ORIGINAL ARTICLE
Pediatric Non-Helicobacter Pylori Atrophic Gastritis A Case Series Jennifer Pogoriler, MD, PhD,* Daniel Kamin, MD,w and Jeffrey D. Goldsmith, MDz
Abstract: Although autoimmune atrophic gastritis is classically a disease of elderly adults, recent studies have described the disease in younger adults, particularly in those with other autoimmune diseases and iron-deficiency anemia. Atrophic gastritis in pediatrics is a rare and possibly underdiagnosed entity that has been primarily reported as single-case reports. This retrospective study of atrophic gastritis not associated with Helicobacter pylori infection was performed to further expand the knowledge of clinical presentation, pathologic findings, and natural history of this disease in the pediatric population. Twelve patients with a histologic diagnosis of atrophic gastritis were identified, with an age range of 8 months to 18 years. Seven had other autoimmune diseases and/or immunodeficiency. Atrophy was confined to the oxyntic mucosa in 10 patients, with intramucosal inflammation in a diffuse or basal-predominant pattern. Active inflammation was present in 7 patients. Pseudopyloric, intestinal, or squamous/mucinous metaplasia was seen at initial biopsy or on follow-up in 8 patients, and enterochromaffin-like cell hyperplasia was seen in 5. One patient developed an adenocarcinoma during the follow-up period of 10 years. Two false-negative diagnoses were retrospectively identified. In the majority of cases, the possibility of atrophic gastritis was not raised by the submitting physician, and the endoscopic findings were not specific. Therefore, accurate diagnosis requires a high degree of suspicion on the part of the pathologist, and the diagnosis should be considered particularly in patients with a clinical history of other autoimmune diseases or iron-deficiency anemia. Key Words: atrophic gastritis, autoimmune gastritis, pediatrics (Am J Surg Pathol 2015;39:786–792)
From the *Department of Pathology and Laboratory Medicine, The Children’s Hospital of Philadelphia, Philadelphia, PA; wDepartment of Gastroenterology and Nutrition; and zDepartment of Pathology, Beth Israel Deaconess Medical Center, Children’s Hospital Boston, and Harvard Medical School, Boston, MA. Presented in abstract form at the 2014 Annual Meeting of the Society of Pediatric Pathologists, San Diego, CA. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Jeffrey D. Goldsmith, MD, Department of Pathology, Beth Israel Deaconess Medical Center, 330 Brookline Ave., Boston, MA 02215 (e-mail: [email protected]). Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
786 | www.ajsp.com
G
astric atrophy is defined as the loss of gastric oxyntic and/or antral glands,1,2 which may be replaced by fibrous tissue and/or metaplastic epithelium, and typically includes increased lamina propria inflammation. The 2 common causes of atrophic gastritis are Helicobacter pylori with antral or multifocal atrophy and autoimmune atrophic gastritis with atrophy of the oxyntic mucosa. Although autoimmune atrophic gastritis is classically a disease of elderly adults of northern European ethnic background with B12 deficiency, recent studies have described the disease in younger adults, particularly in those with other autoimmune diseases and iron-deficiency anemia.3–6 Atrophic gastritis in pediatrics is a rare and possibly underdiagnosed entity. The majority of studies have examined H. pylori-associated antral atrophy in children located in developing countries,7 whereas autoimmune atrophic gastritis has been primarily reported as singlecase reports in patients with immunodeficiencies, some of whom subsequently developed adenocarcinoma.8–10 Most recently, a series of 8 pediatric patients with refractory iron-deficiency anemia and atrophic gastritis was reported11 without associated autoimmune conditions; 4 of these patients showed H. pylori infection. Rare cases of a relatively recently described entity called “atrophic autoimmune pangastritis” have also been reported in children, in which both the antral and oxyntic mucosae demonstrate intense inflammation with complete atrophy of specialized cells.12 This retrospective study of atrophic gastritis not associated with H. pylori was performed to further expand the knowledge of the clinical presentation, pathologic findings as seen on endoscopically obtained mucosal biopsies, and natural history of this disease in the pediatric population.
METHODS With institutional review board approval (protocol number IRB-P00009596), the pathology archives were searched for patients with a diagnosis of atrophic gastritis from 1994 to 2013. Exclusion criteria included any laboratory or histologic evidence of Helicobacter infection and histologic evidence of collagenous gastritis. All gastric and concurrently obtained duodenal biopsies were reviewed. Immunohistochemical analyses for Helicobacter, gastrin, and chromogranin were performed on all biopsies using the Ventana Benchmark Ultra or XT platforms (Tucson, Am J Surg Pathol
Volume 39, Number 6, June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Am J Surg Pathol
Volume 39, Number 6, June 2015
Pediatric Non-H. pylori Atrophic Gastritis
AZ) with the following prediluted antibodies: (1) gastrin (mouse polyclonal; Ventana Medical Systems); (2) helicobacter (rabbit polyclonal; Cell Marque, Rocklin, CA); (3) chromogranin A (mouse clone LK2H10; Ventana Medical Systems). Biopsy location was defined on the basis of an endoscopic report and was confirmed by the presence of G cells in the antrum or the absence of G cells in the oxyntic mucosa using gastrin immunohistochemistry in areas devoid of intestinal metaplasia. Hematoxylin and eosin slides were scored for mucosal atrophy, intestinal metaplasia, pseudopyloric metaplasia, lymphoplasmacytic inflammation, and neutrophilic inflammation on the basis of the visual analogue updated Sydney system.13 The intramucosal distribution of chronic inflammation was evaluated as described14 for basal or fullthickness mucosal inflammation. Increased intraepithelial lymphocytes were defined as >25 intraepithelial lymphocytes/100 epithelial cells.13 Eosinophils were counted in a 400 field in the area of highest density. Enterochromaffin-like cell (ECL) hyperplasia was evaluated on chromogranin immunostain by standard criteria.15 Epithelial apoptosis was quantified as absent, rare ( 100 5 80 51 41 10
ECL Intraepithelial Hyperplasia Apoptosis Lymphocytosis N L, N L L N L, N L, N NA NA L, N
Rare Rare Rare Rare Rare Rare + NA Rare
+B + + NA +
Follow-up biopsies are in gray. +++ indicates marked; ++, moderate; +, mild; B, basal; C, complete; CI, chronic inflammation; D, diffuse; I(c), complete intestinal metaplasia; L, linear; M, mucinous metaplasia; N, micronodular; NA, not available; P, partial; Py, pseudopyloric metaplasia; S, squamous metaplasia; S, superficial.
