Food allergy: Mechanisms, diagnosis, and management in children

PEDIATRIC GASTROENTEROLOGY AND NUTRITION

0031–3955/02 $15.00
.00

FOOD ALLERGY Mechanisms, Diagnosis, and Management in Children Jonathan M. Spergel, MD, PhD, and Nicholas A. Pawlowski, MD

People eat approximately 2 to 3 tons of food in their lifetimes. Most people do not have an adverse reaction to foods, however. Many people—approximately 25% of the United States population—believe that they have an allergic reaction to foods; however, the actual incidence confirmed by history and challenges suggests a prevalence rate closer to 2% to 8% in young infants and less than 2% in adults. The most common food allergies in the United States are milk, egg, peanut, soy, wheat, tree nuts, fish, and shellfish. Milk is the most common food allergy, with approximately 2% of infants having food intolerance or allergy compared with 1.3% for egg and 0.5% for peanut and lesser rates for the other foods. Reactions to foods are not new and have been described for 2000 years. The ancient Greek physician Hippocrates describes a reaction to milk in the first century. Anaphylactic reactions to egg and fish have been described as early as the sixteenth and seventeenth centuries.45 An adverse food reaction is any abnormal response to an ingested food, regardless of the pathophysiology. One classification scheme, based on the universe of possible mechanisms, separates immunologic from nonimmunologic entities.

From the Allergy Section, Division of Immunologic and Infectious Diseases, The Children’s Hospital of Philadelphia; and the Department of Pediatrics, The University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania

PEDIATRIC CLINICS OF NORTH AMERICA VOLUME 49 • NUMBER 1 • FEBRUARY 2002

73

74

SPERGEL & PAWLOWSKI

Non–Immune-Mediated Adverse Reactions Metabolic Disaccharidase deficiency: lactase deficiency Favism: glucose-6-phosphate dehydrogenase deficiency Pancreatic insufficiency: cystic fibrosis Galactosemia Phenylketonuria Pharmacologic Caffeine Histamine Tyramine Toxic Flavorings and preservatives: sodium metabisulfite Dyes: tartrazine Bacterial and fungal toxins: C. botulinum, aflatoxin Seafood toxins: scromboid (tuna, mackerel) Contaminants: heavy metals, pesticides Infectious Parasitic: Giardia sp Bacterial: Salmonella sp Viral: hepatitis Data from Sampson H: Food allergy: II. Diagnosis and management. J Allergy Clin Immunol 103:981–989, 1999.

Immune-Mediated Adverse Reactions Involving the Gastrointestinal (GI) Tract Immunoglobulin E (IgE) mediated Oral allergy syndrome Immediate GI hypersensitivity Non–IgE mediated Food protein–induced enterocolitis, proctitis, enteropathy Celiac disease Mixed immune mechanisms (IgE and T-cell mechanism) Eosinophilic esophagitis (EE), gastroenteritis, gastritis

FOOD ALLERGY

75

Other Food–Induced Immune Disorders Cutaneous Acute urticaria and angioedema Contact urticaria Chronic urticaria and angioedema Atopic dermatitis (AD) Dermatitis herpetiformis Respiratory Allergic rhinitis Asthma Food-induced pulmonary hemosiderosis Anaphylaxis Food triggered Exercise-induced

Immune reactions are subclassified into IgE- and non–IgE-mediated reactions. IgE-mediated reactions are the classic presentations of food allergy, such as hives or anaphylaxis after eating the offending food antigen. Non–IgE-mediated food reactions have been described in the past several years and include food protein–induced enterocolitis syndrome and allergic EE. The principal focus of this article is food hypersensitivity disorders affecting the GI tract.

MECHANISMS OF ADVERSE FOOD REACTIONS Non–Immune-Mediated Reactions Food may cause an adverse reaction because of the toxic nature of the food particle. The classic example is food poisoning from botulism, in which the toxin interacts with cyclic adenosine monophosphate regulatory protein. The most common cause of acute food-borne disease of chemical etiology is ciguatera poisoning. Ciguatera is caused by tropical fish by eating dinoflagellate algae, causing a variety of symptoms that are primarily gastrointestinal and neurologic.10 The most common adverse food reaction is lactose intolerance, caused by the deficiency of ␤-lactase. The lack of the enzyme leads to increased lactose into the colon, where the bacteria convert it to carbon dioxide, hydrogen, and water. Symptoms from the excess gas production include abdominal cramps, flatulence, and diarrhea. Lactose intolerance affects 6% of the US population but affects up to 90% of some ethnic groups.40 Other metabolic disorders include pancreatic insufficiency and favism, caused by glucose-6-phosphate dehydrogenase deficiency (see previous box titled ‘‘Non–Immune-Mediated Adverse Reactions’’).

76

SPERGEL & PAWLOWSKI

Immune-Mediated Reactions For particles of food proteins to create an immune reaction, they must be absorbed, processed, and presented to an active immune system leading to an adverse reaction. The food protein first is processed in the mouth by salivary amylases and mastication and then is processed further down the GI tract by gastric, pancreatic, and intestinal enzymes. After this processing, the food antigen penetrates the intestinal barrier, which consists of epithelial cells, glycocalyx, microvilli, tight junctions, and intestinal peristalsis. Food enters the systemic circulation by one of several mechanisms, including endocytosis or passage through the tight junctions.40 In addition, soluble IgA can block penetration of foods, and food antigen-specific IgA and IgG can clear the food antigen.42 Therefore, only 2% of ingested food antigens are absorbed and transported throughout the body in an ‘‘immunologically’’ intact form or recognizable antigen. Food antigens then can cause three types of immune responses: (1) IgE-mediated reaction, (2) non–IgE-mediated reaction, and (3) tolerance (i.e., no reaction).

IgE-Mediated Reactions The development of an IgE-mediated response to an allergen is the result of a series of interactions involving antigen-presenting cells (APCs), T cells, and B cells. The APCs present the antigen as small peptide fragments in conjunction with major histocompatibility complex (MHC) class II molecules to T cells. T cells bearing the appropriate complementary T-cell receptor bind to the peptide–MHC complex on T cells, leading to T-cell proliferation and cytokine generation. Interleukin 4, a T-helper type 2 (Th-2) cytokine, and interactions with costimulatory molecules, such as CD40–CD40 ligand, cause isotype switching and the production of antigen-specific IgE.3 Once produced, antigen-specific IgE circulates by the bloodstream to the relevant mucosa and tissue sites and binds to Fc⑀R-bearing cells. Food antigen then can bind to the specific IgE on the surface of mast cells, basophils, eosinophils, and macrophages, causing intracellular signaling and the production and release of histamine, prostaglandins, leukotrienes, and cytokines. These molecules create the symptoms of urticaria, wheezing, rhinorrhea, and anaphylaxis that can occur with IgE-mediated reaction. For this IgEmediated reaction to occur, the body must have been exposed to antigen previously. The first exposure leads to the production of the antigenspecific IgE and priming of the system. Interestingly, the first exposure can be in utero or through breast milk.16, 54 Subsequent exposure in the sensitized host can lead to an immediate hypersensitivity reaction in varying target organs.

