those of PGTL in order to reveal clinicopathological characteristics of the latter. Statistical analysis of the survival of the patients with PGTL or PSITL was carried out using the Kaplan-Meier method (15).
CASE REPORTS
Clinical Summary
Case 1A submucosal mass in the mid-portion of the gastric body was found in a 52-year-old man in the course of a routine health examination by a barium meal study of the upper gastrointestinal tract. He had no clinical symptoms. Endoscopic examination and a histological study of the biopsied specimen revealed a malignant lymphoma of diffuse pleomorphic type. No lymphadenopathy was found on physical examination or chest X- ray film. Examination of the abdomen by ultrasonography and CT scan revealed regional lymphadenopathy. Laboratory examination including bone marrow aspiration was normal. Serum antibody for HTLV-1 was negative by the particle agglutination (PA) method (Fuji Rebio, Tokyo, Japan). After surveying systemic involvements by CT scan, gallium scintigram and lymphangiography, he was diagnosed as having a primary gastric lymphoma of stage IIE according to the Ann Arbor classification (16). Total gastrectomy with splenectomy and lymph node dissection was carried out, followed by one cycle of VEPA (vincristine, cyclophosphamide, prednisolone, adriamycin) therapy (17). He was readmitted to the hospital 1 month later because of fever and an abdominal mass. A mass of 17 × 14 cm was palpable in the left upper quadrant. He was treated with one cycle of ABEP (aclarubicin hydrochloride, behenoyl Ara-C, etoposide, prednisolone) therapy (18). Although the mass reduced in size in response to the therapy, he died of peritonitis secondary to perforation of the colon 6 months after the initial diagnosis.
Case 2A 75-year-old man complained of epigastric pain and 10 kg body weight loss over a year. Barium meal studies and endoscopic examination of the upper gastrointestinal tract revealed a lesion in the lesser curvature of the gastric angle that had an appearance of Borrmann III type gastric cancer. Histological examination of the biopsied samples obtained by endoscopy revealed an infiltration of lymphoma cells. No lymphadenopathy was found by physical examination or chest X-ray film. Examination of the abdomen by ultrasonography and CT scan revealed regional lymphadenopathy. Laboratory examination of peripheral blood was normal. Serum antibody for HTLV-1 was negative by the PA method (Fuji Rebio). He was diagnosed as having a primary gastric lymphoma of stage IIE according to the Ann Arbor classification (16) and subjected to total gastrectomy. He has been followed for more than 5 years without chemotherapy or any evidence for recurrence of the disease.
Pathological Findings
Case 1Macroscopically, the lesion showed an appearance similar to Borrmann I type gastric cancer located in the locus B/C and had a size of 3.5 × 3.3, 6.2 × 5.3 cm (Fig. 1a). Microscopic examination revealed an involvement of the regional lymph nodes, with six positives out of 44 nodes examined. Microscopic examination of wedge biopsy samples of the stomach, liver and spleen showed no evidence for an involvement by the lymphoma cells. Histologically, the primary gastric lesion was a diffuse pleomorphic-type lymphoma and an infiltration into the submucosa was found. Neoplastic cells were small, medium-sized lymphocytes with prominent and convoluted nuclei. Eosinophils were conspicuous among neoplastic cells and in the marginal region (Fig. 1b and 1c). Immunohistochemical studies of the resected gastric specimen revealed that the lymphoma cells were positive for Leu-4 (CD3) and Leu-3a (CD4) and negative for Leu-1 (CD5), Leu-2a (CD8), Leu-7 (CD57), Leu-14 (CD22), HLA-DR, Ig M, Ig D, Ig[kappa] and Ig[lambda], indicating that the patient had a T-cell lymphoma with a surface phenotype of helper/inducer T-cells. Staining of formalin-fixed, paraffin-embedded tissue sections by CD3 is shown in Fig. 1d. Autopsy revealed a disseminated infiltration of lymphoma cells into the small intestine, liver, colon, diaphragm, bone marrow and Douglas pouch. A perforation 1.5 cm in diameter in the ileum was found 140 cm from the ileum end. Peritonitis was found in the pelvic cavity associated with 2.5L purulent ascites and lymphoma cell infiltration.
