Lymphocytic gastritis (LG), first described by Haot et al (1) in 1985, is a rare form of gastritis (0.8%–1.6% of cases) with unclear pathogenesis. According to the Updated Sydney system it is classified as a special form and is a histopathologic entity characterized by a dense lymphocytic infiltration (≥25 lymphocytes/100 epithelial cells) of the gastric foveolar and pit epithelium, together with a variable increase in chronic inflammatory cells in the lamina propria of the gastric mucosa (2). The diagnosis is based on the intraepithelial lymphocytes (IELS) count, which is always far beyond the values obtained in other types of gastritis. The disorder usually affects the whole stomach, although the degree of intraepithelial lymphocytosis displays a patchy variation.
There is scanty information about the etiopathogenesis, clinical significance, and evolution of LG. It is thought to be related to gluten sensitivity or to be an atypical reaction to Helicobacter pylori infection (3–6). Other entities such as lymphoma and adenocarcinoma have also been rarely associated with LG (7–9). Some studies have reported LG in almost 45% of cases of celiac disease (CD), with the gastritis regressing in response to a gluten-free diet (3,10,11). Others have reported a correlation of LG with serologically and/or histologically confirmed H pylori infection, with the LG being cured by H pylori eradication. The present study was aimed to characterize, in a prospective controlled manner, the relationship of LG to CD in a large cohort of patients with and without H pylori infection.
PATIENTS AND METHODS
This prospective study included 164 children with CD attending the Pediatric Gastroenterology division of the Department of Superspecialty of Gastroenterology, Postgraduate Institute of Medical Education and Research, Chandigarh, India, between June 2003 and October 2005. Gastric (2 biopsy specimens each from antrum and corpus) and duodenal biopsy specimens were taken simultaneously during the same upper gastrointestinal endoscopic procedure. The diagnoses of CD in these children were based on the modified ESPGHAN criteria (12). A positive serological result for antibodies against human recombinant tissue transglutaminase was taken as supportive evidence (13). In our area, CD has a prevalence of about 1:120 in the pediatric population (14). The control population consisted of 164 children without CD, matched for sex and age (within 1 year), in whom upper gastrointestinal endoscopy was performed for various upper gastrointestinal symptoms during the study period. None of the control children had peptic ulcer disease of the gastroduodenum.
Each biopsy specimen was laid on filter paper, oriented so that the luminal surface was uppermost, and fixed in 10% neutral formalin. The biopsy specimens were subsequently embedded in paraffin wax, cut into 5-μm-thick sections, and stained with hematoxylin and eosin. All of the biopsy specimens were assessed by a single experienced pathologist (K.K.P.), who was blinded to the clinical diagnosis at the time of histological evaluation. The diagnosis of LG was based entirely on the presence of gastric IELS. LG was diagnosed when there were ≥25 lymphocytes/100 gastric columnar epithelial cells. Lymphocytes accumulated in the foveolar and pit epithelium gave an intensely basophilic appearance at low magnification. At higher magnification, the lymphocytes were clearly distinguishable from the background of epithelial cytoplasm and nuclei because of their deeply basophilic nuclear chromatin, convoluted nuclear contours, and inapparent cytoplasm. They were surrounded by a clear rim, possibly because of artefactual retraction (15). H pylori were searched for in the gastric biopsy on hematoxylin and eosin sections, and a modified Giemsa stain was performed for confirmation in specimens suspected of showing H pylori. All of the duodenal histological parameters were evaluated and classified according to the Oberhuber classification (16) after modification as proposed by our group (17,18). A duodenal biopsy specimen with a villous/crypt ratio of >3 (along with a normal IEL count) was considered normal, whereas a decrease of the villous/crypt ratio indicated a diagnosis of varying degrees of atrophy, as follows: 2.5 to 1.5, mild villous atrophy; 1 to 0.5, moderate (partial) villous atrophy; and <0.5, severe (subtotal/total) villous atrophy. All of the parents gave informed consent for the participation of their children in the study. The study was approved by the ethical committee of our institute.