FOOD ALLERGY

77

Non–IgE-Mediated (T Cell–Mediated) Reactions The biology and documentation of food-specific T cell–mediated reactions are currently not as well understood. The strongest evidence comes from the identification of food-specific T cells in AD. Food-specific cutaneous lymphocyte antigen (CLA
)-T cells have been identified in the lesions of milk-allergic patients.1 These patients have an AD flare when challenged by milk, indicating the clinical significance of these reactive T cells. Another example is food patch testing in AD. Patch tests reflect T cell–mediated reactions and correlate with relevant food as confirmed by food challenges31, 32; however, these results could not be extended to in vitro studies because T cells from allergic and nonallergic individuals can proliferate to food antigens.20 The other clinical indications that T cell–mediated food allergies exist arise from clinical experience. Patients have reactions to foods and are skin test–negative, for example, food-protein enterocolitis; therefore, another mechanism must exist. Tolerance or resolution of food allergies also involves T cells (detailed later). Oral Tolerance A host is tolerant when food protein can enter the systemic circulation or contact any immune reactive cell and not trigger an immune reaction. Three mechanisms have been associated with immune tolerance: (1) clonal anergy, (2) clonal deletion, and (3) active suppression. T and B immune responses may be involved in the development of oral tolerance. Detailed reviews of oral tolerance have been published (see reference 53). In simplified terms, the T cells can be deleted actively or passively. In the passive model, T cells are no longer stimulated by the respective food antigen. This lack of stimulation leads to decreased growth and eventually cell death. In the active model, clonal deletion, T cell death occurs after interaction with self-antigens and self-MHC cells that are developing within the thymus.33 These mechanisms of tolerance induction have begun to be elucidated in mice; the existence of tolerance and its regulation in humans are less well understood. The third mechanism for tolerance, active suppression, may be related to the mechanism of antigen presentation in the gut. M cells (i.e., specialized epithelial cells overlying the Peyer’s patches) and intestinal epithelial cells (IECs) are the major sites of immune antigen sampling in the intestine.35 Recent studies, however, indicate that IECs may be the key APCs used for generating immunosuppression in the gut. These ‘‘nonprofessional’’ APCs can take up and process soluble antigen from the apical end and present these antigens to unique regulatory T cells. IECs have unique surface molecules and include the expression of the nonclassic HLA class Ib molecule CD1d and a novel CD8 ligand, gp180. These molecules come together to activate a subpopulation of CD8
regulatory cells the function of which is to suppress immune responses.7

78

SPERGEL & PAWLOWSKI

The most recent theories suggest that soluble antigens are presented primarily by IECs, whereas particulate antigens and intact bacteria, viruses, and parasites are sampled by M cells, leading to active immunity and generation of IgA.17 Infants may have a higher incidence of food allergies because of how food is processed and presented to the immune system. Infants have immature systems, with decreased stomach pH,21 decreased enzyme activity,19 and a low concentration of s-IgA. These factors may lead to increased level of food antigen in the systemic circulation and to possible food allergy. As the intestinal tract matures, the T cells may be exposed to less antigen, leading to less stimulation of food-specific T cells. A similar process may occur as a person develops tolerance to food over time; as a patient eliminates specific dietary antigens, T cells are no longer stimulated and clonal deletion occurs. For the foods that have a low rate of tolerance development, such as shellfish, the food allergen (tropomysin of shellfish) may cross-react with allergens (dust mite-tropomysin) in the environment, leading to repeated stimulation of the T-cell clones and continuing food allergy.

CLINICAL DISORDERS Immune-Mediated Disorders of the Gastrointestinal Tract The GI tract confronts the largest burden of external antigens faced by any organ system. Multiple disorders can be identified based on clinical patterns of presentation (e.g., time of onset, type of symptom, persistence, and severity) and immune mechanism (see previous box titled ‘‘Immune-Mediated Adverse Reactions Involving the Gastrointestinal Tract’’).

IgE-Mediated Gastrointestinal Disorders Oral Allergy Syndrome The oral allergy syndrome is more prevalent in adults than in children. Symptoms are restricted essentially to the oropharynx and include rapid onset of itch or discomfort (e.g., ‘‘tingling’’). Angioedema also may occur, and when accompanied by the sensation of throat tightness, it may be confused with a more severe systemic reaction. Symptoms are usually brief and triggered by various fruits and vegetables. Typical groupings of causative foods suggest IgE cross-reactivity between selected pollens (e.g., ragweed or birch) and foods (e.g., melons and bananas with ragweed; apples, hazelnuts, celery, carrots with birch; cherry, apricot, plum, and peach with brazil nuts).15, 45

FOOD ALLERGY

79

Immediate Hypersensitivity Reactions Isolated immediate hypersensitivity reactions in the GI tract induced by food proteins can occur at any time in life. As expected from the mechanism, they usually develop within minutes of ingestion or as long as 2 hours later. Rapid onset within this time frame is an important diagnostic clue for the presence of an IgE-mediated or true immediate hypersensitivity process. Symptoms include vomiting, nausea, colic, abdominal pain, and diarrhea. Vomiting is the most explosive presentation, but this may be reduced in patients with long-standing exposure, possibly secondary to partial desensitization. In this case, the clinical picture may include only intermittent vomiting or none but instead poor appetite, poor weight gain, and intermittent abdominal pain or colic.59 Detailed evaluation often reveals decreased processing of complex carbohydrates. IgE-mediated reactions involving the GI tract may accompany reactions in other target organs, such as the lungs and skin, during systemic reactions. These reactions are to be differentiated from the oral allergy syndrome, which represents a local reaction without the risk for systemic involvement in its isolated presentation. Constipation and IgE Sensitization to Cow’s Milk A distinctive syndrome consisting of constipation and IgE sensitization to cow’s milk has been described.22 After careful evaluation, most patients had symptoms and endoscopic evidence of proctitis. Constipation was relieved by milk protein elimination.

