Practical Approach to the Pathologic Diagnosis of

Antonia R. Sepulveda, MD, PhD; Madhavi Patil, MD

● Context.—Most types of gastritis can be diagnosed on sentative gastric cases from a gastrointestinal pa- hematoxylin-eosin stains. The most common type of chron- thology practice to demonstrate the practical application ic gastritis is Helicobacter pylori gastritis. Reactive or of basic histopathologic methods for the diagnosis of gas- chemical gastropathy, which is often associated with non- tritis. steroidal anti-inflammatory drug use or reflux, is com- Conclusions.—Limited ancillary tests are usually re- mon in most practices. The diagnosis of quired for a diagnosis of gastritis. In some cases, special can be challenging if few biopsy samples are available and stains, such as acid-fast stains, and immunohistochemical if the location of the in the is not known, stains, such as for H pylori and viruses, can be useful. Hel- such as when random biopsies are sampled in one jar. If icobacter pylori immunohistochemical stains can particu- the biopsy site is not known, immunohistochemical stains, such as a combination of synaptophysin and gastrin, are larly contribute (1) when moderate to severe, chronic gas- useful in establishing the biopsy location. tritis or active gastritis is present but no Helicobacter or- Objective.—To demonstrate a practical approach to ganisms are identified upon hematoxylin-eosin stain; (2) achieving a pathologic diagnosis of gastritis by evaluating when extensive intestinal metaplasia is present; and (3) in a limited number of features in mucosal biopsies. follow-up biopsies, after antibiotic treatment for H pylori. Data Source.—In this article, we present several repre- (Arch Pathol Lab Med. 2008;132:1586–1593)

astritis refers to a group of diseases characterized by present. If the biopsy shows chronic gastritis, the follow- G inflammation of the gastric mucosa. Histologic ex- ing questions should be posed: amination of gastric mucosal biopsies is necessary to es- tablish a diagnosis of gastritis. In clinical practice, the role 1. Are there features of chronic gastritis present? Lym- of the pathologist who evaluates a gastric biopsy for gas- phocytic and plasmacytic inflammatory reaction indicates tritis is to find the cause of gastritis because that will pro- chronic gastritis. vide direct targets toward which therapeutic measures can 2. Are there in the mucosa? The presence be directed. An etiologic classification of gastritis is pre- of neutrophils indicate active gastritis. sented at the end of this section. Comprehensive reviews 3. Is there Helicobacter? of gastritis have been published.1,2,3 The goal of this article 4. Is there glandular atrophy? Is intestinal metaplasia is to present a practical approach to the diagnosis of the present? most common types of gastritis encountered in a large 5. What is the topography of lesions (predominantly in practice of gastrointestinal . The reader will be the oxyntic mucosa of the body and fundus, predomi- presented several cases representative of typical forms of nantly in antrum, or involving both locations)? 6. Are there special features (such as , fo- gastritis; for each case, the reader will be prompted veolar hyperplasia, viral inclusions)? through a series of questions to examine the histologic 7. What ancillary studies are indicated, and what are features of the mucosa, leading to a pattern of answers the results? and to a final diagnosis. The first question is aimed at determining whether or not there are features of chronic or acute (active) gastritis TYPES OF CHRONIC GASTRITIS Infectious Gastritis Accepted for publication January 10, 2008. Helicobacter pylori infection is the most common cause of From the Department of Pathology and Laboratory Medicine, Hos- pital of the University of Pennsylvania, Philadelphia. chronic gastritis. Other forms of infectious gastritis include The authors have no relevant financial interest in the products or the following: Helicobacter heilmannii–associated gastritis; companies described in this article. granulomatous gastritis associated with gastric infections Presented in part at the 47th Annual Meeting of the Houston Society in mycobacteriosis, syphilis, histoplasmosis, mucormyco- of Clinical Pathologists, Houston, Tex, April 21, 2007. sis, South American blastomycosis, anisakiasis or anisak- Reprints: Antonia R. Sepulveda, MD, PhD, Department of Pathology and Laboratory Medicine, University of Pennsylvania, 3400 Spruce St, idosis; chronic gastritis associated with parasitic infec- Founders Six, Philadelphia, PA 19104 (e-mail: [email protected]. tions; and viral infections, such as cytomegalovirus and upenn.edu). herpesvirus infection. 1586 Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil examination reveals diagnostic H pylori bacterial forms in the surface mucus layer in close proximity to the apical aspect of surface epithelial cells. 