The categorical variables were expressed as frequencies and percentages. The continuous variables were expressed as mean ± standard error of the mean. The χ2 test and the Student t test were used for comparison of categorical and continuous variables, respectively. Pearson correlations and logistic regression models were applied as appropriate. A value of P < 0.05 was taken to be significant.
Demographics and Clinical Presentation of Celiac Disease
The mean age of children with CD at presentation was 8.2 ± 0.3 years (range 18 months to 17 years), and the mean age at onset of symptoms was 3.6 ± 0.9 years. The male:female ratio of the children with CD was 3:2 (99:65). The principal clinical manifestations in 164 patients with CD are shown in Figure 1.
The diagnosis of LG was made if ≥25 lymphocytes/100 gastric columnar epithelial cells were present in the gastric biopsy sample (Fig. 2). Patients with CD had a significantly higher prevalence of LG (69/164 [42.1%]) than did control children without CD (8/164 [4.9%]; P < 0.0001). Intergroup comparisons of gastric and duodenal IELs in patients with CD and control children without CD are shown in Figure 3. The gastric IELs were positively correlated with duodenal IELs in patients with CD (Pearson r = 0.601; P < 0.0001), and no correlation was seen in control children without CD (Pearson r = 0.081; P = 0.303). The patients with CD and LG had a mean IEL count of 43.9 ± 1.5/100 surface epithelial cells, compared with a mean of 13.4 ± 0.4 in patients with CD without LG (P < 0.0001) and 7.8 ± 0.5 in control children without CD (P < 0.0001). Control children with LG had a mean IEL count of 28.6 ± 0.2, compared with a mean of 6.7 ± 0.3 in control children without LG (P < 0.0001). Of the 95 patients with CD who did not have LG, 38/95 (40%) had moderately increased numbers of gastric IELs, defined as ≥15 and <25 IELs/100 epithelial cells, with a mean of 18.7 ± 0.3 IELs. A mild increase in numbers of gastric IELs, defined as 5 to <15 IELs/100 epithelial cells, was seen in 57/95 (60%) of non-LG patients with CD, with a mean of 9.9 ± 0.3 IELs. These showed increased gastric IELs compared with a mean of 7.8 ± 0.5 in the control children without CD (P < 0.001). However, the concentration of IELs had a heterogenous distribution and it varied from place to place. There was no strict parallel between the inflammation in the lamina propria and the lymphocytic infiltration of the epithelium. Patients with CD had a significantly lower prevalence of H pylori infection than did control children without CD (P < 0.0001). Nine of 164 (5.5%) patients with CD were positive for H pylori. Of the 69 patients with CD with LG, 4 (5.8%) were H pylori positive, whereas 49 of 164 (29.9%) control children without CD were H pylori positive, and 5 of 49 (10.2%) control children with LG were H pylori positive. The LG in patients with CD had no relation with age, sex, H pylori infection, or clinical manifestations like chronic recurrent diarrhea, iron deficiency anemia, failure to thrive, abdominal distension, recurrent abdominal pain, or short stature (P > 0.05).
Of the 164 patients with CD, 158 (96.3%) had evidence of villous atrophy (type 3a, 3b, or 3c), and 6/164 (3.7%) patients with CD showed a type 2 lesion. LG was observed only in patients with duodenal villous atrophy (type 3a, 3b, or 3c); it was not seen in any patients with CD who had a normal duodenal villous architecture (type 1 or 2 lesions). The patients with CD and LG were significantly more likely to have a classic duodenal histological lesion (type 3c or 3b; P < 0.0001). Patients with CD showed a significantly higher duodenal mean IEL of 80.5 ± 1.9 than did control children without CD, with a mean of 15.7 ± 0.3 (P < 0.0001). In 69 patients with CD and LG, the mean duodenal IEL was 94.1 ± 11.3, compared with 70.6 ± 7.2 in 95 patients with CD without LG (P < 0.0001).