Non–IgE-mediated Gastrointestinal Disorders Food Protein–Induced Enterocolitis Syndrome The symptoms of food protein–induced enterocolitis syndrome (FPIES) include vomiting and diarrhea and can progress into a severe shocklike state.13, 47 Most patients with FPIES present in the first months of life, and the disorder typically resolves by 2 years of age but can persist in rare cases into later childhood. Compared with the immediate reaction seen with IgE-mediated reactions, the onset of symptoms in FPIES is delayed from 1 to 10 hours after the ingestion of food. Symptoms typically start with emesis that often is followed by diarrhea.47 In the authors’ experience, vomiting usually begins within approximately 2 to 4 hours, a time frame that distinctly contrasts with IgE-dependent, classic immediate hypersensitivity reactions. Similar to IgE-mediated reactions, cow’s milk and soy proteins are the most common antigens responsible for FPIES in infants.48 Other foods, however, including egg, wheat, rice, oat, and peas, have been reported to trigger symptoms in older children. The only reported laboratory findings that are observed in FPIES

80

SPERGEL & PAWLOWSKI

are an increase in peripheral blood neutrophil counts during a positive challenge and an alteration in tumor necrosis factor ␣ levels in the feces and secretion of peripheral blood mononuclear cells.4, 13 Because of the lack of positive skin tests and limited laboratory findings, the pathophysiology of FPIES is incompletely understood.48 Because most of these patients have negative skin tests for the offending antigen, FPIES is thought to be T cell–mediated disease. Evaluation for T-cell function has shown that antigen-specific T cells proliferate to milk and soy in patients with milk and soy-induced FPIES,56 but this response also can be seen in healthy individuals. Therefore, no specific diagnostic testing for FPIES currently exists, and diagnosis rests on clinical history. Food Protein–Induced Proctitis Food protein–induced proctitis is found early in infancy, presenting with blood-streaked stools in generally well-looking patients. Blood loss is rarely severe, and bowel lesions (eosinophil and neutrophil infiltration) are limited to the distal large bowel. Most common triggers are milk and soy formulas, but this syndrome also occurs in breastfed infants.28 Food Protein–Induced Enteropathy Food protein–induced enteropathy is another disorder of infancy, presenting early with diarrhea and poor weight gain. Vomiting is common, together with malabsorption. Intestinal biopsy reveals villous atrophy and cellular infiltration likely responsible for the poor absorption of nutrients and protein loss leading to edema in some patients.58 Anemia is less common. Milk sensitivity is the most common trigger, but soy, egg, wheat, and other foods can be associated. A separate disorder is the association of occult blood loss leading to iron deficiency anemia in cow’s milk–fed infants. GI symptoms are not evident. A non-IgE mechanism is postulated here, and the blood loss is reversed quickly by the elimination of milk protein. Most patients become tolerant to milk by 3 years of age. Celiac Disease Celiac disease is a specific food protein–induced enteropathy in which patients react to gliadin, the alcohol-soluble portion of gluten found in wheat, oat, rye, and barley.9 Diagnosis is made by documentation of typical abnormality (villous atrophy and cellular infiltrate), which is reversed by the elimination of gliadin from the diet. Most patients produce IgA antigliadin and antiendomysial antibodies. GI symptoms include weight loss and chronic diarrhea, steatorrhea, and associated abdominal distension. Extraintestinal features include oral ulcers.41

FOOD ALLERGY

81

Combined IgE- and T Cell–Mediated Gastrointestinal Disorders Eosinophilic Gastroenteritis Eosinophilic gastroenteritis has been found in all age groups, with the predominance in early to midadulthood (reviewed by Kelly26). Eosinophilic gastroenteritis presents with a variety of symptoms, including abdominal pain, diarrhea, melena, weight loss, and others; however, many other entities can present with eosinophils in the GI tract, including gastroesophageal reflux disease (GERD)5, parasitic infection, or inflammatory bowel disease. The diagnosis of eosinophilic gastroenteritis can be made only if other pathologic disorders have been eliminated. In this disorder, T-cell and IgE mechanisms may exist. Eosinophilic Esophagitis Eosinophilic esophagitis, primary EE, or idiopathic EE occurs in adults and children and may represent a disease subset or variant of eosinophilic gastroenteritis.38 Patients with EE present with symptoms similar to those of GERD but are unresponsive to antireflux medication and have normal pH probe studies. Vomiting and abdominal pain are the most common symptoms,6 but other common symptoms include anemia (occult blood loss), weight loss,25 achalasia,29 and failure to thrive. An early link between EE and allergies was reported in 1977,12 with a case of a 51-year-old man with asthma and allergies who developed dysphagia and substernal chest pain. Esophageal biopsies revealed a focal, marked eosinophilic infiltrate. Another case report34 associating EE with food allergy involved an adolescent with dysphagia, food allergy, and a peripheral eosinophilia and eosinophilic infiltration of the esophagus. In 1993, Attwood et al2 published one of the first studies comparing patients with isolated EE with patients with GERD. They described 12 patients presenting with dysphagia who had more than 20 esophageal eosinophils per high-powered field (HPF; mean, 56 cells/HPF). All had visually normal esophageal mucosa and no esophageal anatomic abnormality, 11 had normal pH monitoring, and 7 had evidence of an allergic disorder. The investigators compared this group with 90 patients with medically responsive GERD (documented by 24-hour pH probe) and reported that only 43 of these patients had esophageal eosinophils, and to a much lesser degree (mean, 3.3 cells/HPF). He suggested that these patients represented a new clinicopathologic syndrome not previously described, with increased eosinophils in the esophagus and normal pH. In 1995, Vitellas et al57 also reported on 13 male patients with idiopathic EE and found that most of these patients responded to corticosteroids, indicating that this entity is an inflammatory process. The patients’ clinical symptoms included dysphagia (12 of 13), allergic manifestations (10 of 13), peripheral eosinophilia (12 of 13), and proximal esophageal

