Cockburn, Amber N. MD*; Morgan, Christopher J. DO*; Genta, Robert M. MD, FACG*,†,‡
Since the first description of hormone-producing endocrine cells in the gastric mucosa of dogs by Heidenheim in 1870, the field has evolved slowly for almost a century, occasionally revitalized by the discovery of new staining methods that better allowed the visualization and categorization of the endocrine cells in the human gastrointestinal mucosa. In the last several decades the function of many of these cells has been elucidated, simple immunohistochemical methods for their identification have become available, and disease associations have emerged. However, except for the few with a special interest in the neuroendocrine cells of the stomach, pathologists have generally renounced keeping up with the arcane classifications of the proliferative lesions that lead from enterochromaffin-like (ECL) cell hyperplasia to neuroendocrine tumors (NET), formerly known as carcinoid tumors, and neuroendocrine carcinomas. If we consider that most of these proliferations occur in a background of atrophic gastritis—an entity with complex and often unfamiliar diagnostic criteria, we can readily understand why gastric neuroendocrine precursor lesions remain widely underdiagnosed until fully-formed easily recognizable NETs develop.
The aim of this review is to provide the general pathologist with simple, yet rigorous guidelines on how to recognize, classify, and diagnose the neuroendocrine proliferations found in the stomach. We will first present a succinct outline of the types and distribution of the neuroendocrine cells in the normal gastric mucosa. We will then discuss the most common situations in which the pathologist needs to think about gastric neuroendocrine cells. Finally, we will provide example cases and a series of algorithms that will guide through the specific diagnostic criteria for each of the entities discussed. The pathologist who follows our simple steps will be better aware of this neglected area of gastric pathology and will learn to suspect, recognize, and accurately diagnose the most common abnormalities of the neuroendocrine system in the stomach.
NEUROENDOCRINE COMPONENTS OF THE NORMAL STOMACH
There is a wide array of endocrine cells scattered through the gastric epithelium. These include ECL cells, histamine-producing cells that constitute a majority of the neuroendocrine cell population within the fundus, and 15% to 35% of the total gastric neuroendocrine component; G cells (gastrin); D cells (somatostatin); A cells (glucagon); enterochromaffin cells (serotonin); and X cells, also referred to as A-like cells (ghrelin). In addition, there is a number of other known or presumptive endocrine-type cells that have been characterized mainly by histochemical and ultrastructural methods and whose products and function remain unknown.
Only gastrin-producing (G cells) and ECL cells are easily visualized by widely available immunohistochemical methods, and are relevant to the practice of gastric pathology; therefore, we shall focus exclusively on these 2 cell types and refer the interested reader to the detailed works of Solcia’s group.1–3
Most G cells are located in the neck region of the mucus glands of the antrum, pylorus, and transitional zone, where they compose approximately 50% of the neuroendocrine cell mass (Fig. 1A).4 In addition, scattered G cells may also be found in the mucosa of the corpus and fundus in approximately 10% of gastric biopsy specimens, irrespective of the presence of atrophy (Fig. 1B).5,6 Gastrin is the main regulator of acid secretion, a function it accomplishes by 2 mechanisms: direct and indirect. The direct mechanism acts on the parietal cell’s basolateral membrane cholecystokinin B (CCKB) receptors. The indirect mechanism affects the ECL cell’s CCKB receptors that respond by releasing histamine, which in turn stimulates parietal cell acid secretion by binding to the parietal cell’s basolateral histamine2 receptor.7
ECL cells are small, irregularly shaped cells that are heavily argyrophilic with the Grimelius’, Bodian, and Sevier-Munger methods of staining.2,8–12 Although these immunochemical techniques have been crucial in the identification and characterization of endocrine cells, the advent of immunohistochemical staining has made them largely obsolete. In the stomach, synaptophysin, and chromogranin immunostains—although not completely specific, are perfectly adequate to visualize ECL cells and detect their abnormal proliferations. ECL cells, typically found in the gastric corpus, are rare or absent in other compartments of the stomach. However, when measured by immunoreactivity to the histamine-forming enzyme histidine decarboxylase and to the CCKB receptor, ECL cells were also identified in the transitional and pyloric region, and in pseudopyloric and intestinal metaplasia.6 In the oxyntic glands, ECL cells are scattered in the deep and intermediate regions; few are observed in the neck area and none in the surface epithelium (Fig. 2). ECL cells constitute the major, and only clinically relevant, endocrine cell population in the mucosa of the gastric corpus.
