Abstract
A common clinical indication for duodenal biopsy is the exclusion of coeliac disease / gluten sensitive enteropathy. However, a variety of inflammatory and infectious disorders may affect the duodenum, some of which are associated with subtle endoscopic findings. The indications for duodenal biopsy are often the same as the broader indications for upper gastrointestinal endoscopy and include chronic dyspepsia, unexplained anaemia, abdominal pain, bloating, nausea, and diarrhoea
Endoscopic findings associated with inflammatory duodenal biopsies range from normal-appearing duodenal mucosa to mild hyperaemia and congestion of the duodenal bulb to erosions, severe congestion, mucosal haemorrhage, mucosal contact bleeding, and luminal narrowing.
Introduction
A common clinical indication for duodenal biopsy is the exclusion of coeliac disease / gluten-sensitive enteropathy. However, a variety of inflammatory and infectious disorders may affect the duodenum, some of which are associated with subtle endoscopic findings. The indications for duodenal biopsy are often the same as the broader indications for upper gastrointestinal (GI) endoscopy and include chronic dyspepsia, unexplained anaemia, abdominal pain, bloating, nausea, and diarrhoea.
Endoscopic findings associated with inflammatory duodenal biopsies range from normal-appearing duodenal mucosa to mild hyperaemia and congestion of the duodenal bulb to erosions, severe congestion, mucosal haemorrhage, mucosal contact bleeding, and luminal narrowing.
Normal Duodenal Mucosa
Normal duodenal mucosa includes villi and crypts. The villi are shorter than jejunal and ileal villi, with a villous/crypt length of approximately 3–5:1 in the second to fourth parts of the duodenum. For reliable assessment of villous architecture, four consecutive villi should be present for examination. Within the villi, there are blood vessels and lymphatics, and the villi are lined by columnar absorptive cells (enterocytes) and goblet cells. Surface microvilli form a brush border, whose presence or absence should be recorded (see also Chapter 9). Intraepithelial lymphocytes are normal. The upper limit of normal is approximately 20 per 100 enterocytes.
Within the crypts, the common cell types are columnar cells, Paneth cells, goblet cells, and endocrine cells. Undifferentiated crypt cells are also present at the level of the lower fifth of the crypt.
The lamina propria comprises connective tissue with a population of inflammatory cells that includes lymphocytes, histiocytes, mast cells, plasma cells, and eosinophils. The normal inflammatory cell population should be assessed. A paucity or overabundance of a particular inflammatory cell lineage may indicate underlying inflammatory or immunological pathology.
Submucosal Brunner’s glands are present in the well-sampled, well-oriented duodenal biopsy from the duodenal bulb (D1) and may also occur in the second part (D2) but are less common distal to D2. This may help to confirm the origin of the biopsy and, because mild duodenitis is associated with minimal superficial villous alteration, allows for accurate assessment of mild inflammatory change throughout the full thickness of the mucosa.
Duodenal Sampling
There are few specific guidelines for the sampling of duodenal mucosa in the absence of endoscopically visible lesions. The diagnostic yield of biopsies of endoscopically normal duodenum is low,1, 2 with occasional diagnoses of unsuspected coeliac disease or giardiasis. When dyspepsia is the sole indication for endoscopy, biopsy of normal-appearing duodenum is not usual. However, routine biopsy is recommended in the setting of immunosuppression and/or transplantation to exclude graft-versus-host disease and opportunistic infections.3
In additional to the indications for sampling duodenal mucosa, the site and number of biopsy specimens are important considerations. Biopsies from the duodenal bulb (part 1 of the duodenum) are associated with a higher diagnostic yield.4 For conditions such as giardiasis, which have a homogeneous distribution within the duodenum, a small number of biopsies from any site may be sufficient. However, other conditions involve the duodenum in a more heterogeneous pattern. Coeliac disease may cause variable changes that are more prominent in D1 than D2 or vice versa. Therefore, guidance typically advises a minimum of four biopsies, including duodenal bulb and distal duodenal sampling.5
Clinical Information
It is important to have complete clinical information prior to the assessment of any histology specimen. Ideally, a request form accompanying a duodenal biopsy will include the following:
1. Patient sex
2. Patient age
3. Indication for endoscopy, including relevant symptoms or known history of enteropathy
4. Endoscopic appearance, including gastric findings
5. Helicobacter pylori (H. pylori) status, if known
6. Medical history, including drug history
7. Previous histological diagnosis, if relevant
Non-Specific Duodenitis/Peptic Duodenitis
Duodenitis may occur in the setting of antral gastritis, and particularly in association with H. pylori infection or acid injury. A suspected association between duodenitis and drugs, particularly non-steroidal anti-inflammatory drugs (NSAIDs), is difficult to prove.6 Duodenitis most commonly affects the duodenal bulb (first part).