82

SPERGEL & PAWLOWSKI

strictures (10 of 13). This study result was the first to link allergic tendencies in this population. Recently, two case series that examined eosinophilic esophagitis in children were published. Kelly et al27 examined 23 children with classic symptoms of GERD. Their symptoms did not improve on standard treatment for 6 to 78 months, and eosinophil counts in esophageal biopsies were elevated compared with standard GERD, with 15 to 100 eosinophils per HPF, which decreased to 5 to 30 eosinophils per HPF. Twelve children were treated with elimination diet with an enteral nutritional supplement (Neocate) or steroids. Foods were introduced into their diets at home as results were reported. Milk was identified in open-challenge as causing symptoms in 7 patients; soy, in 4 patients; wheat, in 2 patients; peanut, in 2 patients; and egg, in 1 patient. Liacouras et al30 examined 1809 patients prospectively for GERD. Twenty patients had eosinophils in their esophagus, normal pH probe studies, and poor response to standard GERD therapy. The patients who did not respond to standard therapy had significantly greater eosinophilia (34.2  9.6 eosinophils/HPF) compared with that in children with GERD who responded to medical therapy (2.26  1.16 eosinophils/HPF; P ⬍ 0.001). The nonresponders were treated with 1.5 mg/ kg/day oral methylprednisolone, divided into twice-daily doses for 4 weeks, with resolution of their symptoms. These 20 subjects had elevated IgE of 521 plus or minus 824 IU/mL, indicating an atopic predisposition, but no discussion of specific foods was given.30 Elevated IgE and response to steroids indicated that an inflammatory process was the cause of the disease. Another indication that allergic inflammation may be caused by food is that many similar patients respond to elemental diets. Additional evidence that food may have a role in the inflammatory process comes from study at The Children’s Hospital in Boston. Walsh et al examined 28 children with GERD with intraepithelial eosinophils after initial post-treatment for GERD. The patients were divided into three groups (A, B, and C). Groups A and B (n  11 and 10, respectively) had incomplete clinical response, with normal pH probe monitoring results in group A and no pH probes in group B. Group C (n  7) had clinical improvement with aggressive GERD therapy and abnormal pH probe monitoring characteristic of GERD. Groups A and B had significantly higher numbers of eosinophils in the distal esophagus compared with group C, consistent with a diagnosis of EE (A, 31 cells/HPF  19.5; B, 28 cells/HPF  23.7; versus C, 5 cells/HPF  6.7; P  0.009). Groups A and B had a history of allergy in 81% of the patients. Abnormal allergen skin-test results were detected in 8 of 10 group A and B patients tested.60 This study suggested that patients with eosinophilic esophagitis may have food allergies, but this study did not test the association of specific foods and esophageal inflammation by examining the clinical response to dietary elimination. Orenstein et al39 tried more definitively to link food allergies to EE. They studied 30 patients with EE retrospectively based on the diagnosis of elevated eosinophils (⬎ 5 cells/HPF) in biopsies. These patients had

FOOD ALLERGY

83

similar symptoms to other studies, including recurrent esophageal food impactions, vomiting, pain, and dysphagia. This population had a strong atopic background, with 62% of patients having history of allergy. Food allergy testing was done in 28 children by percutaneous skin testing or in vitro testing. Twelve children were negative by either method, and a significant discordance was found between these two methods for positive reactions. Fifteen of 16 patients with documented food allergies were given an elimination diet. Three patients were noncompliant, but 12 patients showed symptomatic improvement, with a poor response in only one child. Unfortunately, the specific trigger foods they identified were not discussed.39 This study confirmed that food allergies can play a significant role in EE for approximately 60% of patients in this group. The current authors also tried to identify the particular food allergies in EE. Patients with the diagnosis of EE based on the finding of more than 10 eosinophils per HPF and poor responses to standard GERD treatment were studied. Food allergies were examined by standard percutaneous skin testing for IgE-mediated reactions and by patch testing for delayed hypersensitivity reactions. Evidence for sensitization to at least one food was found in 73% of patients by skin testing and in 85% of patients by patch testing. The authors found that milk, egg, soy, and peanuts are the most common foods by skin testing and that wheat, soy, beef, and milk are the most common by patch testing. This ongoing study demonstrates that careful elimination of food triggers leads to significant improvement in many patients. This finding confirms the role of food-induced allergic inflammation in this disorder. Recent research in an animal model by Mishra et al gives another potential insight into the pathophysiology of this disorder. In their study, inhalation of Aspergillus fumigatus antigen led to marked levels of esophageal eosinophils, free eosinophil granules, and epithelial cell hyperplasia, while oral ingestion or intragastric administration elevated EE to a lesser degree.37 This study suggests that allergen inhalation may have a role in the disease in mice in addition to its ingestion. Inhalation of A. fumigatus in this model also led to asthmalike and allergic rhinitis– like symptoms.36, 55 Therefore, patients with EE may need all forms of atopy (i.e., asthma, food allergy, and allergic rhinitis) to be controlled to resolve their disease.

Other Food-Induced Immune Disorders Cutaneous Acute Urticaria or Angioedema. The skin is affected commonly in immune-mediated adverse food reactions, as demonstrated by the commonness of acute urticaria or angioedema. Here, the reactions are mediated by antigen-specific IgE triggered by absorbed food proteins delivered to the skin by the circulation. These often occur during severe

84

SPERGEL & PAWLOWSKI

systemic reactions, but cutaneous reactions can represent the sole systemic consequence of food allergen ingestion. Topical Urticarial Reactions. Isolated topical urticarial reactions also may occur. Here, the skin may react only from local contact without systemic absorption. Differentiation between cutaneous reactions caused by limited, local contact versus systemic absorption can prove difficult in infants who may ingest variable amounts of food or none but still cover their faces and other areas of the body with food. This type of clinical scenario obviously requires careful evaluation. Chronic Urticaria and Angioedema. In contrast to acute skin reactions, chronic urticaria and angioedema (duration, ⬎ 6 weeks) rarely are associated with food sensitivity, especially in children.43 Atopic Dermatitis. AD is a chronic inflammatory skin disorder. The role of food allergy in AD has been reviewed extensively.49 Like other allergic diseases, the prevalence of AD seems to be increasing, changing from 3% to 4% in the 1960s to 10% to 15% in the 1980s. Unlike many other diseases, however, AD has no single diagnostic feature or pathognomonic test. The major features include pruritus, typical morphology, and distribution of the lesions. The skin distribution varies with age. In infancy, the face and extensor surfaces of the arms and legs most commonly are affected. In older children and adults, a scaly and lichenified dermatitis on the flexor surfaces of the extremities, neck, and upper trunk is observed.51 IgE and T cells seem to play a central role in the process. Ninety percent of patients with AD have marked elevated total IgE and high levels of specific IgE. Removal of the allergens that react with the specific IgE can cause decreased symptoms.51 The interaction of allergen with IgE molecules on the surface of cutaneous mast cells causes direct activation of mast cells with consequent mediator liberation. The role of T cells is confirmed by skin biopsies showing T cell–infiltrated lesions and expression of CLA, a homing receptor for T lymphocytes to the skin. CLA interacts with E-selection expressed on activated vascular endothelium in affected areas. The Th-2 cells predominate in the acute lesion, whereas Th-1 and Th-2 cells are found in the chronic eczematous lesions.18 Several studies have elucidated the role of food allergen–specific T cells in the inflammatory process underlying AD. The best evidence is that food allergen–specific T cells have been cloned from active skin lesions and healthy skin of patients with AD.11 In addition, patients with milk-induced AD were studied and compared with control subjects with milk-induced GI reactions without AD and with nonatopic control subjects. Casein-reactive T cells from the children with milk-induced AD had a significantly higher expression of CLA than other antigen-specific T cells from the same patients or the control groups.1 Finally, patients with AD are positive to involved foods on patch testing.23 Taken together, these studies elucidating the role of allergic responses to food in the pathogenesis of AD indicate a mixed inflammatory response involving T cells and IgE-mediated reactions. Dermatitis Herpetiformis. Dermatitis herpetiformis is a non-IgE–