WHEN NEUROENDOCRINE CELLS ARE RELEVANT
There are 3 situations in which a good understanding of gastric neuroendocrine pathology is necessary. The simplest one is when an obvious NET (formerly carcinoid), described by the endoscopist as a nodule or polyp, is identified. Although the histopathologic diagnosis of this lesion rarely poses a diagnostic dilemma, its detection warrants an investigation of the background mucosa that has allowed its genesis. The second situation involves the evaluation of gastric atrophy, the milieu in which endocrine cell hyperplasia and dysplasia occur and progress. Finally, at a time when a third of the adults living in the industrialized world use proton pump inhibitors, iatrogenic hypergastrinemia has reached endemic proportions.13 Clinicians occasionally suspect and inquire about ECL-cell hyperplasia; pathologists must be able to confidently state whether the specimens are adequate to assess ECL-cell hyperplasia and—if present—whether it is significant.
The Unmistakable NET
The most evident situation in which a pathologist will need to assess the endocrine cell population in the gastric mucosa is when a NET is found. Irrespective of its size, one should try to answer the question of whether it represents a: NET arising from an atrophic background (type I); NET (type II) associated with Zollinger-Ellison (ZE) syndrome, multiple endocrine neoplasia type I (MEN-I),14,15 or hyperparathyroidism16; or a sporadic type III NET (Table 1).
If the mucosa surrounding a small NET is atrophic, with loss of oxyntic glands, pseudopyloric and intestinal metaplasia, and various degrees of chronic inflammation, then the tumor is likely a type I NET and a thorough evaluation for atrophic gastritis is warranted. Atrophic gastritis is responsible for generating the most gastric NETs; consequently, type I NETs represent 70% to 80% of all gastric NETs.17 Histologically, these tumors are composed of uniform cells with low proliferation index (grade 1, G1), are confined to the oxyntic mucosa, and tend to be multifocal (Fig. 3, G1). They arise in the setting of hypergastrinemia due to loss of negative inhibition caused by parietal cell atrophy.18,19
Type II gastric NETs are frequently detected as part of the work-up for MEN-1 syndrome or ZE syndrome.20 In both instances the tumors are usually small (<1 cm), have a low proliferation index, and show neither infiltrating nor pleomorphic features (Fig. 3, G1). In patients with MEN-1 syndrome the gastric mucosa is normal or mildly inflamed but not atrophic. The fundic mucosa of patients with ZE syndrome is often hypertrophic, with long densely packed oxyntic glands and no significant inflammation. This type of NET is the most uncommon, representing 5% to 8% of gastric NETs.17
Sporadic NETs occur in otherwise healthy gastric mucosa; the adjacent gastric mucosa may be normal or show chronic gastritis, but it is not atrophic. These tumors represent approximately 20% of all gastric NETs.17 Sporadic NETs are usually detected when they become symptomatic, either secondary to mucosal erosion and blood loss or metastasis. As these alerting events tend to occur only after the tumors reach a certain size, these tumors are usually larger than 1 cm, display infiltrating growth patterns with occasional areas of necrosis, and exhibit various degrees of pleomorphism (Fig. 3, G2). They are more aggressive than nonsporadic NETs and, consequently, have a generally poor prognosis.21
The WHO Classification of Tumours of the Digestive System (2010) gastrointestinal diseases emphasizes combining the proliferation index with the morphologic findings when grading a NET.22 The accepted methods of determining the proliferation index include counting mitoses and/or obtaining the Ki-67 index. Typically the NETs or the biopsies are too small to attain the requisite 50 high-power fields (HPF) needed for a mitotic count. If the specimen is adequate, the mitoses are graded as illustrated in Table 2. Because specimens are often small, the Ki-67 index is preferentially used in our practice. It is calculated by examining 500 to 2000 neoplastic cells and determining the percentage of Ki-67–positive neoplastic cells; counting should be performed in the areas of greatest labeling. Grade 1 (G1) tumors have a proliferation rate of ≤2% (Fig. 3, G1). Grade 2 (G2) tumors’ proliferation index ranges from 3% to 20% (Fig. 3, G2) and neuroendocrine carcinomas (G3) have a proliferation index >20% (Fig. 3, G3). An important caveat: as lymphocytes will stain intensely with Ki-67, and are often abundant within NETs, they must be identified and excluded from the assessment, which must be limited to neoplastic cells. Should the mitotic count and Ki-67 index differ, the higher grade is used.