panel of other autoantibody tests and the finding of elevated gastrin levels in the workup for chronic diarrhea. Atrophic gastritis was unexpected in the other 9 patients. Six of the 12 patients had iron-deficiency anemia at presentation, although this was often multifactorial and not attributable solely to gastritis. In the 3 patients who were tested, H. pylori IgG was negative. In no case was the stomach described as atrophic, ulcerated, or eroded at endoscopy. Laboratory markers associated with autoimmune gastritis were ordered inconsistently after biopsy results: antiparietal cell antibodies were positive in 3 of 7 patients, and anti–intrinsic factor antibodies were positive in 2 of 6 patients, with a total of 4 patients positive for at least 1 of these 2 markers. In patients with an unexpected diagnosis of atrophic gastritis, follow-up laboratory testing for vitamin B12 levels, when performed, was within normal limits; however, only 9 patients were tested.
Histologic Data For 10 patients, both the antral and oxyntic mucosa were biopsied at the time of initial endoscopy. The remaining 2 patients had initial biopsies from either the oxyntic (patient #11) or antral (patient #7) mucosa. In the patient with only antral biopsies, a follow-up endoscopy included a biopsy of the corpus (Table 2). On initial biopsy, 10 patients had mucosal atrophy present only in the oxyntic mucosa (Fig. 1). One patient with oxyntic mucosal atrophy (#3) developed atrophy of the antrum on follow-up biopsy. In addition to loss of oxyntic cells, pseudopyloric metaplasia in the oxyntic mucosa was identified in 3 cases at initial diagnosis,
788 | www.ajsp.com
whereas no patient showed intestinal metaplasia at presentation. Patient #7, who had a history of cytomegalovirus gastritis with complete loss of normal antral and oxyntic glands, showed extensive squamous and mucinous metaplasia (Fig. 2). Apoptotic epithelial cells were not a significant histologic finding in any biopsy. Of the 7 cases with follow-up biopsies (taken 9 mo to 10 y after index endoscopy; mean 49 mo), 3 cases showed persistent metaplastic findings (2 cases with pseudopyloric metaplasia and 1 case with squamous/mucinous metaplasia). Three of 4 patients without metaplasia developed complete intestinal metaplasia on follow-up, including 1 patient who developed complete intestinal metaplasia in both the oxyntic and antral mucosae (Fig. 3); incomplete intestinal metaplasia was not seen. Inflammation was usually more severe in the corpus than in the antrum but was occasionally of similar intensity. Antral-predominant inflammation was not identified. In the oxyntic mucosa, the intramucosal distribution of chronic inflammation was always present in a fullthickness or basal-predominant pattern and was moderate or marked in 10 cases. Neutrophilic inflammation was present in 7 patients and eosinophils were increased in all but 1 case (Fig. 4). Diffusely increased intraepithelial lymphocytes were present in 2 patients. Chronic inflammation persisted in all patients who had follow-up biopsies. Immunohistochemistry for Helicobacter was negative in all cases. Linear or micronodular ECL hyperplasia was seen in the oxyntic mucosa in 4 cases initially and in 1 additional patient on follow-up (Fig. 5). Patient #3 developed a gastric adenocarcinoma at 17 years of age, 8 Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Am J Surg Pathol
Volume 39, Number 6, June 2015
Pediatric Non-H. pylori Atrophic Gastritis
TABLE 2. (continued) Antral Mucosa Type of Atrophy Metaplasia 1 1 2 2 3 3 4 4 5 5 6 7 7 8 9 10 10 11 12
P NA P NA NA
I(c) NA S, M NA NA
Extent of Metaplasia
Chronic Inflammation
Distribution of Mucosal CI
Neutrophilic Inflammation
Maximum Eosinophils/ HPF
Apoptosis
Intraepithelial Lymphocytosis
++ NA ++ NA NA
+++ + + +++ + ++ NA + + ++ NA + + + NA +
B S NA NA D D S NA D NA B D NA D D NA S NA D
+ + NA NA NA
78 2 1 4 12 64 36 NA 24 19 2 21 NA 48 6 6 13 NA 1
Rare Rare + NA Rare Rare NA NA
+ NA + NA NA +
years after initial diagnosis. Patient #4 died of graft versus host disease after stem cell transplantation for lymphoma. Gastric neuroendocrine tumors were not seen. In the absence of G cells as displayed immunochemically, false-negative diagnoses were retrospectively identified in 2 cases. These cases, which were designated by the endoscopists as having an origin from either the corpus or fundus, were initially diagnosed as “antral” or “pyloric-type” mucosa with chronic gastritis. In patient #3, this diagnosis was repeated until the development of B12 deficiency, at which time a correct diagnosis of atrophic gastritis was rendered. Retrospective gastrin staining of the previous biopsies on patient #3 was negative, indicating that the corpus/fundus area was sampled, and the previous biopsies did show atrophy. Examination of the concurrently obtained duodenal biopsies was not contributory except for patient #3, whose duodenal biopsies showed changes consistent with celiac disease.
