As pediatric esophagogastroduodenoscopy (EGD) has evolved into a routine outpatient procedure over the last few decades, indications for its use have correspondingly changed. At the Children’s Hospital of Philadelphia, the first-time EGD rate increased 12-fold in a 20-year interval between 1985 and 2005, with isolated abdominal pain replacing GI bleeding as the most frequent indication (Franciosi et al. 2010). As one would expect, the proportion of biopsies with histologic abnormalities also correspondingly declined (Franciosi et al. 2010). In children less than 1 year of age, the most frequent indications for endoscopy were diarrhea, failure to thrive, reflux, and rectal bleeding, according to one recent study (Volonaki et al. 2011). Histologic abnormalities were detected in two-thirds of cases. In current gastrointestinal practice, endoscopic forceps biopsies have replaced the Crosby suction biopsies because of greater ease of the procedure, patient comfort, avoidance of radiation exposure, direct visualization of the GI tract, which allows more precise targeting of the mucosal lesion, and the possibility of obtaining multiple biopsies from several sites. Numerous studies have shown that the histologic quality and interpretation of endoscopic forceps biopsies and capsule biopsies are equivalent (Achkar et al. 1986; Dandalides et al. 1989; Granot et al. 1993; Mee et al. 1985; Thomson et al. 1999). In the study by Volonaki et al. mentioned previously, only 2 % of biopsies in children less than 1 year of age were found to be insufficient for diagnosis (Volonaki et al. 2011). In addition to obtaining biopsies for routine histology, samples may also be snap-frozen (for disaccharidase analysis) or submitted for electron microscopy (for confirmation of microvillus inclusion disease). Biopsies from both proximal and distal duodenum are recommended, including biopsies of endoscopically normal mucosa, as many disorders affecting the duodenum have a focal distribution. In children, for example, the lesions in gluten-sensitive enteropathy (GSE) may be patchy, and villus atrophy may coexist with normal mucosa (Bonamico et al. 2004). Furthermore, villus atrophy may be limited to or most severe in the duodenal bulb at the time of diagnosis (Bonamico et al. 2004; Levinson-Castiel et al. 2011). Sampling endoscopically normal-appearing mucosa is recommended, as it may help assess the “background” features of the mucosa, and histologic examination may reveal clinically relevant findings unsuspected by the endoscopist (i.e., granulomas) (De Matos et al. 2008; Kori et al. 2003). A recent pediatric study found that routine duodenal biopsies performed for indications such as gastroesophageal reflux, vomiting, abdominal pain, anemia and during the evaluation of Crohn’s disease yielded clinically unsuspected pathologic findings in about 17 % of cases; conversely, a normal-appearing mucosa did not rule out abnormal histology (Kori et al. 2003).

