Dady, Ian M.; Thomas, Adrian G.; Miller, Victor; Kelsey, Anna J.*
Inflammatory bowel disease (IBD) in older children is an increasingly recongnised problem in the U.K.; there has been, for example, a threefold increase in the incidence of Crohn's disease between 1968 and 1983 (1). Major problems in children include poor growth, delay of puberty, and psychological effects. Flexible fibreoptic endoscopy allows visualisation of the colon and upper gastrointestinal tract, which, coupled with biopsy and radiological studies, has greatly facilitated the diagnosis of childhood IBD (2,3).
Earlier case reports of infantile IBD were based on autopsy or laparotomy findings or by sigmoidoscopy and barium studies. There are sporadic descriptions of both ulcerative colitis (4-7) and Crohn's disease (8,9), but before the age of 12 months, both are rare. A series of six infants with Crohn's disease were seen within an 18-month period, all with acute onset of small-bowel obstruction (10). At laparotomy segmental transmural inflammation and ulceration was seen in the terminal ileum, with variable degrees of extension into the colon. Recently, four children with Crohn's disease whose symptoms began in early infancy were reported, but all except one responded well to steroids or azathioprine (11).
Food allergy, particularly cow's milk protein intolerance, has been suggested as the major cause of infantile colitis. A study of eight children <2 years of age with endoscopically proven colitis showed that all of them responded to a rigorous exclusion diet (12). However, in 1986 a mixed group of conditions were described in a prospective study of 10 children <3 years old with colitis established by biopsy (13). Five had onset of symptoms in the first years; two of them had cow's milk protein intolerance, and three had indeterminate colitis. Their outcomes were good apart from one who developed small-bowel enteropathy and mouth ulcers after colectomy. This case was later included in a description of five children with “intractable ulcerating enterocolitis of infancy” (14), all of whom had intractable diarrhoea, ulcerating stomatitis, and colitis characterised by flask-shaped large ulcers with overhanging edges. All except one infant had severe perianal disease and partial villous atrophy; failure of medical treatment necessitated subtotal colectomy in all five.
This report describes seven infants born in the northwest and Mersey regions of England between 1990 and 1993 with intractable diarrhoea commencing before 4 months of age. The total number of live births during this period in these two regions was ≈351,000. All seven infants had chronic nonspecific IBD unrelated to infection or cow's milk protein intolerance and were highly resistant to treatment. IBD unrelated to infection or cow's milk protein intolerance was thought to be rare infancy and is an unusual finding in infants with intractable diarrhoea. The prevalence is unknown. A better understanding of the nature and relationship of the two conditions may help define the aetiology or aetiologies and hence direct future treatment strategies.
PATIENTS AND METHODS
The seven patients described in this report were born between January 1990 and June 1993 and subsequently were referred to the regional paediatric gastroenterology unit. Their characteristics and clinical features are shown in Tables 1 and 2, respectively. Initially, at least three stool samples each were sent for virology, bacterial culture, and search for ova, cysts, and parasites, and then again at regular intervals. Investigations for immune deficiency included total assessment of immunoglobulins and IgG subclasses, lymphocyte classes, and nitrobluetetrazolium tests. Gut autoantibodies were assessed at regular intervals throughout each patient's clinical course. Small-bowel biopsy was performed endoscopically or by Watson capsule, and six had a small-bowel enema radiograph. Further radiological examinations were not usually performed because disease activity was evident from the clinical features, inflammatory blood tests, and endoscopy or small-bowel biopsy.
All patients had been given a soya formula before referral, and a hypoallergenic diet was subsequently instituted with one or more of the following: a casein hydrolysate (Pregestimil, Mead Johnson), an elemental feed (Neocate, Scientific Hospital Supplies), or a hospital-prepared modular feed made from either comminuted chicken meat (Cow and Gate) or Vamin 9 glucose (Kabi Pharmacia) (15). An elemental, modular, or modular elemental feed was used for ≥6 weeks, often with calorific supplementation (Table 3). In addition, pectin was used in all infants (except patients 1 and 3) as a source of short-chain fatty acids. The infants were weighed daily, and stool volumes were estimated by weighing soiled nappies. The appearance of the stool was documented, and they were tested for pH, reducing substances, and sugar chromatography.
