Biopsies of gastric mucosa, obtained at endoscopy, are common in routine pathology practice. The material ranges from single random biopsies of macroscopically normal or near-normal mucosa to detailed series mapping processes such as atrophic gastritis. In practice, the majority of cases will fall into the normal/near-normal mucosa, reactive gastritis, or Helicobacter pylori-associated gastritis categories. In this chapter, a practical, systematic approach to reporting gastric biopsies is emphasised to ensure that the pathology report assists in patient management. Three common systems for classification of gastritis – Sydney, ICD-10, Kyoto – are summarised and the key features required in the pathology report discussed. An approach to the differential diagnosis of some commonly encountered histological findings/patterns of mucosal injury, including atrophic gastritis, lymphocytic gastritis, and ‘granulomas’, is provided. There is discussion of the appropriate use of histochemistry/immunohistochemistry in gastric biopsies, noting that ‘routine specials’ add little to the assessment of H&E sections in most cases. Finally, the need for accurate clinical information (including endoscopic appearances, sites of biopsy, medical history and current/previous medications) to aid interpretation of the morphological findings and of the results of ancillary investigations is considered.
Endoscopic examination of the stomach plays an important role in the investigation of patients with dyspepsia, haematemesis and/or melaena, and upper abdominal pain. While biopsy is always necessary when gastric polyps, ulcers, or areas of mucosal irregularity have any features suggesting a neoplastic process,1 there is general agreement that biopsy of the normal stomach does not increase the likelihood of detection of neoplasia and that it is not necessary.2 While some authors recommend gastric biopsy to characterise gastritis,3 there is poor correlation between ‘endoscopic gastritis’ (redness of the mucosa) and demonstration of gastritis on histological assessment of biopsies.4 Moreover, the diagnosis of Helicobacter gastritis can be made by less expensive investigations (e.g. the urease test) and there is no evidence that detection of intestinal metaplasia or grading of chronic active gastritis currently influences management of the patient.5 Recommendations for endoscopy biopsy practice differ between pathology departments and countries, but histological assessment of endoscopic gastric biopsies remains a major component of the histopathology department’s workload and requires close co-operation between the clinician and the pathologist.
Most gastric biopsies are taken for assessment of gastritis, ulceration, or polyps, or for exclusion of dysplasia or malignancy. The standard laboratory approach to the gastric biopsy is to cut routine haematoxylin and eosin (H&E)-stained step sections at two or three levels, with the option of cutting additional levels depending on the initial morphological findings. Some laboratories perform routine stains for Helicobacter pylori and for mucins (to identify intestinal metaplasia) on every gastric biopsy, but there is conflicting evidence over the clinical value and effectiveness of this approach.6, 7 The updated Sydney classification system guidelines (see later) suggested that a histochemical stain (e.g. Giemsa, cresyl fast violet) is appropriate when Helicobacter pylori are not identifiable on H&E examination in the setting of gastric inflammation.8 Others prefer to use immunohistochemistry, particularly after antibiotics or proton pump inhibitor therapy, because it appears to be more sensitive for the detection of small numbers of Helicobacter pylori and the coccoid forms, the latter possibly representing dead organisms.6, 9, 10 The Royal College of Pathologists, UK, recommends a Helicobacter stain if characteristic inflammation is present, no Helicobacter are apparent on H&E, and no clinical test (e.g. urease test) has been performed.11
Other histochemical stains and immunohistochemistry (Table 12.1) may be useful in the appropriate setting, for example a Perls stain to confirm iron tablet debris in iron pill gastritis13 or a Ziehl–Neelsen stain in granulomatous inflammation.
