At Great Ormond Street, there is a decades-long history of investigating Hirschsprung disease (HD), an enteric neuromuscular disease. Around 1950–1960, the absence of enteric nervous system was universally accepted as the cause of HD. It soon became apparent that there were conditions under which the intestine was ganglionic, but the children clearly had abnormal intestinal motility causing functional intestinal obstruction (pseudoobstruction) or constipation (pseudo-HD). It is now known that these disorders most commonly present in the neonatal period or early childhood, and some may present in utero. A paediatric surgeon, Dr Harold Homewood “Nikki” Nixon, was fascinated by these children and inspired Dr Peter Milla to investigate their gut motility by the newly developing manometric methods. Peter, in turn, enthused me to try finding morphological correlates to his manometric findings. It became evident that the abnormal motility resulted from a variety of defects in intestinal innervation or in the enteric smooth muscle. These were either congenital (mostly genetic) or acquired and affected a segment of the intestine or diffusely involved the entire gastrointestinal tract.
In childhood, acquired conditions are not as common as genetic disorders, but they are more readily treatable. In a youngster with previously normal bowel habits who presents later in childhood, particularly after an episode of apparent gastroenteritis, acquired aetiology must be considered. In these children, who may have a lymphocytic ganglionitis (1) or myositis (2), antibody screening for circulating immunoglobulin G–class autoantibodies against enteric neurons (1) or smooth muscle (2) is indicated. Eosinophilic ganglionitis (3), leiomyositis, or mucosal eosinophilia together with increased numbers of mast cells also can result in altered gut motility. Anorectal manometry in these patients shows increased internal anal sphincter pressure, but normal sphincter relaxation. Similarly, abnormal intestinal motor activity secondary to other diseases is less often seen in children than in adults, but is present in some conditions including cystic fibrosis (4) or muscular dystrophies.
Aganglionosis (5) is by far the most common genetic neural abnormality, but a number of other specific histological phenotypes of nerve and muscle diseases have been recognised. Some of these may be acquired (Table 1). Selected phenotypes are discussed below. Accurate identification of phenotypes allows for better prognostic prediction, including genetic implications, and improved potential treatment options.
Surgeons must be discouraged from performing unnecessary laparotomies, but when these are undertaken it is vital that full-thickness intestinal tissue is taken for histological diagnosis. Because the changes are often subtle, this includes paraffin sections in longitudinally orientated full-thickness bowel (preferably at least 2-cm long), which allows for assessment of myenteric neuronal density (7), fibrosis, and inflammation, and can also be used for immunohistochemistry of muscle (8,9,20) and nerve proteins. In addition, a snap-frozen full-thickness longitudinal block can be used for enzyme histochemistry, such as acetyl-cholinesterase (5) and acid phosphatase activities, and can be stored for possible molecular analyses. Full-thickness bowel also should be fixed for electron microscopy. Blocks, however, must be taken so that it is possible to recognise the circular and longitudinal muscle coats. Achieving this good liaison between professionals is crucial.
In any patient suspected to have an enteric neuromuscular disease, it is important that HD is excluded on a suction rectal biopsy (5). Suction/full-thickness rectal biopsy is generally not helpful in the diagnosis of other enteric neuromuscular diseases, which requires full-thickness bowel taken more proximally to avoid the physiologically normal hypoganglionic bowel adjacent to the pectinate line (21,22).
However, any suspicion of an intestinal ganglioneuroma (11,12) (sometimes incorrectly labelled as intestinal neuronal dysplasia, although the histological appearances are specific) in a rectal biopsy may be a sign of transmural ganglioneuromatosis, and must be followed up by germline M918T and/or A883F mutation analysis of the RET protooncogene to exclude or confirm the diagnosis of multiple endocrine neoplasia type 2B. Children with transmural intestinal ganglioneuromatosis and the above-mentioned molecular diagnosis of multiple endocrine neoplasia type 2B inevitably develop medullary thyroid carcinoma (12). Monitoring the calcitonin concentrations and scanning for adrenal and thyroid masses is not enough because microscopic medullary thyroid carcinoma can be present without raised calcitonin concentrations even with pentagastrin stimulation or without identifiable masses on imaging. A prophylactic thyroidectomy is recommended, as well as continued surveillance of the adrenal glands for evidence of phaeochromocytoma (23).