Limited colonoscopy (Olympus PCF 10) with biopsy was performed at the onset of bleeding or perianal disease; more extensive views were often posible during subsequent follow-up procedures. One child (patient 7) was not tested endoscopically until her younger sister (patient 6) showed signs of rectal bleeding and florid colitis at the age of 2 weeks. Inflammatory activity was monitored regularly using the erythrocyte sedimentation rate, serum orosomucoid, and faecal alpha-1-antitrypsin levels. Serum copper, zinc, selenium, and vitamins A and E were measured monthly while the infants were on parenteral nutrition and every 3 months otherwise. Supplementation was given where low levels were recorded.
Steroid therapy was given to all patients except patient 7. Initially, prednisolone was administered orally at a dose of 2 mg/kg per day, but if ther was no response, then intravenous hydrocortisone was commenced. If the infant became cushingoid and or failed to improve, steroid therapy was stopped, and an alternative treatment was considered. Thus patients 2 and 4 were treated with azathioprine, and patients 1, 2, and 4 were given intravenous cyclosporine. Sodium cromoglycate was used in infants with a eosinophilic inflammatory cell infiltrate found on biopsy specimens who failed to recover with steroids (patients 2, 5, and 6). Response to treatment was assessed by the resolution of the diarrhoea and rectal bleeding, improvement in the inflammatory parameters, the ability to thrive when not supported by parenteral nutrition, and a second endoscopy or small-bowel biopsy.
All seven patients had intractable diarrhoea in the first 4 months, four within 2 weeks of birth (Table 1). Four infants had rectal bleeding by 5 months of age, but patients 1 and 2 developed it at 26 and 11 months, respectively. Vomiting and perianal involvement were also seen in several patients (Table 2), and mouth ulcers were present in patients 1 and 4. Patient 1 also had recurrent respiratory infections, skin nodules, and fever; the most striking initial features of patient 4 were persistent vomiting and aspiration, multiple perianal skin tags (Fig. 1) and developmental delay. No patient had eye or joint manifestations or any other features of autoimmune disease. Initial investigations showed negative stool microbiology at onset, apart from infant 3, who had type 3 adenovirus. The results of immune testing are shown in Fig. 2; gut autoantibodies were repeatedly negative.
There was no response to dietary manipulation despite a long duration of treatment with protein exclusion and elemental feeding (Table 3). This finding effectively excluded cow's milk protein intolerance, and parenteral nutrition (PN) was instituted in all seven infants for a mean duration of 65 weeks (range 4-195); patients 2, 6, and 7 remain on long-term PN at home. Reduced trace element and vitamin levels were frequently documented, requiring enteral or parenteral supplementation. Raised inflammatory parameters and hypoalbuminaemia were seen in all seven patients, and five had evidence of protein-losing enteropathy with a raised faecal alpha-1-antitrypsin level (mean 12.2 mg/g dried faeces, range 4.4-33, normal value <2). The histological findings are summarised in Fig. 3. Small-bowel biopsy showed partial villous atrophy in four patients and total villous atrophy in two. In five cases there was an inflammatory cell infiltrate, and two had an excess of eosinophils. In cases 1, 2, 6, and 7, these biopsies were taken while parenteral as well as enteral nutrition was being provided. Neither of the two patients with total villous atrophy were receiving parenteral nutrition at the time. None of the biopsy specimens had the histological features of congenital microvillous atrophy, and where electron microscopy was undertaken (cases 1, 2, and 6) the appearances were normal.