4. Is there glandular atrophy? The biopsy sample avail- able is not adequate for evaluation of atrophic gastritis; multiple biopsies, including samples of gastric body, are necessary for adequate evaluation of glandular atrophy. Is there intestinal metaplasia? Yes. 5. What is the topography of lesions? The chronic gas- tritis in this case involves, at minimum, the gastric antrum; it is advisable to obtain biopsy samples of both gastric Figure 1. Helicobacter pylori–associated chronic active gastritis. A, antrum and body for a better evaluation of gastritis, as Chronic inflammation oriented toward the surface of the mucosa. Neu- recommended by the updated Sydney guidelines4 for clas- trophils cannot be seen at this magnification (original magnification sification of gastritis. ϫ10) but were identified with high power (hematoxylin-eosin stain). B, Circled area shows H pylori organisms within the mucus layer close 6. Are additional special features present? No. to the surface of gastric epithelial cells (hematoxylin-eosin, original 7. Are special stains recommended? No. magnification ϫ40). C, Gastric mucosa with intestinal metaplasia (he- matoxylin-eosin, original magnification ϫ20). Diagnosis. Gastric antral mucosa with H pylori–asso- ciated chronic gastritis, mildly active, and focal intestinal metaplasia. Noninfectious Gastritis H PYLORI–ASSOCIATED CHRONIC GASTRITIS Noninfectious gastritis is associated with autoimmune gastritis; reactive or chemical gastropathy, usually related The Helicobacter species consist of gram-negative rods that infect the gastric mucosa. Helicobacter pylori bacteria to chronic bile reflux or nonsteroidal anti-inflammatory ␮ drug (NSAID) intake; uremic gastropathy; noninfectious are 3.5 m long and are generally comma-shaped or have granulomatous gastritis; lymphocytic gastritis, including slightly spiral forms. Helicobacter heilmannii, a rare agent of chronic gastritis, is a 5- to 9-␮m-long bacterium, with gastritis associated with celiac disease; eosinophilic gas- 5 tritis; radiation injury to the stomach; graft-versus-host a characteristic tightly corkscrew-shaped, spiral form. disease; ischemic gastritis; and gastritis secondary to che- Helicobacter pylori infection usually is acquired during motherapy. childhood, persisting as chronic gastritis if the organism Many cases of gastritis are of undetermined cause and is not eradicated. During progression of gastritis over the present as chronic, inactive gastritis with various degrees years, the gastric mucosa undergoes a sequence of changes of severity.3 that may lead to glandular atrophy, intestinal metaplasia, increased risk of gastric dysplasia and carcinoma,6–9 and TYPES OF ACUTE GASTRITIS mucosa-associated lymphoid tissue lymphoma,10,11 report- ed as extranodal, marginal zone, B-cell lymphoma in the Many of the forms of chronic gastritis may present with 12 an acute form, with progression to chronic gastritis be- World Health Organization classification. cause of persisting injury or sequelae. This is the case of Helicobacter pylori infection is associated with the histo- gastritis associated with long-term intake of aspirin and logic pattern of active and chronic gastritis, reflecting the other NSAIDs and bile reflux into the stomach; excessive presence of neutrophils and mononuclear cells (lympho- consumption; heavy smoking; cancer chemother- cytes and plasma cells) in the mucosa, respectively. The apeutic drugs and radiation; acids and alkali in suicide term active gastritis is preferred to acute gastritis because H attempts; uremia; severe stress (trauma, burns, surgery); pylori gastritis is a long-standing chronic infection with ischemia and shock; systemic infections; mechanical trau- ongoing activity. Lymphoid aggregates and lymphoid fol- ma, such as intubation associated mucosal lesions; and vi- licles may be observed expanding the , and ral infections. rare lymphocytes may enter the epithelium. Helicobacter pylori organisms are found within the gastric mucus layer Case 1 that overlays the apical side of gastric surface cells, and lower numbers are found in the lower portions of the gas- A 60-year-old man underwent esophagogastroduoden- tric foveolae. Helicobacter pylori may be found within the oscopy. A biopsy of gastric antrum was submitted to pa- deeper areas of the mucosa in association with glandular thology to rule out H pylori. The histologic findings are cells in patients on acid blockers, such as the commonly shown in Figure 1, A through C. used proton pump inhibitors.13 Findings. Examination of the biopsy material available Helicobacter pylori–associated gastritis can display differ- gives the following answers: ent levels of severity. The severity of H pylori gastritis ac- 1. Are there features of chronic gastritis? Yes. The gas- tivity may be indicated in a pathology report as mild (rare tric antral mucosa shows expansion of the lamina propria neutrophils seen), moderate (obvious neutrophils within by chronic inflammatory cells, consisting of plasma cells the glandular and foveolar epithelium), or severe (numer- and small lymphocytes, predominantly located toward the ous neutrophils with glandular microabscesses and mu- luminal aspect of the mucosa, a pattern that is suggestive cosal erosion or frank ulceration).4,14 of H pylori infection. Helicobacter pylori–associated chronic gastritis can mani- 2. Are there neutrophils in the mucosa? Yes. Therefore, fest as a pangastritis involving the area from the pylorus this represents active gastritis. This is a mild form of active to the gastric body and cardia, or it may predominantly gastritis. involve the antrum. Patients with gastric ulcers generally 3. Is there Helicobacter? Yes. Hematoxylin-eosin (H&E) have antral-predominant gastritis, whereas pangastritis, Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil 1587 Findings. Examination of the biopsy material results in the following pattern of answers: 1. Are there features of chronic gastritis? Yes. The gas- tric antral mucosa shows expansion of the lamina propria by chronic inflammatory cells, consisting of admixed plas- ma cells and small lymphocytes, throughout the thickness of the mucosa. 2. Are there neutrophils in the mucosa? Yes, with an occasional glandular abscess; therefore, there is active gas- tritis. Of note, the active gastritis has a patchy distribution. 3. Is there Helicobacter? No. Examination with H&E stain does not reveal such bacterial forms. Immunohisto- chemical stain is performed. 