Summary: Background: Lymphoid aggregates are normally found throughout the small and large intestine. Known as lymphoid nodular hyperplasia (LNH), these aggregates are observed especially in young children and are not associated with clinical symptoms being considered 'physiological'. In children presenting with gastrointestinal symptoms the number and size of the lymphoid follicles are increased. Patients suffering from gastrointestinal symptoms (i.e. recurrent abdominal pain) should systematically undergo gastroduodenoscopy and colonoscopy. With these indications LNH, especially of the upper but also of the lower gastrointestinal tract has been diagnosed, and in some children it may reflect a food hypersensitivity (FH) condition. Aim: To review the literature about the relationship between LNH and FH, particularly focusing on the diagnostic work‐up for LNH related to FH. Methods: We reviewed literature using Pubmed and Medline, with the search terms 'lymphoid nodular hyperplasia', 'food hypersensitivity', 'food allergy' and 'food intolerance'. We overall examined 10 studies in detail, selecting articles about the prevalence of LNH in FH patients and of FH in LNH patients. Results: Collected data showed a median of 49% (range 32–67%) LNH in FH patients and a median of 66% (range 42–90%) FH in LNH patients. Literature review pointed out that the most important symptom connected with LNH and/or FH was recurrent abdominal pain, followed by diarrhoea and growth retardation. Both LNH and FH are associated with an increase in lamina propria γ/δ+ T cells, but the mechanisms by which enhanced local immune responses causing gastrointestinal symptoms still remain obscure. Conclusions: When assessing FH, we rely on clinical evaluation, including elimination diet and challenge tests, and endoscopic and immunohistochemical findings. Considering the possible co‐existence of duodenal and ileo‐colonic LNH, upper endoscopy can be recommended in children with suspected FH, especially in those presenting with additional upper abdominal symptoms (i.e. vomiting). Likewise, lower endoscopy might be additionally performed in patients with suspected FH and LNH of the duodenal bulb, also presenting with lower abdominal symptoms (i.e. recurrent abdominal pain).
Lymphoid follicles, with or without germinal centres, are normally found throughout the small and large intestine. In the terminal ileum these coalesce to form Peyer's patches. In the colon, the number of lymphoid structures increases from the caecum to the rectum. Lymphoid tissue is particularly prominent in the anorectal region, just above the dentate line of the anal canal.[[
In the gastrointestinal tract, the key step to antigen processing, evolving either into oral tolerance or a strong immune response against antigens administered orally, involves its initial uptake from the gut lumen by specialised follicle‐associated epithelium called 'M' cells. M cells originate from adjacent crypt epithelium and are interspersed between the absorptive epithelial cells in the follicle‐associated epithelium. These cells take up macromolecules, viruses, bacteria and protozoa within 30 min from the initial presentation of the antigen in the intestinal lumen. After the initial uptake of antigen by M cells, the antigens are transported into the follicular areas to be processed by dendritic cells and brought into close contact with the antigen‐specific precursors for IgA‐secreting plasma cells. The final result of M‐cell processing is the production of a vigorous secretory IgA response and local cell‐mediated immunity with suppression of a systemic IgG, IgE and delayed‐type hypersensitivity to orally administered antigens.[
Lymphoid nodular hyperplasia presents endoscopically as smooth, yellow‐white nodules up to 2 mm in diameter. They are observed more frequently in young children and are not associated with clinical symptoms, in which form LNH is considered to be 'physiological'. In children presenting with gastrointestinal symptoms, i.e. abdominal pain and rectal bleeding, the number and size of lymphoid follicles is increased and the anatomical distribution of enlarged lymphoid aggregates is age‐dependent. It has been demonstrated that upper gastrointestinal symptoms in children could be related to food allergy. However, it is also known that the prevalence of food allergy declines with increasing age in adult life and this could be linked to the marked decrease in number of Peyer's patches and in LNH. In addition, adult subjects showing LNH at endoscopy examination have been reported to suffer more frequently from food allergy.[
Intestinal LNH has also been associated with viral and bacterial infection and immune deficiency status, thus it exists as both a physiological and as a pathological lesion.[
Food hypersensitivity (FH) is the umbrella term used to describe both food allergy, which involves the immune system (i.e. IgE and non‐IgE‐mediated mechanisms), as well as food intolerances, which do not.[
Food hypersensitivity may produce a wide variety of patchy or diffuse mucosal lesions in any part of the gastrointestinal tract, depending on the genetic characteristics of the subject and the type of the intervening immunological reaction. The changes may manifest clinically as buccal rash, aphthous wounds of the oral cavity, gastro‐oesophageal reflux or oesophagitis, gastritis in the stomach, villous atrophy in the small intestine or colitis and may finally appear as an itchy rash around the anus. Although most of the lesions are limited to certain areas, the changes may occasionally spread throughout the whole gastrointestinal tract, a condition which is very often characterised by eosinophilic mucosal infiltrate and is termed 'eosinophilic gastroenteropathy'.[[