FOOD ALLERGY

85

mediated skin disorder that presents as a pruritic rash distributed over extensor surfaces and buttocks that may be mistaken for AD. It can be associated with celiac disease with sensitivity to gluten. Dermatopathology of the skin reveals IgA deposits in the dermoepidermal junctions, whereas GI lesions resemble celiac disease.41 Respiratory Nasal and Bronchial. Nasal and bronchial reactions to foods are uncommon unless multiple target organs (e.g., skin and GI tract) are involved. The rarity of isolated nasal reactions has been documented in challenge studies. The rate of respiratory reactions documented in foodchallenge studies ranges from 6% to 25% of various populations, with the highest incidence among children with AD and asthma.24, 43, 45 Food-Induced Pulmonary Hemosiderosis. The immune mechanisms underlying food-induced pulmonary hemosiderosis (Heiner ’s syndrome) are unknown. Most often associated with sensitivity to milk or egg, the syndrome is characterized by pulmonary infiltrates associated with hemosiderosis, GI blood loss, anemia, and failure to thrive. Anaphylaxis Syndromes Generalized Anaphylaxis. Food is a common trigger for generalized anaphylaxis, causing approximately 100 deaths per year in the US.61 Cardiovascular (i.e., hypotension and dysrhythmia) and respiratory compromise are related most directly to severe morbidity and mortality. The pathophysiology of anaphylaxis is not well understood but generally is believed to be consequent to massive mediator release. As with allergic inflammation in the nose, lungs, and skin, biphasic reactions have been reported to occur. Here, an initial or ‘‘immediate’’ reaction triggered by an IgE-mediated mechanism is followed hours later by a ‘‘late-phase’’ reaction caused by a second wave of inflammation. The delayed resurgence of possibly severe symptoms has obvious treatment implications.46 Anaphylaxis can be associated with exercise, and, in an unusual form, the ingestion of a specific food(s) 2 to 4 hours before exercise is a necessary condition for anaphylaxis to occur. This rare condition usually occurs in teenagers and younger adults who are generally atopic and some of whom have a history of food sensitivity in earlier years.8 Food triggers do not cause symptoms during rest, but they do elicit IgEmediated skin-test reactivity. DIAGNOSIS The reader should refer to Figures 1 to 3, which summarize diagnostic strategies for IgE-mediated, T cell–mediated (FPIES), and mixed immunologic (EE) disorders.

86

SPERGEL & PAWLOWSKI

Figure 1. Evaluation of IgE-mediated gastrointestinal disorders.

FOOD ALLERGY

87

Figure 2. Evaluation of non-IgE–mediated gastrointestinal disorders.

History The patient’s history can be a powerful diagnostic tool, especially if the patient and family are objective historians. The family’s own perceptions and knowledge often influence history, however. Food allergy clearly is suspected more often than it is found by accurate diagnostic procedures and is confirmed by challenges in fewer than 20% of cases. In general, the history can be more helpful in IgE-mediated disorders because these reactions occur so soon after food ingestion and because multiple target organs are affected. In the authors’ experience, if a parent can recall distinctly that a specific food has been ingested more than once without consequence, it rarely emerges later as a confirmed trigger. Thus, a systematic review of the patient’s diet is a highly useful first step. Important historical considerations include: 1. Is the reaction reproducible? That is, does it occur each time the food is ingested? If not, the food is an unlikely trigger. The dose may be crucial, but this is not a common issue. If high doses are required, the level of reactivity is generally low.

88

SPERGEL & PAWLOWSKI

Figure 3. Evaluation of mixed immune-mediated gastrointestinal disorders. Eosinophilic esophagitis.

FOOD ALLERGY

89

2. What is the time frame for the reaction? Immediate hypersensitivity reactions generally occur rapidly, often within minutes and virtually always within 2.5 hours. This time frame is confirmed in the authors’ experience with food challenge.50 Mixed and T cell–mediated reactions have a characteristically delayed onset. Therefore, patients with FPIES typically begin to have symptoms later than 1.5 hours after ingestion. An additional consideration for FPIES is that the youngest patients have more severe symptoms and more rapid onset. Thus, a 2-month-old patient may vomit within 1.5 hours after food ingestion and show signs of hypotension. After abstinence for a long period of time, the onset of symptoms becomes more delayed and less severe and potentially confusing. The patient may show no reaction on challenge, but after daily feedings for an additional 1 to 5 days, diarrhea, abdominal pain, or bleeding may reemerge. 3. Do the clinical manifestations fit the expected pathophysiology? The clinical history should evolve toward a picture consistent with one or more entities outlined in the previous boxes. As noted in the first box, not all reactions have an immunologic basis. In addition to metabolic, pharmacologic, toxic, and infectious causes, the clinician must include structural abnormalities of the GI tract (e.g., pyloric stenosis, hiatal hernia, or tracheoesophageal fistula), malignancy (e.g., obstruction or mediator release), and certainly psychologic reactions (e.g., feigned, disbelief, or other) in the differential diagnosis. Clues to IgE-mediated diagnoses include: Skin Urticaria or angioedema or both AD or eczema Generalized flush and pruritus (may precede eruptions of eczema) Nose Congestion, rhinorrhea, pruritus, sneeze Isolated nasal symptoms uncommon Larynx Wheeze, cough, chest tightness, shortness of breath Lower respiratory symptoms often are preceded by nasal symptoms Gastrointestinal Edema or pruritus of lips, tongue, mucosa (most specific) Vomiting Nausea Abdominal pain, cramping, colic Diarrhea (less common except in eosinophilic gastroenteritis) Cardiovascular Hypotension in anaphylaxis (may be present in severe FPIES) CNS Nonspecific symptoms associated with anaphylaxis

