A mass of lymphoid tissue or lymphonodular hyperplasia (LNH) in the terminal ileum (TI) is a common finding during pediatric colonoscopies, so much so that it has even been considered normal or age related (1–2). However, its significance has remained obscure. We recently showed that it is associated with food allergy (FA) (3–4) when located on the mucosa of the other extremity of the small intestine, the bulb of the duodenum. Lymphonodular hyperplasia or increased lymphoid tissue on the mucosa of the colon is rarely seen and has been described in the literature in case reports or small series (5–7). In some patients, LNH may be related to inflammatory bowel disease (IBD), whereas in other patients FA is an associated disorder (8). These findings lead to the assumption that LNH may be an expression of locally up-regulated immune mechanisms and thus associated with immune-mediated disorders. In this study, we sought to investigate the prevalence of LNH on the mucosa of the colon and the TI and to see how often this lesion is associated with IBD or FA.
PATIENTS AND METHODS
We enlisted all children who had undergone a full colonoscopic examination, performed by one pediatric gastroenterologist (J. K.) at the Department of Pediatrics, Oulu University Hospital, during the years 1997 to 2000. The 4-year cohort included 140 children (mean age, 10.4 ± 3.9 years; range, 3–16 years; 51 girls) who underwent a total of 190 colonoscopies. The first endoscopic examination of each patient was used in the analysis of the data. Although the aim in each case was to reach the TI, the TI was assessed only in 74 patients (54%). In our clinical practice, all scopic examinations are performed during general anesthesia and in an operating room. If a patient intended to undergo colonoscopy had upper gastrointestinal symptoms or if Crohn disease was suspected, gastroduodenoscopy was scheduled at the same session. Accordingly, the foregut of 102 (72%) subjects also was studied.
Based on clinical information and on endoscopic and histopathologic findings, the patients were grouped into the following categories:
Those with IBD, chronic colitis. Patients were classified into two main groups according to distribution of inflammation in the endoscopic and histologic examinations: 1) typical ulcerative colitis with left-sided involvement, the cecum and ascending colon being fully normal; and 2) right-sided or pancolitis, both either exhibiting the involvement of the cecum and ascending colon or with inflammation extending all through the colon.
Those with IBD, Crohn disease. Diagnosis was made in this study only if aphthous ulcerations with epithelioid granulomas were seen in histologic examination. Accordingly, as seen during endoscopic examination, all patients had segmental colitis with normal mucosa between.
Those with inflammatory conditions secondary to local or systemic disorders, such as systemic vasculitis.
Those patients showing lymphoid mass or LNH of the colon or the TI without other endoscopic findings or histologic alterations compatible with colitis.
Those children who were gastrointestinally healthy, with no significant findings during endoscopic study.
Diagnosis of Food Allergy
All subjects in whom the patient or the parents related exacerbation of symptoms to food ingestion or any other suspicion of food-related symptoms (milk allergy in infancy, abdominal pains or diarrhea after eating, self-diagnosed lactose intolerance) were designated for FA evaluation. Until July 30, 1999, we used open oral challenges (n = 42); thereafter, we used masked (n = 19) oral challenges. The test protocols have been described in previous reports (3,10). In short, the earlier diagnosis of FA was only accepted if the patient had immunoglobulin E (IgE)–specific antibodies and a positive elimination challenge test result, or if in IgE-specific antibody-negative cases, two separate challenges were positive. If the diagnosis was not established before endoscopic examination, the subjects were scheduled, after a period of 2 weeks' elimination, for oral challenges with milk and cereals (gluten-containing cereals and oats separately). Each period was continued for 1 week if symptoms did not manifest earlier. Abdominal pain, diarrhea, loose mucous stools, and exacerbation of dermatitis were considered positive results. In the masked challenges, rice flour was used as a carrier, and the subjects kept a daily diary about the occurrence of their symptoms and symptom intensity.
