Philippe A. Eigenmann, MD Hôpital des Enfants 6, rue Willy-Donzé 1211 Geneva 14 Switzerland
Up to 5% of young children and 2% of adults suffer from food allergy. Among them many have immunoglobulin E (IgE)-mediated food allergy, a condition with potentially fatal allergic reactions. Several studies have addressed possible definite treatment options for food allergy. Immunotherapy, by the oral route or by systemic injections shows promising preliminary results, but current interpretation of these therapeutic options are mostly handicapped by studies with insufficient scientific support, or by severe side-effects. Currently, no studies can support pharmacotherapy. Finally, most promising results were recently published with anti-IgE antibodies in a human trial, or various approaches in a mouse model of food allergy (chinese herbal medicine, specific modulation of the T cell response). Rapidly evolving findings might provide hope for a cure of food allergy in the near future.
Allergy has become a major burden in westernized societies during the 20th century, and among allergic manifestations, food allergy accounts for a substantial part of the morbidity rate. Patients with food allergy are handicapped by a strict avoidance diet and a potentially lifelong-lasting disease. Furthermore, in some patients with exquisite food allergy, the risk of fatal reactions is present daily.
Current figures show that up to 5% of young children might suffer from immunoglobulin E (IgE)-mediated food hypersensitivity (1–5). In adults, probably <2% of the population suffers from food allergy (6, 7). In westernized societies, the foods most commonly responsible for allergic reactions are: egg, peanuts, milk, tree nuts, fish, as well as grains, soy and many others, as potentially any food can provoke an IgE-mediated reaction.
A majority of reactions are self-limited, or easily treated with an oral antihistamine treatment and/or injection of adrenaline in potentially life-threatening reactions, but major public concern has increased after reports of fatal allergic reactions caused by foods. Correct numbers of fatalities are hard to determine but projections can help to visualize the size of the problem. In a 4-year time-period community base study in the USA, an anaphylaxis occurrence rate of 30 of 100 000 persons per year was reported, and allergic reactions to foods were the leading identifiable cause of reactions (8). From this data, the number of 29 000 anaphylactic episodes because of food in the USA per year could be extrapolated, with approximately 150 expected deaths per year (9). Reported to the European population this number would be above 200 deaths per year. Knowing that, despite strict avoidance, most food allergic patients carry a high risk of reacting to commercially processed foods contaminated with common food allergens, a pro-active treatment of food allergy is strongly needed (10).
Mechanisms of food hypersensitivity
According to the revised nomenclature of the European Academy of Allergy and Clinical Immunology (EAACI), the denomination ‘food hypersensitivity’ is common to nonallergic food reactions and to reactions caused by food allergy (11). Among these, it can be further discriminated between IgE and non-IgE-mediated food allergy. On the contrary, partial lactase deficiencies are probably the most common nonallergic food hypersensitivities, as the majority of the world population is partially lactase deficient. However, this and other nonallergic food hypersensitivity do not cause potentially fatal food related reactions. Among immunologically mediated reactions, non-IgE-mediated food hypersensitivity is often seen in young children mostly with gastro-intestinal symptoms (12). Again, avoidance is the only recommendation to provide to these patients. While an impairment in daily life is present in all patients with food hypersensitivity, those with IgE-mediated food allergy, accounting for the most severe manifestations of food hypersensitivity, would benefit most from a pro-active food allergy treatment.
Treatment of food allergy, historical and current clinical aspects
Recognition of the potential hazards of food hypersensitivity has prompted allergists to seek for effective treatments of this condition already decades ago. Oral desensitizations to foods, mostly to milk, have been reported in numerous limited series but with various results (13–16). It is difficult to draw any conclusion from these early isolated reports. However, it can be stated that while many patients were reporting to tolerate increased amounts of foods, most were still reactive to larger quantities. Furthermore, the effectiveness of the procedure relies mostly on a daily ingestion of a certain amount to keep the partial tolerance achieved by this procedure.
Recently, Patriarca et al. published a series of 59 patients with food allergy who underwent an oral desensitization protocol (17). This population consisted of a mixed sample of adults and children. Twenty-nine were allergic to milk, 18 to egg, 11 to fish and nine to other foods. The diagnosis relied on skin prick tests (SPT) and/or assessment of serum specific IgE levels. Double-blind placebo control food challenge (DBPCFC) procedures were also used in some patients. Specific oral desensitization protocols for the different foods were then applied to all subjects. They were stretched over 60–84 days up to a maintenance dose of 120 ml of milk, 50 ml of egg or 160 mg of boiled codfish. The authors reported successful desensitizations in 24 of 29 patients with milk allergy, 13 of 15 patients with egg allergy and eight of 11 patients with fish allergy. However, the authors reported that desensitization procedures could not be completed in some patients because of recurrent urticaria or gastro-intestinal symptoms. Unfortunately, this study was not validated by a control group and was not blinded.