DISCUSSION In contrast to previous pathologic case series in developing countries that focused on H. pylori-associated pediatric atrophic gastritis, we identified 12 H. pylorinegative pediatric patients with atrophic gastritis predominantly affecting the corpus, which is, to our knowledge, the largest retrospective case series to date. This is a very rare diagnosis in the pediatric population, with only 12 patients identified over 20 years in an institution with a large gastrointestinal biopsy volume (approximately 3200 gastrointestinal cases accessioned in 2013). In the majority of cases, the possibility of atrophic Copyright
r
gastritis was not raised by the submitting physician, and the endoscopic findings were not specific. Therefore, accurate diagnosis requires a high degree of suspicion on the part of the pathologist, and the diagnosis should be considered particularly in patients with a clinical history of other autoimmune diseases or iron-deficiency anemia. The largest subset of patients in our series was diagnosed with autoimmune atrophic gastritis, but even among this cohort an exact diagnosis was not always possible. Large adult series often use the presence of body-predominant inflammation with the detection of anti–parietal cell or anti–intrinsic factor antibodies as inclusion criteria for autoimmune atrophic gastritis14,16; however, autoantibodies are not 100% sensitive,17 and patients who are initially negative may test positive later in their disease course.16 A small series of patients with “atrophic autoimmune pangastritis” has been described12 and includes several pediatric patients. The main findings in this entity included a marked inflammatory infiltrate in both the antrum and corpus with neutrophilic activity and increased apoptotic cells despite a complete absence of oxyntic cells, the absence of ECL hyperplasia, and a lack of H. pylori infection. Although several of our cases showed no evidence of ECL hyperplasia, a diagnosis of autoimmune pangastritis is excluded by the absence of atrophy and significant inflammation of the antrum in these cases. In our study, antral inflammation was variable, and in patients who had follow-up biopsies the degree of acute and chronic inflammation was often different from the original biopsy findings, making it concerning to use this as a criterion for disease classification.
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
www.ajsp.com |
789
Pogoriler et al
Am J Surg Pathol
Volume 39, Number 6, June 2015
FIGURE 1. Typical case of atrophic gastritis as seen in this biopsy from the gastric corpus. Complete loss of oxyntic cells are seen with full-thickness lymphoplasmacytic inflammation. A gastrin immunostain on this biopsy was negative, which confirms the site of origin (hematoxylin and eosin).
However, in no case was there an improvement in atrophy; indeed, on follow-up more patients developed metaplasia, which is consistent with the known progression of typical autoimmune atrophic gastritis. Clinical follow-up of our patients was quite variable. Those with other autoimmune diseases often had follow-up biopsies; however, even in these patients, assays for autoantibodies, serum gastrin, or B12 were often not performed. In 1 case, the atrophy was attributed to excess NSAID use, and further follow-up was not performed. We have not identified any cases in the literature in which chronic NSAID use was a cause for atrophic gastritis.
FIGURE 2. This unusual pattern of “squamo-mucinous” metaplasia was seen in case number 7; the patient had a history of immunodeficiency and chronic cytomegalovirus infection. Normal oxyntic glands were not present (hematoxylin and eosin).
790 | www.ajsp.com
FIGURE 3. This biopsy of the gastric corpus showed combined complete intestinal metaplasia and pseudopyloric metaplasia (hematoxylin and eosin).
One unique patient with a history of cytomegalovirus gastritis, immunodeficiency, and gastro-cutaneous fistulas showed a highly unusual pattern of squamous and mucinous metaplasia, which was originally identified in the antrum and persisted on follow-up. Whether this
FIGURE 4. Few cases showed patchy increase in lamina propria eosinophils as seen in the center-bottom portion of this biopsy of the corpus (hematoxylin and eosin). Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Am J Surg Pathol
Volume 39, Number 6, June 2015
Pediatric Non-H. pylori Atrophic Gastritis
FIGURE 5. Entrochromaffin-like cell (ECL) hyperplasia was seen in approximately half of the cases. A, In a few cases, micronodular hyperplasia was appreciated on the H&E stain (H&E). B, This impression was confirmed using chromogranin immunohistochemistry, which confirmed the presence of both linear and micronodular ECL hyperplasia (chromogranin immunohistochemistry). H&E indicates hematoxylin and eosin.
pattern reflects her young age or the unusual clinical circumstances is unknown. A limitation of this study was our dependence on the prior diagnosis of atrophy to identify cases, and, thus, we likely overlooked cases that were misdiagnosed. The majority of patients in our series had complete or subtotal atrophy, and specimens with partial atrophy would likely be overlooked. On the basis of history and follow-up biopsies, we did identify 2 patients with complete atrophy who did not receive a correct histologic diagnosis on initial biopsy. Beyond the known interobserver variability in grading atrophy,1,18 the ability to make a diagnosis of atrophy can be compromised by the receipt of specimens designated simply as “stomach” as well as by poor orientation or sampling of the oxynto-antral transition zone. With complete loss of oxyntic glands, atrophic corpus may easily be mistaken for antral mucosa with chronic gastritis, and areas with partial atrophy may be indistinguishable from areas of transitional mucosa with chronic gastritis. Given the relative rarity of atrophic gastritis in pediatric biopsies, the finding of “antral” mucosa in a specimen labeled “corpus” may be more likely to reflect inaccurate sampling than atrophic gastritis of the corpus, particularly in the absence of chronic inflammation. However, findings that should raise the suspicion for atrophic gastritis include a full-thickness or basal-predominant pattern of chronic inflammation with or without intestinal metaplasia. In many cases, gastrin immunohistochemistry can be helpful to clarify the location of the biopsy, as the antrum should have a prominent mid-zone band of gastrin-positive cells. Some limitations to this approach include the facts that rare, scattered gastrin-positive cells may be present in the oxyntic mucosa, and variable numbers of these cells may be present in the oxynto-antral transition zone. In addition, Copyright