Colonoscopies were performed using an Olympus GF-Q1401 (Olympus, Tokyo, Japan) during general anesthesia. Biopsy specimens were taken for routine histology from at least three separate areas of the colon, each from the site where the most demonstrative local disease was seen. A fourth routine biopsy specimen was obtained from the TI if that level was reached. Upper intestinal endoscopies were performed using an Olympus GIF-XQ 140 gestroscope (Olympus, Tokyo, Japan). Biopsy specimens were taken for routine histology from the duodenum below the sphincter of Oddi, the antrum of the stomach, and the lower part of the esophagus, each from the site where the most demonstrative local disease was seen. All biopsy specimens were evaluated according to the established general criteria as clinical samples intended to establish a definite clinical diagnosis.
Assessing Lymphonodular Hyperplasia
In this study, assessing lymphoid nodularity on the mucosa of the colon or the TI was based on endoscopic evaluation after filling the area to be inspected with air. Only a cluster of lymphoid nodules (n > 10 nodules) was considered significant (Fig. 1 and 2). In the TI, a mass of lymphoid tissue and the presence of LNH were inspected in the 10-cm to 15-cm area that could be seen during the procedure.
The data collected from the history notes and food-challenge diaries were analyzed using the SPSS package (SPSS, Chicago, IL, U.S.A.). The Student t test and χ 2 test were used to estimate the significance of the differences between the subgroups. The Ethical Committee of Oulu University Hospital accepted the study protocol.
The final study group included seven subjects with left-sided colitis (three girls; mean age, 13.2 ± 2.0 years), 31 subjects with right-sided colitis with cecum and ascending colon affected or pancolitis (14 girls; mean age, 11.5 ± 3.2 years), 8 subjects with Crohn disease (5 girls; mean age, 13.5 ± 1.8 years), and 36 subjects with LNH on the mucosa of the colon or TI without colitis (11 girls; mean age, 8.6 ± 3.6 years), whereas the remaining 58 subjects (18 girls; mean age, 9.2 ± 4.1 years) had no endoscopic findings at all (Table 1).
Three patients with right-sided colitis had an associated disorder: chronic aggressive hepatitis in two patients and polyarthritis in one. Moreover, another patient had polyarthritis and one patient had vasculitis who had LNH of the TI but no colitis.
LNH of the Colon
Among the 140 subjects, LNH of the colon was diagnosed in 46 subjects (32%), as shown in Figure 1. Only in 4 subjects was the lesion dispersed throughout the colon, being patchy in one or more of the four categorized areas in the rest. In 23 subjects, the rectum was affected, the sigma in 27, the transverse colon in 23, and the cecum in the other 16.
Lymphonodular hyperplasia was associated with IBD in 10 of 46 patients, being more prevalent among the colitis patients (9/38) than among the Crohn patients (1/8).
LNH of the Terminal Ileum
Lymphoid mass or LNH of the TI (Fig. 2) was diagnosed in 50 of 74 (68%) subjects in whom this endpoint was reached. The valve of Bauchini markedly protruded inside the colon in three subjects, all with an abundance of lymphoid tissue at the end of the TI. In general, LNH of the TI was found in all endoscopic groups. It was demonstrated in half of the subjects with colitis, whereas it was strikingly rare in subjects with Crohn disease.
Upper Gastrointestinal Findings
A total of 22 of 102 subjects undergoing gastroduodenoscopy (22%) had LNH on the duodenal bulb (Table 2). It was found significantly more often in subjects with LNH of the colon (χ 2 = 12.4, P < 0.0001) than in those without. However, LNH of the TI showed no significant association with LNH of the foregut.
A total of 47 subjects showed at least one endoscopic alteration in the gastroduodenoscopy, the lesions being most prevalent in subjects with Crohn disease (4/6) or with diagnosed FA (17/27). Of the various specific lesions, gastric erosions were diagnosed in 23 subjects, esophagitis in 17, and duodenitis in 10.
