Inflammatory Disorders of the Stomach

Neutrophil Infiltration

The lamina propria of normal gastric mucosa may contain rare neutrophils. However, infiltration of the epithelium by neutrophils always represents a pathological response, and when present always constitutes an active component of gastritis. The term active is used to indicate an ongoing inflammatory process. In industrialized nations, active inflammation is most commonly caused by H. pylori gastritis, but many other infectious and inflammatory conditions (e.g., syphilis, Crohn’s disease ) may cause neutrophil infiltration as well.

The intensity and location of neutrophil infiltration may help differentiate various types of gastritis. For instance, the acute phase of infectious gastritis (such as phlegmonous gastritis) is likely to present with an abundant neutrophilic infiltrate, while the active phase of Helicobacter -induced gastritis may have moderate to severe levels of neutrophils, depending on a variety of factors. For instance, the density of neutrophils detected in H. pylori gastritis usually (but not universally) correlates with the density of H. pylori infection. Conversely, in cases with rare neutrophils but with features otherwise suggestive of H. pylori infection, such as a plasma cell-rich infiltrate, an intense hunt for the bacteria is usually warranted, either with or without ancillary tests. Acute hemorrhagic gastritis resulting from nonsteroidal antiinflammatory drug (NSAID)– or alcohol-induced chemical injury often shows only a minimal neutrophilic infiltrate, unless an erosion is present. In such cases, biopsies are typically devoid of a rich background of plasma cells. Mild and focal neutrophilic infiltration can be seen in some phases of autoimmune gastritis.

Mononuclear Infiltration

Antral mucosa normally contains a few mononuclear cells in the lamina propria (i.e., lymphocytes and plasma cells), whereas normal corpus mucosa contains virtually none. However, the density of “normal” mononuclear cells in the lamina propria of patients without disease varies considerably according to a number of factors, such as their geographic location and ethnicity, their age, and, of course, their diet and medication profile, among others (see Fig. 15.2 ). It is generally recommended that “gastritis” not be diagnosed unless there are at least several clusters of about five or more mononuclear cells in the lamina propria, or the infiltrate is diffuse and bandlike. In general, if the epithelium is healthy and does not reveal either regenerative or degenerative changes, then it is unlikely that a mononuclear infiltrate in the lamina propria is considered “pathological” or abnormal. In the corpus and fundus, rare single mononuclear cells in the lamina propria are considered a normal finding. Under normal conditions, intraepithelial lymphocytes are not present anywhere in the gastric epithelium.

Mucosal infiltration by plasma cells, lymphocytes, and often some eosinophils and mast cells is characteristic of chronic H. pylori gastritis. This infiltrate is typically bandlike and located mainly in the superficial portion of the lamina propria. It usually starts in the antrum, but it may progress proximally in the stomach, particularly in patients receiving proton pump inhibitors (PPIs) and without proper antibiotic therapy. In autoimmune gastritis, the corpus and fundus of the stomach are selectively involved, the infiltrate is more diffuse (usually centered around the deep portions of the mucosa), and the infiltrate consists mainly of lymphocytes rather than plasma cells.

Lymphoid Aggregates and Follicles

Normal gastric mucosa, particularly the corpus, may contain occasional, small lymphoid aggregates, usually basally located close to the muscularis mucosae. In contrast, lymphoid follicles (i.e., with germinal centers) are rare in normal gastric mucosa of H. pylori –negative adults. In studies with extensive biopsy sampling, lymphoid follicles or aggregates have been detected in almost all individuals with H. pylori gastritis. Subsequently, detection of lymphoid follicles is highly suggestive of H. pylori infection ( Fig. 15.3 ). In H. pylori –infected children and young adults, lymphoid follicles may be quite numerous, and it may in fact produce an endoscopic appearance of nodularity that is often referred to as follicular gastritis . ^,

Eosinophil Infiltration

Rare, scattered eosinophils may be present in the gastric mucosa of normal healthy patients, particularly in individuals who live in a suboptimal public health environment. However, prominent eosinophilic infiltration usually represents a pathological process, such as eosinophilic gastritis or gastroenteritis. ^, Eosinophil infiltration may occur in a wide variety of disorders, such as gastric anisakiasis and other granulomatous or parasitic infections of the stomach as well.