4. Is there atrophy? The biopsy sample available is not adequate for evaluation of atrophic gastritis because the biopsy material is only from the gastric antrum; multiple gastric body biopsies are necessary for adequate evalua- tion of glandular atrophy. There is no intestinal metapla- sia. 5. What is the topography of lesions? The chronic gas- Figure 2. Chronic active gastritis in a patient with Crohn disease. A tritis involves, at minimum, the gastric antrum. glandular abscess is shown; additionally, there are many neutrophils in 6. Are additional special features seen? No. Although the lamina propria admixed with a background of chronic inflamma- tion (hematoxylin-eosin, original magnification ϫ20). in a case of Crohn disease gastritis, epithelioid granulo- mas may be present; in this case, no granulomas were seen. or at least multifocal gastritis, is more common in patients 7. Are special stains recommended? Yes. Helicobacter py- with gastric carcinoma. The latter generally have signifi- lori immunohistochemical stain, which is helpful in cases cant intestinal metaplasia and gastric oxyntic glandular where Crohn disease is suspected because the absence of atrophy coexisting in the background stomach. It is im- H pylori organisms in chronic active gastritis is consistent portant to make a pathologic diagnosis of atrophic gastri- with Crohn disease. The H pylori immunohistochemical tis because gastric atrophy is associated with increased stain in this case is negative. risk of gastric cancer.15,16 Patients with chronic atrophic Diagnosis. Gastric antral mucosa with chronic active gastritis may have up to a 16-fold increased risk of devel- gastritis, moderately active, patchy. No H pylori organisms oping gastric carcinoma, compared with the general pop- are identified by H&E or immunohistochemistry. Note: ulation.15,17 These features are consistent with Crohn disease–associ- When large numbers of H pylori are present in the mu- ated gastritis. cosa, the identification of typical organisms is generally possible on H&E stains. However, there are cases of chron- CROHN DISEASE–ASSOCIATED GASTRITIS ic, active gastritis with features suggestive of H pylori gas- The hallmark histopathologic features of Crohn disease– tritis in which the organisms are not detected. Several spe- associated gastritis are the presence of patchy, acute in- cial stains have been extensively used to help identify H flammation with possible gastric pit or glandular abscess- pylori organisms in the gastric mucosa, including modi- es, commonly with a background with lymphoid aggre- fied-Giemsa, Genta, thiazine stains, and immunohisto- gates. Recent studies20 reported the presentation of gastri- chemistry against Helicobacter antigens. The selection of the tis in patients with Crohn disease as a focally enhanced special stain used is largely dependent on preferences re- gastritis, characterized by small collections of lymphocytes lated to individual practices. Although, overall, no major and histiocytes surrounding a small group of gastric fo- differences in sensitivity and specificity have been report- veolae or glands, often with infiltrates of neutrophils. In ed, studies have recommended immunohistochemical severe cases, there may be diffuse inflammation in the stains in a subset of cases.18,19 In our practice, we prefer to lamina propria, with variable glandular loss, fissures, ul- use immunohistochemical stains for detection of H pylori cers, transmural inflammation, and fibrosis. Noncaseating if organisms are not found on H&E stains in the following epithelioid granulomas may be present in about one third cases: (1) if moderate to severe chronic gastritis or any of cases of Crohn disease gastritis but are often not seen, grade of active gastritis is present but no Helicobacter or- at least in part, because of limited tissue sampling. ganisms are identified on H&E; (2) when extensive intes- When granulomas are identified, the differential diag- tinal metaplasia is present because H pylori density is re- nosis includes other forms of granulomatous gastritis. duced in areas of intestinal metaplasia; and (3) during There are infectious and noninfectious causes of granulo- follow-up biopsies after antibiotic treatment for H pylori. matous gastritis. Noninfectious diseases represent the usu- Helicobacter heilmannii may cause similar pathology, and al cause of gastric granulomas and include Crohn disease, the treatment is similar to H pylori.5 sarcoidosis, and isolated granulomatous gastritis. Sarcoid- like granulomas may be observed in cocaine users, and Case 2 foreign material is occasionally observed in the granulo- A 45-year-old man is seen to rule out H pylori. He pre- mas. Sarcoidosis of the stomach is usually associated with sents with a history of Crohn disease. The histologic find- granulomas in other organs, especially the lungs, hilar ings are shown in Figure 2. nodes, or salivary glands. A diagnosis of idiopathic, iso- 1588 Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil 4. Is there atrophy? Yes. There is a reduced number of oxyntic glands in the biopsy. There is no intestinal meta- plasia. 5. What is the topography of lesions? The chronic gas- tritis involves, at minimum, the gastric body. 6. Are additional special features seen? No. 7. Are special stains recommended? Yes. Helicobacter py- lori immunohistochemical stain, which is positive. Diagnosis. Gastric oxyntic mucosa with H pylori–as- sociated chronic active gastritis and glandular atrophy, moderate. No intestinal metaplasia is identified. Helicobac- ter organisms are identified by immunohistochemistry.