In the whole study group, 37 children (26%) could be defined as having FA. In 11 subjects, the tests had been performed before admission for the endoscopic examination and they were on elimination diets according to the foodstuff they reacted to. Five of them had been positive for IgE-specific antibodies. At the time of the study, 50 subjects had undergone elimination challenge tests, 26 of whom showed a positive reaction, whereas the other 24 tested negative. Milk was the foodstuff causing the reaction in 23 patients and wheat in 4, whereas 10 subjects reacted to both wheat and milk. In seven subjects, FA was related to right-sided colitis (n = 4) or pancolitis (n = 3), whereas no patients with left-sided colitis or Crohn disease could be designated as having FA.
Overall, treated or untreated FA showed a highly significant association with the presence of LNH on the colon (χ 2 = 27.2, P < 0.0001;Table 3). This association was significant even in the ileoscopic material (χ 2 = 5.3, P < 0.05), but not strikingly so. Of the 22 children with LNH of the bulb of the duodenum, 13 either had or had had FA (χ 2 = 8.4, P < 0.004).
This retrospective analysis of a clinical and endoscopic series shows that LNH is a common endoscopic bystander on the mucosa of the lower GI tract. Our results also support the view that this is not just a silent finding but is related to enhanced immunologic activity, FA being the most common underlying state. Indeed, if present on the mucosa of the colon, it shows an association with FA similar to that in the foregut. If present on the mucosa of the TI, immune-mediated disorders other than FA may also have produced the lesion. Within the spectrum of IBD, FA could be diagnosed in some patients with colitis but in none with Crohn disease. Thus, subjects diagnosed as having LNH of the lower GI tract should be carefully assessed for the possibility of hypersensitivity to basic foodstuffs by whatever techniques are appropriate, including long oral challenges.
As far as we understand and could see from the literature, this is one of the first extensive studies to show the prevalence and significance of LNH or increased lymphoid tissue on the mucosa of the lower gastrointestinal tract in endoscopic examinations. The other accounts of LNH in the literature are mostly case reports or small series (5–7). Colon et al. (8) found the association with LNH of the colon and IBD in 13 subjects in a 10-year retrospective study in which patients were examined mostly with radiography, whereas 84 had no underlying disorder. However, Wakefield et al. (9) very recently reported an endoscopic finding of LNH of the TI with histologically reactive follicular hyperplasia in most children with developmental disorders (93%). The same group later showed enhanced local immune response with a high density of intraepithelial gamma-delta positives T cells and other markers of gut epithelial dysfunction (11). Our conjecture is that LNH of the colon may have remained unreported because most of the patients usually have no underlying colitis, which is usually looked for. However, as also observed in this report, LNH of the TI is such a common finding that it has been considered a normal state (7). The true prevalence of LNH in healthy subjects remains unknown because colonoscopy is only indicated in patients with a high probability of specific diseases and with severe symptoms, such as blood in the stools or long-term diarrhea.
The prevalence of FA as diagnosed by history and positive elimination-challenge test was significantly higher in patients with LNH of the colon than in those without FA. The association between LNH and FA was even more apparent in the colon than in the foregut, in which case we recently reported a 75% coincidence. Even the presence of LNH on the mucosa of the TI showed a significant association with FA, but not as significant as LNH of the colon. Of note, the FA in our patients were mostly of the gastrointestinal type with a local reaction and without dermatitis or other general symptoms relating to IgE-mediated reactions. In a recent study, we found evidence that cell-mediated immunity may be active and may cause the symptoms (4). In these cases, as in the current series, the diagnosis may be evidenced only by an elimination and challenge test. However, as the symptoms of FA beyond infancy seem to be dependent on the dose and the duration of the antigen exposure, the 1-week challenge may be too short for lower gastrointestinal symptoms to become obvious. Thus the prevalence of FA in this series should be considered as a minimum.
Two thirds of patients with LNH of the colon showed similar lesions on the mucosa of the duodenum. These patients also showed a high prevalence of FA. Of note, the coexistence of LNH in these two locations could not be demonstrated in all patients, confirming the view that FA in children may produce patchy or diffuse LNH in any part of the gastrointestinal tract (12). Lymphonodular hyperplasia of the colon was seen even in children being diagnosed for FA and treated with an elimination diet, supporting the view that the increment in the lymphoid tissue remains for a long time, even after the triggering foodstuffs have been eliminated. We found the same in the duodenum (4). Moreover, in a very recent report, we found that the rise in intraepithelial gamma-delta positive T cells also showed a close association with untreated FA and LNH, but that subjects on a diet at the time of the examination showed low counts (4). Thus the γς seem to react more rapidly and arise only in active and untreated cases. Lymphonodular hyperplasia could represent a more long-standing or even permanent up-regulation of immunologic response to food antigens.