LG is recognized as a special type of chronic gastritis characterized by an intense lymphocytic infiltration of the gastric foveolar and pit epithelium in the antral or corpus mucosa. LG rarely exceeds 5% of the histological diagnosis in gastric biopsy specimens (11,19). This diagnosis can be easily made by IEL counts in preparations stained with hematoxylin and eosin; the IEL counts are far beyond the values obtained in other types of gastritis. Little is known about the etiopathogenesis, clinical significance, and evolution of the disease. A possible relation of LG exists with H pylori infection (5,6) and CD (3,4). LG may also complicate mucosa-associated lymphoma tissue lymphomas and gastric carcinoma (7–9). The prevalence of LG in patients with CD has been reported to vary between 10% and 45% (3,10,11,20), with the gastritis regressing in response to a gluten-free diet (21); in our series 69 of 164 patients (42%) had LG, with no difference between H pylori–negative and H pylori–positive patients. The data from the present study indicate that CD is associated with a lower prevalence of H pylori infection detected by direct determination of the bacterium in gastric tissue, compared with control children without CD. Similar findings have been reported in a previous study (22). It was beyond the scope of this study to compare the data on H pylori infection in patients with CD with data from the general population. It may be hypothesized that the modified gastric mucosal milieu in patients with CD is less favorable for H pylori infection. The gastric immunoactivation that occurs as an immune response to gluten, reduced gastric acid output as a result of the alteration in gastric hormone release, or other currently unknown factors may constitute a host defense against H pylori(22).
The pathogenesis of LG is uncertain. LG is speculated to be a consequence of 1 or more of the following mechanisms: an independent immunological disorder produced by a gastric mucosal intolerance to various environmental factors (gliadin, H pylori, viral agents, or other factors) or as part of a generalized autoimmune process associated with CD (23–27). The frequent association of LG with H pylori infection and CD suggests that IELs constitute a local immune response to luminal antigens (11). The IELs are predominantly CD8+ cytotoxic/suppressor T lymphocytes (28,29). CD8+ intraepithelial T lymphocyte population lacks perforin and granzyme B, undergoes apoptosis, and is not associated with substantial damage to the epithelial cells (30). CD is associated with the human leukocyte antigen class I antigens A1 and B8 and also with class II antigens D/DR3, 7, and DC3 (31–33). Class I antigens are important in the process of self-recognition, guiding the cytolytic action of T killer cells, whereas class II antigens are thought to regulate the functions of T helper and suppressor lymphocytes and may have a role in the intracellular transport of foreign macromolecules. Kushima and Borchard (34) have suggested that the expression of major histocompatibility complex class I is closely related to the pathogenesis of LG. Therefore, some intracellular autoantigens, molecularly mimicking viral antigens, may play a role in its pathogenesis. These autoantigens may be associated indirectly with H pylori infection but also with viral infection, food allergen, bile reflux, or gastric tumor (24–27,34).
The association of CD with morphological gastric changes, such as LG (3,4,10,11,35), follicular gastritis (36), or H pylori infection (22,37), has been a matter of investigation for several years. Most of the published studies available have shown methodological defects. Some studies were based on retrospective analysis with inappropriate or inadequate controls; other studies had a selection bias or considered a small number of cases. To the best of our knowledge, this is the first prospective controlled study with the largest cohort of CD in children reported from a single center in the English literature. The recent description of a high prevalence of LG in patients with CD is in agreement with our present findings (3,10,11,34). Gastric involvement in CD is not surprising and confirms the theory that gluten intolerance may involve any portion of the gastrointestinal tract (38). Furthermore, CD-associated LG may not be a distinct entity but an inconsistent reflection of the true pathological changes of the condition, which is found in the small intestine and may be the less florid gastric equivalent of the small intestinal response to gluten (20). We observed that H pylori infection in the setting of CD conferred no risk for LG. The LG was seen only in patients with duodenal villous atrophy; it was not seen in any of the patients who had normal duodenal villous architecture. This finding was comparable to the findings of Feeley et al (20). The duodenal IEL count in patients without LG was high and it differed significantly from those in the patients with LG.