ATROPHIC GASTRITIS Several publications, including those reporting the Syd- ney system and the updated Houston classification of gas- tritis, have proposed criteria for the evaluation of atrophic gastritis. Interobserver variability is significant, especially in the evaluation of antral atrophy.4,21 Recent advances that appear to decrease the interobserver variation in the as- sessment of gastric atrophy have been reported.14 Atrophy is more accurately assessed after resolution of severe in- flammation of the mucosa; therefore, if there is H pylori gastritis, the infection should be eradicated before atrophy is difinitively evaluated. When marked inflammation is present, a diagnosis of indefinite for atrophy may be of- fered, especially if there is no intestinal metaplasia. The recommended definition of atrophy is the loss of appropriate glands, and atrophy can be scored according to the degree of severity as mild, moderate, or severe.22 In this definition, intestinal metaplasia represents a form of atrophy described as metaplastic atrophy (or gastric glan- dular atrophy with intestinal metaplasia). Gastric atrophy is usually associated with intestinal Figure 3. Helicobacter pylori–associated atrophic gastritis. A, Chronic metaplasia. However, in limited endoscopic biopsies, in- active gastritis involving the gastric oxyntic mucosa with glandular at- testinal metaplasia might not be sampled, whereas the rophy (hematoxylin-eosin, original magnification ϫ10). Inset shows mucosa shows definitive atrophy. Usually gastric atrophy rare neutrophils into the glandular epithelium (white arrows). B, Im- and intestinal metaplasia occur on a background of chron- munohistochemical stain for H pylori shows a small area with organ- atrophic gastritis. isms attached to the surface epithelium. Insets show individual Heli- ic gastritis, hence the term cobacter bacteria (thin arrows) with characteristic elongated, slightly Sampling of the mucosa for evaluation of atrophy and spiral S shape or clusters of packed bacteria (thick arrowhead) closely gastritis is generally adequate by using the 5 biopsies rec- adherent to the surface of epithelial cells. Immunohistochemical stains ommended by the Sydney system, including 2 biopsies are useful in cases such as this one, when H pylori organisms are from the antrum, 2 from the corpus or body, and 1 from closely associated with the surface epithelial cells making it difficult to the incisura angularis.4,21 It is essential for the pathologist ascertain the characteristic bacterial morphology on hematoxylin-eosin stains (original magnification ϫ40). to have a means of determining the specific site in the stomach where a biopsy is sampled from because specific topography of atrophy characterizes the different types of lated, granulomatous gastritis is rendered when known atrophic gastritis. In atrophic gastritis associated with H entities associated with granulomas are excluded. pylori, glandular atrophy and intestinal metaplasia involve both the gastric antrum and body, whereas in autoim- Case 3 mune atrophic gastritis, the disease is essentially restricted A 60-year-old man presents with a nodularity of the to the gastric body. Ideally, the precise location is indicated gastric body to rule out H pylori. Esophagogastroduoden- by the endoscopist, and the biopsies from different sites oscopy with biopsy of the nodular areas was performed. are submitted in separate containers. However, using spe- The histologic findings are shown in Figure 3, A and B. cial stains can help the pathologist determine the location Findings. Examination of the biopsy material results of the biopsy fragments received. This approach is ex- in the following pattern of answers: emplified in case 5. 1. Are there features of chronic gastritis? Yes. Gastric atrophy and intestinal metaplasia are associated 2. Are there neutrophils in the mucosa? Yes. There are with increased gastric cancer risk, but unlike the intestinal neutrophils in the mucosa, representing active gastritis. metaplasia of Barrett syndrome, no specific recommen- 3. Is there Helicobacter? No. Examination with H&E dations for surveillance have been established in the Unit- stain does not reveal H pylori bacterial forms. Immunohis- ed States, although published data in other populations tochemical is performed. have suggested a benefit.23 In that study,23 patients with Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil 1589 itive cells, whereas the gastrin stain is negative. Because gastrin is negative, the biopsy is not from the gastric an- trum (G cells are characteristically located in the antrum and pylorus), and therefore, it can be established that the biopsy is of oxyntic mucosa with reduced oxyntic glan- dular profiles, establishing a diagnosis of atrophy. The lin- ear arrays of synaptophysin-positive cells represent en- terochromaffin-like cell hyperplasia. Enterochromaffin- like cell hyperplasia occurs in response to hypergastri- nemia that results from hypochlorhydria associated with gastric oxyntic cell atrophy. 6. Are additional special features seen? No. 7. Is immunohistochemical stain for H pylori positive? No. Diagnosis. Gastric oxyntic mucosa with chronic active gastritis and glandular atrophy, severe. No intestinal meta- plasia is identified. No Helicobacter organisms are identi- fied. Note: These features are most suggestive of autoim- mune gastritis.