GASTRIC MUCOSAL ATROPHY AND ATROPHIC GASTRITIS
Gastric mucosal atrophy is defined as the decrease of the structures present in the normal stomach.23,24 When the vanished structures (eg, oxyntic glands) are replaced by loose connective tissue or fibrosis the atrophy is classified as nonmetaplastic (Fig. 4A). Metaplastic atrophy is characterized by the replacement of the native glands with either intestinal-type epithelium (intestinal metaplasia, Fig. 4B) or pyloric-type glands (pseudopyloric metaplasia, Fig. 4C). Mucosal atrophy is simply a histopathologic finding.25,26 In contrast, atrophic gastritis is a condition of the stomach characterized by extensive areas of atrophy, metaplastic or not, that may be patchy and involve both the antrum and the corpus (multifocal atrophic gastritis, almost always a consequence of long-standing Helicobacter infection, Fig. 5A) or exclusively the corpus (corpus-restricted atrophic gastritis, virtually synonymous with autoimmune atrophic gastritis, Fig. 5B).24,27,28 The diagnosis of atrophic gastritis requires that at least 2 identifiable specimens from the antrum and 2 from the corpus be available. As the separate submission of specimens as per the Sydney System guidelines rarely occurs in clinical practice and particularly difficult situations occur when a clinician only submits multiple gastric biopsy fragments in a single formalin container, we have prepared a set of simple algorithms that should help decide how to approach most sets of suboptimally sampled gastric biopsy specimens (Fig. 6).
When adequate sampling (2 separate and topographically labeled specimens each from antrum and corpus) is available, foci of atrophy, metaplastic or not, in both gastric compartments indicate multifocal atrophic gastritis, likely related to H. pylori infection. In these cases ECL-cell proliferations are exceedingly rare and, unless obvious neuroendocrine nests are seen on the hematoxylin and eosin–stained sections, there is no need to perform special stains for neuroendocrine cells. If the antrum is either normal or shows reactive gastropathy with mild or moderate chronic inflammation only,25 and specimens from the corpus show atrophy, particularly with intestinal and pseudopyloric metaplasia, then a diagnosis of autoimmune atrophic gastritis is extremely likely. As autoimmune gastritis is often associated with ECL-cell proliferations and NETs, a work-up with immunohistochemical staining of both antrum and corpus is suggested.29–31
If only 1 sample from a stomach is available and it shows mucosal atrophy with or without metaplasia: if focal it should be diagnosed as “gastric mucosal atrophy,” whereas if diffuse it may be called “chronic atrophic gastritis.” If the specimen is labeled antrum, and its histologic features corroborate its antral origin, there is little that can be added regarding specific etiology, as it may be from the site of a healed ulcer, be associated with reactive gastropathy, or be part of multifocal atrophic gastritis. In contrast, if the specimen is clearly of corpus origin (with residual oxyntic glands) or shows pseudopyloric metaplasia, the possibility of autoimmune atrophic gastritis should be mentioned and additional sampling suggested to the clinician.