The present study also shows that patients with CD without LG nevertheless contained significantly increased numbers of IELs in the gastric epithelium, compared with control children without CD, as reported by other workers (4,20). The IEL counts between the patients with CD and control children without CD overlapped, so this increase cannot be regarded as a specific diagnostic feature. On the contrary, it may be an indication for a duodenal mucosal biopsy, especially when typical features of CD are not present, the clinical presentations of CD often being diverse and/or extraintestinal. We therefore suggest that obtaining a biopsy specimen from the duodenal mucosa may be considered in patients in whom LG is identified and the clinical findings are remotely suggestive of a malabsorption syndrome. Our study did not entail a follow-up of patients with CD with LG after the institution of a gluten-free diet. We evaluated the gastric antral and duodenal biopsy specimen in patients with CD only at the initial presentation. Further studies are needed to document the alterations of the gastric mucosa in patients with CD with LG after gluten is removed from the diet.
In conclusion, our study demonstrates that LG in a cohort of Indian children is associated with CD. An interesting observation in our CD population was the increased gastric IEL concentration even in the absence of LG. In the absence of CD, LG may be caused by a sensitivity of the host to various luminal antigens, such as H pylori, not unlike the response to gluten in patients with CD. Histopathologists should be alert to the distinct appearance of LG and should suggest a duodenal biopsy in positive cases to identify patients with atypical CD.
1. Haot J, Jouret-Mourin A, Hardy N. La gastrite “à lymphocytes.” Une nouvelle entité? Acta Endosc 1985; 15:187–188.
2. Dixon MF, Genta RM, Yardely JH, et al
. Classification and grading of gastritis: the updated Sydney system. Am J Surg Pathol 1996; 20:1161–1181.
3. De Giacomo C, Gianatti A, Negrini R, et al
. Lymphocytic gastritis
: a positive relationship with celiac disease
. J Pediatr 1994; 124:57–62.
4. Alsaigh N, Odze R, Goldman H, et al
. Gastric and esophageal intraepithelial lymphocytes in pediatric celiac disease
. Am J Surg Pathol 1996; 20:865–870.
5. Sundaram KK, Mendall MA. Lymphocytic gastritis
and Helicobacter pylori
: reluctant mucosal partners? Helicobacter 2000; 5:248–249.
6. Muller H, Volkholz H, Stolte M. Healing of lymphocytic gastritis
by eradication of Helicobacter pylori
. Digestion 2001; 63:14–19.
7. Heitzmann A, Michenet P, Esperandieu O, et al
. Lymphocytic gastritis
associated with gastric MALT lymphoma. Gastroenterol Clin Biol 1997; 21:433–434.
8. Cammarota G, Tursi A, Fedell G, et al
. Association of Helicobacter pylori
infection, lymphoid follicles, and lymphocytic gastritis
: a risk factor for the development of primary gastric lymphoma? Gut 1996; 38:792.
9. Griffiths AP, Wyatt J, Jack AS, et al
. Lymphocytic gastritis
, gastric adenocarcinoma and primary gastric lymphoma. J Clin Pathol 1994; 47:1123–1124.
10. Wolber R, Owen D, Del Buono L, et al
. Lymphocytic gastritis
in patients with celiac sprue or spruelike intestinal disease. Gastroenterology 1990; 98:310–315.
11. Wu TT, Hamilton SR. Lymphocytic gastritis
: association with etiology and topology. Am J Surg Pathol 1999; 23:153–158.
12. Walker-Smith JA, Guandalini S, Schmitz J, et al
. Revised criteria for diagnosis of coeliac disease. Arch Dis Child 1990; 65:909–911.
13. Baldas V, Tommasini A, Trevisiol C, et al
. Development of a novel rapid non-invasive screening test for coeliac disease. Gut 2000; 47:628–631.
14. Lal S, Nain CK, Prasad KK, et al
. Celiac disease
in Chandigarh: a population survey. Indian J Gastroenterol 2005; 24(Suppl 1):A39.
15. Haot J, Hamichi L, Wallez L, et al
. Lymphocytic gastritis
: a newly described entity: a retrospective endoscopic and histological study. Gut 1988; 29:1258–1264.
16. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol 1999; 11:1185–1194.
