Non-IgE-Mediated Gastrointestinal Food Protein-Induced Allergic Disorders: Comparison
Please note this is a comparison between Version 1 by Marina Perez-Gordo and Version 2 by Vivi Li.

Non-IgE-mediated gastrointestinal food allergy (non-IgE-GI-FA) is the name given to a series of pathologies whose main entities are food protein-induced allergic proctocolitis (FPIAP), food protein-induced enteropathy (FPE), and food protein-induced enterocolitis syndrome (FPIES). These are more uncommon than IgE-mediated food allergies, their mechanisms remain largely unknown, and their diagnosis is mainly done by clinical history, due to the lack of specific biomarkers. 

  • food allergy
  • non-IgE-mediated gastrointestinal food allergy
  • Cow’s milk
  • Food protein-induced enterocolitis syndrome
  • Food protein-induced allergic proctocolitis
  • Food protein-induced enteropathy
  • Fecal calprotectin
  • omics

1. Digestive Manifestations of Food Allergy

Food allergy (FA) is a specific immunological response that occurs reproducibly on exposure to a given food. This immune reaction can be due to an IgE-mediated mechanism or a non-IgE-mediated mechanism. IgE-mediated FA is much more common than non-IgE-mediated FA, the symptoms are easily recognized, and the underlying mechanisms are better established [1].
Although the relationship between food allergy and gastrointestinal dysfunction has been known for several years, the exact mechanisms involved have not yet been clearly defined. The diagnosis of non-IgE-mediated gastrointestinal food allergy (non-IgE-GI-FA) is clinical and—due to a lack of a specific biomarker—not always easy. The diagnosis is based on compatible symptoms, the exclusion of other diseases, resolution of the symptoms once the offending food is removed from the diet, and the reappearance of the symptoms when it is reintroduced.
Non-IgE-mediated gastrointestinal food allergy typically comprises three entities: (i) food protein-induced allergic proctocolitis (FPIAP); (ii) food protein-induced enteropathy (FPE), and (iii) food protein-induced enterocolitis syndrome (FPIES).

1.1. Food Protein-Induced Allergic Proctocolitis (FPIAP)

Food protein-induced allergic proctocolitis (FPIAP) is a non-IgE-mediated food allergy characterized by bloody stools in a healthy infant. The symptoms begin within the first months of life; usually in the first two months. It affects both exclusively breastfed infants and formula-fed infants; although it appears to be slightly more common in exclusively breastfed infants (60%) [2].
The most common offending triggers are cow’s milk (CM) proteins (76%), followed by hen’s egg (16%), soy (6%), and corn (2%). In 8% of the cases, the etiology could be multiple [3]. Any food can induce FPIAP and, once the food is removed from the diet, symptoms disappear within a fortnight. Therefore, if the symptoms persist, it is necessary to continue withdrawing food or looking for another diagnosis [4].
The pathophysiological mechanism of FPIAP is not well established. It seems that it involves CD8+ lymphocytes, Th-2 type lymphocytes, and eosinophil infiltrate in all layers of the colonic mucosa [5]. Additionally, some studies suggest the involvement of T-cells in the pathogenesis of this entity, associated with secretion of tumor necrosis factor-alpha (TNF-α) by activated lymphocytes [6].
The typical clinical manifestations are mucus bloody stools in a healthy appearing infant. The bleeding is usually visible although it can appear as occult blood in the stools. Some patients have iron-deficiency anaemia and less frequent hypoalbuminemia and peripheral eosinophilia [7]. Symptoms such as vomiting, frank diarrhoea, or unwell appearance are not present, unlike FPE and FPIES.
FPIAP is a very common cause of rectal bleeding in infants although the real prevalence is unknown. The population prevalence in infants is approximately 1% to 2%. The prevalence among infants with rectal bleeding ranges from 18% to 64% [8]. Elizur et al. found an accumulative incidence in 13,019 newborns over a two-year period of 0.16% [9]. Other studies have reported that FPIAP represents between 18% and 64% of children with rectal bleeding when the diagnosis is based on withdrawal of the offending food or by biopsy findings from the flexible sigmoidoscopy [10][11][10,11].
The diagnosis is clinical and based on the following criteria: (a) small amount of rectal bleeding in an otherwise healthy infant; (b) disappearance of symptoms after all causal food is removed from diet; and (c) exclusion of other causes of rectal bleeding [12]. Although the histopathological findings are not pathognomonic of the disease, Odze et al. suggest that the finding of ≥60 eosinophils/10 high-power fields is enough to establish the diagnosis in 97.4% of the cases [13].
The differential diagnosis includes anal fissure, which is the most common cause of rectal bleeding in children under 12 months; swallowed maternal blood, or other diseases, such as enteric infection, malrotation with volvulus or Hirschsprung disease—but these usually affect the general condition [8].
FPIAP can be an over-diagnosed entity, which is why some authors advocate the performance of an early oral food challenge (OFC) to avoid unnecessary elimination diets [14].
The prognosis is excellent and nearly all infants outgrow the disease by 1 to 3 years of age [15].