THE PATIENT ON LONG-TERM PROTON PUMP INHIBITORS
The normal range of serum gastrin levels depends on the method and laboratory, but is generally <150 pg/mL. Very high levels of gastrin (eg, >400 pg/mL) were previously considered indicative of ZE syndrome or end-stage atrophic gastritis, where the gastric corpus has lost most or all the parietal cells and antral G cells secrete ever increasing quantities of gastrin in a futile attempt to stimulate acid secretion.13 In most subjects, chronic use of histamine2-receptor antagonists and proton pump inhibitors (PPI) is associated with a slight increase in serum gastrin. However, about 20% to 25% of chronic PPI users develop modest degrees of hypergastrinemia (200 to 400 pg/mL) and approximately 1% per year, particularly patients with H. pylori infection, develop significant hypergastrinemia (>400 pg/mL).32 Histologically, most chronic PPI users have no significant changes in either G cells or ECL cell density and distribution. However, some patients develop ECL cell hyperplasia (usually no greater than linear, rarely micronodular). NETs may occur in PPI users, and some authors have speculated that these NETs, and even neuroendocrine carcinomas, are caused by the acid suppression.33,34 Given the billions of PPI prescriptions filled in the last 25 years by millions of patients worldwide and the lack of a corresponding documented increase in the incidence of gastric NETs, it seems more likely that the few reported cases represent the chance occurrence of 1 rare (NET) and 1 common (PPI use) event.
The density of G cells may increase, both in the stomach and the duodenum, as a consequence of acid suppression.35 The changes, however, are essentially impossible to detect without applying sophisticated counting techniques and an extensive bank of normal data collected with identical methods. Therefore, we suggest that no comments be made on the numbers or density of G cells in response to clinicians’ requests to evaluate the effects of PPIs on neuroendocrine cells.
ECL CELL PROLIFERATIONS IN ATROPHY—HYPERPLASIA, DYSPLASIA, AND NEUROENDOCRINE TUMORS
In chronically hypergastrinemic states, gastrin stimulates ECL cells to proliferate. The proliferations run the spectrum from simple hyperplasia to frank NET (previously discussed) that is described by the Solcia classification published in 1988.36
ECL-cell simple (or diffuse) hyperplasia is defined as ECL-cell density >2 SDs of the normal density in matched controls. The cells are hypertrophied and arranged singly or in clusters of fewer than 5 cells, usually in the lower third of the gastric pits. Because quantification is difficult, the causes are numerous and a visual understanding of “normal” is not readily available, we do not estimate or comment on simple hyperplasia. ECL-cell linear hyperplasia is defined as at least 2 linear groups of 5 consecutive neuroendocrine cells lining a gland per millimeter, or 2 linear groups in 1 HPF (Fig. 7A). This is typically found in the base or lower portions of the pits, but may be seen in the glandular neck region. Linear hyperplasia is highlighted by neuroendocrine immunohistochemical stains and should be mentioned in the pathology report. ECL-cell micronodular hyperplasia (Fig. 7B) consists of clusters of 5 or more neuroendocrine cells, bounded by basement membrane, that do not exceed the diameter of a gastric gland (<150 μm). The clusters may be grouped or scattered through the lamina propria and typically have a wide distribution in the oxyntic mucosa. When aggregates of 5 or more micronodules cluster it is termed adenomatoid hyperplasia (Fig. 7C). Adenomatoid hyperplasia typically forms in the lower third of the mucosa. Several studies have concluded that linear, micronodular, and adenomatoid hyperplasia have low potential for progression to NETs.10,37
Dysplasia occurs when the micronodules fuse with loss of the basement membrane or an individual nodule is >150 μm. The dysplastic nodules may progress to become microinvasive when newly formed stroma can be seen within the lamina propria (Fig. 7D). Measurement may be performed using a calibrated micrometer eyepiece and calculating the size for the specific objective or using a Vernier calibrated gauge. As these lesions have the potential to progress to NETs, they should be identified and mentioned in the pathology report. Figure 8 shows an atrophic oxyntic mucosa in which all types of hyperplasia and dysplasia are present.
Micro-NETs (microcarcinoids) are nodules of ECL cells with a diameter ranging from 500 μm (0.5 mm) to 0.5 cm that are not identified by endoscopy. They are often found in biopsies labeled as “representative,” “random,” or equivalent generic designations. The identification of a NET in one of these should prompt further investigations, including a thorough evaluation of the adjacent mucosa.