r
gastrin-positive cells may be increased in areas of pseudopyloric metaplasia, and they are often absent in areas of intestinal metaplasia.19 Conversely, gastrin-positive cells in the antrum may be slightly decreased in areas of chronic gastritis.20 Therefore, even with a gastrin stain, the appropriate localization of a biopsy may be challenging. Although eosinophils are often prominent in pediatric inflammatory disorders and were increased in the majority of the biopsies from our patients with atrophic gastritis, this feature is not specific and is likely secondary to the markedly increased lymphoplasmacytic infiltrate in a majority of study patients. A recent series of adult biopsies with autoimmune atrophic gastritis noted >30 eosinophils per high-power field in 46% of patients and neutrophilic cryptitis in 44%.14 Long-term consequences of atrophic gastritis include adenocarcinoma and neuroendocrine tumors. In adults with atrophic gastritis, the risk for adenocarcinoma is estimated to be increased by 3-fold and the risk for gastric neuroendocrine tumor by 13-fold.17 Consistent with this rare event, 1 of our patients developed an adenocarcinoma during the follow-up period of 10 years. This was the patient in whom the diagnosis of atrophy was initially missed but who subsequently received frequent follow-up and monitoring for dysplasia. Rare cases of gastric adenocarcinoma in teenagers have been reported10 that arose in a background of atrophic gastritis. ECL hyperplasia in the gastric corpus is a result of the hypergastrinemia induced by the hypochlorhydric state and is the histologic precursor of neuroendocrine tumors that arise in the context of atrophic gastritis. The elevated serum gastrin levels are secondary to a loss of the negative signal imposed on G cells by the presence of gastric acid. Even in early atrophic gastritis before the complete loss of oxyntic glands is apparent, ECL hyper-
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
www.ajsp.com |
791
Pogoriler et al
Am J Surg Pathol
plasia is often found.16 Some level of ECL hyperplasia was appreciated in 5 of our study patients; none of our study patients showed evidence of neuroendocrine tumors. In conclusion, atrophic gastritis is an extremely rare diagnosis in the pediatric population whose histology is reminiscent to that in adults. In those cases in which H. pylori infection has been excluded, autoimmunity is the most likely etiology. Given its extreme rarity in children, a high level of suspicion is required for accurate diagnosis. REFERENCES 1. Rugge M, Genta RM. Staging and grading of chronic gastritis. Hum Pathol. 2005;36:228–233. 2. Rugge M, Correa P, Dixon MF, et al. Gastric mucosal atrophy: interobserver consistency using new criteria for classification and grading. Aliment Pharmacol Ther. 2002;16:1249–1259. 3. Centanni M, Marignani M, Gargano L, et al. Atrophic body gastritis in patients with autoimmune thyroid disease: an underdiagnosed association. Arch Intern Med. 1999;159:1726–1730. 4. De Block CE, De Leeuw IH, Van Gaal LF. High prevalence of manifestations of gastric autoimmunity in parietal cell antibodypositive type 1 (insulin-dependent) diabetic patients. The Belgian Diabetes Registry. J Clin Endocrinol Metab. 1999;84:4062–4067. 5. Marignani M, Delle Fave G, Mecarocci S, et al. High prevalence of atrophic body gastritis in patients with unexplained microcytic and macrocytic anemia: a prospective screening study. Am J Gastroenterol. 1999;94:766–772. 6. Hershko C, Hoffbrand AV, Keret D, et al. Role of autoimmune gastritis, Helicobacter pylori and celiac disease in refractory or unexplained iron deficiency anemia. Haematologica. 2005;90:585–595. 7. Ricuarte O, Gutierrez O, Cardona H, et al. Atrophic gastritis in young children and adolescents. J Clin Pathol. 2005;58:1189–1193. 8. Vaterlaws AL. Gastric atrophy in childhood. Arch Dis Child. 1969;44:710–714.
792 | www.ajsp.com
Volume 39, Number 6, June 2015
9. Russell AC, Black JO, Schwartz DA, et al. 15-year-old girl with metaplastic atrophic gastritis and enterochromaffin-like cell hyperplasia. J Pediatr Gastroenterol Nutr. 2012;55:e148–e151. 10. Slotta JE, Heine S, Kauffels A, et al. Gastrectomy with isoperistaltic jejunal parallel pouch in a 15-year-old adolescent boy with gastric adenocarcinoma and autosomal recessive agammaglobulinemia. J Pediatr Surg. 2011;46:e21–e24. 11. Miguel N, Costa E, Santalha M Jr, et al. Refractory iron-deficiency anemia and autoimmune atrophic gastritis in pediatric age group: analysis of 8 clinical cases. J Pediatr Hematol Oncol. 2014;36: 134–139. 12. Jevremovic D, Torbenson M, Murray JA, et al. Atrophic autoimmune pangastritis: a distinctive form of antral and fundic gastritis associated with systemic autoimmune disease. Am J Surg Pathol. 2006;30:1412–1419. 13. Dixon MF, Genta RM, Yardley JH, et al. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol. 1996;20:1161–1181. 14. Bettington M, Brown I. Autoimmune gastritis: novel clues to histological diagnosis. Pathology. 2013;45:145–149. 15. Solcia E, Fiocca R, Villani L, et al. Hyperplastic, dysplastic, and neoplastic enterochromaffin-like-cell proliferations of the gastric mucosa. Classification and histogenesis. Am J Surg Pathol. 1995; 19(suppl 1):S1–S7. 16. Torbenson M, Abraham SC, Boitnott J, et al. Autoimmune gastritis: distinct histological and immunohistochemical findings before complete loss of oxyntic glands Mod Pathol. 2002;15:102–109. 17. Park JY, Lam-Himlin D, Vemulapalli R. Review of autoimmune metaplastic atrophic gastritis. Gastrointest Endosc. 2013;77:284–292. 18. Talebkhan Y, Mohammadi M, Rakhshani N, et al. Interobserver variations in histopathological assessment of gastric pathology. Pathology. 2009;41:428–432. 19. Rubio CA, Jaramillo E, Suzuki G, et al. Antralization of the gastric mucosa of the incisura angularis and its gastrin expression. Int J Clin Exp Pathol. 2009;2:65–70. 20. Sloan JM, Buchanan KD, McFarland RJ, et al. A histological study of the effect of chronic gastritis on gastrin cell distribution in the human stomach. J Clin Pathol. 1979;32:201–207.