Food hypersensitivity in infants is more often generalised, with symptoms of dermatitis and/or mucosal reactions of the respiratory tract to corresponding allergies. In school‐age children or adults, in whom gastrointestinal symptoms are more prominent, symptoms vary from malabsorption with loose stools or growth retardation and anaemia, to regurgitation, discomfort and abdominal pain and constipation.[
The diagnosis is nowadays based on elimination diets and on repeated open oral provocation tests, but the 'gold diagnostic standard' is the double‐blind placebo‐controlled challenge.[
As most subjects with FH are examined only for the main and more severe symptoms, the overall extent and macroscopic and histological characteristics of the related lesions have often remained obscure. In particular, infants and children with FH have not been systematically scheduled for endoscopic examination. Knowledge of mucosal abnormalities has been based on small intestinal biopsies from limited selected patient series. Colonoscopy has also been indicated in selected patients with severe chronic or recurrent abdominal pain, or presenting with chronic constipation refractory to laxative therapy or blood in stools. With these indications LNH, especially of the mucosa of the upper gastrointestinal tract (i.e. duodenal bulb), has been diagnosed and in some children it may reflect a condition of overt FH. On the other hand, it must be underlined that in most of these studies, the diagnosis of FH was not properly managed (nonblinded, not placebo‐controlled). However, previous reports have demonstrated this lesion in patients without FH, presenting with common variable immunodeficiency, hypogammaglobulinaemia, IgA deficiency, HIV infection, HP infection, giardiasis, juvenile idiopathic arthritis and connective tissue disease or intestinal malignant lymphoma, as well as in healthy subjects.[[
Some studies have demonstrated a significant association between FH and LNH distributed diffusely on the duodenal bulb. Lymphoid nodules on the duodenal mucosa are distinctly more numerous in subjects with FH than in patients in whom any other state or coeliac disease is diagnosed. FH in duodenal LNH subjects was mostly of the gastrointestinal type. Even at onset it manifested with symptoms of vomiting or recurrent abdominal pain, and after oral challenge gastrointestinal symptoms were the first to be manifested, although in some patients dermatitis was exacerbated later on. Moreover, in most duodenal LNH patients it is not possible to demonstrate atopic allergy by skin prick tests or elevated serum IgE values. Thus, the patients with duodenal bulb LNH mostly present a local reactivity of delayed‐type FH.[[
It has been also demonstrated that an intensive humoral immune response to cow's milk proteins and their fractions may be associated with gastrointestinal symptoms (i.e. recurrent abdominal pain) and duodenal LNH, even in children over the age of 3 years. Endoscopic findings in these patients included local and patchy hyperplasia of the lymphoid tissue and immunological evidence showed that the untreated cow's milk allergy children had significantly higher levels of IgA and IgG class antibodies to whole cow's milk and its specific fractions than age‐matched controls.[
Cow's milk allergy can also exist in school‐age children. Moreover, there is increasing evidence that this allergy may also be a common problem in young adults with undefined and prolonged gastrointestinal complaints. Both endoscopic and histological alterations are most prominently found in the duodenum bulb, where LNH or lymphoid follicles are commonly seen. As in coeliac disease, the increase in γ/δ+ T‐cells is characteristic of cow's milk allergy. However, unlike in coeliac disease, patients with cow's milk allergy do not present antiendomysium antibodies, HLA‐DQ2 and ‐DQ8 genotypes, abnormal HLA‐DR expression or duodenal villous changes. Taken together, these results indicate that the mucosal changes are because of a different pathogenetic mechanism in cow's milk allergy than in coeliac disease. Moreover, this histological picture cannot be considered as precursor to coeliac disease, because its striking features are mucosal lymphoid nodules with normal villous architecture.[
Unlike duodenal LNH, LNH of the mucosa of the lower gastrointestinal tract (i.e. terminal ileum and colon) in children was for several years considered a serendipitous finding without any clinical significance, as an expression of a mucosal response to nonspecific stimuli, most often infection. It was consequently regarded as a pathophysiological phenomenon during infancy and childhood, or, in a lower percentage of cases, as a sign within the spectrum of inflammatory bowel diseases.[[
More recent studies[
In agreement with previous reports on the frequency of colonic LNH in both children and adult patients with haematochezia,[[
In adults suffering from recurrent rectal bleeding, we demonstrated that LNH not associated to other endoscopic signs was very often caused by multiple food hypersensitivity.[
Other symptoms associated with endoscopic evidence of isolated colonic LNH were anaemia and growth deficiency, recently recognised as part of the clinical spectrum of FH.[
Iacono et al.[
Finally, Kokkonen et al.[
Tables and summarised the study regarding LNH and FH. Figure shows a typical picture of colonic LNH in a patient suffering from FH, diagnosed at the 'Di Cristina' Pediatric Hospital of Palermo.