Invasive NETs involve the mucosa and penetrate the muscularis mucosae to infiltrate the submucosa. However, gastric biopsies often include mucosa, limited muscularis mucosae and minimal or no submucosa, essentially precluding a diagnosis of invasive NET. The most useful and clinically relevant information we recommend including in all reports regarding NETs, regardless of the etiology, comprises: size, proliferation index, grade, and—when sufficient submucosa is available—invasion. We also state whether the NET extends to the biopsy edge. Neuroendocrine proliferations that are >500 μm and endoscopically identified are classified as NETs (see the Neuroendocrine components of the normal stomach section).
Practical Suggestions for the Evaluation of Neuroendocrine Proliferations
To work-up suspected proliferations, we recommend ordering immunostains. A single neuroendocrine marker and gastrin are sufficient for working up ECL-cell hyperplasia. In the evaluation of ECL-cell hyperplasia or dysplasia it is not necessary to assess the proliferation index. In contrast, the work-up of NETs includes ordering Ki-67 and chromogranin, synaptophysin, or CD56.
Received 4 biopsies in 1 container labeled “antrum and body.” No fragments have unequivocal oxyntic mucosa.
Approach: We recommend ordering gastrin and at least 1 neuroendocrine immunostain. Gastrin stain will help distinguish antrum from body. Synaptophysin or chromogranin will highlight endocrine cell hyperplasia. These stains will also highlight G cells and should be compared with the gastrin stain to avoid incorrectly diagnosing linear hyperplasia in biopsies from the antrum.
Suggested diagnosis: Chronic atrophic gastritis. See comment.
Comment: Sections show body-type mucosa (confirmed by gastrin stain) with loss of oxyntic glands, pyloric gland metaplasia, and intestinal metaplasia. Synaptophysin and chromogranin stains highlight linear and micronodular hyperplasia. The antral mucosa shows no significant histopathologic abnormalities. This pattern of atrophy likely represents autoimmune atrophic gastritis. Correlation with clinical and serologic findings, including antiparietal and anti-intrinsic factor antibodies, is recommended.
Received biopsies labeled “gastric nodules.” The endoscopy report does not quantify the number of nodules seen and the gastroenterologist is unavailable. Multiple sections show nodular neuroendocrine proliferations >500 µm. No biopsies are available to assess the background mucosa.
Approach: Order gastrin, synaptophysin, chromogranin, and Ki-67. Determine proliferation index.
Suggested Diagnosis: NET, G1 (carcinoids). See comment.
Comment: Sections show 3 fragments of body-type mucosa (confirmed by gastrin stain) with NET(s) ranging in size from <1 to 2 mm. Each tumor extends to the biopsy edge. The Ki-67 proliferation index is <2%, confirming grade 1. No muscularis mucosae is available for assessment of invasion. These biopsies may represent multiple NET or fragments of a single mass. Multiple NET may arise in the setting of autoimmune chronic atrophic gastritis, ZE syndrome, or MEN-1 syndrome. Clinical and endoscopic correlation and topographically defined biopsy sampling of both antrum and corpus mucosa are recommended.
Received biopsies from a single gastric nodule with accompanying, separately submitted mucosal biopsies. Sections show fragments of NET with uniform nuclei and cells arranged in nests. There is no obvious invasion, however the muscularis mucosae is limited. The separately submitted mucosa shows unremarkable oxyntic mucosa.
Approach: Order synaptophysin, chromogranin, and Ki-67 (if fewer than 50 HPF available for assessment of proliferation index). Ki-67 proliferation index is 3% to 20%.
Suggested Diagnosis: NET. See comment.
Comment: Sections show neuroendocrine cells, highlighted by synaptophysin and chromogranin stains, with small, uniform nuclei arranged in a nested pattern. No invasion is identified; however the specimen is limited by scant muscularis mucosae. The tumor extends to the biopsy edges. The Ki-67 proliferation index is 5%, qualifying this tumor as grade 2. Grade 2 tumors are considered low-grade malignant tumors, despite the histologic differentiation. Patient work-up should include ruling out local metastases.