1.2. Food Protein-Induced Enteropathy (FPE)

Food protein-induced enteropathy (FPE) is also a non-IgE-GI-FA. It is probably the most forgotten disorder of the three, and in recent years there are hardly reports of FPE in the literature. The symptoms begin in the first year of life and manifest as chronic diarrhea and intermittent vomiting. If the symptoms persist, they will lead to poor weight gain, failure to thrive, and villous atrophy. Some children develop an iron-deficiency anemia associated with the presence of occult blood in the stools [16]. The pathophysiologic mechanisms are unknown, as is the real prevalence.
The most common offending triggers are CM proteins and soy. The diagnosis is based on a compatible clinical history, the anatomopathological findings, and the disappearance of the symptoms once the food is removed from the diet. Intestinal lesions are patchy with villus atrophy, crypt hyperplasia, increased number of intraepithelial lymphocytes, and eosinophilic infiltration [12].
The main differential diagnosis is with celiac disease. Clinically and pathologically, they are indistinguishable, but the typical celiac disease antibodies are absent in FPE [17].

1.3. Food Protein-Induced Enterocolitis Syndrome (FPIES)

FPIES has been considered a rare disease, whose cumulative incidence rates are between 0.15–0.7% [18]. However, it might be under-diagnosed since the epidemiological data are scarce and variable.
FPIES is a non-IgE-GI-FA and is considered a heterogenous disorder with a wide spectrum of clinical phenotypes. These are based on: (i) time and duration of the symptoms, (ii) age of presentation, and (iii) specific IgE-positive of the causal food. Thus, we can classify FPIES into acute, chronic, later onset, and atypical [19].
Acute FPIES usually presents in infancy, with repetitive protracted vomiting that begins 1 to 4 h after food ingestion. Vomiting is often accompanied by lethargy and pallor and can be followed by diarrhea. Some patients have an isolated fever peak and severe cases can progress to hypothermia, methemoglobinemia, acidemia, and hypotension, mimicking sepsis [20]. Children are well between episodes and have normal growth, and these typical symptoms commonly disappear within 24 h.
Chronic FPIES is less well characterized. In this presentation, the symptoms occur over several days or weeks and do not have a clear relationship with the offending food. Children have intermittent vomiting and/or diarrhea, failure to thrive, or poor weight gain. The severe form can develop dehydration and hypovolemic shock. In chronic FPIES, once the offending food is removed from the diet, the symptoms subside within two weeks, but re-exposure to it triggers the typical symptoms of acute FPIES. When the disease onsets in adolescence or adulthood, the symptoms appear after years of tolerance to the food and the predominant symptoms are, in most cases, colic abdominal pain and diarrhea [19][20][21][22][19,20,21,22].
Although FPIES is considered a non-IgE-mediated FA, in some patients, it is possible to detect specific IgE against the causal food, either at the beginning or during the course of the disease. This form of presentation is known as atypical FPIES and it seems to have a protracted course, with a further risk of developing IgE-mediated symptoms. In adults, it has been also described that severe FPIES phenotypes developed IgE and this is linked to the worst prognosis of the disease [19][21][23][24][25][26][19,21,23,24,25,26].
Any food can trigger FPIES; however, there are geographical differences, and some foods are more prevalent than others. In the USA, the most causative foods are CM proteins and soy. Children with either CM- or soy-induced FPIES can also react to both foods and have an increased likelihood of reacting to a solid food, most commonly oats and rice. In Asia, CM, soy-based formulas, and rice are the most common causal foods. Soy is especially frequent in Korea, but it has a better prognosis and earlier resolution than in other countries. Rice has been observed to be the main trigger of FPIES among the Australian pediatric population, and in European countries, CM and fish are the most frequent triggers of FPIES [27][28][29][27,28,29].
The development of tolerance varies by type of trigger and geographic location. CM- and soy-induced FPIES are reported to occur at an earlier age than other foods. The average age of tolerance of grains-FPIES is 35 months and fish-FPIES usually resolves within the first 5 years [27][28][27,28].
Nowadays, there is not a specific biomarker for FPIES that allows for confirmation of the diagnosis or to predict when the tolerance to the offending food has been achieved. For this reason, all patients with FPIES should undergo an oral food challenge (OFC) at least once in life. The OFC is a risky procedure and patients can develop moderate or severe symptoms after it, so it should be always performed under medical supervision [20][30][20,30].
The diagram shown in Figure 1 summarizes similarities and differences of FPIAP, FPE, and FPIES.
Figure 1. Description of differences and similarities between FPIES, FPE, and FPIAP. FPIES: food protein-induced enterocolitis syndrome; FPE: food protein-induced enteropathy; FPIAP: food protein-induced allergic proctocolitis. * Laboratory findings (e.g., anemia, iron deficiency, eosinophilia, rectal bleeding, etc.) are similar as long as the food is not withdrawn from the diet.