Villi appear in the duodenum during the 8th week and crypts of Lieberkuhn are noted during the 9th week (Fig. 4.2), spreading caudally and arriving in the distal ileum by the 14th week. The definitive histologic features of the duodenum are established by the 14th week post-fertilization, and its histology closely resembles that of the newborn by 20 weeks (Fig. 4.3). The transition from pyloric to duodenal mucosa is gradual in some cases, and duodenal-like villi may be found in the distal pylorus and pyloric-like epithelium may be found in the duodenum. An adequate biopsy should have four to five well-oriented villi in a row, and observation of four normal villi in a row is generally an indication that the specimen is normal (Gramlich and Petras 2007) (Fig. 4.4). The normal villus height–crypt depth ratio in the duodenum is about 3:1 to 5:1. It is somewhat greater in adults (Whitehead 1997) and higher in the distal as compared to the proximal duodenum (Korn and Foroozan 1974) (Fig. 4.5). Whether genetically determined or resulting from environmental factors such as endemic infections, “normal” histologic findings in the small bowel also vary according to geography and race. Somewhat shorter villi and increased lamina propria mononuclear cells are reported in healthy individuals from developing countries as compared to industrialized nations (Bennett et al. 1985; Cook et al. 1969; Lindenbaum et al. 1971). The villus epithelium consists of absorptive enterocytes, characterized by basally located nuclei and a smooth brush border demonstrable by the periodic acid–Schiff (PAS) stain, as well as by immunohistochemical markers such as CD10. Goblet cells are interspersed between the enterocytes and contain an Alcian-blue-positive sialomucin (Fig. 4.6). Paneth cells are located at the base of the intestinal crypts and are easily recognized by the presence of eosinophilic cytoplasmic granules. Paneth cells form part of the innate immune system of the bowel, secreting substances such as defensins, lysozyme, and tumor necrosis factor-α (TNF-α), which play a role in protecting the gut from pathogenic microorganisms. Enteroendocrine cells are best demonstrated by immunohistochemical stains which label nonspecific neurosecretory granules such as chromogranin or specific neurosecretory products. These latter three subsets of mucosal cells appear to differentiate prior to 14 weeks under the control of transcription factors such as Math1 (Yang et al. 2001; Shroyer et al. 2007). The upper limit of intraepithelial lymphocytes (IELs) in the duodenum is estimated to be 20/100 epithelial cells by H+E staining, in contrast to the normal upper limit of 40 per 100 epithelial cells established 30 years ago mainly on the basis of studies of the jejunum (Ferguson and Murray 1971). Using immunostaining for CD3 to identify IELs, the upper limit of normal is reported as 25/100 epithelial cells in biopsies from the distal duodenum, with 26–29 CD3+ IELs considered borderline and 30 or more CD3+ IELs considered pathological (Veress et al. 2004) (Fig. 4.7). These CD3+ IELs are CD8+ T lymphocytes which belong to the MALT (mucosa-associated lymphoid tissue) system. The vast majority normally express the αβ T-cell surface receptor (TCR), and a small minority (<10 %) the γδ receptor (Kutlu et al. 1993). TCRγδ-expressing IELs are believed to mediate induction of oral tolerance. The lamina propria contains a mixture of plasma cells (mostly IgA secreting), CD4+ T lymphocytes, and lesser numbers of eosinophils, histiocytes, and mast cells. Newborns usually lack plasma cells in the first week of life, gradually acquiring them during the first month of life, with IgM-containing plasma cells predominating. By 3 months of life, IgA plasma cells predominate. Normal numbers of plasma cells and ratios of IgA/IgM/IgG are attained by the first year of life (Perkkio and Savilahti 1980).

Histologic evaluation of the small bowel biopsy should proceed systematically, to include (1) adequacy of the sampling, as determined by the number and site of the biopsy specimens; (2) assessment of the villus architecture with evaluation of the villus to crypt (V:C) ratio; (3) evaluation of the crypts (hyperplasia, hypoplasia); (4) the appearance of the surface enterocytes and their brush border; (5) intraepithelial lymphocytes; (6) the cellularity of the lamina propria (plasma cells may be absent or decreased in some primary immunodeficiencies); and (7) the presence of microorganisms (Giardia, cryptosporidia). Accurate evaluation of villus morphology requires careful attention to proper orientation of the specimen. Some laboratories have found that orientation of the specimen on filter paper prior to fixation is helpful (Serra and Jani 2006). We have found that rapid immersion of the specimen in fixative is preferable to having the endoscopist place the biopsy on filter paper, which may result in excessive stretching or drying of the specimen. Poor quality of the specimen, procedural artifacts, and tangential sectioning may lead to misinterpretations (Table 4.2) (Fig. 4.8).

Intestinal biopsy findings in some entities, such as congenital transport disorders, are associated with a normal biopsy, whereas others, such as autoimmune enteropathy or celiac disease, have variable degrees of villus atrophy with or without inflammation. Some diseases, such as abetalipoproteinemia or microvillus inclusion disease, present characteristic findings on an intestinal biopsy (Table 4.3).

Carbohydrate absorption begins with the breakdown of complex carbohydrates by salivary and gastric enzymes into oligosaccharides, which are then hydrolyzed to monosaccharides by specific disaccharidases located at the enterocyte brush border. Internalization of the hexose molecule is then mediated by “active” glucose absorption across the brush-border membrane by the Na+− dependent sodium/glucose cotransporter (SGLT1). Absence of SGLT1 results in glucose–galactose malabsorption (Hediger et al. 1989, 1995). There are also facilitative hexose transporters located at the basolateral membrane. The latter have been cloned and functionally characterized in several different tissues (GLUT1, erythrocyte; GLUT2, hepatocyte; GLUT3, brain; GLUT4, muscle and fat; GLUT5, small intestine) (Corpe et al. 1999).