Small-bowel enema studies showed bowel wall thickening in the jejunum and proximal ileum in two cases, patient 1 developed a stricture of the terminal ileum following colectomy. In the remaining three children there was no radiological abnormality of the small bowel. Endoscopy showed the macroscopic features of proctocolitis in all seven and upper-gastrointestinal-tract inflammation in two. The inflammation was initially confluent and variable in intensity, being most severe in the rectum and sigmoid colon. Patient 2 later developed punched-out colonic ulcers. Appearances also progressed in patients 1 and 4 with pseudopolyp formation, but in the latter patient the ulcers were small and localised with normal intervening mucosa. The response to treatment is shown in Fig. 4 and the disease progression for each patient in Fig. 5. Prolonged steroid therapy was used, except in patient 7, who had very mild colitis. Four of the remaining six failed to improve, but patient 3 is now in complete remission. In patient 5 there was some reversal of the inflammation and villous atrophy, although he still had diarrhoea and poor weight gain. Those patients given azathioprine and intravenous cyclosporine showed no clinical improvement, and mesalazine and cromoglycate were also ineffective.
Two patients have undergone colectomy for severe colitis; although patient 2 showed a temporary improvement, both have subsequently relapsed with small-bowel disease. Currently, apart from patient 3, all the surviving children continue to have diarrhoea; three are on long-term parenteral nutrition, and the rest have failed to thrive. In addition, three patients have experienced serious complications of long-term central venous access in the form of superior vena caval thrombosis or obstruction.
To date, two children have died—one (patient 4) suddenly at the age of 19 months while being treated with hydrocortisone and cyclosporin for extensive colitis and one (patient 1) at the age of 56 months from chronic renal failure. Postmortem consent was refused for the latter child, but detailed histological examination of his colectomy specimen had already been performed (Fig. 3). In the other child (patient 4) an E. coli infection was identified from blood and lung cultures with histological evidence of pneumonia, presumably as a complication of immunosuppressive therapy. There was also marked brain atrophy, involution of the thymus, and diffuse colitis affecting the whole of the large bowel with severe perineal ulceration and skin tags. Mucosal ulceration and crypt abscesses were seen (Fig. 6), but there were no granulomata. In the proximal colon these changes were mild and confined to the mucosa, but distally more severe inflammation was seen (Fig. 7) with fissuring and focal inflammation extending to the serosa. Macroscopic examination of the upper gastrointestinal tract and small bowel appeared normal, but microscopic appearances were marred by postmortem autolysis.
Seven children were referred over a 3-year period with intractable diarrhoea of infancy, in whom nonspecific inflammatory bowel disease was subsequently diagnosed; this combination had not been seen by us before 1990. None responded to exclusion of cow's milk or soya, and infection was identified in only one patient (Adenovirus). All required parenteral nutrition, and only one achieved remission with steroids. The response to colectomy was also disappointing.
The prevalence and significance of IBD in children with intractable diarrhoea are unknown, although the case of an infant with congenital diarrhoea and intestinal inflammation had been reported (16). Also increased mucosal T-cell numbers have been identified in a subgroup of seven of 13 children with intractable diarrhoea and villous atrophy (17).
Within our cohort of patients there was considerable difference in the clinical features, distribution, nature, and severity of the histological changes. The onset of IBD also varied as the delay between presentation with intractable diarrhoea and rectal bleeding ranged from 0-26 months. In patient 7 intractable diarrhoea was the main clinical problem, and endoscopy was not contemplated until her younger sister developed a florid colitis shortly after birth.
Where pathological examination of the whole colon was possible after colectomy or at postmortem (cases 1, 2, and 4), severe nonspecific inflammatory bowel disease was found, confirming the biopsy findings. There were insufficient criteria for a firm diagnosis of Crohn's disease to be made, although all three had either previously or subsequently developed small-bowel or upper-gastrointestinal-tract pathology. All seven patients were classified as having chronic nonspecific IBD, and although patient 1 had some features of intractable ulcerating enterocolitis of infancy (14), the ulcers were not large or flask-shaped, as described in this condition. None of the seven infants fulfilled clincal or histological criteria for Behcet's disease (18) or collagenous colitis (19). Microscopic colitis (20) was excluded by the abnormal macroscopic appearances of our patients at endoscopy.