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Pediatric Non-Helicobacter Pylori Atrophic Gastritis A Case Series Jennifer Pogoriler, MD, PhD,* Daniel Kamin, MD,w and Jeffrey D. Goldsmith, MDz
Abstract: Although autoimmune atrophic gastritis is classically a disease of elderly adults, recent studies have described the disease in younger adults, particularly in those with other autoimmune diseases and iron-deficiency anemia. Atrophic gastritis in pediatrics is a rare and possibly underdiagnosed entity that has been primarily reported as single-case reports. This retrospective study of atrophic gastritis not associated with Helicobacter pylori infection was performed to further expand the knowledge of clinical presentation, pathologic findings, and natural history of this disease in the pediatric population. Twelve patients with a histologic diagnosis of atrophic gastritis were identified, with an age range of 8 months to 18 years. Seven had other autoimmune diseases and/or immunodeficiency. Atrophy was confined to the oxyntic mucosa in 10 patients, with intramucosal inflammation in a diffuse or basal-predominant pattern. Active inflammation was present in 7 patients. Pseudopyloric, intestinal, or squamous/mucinous metaplasia was seen at initial biopsy or on follow-up in 8 patients, and enterochromaffin-like cell hyperplasia was seen in 5. One patient developed an adenocarcinoma during the follow-up period of 10 years. Two false-negative diagnoses were retrospectively identified. In the majority of cases, the possibility of atrophic gastritis was not raised by the submitting physician, and the endoscopic findings were not specific. Therefore, accurate diagnosis requires a high degree of suspicion on the part of the pathologist, and the diagnosis should be considered particularly in patients with a clinical history of other autoimmune diseases or iron-deficiency anemia. Key Words: atrophic gastritis, autoimmune gastritis, pediatrics (Am J Surg Pathol 2015;39:786–792)
From the *Department of Pathology and Laboratory Medicine, The Children’s Hospital of Philadelphia, Philadelphia, PA; wDepartment of Gastroenterology and Nutrition; and zDepartment of Pathology, Beth Israel Deaconess Medical Center, Children’s Hospital Boston, and Harvard Medical School, Boston, MA. Presented in abstract form at the 2014 Annual Meeting of the Society of Pediatric Pathologists, San Diego, CA. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Jeffrey D. Goldsmith, MD, Department of Pathology, Beth Israel Deaconess Medical Center, 330 Brookline Ave., Boston, MA 02215 (e-mail: [email protected]). Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
786 | www.ajsp.com
G
astric atrophy is defined as the loss of gastric oxyntic and/or antral glands,1,2 which may be replaced by fibrous tissue and/or metaplastic epithelium, and typically includes increased lamina propria inflammation. The 2 common causes of atrophic gastritis are Helicobacter pylori with antral or multifocal atrophy and autoimmune atrophic gastritis with atrophy of the oxyntic mucosa. Although autoimmune atrophic gastritis is classically a disease of elderly adults of northern European ethnic background with B12 deficiency, recent studies have described the disease in younger adults, particularly in those with other autoimmune diseases and iron-deficiency anemia.3–6 Atrophic gastritis in pediatrics is a rare and possibly underdiagnosed entity. The majority of studies have examined H. pylori-associated antral atrophy in children located in developing countries,7 whereas autoimmune atrophic gastritis has been primarily reported as singlecase reports in patients with immunodeficiencies, some of whom subsequently developed adenocarcinoma.8–10 Most recently, a series of 8 pediatric patients with refractory iron-deficiency anemia and atrophic gastritis was reported11 without associated autoimmune conditions; 4 of these patients showed H. pylori infection. Rare cases of a relatively recently described entity called “atrophic autoimmune pangastritis” have also been reported in children, in which both the antral and oxyntic mucosae demonstrate intense inflammation with complete atrophy of specialized cells.12 This retrospective study of atrophic gastritis not associated with H. pylori was performed to further expand the knowledge of the clinical presentation, pathologic findings as seen on endoscopically obtained mucosal biopsies, and natural history of this disease in the pediatric population.
METHODS With institutional review board approval (protocol number IRB-P00009596), the pathology archives were searched for patients with a diagnosis of atrophic gastritis from 1994 to 2013. Exclusion criteria included any laboratory or histologic evidence of Helicobacter infection and histologic evidence of collagenous gastritis. All gastric and concurrently obtained duodenal biopsies were reviewed. Immunohistochemical analyses for Helicobacter, gastrin, and chromogranin were performed on all biopsies using the Ventana Benchmark Ultra or XT platforms (Tucson, Am J Surg Pathol
Volume 39, Number 6, June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Am J Surg Pathol
Volume 39, Number 6, June 2015
Pediatric Non-H. pylori Atrophic Gastritis
AZ) with the following prediluted antibodies: (1) gastrin (mouse polyclonal; Ventana Medical Systems); (2) helicobacter (rabbit polyclonal; Cell Marque, Rocklin, CA); (3) chromogranin A (mouse clone LK2H10; Ventana Medical Systems). Biopsy location was defined on the basis of an endoscopic report and was confirmed by the presence of G cells in the antrum or the absence of G cells in the oxyntic mucosa using gastrin immunohistochemistry in areas devoid of intestinal metaplasia. Hematoxylin and eosin slides were scored for mucosal atrophy, intestinal metaplasia, pseudopyloric metaplasia, lymphoplasmacytic inflammation, and neutrophilic inflammation on the basis of the visual analogue updated Sydney system.13 The intramucosal distribution of chronic inflammation was evaluated as described14 for basal or fullthickness mucosal inflammation. Increased intraepithelial lymphocytes were defined as >25 intraepithelial lymphocytes/100 epithelial cells.13 Eosinophils were counted in a 400 field in the area of highest density. Enterochromaffin-like cell (ECL) hyperplasia was evaluated on chromogranin immunostain by standard criteria.15 Epithelial apoptosis was quantified as absent, rare ( 100 5 80 51 41 10
ECL Intraepithelial Hyperplasia Apoptosis Lymphocytosis N L, N L L N L, N L, N NA NA L, N
Rare Rare Rare Rare Rare Rare + NA Rare
+B + + NA +
Follow-up biopsies are in gray. +++ indicates marked; ++, moderate; +, mild; B, basal; C, complete; CI, chronic inflammation; D, diffuse; I(c), complete intestinal metaplasia; L, linear; M, mucinous metaplasia; N, micronodular; NA, not available; P, partial; Py, pseudopyloric metaplasia; S, squamous metaplasia; S, superficial.