Graph: Typical picture of colonic lymphoid nodular hyperplasia in a patient suffering from food hypersensitivity.
Lymphoid nodular hyperplasia (LNH) in food hypersensitivity (FH) patients
Untreated FH: 22/22 (100%) Treated FH: 14/14 (100%) Untreated FH: 11/22 (50%) Treated FH: 5/14 (36%) Total FH: 21/36 (58%) Recurrent abdominal pain: 6/44 (14%) Total: 27/134 (34%) Cow's milk FH: 15 Suspected cow's milk FH: 12 Coeliac disease: 11 Controls: 12 Total: 50 Cow's milk FH: 6–4 Suspected cow's milk FH: 7–14 Coeliac disease: 8–15 Controls: 6–15 Cow's milk FH: 15/15 (100%) Suspected cow's milk FH: 12/12 (100%) Cow's milk FH: 9/15 (60%) Suspected cow's milk FH: 8/12 (67%) Coeliac disease: 1/11 (9%) Controls: 2/12 (17%) Total: 20/50 (40%) Diagnosed FH: 36/102 (35%) Suspected FH: 24/102 (24%) 18/36 with diagnosed FH (50%) 8/24 with suspected FH (33%) Untreated FH: 20 Treated FH: 17 Coeliac Disease: 12 Controls: 12 Total: 61 Untreated FH: 20 (33%) Treated FH: 17 (28%) Untreated FH: 14/20 (70%) Treated FH: 8/17 (47%) Coeliac Disease: 0 Controls: 0 Total: 22/61 (36%)Text reference Authors Clinical presentation Cases number Age (years range) Cases number and % of patients affected with FH Cases number and % of patients affected with LNH Kokkonen J Recurrent abdominal pain 84 1.6–15 28/84 (33%) 9/28 (32%) Kokkonen J Recurrent abdominal pain or growth retardation Untreated FH: 22 Treated FH: 14 Recurrent abdominal pain without FH: 44 Healthy controls: 54 Total: 134 3–15 Kokkonen J Recurrent abdominal pain and diarrhoea Turunen S Constipation 35 3–15 12/35 (34%) 8/12 (67%) Kokkonen J Recurrent gastrointestinal symptoms and/or growth retardation 102 1–15 Kokkonen J Gastrointestinal symptoms 2–15
Food hypersensitivity (FH) in lymphoid nodular hyperplasia (LNH) patients
Total: 73/245 (30%) Isolated LNH without IBD: 52/245 (21.2%) Total FH: 103/245 (42%) FH with isolated LNH without IBD: 43/52 (83%) Colonic LNH: 46/140 (33%) 36 without colitis (26%) 9 with colitis (6%) 2 with Crohn (0.7%) Ileocolonic LNH: 50/74 (67%) 35 without colitis (47%); 13 with colitis (17%); 2 with Crohn (3%). Duodenal bulb LNH: 22/102 (22%) Colonic LNH: 25/46 (54%) Ileocolonic LNH: 18/50 (36%) Duodenal bulb LNH: 13/22 (59%)Text reference Authors Clinical presentation Cases number Age (years range) Cases number and % of patients affected with LNH Cases number and % of patients affected with FH Kokkonen J Recurrent abdominal pain 63 1.9–15 12/63 (19%) 9/12 (75%) Iacono G Severe chronic or recurrent abdominal pain, constipation, diarrhoea, bloody diarrhoea, growth retardation, chronic vomiting 245 3–10 Carroccio A Rectal bleeding 64 18–79 10/64 (16%) 9/10 with LNH (90%) Kokkonen J Chronic colitis, IBD, systemic vasculitis 140 3–16
An increase in the density of γ/δ+ T‐cell receptor‐bearing intraepithelial lymphocytes is considered to be specific to gluten‐sensitive enteropathy, although it must be recognised that the real role of γ/δ+ T‐cells in this disease is still enigmatic (i.e. they may remain elevated even on a gluten‐free diet or decrease on elimination diet). In addition, their normal density has not yet been precisely determined.[[
However, in children with FH and LNH of the duodenum bulb it is possible to demonstrate a higher than normal density of γ/δ+ intraepithelial lymphocytes, together with a higher density of IFN‐γ positive cells in the lamina propria and a higher proportion of crypt cells in mitosis, as well as a rise in intraepithelial γ/δ+ T‐cells in untreated FH, whereas patients on an elimination diet at the time of the examination show low counts. Thus, γ/δ+ T‐cells, which are considered to be markers of FH, seem to react more rapidly during dietary therapy and only increase in active and untreated cases. LNH could represent a more long‐standing or even permanent up‐regulation of immunological response to food antigens, persisting even after the deleterious antigen has been removed from the diet.[[