In H. pylori gastritis, if detected, eosinophil infiltration is usually mild. However, a greater eosinophilic component of the inflammatory infiltrate may manifest in children with H. pylori infection ( Fig. 15.4 ). After H. pylori eradication, eosinophils may persist for a long time, similar to mononuclear cells. Eosinophils can be a prominent component of the inflammatory infiltrate in autoimmune gastritis as well.

Mucosal Hyperemia

Mucosal hyperemia (congestion) is frequently an indicator of chemical injury. Interestingly, a significant correlation has been reported between the degree of hyperemia and the concentration of bilirubin within gastric fluid. However, congestion may also be detected in many other conditions, such as H. pylori gastritis, and may be related to an increased infiltration of mast cells ( Fig. 15.5 ). ^,

Surface Epithelium Regeneration and Degeneration

Surface epithelium regeneration and degeneration (or both) represents a nonspecific cellular response to a variety of injuries to the mucosa. For instance, degenerative changes may result in the emergence of cuboidal (rather than columnar) cells and depletion of mucin. Regenerative changes are particularly prominent in a variety of conditions, such as chemical gastropathy (resulting from bile reflux, ethanol, or NSAIDs), vascular gastropathies, and H. pylori gastritis as well. When prominent, epithelial regeneration in H. pylori gastritis may result in the accumulation of tufts and buds of cells at the surface of the mucosa ( Fig. 15.6 ).

Surface Erosion

Surface erosions are the result of severe epithelial injury that leads to necrosis. By definition, erosions are ulcers that do not extend beyond the muscularis mucosae. Chemically-induced flat surface erosions are characterized by a relatively paucicellular lamina propria, neutrophilic infiltration, hyalinization of the lamina propria, and a withering pit and gland epithelium, giving the mucosa an ischemic appearance. They often result from the acute effects of drugs, alcohol, bile reflux, or ischemia ( Fig. 15.7 ). ^, Elevated surface erosions are typically associated with H. pylori gastritis. These erosions are usually characterized by a superficial layer of fibrinopurulent debris (i.e., fibrinoid necrosis, neutrophils, and cellular debris) associated with a plasma cell–rich lamina propria infiltrate and hyperplastic, regenerative-appearing epithelium at the margins ( Fig. 15.8 ).

Foveolar Hyperplasia

Foveolar hyperplasia is defined as proliferation, elongation, and tortuosity of gastric pits that results in a corkscrew configuration to the pits. It is a regenerative phenomenon. It represents a compensatory tissue response to increased exfoliation of the surface epithelium, and it is a visual surrogate for increased cell proliferation and cell turnover ( Fig. 15.9 ). Foveolar hyperplasia may be diagnosed readily when several cross-sections of the same gastric pit are visualized in a single, well-oriented gastric biopsy specimen. ^, Other features of foveolar hyperplasia include hyperchromatic nuclei, high nucleus-to-cytoplasm ratio, upper pit mitoses, mucin depletion, and cuboidalization of the epithelial cells.

Foveolar hyperplasia is a characteristic feature of reactive gastropathy, including bile reflux and NSAID gastritis, particularly in long-term users. ^{, ,} The degree of foveolar hyperplasia (together with smooth muscle hyperplasia and congestion) has been used to score the severity of reactive gastropathy. A mild degree of foveolar hyperplasia is common in patients with H. pylori gastritis; however, marked hyperplasia often indicates coexistent chemical injury ( Fig. 15.10 ). Foveolar hyperplasia may also be prominent in autoimmune gastritis, sometimes with the development of hyperplastic polyps. Finally, the use of PPIs can also lead to foveolar hyperplasia.