Figure 1. Pathological findings of the resected stomach of case 1. (a) Gross appearance of the stomach showing two Borrmann I type gastric cancer-like lesions (arrows). (b) Low-power view of gastric tumor shows diffuse type lymphoma, infiltrating into the submucosa (hematoxylin-eosin). (c) Pleomorphic morphology of neoplastic cells with eosinophil leukocyte infiltration (hematoxylin-eosin). (d) CD3 immuno-staining revealing positive reactivity on neoplastic cells. Case 2 Macroscopically, mucosa of the whole resected stomach (17.5 × 11cm) was infiltrated by lymphoma cells with an appearance of a Borrmann IV type gastric cancer (Fig. 2a). Ulceration with a slightly elevated margin was found at the center of the lesion which appeared like a Borrmann III type gastric cancer by a barium meal study. Lymphoma cells infiltrated into the submucosa in most areas examined, along with partial infiltration into the serosa. Involvement of the regional lymph nodes was noted by microscopic examination (29 positives out of 45 nodes examined) (Fig. 2b and 2c). Immunohistochemical studies revealed that the lymphoma cells were strongly positive for Leu-4 (CD3) and partially for Leu-3a (CD4) and negative for Leu-2a (CD8) or L26 (CD20), indicating that the patient had a T-cell lymphoma with a surface phenotype of CD3 positive T-cells. Staining of formalin-fixed, paraffin-embedded tissue sections by CD3 is shown in Fig. 2d. Figure 2. Pathological findings of the resected stomach of case 2. (a) Ulceration with a slightly elevated margin was found at the center of the lesion showing Borrmann IV type gastric cancer in the lesser curvature of the corpus (the hole in the ulcer bed was an artifact). (b) Low-power view of the gastric tumor shows infiltration of the neoplastic cells into the sub mucosa. (c) Neoplastic cells show mixed small and large cell morphology. (d) CD3 immuno-staining revealing positive reactivity on neoplastic cells. Including the two cases presented here, clinicopathological features of 23 cases of PGTL were reviewed (19-30). Those reports without detailed clinical data were not included here (31-36). The serological status for HTLV-1 antibody was available in 17 cases, of which seven were seronegative for HTLV-1 and hereafter referred to as PGTL without HTLV-1 (cases 6-12 in Table 1). Among the 10 seropositive cases, those who had other evidence of HTLV-1 infection, such as detection of the provirus DNA by polymerase chain reaction (PCR) or Southern blot analysis, were included in PGTL with HTLV-1, excluding case 13 because of the lack of other evidence (21). On the other hand, case 15 was included in this group despite the absence of serological information, because HTLV-1 provirus integration was demonstrated in the lymphoma cells by in situ hybridization (26). These 10 cases will be referred to as PGTL with HTLV-1. All the HTLV-1-infected cases were found in Japan. No information as to HTLV-1 infection was available in the rest of the cases (cases 1-5). A summary of the clinical information of the PGTL cases reviewed is presented in Table 1. Table 1. Summary of clinical information on primary gastric T-cell lymphomas We also reviewed the literature on gastrointestinal T-cell lymphomas (PGITL) other than PGTL. We found reports on 21 cases with T-cell lymphoma of the small intestine (37-49), three of the colon (50) and three of multicentric origin (51, 52). We could not find reports on esophageal T-cell lymphoma. These results indicated that the PGITL other than PGTL is most frequently found in the small intestine (PSITL). No information as to HTLV-1 infection was available in most patients with PSITL. Negative results of the serological test for HTLV-1 antibody were documented in a few cases. Thus, no PSITL cases have been reported that have definitive association with HTLV-1 infection. The median age of patients with PGTL was 58 years (range 18-75 years). The male (M)-to-female (F) ratio was 2.3:1. The median age of PGTL with HTLV-1 was younger (44, range 31-70 years) than that of PGTL without HTLV-1 (58, range 46-75 years). Equal numbers of male and female patients were reported in PGTL with HTLV-1 (M:F = 5:5), whereas all the patients with PGTL without HTLV-1 were male. PSITL cases showed a similar median age and gender ratio: median age 58, range 29-75 years; M:F = 2.1:1. Cardinal symptoms at presentation in PGTL were upper abdominal pain and/or discomfort (76%) and body weight loss (29%), whereas those in PSITL were body weight loss (61%), diarrhea (38%) and abdominal pain (38%). The primary site of PGTL was described in 18 cases. Among them, that of 14 PGTL was located within the corpus to antrum, whereas four cases had lesions in the fundus or areas including the fundus. Gastric lesions showed various gross appearances; however, ulceration was most frequently observed and polypoid or complex ulceration and polypoid lesions were occasionally found. The median size of the gastric lesion was mostly 5-10 cm, using the largest single dimension. No clear differences in the gross appearance of the lesion were found between PGTLs with or without HTLV-1. As for PSITL, the jejunum was the most frequently observed primary site (71%). Intestinal lesions of PSITL also showed various gross appearances. Multiple and diffuse infiltration were often documented in PSITL, which appears to be different from that of PGTL. Involvement of the lymphoma was limited to the primary site in three of 23 PGTL cases and that of only abdominal lymph nodes was documented in 14 of these, which was described as that of local (surrounding, regional) or perigastric or epigastric lymph nodes. Four had involvement of other organs and one had massive ascites. A systemic involvement of lymph nodes was found in one case. These findings indicated that in most patients PGTL presented with involvement of primary site and abdominal lymph nodes (61%). No apparent differences were discerned between PGTLs with and without HTLV-1. Most reports did not refer to the criteria by which histological classification was made. Therefore, the original description of the histological classification is listed in Table 1. The cellular morphology of the diffusely infiltrating neoplastic cells was variously documented, such as medium, mixed, large, immunoblastic, anaplastic large and pleomorphic cell type. Among seven cases of PGTL without HTLV-1, neoplastic cells were classified as large cell type in four, mixed cell type in two and pleomorphic cell type in one. In the 10 cases of PGTL with HTLV-1, four were with pleomorphic cell components, two had immunoblastic cell type, one medium cell type, one small cleaved cell type, one large cell type and one anaplastic large cell type. Taken together, the lymphoma cells were mainly of the large cell type in PGTL without HTLV-1 and of pleomorphic cell type in those with HTLV-1. However, no consistent characteristics of cellular morphology could be revealed for either group of PGTL.