As regards to LNH of the terminal ileum, some authors have demonstrated that this pathology is associated with increased densities of γ/δ+ T‐cells, as in children with LNH of the duodenal bulb and untreated FH. Surprisingly, they found not only this association but also a close association between an increase in these cells and right‐sided colitis and pancolitis. In contrast, neither Crohn's disease nor typical left‐sided ulcerous colitis showed any association. Thus, it would be possible to classify colitis in children into γ/δ+ or 'allergic' and γ/δ‐ or 'non‐allergic diseases'. The former seems to be an entity starting from the terminal ileum as lymphoid hyperplasia and continuing into the caecum as chronic colitis, extending in a few cases as far as the rectum. Clinically, this disease resembles what is traditionally classified as 'indeterminate colitis'. The other form of colitis seems to start locally on the left side, probably on the rectal mucosa, and extends proximally. Clinically, this entity is compatible with typical ulcerous colitis. A differential diagnosis might be made by simply counting the density of γ/δ+ T‐cells on the terminal ileum mucosa. Thus, it seems evident that in Crohn's disease and left‐sided colitis, different cytokine cascades and pathogenetic mechanisms to those involved in right‐sided colitis become activated.[
At present, no immunohistochemical studies on colonic biopsies are available of patients with colonic LNH, either related or unrelated to FH.
Finally, the mechanisms by which the enhanced local immune responses (i.e. LNH and γ/δ+ intraepithelial T‐cells) cause gastrointestinal symptoms remain mostly obscure. FH symptoms have been suggested as being due to a cytokine imbalance. Children with delayed‐type cow's milk allergy showed lower IL‐2 and IL‐18 mRNA expression in the duodenum and higher CCR‐4 and IL‐6 mRNA expression in the terminal ileum, compared with controls and children with coeliac disease. The mRNA expression levels of regulatory cytokines (i.e. transforming growth factor‐beta and IL‐10) remained similar in patients and controls. Children with delayed‐type gastrointestinal cow's milk allergy showed a unique pattern of local intestinal hypersensitivity with TH2 response‐related characteristics, a profile clearly differing from those with coeliac disease.[[
The autistic spectrum disorder, or ASD, is a neurodevelopmental disorder characterised by socially aloof behaviour and impairment of language and social interaction. Gastrointestinal symptoms are described in 9–54% of autistic children, the most common of which are constipation, diarrhoea and abdominal distension. The gastrointestinal abnormalities reported in autism include adverse reaction to foods (i.e. non‐IgE‐mediated FH), inflammation (oesophagitis, gastritis, duodenitis, enterocolitis), increased intestinal permeability, low activities of disaccharidase enzymes, dysbiosis with bacterial overgrowth, impairment of detoxification (i.e. defective sulfation of ingested phenolic amines) and exorphin intoxication (by opioids, i.e. caseomorphine and gluteomorphine derived from casein and gluten). A beneficial effect of dietary intervention on the behaviour and cognition of some autistic children indicates a functional relationship between the alimentary tract and the central nervous system.[[
In addition to the above‐mentioned gastrointestinal abnormalities, some authors also demonstrated that several ileo‐colonic LNH in children with ASD, referred to a gastroenterology clinic for chronic gastrointestinal symptoms, warranted investigation by ileo‐colonoscopy.[[
If the data on gastrointestinal manifestations in autistic children continue to mount, then the basis for a link will need to be thoroughly investigated to ascertain whether the intestinal and cognitive manifestations are merely different components of a complex syndrome of unknown pathogenesis, or could be causally related.