Intestinal Metaplasia

Intestinal metaplasia is defined as replacement of gastric-type foveolar epithelium with intestinal cells (i.e., goblet cells, enterocytes with brush border, and Paneth cells) ( Fig. 15.11 ). Two types of intestinal metaplasia (I and II) have been described. Metaplastic epithelium that closely resembles normal small intestinal epithelium, containing acid mucin–producing goblet cells and absorptive enterocytes with a brush border, is considered complete metaplasia (type I). Incomplete metaplasia (type II) shows a disorderly mixture of irregularly shaped goblet cells (intestinal and immature intermediate mucous cells) that contain acidic sialomucins and sulfomucins. Although the clinical importance of their distinction has long been debated, a recent meta-analysis reemphasized that incomplete intestinal metaplasia is associated with a higher risk of gastric cancer. Incomplete metaplasia is further subdivided into type IIa and type IIb (also referred to as type III) by the presence of sulfomucins within nongoblet (mucinous) cells in the latter. Although type IIb has been shown to be associated with an even higher increased cancer risk, the practicability of subtyping incomplete intestinal metaplasia is limited in clinical practice. In fact, most patients with intestinal metaplasia have a mixture of the incomplete and complete types when extensive areas of mucosa are sampled, and the degree of incomplete intestinal metaplasia (including type IIb) parallels the extent of intestinal metaplasia in general. Furthermore, one meta-analysis indicated that intestinal metaplasia in the antrum and corpus is associated with a higher cancer risk than when it is limited to the antrum, which underscores a positive correlation between the extent of any type of intestinal metaplasia and risk of progression to carcinoma.

Intestinal metaplasia can be identified and its extent evaluated with the use of hematoxylin and eosin (H&E) stain, although specific mucin histochemical stains can be used to highlight metaplastic areas. For instance, the Alcian blue/periodic acid–Schiff (AB/PAS) stain at pH 2.5 is an excellent method for demonstrating the type and extent of intestinal metaplasia, particularly when used in combination with a hematoxylin counterstain ( Fig. 15.12 ). Mucin histochemical stains, some of which reported are potentially toxic (e.g., high-iron diamine), which were traditionally used to determine the specific type of intestinal metaplasia by helping to detect sulfated mucins, have been largely replaced by immunohistochemical stains that identify proteins associated with particular mucin-encoding genes. Although more than 20 such mucin genes have been identified, only a few (i.e., MUC1, MUC2, MUC5AC, and MUC6 ) are commercially available, and even these are used mainly in the research setting ( Table 15.2 ).

Intestinal metaplasia may develop in a variety of pathological settings, but its presence always indicates an underlying chronic, and often atrophic, gastritis. Intestinal metaplasia occurs frequently in patients with H. pylori gastritis. ^, Because H. pylori organisms do not normally adhere to intestinal-type epithelium (see Fig. 15.11 ), they commonly disappear in mucosa with extensive intestinal metaplasia and atrophy. Intestinalized epithelium may provide additional defense against H. pylori through changes in the composition of the gastric mucus resulting from metaplasia.

Intestinal metaplasia is also frequently found in the corpus of patients with autoimmune gastritis, and this is usually associated with pseudopyloric metaplasia as well. Foci of intestinal metaplasia are common in patients with reactive gastropathy after Billroth II surgery. Metaplasia can be detected in individuals with an otherwise completely normal stomach.

Atrophy

Gastric atrophy is defined as loss of gastric glands. ^, Atrophy is a histological finding, not a nosological entity. When the gastric mucosa is damaged, regardless of the cause, it may (1) regenerate to normal (restitutio ad integrum) or (2) undergo an adaptive reparative change with replacement of native glands with other types of tissue (i.e., metaplasia) ( Fig. 15.13 ). When injured glands fail to regenerate, the stromal space they previously occupied may be replaced by fibroblasts and extracellular matrix (i.e., fibrosis) ( Fig. 15.14 ). Ultimately, the result of metaplasia and fibrosis is loss of functional epithelium (i.e., atrophy).