AUTOIMMUNE ATROPHIC GASTRITIS This form of gastritis (reviewed in Sepulveda et al1 and Capella et al24) is caused by antiparietal cell and anti-in- trinsic factor antibodies and presents as a chronic gastritis with oxyntic cell injury, and glandular atrophy essentially restricted to the oxyntic mucosa of the gastric body and fundus. The histologic changes vary in different phases of the disease. During the early phase, there is multifocal infiltration of the lamina propria by mononuclear cells and Figure 4. Autoimmune gastritis. A, The gastric biopsy location in eosinophils and focal T-cell lymphocyte infiltration of ox- stomach was not provided in this case (hematoxylin-eosin, original magnification ϫ20). B, Immunohistochemical stain for synaptophysin yntic glands with glandular destruction. Focal mucous demonstrates enterochromaffin-like cell hyperplasia (original magnifi- neck cell hyperplasia (pseudopyloric metaplasia), and hy- cation ϫ20). The inset shows individual enterochromaffin-like cells pertrophic changes of parietal cells are also observed. staining brown, indicated by the arrow. C, Immunohistochemical stain During the florid phase, there is increased lymphocytic for gastrin is negative, indicating the biopsy site to be from the gastric inflammation, oxyntic gland atrophy, and focal intestinal ϫ oxyntic mucosa (original magnification 20). metaplasia. The end stage is characterized by diffuse in- volvement of the gastric body and fundus by chronic atro- phic gastritis associated with multifocal intestinal meta- extensive atrophic gastritis and intestinal metaplasia had plasia. In contrast to the gastric body, the antrum is an 11% risk of gastric malignancy. spared. Recently, a distinct form of autoimmune gastritis, Case 4 characterized by atrophic pangastritis, was reported in a small group of patients with systemic autoimmune dis- Esophagogastroduodenoscopy of a 60-year-old man orders.25 shows gastritis. The pathologist needs to rule out H pylori Autoimmune gastritis is a relatively rare disease but and gastric atrophy. The gastric site of the biopsy is not represents the most frequent cause of pernicious anemia specified. Figure 4, A through C, represents the histologic in temperate climates. The risk of gastric adenocarcinoma findings. was reported to be at least 2.9 times higher in patients Findings. Examination of the biopsy material results with pernicious anemia than in the general population, in the following pattern of answers: and there is also an increased risk of gastric carcinoid tu- 1. Are there features of chronic gastritis? Yes. mors. 2. Are there neutrophils in the mucosa? Yes. There are neutrophils in the mucosa; therefore, there is a component Case 5 of active gastritis. A 47-year-old woman presents with a history of celiac 3. Is there Helicobacter? No. Examination with H&E disease. Esophagogastroduodenoscopy was performed, stains do not reveal H pylori bacterial forms. Immunohis- with biopsy of gastric antrum. The pathologist needs to tochemical stain is performed. rule out H pylori. Figure 5, A and B, illustrates the histo- 4. Is there atrophy? If the biopsy is from gastric oxyntic logic findings. mucosa then there is atrophy, however, if the specimen is Findings. Examination of the biopsy material results from the antrum, it may represent chronic gastritis with- in the following pattern of answers: out atrophy. There is no intestinal metaplasia. 5. Are special stains recommended? Yes. Immunohis- 1. Are there features of chronic gastritis? Yes. There are tochemical stains for synaptophysin and gastrin are per- large numbers of intraepithelial lymphocytes. formed. Immunohistochemical stains for synaptophysin 2. Are there neutrophils in the mucosa? No. (Figure 4, B), show a linear pattern of synaptophysin-pos- 3. Is there Helicobacter? No. Examination with H&E 1590 Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil LYMPHOCYTIC GASTRITIS Lymphocytic gastritis is a type of chronic gastritis char- acterized by marked infiltration of the gastric surface and foveolar epithelium by T lymphocytes and by chronic in- flammation in the lamina propria. A diagnosis can be ren- dered when 30 or more lymphocytes per 100 consecutive epithelial cells are observed, and the counts are recom- mended in biopsies from the gastric corpus. The endo- scopic pattern is, in some cases, described as varioliform gastritis. The cause of lymphocytic gastritis is usually un- known, but some cases are seen in patients with gluten- sensitive /celiac disease and in Me´ne´trier disease. Smaller numbers of intraepithelial lymphocytes can also be seen in H pylori gastritis, but the diagnosis of lympho- cytic gastritis should be reserved for cases with marked intraepithelial lymphocytosis in the absence of active H pylori gastritis. Lymphocytic gastritis can be observed in children but is usually detected in late adulthood, with average age of diagnosis of 50 years. Case 6 A 75-year-old woman presents after esophagogastro- duodenoscopy. Gastric antrum shows gastritis; the pa- thologist is asked to rule out H pylori. The histologic find- ings are shown in Figure 6. Findings. Examination of the biopsy material results in the following pattern of answers: 1. Are there features of chronic gastritis? There is min- imal chronic gastritis. 