A large number of children in low- and middle-income countries experience weight loss and chronic diarrhea, which is defined as diarrhea persisting for >12 weeks (1). Many of them remain undiagnosed and the etiology is often obscured (2). Microscopic colitis (MC) has been reported as a possible cause of chronic diarrhea in adults, but its role in children with nonbloody diarrhea has not yet been investigated extensively (3–5).
In this study, our objectives were to determine whether MC was one of the reasons for chronic nonbloody diarrhea in children with a normal colonic endoscopy, in which other established etiologies have been excluded, and to characterize the colonic pathology with special reference to the MC in these children. Being one of the earliest studies of this kind in children with diarrhea in India, we designed it as a descriptive study. It was hoped that this preliminary study would pave the way for subsequent studies on MC prevalence, pathogenesis, and other aspects of the disease, if found to be significant in this age group.
The study was conducted on patients with chronic diarrhea visiting the pediatric gastroenterology outpatient clinic of the All India Institute of Medical Sciences for a period of 1 year. One hundred consecutive patients ages 3 to 12 years with nonbloody diarrhea (defined as passage of ≥3 loose stools per day) of >12 weeks’ duration were investigated for a possible infective, metabolic, or biochemical cause of chronic diarrhea. The workup included stool examination including wet smears and modified Kinyoun acid-fast stain, stool culture for infectious causes, anti-endomysial antibody and anti-tissue transglutaminase for celiac disease, esophagogastroduodenoscopy and intestinal biopsies for celiac disease, abetalipoprotenemia and lymphangiectesia along with sweat chloride tests, and radiological studies wherever indicated. These patients were enrolled in the study if no specific etiology could be found and colonoscopy did not reveal any abnormality. Those enrolled in the study were evaluated for the presence of MC, that is, lymphocytic colitis (LC) and collagenous colitis (CC) as defined below. Colonic biopsies from patients with MC were compared with those who had chronic diarrhea with no evidence of MC.
For the second comparison, 100 consecutive children presenting to the clinic with bleeding per rectum were examined by colonoscopy and were included as a comparison group if they had colonic mucosal changes on the endoscopy in the absence of vascular or polyp changes. Colonic biopsies from these patients were compared separately with patients with MC.
For the ease of analysis, these 3 groups were labeled as group A (patients of chronic diarrhea with MC), group B (chronic diarrhea patients without MC), and group C (patients with bleeding per rectum). Informed consent was taken from guardians of both subjects and comparison group. Ethical clearance was obtained from the institutional ethics committee.
A lower gastrointestinal (GI) endoscopy up to the level of the sigmoid colon was performed under sedation using a standard protocol followed in the pediatric ward. At least 4 mucosal biopsy fragments were taken from the rectum and rectosigmoid junction in each case. Two of these biopsies were oriented on black filter papers in such a way that there was no rolling of tissue and then fixed in 10% buffered formalin for 3 to 8 hours before the paraffin blocks were prepared. One piece was collected in tinfoil and transported in an ice and salt mixture to the tissue culture laboratory immediately and stored at −70°C. The fourth piece was collected in 4% glutaraldehyde after cutting into smaller pieces of approximately 1-mm3 size and were kept in cacodylate buffer at 4°C until they were further processed for transmission electron microscopic (TEM) examination. Postfixation was performed in 1% osmium tetroxide. After following the standard protocol, blocks were prepared and ultrathin sections were stained with lead acetate and lead citrate for the final TEM viewing.
The detailed histological analysis was performed on hematoxylin and eosin (H&E)–stained slides. It was decided a priori that 2 pathologists blinded to the case would examine the biopsies and any diagnosis would be confirmed only when both pathologists agreed to the same findings. Colonic biopsies of all of the subjects and comparison group were specifically assessed by 2 histopathologists with experience in evaluating GI mucosal biopsies for the presence or absence of LC and CC, the 2 histological subtypes of MC. LC was defined when the colonic mucosa showed an increase in the intraepithelial lymphocytes (IEL) count of >20/100 (normal: 5/100) colonic epithelial cells and there was a diffuse inflammatory cell infiltrate in the lamina propria (6–8). Similarly, CC was diagnosed if the colonic mucosa showed a subepithelial thick collagenous band (ie, >10 μm) and evidence of colitis, that is, inflammatory cell infiltrate in colon (6–8).