17. Prasad KK, Thapa BR, Nain CK, et al. Assessment of the diagnostic value of duodenal bulb histology in patients with celiac disease
, using multiple biopsy sites. J Clin Gastroenterol
18. Prasad KK, Thapa BR, Nain CK, et al
. Brush border enzyme activities in relation to histological lesion in pediatric celiac disease
. J Gastroenterol Hepatol 2008; 23:e348–e352.
19. Ribeiro VL, Barbosa AJ. Lymphocytic gastritis
: a study of its frequency and review of the literature. Arq Gastroenterol 1998; 35:26–31.
20. Feeley KM, Heneghan MA, Stevens FM, et al
. Lymphocytic gastritis
and coeliac disease: evidence of a positive association. J Clin Pathol 1998; 51:207–210.
21. Villanacci V, Bassotti G, Liserre B, et al
. Helicobacter pylori
infection in patients with celiac disease
. Am J Gastroenterol 2006; 101:1880–1885.
22. Ciacci C, Squillante A, Rendina D, et al
. Helicobacter pylori
infection and peptic disease in coeliac disease. Eur J Gastroenterol Hepatol 2000; 12:1283–1287.
23. Diamanti A, Maino C, Niveloni S, et al
. Characterization of gastric mucosal lesions in patients with celiac disease
: a prospective controlled study. Am J Gastroenterol 1999; 94:1313–1319.
24. Stene LC, Honeyman MC, Hoffenberg EJ, et al
. Rotavirus infection frequency and risk of celiac disease
autoimmunity in early childhood: a longitudinal study. Am J Gastroenterol 2006; 101:2333–2340.
25. Troncone R, Auricchio S. Rotavirus and celiac disease
: clues to the pathogenesis and perspectives on prevention. J Pediatr Gastroenterol Nutr 2007; 44:527–528.
26. Fine KD, Ogunji F, Saloum Y, et al
. Celiac sprue: another autoimmune syndrome associated with hepatitis C. Am J Gastroenterol 2001; 96:138–145.
27. Colombo E, Cermesoni L, Morganti D. Celiac sprue: another autoimmune syndrome associated with hepatitis C? Dig Liver Dis 2003; 35:64–65.
28. Freedman AR, Macartney JC, Nelufer JM, et al
. Timing of infiltration of T lymphocytes induced by gluten into the small intestine in coeliac disease. J Clin Pathol 1987; 40:741–745.
29. Olives JP, Voigt JJ, al Saati T, et al
. T lymphocyte populations of the intestinal mucosa in celiac disease
in children: immunohistochemical study. Gastroenterol Clin Biol 1990; 14:33–40.
30. Drut R, Drut RM. Lymphocytic gastritis
in pediatric celiac disease
: immunohistochemical study of the intraepithelial lymphocytic component. Med Sci Monit 2004; 10:CR38–CR42.
31. Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunologic approach to the spectrum of gluten sensitivity (“celiac sprue”). Gastroenterology 1992; 102:330–354.
32. Stepniak D, Koning F. Celiac disease
: sandwiched between innate and adaptive immunity. Hum Immunol 2006; 67:460–468.
33. Sollid LM, Markussen G, Ek J, et al
. Evidence for a primary association of celiac disease
to a particular HLA-DQ alpha/beta heterodimer. J Exp Med 1989; 169:345–350.
34. Kushima R, Borchard F. Lymphocytic gastritis
: autoimmune disease or variant of Helicobacter gastritis? Verh Dtsch Ges Pathol 1996; 80:208–211.
35. Vogelsang H, Oberhuber G, Wyatt J. Lymphocytic gastritis
and gastric permeability in patients with celiac disease
. Gastroenterology 1996; 111:73–77.
36. Cammarota G, Fedell G, Tursi A. Coeliac disease and follicular gastritis. Lancet 1996; 347:268.
37. Santarelli L, Gabrielli M, Santoliquido A, et al
. Interaction between Helicobacter pylori infection and untreated coeliac disease on gastric histological pattern. Scand J Gastroenterol 2006; 41:532–535.
38. Austin L, Dobbins W. Studies of the rectal mucosa in celiac sprue: the intraepithelial lymphocyte. Gut 1988; 29:200–205.