Review of the Cases
A surface phenotype corresponding to helper/inducer T-cells (CD4+8-) was most commonly observed in PGTL (16 of 19 cases). CD4-8+ phenotype was found in one and double-negative cells (CD4-8-) in two. All eight cases of PGTL with HTLV-1 and three of five cases without HTLV-1 had CD4+8- surface phenotype. The double negative (CD4-8-) phenotype was found in the rest of the cases of PGTL without HTLV-1 (two of five). In contrast, these three surface phenotypes were almost equally reported in 19 evaluable cases of PSITL as follows: CD4+8- in seven, CD4-8+ in four, CD4-8- in seven and CD4+8+ in one.
TreatmentTreatments were described in 19 PGTL patients. Eleven of them were treated by gastrectomy followed by chemotherapy, three by gastrectomy alone, two by chemotherapy followed by gastrectomy and three by chemotherapy alone. In the seven PGTL patients with HTLV-1, four were treated by gastrectomy followed by chemotherapy, two by chemotherapy alone and one by chemotherapy followed by gastrectomy. In the six PGTL cases without HTLV-1, two were treated by surgical resection alone, three by surgical resection followed by chemotherapy and one by chemotherapy followed by surgical resection. Similarly to the PGTL cases, surgical resection of the small intestine was the main modality of initial treatment for most PSITL patients. Treatments for all 21 PSITL patients were described in the reviewed literature. Eleven of these were treated by surgical resection alone, four by surgical resection and chemotherapy, three by chemotherapy alone and three by surgery and/or chemotherapy plus radiation.
The protocols of the chemotherapy were described in about half of the PGTL cases, most of which included all or combinations of the following drugs: cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate and prednisolone. No clear differences were apparent as to the chemotherapeutic protocols used between PGTL and PSITL.
PrognosisThe prognosis of 21 PGTL patients was evaluable based on the information in the literature. Among them, 12 patients were alive 6-66 months after the diagnosis. In contrast, most patients with PSITL died within 1 year after diagnosis (1 year survival: 73.0% in PGTL versus 11.1% in PSITL). At the time of the case reports, two of nine and five of six PGTL patients with and without HTLV-1, respectively, were alive. Survival curves for the patients with PGTL and those with PSITL were calculated by the Kaplan-Meier method based on information provided in the literature (Fig. 3). Despite the diversity of the treatment protocols used, the results suggested a significantly poorer prognosis for PSITL than PGTL (Cox-Mantel test : P < 0.01). When the survival of PGTL patients was analyzed in those with or without HTLV-1, HTLV-1-related cases clearly showed a poorer prognosis than those without HTLV-1, although the difference was not statistically significant (P = 1.07) mainly because of the short observation period. (Fig. 3).