IEL count was done by counting the number of lymphocytes per 100 epithelial cells from the surface epithelium and crypts manually, by a pathologist specially trained in interpreting GI biopsies. For determination of IEL count, all of the biopsy fragments in a particular case were evaluated. A minimum number of approximately 2000 epithelial cells was counted to reach an average value. Semiquantitation of various inflammatory cells such as neutrophils and lymphocytes in the lamina propria was done by counting at least 500 inflammatory cells in both the upper and lower halves of the lamina propria and was expressed as percentages. Mast cells were similarly counted after toluidine blue staining.
Immunohistochemical staining was used to characterize the mononuclear inflammatory cells and enteroendocrine cells in 5 biopsies of MC and randomly selected 6 biopsies from group B and 4 from group C. Standardization of the primary antibodies for immunohistochemistry was done on tissues as follows: lymph node and tonsil were used for standardization of CD3 (1:150; Novocastra, Newcastle, UK), CD4 (1:10; Novocastra [Citrate pH 8]), CD8 (1:100; Novocastra), CD20 (1:200; Novocastra), CD68 (1:1000; BioSB, Santa Barbara, CA), and immunoglobulin G (IgG) (1:300; DAKO, Glostrup, Denmark); nasal polyp was used for IgE (1:50; LifeSpan Biosciences, Inc, Seattle, WA); large intestine for IgA (1:200; DAKO), S-100 (1:200; DAKO), PGP-9.5 (1:100; DAKO); and adrenal medulla for synaptophysin (1:20; DAKO). PGP 9.5 and synaptophysin were used as markers of the enteroendocrine cells. Approximately 1000 mononuclear cells in the lamina propria of each biopsy were examined under the microscope for positive cytoplasmic immunohistochemical staining and results were expressed as percentages.
Morphometry and image analysis were done in all 5 cases of MC, 20 cases from group B, and in randomly selected 24 patients from group C. Biopsies that had well-oriented sections with at least 4 to 5 crypts arranged parallel to each other were considered optimal for morphometry and image analysis. Image analysis was done using image analyzer software, Image Proplus 6 (Media Cybernetics, Rockville, MD), to study the following 4 parameters: total height of mucosa, height of crypts, basement membrane thickness, and epithelial cell thickness. The crypt and mucosal heights were studied on H&E-stained sections; basement membrane thickness was measured on sections stained with Sirius red stain. Alcian blue periodic acid-Schiff stain was used to highlight the mucosal epithelial cells, for morphometric measurements of the epithelial cell heights. First, as stated, the most oriented part of the biopsy fragments was identified and the basement membrane thickness was measured under the surface mucosa. In a precalibrated image analysis software, the thickness was measured at a total magnification of ×400.
Stool samples and colonic contents collected during colonoscopy by an infant feeding tube of both subjects and comparison group were subjected to routine microscopic examination and culture and sensitivity. A detailed TEM study was done on biopsies from 12 subjects, that is, randomly selected 4 subjects each from the 3 groups. Histology, immunohistochemistry, and electron microscopy findings of group A were compared with group B and group C separately. Nonparametric tests were used for analysis and results were considered significant if P value was <0.05.
Of the 26 patients of chronic diarrhea studied, 5 cases had MC, including 3 LC and 2 CC (Fig. 1). There was no subject in which there was a disagreement on the biopsy findings between the 2 pathologists. The median ages (range) of groups A, B, and C were 101.2 (43–132), 70 (80–130), and 130 (90–140) months, respectively. Patients with MC (group A) had no evidence of inflammatory bowel disease in the form of cryptitis, crypt abscesses, or crypt distortion, nor did they have any granuloma. The remaining 21 cases (group B) showed chronic nonspecific colitis defined as chronic lymphoplasmacytic infiltration of colonic mucosa without any specific pattern or distribution of inflammation. Epithelial cell damage and regenerative changes were not found in any of the biopsies studied. No pathogens were identified.