There are four major disaccharidase enzymes: sucrase–isomaltase, glucoamylase–maltase, trehalase, and lactase–phlorizin hydrolase (or lactase-glycosylceramidase) (Auricchio and Troncone 2000). Sucrase–isomaltase is responsible for the degradation of dietary sucrose. Glucoamylase–maltase hydrolyses starch into small polymers. Lactase–phlorizin hydrolase degrades lactose, the main carbohydrate in milk. Trehalose occurs mainly in mushrooms. Trehalase deficiency is extremely rare, encountered mainly in Greenland natives and in Finns (Gudmand-Hoyer et al. 1988). The expression of these enzymes on the intestinal brush border appears to follow a time-dependent sequence. Lactase–phlorizin hydrolase, highly expressed at birth, declines in early life in most people (though it persists through adult life in Caucasians) (Welsh et al. 1978). Sucrase–isomaltase, undetectable at birth, reaches adult levels in the first few months of life. These changes coincide with the switch from milk to a solid diet.

The clinical picture of congenital disaccharidase deficiency and carbohydrate malabsorption is an osmotic diarrhea due to the unabsorbed solute in the ileum. The more rapid transit through the GI tract in the child results in a more severe diarrhea than in the adult. Symptoms are also related to factors such as residual enzyme activity, the amount of fed carbohydrate, and the degree of fermentation of the unabsorbed carbohydrate by colonic bacteria. The appearance of the small intestine in these cases is usually unremarkable, and the diagnosis arrived at by the determination of disaccharidase activities in homogenates of small bowel biopsies or by breath testing. Congenital sucrase–isomaltase deficiency is an autosomal recessive disorder occurring mainly in individuals of European descent, in whom the prevalence of heterozygosity reaches 2–9 % (Treem 1995). A diversity of different mutations underlies clinical heterogeneity, with the presentation of disease ranging from osmotic diarrhea, dehydration, and failure to thrive in infants to much milder symptoms in adults, in whom the diagnosis may be considerably delayed (Treem 1996). In most cases the disease becomes apparent when the child starts to consume fruits and juices (Canani and Terrin 2011). Lactose intolerance can be divided into three different groups: the rare congenital lactase deficiency characterized by complete absence of the enzyme from birth, manifesting as a severe osmotic diarrhea and dehydration; the very common hypolactasia which is the most common form of lactose malabsorption and which seems to be a consequence of a genetically mediated downregulation of the lactase enzyme (Campbell et al. 2009); and secondary lactase deficiency consequent to mucosal injury as in infectious diarrhea, celiac, or Crohn’s disease (Schulzke et al. 2009). Maltase–glucoamylase deficiency is as yet poorly defined and which appears closely related to sucrase–isomaltase deficiency (Canani and Terrin 2011).

Glucose–galactose malabsorption is a rare autosomal recessive disorder resulting from a mutation in the gene for SGLT 1 (also known as sodium/glucose cotransporter, solute carrier family 5, member 1, SLC 5A1), located on chromosome 22, which can cause severe chronic diarrhea (Xin and Wang 2011). Congenital fructose malabsorption is the result of a mutation in the gene for the hexose transporter isoform, GLUT5, located on chromosome 1 (Burant et al. 1992; Davidson et al. 1992). In clinical practice, malabsorption of fructose and that of a closely related sugar, sorbitol, may also result from overfeeding with juices containing high levels of these sugars.

Most clinical disorders of fat malabsorption result from either severe liver disease, pancreatic disease (such as cystic fibrosis), or extensive ileal resection (as in Crohn’s disease) with loss of the enterohepatic circulation of bile acids. Intestinal biopsies play a limited role in the diagnosis of these disorders. However, primary abnormalities involving abnormalities of fat transport within the enterocyte, though much less frequent, can result in a characteristic vacuolization of the enterocyte in intestinal biopsies.

Also known as autoimmune polyendocrinopathy candidiasis ectodermal dystrophy (APECED), this is an autosomal recessive disorder with heterogeneous clinical manifestations. It is caused by mutations in the AIRE gene, which codes for a transcription factor primarily expressed in medullary thymic epithelial cells where negative selection is thought to occur (Kampe 2009). The type I form is characterized by various endocrinopathies of an autoimmune nature, often beginning in childhood or early teenage years, by chronic mucocutaneous candidiasis due to a T-cell defect, starting soon after birth, and by dystrophy of ectodermal tissues (Betterle et al. 1998). Malabsorption can occur in up to 25 % of patients and appears to result from destruction of intestinal endocrine cells (Posovszky et al. 2012). A specific deficiency of cholecystokinin-producing enteroendocrine cells has been reported in one patient, with seeming reappearance of the cells when the diarrhea abated (Hogenauer et al. 2001). Small bowel biopsies show mild changes or can even be normal, in contrast with the crypt-destructive inflammation usually seen in autoimmune enteropathy.