2. Are there neutrophils in the mucosa? No. 3. Is there Helicobacter? No. Examination of H&E stains does not reveal H pylori bacterial forms. Figure 5. Lymphocytic gastritis. A, Gastric mucosal surface epithelium 4. Is there atrophy? No. There is no atrophy or intestinal is studded with intraepithelial small lymphocytes (hematoxylin-eosin, metaplasia. original magnification ϫ20). B, Immunohistochemistry highlights nu- 5. What is the topography of lesions? The chronic gas- merous CD3-positive T lymphocytes staining dark brown (original mag- tritis involves, at minimum, the gastric antrum. ϫ nification 20). The lymphocytes predominantly infiltrate the surface 6. Are additional special features seen? Yes. There are and foveolar epithelium. The thick arrow points to light-brown back- ground stain found in some glandular cells. The inset shows a higher diagnostic special features, including foveolar hyperplasia magnification of the surface epithelium containing many intraepithelial with a corkscrew appearance of the foveolae. The foveolar lymphocytes (original magnification ϫ40). One individual lymphocyte epithelium shows reactive cytologic features, including re- is indicated by the arrow. T lymphocytes stain dark brown, whereas duced cytoplasmic mucin. The lamina propria shows con- the nucleus of the epithelial cell nuclei stain with the blue counterstain. gestion and smooth muscle hyperplasia, with prominent muscularization of the most superficial mucosa. 7. Are special stains recommended? No ancillary tests are performed. stain does not reveal H pylori bacterial forms. Immunohis- tochemical is performed. Diagnosis. Gastric antral mucosa with features con- 4. Is there atrophy? No. There is no glandular atrophy sistent with reactive gastropathy. No H pylori organisms and no intestinal metaplasia. are identified. 5. What is the topography of lesions? The chronic gas- CHRONIC, REACTIVE (CHEMICAL) GASTROPATHY tritis involves, at minimum, the gastric antrum. Chronic reactive gastropathy (also know as chemical 6. Are additional special features seen? Yes. The specific gastropathy) is very common in current clinical practice. features in this biopsy include a characteristic intraepithe- The mucosal changes are usually more prominent in the lial lymphocytosis. Immunohistochemical stain for CD3 is prepyloric region, but they may extend to involve the ox- positive, highlighting a population of T lymphocytes in yntic mucosa. The usual underlying causes include chron- the mucosa and, typically, many intraepithelial lympho- ic bile reflux and long-term NSAID intake. The histopath- cytes. ologic features include mucosal , congestion, fibro- 7. Are special stains recommended? Yes. Immunohis- muscular hyperplasia in the lamina propria, and foveolar tochemical stain for H pylori, which is negative. hyperplasia with a corkscrew appearance in the most se- vere forms. The foveolar epithelium characteristically Diagnosis. Chronic gastritis with increased intraepi- shows reactive nuclear features and reduction of mucin. thelial T lymphocytes. No Helicobacter organisms are iden- The epithelial changes occur with little background chron- tified. Note: These features are consistent with lympho- ic inflammation. However, if there is erosion of the mu- cytic gastritis–associated with celiac disease. cosa, superficial neutrophils may be present. Erosive gas- Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil 1591 tritis (Figure 7, A) can present clinically as acute gastritis, often associated with NSAID intake. The features associated with bile reflux are typically found in patients with partial gastrectomy, in whom, the lesions develop near the surgical stoma. However, alter- ations induced by bile reflux also affect the intact stomach. A recent study26 reported altered mucin expression in re- active gastropathy, including aberrant expression of MUC5Ac in pyloric glands. Evaluation of mucin-expres- sion patterns can be useful to support a diagnosis of re- active gastropathy; however, additional studies are war- ranted to validate this potential application of mucin im- munohistochemistry.