It is possible to hypothesise a primitive, poorly understood immune dysfunction that predisposes autistic children to a broad spectrum of immunologically based disorders, i.e. eczema, upper respiratory tract infections, FH, as well as ileo‐colonic LNH. The impaired cognitive function could then be explained on the basis of an 'entero‐colonic encephalopathy', analogous to hepatic encephalopathy. Specifically, the intestinal lesion results in an increased intestinal permeability to exogenous peptides of dietary origin, ultimately leading to the disruption of neuroregulatory mechanisms and normal brain development. These possibilities should be investigated further.[[
In the presence of LNH, demonstrated during upper and/or lower endoscopy performed for severe gastrointestinal symptoms, the possible aetiologies and the related diagnostic procedures are shown in Table , where we considered FH together with many other diseases. For diagnostic purposes, the presence of lymphoid follicles in the biopsy specimen and a rise in γ/δ+ T‐cells alone would not be sufficient in patients with suspected FH showing endoscopic picture of LNH of the duodenal bulb and/or of the terminal ileum and the colon. The diagnostic problem is the patchy nature of all these signs, as they can be dispersed at any one site along the gastrointestinal tract from the oral cavity to the anus. In addition, there is a significant lack of serological markers for a diagnostic assessment of mucosal abnormality. At present, when clinically assessing FH, we rely on a clinical evaluation, including a lengthy elimination diet and challenge tests, together with endoscopic and/or immunohistochemical findings as a sign of enhanced immune activity.
Lymphoid nodular hyperplasia: aetiologies and diagnostic procedures
Aetiologies Diagnostic procedures IgA deficiency Laboratory measurement of IgA level in the blood Common variable immunodeficiency Low serum IgG concentrationReduced serum concentrations of other immunoglobulins, i.e. IgA or IgMLow CD4+ count, associated with a normal to increased CD8+ count Food hypersensitivity Prick tests Radioallergosorbent test (RAST) Elimination diet and oral food challenge Human immunodeficiency virus (HIV) infection HIV enzyme‐linked immunosorbent assay (ELISA) HIV Western blot test Human T‐cell lymphotropic viruses (HTLV)‐I & II infection HTLV ELISA HTLV Western blot test Detection of proviral DNA by Polymerase Chain Reaction (PCR) Quantitative PCR assay Cytomegalovirus (CMV) infection Serological assays, i.e. CMV IgM and IgG antibodies by ELISA and Radio Immune Assay (RIA) CMV IgG avidity Epstein‐Barr Virus (EBV) Infection Serological assays for viral capsid antigen (IgG and IgM), early antigen (IgG) and EBV uclear antigen (IgG) antibodies Recovery of ELISA detection of Giardiasis Stool ova and parasites detection Juvenile idiopathic arthritis (JIA) History Physical examination findings Laboratory studies help to exclude underlying disorders, classify type of arthritis and evaluate for extra‐articular manifestations Imaging of affected joints is usually indicated Other connective diseases History Physical examination findings Non‐organ‐specific autoantibodies Laboratory assays and imaging techniques to detect e type organ injury Intestinal lymphoma Endoscopy and biopsy Wireless video capsule camera Abdominal computerised tomography (CT) with intravenous contrast Arteriography of the intestinal arteries (in case of tumour bleeding) Exploratory surgery
Considering the possible co‐existence of duodenal and ileo‐colonic LNH, upper endoscopy can also be recommended in children with suspected FH and ileo‐colonic LNH, especially in those presenting with additional upper abdominal symptoms (i.e. vomiting and dyspepsia). Likewise, lower endoscopy might be additionally performed in patients with suspected FH and endoscopically diagnosed LNH of the duodenal bulb also presenting with lower abdominal symptoms (i.e. severe chronic or recurrent abdominal pain, or chronic constipation refractory to laxative therapy or presence of blood in stools).[
Declaration of personal and funding interests: None.
By P. Mansueto; G. Iacono; A. Seidita; A. D'Alcamo; D. Sprini and A. Carroccio
Reported by Author; Author; Author; Author; Author; Author