Among the 45 patients included in group C with bloody diarrhea, 3 patients had ulcerative colitis identified by chronic nongranulomatous inflammation and ulceration of mucosa with cryptitis, crypt abscesses, and crypt distortion. Three subjects showed ulcerated lesion with crypt disarray and splaying of the smooth muscle fibers and were diagnosed as having solitary rectal ulcer. The remaining 39 subjects had chronic nonspecific colitis.
The median duration (range) of chronic diarrhea among patients diagnosed as having MC was 20 weeks (16–36 weeks). The median number of stool frequency (range) was 6 (4–8) and consistency was liquid in all of the patients. All 5 patients were severely underweight (weight for age <3rd percentile) and height severely stunted (height for age <3rd percentile). They did not experience any autoimmune or allergic conditions. None of the subjects had a history of proton pump inhibitors or nonsteroidal anti-inflammatory drug intake; however, all of these patients had received multiple courses of antibiotics for chronic diarrhea before visiting our clinic, this being a common clinical practice in developing countries.
IEL count (median [range]) was significantly higher in group A as compared with group C (12 [4–32] vs 4 [0–24] in surface epithelium, P = 0.0008 and 18 [3–28] vs 3 [1–17], P = 0.0008 in crypt wall, respectively) (Table 1). IEL count [median (range)] was significantly higher in patients diagnosed as having LC in group A as compared with group C (22 [21–32] vs 4 [0–24] in surface epithelium, P < 0.001 and 21 [20–28] vs 3 [1–17], P < 0.001 in crypt wall, respectively). The minimum IEL count in patients with CC was 4.
Semiquantitation of the inflammatory cells in the lamina propria revealed that neutrophils were significantly higher in the upper half of the lamina propria among group A as compared with group B (median [range]) (13.8% [5.4%–20.6%] vs 7.2% [0%–19.6%], P = 0.03) as well as group C (13.8% [5.4%–20.6%] vs 4 [0%–13.4%], P = 0.007); however, macrophages were found to be significantly higher among group C as compared with group A (31% [11.8%–88%] vs 19.6% [11.6%–22.2%], P = 0.009) (Table 1). Infiltration of mast cells in both the upper and lower halves of the lamina propria was similar in all 3 groups (Table 1); however, these characteristic findings were not present in all of the biopsies from a patient, and even in a mucosal fragment the lesions were patchy.
Image analysis and morphometry of these biopsies revealed increased thickness of the basement membrane in the 2 subjects with CC. The basement membrane thickness in these 2 patients was 10.69 and 10.88 μm; however, there was no significant difference in total height of the mucosa, height of the surface epithelial cells, and crypts among the 3 groups (Table 2).
On the contrary, immunohistochemical studies with CD3, CD8, CD68, IgE, and IgG did not reveal any specific pattern of distribution. The immunohistochemical markers were not expressed differently across the 3 groups. IgE was only occasionally positive across the groups. CD4 was negative in all 3 groups. Immunohistochemical staining for S100, PGP9.5, and synaptophysin expression also did not differ across the groups. The microbiological evaluation of stools and colonic contents did not reveal any significant findings.
Electron microscopic study of the biopsies of group A showed a varied spectrum of changes, ranging from normal ultrastructure to increase in IELs, mononuclear cell infiltration (macrophages, plasma cells, scattered mast cells, and lymphocytes) in the lamina propria, along with distortion of microvilli, dilated intercellular spaces, lysosomal abundance in the cells, and collagen fiber deposition in the subepithelial region, when compared with the other 2 groups. In addition to plasma cells, eosinophils were also seen in the lamina propria. Group B biopsies also showed a significant mononuclear cell infiltration in the lamina propria with focal increase of IELs along with irregular microvillus architecture. Group C biopsies showed normal villous and epithelial architecture with no distortion of microvilli and normal intercellular spaces. The lamina propria showed a few inflammatory cells such as macrophages, plasma cells, lymphocytes, and eosinophils.