Mild duodenitis may be subtle, is not associated with villous change, and in practice may be difficult to distinguish macroscopically and histologically from normality. In more severe duodenitis, there may be villous blunting (Figure 14.1A).
(A) Villous blunting may occur, potentially causing confusion with coeliac disease. However, intraepithelial lymphocyte numbers are not increased and intraepithelial neutrophils are usually present, assisting with the distinction from coeliac disease.
(B) Gastric metaplasia is common in duodenitis (thick arrows) and shows apical periodic acid–Schiff-positive mucin, contrasting with adjacent non-metaplastic absorptive cells and goblet cells (thin arrows).
In mild duodenitis, the inflammatory changes may be confined to the lamina propria with oedema and an increase in chronic inflammatory cells (e.g. plasma cells and eosinophils). However, the latter is somewhat subjective. Intraepithelial neutrophils constitute more objective evidence of inflammation. Indeed, many experts require intraepithelial neutrophils for a definite diagnosis of ‘duodenitis’, favouring terms such as ‘active duodenitis’ or ‘erosive duodenitis’ and avoiding the terms chronic duodenitis or peptic duodenitis, partly because a subjective increase in chronic inflammatory cells in the lamina propria is not sufficient, in their opinion, for a diagnosis of duodenitis.6 Neutrophils in the lamina propria are usually an abnormal finding unless they are very sparse and should prompt a search for additional evidence of duodenitis, and particularly a search for intraepithelial neutrophils. Epithelial cell abnormalities, e.g. mucin depletion, flattening, basophilia, and regenerative nuclear changes may occur in duodenitis (see Fact Sheet 14.1). Any of these changes may be focal or patchy. At low-power examination, villous changes and mucin depletion may be a clue that the duodenal biopsy is not normal.
Features
Associated with antral gastritis, particularly Helicobacter pylori–associated gastritis
Duodenal bulb affected most often
Histology
If mild, inflammation may be only in the lamina propria, making histological confirmation difficult
Intraepithelial neutrophils
Villous blunting (if moderate/severe)
Mucin depletion
Gastric metaplasia
Duodenitis versus Coeliac Disease
May share histological features, e.g. villous blunting
Intraepithelial lymphocytes usually increased in coeliac disease and not duodenitis
Intraepithelial neutrophils more numerous and common in duodenitis than coeliac disease
Consider coeliac serology if ‘duodenitis’ occurs distal to D1 or if interpretation of duodenitis is difficult
Gastric metaplasia is common and is strongly associated with duodenitis (Practice Points 14.1). It occurs more often in the duodenal bulb than elsewhere and is characterised by foci of gastric foveolar-type epithelial cells (Figure 14.1B; see also Figure 9.22). Apical mucin in the metaplastic cells can be highlighted by a periodic acid–Schiff (PAS) stain (Figure 14.1B) and is absent from the adjacent normal duodenal epithelium. In addition, gastric metaplasia shows loss of the normal duodenal epithelial brush border.7 Normal duodenal epithelial cells are interspersed between the metaplastic cells. H. pylori may colonise the metaplastic areas.8
Features
Most often affects duodenal bulb (D1)
Associated with duodenitis
Histology
Foci of mucin-producing gastric foveolar-type epithelial cells
Apical periodic acid–Schiff-positive mucin
May be colonised by H. pylori
Loss of the brush border
Interspersed normal duodenal epithelial cells
Distinction from Gastric Heterotopia
Metaplasia does not involve full mucosal thickness
Metaplasia does not include parietal cells
Metaplasia is often accompanied by clinical duodenitis and/or histological inflammation and/or H. pylori
Distinction between gastric metaplasia and gastric heterotopia (see Figure 9.27A) in a biopsy can be difficult, partly because both may have surface epithelium of gastric type. In general, heterotopic gastric mucosa is visible on endoscopy, forming a mass lesion. Gastric metaplasia may not be visible on endoscopy and is more diffuse. In contrast to heterotopia, gastric metaplasia does not involve the full thickness of the duodenal mucosa and does not include parietal cells. Duodenitis and H. pylori infection are more likely to be present in gastric metaplasia than in gastric heterotopia. Despite assertions by some authors of a pathogenetic continuum between duodenitis, gastric metaplasia of the duodenum, and gastric heterotopia of the duodenum, the latter is probably congenital rather than acquired.6