panel of other autoantibody tests and the finding of elevated gastrin levels in the workup for chronic diarrhea. Atrophic gastritis was unexpected in the other 9 patients. Six of the 12 patients had iron-deficiency anemia at presentation, although this was often multifactorial and not attributable solely to gastritis. In the 3 patients who were tested, H. pylori IgG was negative. In no case was the stomach described as atrophic, ulcerated, or eroded at endoscopy. Laboratory markers associated with autoimmune gastritis were ordered inconsistently after biopsy results: antiparietal cell antibodies were positive in 3 of 7 patients, and anti–intrinsic factor antibodies were positive in 2 of 6 patients, with a total of 4 patients positive for at least 1 of these 2 markers. In patients with an unexpected diagnosis of atrophic gastritis, follow-up laboratory testing for vitamin B12 levels, when performed, was within normal limits; however, only 9 patients were tested.
Histologic Data For 10 patients, both the antral and oxyntic mucosa were biopsied at the time of initial endoscopy. The remaining 2 patients had initial biopsies from either the oxyntic (patient #11) or antral (patient #7) mucosa. In the patient with only antral biopsies, a follow-up endoscopy included a biopsy of the corpus (Table 2). On initial biopsy, 10 patients had mucosal atrophy present only in the oxyntic mucosa (Fig. 1). One patient with oxyntic mucosal atrophy (#3) developed atrophy of the antrum on follow-up biopsy. In addition to loss of oxyntic cells, pseudopyloric metaplasia in the oxyntic mucosa was identified in 3 cases at initial diagnosis,
788 | www.ajsp.com
whereas no patient showed intestinal metaplasia at presentation. Patient #7, who had a history of cytomegalovirus gastritis with complete loss of normal antral and oxyntic glands, showed extensive squamous and mucinous metaplasia (Fig. 2). Apoptotic epithelial cells were not a significant histologic finding in any biopsy. Of the 7 cases with follow-up biopsies (taken 9 mo to 10 y after index endoscopy; mean 49 mo), 3 cases showed persistent metaplastic findings (2 cases with pseudopyloric metaplasia and 1 case with squamous/mucinous metaplasia). Three of 4 patients without metaplasia developed complete intestinal metaplasia on follow-up, including 1 patient who developed complete intestinal metaplasia in both the oxyntic and antral mucosae (Fig. 3); incomplete intestinal metaplasia was not seen. Inflammation was usually more severe in the corpus than in the antrum but was occasionally of similar intensity. Antral-predominant inflammation was not identified. In the oxyntic mucosa, the intramucosal distribution of chronic inflammation was always present in a fullthickness or basal-predominant pattern and was moderate or marked in 10 cases. Neutrophilic inflammation was present in 7 patients and eosinophils were increased in all but 1 case (Fig. 4). Diffusely increased intraepithelial lymphocytes were present in 2 patients. Chronic inflammation persisted in all patients who had follow-up biopsies. Immunohistochemistry for Helicobacter was negative in all cases. Linear or micronodular ECL hyperplasia was seen in the oxyntic mucosa in 4 cases initially and in 1 additional patient on follow-up (Fig. 5). Patient #3 developed a gastric adenocarcinoma at 17 years of age, 8 Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Am J Surg Pathol
Volume 39, Number 6, June 2015
Pediatric Non-H. pylori Atrophic Gastritis
TABLE 2. (continued) Antral Mucosa Type of Atrophy Metaplasia 1 1 2 2 3 3 4 4 5 5 6 7 7 8 9 10 10 11 12
P NA P NA NA
I(c) NA S, M NA NA
Extent of Metaplasia
Chronic Inflammation
Distribution of Mucosal CI
Neutrophilic Inflammation
Maximum Eosinophils/ HPF
Apoptosis
Intraepithelial Lymphocytosis
++ NA ++ NA NA
+++ + + +++ + ++ NA + + ++ NA + + + NA +
B S NA NA D D S NA D NA B D NA D D NA S NA D
+ + NA NA NA
78 2 1 4 12 64 36 NA 24 19 2 21 NA 48 6 6 13 NA 1
Rare Rare + NA Rare Rare NA NA
+ NA + NA NA +
years after initial diagnosis. Patient #4 died of graft versus host disease after stem cell transplantation for lymphoma. Gastric neuroendocrine tumors were not seen. In the absence of G cells as displayed immunochemically, false-negative diagnoses were retrospectively identified in 2 cases. These cases, which were designated by the endoscopists as having an origin from either the corpus or fundus, were initially diagnosed as “antral” or “pyloric-type” mucosa with chronic gastritis. In patient #3, this diagnosis was repeated until the development of B12 deficiency, at which time a correct diagnosis of atrophic gastritis was rendered. Retrospective gastrin staining of the previous biopsies on patient #3 was negative, indicating that the corpus/fundus area was sampled, and the previous biopsies did show atrophy. Examination of the concurrently obtained duodenal biopsies was not contributory except for patient #3, whose duodenal biopsies showed changes consistent with celiac disease.