In summary, oral desensitization is a conceptually attractive procedure. However, well-designed placebo controlled studies are needed for proper assessment of the method. Furthermore, such procedures are hampered by potential severe side effects, although uncommon (18), and can probably not be applied in the patients who would most desperately need a treatment of food allergy, the patients with potentially fatal food allergic reactions.
Peanut allergy is a strongly increasing phenomenon (19), and the severity of reaction in peanut allergic patients led investigators in the early 1990s to start a trial with peanut-rush immunotherapy (20). This study enrolled 11 subjects with a history of systemic reactions to peanuts. Measures of efficacy included a symptom score with DBPCFC and titrated SPT results. In the three subjects treated with peanut immunotherapy who completed the studies, a decrease in symptoms of 67–100% measured during DBPCFC could be seen. However, the rate of systemic reactions with rush immunotherapy was 13.3%, and unfortunately one patient died after immunotherapy injection. These severe side effects prompted the authors not to recommend this procedure for routine treatment of peanut allergy.
The same group investigated in a second study 12 patients with exquisite sensitivity to peanuts (21). Half of them underwent immunotherapy with aqueous peanut extracts during 1 year. All treated patients experienced increased tolerance to DBPCFC challenge to peanut. However, systemic reactions were common in this group of patients and half of them required a dose reduction. After dose reduction, the partial tolerance to peanuts by oral challenge was partially or completely lost.
From these studies, it can be concluded that systemic immunotherapy with peanut extract carry a high risk of severe side effects and that they are only partially effective. Future studies should consider modified allergen extracts with a predominant T cell epitope immuno-modulating effect. Preliminary results with such modified allergens have already been published, mostly using peanut antigens, with most encouraging results (22, 23). In such a study, the cDNA of Ara h 3, one of the major peanut allergens, could be modified in order to obtain a decrease of 35–85% of binding to IgE from patients with peanut allergy. Such an genetically engineered protein might be used for immunotherapy without potential immediate side-effects (24).
Few studies with pharmacological agents were performed in food hypersensitivity. Oral sodium cromolyn has been studied by several authors, however, only few trials used a well-designed method allowing proper conclusions (25–27). In one of these studies, Burks and Samson enrolled 10 patients who underwent a double-blind, placebo controlled cross-over trial with oral cromolyn (26). In the patients who reacted to the subsequent food challenge, there was no significant difference in the threshold level, or the severity of the symptoms in the groups with or without treatment. The authors concluded that oral cromolyn sodium has no benefit in the treatment of children with atopic dermatitis and food hypersensitivity.
One of the most common manifestation of food allergy is the oral allergy syndrome. These patients exhibit allergic rhinitis and conjunctivitis, to tree, grass, or weed pollens, and present pharyngeal and laryngeal symptoms after ingestions of cross-reacting foods. Several authors have tried pollen immunotherapy in order to decrease oral symptoms in these patients. Overall, the results are mostly inconclusive, with some studies showing a benefit but others showing only deceiving results of specific immunotherapy in oral allergy syndrome (28–30). Again, modification of native allergens or usage of recombinant allergens might improve the treatment of pollen associated oral allergy syndrome (Table 1).
Table 1. Potential treatments for food allergy and their limitations
New strategies for treatment of food allergy in humans
Infusion of anti-IgE antibodies
Most recently, a multi-center trial with anti-IgE therapy in patients with peanut allergy has shown some effectiveness. Leung and other investigators conducted a double-blind, randomized trial in 84 patients with a history of immediate hypersensitivity to peanuts (31). The level of clinical sensitivity was determined by a DBPCFC prior to the study. The patients received 150, 300, 450 mg of an anti-IgE antibody (TNX-901; Tanox, Houston, TX) or placebo. Injections were given sub-cutaneously every 4 weeks for a total of four doses. The DBPCFC were then performed to assess the possible change in the level of sensitivity. The mean threshold dose of peanut eliciting symptoms significantly rose in a dose dependent manner in the patients receiving the active treatment. Although this procedure did not completely cure these patients with food allergy, it provided them an interesting option to diminish the risk of potentially severe or fatal reaction after an accidental ingestion. However, several questions remain open. Long term assessment of these patients will be most interesting to show if their change in sensitivity subsists over time. Furthermore, this treatment might be of peculiar interest to patients with multiple food allergies in whom an strict diet is difficult to implement.
Numerous evidences link IgE-mediated food hypersensitivity to a TH2-type disease. The origin of the TH2 skewing of the immune system might be linked to a specific microbial environment present early in life (according to the hygiene hypothesis). This has prompted several investigators to consider the use of micro-organisms to treat or prevent food allergy. Although most studies have been performed in animals, a report on a small number of patients with cows milk allergy treated with Lactobacillus GG in an extensively hydrolyzed formula has shown an improvement in the clinical score for atopic dermatitis during the 1 month study period (32). Such results are most promising, however more studies are needed before definite conclusions can be drawn.