The terms ‘peptic duodenopathy’ and ‘peptic duodenitis’ have various definitions and may cause controversy. According to one scheme, peptic duodenopathy shows gastric foveolar metaplasia but no active inflammation whereas peptic duodenitis requires both foveolar metaplasia and active inflammation. No parietal or chief cells are present in either setting.6
Duodenitis and coeliac disease may show similar features, including villous architectural change, epithelial mucin depletion, and intraepithelial polymorphs (Figure 14.1A). The most important distinguishing histological feature is the number of intraepithelial lymphocytes (IELs). An increase in IELs is more specific than villous atrophy for a diagnosis of coeliac disease and, according to some authors, is a requirement for a diagnosis, or a suggestion of a diagnosis, of coeliac disease on the basis of histology. Particularly if the biopsies are from D2 or a more distal location, villous atrophy or inflammation should prompt consideration of coeliac disease if they are present. Correlation with coeliac serology may be helpful if there are features of possible peptic duodenitis beyond the duodenal bulb or if histological interpretation is difficult.9
Duodenal Ulcer/Peptic Ulcer
The prevalence and severity of duodenal ulceration have decreased over the past four decades. It generally occurs in older people and is more common in men, although the proportion of female patients is now higher than in the past.10, 11 Associations with duodenal ulceration include H. pylori infection, gastric metaplasia, duodenitis, NSAIDs, and tobacco smoking.12 Crohn’s disease and hypergastrinaemia may also be associated. In one report, independent predictors of duodenal ulceration included H. pylori gastritis, and the combination of active duodenitis and gastric metaplasia.13 The recurrence rate of duodenal ulceration is considerably lower if there is successful eradication of H. pylori. Other risk factors for recurrence include smoking, alcohol, and NSAID use.14 However, the exact relationship between H. pylori infection, acid levels, and duodenal ulceration is controversial and the role of NSAIDs is uncertain.15 Furthermore, the term ‘peptic ulcer’ is used variably, e.g. to include cases secondary, or presumably secondary, to H. pylori infection and/or cases caused by inappropriate acid secretion.
A duodenal peptic ulcer is typically solitary, well defined, less than 10 mm in diameter, and within 20 mm of the pylorus. The usual reason for biopsy is to exclude neoplasia. Histological features include an abrupt transition with adjacent normal mucosa, gastric metaplasia, and chronic duodenitis. Secondary villous change may also be present and there is associated Brunner’s gland hyperplasia.
Intraepithelial Lymphocytosis (‘Lymphocytic Duodenitis’/Microscopic Enteritis)
‘Lymphocytic duodenitis’ (or duodenal lymphocytosis or microscopic enteritis) is characterised by normal intestinal architecture associated with an increase in intraepithelial lymphocytes (Figure 14.2).16, 17 A description of the increase in intraepithelial lymphocytes may be more appropriate than a term such as ‘lymphocytic duodenitis’, as the latter is not a diagnosis but a description of a pattern. The pattern may represent latent coeliac disease (see also Chapter 15). Therefore, the pathologist should recommend coeliac serology. However, there may be no obvious cause. Some examples are secondary to H. pylori infection, especially if involving the duodenal bulb. Other causes or associations include NSAIDs, other drugs, inflammatory bowel disease, and autoimmune conditions such as rheumatoid arthritis.
Figure 14.2 Intraepithelial lymphocytosis with normal or near-normal villous architecture. There is a differential diagnosis for this appearance, including coeliac disease (which may be latent), infection, drugs (especially NSAIDs), systemic autoimmune disease, autoimmune enteropathy, and idiopathic inflammatory bowel disease.