Diffuse atrophy of the corpus and fundus typically occurs in autoimmune gastritis as a consequence of immune-mediated destruction of oxyntic epithelium. Less severe and more focal atrophy, which is usually limited to the antrum, may also occur in reactive gastropathy. H. pylori gastritis usually triggers multifocal atrophy in the antrum, which spreads to the oxyntic mucosa in the advanced stages of the disease.

The updated Sydney system recognizes multifocal atrophic gastritis (MAG; referred to as environmental atrophic gastritis in the pre– H. pylori era) as an entity distinct from nonatrophic gastritis and autoimmune atrophic gastritis. ^, The distinction between gastritis with focal atrophy and atrophic gastritis has not been well defined ( Fig. 15.15 ). This issue is important because scattered foci of intestinal metaplasia are found in the antrum of most patients with H. pylori gastritis and in a small percentage of noninfected adults. It is not appropriate to classify these individuals with “atrophic” gastritis, a diagnosis that implies altered gastric function and an increased risk of cancer. In the absence of established guidelines, we suggest that a diagnosis of MAG should be made only when there is evidence of atrophy either with or without intestinal metaplasia in at least 50% of a generously biopsied stomach (i.e., minimum of two samples from the antrum and two from the corpus or fundus).

Several pathology workshops have been devoted to the development of a reproducible method for grading atrophy in mucosal biopies. ^, Pathologists have recommended that atrophy be evaluated according to its two subtypes: nonmetaplastic or metaplastic . The subtype known as nonmetaplastic atrophy (see Fig. 15.14 ) is an area of mucosa with true glandular loss, replaced by stromal elements (usually fibrosis). In metaplastic atrophy (i.e., metaplasia equals atrophy), the native glands that are physiological for that site may be replaced by those with a pyloric phenotype (i.e., pyloric or pseudopyloric metaplasia) or an intestinal phenotype (i.e., intestinal metaplasia) comprising goblet cells and absorptive cells (either with or without a brush border) (see Fig. 15.13 ). Intestinal metaplasia may develop anywhere in the stomach, whereas pyloric (pseudopyloric) metaplasia occurs in the corpus or cardia. In both types of atrophy, the degree of gland loss may be broadly graded as mild, moderate, or severe, which corresponds to a scale of 1 to 3 ( Fig. 15.16 ).

Problems may arise when gastric biopsies are labeled as “stomach” or “antrum-body,” in which case it may be difficult to determine whether there is true pyloric metaplasia or the biopsy is simply representative of the antrum or antral-fundic transition. In this instance, immunohistochemical staining for gastrin-positive endocrine cells can help identify mucosa derived from the antrum. However, this method is not fool-proof because gastrin-positive cells may, on occasion, be present in metaplastic and atrophic areas of the fundus. Thus it is not just the finding of a gastrin-positive cell that should be considered, but also its quantity and location in the epithelium and its association with other mucosal structures. The algorithm depicted in Figure 15.17 summarizes an approach to the evaluation of atrophy in gastric biopsies.

Endocrine Cell Proliferation

Endocrine cell hyperplasia develops as a consequence of functional changes in the stomach and is most prominent in autoimmune atrophic gastritis. In this condition, hypochlorhydria or achlorhydria may lead to antral G-cell hyperplasia and a secondary elevation of serum gastrin levels. ^, Hypergastrinemia causes histamine-producing enterochromaffin-like (ECL) cells within oxyntic mucosa to proliferate. Neuroendocrine cell proliferation can be detected in patients with advanced atrophic gastritis and in biopsies of patients who take PPIs. Although antral G-cell hyperplasia can be detected easily on H&E–stained tissue sections ( Fig. 15.18 ), hyperplasia of ECL cells in oxyntic mucosa is best visualized and quantified with specific immunostains ( Fig. 15.19 ).