All 5 patients were treated with sulfasalazine at the dosage of 2 to 4 g/day orally in divided doses and all of them responded with abatement of diarrheal symptoms in 4 to 6 weeks. Patients remained on this therapy for 6 to 8 weeks and then dose gradually tapered for next 4 to 6 months depending on their clinical response. A low-residue diet was also adapted and dietary secretagogues such as caffeine and dairy products were eliminated from the diet. Patients did not undergo biopsies to confirm histological response.
Chronic diarrhea is an important cause of morbidity and mortality among children in low- and middle-income countries (9,10). Because in most of these patients no specific etiology is found, they remain deprived of any specific targeted treatment. With our study in children with chronic diarrhea of unknown etiology, we found that MC can be an important association in such patients in appropriate settings.
MC is typically a disease of the older age group; therefore, the literature describing MC in children is scarce. Presently, few cases of MC have been reported in children, some of whom had an association with collagenous gastritis, celiac disease, immunodeficiency, juvenile scleroderma, eosinophilic gastritis, Crohn disease, autism, and Aeromonas hydrophila(11–21). In addition to these case reports, there are 2 retrospective studies describing case series of 5 and 11 children with MC, respectively (16,22). This is the first prospective study describing the presence of MC in children with chronic diarrhea, to the best of our knowledge.
Although the diagnosis of MC in adult populations is based on specific diagnostic criteria, in the pediatric population, there is no clarity on its diagnosis. Studies have suggested an association of microscopic changes such as nonspecific inflammation and chronic nonspecific inflammatory bowel disease changes, with chronic diarrhea in the pediatric population (22,23). Of note, both these studies did not report LC or CC as the microscopic changes.
In the index study in children with chronic diarrhea without any identifiable etiology, increased IELs were found to be an important marker of MC, especially in LC. Similarly, subepithelial collagen was an important marker of CC. Besides LC and CC, other atypical forms of MC such as clear cell colitis and eosinophilic colitis have been described in children (20,24). We, however, did not identify such cases in our study.
Electron microscopy of the biopsies studied confirmed the finding of histopathological examination, that is, increased IEL and subepithelial collagen deposition in cases with MC. It also documented the presence of epithelial injury in the form of distortion in microvilli orientation, dilation of intercellular spaces in the crypt epithelium, and increased number of lysosomes in the cells. Similar electron microscopic findings of widened intercellular spaces between epithelial cells and the presence of subepithelial collagen have been reported in other similar studies (3,25); however, immunohistochemical studies with CD3, CD4, CD8, CD20, CD68, IgE, IgA, and IgG did not reveal any specific pattern of distribution of these inflammatory cells. Enterochromaffin cells have been reported to play a role in mucosal immunity and inflammation (26). An attempt was made to identify changes, if any.
Helal et al (27)reported Escherichia coli in 18 of 20 colonic biopsy cultures taken from patients with LC, and case reports have associated MC with Campylobacter jejuni, Clostridium difficle, Yersinia enterocolitica, and Aeromonas(21,28); however, no specific pathogen has been consistently identified in such patients, and stool cultures remained negative (28). We also did not find any significant findings from microbiology studies performed on stool and colonic contents either in the patients with MC or in the disease controls without MC and in controls.
In summary, it appears that MC is present in children with chronic diarrhea in our setting and needs to be carefully investigated. The limitation of this study is in numbers of subjects in each group; however, this could not be avoided, given the preliminary nature of this study. Another limitation of this study was that a full-length colonoscopy was not performed and biopsies were not taken from the proximal colon in these patients. We could have missed a few cases of MC along with other diagnoses such as Crohn disease by this approach. Bile acid malabsorption is known to coexist with MC in adults (29); however, it was not evaluated in these children. Nevertheless, the study highlights the need for carefully evaluating cases of chronic diarrhea in children, especially when the etiology is not apparent after conventional investigations.
The authors are grateful to the Indian Council of Medical Research for supporting this study.