Received biopsies from an ulcerated mass in the fundus. Sections show a diffuse infiltrate of single cells with large nuclei, stippled chromatin, prominent nucleoli, and a moderate amount of cytoplasm. No definite adenocarcinoma is identified. No background mucosa is available for assessment.
Approach: Order synaptophysin, chromogranin, and Ki-67. Depending on results, pancytokeratin, CD-56, and melanoma markers (S100, MART-1) may be helpful.
Suggested diagnosis: Neuroendocrine carcinoma. See Comment.
Comment: Sections show a diffuse infiltrate of single cells with large nuclei, stippled chromatin, prominent nucleoli, and a moderate amount of cytoplasm. No adenocarcinoma is identified. The neoplastic cells are immunoreactive for synaptophysin and chromogranin. The Ki-67 proliferation index is 80%. Overall, these findings are consistent with a neuroendocrine carcinoma. No angioinvasion is identified.
Neuroendocrine proliferations may be daunting and the terminology is often confusing, especially with the new classification. It is important to approach gastric specimens systematically. To do so, one must remain alert for the mucosal milieu in which ECL-cell hyperplasia will most likely occur. Robertson Davies wrote “the eye sees only what the mind is prepared to comprehend.”38 This review should aid in the understanding of ECL cell proliferations so that the pathologist may recognize and appreciate the importance of ECL cell proliferations.
1. La RS, Inzani F, Vanoli A, et al. Histologic characterization and improved prognostic evaluation of 209 gastric neuroendocrine neoplasms. Hum Pathol. 2011;42:1373–1384
2. Solcia E, Capella C, Vassallo G, et al. Endocrine cells of the gastric mucosa. Int Rev Cytol. 1975;42:223–286
3. Solcia E, Fiocca R, Rindi G, et al. The pathology of the gastrointestinal endocrine system. Endocrinol Metab Clin North Am. 1993;22:795–821
4. Owen DASternberg SS Histology for Pathologists. 19972nd ed Philadelphia Lippincott Williams & Wilkins:481–494
5. 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
6. Tanaka-Shintani M, Watanabe M. Immunohistochemical study of enterochromaffin-like cell in human gastric mucosa. Pathol Int. 2007;57:572–583
7. Zhao CM, Chen D. The ECL cell: relay station for gastric integrity. Curr Med Chem. 2012;19:98–108
8. Capella C, Vassallo G, Solcia E. Light and electron microscopic identification of the histamine-storing argyrophil (ECL) cell in murine stomach and of its equivalent in other mammals. Z Zellforsch Mikrosk Anat. 1971;118:68–84
9. Solcia E, Capella C, Vassallo G. On the staining of the gastrin cell. Gastroenterology. 1971;61:794–796
10. Solcia E, Fiocca R, Villani L, et al. Morphology and pathogenesis of endocrine hyperplasias, precarcinoid lesions, and carcinoids arising in chronic atrophic gastritis. Scand J Gastroenterol Suppl. 1991;180:146–159
11. Vassallo G, Capella C, Solcia E. Grimelius’ silver stain for endocrine cell granules, as shown by electron microscopy. Stain Technol. 1971;46:7–13
12. Vassallo G, Capella C, Solcia E. Endocrine cells of the human gastric mucosa. Z Zellforsch Mikrosk Anat. 1971;118:49–67
13. Graham DY, Genta RM. Long-term proton pump inhibitor use and gastrointestinal cancer. Curr Gastroenterol Rep. 2008;10:543–547
14. Hemminki K, Li X. Familial carcinoid tumors and subsequent cancers: a nation-wide epidemiologic study from Sweden. Int J Cancer. 2001;94:444–448
15. Pieterman CR, Vriens MR, Dreijerink KM, et al. Care for patients with multiple endocrine neoplasia type 1: the current evidence base. Fam Cancer. 2011;10:157–171