The most widely used classification system of ECL cell hyperplasia consists of simple or diffuse, linear, micronodular, and adenomatoid hyperplasia; ECL cell dysplasia; and neoplasia (i.e., neuroendocrine tumors). These are reviewed later in this chapter in the Autoimmune Gastritis section and in Table 15.3 . A mild to moderate degree of ECL cell hyperplasia may occur in patients who use PPIs for extended periods. However, in these patients, the oxyntic mucosa is not atrophic and commonly displays parietal cell hypertrophy and oxyntic gland dilation, both frequently seen in chronic PPI users. Alternatively, in the setting of atrophic gastritis, identification of ECL cell hyperplasia (particularly micronodular type) is a diagnostic feature of corporal autoimmune atrophic gastritis.

Parietal Cell Alterations

Apocrine-like protrusion and pseudohypertrophy of oxyntic cells are frequently attributed to chronic use of PPIs ( Fig. 15.20A ). ^, However, identical histological changes have been reported in other clinical settings; therefore the findings are not pathognomonic. Use of PPIs may lead to dilation of oxyntic glands, which in extreme cases may produce multiple fundic gland–type polyps that impart an appearance of the gastric mucosa that inspired the picturesque description of gastric acne . ^, Cytoplasmic vacuolation of parietal cells has been reported in patients using PPIs as well as rare cases with pseudo–signet ring cell morphology ( Fig. 15.20B ). Because most chronic users of NSAIDs also take PPIs to reduce the risk of NSAID-induced ulceration, oxyntic gland alterations are frequently seen in conjunction with reactive (chemical) gastropathy. ^{, ,}

Interfoveolar Smooth Muscle Hyperplasia

In the normal stomach, smooth muscle fibers are generally confined to the muscularis mucosae, where they run parallel to the mucosal surface. Several studies have shown that reactive (chemical) gastropathy is often associated with proliferation of smooth muscle fibers that run perpendicular to the muscularis mucosae within the interfoveolar lamina propria (see Fig. 15.9 ). The changes may be caused by the pulling effect of prolapsing mucosa, similar to that seen in solitary rectal ulcer syndrome and other prolapse-type polyps. Another proposed mechanism is the release of platelet-derived growth factor (PDGF), a known smooth muscle stimulant, as a result of epithelial damage. However, smooth muscle hyperplasia is not specific for NSAID-induced gastritis and may be seen in a variety of other conditions, such as gastric antral vascular ectasia (GAVE), bile reflux gastritis, and even in simple reactive mucosa adjacent to ulcers. ^,

Biopsy Protocol

The biopsy protocol depicted in Figure 15.21 is recommended to obtain satisfactory mucosal sampling. Specimens from three compartments (i.e., antrum, incisura angularis, and corpus) should be separately designated when submitted to the pathology laboratory. Proper specimen orientation is critical for optimal evaluation and is best accomplished at the time of tissue embedding. Unfortunately, in routine clinical practice, gastroenterologists do not often adhere to such a rigorous sampling and labeling protocol (see Reporting Gastritis in the Absence of a Complete Biopsy Set), and this limitation is not addressed by the Sydney classification.

Evaluation of Histological Variables

Each mucosal biopsy specimen should be assessed for its suitability for pathological examination. An acceptable slide is one that shows several well-oriented sections with the mucosal surface and the muscularis mucosae visible. Each relevant pathological feature (e.g., density of H. pylori , intensity of neutrophilic and mononuclear inflammation, atrophy of the antrum and corpus, intestinal metaplasia) should be graded on a standardized visual analogue scale ( Fig. 15.23 ). Each feature is assigned a numeric or descriptive value: 0 for absent, 1 for mild, 2 for moderate, and 3 for marked or severe. The values of each specimen are determined separately for each anatomic compartment (i.e., antrum and corpus). A minimum of two specimens from the antrum, one from the incisura angularis, and two from the corpus, should be evaluated.

Determination of Gradient of Inflammation

After evaluation of the histological variables, the next step in the Sydney system is to document the degree of inflammation in the two main gastric compartments to determine whether the inflammation is similar in intensity (i.e., pangastritis) or more severe in the antrum (i.e., antrum-predominant gastritis) or the corpus (i.e., corpus-predominant gastritis). To conclude that inflammation in one compartment is predominant, the difference in the inflammatory variables should be at least two grades. This helps minimize the effect of interobserver variability. The degree of atrophy and metaplasia can also be assessed according to the Atrophy 2000 guidelines. The last step in the Sydney classification is to decide whether focal atrophy or diffuse atrophy (metaplastic or nonmetaplastic) is present.