Autoimmune enteropathy is worth considering if the increased lymphocyte population is predominantly in crypts rather than in surface epithelium. This clinical condition is associated with severe diarrhoea and malabsorption and tends to present in young children and infants (see Chapter 22).
If a raised intraepithelial lymphocyte count is present within architecturally normal duodenal mucosa, and if coeliac disease (gluten-sensitive enteropathy) has been excluded, a description of the pattern and correlation with the clinical picture are appropriate. Immunohistochemistry can subtype intraepithelial lymphocytes if neoplasia is a possibility, but is not generally used to calculate lymphocyte numbers (see Chapter 15).
Eosinophilic Duodenitis
Eosinophils are a component of the normal lamina propria of the duodenum and may occasionally infiltrate crypt epithelium in small numbers. In eosinophilic duodenitis, eosinophils are present in clusters, may infiltrate crypts more widely, and are more numerous. Some authors suggest a threshold of more than 50 eosinophils per high-power field for a significant increase. Eosinophil degranulation is a feature. Eosinophilic abscesses are unusual. There may also be an increase in other inflammatory cells, with admixed lymphocytes and plasma cells. Villous atrophy and crypt hyperplasia may occur. The deeper layers of the bowel may also be affected, sometimes in the absence of mucosal disease. Multiple biopsies are advisable because eosinophilia is often patchy.
A diagnosis of eosinophilic duodenitis is not possible in the absence of clinical information. To make the diagnosis, there must be GI symptoms, a marked eosinophilic infiltrate, and the absence of any other disease or condition that may cause an increase in eosinophils.18 Exclusion of parasitic infection is part of the standard assessment of duodenal mucosa. Other causes of eosinophilia include systemic diseases, e.g. asthma and atopy, and drugs (see Chapters 2, 7, and 13 and Fact Sheet 13.1).
Infectious Causes of Inflammation in the Duodenum
Giardiasis
The infectious organism is Giardia intestinalis (syn. Giardia lamblia; Giardia duodenalis), a unicellular flagellated protozoan spread by faecally contaminated water. The incidence rate in the United Kingdom is 7–8/100 000,19 of which 7% is associated with travel.20 It may affect immunocompetent or immunosuppressed individuals. The clinical differential diagnosis includes viral gastroenteritis (e.g. norovirus), lactose intolerance, irritable bowel syndrome, tropical sprue, inflammatory bowel disease, other infections such as cryptosporidiosis, and coeliac disease.21 Diagnosis usually relies on assessment of stool samples rather than biopsy.21 Cytology preparations are sometimes more informative than biopsy.22
On low-power examination of the affected duodenal mucosa, architectural distortion is often absent or minimal although there may be mild or more severe villous blunting. Intraepithelial lymphocyte numbers are normal or mildly increased and the crypts are normal and show no hyperplasia. There is often an increase in inflammatory cells in the lamina propria. However, the mucosa is sometimes completely normal.22
Giardia cysts are ovoid and measure 8–14 μm in maximum dimension, comparable to the size of an enterocyte nucleus. The organism does not invade and is present on the surface epithelium, often within crypts or between villi. It has a characteristic teardrop or pear shape (Figure 14.3). It may also be found in the stomach, ileum, or colon. Rarely, it occurs at non-duodenal sites in the absence of duodenal involvement. Giemsa staining and immunohistochemistry for CD117 help detect or confirm Giardia (Fact Sheet 14.2).
Setting
May affect immunocompetent or immunosuppressed subjects
Higher risk after travel
Usually duodenal
May also involve stomach, ileum, or colon
Non-duodenal involvement in the absence of duodenal involvement is rare
Histology
Architectural distortion usually absent or minimal
Villous blunting may occur
Intraepithelial lymphocyte numbers normal or mildly increased
Crypts normal
Lamina propria inflammation
Giardia Cysts
Arch/teardrop shape in tangential section and a binucleate pear shape en face
8–14 μm, similar in size to an enterocyte nucleus
Non-invasive
On the surface epithelium; +/− within crypts or between villi
Giemsa and CD117 help with detection