1. Bhatnanagar S, Bhan MK. Persistent diarrhea and malabsorption in children. Indian J Gastroenterol
2. Yachha SK, Misra S, Malik AK, et al. Spectrum of malabsorption syndrome in north Indian children. Indian J Gastroenterol
3. Veress B, Löfberg R, Bergman L. Microscopic colitis
4. Yen EF, Pardi DS. Review article: microscopic colitis
—lymphocytic, collagenous and ‘mast cell’ colitis. Aliment Pharmacol Ther
5. Williams JJ, Beck PL, Andrews CN, et al. Microscopic colitis
: a common cause of diarrhoea in older adults. Age Ageing
6. Mohamed N, Marais M, Bezuidenhout J, et al. Microscopic colitis
as a missed cause of chronic diarrhea. World J Gastroenterol
7. Erdem L, Yildirim S, Akbayir N, et al. Prevalence of microscopic colitis
in patients with diarrhea of unknown etiology in Turkey. World J Gastroenterol
8. Pardi DS, Loftus EV Jr, Smyrk TC, et al. The epidemiology of microscopic colitis
: a population based study in Olmsted County, Minnesota. Gut
9. Mathai J, Raju B, Bavdekar A. Pediatric Gastroenterology Chapter, Indian Academy of Pediatrics. Chronic and persistent diarrhea in infants and young children: status statement. Indian Pediatr
10. El Mouzan MI. Chronic diarrhea in children
: part I. Physiology, pathophysiology, etiology. Saudi J Gastroenterol
11. Kuhn BR, Mezoff AG. Pediatric lymphocytic colitis
presenting with intractable diarrhea. J Pediatr Gastroenterol Nutr
12. Haafiz A, Hu Y, Jolley C, et al. J Pediatr Gastroenterol Nutr
13. Vanderhoof JA, Goble K, Young RJ. Collagenous colitis
in a 4-year-old child: response to budesonide. J Pediatr Gastroenterol Nutr
14. Camarero Salces C, Enes Romero P, Redondo C, et al. Collagenous colitis
and collagenous gastritis in a 9 year old girl: a case report and review of the literature. Acta Gastroenterol Belg
15. Ozturk Y, Soylu OB, Ozer E. Lymphocytic colitis
in a child with non-responsive celiac disease. Acta Gastroenterol Belg
16. El-Matary W, Girgis S, Huynh H, et al. Microscopic colitis
in children. Dig Dis Sci
17. Benchimol EI, Kirsch R, Viero S, et al. Collagenous colitis
and eosinophilic gastritis in a 4-year old girl: a case report and review of the literature. ActaPaediatr
18. O’Beirne JP, Ireland A. Progression of collagenous colitis
to Crohn's disease. Eur J GastroenterolHepatol
19. Esselinckx W, Brenard R, Colin JF, et al. Juvenile scleroderma and collagenous colitis
. The first case. J Rheumatol
20. Chen B, Girgis S, El-Matary W. Childhood autism and eosinophilic colitis. Digestion
21. Camarero C, Leon F, Colino E, et al. Collagenous colitis
in children: clinicopathologic, microbiologic, and immunologic features. J Pediatr Gastroenterol Nutr
22. Mashako MN, Sonsino E, Navarro J, et al. Microscopic colitis
: a new cause of chronic diarrhea in children
? J Pediatr Gastroenterol Nutr
23. Heyman MB, Perman JA, Ferrell LD, et al. Chronic nonspecific inflammatory bowel disease of the cecum and proximal colon in children with grossly normal-appearing colonic mucosa: diagnosis by colonoscopic biopsies. Pediatrics
24. Józefczuk J, Wozniewicz BM. Clear cell colitis: a form of microscopic colitis
in children. World J Gastroenterol
25. Balázs M, Egerszegi P, Vadász G, et al. Collagenous colitis
: an electron microscopic study including comparison with the chronic fibrotic stage of ulcerative colitis. Histopathology
26. Khan WI, Ghia JE. Gut hormones: emerging role in immune activation and inflammation. Clin Exp Immunol
27. Helal TE, Ahmed NS, El Fotoh OA. Lymphocytic colitis
: a clue to bacterial etiology. World J Gastroenterol
28. Pardi DS, Kelly CP. Microscopic Colitis
29. Ung KA, Gillberg R, Kilander A, et al. Role of bile acids and bile acid binding agents in patients with collagenous colitis