The association between IBD and definitely diagnosed FA was another highlight of the study. Seven of 31 patients with right-sided or pancolitis could be designated as having FA, whereas none with left-sided colitis or Crohn disease exhibited the same association. This observation was a surprise, because in separate and well-controlled studies, antigen avoidance or an exclusively whole protein diet has been suggested as the primary therapy for children with Crohn disease (13). If we consider LNH as an immunologic response to luminal antigens, this study suggests that patients with right-sided or pancolitis, but not with Crohn disease, would benefit from the elemental diet. This postulation, of course, needs to be tested by a real study. At any rate, according to this study and even immunologically, Crohn disease with epithelioid granulomas seems to be an entirely different entity in the spectrum of IBD than is right- or left-sided colitis.
In conclusion, we found formal evidence that LNH on the mucosa of the colon or TI is not just an innocent bystander. If detected on the colon, it seems more suggestive of gastrointestinal FA. Being diagnosed in the TI, it may be related to a variety of immunologic states. The analysis also presents options to pediatric endoscopists. The TI should always be reached.
1. Bines JE, Winter HS. Lower endoscopy. In: Walker WA, Durie PR, Hamilton JR, et al., eds. Pediatric Gastroenterology. Philadelphia: Decker; 1991: 1256–69.
2. Rossi T. Endoscopic examination of the colon in infancy and childhood. Pediatr Clin North Am 1988; 35: 331–56.
3. Kokkonen J, Karttunen T, Niinimäki A. Lymphonodular hyperplasia as a sign of food allergy in children. J Pediatr Gastroenterol Nutr 1999; 29: 57–62.
4. Kokkonen J, Holm K, Karttunen TJ, Mäki M. Children with untreated food allergy express a relative increment in the density of duodenal γς T-cells. Scand J Gastroenterol 2000; 35: 1137–42.
5. Kaplan B, Benson J, Rothstein F, et al. Lymphonodular hyperplasia of the colon as a pathologic finding in children with lower intestinal bleeding. J Pediatr Gastroenterol Nutr 1984; 3: 704–8.
6. Gottrand F, Erkan T, Fabriaux J-P, et al. Food-induced bleeding from lymphonodular hyperplasia of the colon. Am J Dis Child 1993; 147: 821–3.
7. Beaoui M, Guezmir M, Hamdi M, et al. Lymphoid hyperplasia of the intestine in children. 15 cases. Ann Pediatr (Paris) 1992; 39: 359–64.
8. Colon A, DiPalma J, Leftridge C. Intestinal lymphonodular hyperplasia of childhood: patterns of presentation. J Clin Gastroenterol 1991; 13: 163–6.
9. Wakefield AJ, Anthony A, Murch SH, et al. Enterocolitis in children with developmental disorders. Am J Gastroenterol 2000; 95: 2154–6.
10. Kokkonen J, Tikkanen S, Savilahti E. Residual intestinal disease after milk allergy in infancy. J Pediatr Gastroenterol Nutr 2001; 32: 156–61.
11. Furlano RI, Anthony A, Day R, et al. Colonic CD8 and gamma delta T-cell infiltration with epithelial damage in children with autism. J Pediatr 2001; 138: 366–72.
12. Crowe S, Perdue M. Gastrointestinal food hypersensitivity: basic mechanisms of pathophysiology. Gastroenterology 1992; 103: 1075–95.
13. Ruuska T, Savilahti E, Mäki M, et al. Exclusive whole protein enteral diet versus prednisolone in the treatment of acute Crohn's disease in children. J Pediatr Gastroenterol Nutr 1994; 19: 175–80.
© 2002 Lippincott Williams & Wilkins, Inc.