A specific condition that appears to be familial is hypertrophic eosinophilic gastroenteropathy (15). This entity is characterised by profound hyperplasia/hypertrophy of the intestinal villi with grossly increased villous/crypt ratio and prominent mucosal eosinophilia. Increased eosinophils also are present in the muscularis propria and/or submucosa. The most remarkable finding is that apoptotic enterocytes are undetectable in the elongated villi. It seems likely that a defect in the regulation of apoptosis of the epithelium occurs, which could explain the remarkably lengthened villi seen.
Degenerative neural conditions are common in adults, but rarely encountered in the paediatric population. Neuronal degeneration has been seen in a large Italian kindred (13). On ultrastructural examination, myenteric neurons appear shrunken and condensed. On silver staining, abnormal irregular neurons are noted in older children. This technique is unreliable in infants, however, because normal young babies often lack argyrophilic neurons (24). The genetic defect in these patients is not known, but the condition links to Xq28 (13).
Enteric myopathies include those diseases in which there is the absence of a specific muscle coat or a presence of an additional muscle layer in the bowel wall (6). Others are characterised particularly by circular muscle fibrosis and vacuolation and/or atrophy of myocytes (6), as well as alterations in contractile proteins. The majority of contractile protein abnormalities are acquired, but it is conceivable that congenital/genetic defects may exist. The cocktails of contractile protein isoforms vary in different muscles. For instance, the ratio of actin isoforms is different in normal vascular (tonic) smooth muscle than in intestinal (phasic) smooth muscle. The ratio of α-smooth muscle actin to γ-smooth muscle actin is far greater in tonic than in phasic muscles. Similarly, during gut development actin-isoform ratios in enteric smooth muscles change. In early myogenesis the α -smooth muscle actin isoform predominates, but later the predominant actin isoform is γ -smooth muscle actin.
Two particular phenotypes of enteric muscle disease are recognised. In 1 there is an abnormal distribution of myocyte nuclei with areas of crowding and others devoid of nuclei in the circular muscle coat. A pink blush is detected in the areas of nuclear paucity on picrosirius staining. Ultrastructural examination reveals separation of leiomyocytes from each other by a granular proteinaceous material in which there are occasional collagen fibres. Consequently, this type is referred to as pink-blush myopathy with nuclear crowding (6). It is postulated that the proteinaceous material results from alterations of smooth-muscle phenotype from contractile to secretory. The second phenotype shows prominent fibrosis together with atrophy and dropout of leiomyocytes, and on acid phosphatase histochemistry a presence of active lysosomes in the smooth-muscle cells is seen. On electron microscopy, abnormal vacuolated myocytes are evident. The vacuoles are dilated lysosomes, which contain electron-dense degradation products. This phenotype is known as myopathy with autophagic activity (6).
Because these appearances may be subtle, immunohistochemistry and ultrastructural examination is needed. These features are seen in congenital disorders, but may also be the result of inflammation with gross changes in the end stage. Prognosis is generally poorer in myopathies than in neural disorders because the abnormalities often are present throughout the gastrointestinal tract (25). Decompression ileostomy usually is required and full-thickness samples can be taken at the time of stoma fashioning. Many require parenteral nutrition, but in recent years when this has been given at home problems have been greatly reduced. A small number of patients, however, proceed to small-bowel transplantation because of restricted venous access or liver complications.
During the past couple of decades, our systematic approach has provided improved understanding and more reliable diagnosis and treatment of intestinal neuromuscular diseases in children. This is only a beginning, and there is still much more to be learned. The future includes continued careful phenotypic descriptions of disease processes, including histopathology. Because these conditions are rare, in each phenotype one is likely to encounter only a small number of affected individuals. Therefore, data pooling between centres worldwide using carefully agreed-on parameters may be necessary. In addition, continuing basic science research into developmental processes and genetic pathways will provide information helpful in identifying genetic defects and formulating innovative treatment options.
1. Smith VV, Gregson N, Foggensteiner L, et al
. Acquired intestinal aganglionosis and circulating auto-antibodies without neoplasia or other neural involvement. Gastroenterology 1997; 112:1366–1371.
2. Ruuska TH, Karikoski R, Smith VV, et al
. Acquired myopathic intestinal pseudo-obstruction may be due to autoimmune enteric leiomyositis. Gastroenterology 2002; 122:1133–1139.
3. Schäppi MG, Smith VV, Milla PJ, et al
. Eosinophilic myenteric ganglionitis
is associated with functional intestinal obstruction. Gut 2003; 52:752–755.
4. Schaeppi MG, Smith VV, Milla PJ, et al
. Distal ileal obstruction syndrome in cystic fibrosis is associated with myenteric lymphocytic ganglionitis
. Neurogastroenterol Motil 2003; 15:232.