Figure 3. Survival curve for primary gastric T-cell lymphoma (n = 21), with HTLV-1 (n = 9), without HTLV-1 (n = 6) and primary small intestinal T-cell lymphoma (n = 18). We have reported two cases of PGTL without HTLV-1 and reviewed the literature of primary gastrointestinal T-cell lymphomas (PGITL) with emphasis on that of the stomach (PGTL). The digestive tract is one of the most frequent sites involved in extranodal non-Hodgkin's lymphomas; however, most of them are of B-cell origin. T-cell lymphomas of the gastrointestinal tract are rare and most of the reported cases originated in the small intestine (6-8). Thus, only a limited number of cases with PGTL have been reported. As such, their clinicopathological features remained to be characterized. Despite the small number of reported PGTL cases, our review of the clinicopathological data provided some insights into the characteristics of this rare lymphoma. Almost two thirds of the reported PGTL cases (10 of 17) were infected by HTLV-1. Cardinal presenting symptoms of PGTL were upper abdominal pain and/or discomfort and body weight loss. Diarrhea was a rare symptom in PGTL, whereas it was one of the main presenting symptoms of PSITL. The review of the literature did not reveal any symptoms specific to PGITL. Diagnoses of PGTL were usually made by endoscopic biopsy, which can be performed non-invasively. However, that of PSITL was more difficult and, occasionally, the initial presentation was perforation of the intestine. Diagnostic surgery was required in some of the PSITL cases. A previous history of celiac disease may help to reach a correct diagnosis. It was shown that PGITL occurred in middle to old age, whereas HTLV-1-associated PGTL had a younger median age of onset. Despite the small number of patients analyzed, it was suggested that males predominated in PGTL without HTLV-1 and also in PSITL, but not in PGTL with HTLV-1. Laboratory examinations of peripheral blood were not informative in most patients. Slight increases in white blood cell counts were reported in some cases. In some PGTLs with HTLV-1, elevation of LDH was found and atypical lymphocytes in peripheral blood and skin infiltration were observed in the course of the disease, as was found in the lymphoma type ATL. The primary site in most PGTL was in the region from corpus to antrum and the fundus was not a common primary site. The majority of PGTL presented as a localized disease with or without involvement of local lymph nodes. The gross appearance of PGTL was a large ulcerated lesion in most cases and polypoid lesion or multiple lesions of infiltration were not common, which differed from those of PSITL. Perforation of the stomach has not been reported. No clear differences in gross appearance of the lesions were found between PGTLs with HTLV-1 and those without HTLV-1. No characteristic morphology of the neoplastic cells of PGTL was revealed irrespective of HTLV-1 infection. Eosinophilia has been considered a feature of some T-cell lymphomas such as ATL (32,53), which has been ascribed to constitutive overexpression of cytokines and their receptors in lymphoma cells (54). However, the present review revealed only two cases of PGTL with tissue eosinophilia, including one of the two cases presented here, and no cases of PGTL with HTLV-1 showed eosinophilia. Surgical resection of the primary organ was the modality of treatment for both PGTL and PSITL. Gastrectomy alone or combined with chemotherapy was most commonly used to treat PGTL. Surgical resection appeared to be chosen for the following purposes: a means for diagnosis, prevention of the complication and debulking of the tumor mass. Since T-cell lymphomas are generally resistant to chemotherapy, surgical resection, especially if it is curative, appeared to be an effective approach (55-58). Most patients with PSITL died within 1 year after the diagnosis, so that PSITL has been considered to have an extremely poor prognosis. However, the prognosis of PGTL appeared to be better than that of PSITL, although a few cases showed a short survival. The major cause of death in patients with PSITL was organ infiltration of the lymphoma cells. In addition, complications such as intestinal bleeding, perforation, peritonitis and bowel obstruction, which occurs post-operatively or during chemotherapy, were the causes of death in PSITL patients. In contrast, these serious complications were not common in PGTL. Only one of the six PGTL cases without HTLV-1 died of perforation of the colon, which is worthy of note in treating PGTL patients. As was evident in patients with PSITL, infiltration of the lymphoma cells into the intestinal wall constituted a potential risk of perforation, which warrants careful clinical management of PGITL. We attempted to understand more fully the clinicopathological features of PGTL, which is a very rare clinical condition. A retrospective review of the cases in the literature was therefore necessary. Despite these limitations, some of the characteristics of PGTL could be discerned. PGTL had a better prognosis than PSITL; however, HTLV-1 infection of PGTL predicted a very poor prognosis similar to those of PSITL and ATL. Hence it appears very important to describe morphological and immunological phenotypes of the lymphoma cells in addition to serological and molecular biological studies of HTLV-1 infection. Accumulation of such clinical information will help obtain an insight into the lymphomagenesis of PGTL. We thank Drs N. Mohri and S. Mori for their valuable comments and discussions. This work was supported in part by a Grand-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture, Japan, to T. Watanabe
DISCUSSION
Acknowledgments
References
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 19 Mar 1999
Copyright© 1999 Foundation for Promotion of Cancer Research.
CiteULike 
Connotea 
Del.icio.us What's this?
J. H. Cherem, J. Malagon, and H. Nellen
Cimetidine and Acute Intermittent Porphyria
Ann Intern Med,
November 1, 2005;
143(9):
694 - 695.
[Full Text]
[PDF]
This Article 
Abstract
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (10)
Request Permissions
Google Scholar
Articles by Horie, R
Articles by Watanabe, T
Search for Related Content
PubMed
PubMed Citation
Articles by Horie, R
Articles by Watanabe, T
Social Bookmarking
What's this?