Small-bowel mucosal injury and disaccharidase deficiency are a common accompaniment to intractable diarrhoea of infancy (21). In our series there was an inflammatory response in five and villous atrophy of variable severity in six. Villous atrophy has been described in animals exclusively given total parenteral nutrition; however, all our patients were given some enteral nutrition during these periods of intravenous feeding. Moreover, patients 3 and 5, who had the most severe villous atrophy, were not receiving parenteral nutrition at the time of their small-bowel biopsies. Parenteral nutrition thus does not appear to be the primary cause of the villous atrophy found in these infants with IBD. In addition, there was no evidence to support a diagnosis of autoimmune enteropathy or congenital microvillous atrophy. An eosinophilic infiltration was noted in four of our patients, but the lack of symptom periodicity, peripheral eosinophilia, and prompt response to steroids typically seen in eosinophilic enteropthy (22) made this diagnosis untenable.
Comparing patients 3 and 5 with the other children, it appears that both had total villous atrophy, yet less severe proctocolitis and a better response to steroid therapy. Their outcomes are also superior since the other five have either died or required long-term parenteral nutrition at home. Apart from a later onset, there is little to differentiate them from the others in their clinical features, family history, or early feeding practices. They shared the common endoscopic findings of nonspecific proctocolitis and so may represent the better end of a single disease spectrum. If one is to postulate a single disease process for all these cases, then it must have considerable clinical and pathological heterogeneity. Although they are rare, Crohn's disease and ulcerative colitis can both occur in infancy, but none of our patients had the subacute obstruction that has been described in the infantile form of Crohn's (10). Perhaps given time to evolve, the classic discontinuous panintestinal and transmural inflammatory changes would manifest.
Alternately, there may be more than one disease process. IBD may occur insidiously in a proportion of patients and immediately in others. We speculate that these infants with intractable diarrhoea who are found to have changes reflective of IBD well after the original symptoms began could have developed it as a consequence of nutritional deficiency. Recently, the importance of glutamine and short chain fatty acids (SCFA) has been recognised in small-and large-bowel function, respectively (23). SCFA, predominantly butyrate, has been shown to alter colonic sodium and water absorbtion, mucosal blood flow, and mucin production (24). SCFA deficiency has been proposed as a cause of the colitis and diarrhoea seen in time of famine, and absent luminal SCFA supplies after defunctioning colostomy may lead to diversion colitis. Irrigation of SCFA into the rectal stump of a patient with diversion colitis has reversed the macroscopic and histological evidence of inflammation (25).
All patients, except patients 1 and 3, were given pectin, which is converted to SCFA by bacterial degradation in the colon. However, the age at which this diet was commenced varied from 6 to 24 months, by which time the colitis associated with SCFA deficiency might have already become well established.
In conclusion, seven infants with intractable diarrhoea of infancy had features of chronic nonspecific IBD. All had proctocolitis, and six had small-bowel or upper-gastrointestinal-tract involvement. The heterogeneity in their presentation and histology may indicate more than one disease process. The true prevalence of the condition will be known only if endoscopy is undertaken in cases of intractable diarrhoea. Further work is required to investigate the pathophysiology and possible significance of nutritional deficiencies in this condition.
Acknowledgment: We gratefully acknowledge the help of John Stanton in the nutritional treatment of these patients.
1. Barton JR, Gillon S, Ferguson A. Incidence of inflammatory bowel disease in Scottish children between 1968 and 1983: marginal fall in ulcerative colitis, three-fold rise in Crohn's disease. Gut 1989;30:618-22.
2. Chong SKF, Bartram C, Campbell CA, Williams CB, Black-shaw AJ, Walker-Smith JA. Chronic inflammatory bowel disease in childhood. Br Med J 1982;284:101-3.