90

SPERGEL & PAWLOWSKI

As mentioned previously, T cell–predominant disorders usually include only GI symptoms. Mixed disorders, such as eosinophilic gastroenteritis, can offer clues for some of the pathophysiologic elements. The patient may show rapid onset of symptoms (e.g., abdominal discomfort or throat itch) after food ingestion, but the manifestations may be limited to the GI tract. These considerations, together with a careful review of the clinical symptoms, usually point to a limited set of diagnostic possibilities. Additional clinical history elements can be helpful. Timing of the first and last occurrences can reveal whether sensitivity is increasing or waning. These considerations, together with the quantity necessary to trigger a reaction, are helpful for planning diagnostic challenge procedures as well. Occasionally, the history is complicated by the fact that trace amounts of foods may occur in certain products. For example, milk may contaminate fruit juices packaged in plants that use the same apparatus to prepare milk products. In other situations, food may be added intentionally to products as flavor enhancers. In these instances, experience and careful, educated label reading are extremely important. Careful, educated label reading includes looking for the common use of ‘‘natural flavoring,’’ which can mean that milk, soy, or other flavorings have been added. In addition, a family may not recognize that ‘‘caseinate’’ signals the presence of milk proteins. Finally, because recall can be imperfect, diet diaries may be used to clarify the history. This approach provides the necessary detail about approximately what foods are consumed, and temporal relationships emerge. Finally, families can help document that, on certain occasions, the suspected food is tolerated, narrowing the list of possible foods. Physical Examination A complete physical examination is indicated. Special attention should be directed toward uncovering allergic stigmata in any target organ, the patient’s general nutritional status, and the markers of any other chronic disease. The examination, together with the general clinical history, helps to reveal concurrent disorders that may influence the clinical picture. For example, patients with asthma and environmental risk factors causing persistent symptoms offer additional historical and management challenges. Persistent symptoms can be confusing to the family, possibly leading to the conclusion that many food triggers exist. Furthermore, persistent cough can exacerbate GERD. The triad of asthma, GERD (with vomiting), and food allergy is common. Cough and gag from persistent asthma may trigger vomiting that otherwise may be construed as food induced. Food refusal may be caused by true food allergy or persistent esophagitis. In these instances, asthma and GERD of course must be managed aggressively. Often, the identity of the true food triggers does not emerge until these concurrent problems are addressed.

FOOD ALLERGY

91

Laboratory Studies The general history and examination lead to a list of suspected foods, if any, and the likelihood for IgE- or T cell–mediated processes. Immediate hypersensitivity skin tests probe for the presence of food protein–specific IgE. Standard procedure involves the prick skin test. Multiple devices are available for this procedure, but the authors’ preference is a bifurcated steel needle because of enhanced sensitivity and specificity. Intradermal skin testing is not used because of increased nonspecificity and the risk for systemic reactions. Controls used include the glycerinated saline diluent (to rule out nonspecific or dermatographic reactions) and histamine (to screen for the presence of residual antihistamines in the tissues). Positive responses are dependent on the release of endogenous histamine; therefore, histamine receptor 2 (H2) blockers may blunt the normal wheal and flare response but not as substantially as H1 blockers. Prolonged systemic or topical steroid use may decrease skin mast cell numbers, thus reducing the capacity to elicit a response. To control for this factor, codeine, a direct mast cell activator, is used. Normally, histamine and codeine responses should be equivalent. If histamine responsiveness is intact but codeine elicits no response, skintest results are invalid. In general, skin tests have positive predictive accuracies of approximately 50%, but their negative predictive values are in excess of 95%.14 The authors inform their patients and families that a positive skin test only makes the patient eligible to have IgEmediated reactions. True clinical sensitivity must be proved with a food challenge, as discussed later; however, a negative skin test for IgEmediated problems is helpful if false-negative reactions are taken into consideration. False-negative reactions can occur secondary to the effect of drugs (H1 and H2 blockers and topical or systemic steroids) and with commercial extracts that may not preserve adequately the proteins needed to interact with IgE molecules. When this is suspected, testing should be repeated with fresh food, in particular for fruit and vegetable sensitivity. When this is performed, a control subject is helpful to rule out nonspecific reactions. An alternative method to detect food protein–specific IgE is by in vitro methods. Radioallergosorbent tests detect IgE in blood samples, but usually sensitivity and specificity are reduced compared with traditional skin testing. Some investigators may prefer to use in vitro testing when persistent dermatographism (rare) or severe eczema is present or when families are reluctant to discontinue H1 blockers. A newer in vitro test (CAP-RAST FEIA) allows for improved quantification of allergenspecific IgE; higher levels of reactivity in the six foods studied correlated with an increased risk for clinical sensitivity.44 More studies are needed to confirm the predictive capacity of this test. Basophil histamine release assays are one additional means to assess IgE-mediated sensitivity in vitro, but this protocol is laborious and less well explored and therefore remains a research tool. For non–IgE-mediated disorders, fewer laboratory diagnostic tools

92

SPERGEL & PAWLOWSKI

exist. As noted earlier, food patch testing is just now being explored. Eosinophils in the blood or stool may point to an ongoing enteropathy, but these findings are certainly nonspecific. Serum levels of allergenspecific IgG are not helpful. Endoscopy followed by examination of biopsy specimens can point to immunologic processes and therefore are the most important tools in non–IgE-mediated disorders. Challenges are needed to identify specific food triggers in all cases. Oral Food Challenges Often, an elimination diet provides diagnostic information and symptomatic relief. If not, it is possible that not all responsible foods have been eliminated (not all detected in diagnostic workup, hidden goods or contaminants present, or sufficient allergen is contained in ‘‘hypoallergenic’’ casein hydrolysate formula in milk-sensitive patients). Elemental diets (e.g., Neocate or EleCare) also may be helpful because they avoid all protein allergens. If the elimination diet is successful, food challenges are indicated to confirm the diagnosis and clarify the individual food triggers. For GI disorders, biopsy after elimination (normalization) and then after reintroduction (inflammatory response) can help to identify responsible food triggers. Oral food challenges are the key to establishing the identity of specific food triggers. The most rigorous method is double-blind and placebo-controlled, but single-blind (i.e., the patient) and open-labeled challenges can be performed. The least time-intensive procedure is the open-labeled challenge. If endpoints are defined and documented specifically, this procedure is satisfactory for diagnostic purposes.50 Doubleblind, placebo-controled challenges are indicated when the endpoints are subjective complaints (bias is possible) or there are specific research objectives. Before challenge, intended food is eliminated for 7 days, and H1 blockers (i.e., diphenhydramine) are discontinued for 3 days. The challenge procedure involves giving increasing doses at intervals during constant observation. The starting dose is sufficiently low not to trigger a severe reaction (e.g., ⬇ 100–500 mg). Intervals are shorter (⬇ 20 minutes) when testing for IgE-mediated processes rather than T cell– mediated processes (hours; e.g., FPIES). Once the top dose is reached, the observation period varies, with 2.5 hours for IgE-mediated reactions and 4 hours for T cell–mediated processes (e.g., FPIES). Longer time periods and multiple doses may be required to elicit reaction in some disorders (e.g., eosinophilic gastroenteritis). For safety, oral challenges should be performed only in a setting in which severe reactions (e.g., anaphylaxis or FPIES) can be treated (i.e., in a hospital). Other non–IgE-mediated disorders may be challenged on an outpatient basis if the diagnosis and risks are defined clearly. Because most food sensitivities resolve after prolonged abstinence, challenges are repeated over time (e.g., each year) to determine the onset of tolerance. Even some food sensitivities long considered to be lifelong

FOOD ALLERGY

93

(e.g., peanut) may resolve.52 Physicians continue to learn by performing food challenges.