5. Smith VV, Tozzi A. La biopsia da suzione rettale nella diagnosi della malattia di Hirschsprung: stato dell'arte. Il bollettino sigep. Soc Ital Gastroenterologia Pediatr 2003; 11:2–5.
6. Smith VV, Milla PJ. Histological phenotypes of enteric smooth muscle disease causing functional intestinal obstruction in childhood. Histopathology 1997; 31:112–122.
7. Smith VV. Intestinal neuronal density in childhood: a baseline for the objective assessment of hypo- and hyperganglionosis. Pediatr Pathol 1993; 13:225–237.
8. Smith VV, Lake BD, Kamm MA, et al
. Intestinal pseudo-obstruction with deficient smooth muscle alpha actin. Histopathology 1992; 21:535–542.
9. Gattuso JM, Smith VV, Kamm MA. Altered contractile proteins and neural innervation in idiopathic megarectum and megacolon. Histopathology 1998; 33:34–38.
10. Navarro J, Sonsino E, Boige N, et al
. Visceral neuropathies responsible for chronic intestinal pseudo-obstruction syndrome in pediatric practice: analysis of 26 cases. J Pediatr Gastroenterol Nutr 1990; 11:179–195.
11. Eng C, Marsh DJ, Robinson BG, et al
. Germline RET codon 918 mutation in apparently isolated intestinal ganglioneuromatosis. J Clin Endocrinol Metab 1998; 89:4191–4194.
12. Smith VV, Eng C, Milla PJ. Intestinal ganglioneuromatosis and multiple endocrine neoplasia type 2B: implications for treatment. Gut 1999; 45:143–146.
13. Auricchio A, Brancolini V, Casari G, et al
. The locus for a novel syndromic form of neuronal intestinal pseudo-obstruction maps to Xq28. Am J Hum Genet 1996; 58:743–748.
14. Erdohazi M. Retarded development of enteric nerve cells. Dev Med Child Neuro 1974; 16:365–368.
15. Cuperuus R, Schäppi MG, Shah N, et al
. Hypertrophic eosinophilic gastroenteropathy is associated with reduced enterocyte apoptosis. Histopathology 2005; 46:73–80.
16. Hogaboam CM, Hewlett B, Stead RH, et al
. Class 2 major histocompatibility antigen complex (MCH2) and intracellular adhesion molecule (ICAM-1) expression on murine intestinal smooth muscle cells. Clin Invest Med 1994; 17:236.
17. Khan I, Blennerhassett MG, Kataeva GV, et al
. Interleukin 1β induces the expression of interleukin 6 in rat intestinal smooth muscle cells. Gastroenterology 1995; 108:1720–1728.
18. Smith VV, Schaeppi MG, Lindley KJ, et al
. Neuronal proteinase-activated receptors are involved in mast-cell-neuronal interactions within the stomach. Neurogastroenterol Motil 2003; 15:233–234.
19. Borrelli O, Schaeppi MG, Knafelz D, et al
. Mast cell-nerve interaction is critical for food allergic intestinal motility. J Pediatr Gastroenterol Nutr 2001; 32:348.
20. Richardson CE, Morgan JM, Jasani B, et al
. Megacystis-microcolon-intestinal hypoperistalsis syndrome (MMIHS) and the absence of the α3 nicotinic acetylcholine receptor subunit. Gastroenterology 2001; 121:350–357.
21. Milla PJ, Smith VV. Intestinal neuronal dysplasia. J Pediatr Gastroenterol Nutr 1993; 17:356–357.
22. Cord-Udy C, Smith VV, Ahmed S, et al
. An evaluation of the role of suction rectal biopsy in the diagnosis of intestinal neuronal dysplasia. J Pediatr Gastroenterol Nutr 1997; 24:1–6.
23. Ashworth M. The pathology of pre-clinical medullary thyroid carcinoma. In: Vizjak A, Ferluga D, Bussolati G (eds). Update in Pathology. Proceedings of the 19th European Congress of Pathology
. Ljubljana, Slovenia: University of Ljubljana; 2003:317–8.
24. Smith VV, Milla PJ. Argyrophilia in the developing myenteric plexus. Br J Biomed Sci 1996; 53:278–283.
25. Heneyke S, Smith VV, Spitz L, et al
. Chronic intestinal pseudo-obstruction: treatment and long term follow-up of 44 patients. Arch Dis Child 1999; 81:21–27.