16. Christopoulos C, Balatsos V, Rotas E, et al. The syndrome of gastric carcinoid and hyperparathyroidism: a family study and literature review. Eur J Endocrinol. 2009;160:689–694
17. Kulke MH, Anthony LB, Bushnell DL. NANETS treatment guidelines: well-differentiated neuroendocrine tumors of the stomach and pancreas. Pancreas. 2010;39:735–752
18. Freston JW, Hisada M, Peura DA, et al. The clinical safety of long-term lansoprazole for the maintenance of healed erosive oesophagitis. Aliment Pharmacol Ther. 2009;29:1249–1260
19. Haber MM, Hunt B, Freston JW, et al. Changes of gastric histology in patients with erosive oesophagitis receiving long-term lansoprazole maintenance therapy. Aliment Pharmacol Ther. 2010;32:83–96
20. von Rosenvinge EC, Wank SA, Lim RM. Gastric masses in multiple endocrine neoplasia type I-associated Zollinger-Ellison syndrome. Gastroenterology. 2009;137:537
21. Delle FG, Capurso G, Annibale B, et al. Gastric neuroendocrine tumors. Neuroendocrinology. 2004;80(suppl 1):16–19
22. Solcia E, Arnold R, Capella C, et al.Bosman FT, Carneiro F, Hruban RH, et al. Tumours of the stomach: neuroendocrine neoplasms of the stomach. WHO Classification of Tumours of the Digestive System. 20102nd ed Lyon IARC:64–68
23. Rugge M, Correa P, Di MF, et al. OLGA staging for gastritis: a tutorial. Dig Liver Dis. 2008;40:650–658
24. Ruiz B, Garay J, Johnson W, et al. Morphometric assessment of gastric antral atrophy: comparison with visual evaluation. Histopathology. 2001;39:235–242
25. Genta RM. Atrophy and atrophic gastritis: one step beyond the Sydney system. Ital J Gastroenterol Hepatol. 1998;30(suppl 3):S273–S275
26. Genta RM. Review article: gastric atrophy and atrophic gastritis--nebulous concepts in search of a definition. Aliment Pharmacol Ther. 1998;12(suppl 1):17–23
27. 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
28. 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
29. Annibale B, Azzoni C, Corleto VD, et al. Atrophic body gastritis patients with enterochromaffin-like cell dysplasia are at increased risk for the development of type I gastric carcinoid. Eur J Gastroenterol Hepatol. 2001;13:1449–1456
30. Genta RM. One more dysplasia. Eur J Gastroenterol Hepatol. 2001;13:1411–1414
31. Vannella L, Lahner E, Annibale B. Risk for gastric neoplasias in patients with chronic atrophic gastritis: a critical reappraisal. World J Gastroenterol. 2012;18:1279–1285
32. Orlando LA, Lenard L, Orlando RC. Chronic hypergastrinemia: causes and consequences. Dig Dis Sci. 2007;52:2482–2489
33. Jianu CS, Lange OJ, Viset T, et al. Gastric neuroendocrine carcinoma after long-term use of proton pump inhibitor. Scand J Gastroenterol. 2012;47:64–67
34. Jianu CS, Fossmark R, Viset T, et al. Gastric carcinoids after long-term use of a proton pump inhibitor. Aliment Pharmacol Ther. 2012;36:644–649
35. Merchant SH, VanderJagt T, Lathrop S, et al. Sporadic duodenal bulb gastrin-cell tumors: association with Helicobacter pylori
gastritis and long-term use of proton pump inhibitors. Am J Surg Pathol. 2006;30:1581–1587
36. Solcia E, Rindi G, Paolotti D, et al. Natural history, clinicopathologic classification and prognosis of gastric ECL cell tumors. Yale J Biol Med. 1998;71:285–290
37. Solcia E, Fiocca R, Rindi G, et al. Endocrine tumors of the small and large intestine. Pathol Res Pract. 1995;191:366–372
38. Davies R Tempest-Tost. 2006 London Penguin Books Ltd.
© 2013 Lippincott Williams & Wilkins, Inc.