GENERAL THERAPEUTIC CONSIDERATIONS The only proved therapy for food allergy is food elimination, which may require an intensive learning process and work on the patient’s and caregiver’s parts because of pervasiveness or hidden allergens. Certain situations bear a high risk for adverse reaction, such as eating foods prepared by others in cafeterias or restaurants. Finally, dietary elimination may lead to nutritional deficiency. Therefore, nutritional support services and continued diagnostic procedures to further limit the list of foods to be avoided may be crucial. The Food Allergy Network (www.foodallergy.org; 800-929-4040) is an essential resource for consultation. Drugs generally are used to treat only specific atopic symptoms. Occasionally, systemic oral steroids are indicated for severe enteropathy or nutritional risk. All patients at risk for anaphylaxis must be trained to identify early symptoms and be prepared to treat appropriately. Autoinjectable epinephrine, together with education to help identify avoidable risks, is essential. New strategies for treatment are on the horizon. Although ‘‘standard immunotherapy’’ (analogous to that for pollen sensitivity) so far has been disappointing, anti-IgE therapy using injections of humanized anti-IgE monoclonal antibodies is under intensive investigation. On the horizon is the prospect that the responsible antigens may be modified and made safe for immunomodulatory purposes. Likewise, oligonucleotide immunostimulatory sequences may be useful adjuvants to sway systemic responses away from the Th-2 or allergic phenotype. Assiduous food elimination is the most important treatment modality. Sufficient data (reviewed by Sampson43) indicate that tolerance develops for most IgE-mediated disorders and FPIES. Eosinophilic gastroenteritis has neither been defined nor studied sufficiently to know its potential for resolution. Unfortunately, some immunologically based disorders, such as celiac disease, are lifelong but still can be managed effectively by dietary elimination.

References 1. Abernathy-Carver KJ, Sampson HA, Picker LJ, et al: Milk-induced eczema is associated with the expansion of T cells expressing cutaneous lymphocyte antigen. J Clin Invest 95:913–918, 1995 2. Attwood S, Smyrk T, Demeester T, et al: Esophageal eosinophilia with dysphagia: A distinct clinicopathologic syndrome. Dig Dis Sci 38:109–116, 1993 3. Bacharier LB, Geha RS: Molecular mechanisms of IgE regulation. J Allergy Clin Immunol 105(suppl):547–558, 2000 4. Benlounes N, Dupont C, Candalh C, Blaton MA, et al: The threshold for immune cell

94

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

SPERGEL & PAWLOWSKI

reactivity to milk antigens decreases in cow’s milk allergy with intestinal symptoms. J Allergy Clin Immunol 98:781–789, 1996 Black DD, Haggitt RC, Orenstein SR, et al: Esophagitis in infants: Morphometric histological diagnosis and correlation with measures of gastroesophageal reflux. Gastroenterology 98:1408–1414, 1990 Brown K: Eosinophilic enteritis. In Altschuler S, Liacouras C (eds): Clinical Pediatric Gastroenterology. Philadelphia, Churchill Livingstone, 1998 Campbell NA, Kim HS, Blumberg RS, et al: The nonclassical class I molecule CD1d associates with the novel CD8 ligand gp180 on intestinal epithelial cells. J Biol Chem 274:26259–26265, 1999 Castells M, Horan R, Sheffer A: Exercise-induced anaphylaxis (EIA). Clin Rev Allergy Immunol 17:413–424, 1999 Collin P, Kaukinen K, Maki M: Clinical features of celiac disease today. Dig Dis Sci 17:100–106, 1999 Crump JA, McLay CL, Chambers ST: Ciguatera fish poisoning. Postgrad Med J 75:678– 679, 1999 de Jong EC, Spanhaak S, Martens BP, et al: Food allergen (peanut)-specific TH2 clones generated from the peripheral blood of a patient with peanut allergy. J Allergy Clin Immunol 98:73–81, 1996 Dobbins J, Sheahan D, Behar J: Eosinophilic gastroenteritis with esophageal involvement. Gastroenterology 72:1312–1316, 1977 Dupont C, Heyman M: Food protein-induced enterocolitis syndrome: Laboratory perspectives. J Pediatr Gastroenterol Nutr 30(suppl):50–57, 2000 Eigenmann PA, Sampson HA: Interpreting skin prick tests in the evaluation of food allergy in children. Pediatr Allergy Immunol 9:186–191, 1998 Fritsch R, Bohle B, Vollmann U, et al: Bet v 1, the major birch pollen allergen, and Mal d 1, the major apple allergen, cross-react at the level of allergen-specific T helper cells. J Allergy Clin Immunol 102:679–686, 1998 Fukushima Y, Kawata Y, Onda T, et al: Consumption of cow milk and egg by lactating women and the presence of beta-lactoglobulin and ovalbumin in breast milk. Am J Clin Nutr 65:30–35, 1997 Goodrich ME, McGee DW: Preferential enhancement of B cell IgA secretion by intestinal epithelial cell-derived cytokines and interleukin-2. Immunol Invest 28:67–75, 1999 Grewe M, Walther S, Gyufko K, et al: Analysis of the cytokine pattern expressed in situ in inhalant allergen patch test reactions of atopic dermatitis patients. J Invest Dermatol 105:407–410, 1995 Hamosh M: Digestion in the newborn. Clin Perinatol 23:191–209, 1996 Hoffman KH, Ho DG, Sampson HA: Evaluation of the usefulness of lymphocyte proliferation assays in the diagnosis of allergy to cow’s milk. J Allergy Clin Immunol 99:360–367, 1997 Hyman P, Clarke D, Everett S, et al: Gastric acid secretory function in preterm infants. J Pediatr 106:467–471, 1985 Iacono G, Carroccio A, Cavataio F, et al: Gastroesophageal reflux and cow’s milk allergy in infants: A prospective study. J Allergy Clin Immunol 97:822–827, 1996 Isolauri E, Turjanmaa K: Combined skin prick and patch testing enhances identification of food allergy in infants with atopic dermatitis. J Allergy Clin Immunol 97:9–15, 1996 James JM, Eigenmann PA, Eggleston PA, et al: Airway reactivity changes in asthmatic patients undergoing blinded food challenges. Am J Respir Crit Care Med 153:597– 603, 1996 Katz A, Twarog F, Zeiger R, et al: Milk-sensitive and eosinophilic gastroenteropathy: Similar clinical features with contrasting mechanisms and clinical course. J Allergy Clin Immunol 74:72–78, 1984 Kelly K: Eosinophilic gastroenteritis. J Pediatr Gastroenterol Nutr 30(suppl):28–35, 2000 Kelly K, Lazenby A, Rowe P, et al: Eosinophilic esophagitis attributed to gastroesophageal reflux: Improvement with an amino acid-based formula. Gastroenterology 109:1503–1512, 1995