DISCUSSION In contrast to previous pathologic case series in developing countries that focused on H. pylori-associated pediatric atrophic gastritis, we identified 12 H. pylorinegative pediatric patients with atrophic gastritis predominantly affecting the corpus, which is, to our knowledge, the largest retrospective case series to date. This is a very rare diagnosis in the pediatric population, with only 12 patients identified over 20 years in an institution with a large gastrointestinal biopsy volume (approximately 3200 gastrointestinal cases accessioned in 2013). In the majority of cases, the possibility of atrophic Copyright
r
gastritis was not raised by the submitting physician, and the endoscopic findings were not specific. Therefore, accurate diagnosis requires a high degree of suspicion on the part of the pathologist, and the diagnosis should be considered particularly in patients with a clinical history of other autoimmune diseases or iron-deficiency anemia. The largest subset of patients in our series was diagnosed with autoimmune atrophic gastritis, but even among this cohort an exact diagnosis was not always possible. Large adult series often use the presence of body-predominant inflammation with the detection of anti–parietal cell or anti–intrinsic factor antibodies as inclusion criteria for autoimmune atrophic gastritis14,16; however, autoantibodies are not 100% sensitive,17 and patients who are initially negative may test positive later in their disease course.16 A small series of patients with “atrophic autoimmune pangastritis” has been described12 and includes several pediatric patients. The main findings in this entity included a marked inflammatory infiltrate in both the antrum and corpus with neutrophilic activity and increased apoptotic cells despite a complete absence of oxyntic cells, the absence of ECL hyperplasia, and a lack of H. pylori infection. Although several of our cases showed no evidence of ECL hyperplasia, a diagnosis of autoimmune pangastritis is excluded by the absence of atrophy and significant inflammation of the antrum in these cases. In our study, antral inflammation was variable, and in patients who had follow-up biopsies the degree of acute and chronic inflammation was often different from the original biopsy findings, making it concerning to use this as a criterion for disease classification.
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
www.ajsp.com |
789
Pogoriler et al
Am J Surg Pathol
Volume 39, Number 6, June 2015
FIGURE 1. Typical case of atrophic gastritis as seen in this biopsy from the gastric corpus. Complete loss of oxyntic cells are seen with full-thickness lymphoplasmacytic inflammation. A gastrin immunostain on this biopsy was negative, which confirms the site of origin (hematoxylin and eosin).
However, in no case was there an improvement in atrophy; indeed, on follow-up more patients developed metaplasia, which is consistent with the known progression of typical autoimmune atrophic gastritis. Clinical follow-up of our patients was quite variable. Those with other autoimmune diseases often had follow-up biopsies; however, even in these patients, assays for autoantibodies, serum gastrin, or B12 were often not performed. In 1 case, the atrophy was attributed to excess NSAID use, and further follow-up was not performed. We have not identified any cases in the literature in which chronic NSAID use was a cause for atrophic gastritis.
FIGURE 2. This unusual pattern of “squamo-mucinous” metaplasia was seen in case number 7; the patient had a history of immunodeficiency and chronic cytomegalovirus infection. Normal oxyntic glands were not present (hematoxylin and eosin).
790 | www.ajsp.com
FIGURE 3. This biopsy of the gastric corpus showed combined complete intestinal metaplasia and pseudopyloric metaplasia (hematoxylin and eosin).
One unique patient with a history of cytomegalovirus gastritis, immunodeficiency, and gastro-cutaneous fistulas showed a highly unusual pattern of squamous and mucinous metaplasia, which was originally identified in the antrum and persisted on follow-up. Whether this
FIGURE 4. Few cases showed patchy increase in lamina propria eosinophils as seen in the center-bottom portion of this biopsy of the corpus (hematoxylin and eosin). Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Am J Surg Pathol
Volume 39, Number 6, June 2015
Pediatric Non-H. pylori Atrophic Gastritis
FIGURE 5. Entrochromaffin-like cell (ECL) hyperplasia was seen in approximately half of the cases. A, In a few cases, micronodular hyperplasia was appreciated on the H&E stain (H&E). B, This impression was confirmed using chromogranin immunohistochemistry, which confirmed the presence of both linear and micronodular ECL hyperplasia (chromogranin immunohistochemistry). H&E indicates hematoxylin and eosin.
pattern reflects her young age or the unusual clinical circumstances is unknown. A limitation of this study was our dependence on the prior diagnosis of atrophy to identify cases, and, thus, we likely overlooked cases that were misdiagnosed. The majority of patients in our series had complete or subtotal atrophy, and specimens with partial atrophy would likely be overlooked. On the basis of history and follow-up biopsies, we did identify 2 patients with complete atrophy who did not receive a correct histologic diagnosis on initial biopsy. Beyond the known interobserver variability in grading atrophy,1,18 the ability to make a diagnosis of atrophy can be compromised by the receipt of specimens designated simply as “stomach” as well as by poor orientation or sampling of the oxynto-antral transition zone. With complete loss of oxyntic glands, atrophic corpus may easily be mistaken for antral mucosa with chronic gastritis, and areas with partial atrophy may be indistinguishable from areas of transitional mucosa with chronic gastritis. Given the relative rarity of atrophic gastritis in pediatric biopsies, the finding of “antral” mucosa in a specimen labeled “corpus” may be more likely to reflect inaccurate sampling than atrophic gastritis of the corpus, particularly in the absence of chronic inflammation. However, findings that should raise the suspicion for atrophic gastritis include a full-thickness or basal-predominant pattern of chronic inflammation with or without intestinal metaplasia. In many cases, gastrin immunohistochemistry can be helpful to clarify the location of the biopsy, as the antrum should have a prominent mid-zone band of gastrin-positive cells. Some limitations to this approach include the facts that rare, scattered gastrin-positive cells may be present in the oxyntic mucosa, and variable numbers of these cells may be present in the oxynto-antral transition zone. In addition, Copyright