Case 7 A 45-year-old woman presents with a history of bone marrow transplant. Esophagogastroduodenoscopy shows gastric erosion. The histologic findings are represented in Figure 7, B. Findings. Examination of the biopsy material results in the following pattern of answers: 1. Are there features of chronic gastritis? Yes. The sam- ple of gastric mucosa reveals mucosal erosion with gran- ulation tissue and associated chronic and acute inflam- mation. 2. Are there neutrophils in the mucosa? Yes. There are superficial neutrophils in the mucosa, but they are limited to the area of mucosal erosion. 3. Is there Helicobacter? No. Examination with H&E stain does not reveal such bacterial forms. 4. Is there atrophy? No. There is no atrophy or intestinal metaplasia. 5. What is the topography of lesions? Away from the areas of erosion, there is no evidence of gastritis; therefore, the location of the biopsy is not contributory in this case. 6. Are additional special features seen? Yes. There are special features including enlarged cells, arousing suspi- cion of cytomegalovirus inclusions in the granulation tis- sue. 7. Are special stains recommended? Yes. Immunohis- tochemical stain for cytomegalovirus reveals rare but char- acteristic viral inclusions (not shown). Diagnosis. Gastric antral mucosa with erosion and cy- tomegalovirus inclusions, consistent with cytomegalovirus- associated gastritis.

CYTOMEGALOVIRUS GASTRITIS Cytomegalovirus infection of the stomach is observed in patients with underlying immunosuppression. Histo- logically, intranuclear eosinophilic inclusions and smaller intracytoplasmic inclusions in enlarged cells are charac- teristic. A patchy, mild inflammatory infiltrate is observed in the lamina propria. Viral inclusions are present in en- dothelial or mesenchymal cells in the lamina propria and may be seen in gastric epithelial cells. Severe activity may result in mucosal ulceration.

Figure 6. Reactive gastropathy (hematoxylin-eosin, original magnifi- cation ϫ20). ← Figure 7. Erosive gastritis and cytomegalovirus-associated gastritis. A, Gastric mucosa with erosion in a patient with history of nonsteroidal cytomegalovirus inclusion. This single inclusion is identified by the ar- anti-inflammatory drug use (hematoxylin-eosin, original magnification row on the background granulation tissue (hematoxylin-eosin, original ϫ20). B, Gastric mucosal erosion with granulation tissue. Inset shows magnifications ϫ20 and ϫ40 [inset]). 1592 Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil COMMENT 7. Uemura N, Okamoto S, Yamamoto S, et al. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med. 2001;345:784–789. Most types of gastritis can be diagnosed with H&E 8. Asaka M, Sugiyama T, Nobuta A. et al. Atrophic gastritis and intestinal meta- plasia in Japan: results of a large multicenter study: Helicobacter. 2001;6:294– stains. To reach a determination of etiology and a specific 299. diagnostic entity, a limited list of questions can be used 9. Correa P, Haenszel W, Cuello C, Tannenbaum S, Archer M. A model for to evaluate the histopathology of gastric biopsies, which gastric cancer epidemiology. Lancet. 1975;2:58–60. 10. Parsonnet J, Hansen S, Rodriguez L, et al. Helicobacter pylori infection can lead to a pattern of answers that corresponds to a and gastric lymphoma. N Engl J Med. 1994;330:1267–1271. specific diagnosis of the most common types of gastritis. 11. Wotherspoon AC, Ortiz-Hidalgo C, Falzon MR, Isaacson PG. Helicobacter Although not ideal, the diagnosis of gastritis can be pylori–associated gastritis and primary B-cell gastric lymphoma. Lancet. 1991; reached from limited biopsy material, even when the lo- 338:1175–1176. 12. Jaffe ES, Harris NL, Stein H, Vardiman JW. Pathology and Genetics of Tu- cation of the biopsy is not indicated. If the biopsy site is mours of Hematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001. not known, immunohistochemical stains for synaptophy- World Health Organization Classification of Tumours; vol 3. sin and gastrin can help determine the biopsy location, 13. Tagkalidis P, Royce S, Macrae F, Bhathal P. Selective colonization by Hel- icobacter pylori of the deep gastric glands and intracellular canaliculi of parietal permitting a specific diagnosis of atrophic gastritis type. cells in the setting of chronic proton pump inhibitor use. Eur J Gastroenterol Helicobacter pylori immunohistochemical stains can be par- Hepatol. 2002;14:453–456. ticularly useful when moderate to severe chronic gastritis 14. Rugge M, Genta RM. Staging and grading of chronic gastritis. Hum Pathol. 