2. Approaches for the Study of Non-IgE-GI-FA

The most important diagnostic tool in the evaluation of non-IgE-GI-FA is a careful clinical history. However, when the clinical history is not sufficiently thorough, the OFC should be used as the gold standard to confirm the diagnosis of this disease, as has been mentioned above [21][31][32][21,31,32].
There are no other specific diagnostic procedures for these diseases, mainly because their pathophysiological mechanisms are poorly understood. However, there are a variety of laboratory tests to support the diagnosis and, more importantly, to rule out other conditions such as lactose intolerance, sepsis, necrotizing enterocolitis, anaphylaxis, gastrointestinal reflux disease, and celiac disease, among others [20]. In this sense, animal models are also playing an important role in the understanding of the pathophysiological mechanisms involved in these pathologies.
For the in-depth study of non-IgE-GI-FA, feces are the sample of choice. This is due to the fact that the main system affected is the gastrointestinal tract and to their non-invasive collection [33]. There are fecal biomarkers, such as immune cell derived proteins and/or gut microbiota metabolites, defined as biological entities capable of predicting an individual’s physiological state or pathological condition. The detection of these biomarkers by analytical techniques is on the rise to learn more about the underlying mechanisms of these pathologies [16][34][16,34]. In this review, different emerging approaches, such as animal models, omics techniques, and identification of fecal biomarkers are described to deepen the understanding of non-IgE-GI-FA (Figure 2).
Figure 2. Approaches for the study of non-IgE-mediated gastrointestinal food allergy. SCFAs: short-chain fatty acids.

2.1. Animal Models

Animal models are a useful tool to study allergic diseases in vivo. Several mouse models have been developed to study the gastrointestinal reactions related to food allergy [35][36][37][38][39][40][41][42][43][35,36,37,38,39,40,41,42,43]. Unfortunately, while several IgE-mediated food allergy models have been developed [35][41][42][43][44][45][35,41,42,43,44,45], the generation of non-IgE-mediated murine models has been harder to accomplish. For example, to our knowledge, no animal models have been successfully developed for FPIES, FPIAP, and FPE. However, Burggraf et al. generated a mouse model with intestinal symptoms of a food allergy which resemble these pathologies to study their pathomechanism. This model is known as allergic enteritis (AE), and it mimics in mice some traits observed in patients with non-IgE-GI-FA, such as intestinal inflammation characterized by thicker basal layer, crypt elongation, villous atrophy, and granulocyte infiltration [37]. The initial stimulus leading to the development of allergic enteritis remains to be elucidated, but it is postulated that delayed Th2 cell mediated responses are involved in the development of the inflammation [21][23][24][25][26][21,23,24,25,26]. In this model, mice are intraperitoneally sensitized with ovalbumin (OVA) and alum as adjuvant and then challenged with a diet containing high amounts of egg white (EW-diet) [37]. After 4 weeks, tolerance is achieved, but chronic inflammation is still present, despite the lack of allergic signs and the OVA-specific IgE reduction [37]. In a recent study, this AE murine model was used to elucidate the molecular and cellular mechanisms underlying AE [46]. A CC chemokine receptor 8 (CCR8) knock-out mouse was used. CCR8 deficiency translated in a reduced inflammatory profile, characterized by a reduction in the number of infiltrating eosinophils and an enhanced number of neutrophils in the lamina propria of the CCR8KO mice [46]. This revealed that the chemokine CCL1 and its receptor CCR8 are essential for the recruitment of eosinophils in AE sites [46].