r
gastrin-positive cells may be increased in areas of pseudopyloric metaplasia, and they are often absent in areas of intestinal metaplasia.19 Conversely, gastrin-positive cells in the antrum may be slightly decreased in areas of chronic gastritis.20 Therefore, even with a gastrin stain, the appropriate localization of a biopsy may be challenging. Although eosinophils are often prominent in pediatric inflammatory disorders and were increased in the majority of the biopsies from our patients with atrophic gastritis, this feature is not specific and is likely secondary to the markedly increased lymphoplasmacytic infiltrate in a majority of study patients. A recent series of adult biopsies with autoimmune atrophic gastritis noted >30 eosinophils per high-power field in 46% of patients and neutrophilic cryptitis in 44%.14 Long-term consequences of atrophic gastritis include adenocarcinoma and neuroendocrine tumors. In adults with atrophic gastritis, the risk for adenocarcinoma is estimated to be increased by 3-fold and the risk for gastric neuroendocrine tumor by 13-fold.17 Consistent with this rare event, 1 of our patients developed an adenocarcinoma during the follow-up period of 10 years. This was the patient in whom the diagnosis of atrophy was initially missed but who subsequently received frequent follow-up and monitoring for dysplasia. Rare cases of gastric adenocarcinoma in teenagers have been reported10 that arose in a background of atrophic gastritis. ECL hyperplasia in the gastric corpus is a result of the hypergastrinemia induced by the hypochlorhydric state and is the histologic precursor of neuroendocrine tumors that arise in the context of atrophic gastritis. The elevated serum gastrin levels are secondary to a loss of the negative signal imposed on G cells by the presence of gastric acid. Even in early atrophic gastritis before the complete loss of oxyntic glands is apparent, ECL hyper-
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
www.ajsp.com |
791
Pogoriler et al
Am J Surg Pathol
plasia is often found.16 Some level of ECL hyperplasia was appreciated in 5 of our study patients; none of our study patients showed evidence of neuroendocrine tumors. In conclusion, atrophic gastritis is an extremely rare diagnosis in the pediatric population whose histology is reminiscent to that in adults. In those cases in which H. pylori infection has been excluded, autoimmunity is the most likely etiology. Given its extreme rarity in children, a high level of suspicion is required for accurate diagnosis. REFERENCES 1. Rugge M, Genta RM. Staging and grading of chronic gastritis. Hum Pathol. 2005;36:228–233. 2. Rugge M, Correa P, Dixon MF, et al. Gastric mucosal atrophy: interobserver consistency using new criteria for classification and grading. Aliment Pharmacol Ther. 2002;16:1249–1259. 3. Centanni M, Marignani M, Gargano L, et al. Atrophic body gastritis in patients with autoimmune thyroid disease: an underdiagnosed association. Arch Intern Med. 1999;159:1726–1730. 4. De Block CE, De Leeuw IH, Van Gaal LF. High prevalence of manifestations of gastric autoimmunity in parietal cell antibodypositive type 1 (insulin-dependent) diabetic patients. The Belgian Diabetes Registry. J Clin Endocrinol Metab. 1999;84:4062–4067. 5. Marignani M, Delle Fave G, Mecarocci S, et al. High prevalence of atrophic body gastritis in patients with unexplained microcytic and macrocytic anemia: a prospective screening study. Am J Gastroenterol. 1999;94:766–772. 6. Hershko C, Hoffbrand AV, Keret D, et al. Role of autoimmune gastritis, Helicobacter pylori and celiac disease in refractory or unexplained iron deficiency anemia. Haematologica. 2005;90:585–595. 7. Ricuarte O, Gutierrez O, Cardona H, et al. Atrophic gastritis in young children and adolescents. J Clin Pathol. 2005;58:1189–1193. 8. Vaterlaws AL. Gastric atrophy in childhood. Arch Dis Child. 1969;44:710–714.
792 | www.ajsp.com
Volume 39, Number 6, June 2015
9. Russell AC, Black JO, Schwartz DA, et al. 15-year-old girl with metaplastic atrophic gastritis and enterochromaffin-like cell hyperplasia. J Pediatr Gastroenterol Nutr. 2012;55:e148–e151. 10. Slotta JE, Heine S, Kauffels A, et al. Gastrectomy with isoperistaltic jejunal parallel pouch in a 15-year-old adolescent boy with gastric adenocarcinoma and autosomal recessive agammaglobulinemia. J Pediatr Surg. 2011;46:e21–e24. 11. Miguel N, Costa E, Santalha M Jr, et al. Refractory iron-deficiency anemia and autoimmune atrophic gastritis in pediatric age group: analysis of 8 clinical cases. J Pediatr Hematol Oncol. 2014;36: 134–139. 12. Jevremovic D, Torbenson M, Murray JA, et al. Atrophic autoimmune pangastritis: a distinctive form of antral and fundic gastritis associated with systemic autoimmune disease. Am J Surg Pathol. 2006;30:1412–1419. 13. Dixon MF, Genta RM, Yardley JH, et al. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol. 1996;20:1161–1181. 14. Bettington M, Brown I. Autoimmune gastritis: novel clues to histological diagnosis. Pathology. 2013;45:145–149. 15. Solcia E, Fiocca R, Villani L, et al. Hyperplastic, dysplastic, and neoplastic enterochromaffin-like-cell proliferations of the gastric mucosa. Classification and histogenesis. Am J Surg Pathol. 1995; 19(suppl 1):S1–S7. 16. Torbenson M, Abraham SC, Boitnott J, et al. Autoimmune gastritis: distinct histological and immunohistochemical findings before complete loss of oxyntic glands Mod Pathol. 2002;15:102–109. 17. Park JY, Lam-Himlin D, Vemulapalli R. Review of autoimmune metaplastic atrophic gastritis. Gastrointest Endosc. 2013;77:284–292. 18. Talebkhan Y, Mohammadi M, Rakhshani N, et al. Interobserver variations in histopathological assessment of gastric pathology. Pathology. 2009;41:428–432. 19. Rubio CA, Jaramillo E, Suzuki G, et al. Antralization of the gastric mucosa of the incisura angularis and its gastrin expression. Int J Clin Exp Pathol. 2009;2:65–70. 20. Sloan JM, Buchanan KD, McFarland RJ, et al. A histological study of the effect of chronic gastritis on gastrin cell distribution in the human stomach. J Clin Pathol. 1979;32:201–207.
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