2005;36:228–233. or any active gastritis is present but no Helicobacter organ- 15. Huang JQ, Sridhar S, Chen Y, Hunt RH. Meta-analysis of the relationship isms are identified on H&E stains, when extensive intes- between Helicobacter pylori seropositivity and gastric cancer. . tinal metaplasia is present, and to evaluate follow-up bi- 1998;114:1169–1179. 16. Correa P, Houghton J. Carcinogenesis of Helicobacter pylori. Gastroenter- opsies after antibiotic treatment for H pylori. ology. 2007;133:659–72. At the end of the day, there are a number of cases with 17. Sepulveda AR, Coelho LG. Helicobacter pylori and gastric malignancies. a diagnosis of chronic inactive gastritis, generally mild, for Helicobacter. 2002;7(suppl 1):37–42. 18. Jonkers D, Stobberingh E, de Bruine A, Arends JW, Stockbrugger R. Eval- which a specific etiology cannot be determined by histo- uation of immunohistochemistry for the detection of Helicobacter pylori in gastric pathologic examination alone. This may be accounted for mucosal biopsies. J Infect. 1997;35:149–154. by limited tissue sampling, nonspecific focal, mild, chronic 19. Toulaymat M, Marconi S, Garb J, Otis C, Nash S. Endoscopic biopsy pa- inactive gastritis associated with various systemic disor- thology of Helicobacter pylori gastritis: comparison of bacterial detection by im- munohistochemistry and Genta stain. Arch Pathol Lab Med. 1999;123:778–781. ders, or as yet uncharacterized forms of gastritis. 20. Xin W, Greenson JK. The clinical significance of focally enhanced gastritis. Am J Surg Pathol. 2004;28:1347–1351. References 21. Price A. The Sydney System: histological division. J Gastroenterol Hepatol. 1. Sepulveda AR, Dore MP, Bazzoli F. Chronic gastritis. Available at: http:// 1991;6:209–222. www.emedicine.com. Accessed November 27, 2007. 22. Rugge M, Correa P, Dixon MF. et al. Gastric mucosal atrophy: interob- 2. Srivastava A, Lauwers GY. Pathology of non-infective gastritis. Histopathol- server consistency using new criteria for classification and grading. Aliment Phar- ogy. 2007;50:15–29. macol Ther. 2002;16:1249–1259. 3. McKenna BJ, Appelman HD. Primer: histopathology for the clinician—how 23. Whiting JL, Sigurdsson A, Rowlands DC, Hallissey MT, Fielding JW. The to interpret biopsy information for gastritis. Nat Clin Pract Gastroenterol Hepatol. long term results of endoscopic surveillance of premalignant gastric lesions. Gut. 2006;3:165–171. 2002;50:378–381. 4. Dixon MF, Genta RM, Yardley JH, Correa P, the participants in the Inter- 24. Capella R, Fiocca C, Cornaggia M. Autoimmune gastritis. In: Graham DY, national Workshop on the Histopathology of Gastritis, Houston 1994. Classifi- Genta RM, Dixon MF, eds. Gastritis. Philadelphia, Pa: Lippincott Williams; 1999: cation and grading of gastritis: the updated Sydney System. Am J Surg Pathol. 79–96. 1996;20:1161–1181. 25. Jevremovic D, Torbenson M, Murray JA, Burgart LJ, Abraham SC. Atrophic 5. Singhal AV, Sepulveda AR. Helicobacter heilmannii gastritis: a case study autoimmune pangastritis: a distinctive form of antral and fundic gastritis associ- with review of literature. Am J Surg Pathol. 2005;29:1537–1539. ated with systemic autoimmune disease. Am J Surg Pathol. 2006;30:1412–1419. 6. Sipponen P, Kekki M, Haapakoski J, Ihamaki T, Siurala M. Gastric cancer 26. Mino-Kenudson M, Tomita S, Lauwers GY. Mucin expression in reactive risk in chronic atrophic gastritis: statistical calculations of cross-sectional data. Int gastropathy: an immunohistochemical analysis. Arch Pathol Lab Med. 2007;131: J Cancer. 1985;35:173–177. 86–90.

Archives of Pathology & Laboratory Medicine and Archives of Ophthalmology will publish a joint theme issue on ophthalmic pathology in August 2009. Articles on diagnostic procedures, path- ologic mechanistic pathways, and translational research in retinoblastoma, melanoma, lympho- ma, orbital, and adnexal tumors in ophthalmic pathology will have the best chance for consid- eration in this theme issue. Manuscripts must be submitted no later than February 1, 2009 for consideration in the joint theme issue. All submissions will undergo our usual peer review process. Important: When submitting a manuscript for this theme issue, be certain to mention this in both the cover letter and the Comment section within the AllenTrack submission system. To view our Instructions for Authors, visit http://arpa.allenpress.com/pdf/instructionsforauthors.pdf

Arch Pathol Lab Med—Vol 132, October 2008 Pathologic Diagnosis of Gastritis—Sepulveda & Patil 1593