3. Chong SKF, Blackshaw AJ, Boyle S, Williams CB, Walker-Smith JA. Histological diagnosis of chronic inflammatory bowel disease in childhood. Gut 1985;26:55-9.
4. Gryboski JD. Ulcerative colitis in children 10 years old or younger. J Pediatr Gastroenterol Nutr 1993;17:24-31.
5. Ein SH, Lynch MJ, Stephens CA. Ulcerative colitis in children under one year: a twenty year review. J Pediatr Surg 1971;6:264-71.
6. Enzer NB, Hijmans JC. Ulcerative colitis beginning in infancy. J Pediatr 1963;63:437-43.
7. Fonkalsrud EW, Barker WF. Ulcerative colitis in infancy. Surgery 1963;54:819-24.
8. Koop CE, Perlingiero JG, Weiss W. Cicatrizing enterocolitis in a newborn infant. Am J Med Sci 1947;214:27-32.
9. Borer F. Crohn's disease in infants. Gastroenterologia 1961;95:48-56.
10. Miller RC, Larsen E. Regional enteritis in early infancy. Am J Dis Child 1971;122:301-11.
11. Deslandres C, Paradis K, Rasquin A, Roy C, Seidman E. Outcome of patients with inflammatory bowel disease diagnosed before the age of two. Gastroenterology 1993;105:A690.
12. Jenkins HR, Pincott JR, Soothill JF, Milla PJ, Harries JT. Food allergy: the major cause of infantile colitis. Arch Dis Child 1984;59:326-9.
13. Chong SKF, Blackshaw AJ, Morson BC, Williams CB, Walker-Smith JA. Prospective study of colitis in infancy and early childhood. J Pediatr Gastroenterol Nutr 1986;5:352-8.
14. Sanderson IR, Risdon RA, Walker-Smith JA. Intractable ulcerating enterocolitis of infancy. Arch Dis Child 1990;65:295-9.
15. Thomas AG, Stanton JC, Miller V. The nutritional management of chronic diarrhoea and/or malabsorbtion. J Pediatr Gastroenterol Nutr 1990;1:142-3.
16. Kanof ME, Rance NE, Hamilton SR, Luk GD, Lake AM. Congenital diarrhoea with intestinal inflammation and epithelial immaturity. J Pediatr Gastroenterol Nutr 1987;6:141-6.
17. Cuenod B, Brousse N, Goulet O, et al. Classification of intractable diarrhoea in infancy using clinical and immunohistological criteria. Gastroenterology 1990;99:1037-43.
18. Chong SKF, Wright VM, Nishigame T, et al. Infantile colitis: a manifestation of intestinal Behcet's syndrome. J Pediatr Gastroenterol Nutr 1988;7:622-7.
19. Perisic VN, Kokai G. Diarrhoea caused by collagenous colitis. Arch Dis Child 1989;64:867-9.
20. Mashako MNL, Sonsino E, Navarro J, et al. Microscopic colitis: a new cause of chronic diarrhoea in children? J Pediatr Gastroenterol Nutr 1990;10:21-6.
21. Rossi TM, Lebenthal E, Nord KS, Fazili RR. Extent and duration of small intestinal mucosal injury in intractable diarrhoea of infancy. Pediatrics 1980;66:730-5.
22. Steffen RM, Wylie R, Petras RE, et al. The spectrum of eosinophilic gastroenteritis. Clin Pediatr 1991;30:404-11.
23. Roediger WEW. The starved colon: diminished mucosal nutrition, diminished absorbtion, and colitis. Dis Colon Rectum 1990;33:858-62.
24. Scheppach W. Effects of short chain fatty acids on gut morphology and function. Gut 1994;suppll:S35-8.
25. Harig JM, Soergel KH, Komorowski RA. Treatment of diversion colitis with short chain fatty acid irrigation. N Engl J Med 1989;320:23-8.