|Intestinal metaplasia||PASD, Alcian blue|
|Helicobacter pylori Helicobacter heilmannii||Giemsa, cresyl fast violet, Warthin–Starry, modified Steiner, Diff-Quik, Genta, Leung’s Alcian yellow, toluidine blue, acridine orange, Helicobacter pyloriimmunohistochemistry|
|Lymphocytic gastritis||CD3/CD8 immunohistochemistry|
|Autoimmune ECL-cell hyperplasia||Synaptophysin and chromogranin immunohistochemistry|
|Distinguish antrum from atrophic corpus||Gastrin immunohistochemistry|
|Granulomatous, infectious||PAS, PASD, Alcian blue, Ziehl–Neelsen, Giemsa, GMS|
|CMV gastritis||CMV immunohistochemistry|
|Herpetic gastritis||HSV immunohistochemistry|
|EBV gastritis||EBV in situ hybridization|
|Fungal gastritis||PASD, GMS, Masson–Fontana|
|Iron pill gastritis||Perls|
|Mott cell/Russell body gastritis||CD38/CD138/CD79a immunohistochemistry|
|Xanthogranulomatous gastritis||Cytokeratin/CD68 immunohistochemistry|
Classification of Gastritis
There have been many different classifications of gastritis. The current ICD-10 classification (Table 12.2) is a mixture of aetiology and phenotype,14 while the more recently proposed Kyoto classification is based entirely on aetiology (Table 12.3).15 Pathologists will be far more familiar with the Sydney System for classification and grading of gastritis (Table 12.4).8, 16
|Type of gastritis||Aetiological factors||Gastritis synonyms|
|Other infectious gastritides||Phlegmonous|
The Sydney System
The Sydney System for classification and grading of gastritis was introduced in 1990 and incorporated the pathology, endoscopy, microbiology, autoimmunity, and epidemiology of chronic gastritis.17 It replaced a number of different classification systems, and the pathology component or ‘histological division’ combined morphology, topography, and aetiology16 into the final pathology report on endoscopic biopsies. Topography is considered the core of the classification (e.g. gastritis confined to the antrum, gastritis confined to the corpus, or pangastritis). The aetiology of gastritis, if known, is added as a prefix (e.g. Helicobacter pylori antral gastritis, autoimmune corpus gastritis) and the five key graded morphological variables (chronic inflammation, acute inflammation, intestinal metaplasia, atrophy, and Helicobacter pylori) are added as a suffix. The authors recommended semi-quantitative grading of these morphological parameters as absent (i.e. normal), mild, moderate, or severe.16 The final pathology report would then convey the type, severity, and extent of the gastritis, linked to the aetiology where possible.17
The Sydney System is a practical guideline designed to encourage a standard methodology for the reporting of the appearances of gastritis and depends on separate assessment (i.e. separate biopsies) of the antrum and corpus.17 The initial recommendation was for a minimum of four biopsies (two from the antrum and two from the corpus), plus biopsies of any discrete lesions. Examination of these four biopsies has a high probability of establishing the correct Helicobacter pylori status and the background degree and distribution of gastritis.18, 19 The Sydney System also recommended the use of at least one special stain to aid identification of Helicobacter pylori.
The Updated Sydney System
Following publication of the Sydney System, there was some criticism that it did not encompass all of the gastritides and therefore was not a true classification.20, 21 An ‘Updated Sydney System’, published in 1996,8 added the recommendation that biopsies should also be taken from the incisura angularis of the stomach. The maximal degrees of atrophy and intestinal metaplasia occur consistently in the incisura angularis, which is also the site most likely to reveal dysplasia.22 Therefore, a fifth biopsy (or more) from the incisura angularis was recommended. This sampling protocol is also now recommended by the Operative Link for Gastritis Assessment (OLGA) and the Operative Link on Gastric Intestinal Metaplasia (OLGIM) staging systems (used for assessing gastric cancer risk in gastritis), together with sampling of any visible lesions.23, 24
The updated Sydney System also added ‘visual analogue scales’ to aid grading of the histological (morphological) parameters and tabulated the updated classification of chronic gastritis (Table 12.4) which is still based on topography, morphology, and aetiology.8 Although some authors have shown a high level of interobserver agreement using the Sydney System,25 interobserver variation remains an issue for others.26