Diagnosis

The final diagnosis according to the Sydney system represents a synthesis of the observations outlined previously, including information regarding possible cause. Examples of diagnostic reports are “ H. pylori antrum-predominant gastritis” and “corpus-restricted atrophic gastritis without H. pylori infection, suggestive of autoimmune gastritis” ( Fig. 15.24 ).

Reporting Gastritis in the Absence of a Complete Biopsy Set

The Sydney system guidelines can be applied only when a full set of biopsy specimens is available (see Fig. 15.21 ). In routine practice, pathologists are usually asked to make a diagnosis based on only one or two biopsies and often from unspecified sites. In these cases, an empiric approach is recommended.

Most types of gastritis, including those caused by H. pylori, can be diagnosed without extensive tissue sampling. In contrast, assessment of the degree of atrophy requires adequate sampling and knowledge of the site of origin of each specimen to distinguish MAG from autoimmune metaplastic atrophic gastritis. A diagnosis of gastritis based on only a few nonrepresentative or inadequately identified biopsy specimens should not include any statements regarding topographic distribution of disease. If a specimen labeled as “corpus” contains pyloric-appearing glands and lacks parietal and chief cells, suggesting that it may be derived from the antrum, an immunostain for gastrin can help resolve the issue. The finding of numerous gastrin-positive cells, typically in the midgland region, is evidence in favor of the antrum, whereas atrophic pyloric-type glands from metaplastic corpus may contain either rare, scattered, or no gastrin-positive cells at all. In addition, recognition of nodular ECL-cell hyperplasia will raise the possibility of autoimmune gastritis. If uncertainty persists, a less-specific diagnosis (e.g., chronic inactive gastritis with atrophy and intestinal metaplasia) may be used, along with a comment regarding the differential diagnosis, including a suggestion that a more extensive sampling would be helpful to determine an exact etiology. Other information, such as clinical evidence of pernicious anemia, hypergastrinemia, achlorhydria, and the presence or absence of circulating anti–parietal cell antibodies, may help establish a precise diagnosis of autoimmune metaplastic atrophic gastritis.

Acute Hemorrhagic Gastritis

Pathological Features

Grossly, acute hemorrhagic gastritis is characterized by hyperemic edematous-appearing mucosa, surface erosions, and active bleeding ( Fig. 15.25 ). The damage is usually centered on the antrum. The clinical history (e.g., shock, burns, ingestion of large doses of NSAIDs) rather than the pathological features helps determine the precise cause. In addition, multiple subepithelial petechiae may develop after large quantities of strong alcoholic beverages are consumed over a short period. ^,

Microscopically, acute hemorrhagic gastritis, regardless of its cause, is characterized by dilation and congestion of mucosal capillaries, edema, and interstitial hemorrhage within the lamina propria. Surface erosions are typically usually small and appear ischemic, showing superficially hyalinized lamina propria. Aggregates of fibrin and neutrophils replace eroded epithelium and often project above the mucosal surface to form small, elevated clumps of necrotic debris ( Fig. 15.26A ). The foveolar epithelium underneath and adjacent to erosions and ulcers show regenerative changes reminiscent of epithelial “withering” observed in ischemia, but also elongation, increased tortuosity, and mucin depletion. Enlarged and hyperchromatic nuclei and increased mitoses are common features and should not be misinterpreted as dysplasia. Recalling the location of the normal regenerative zones in the antrum (i.e., gland base) and corpus (i.e., mucous neck cell region) can help mitigate an overdiagnosis of dysplasia. In the absence of concurrent H. pylori infection, unaffected areas of the stomach do not usually show increased chronic inflammation. In the setting of H. pylori infection, the pathologist may see chronic active inflammation in the mucosa, which may obscure or worsen the changes caused by NSAIDs or alcohol.