2.2. Gut Microbiota and Omics Techniques

The human microbiota and its genetic composition—known as the human microbiome—began to be studied on a large scale at the beginning of the 21st century within the framework of major projects such as the Human Microbiome Project [47][52]. Among the microbial ecosystems of the human body, the gut microbiota, present in the gastrointestinal tract, is considered a highly important “superorganism” due to its diversity and complexity [48][49][53,54].
Approximately 70% of the primary colonization of the gut microbiota of newborns is vertically transmitted, occurring directly from mother to offspring. However, the composition of the gut microbiota is dynamic and changes throughout life due to factors such as delivery mode, age, diet, antibiotic consumption, ethnicity, lifestyle, environmental factors, and diseases and disorders, among others [33][50][51][52][33,55,56,57].
The development of the gut microbiota during the first years of life correlates with the development and maturation of the immune system. Moreover, the gut microbiota and the production of gut microbiota-derived short-chain fatty acids (SCFAs: acetic, propionic, and butyric acids) have been shown to be crucial for the proper expression and function of regulatory T cells (Treg), which are important in the induction and maintenance of oral tolerance to food antigens in humans [53][54][55][56][57][58][59][60][58,59,60,61,62,63,64,65]. Trompette et al. described in 2014 how high-fiber diet intake in mice offered protection against allergic inflammation in the lungs due to an increased level of circulating SCFAs compared to normal diet-fed mice [61][66]. Bunyavanich et al. studied the association between CM allergy resolution and early-life gut microbiota, showing an association between gut microbiome at early stages of life (from 3 to six months of age) and milk allergy resolution by age 8 years. This fact was due to the presence of bacterial taxa Clostridia and Firmicutes, suggesting them as potential probiotics candidates for milk allergy treatment [56][61]. For the analysis and understanding of the gut microbiota, omics are the techniques of election. In the last decades, advances in bioinformatics, as well as the development of high-throughput technologies, have contributed to the development of omics sciences for the study of multifactorial diseases, including allergy [62][67]. This group of sciences encompasses different disciplines such as genomics (the genome and its function), transcriptomics (mRNA and gene expression), proteomics (structure, function, location, and interaction of proteins), and metabolomics (metabolites and metabolism) (Figure 3) [63][68].
Figure 3. Main omics techniques: genomics, transcriptomics, proteomics, and metabolomics.
The trend in publications of studies analyzing the intestinal microbiota through omics techniques in patients with non-IgE-GI-FA is growing. For this reason, this field is likely to evolve very rapidly in the coming years. Table 1 lists all the studies to date and summarizes what has been found in terms of genomics and metabolomics.
Table 1. Summary of described studies using genomics and metabolomics to study non-IgE-GI-FA.
Disease Description Diet of Allergic Infants Microbial Communities

(Genomics)
Metabolites

(Metabolomics)
Reference
Non-IgE-mediated CMA Allergic vs. healthy children No dietary restriction (collection at diagnosis) ↑ Phylum Bacteroidetes

↑ Genera BacteroidesAlistipes, and Sarcina
↓ Fecal butyrate Berni Canani et al. [58][63]
Non-IgE-mediated CMA Allergic vs. healthy infants (unrestricted conventional diet) Milk exclusion diet for six months:

12 EHF
2 SPBF
3 HRF
EHF, SPBF & HRF:

↑ Families Lachnospiraceae and Coriobacteriaceae

↓ Genus Bacteroides

HRF:

↓ Genus Bifidobacteria
↑ Isobutyric and isovaleric acids = Acetic, propionic, and butyric acids Díaz et al. [64][72]
Non-IgE-mediated CMA Allergic with test formula vs. allergic with control formula vs. HBI at week 8 Amino acid-based formula for 8 weeks:

35 test formula (synbiotic-containing)
36 control formula
Bifidobacteria:

HBI > test formula > control formula

Eubacterium rectale/Clostridium coccoides:

HBI < test formula < control formula
- Candy et al. [65][73]
Non-IgE-mediated CMA Allergic with test formula vs. allergic with control formula at week 26 Formula for 26 weeks:

26 test formula (synbiotic-containing AAF)
28 control formula (AAF)
2 control formula (CMF)
Bifidobacteria:

test formula > control formula

Eubacterium rectale/Clostridium coccoides:

test formula < control formula
- Fox et al. [66][74]
Non-IgE-mediated CMA Allergic with test formula vs. allergic with control formula vs. HBI Amino acid-based formula:

35 test formula (synbiotic-containing)
36 control formula
Test vs. control:

↑ Species of Bifidobacterium and Veillonella families

↓ Species of Lachnospiraceae family and Ruminococcus and Alistipes genera

HBI vs. test formula:

≈ Bifidobacterium spp. and Lachnospiraceae spp.
Test vs. control:

↑ L-lactate

↓ Valeric and isobutyric acids = Acetate, propionate, butyrate, iso-valerate, and

D-lactate
Wopereis et al. [67][75]
Non-IgE-mediated CMA Allergic children before and after milk challenges Hypoallergenic hydrolyzed formulas ↑ Lactic acid bacteria (genus Lactobacillus)

↓ Bacterial groups of Lachnospiraceae and Ruminococcaceae families
↑ Skatole

↓ p-cresol and BCFAs = Acetic, propionic, and butyric acids
Guadamuro et al. [68][76]
Gastrointestinal disorders Colic vs. non-IgE-mediated CMA vs. proctocolitis vs. healthy controls Exclusive breastfeeding Fecal microbiota in non-IgE-mediated CMA:

↑ Families EggerthellaceaeLachnospiraceae, and Peptostreptococcaceae

↑ Genera Intestinibacter (more increased in proctocolitis) and Rothia↓ Genera Erysipelatoclostridium (higher in controls) and Bifidobacterium

Breast milk:

↑ Family Eggerthellaceae in non-IgE-mediated CMA

↑ Family Prolixibacteraceae in proctocolitis
- Aparicio et al. [69][77]
Footnote. CMA: cow’s milk allergy; EHF: extensively hydrolyzed formula; SPBF: soy protein-based formula; HRF: hydrolyzed rice formula; HBI: healthy breastfed infants; AAF: amino acid-based formula; CMF: cow’s milk formula; BCFAs: branched-chain fatty acids. Arrow ↑ stands for increased levels of microbial communities and/or metabolites. Arrow ↓ stands for decreased levels of microbial communities and/or metabolites.

2.3. Fecal Biomarkers. Protein Analysis

The attempts to evaluate the use of fecal biomarkers for the diagnosis of non-IgE FAs have been reviewed recently [16]. The main candidates are fecal calprotectin (FC), α-1 antitrypsin, β-defensin, tumor necrosis factor-α (TNF-α), fecal IgA, eosinophil-derived neurotoxin (EDN), and eosinophilic cationic protein (ECP). From these, FC has been the most discussed, and it has been the subject of a recent systematic review for its use as a biomarker for CMA, including IgE- and non-IgE-mediated forms [70][78].
Calprotectin is an immunomodulatory, antimicrobial, and antiproliferative protein that is present in the cytoplasm of neutrophils, in the membranes of macrophages, in activated monocytes, and in mucosal epithelial cells [71][79]. The measurement of FC has been evaluated as a non-invasive marker of gastrointestinal inflammation, as it is known that its concentrations are correlated with the level of intestinal mucosal inflammation [72][80]. Based on this fact, it has been used for the monitoring of intestinal conditions such as inflammatory bowel disease (IBD) [16][73][16,81] and distinguishing it from irritable bowel syndrome (IBS) [74][82]. As for its measurement for CMA, a remarkable example is the study by Beşer and collaborators, which analyzed children with CMA (24 IgE-mediated, 8 non-IgE-mediated) and compared them to 39 controls of the same age range, before and after an elimination diet [75][83]. These authors found that FC levels were highest in children with non-IgE-mediated CMA, from which they suggest its use as a biomarker. However, the evidence is conflicting, since other authors found no significant differences [64][76][72,84]. Systematic reviews have concluded that the available evidence is not enough to encourage its use as a biomarker for monitoring CMA, neither IgE- nor non-IgE-mediated [16][70][16,78]. Altogether, authors agree that more well-defined studies are needed, and stress the need for establishing the exact cut-off values of FC in different groups taking into account factors such as age, feeding patterns, clinical symptoms, and immune mechanism type [70][78].
On the other hand, α-1 antitrypsin is an inhibitor of serine proteases present in serum which is synthesized in the liver [77][85]. It can be measured to assess the excessive GI protein losses in patients with protein-losing enteropathy (PLE) [78][79][86,87]. It should be noted that PLE must not be confused with α-1 antitrypsin deficiency, a separate condition, as has sometimes been reported in the literature [79][87]. In the case of CMA, there is very little data: to the best of our knowledge, only one paper reports its measurement in a study of CMA, alongside a panel of other markers, and did not find significant differences for this protein [80][88].