FOOD ALLERGY

95

28. Lake AM: Beyond hydrolysates: Use of L-amino acid formula in resistant dietary protein-induced intestinal disease in infants. J Pediatr 131:658–660, 1997 29. Landres R, Kuster G, Strum W: Eosinophilic esophagitis in a patient with vigorous achalasia. Gastroenterology 74:1298–1301, 1978 30. Liacouras CA, Wenner WJ, Brown K, et al: Primary eosinophilic esophagitis in children: Successful treatment with oral corticosteroids. J Pediatr Gastroenterol Nutr 26:380– 385, 1998 31. Majamaa H, Moisio P, Holm K, et al: Cow’s milk allergy: Diagnostic accuracy of skin prick and patch tests and specific IgE. Allergy 54:346–351, 1999 32. Majamaa H, Moisio P, Holm K, et al: Wheat allergy: Diagnostic accuracy of skin prick and patch tests and specific IgE. Allergy 54:851–856, 1999 33. Marrack P, Hugo P, McCormack J, et al: Death and T cells. Immunol Rev 133:119–129, 1993 34. Matzinger M, Daneman A: Esophageal involvement in eosinophilic gastroenteritis. Pediatr Radiol 13:35–38, 1983 35. Mayer L: Mucosal immunity and gastrointestinal antigen processing. J Pediatr Gastroenterol Nutr 30(suppl):4–12, 2000 36. Mehlhop PD, van der Rijn M, Goldberg AB, et al: Allergen-induced bronchial hyperreactivity and eosinophilic inflammation occur in the absence of IgE in a mouse model of asthma. Proc Natl Acad Sci U S A 94:1344–1349, 1997 37. Mishra A, Hogan SP, Brandt EB, et al: An etiological role for aeroallergens and eosinophils in experimental esophagitis. J Clin Invest 107:83–90, 2001 38. Moon A, Kleinman R: Allergic gastroenteropathy in children. Ann Allergy Asthma Immunol 74:5–9, 1995 39. Orenstein SR, Shalaby TM, Di Lorenzo C, et al: The spectrum of pediatric eosinophilic esophagitis beyond infancy: A clinical series of 30 children. Am J Gastroenterol 95:1422–1430, 2000 40. Patel YT, Minocha A: Lactose intolerance: Diagnosis and management. Compr Ther 26:246–250, 2000 41. Poon E, Nixon R: Cutaneous spectrum of coeliac disease. Aust J Dermatol 42:136– 138, 2001 42. Sampson H: Food allergy. In Kay AB (ed): Allergy and Allergic Diseases. London, Blackwell Science, 1997, pp 1517–1549 43. Sampson H: Food allergy: II. Diagnosis and management. J Allergy Clin Immunol 103:981–989, 1999 44. Sampson H: Utility of food-specific IgE concentrations in predicting symptomatic food allergy. J Allergy Clin Immunol 107:891–896, 2001 45. Sampson HA: Food allergy: I. Immunopathogenesis and clinical disorders. J Allergy Clin Immunol 103:717–728, 1999 46. Sicherer S: Determinants of systemic manifestations of food allergy. J Allergy Clin Immunol 106(suppl):251–257, 2000 47. Sicherer S: Food protein-induced enterocolitis syndrome: Clinical perspectives. J Pediatr Gastroenterol Nutr 30(suppl):45–49, 2000 48. Sicherer S, Eigenmann P, Sampson H: Clinical features of food protein-induced enterocolitis syndrome. J Pediatr 133:214–219, 1998 49. Sicherer SH, Sampson HA: Food hypersensitivity and atopic dermatitis: Pathophysiology, epidemiology, diagnosis, and management. J Allergy Clin Immunol 104(suppl): 114–122, 1999 50. Spergel J, Beausoleil J, Williams J, et al: The use of open food challenges to demonstrate the prevalence of food allergy in children [abstract]. J Allergy Clin Immunol 107:358, 2001 51. Spergel J, Schneider L: Atopic dermatitis [online]. The Internet Journal of Asthma, Allergy and Immunology 1, 1997 52. Spergel JM, Beausoleil JL, Pawlowski NA: Resolution of childhood peanut allergy. Ann Allergy Asthma Immunol 85:473–476, 2000 53. Strober W: Interactions between epithelial cells and immune cells in the intestine. Ann NY Acad Sci 859:37–45, 1998

96

SPERGEL & PAWLOWSKI

54. Vadas P, Wai Y, Burks W, et al: Detection of peanut allergens in breast milk of lactating women. JAMA 285:1746–1748, 2001 55. van de Rijn M, Mehlhop P, Judkins A, et al: A murine model of allergic rhinitis: Studies on the role of IgE in pathogenesis and analysis of the eosinophil influx elicited by allergen and eotaxin. J Allergy Clin Immunol 102:65–74, 1998 56. Van Sickle GJ, Powell GK, McDonald PJ, et al: Milk- and soy protein-induced enterocolitis: Evidence for lymphocyte sensitization to specific food proteins. Gastroenterology 88:1915–1921, 1985 57. Vitellas K, Bennett W, Bova J, et al: Radiographic manifestations of eosinophilic gastroenteritis. Abdom Imaging 20:406–413, 1995 58. Walker-Smith J: Food sensitive enteropathy: Overview and update. Acta Paediatr Jpn 36:545–549, 1994 59. Walker-Smith JA, Ford RP, Phillips AD: The spectrum of gastrointestinal allergies to food. Ann Allergy 53:629–636, 1984 60. Walsh SV, Antonioli DA, Goldman H, et al: Allergic esophagitis in children: A clinicopathological entity. Am J Surg Pathol 23:390–396, 1999 61. Yocum MW, Khan DA: Assessment of patients who have experienced anaphylaxis: A 3-year survey. Mayo Clin Proc 69:16–23, 1994 Address reprint requests to Jonathan M. Spergel, MD, PhD Allergy Section Wood 5314 The Children’s Hospital of Philadelphia 34th Street and Civic Center Boulevard Philadelphia, PA 19104–4399 e-mail: [email protected]