Stress-Induced Gastritis

Phlegmonous and Emphysematous Gastritides

Reactive (Chemical) Gastropathy

Etiological Factors and Pathogenesis

Duodenopancreatic (Bile) Reflux

Historically, bile reflux gastritis was initially recognized as a complication of partial (Billroth II) gastrectomy, which was a procedure used for benign conditions such as peptic ulcer disease. With the discovery of H. pylori as the principal cause of peptic ulcer disease and with the use of effective medical treatment, Billroth II antrectomies for benign disease have become increasingly uncommon. Subsequently, the prevalence of anastomotic and stomal gastropathy is expected to decline. Some patients have developed polypoid growths at the anastomotic site and gastric outlet obstruction when the lesions attain a large size, prompting endoscopic assessment.

Reflux of bile, or other luminal components of the duodenum, into a normal, nonoperated stomach is difficult to demonstrate in clinical practice, and the relevance of this condition needs further clarification. However, reactive gastritis, in the absence of known NSAID therapy or any other specific cause, is most commonly induced by duodenogastric bile reflux. Some degree of bile reflux is believed to occur in most healthy individuals and particularly those who smoke and have a duodenal ulcer, or those who have chronic respiratory disease, suffer from alcohol abuse, or are postcholecystectomy. Gastric mucosa chronically exposed to bile may undergo pathological changes similar to those that occur in a gastric stump. Patients with bile reflux documented by 24-hour ambulatory monitoring of intragastric bilirubin absorbance and pH values were found to have significantly more active and chronic inflammation, intestinal metaplasia, atrophy, and H. pylori infection (particularly in the corpus and incisura) than dyspeptic patients without bile reflux. A high prevalence of H. pylori infection was reported for antrectomized patients in Israel. ^, These data contrast with previous observations that bile reflux tends to eliminate, or suppress, H. pylori infection. ^,

Duodenogastric reflux with alkaline pancreatic and duodenal secretions, acids, bile salts, and lysolecithin results in disruption of the mucosal barrier and direct chemical damage to gastric surface epithelium. Loss of this barrier allows back-diffusion of hydrogen ions and secondary injury. This combined injury leads to accelerated exfoliation of surface epithelium and a histamine-mediated vascular response that manifests histologically as edema and hyperemia. Persistent epithelial damage promotes the release of other proinflammatory agents, such as PGDF, which stimulates smooth muscle and fibroblastic proliferation.

Medications

The most commonly used drugs and drug categories that can damage gastric mucosa when consumed at therapeutic doses are listed in Table 15.6 . The most common pattern of drug-induced gastric mucosal injury is reactive gastritis ( Fig. 15.27 ), although some drugs induce other patterns of injury. The most common forms of drug-induced gastritis are discussed in more detail in the next dedicated section.

TABLE 15.6

Drug-Induced Gastric Injury

Drug or Drug Family	Predominant Morphology of Gastric Injury
NSAIDs, aspirin, alcohol	Erosions, ulcers, reactive gastropathy
Proton pump inhibitors	Parietal cell hypertrophy and hyperplasia, fundic gland cysts and polyps
Iron	Reactive gastropathy and erosions with Fe 2+ deposits
Doxycycline	Reactive gastropathy and erosions; characteristic eosinophilic necrosis of the wall of superficial capillaries with luminal microthrombi and neutrophils
Sodium polystyrene sulfonate sorbitol (Kayexalate)	Crystal deposition (i.e., rhomboid or triangular, nonpolarizable, basophilic crystals adherent to the surface epithelium or within sloughed inflammatory exudates)
Cholestyramine	Crystal deposition (similar to Kayexalate crystals)
Colchicine or taxane-based chemotherapy	Abundant metaphase mitoses (especially ring mitoses), epithelial pseudostratification, loss of polarity, increased apoptosis in pit epithelium
Chemotherapy	Mucosal sloughing, enlarged gland cells with normal N:C ratio, gland loss
Immune checkpoint inhibitors	Chronic active gastritis with increased intraepithelial lymphocytosis and prominent apoptosis or focally enhanced gastritis
HAIC or SIR-Spheres therapy	Ulceration with nuclear atypia; numerous enlarged, bizarrely shaped nuclei with vesicular chromatin and large, irregular nucleoli
Bisphosphonates (e.g., alendronate)	Ulcerations (rare in the stomach, most common in the esophagus)
Corticosteroids	Possibly increased acid secretion, synergistic ulcerogenic effect with aspirin and NSAIDs

Fe ²⁺ , Ferrous iron ion; HAIC , hepatic arterial infusion chemotherapy; N:C ratio, nucleus-to-cytoplasm ratio; NSAIDs , nonsteroidal antiinflammatory drugs; SIR-Spheres , selective internal radiation microspheres.

Pathological Features

In patients with duodenogastric bile reflux, the gastric mucosa may exhibit congestion, edema, and surface erosion. In postgastrectomy patients with bile reflux, gross examination of the anastomotic site may also reveal polypoid lesions. ^, Nonspecific superficial erosions may occur in the more proximal areas of the gastric stump.

Microscopically, reactive gastritis is characterized by a variable degree of foveolar hyperplasia distinguished by tortuosity of the surface epithelium, regenerative changes including mucin depletion, thinning and loss of the columnar shape of the foveolar epithelium (i.e., cuboidalization), increased mitoses, nuclear hyperchromasia, occasional superficial erosions (see Fig. 15.7 ), and congestion, edema, and smooth muscle proliferation in the lamina propria. ^, Although a diagnosis of chemical gastritis can be suspected when these features are detected, a firm diagnosis necessitates correlation with supportive clinical data (e.g., medication history) and exclusion of H. pylori infection. The algorithm in Fig. 15.28 provides a common-sense approach to the evaluation of reactive gastritis. Usually, the features of reactive gastritis (see Fig. 15.27 ) are limited to the antrum. The corpus of previously nonoperated stomachs only rarely show findings of reactive gastritis. However, corpus mucosal biopsies are usually the only ones available for study from patients who have had Billroth II surgery.

Between 10% and 30% of patients who undergo a partial gastrectomy will develop mucosal lesions adjacent to the anastomosis or stoma site. These cystic lesions are formed from dilated, irregular, and cystic foveolae that frequently show regenerative changes ( Fig. 15.29 ). When the lesion forms a polyp, it is referred to as gastritis cystica polyposa; when the lesion is predominantly inverted, forming a prominent submucosal lesion or mass, the term gastritis cystica profunda is used (see Chapter 20 for details).

Drug-Induced Gastritis and Other Iatrogenic Mucosal Changes

The list of medications known to cause injury to gastric mucosa is ever-increasing. Some of these medications (e.g., NSAIDs) are available over the counter in many countries. In addition to NSAIDs, other agents, such as iron, potassium, gold, and corticosteroids may cause gastric mucosal damage, with some associated with more distinctive histological findings. The most commonly used drugs and drug categories that can damage gastric mucosa when consumed at therapeutic doses are listed in Table 15.6 . Although the most common pattern is reactive gastritis (see Fig. 15.27 ), some drugs induce other patterns of injury as well. For example, immune checkpoint inhibitors most frequently cause active chronic gastritis, or rarely, focally enhanced gastritis. A graft-versus-host disease (GVHD)–like pattern of injury, with increased apoptosis, may be observed in patients treated with mycophenolate, immune checkpoint inhibitors, colchicine, or taxane-based chemotherapy ( Fig. 15.30 )

Confirmation of the etiology of a suspected iatrogenic gastritis may be impossible without knowledge of the patient’s medication history. However, some histological feature such as ischemic-like erosions, brisk apoptotic activity, and ring mitoses should prompt a review of the patient’s prescribed drugs. In a minority of cases, the specific drug (e.g., Kayexalate resin) itself may be microscopically identified based on its characteristic morphology ( Fig. 15.31 ).