Studies in animal models
The difficulty of studying immune mechanisms in the gut of patients with food allergy has prompted several investigators to set-up animal (dogs, rats and mice) models (33–36). Most recently published studies rely on a mouse model initially reported by Snyder et al. (37). A specific strain of mice, the C3H-HeJ strain is orally sensitized with common food allergen in presence of cholera toxin used as an adjuvant. Upon subsequent oral challenge, these mice elicit symptoms comparable with the clinical picture of food allergy in humans.
Several groups have used this model to test various hypotheses for the treatment of food allergy. Our group has used a high-molecular-weight polysaccharide, λ-carrageenan, for the prevention or the treatment of food allergy (38). We found that the mice preventively fed λ-carrageenan together with β-lactoglobulin were prevented from developing cow's milk allergy. Investigating possible mechanisms, we observed that mesenteric lymph nodes and spleen cells were rendered anergic in these mice. However, this procedure was not effective for the treatment of already sensitized mice.
Vaccination with plasmid DNA
Li et al. investigated the potential application of DNA immunization (39). Interestingly, they found that the response to plasmid DNA coding for a major peanut allergen (Ara h 2), was mostly strain dependent. In the C3H strain, the mice reacted more severely after treatment while AKR/J or BALB/c mice did not react after immunization. These results indicated that this strategy could be hazardous in human.
Complementary medicine such as Chinese herbs
More promising results were obtained by the same group using a Chinese herbal tee formula (FAHF-1) (40). When FAHF-1 treatment was initiated after sensitization of the mice for a total of 7 weeks, treated mice did not show any sign of anaphylaxis after peanut challenge. Peanut specific serum IgE levels were significantly reduced and the peanut antigen-induced TH2 cytokine production was diminished. Such strategies might be most welcome in humans. Current efforts aim to characterize the active ingredient in this herbal tea.
Cytokines and cytokine modifiers
Modulation of the immune system by promoting a TH1-type environment might be effective in treating food allergy. Lee et al. administered liposome-encapsulated rIL-12 and showed that this treatment could prevent and reverse peanut hypersensitivity in the mouse model (41). Oral IL-12 treatment reduced peanut specific serum IgE and IgG-1. The TH1 type cytokyne interferon-γ was increased but the levels of IL-4 or IL-5 did not change. A most promising approach has recently been used successfully in mice given a combination of oral DNA coding for Ara h 2 (a major peanut allergen) and chitosan, a natural biocompatible polysaccharide, resulting in prevention of subsequent sensitization with peanut antigen (42).
Together, these mice studies provide most exciting therapeutic options potentially applicable to human. However, the data gathered in the study with naked DNA show that data obtained in a specific strain of mice might not be applicable in other strain or in other mammals, i.e. in humans.
Future potential treatments aiming at T cells modifications
Adhesion molecule antagonists
In the pathogenesis of food allergic reactions, cytokines secreted by T cells play a major role for inflammatory cell recruitment, maturation, and for IgE production facilitation. Trials targeting cytokines were mainly focused on respiratory allergy. Anti-soluble IL-4 receptors Abs given in a phase 2 study in asthmatic patients diminished inhaled corticosteroid dependency, and might have acted also on VCAM-1 expression, the endothelial ligand for α4β1 (43). Recently, a clinical trial with an anti-IL-5 Ab (SB-240563) showed a reduction of blood and sputum eosinophils in atopic asthmatic subjects, but failed to inhibit airway hyperresponsiveness and late phase bronchoconstriction (44).
Directed trafficking of food-antigen primed T cells might be crucial for initiation of allergic reactions. It is tempting to imagine a therapeutic intervention blocking T cells or other inflammatory cells from trafficking to target organs (Fig. 1). Relatively few studies have addressed this question. Monoclonal antibodies against the α4β7-integrin have reversed experimentally-induced colitis in animals as expressed by a reduction of the inflammatory infiltrate on biopsies, as well as rapid improvement of stool consistency (45). In an ocular model of allergic diseases, anti-ICAM-1 and anti-LFA-1 monoclonal antibodies were able to reduce the clinical signs of allergic conjunctivitis (46). Clinical studies will have to confirm these promising results.
Chemokine and chemokine receptor antagonists
Chemokines and chemokine receptors play a prime role in the pathogenesis of allergy. Another promising direction for therapeutic action in food allergy might be achieved by blocking receptors for ‘allergy’ chemokines. Interventions on these mediators of the allergic reaction might be particularly efficient, as chemokines can influence inflammatory reactions at several distinct levels, such as cell recruitment, integrin expression, and cell maturation. A monoclonal Ab against CCR3, the receptor for eotaxin, was able to block eosinophil migration, thus confirming the function of CCR3 and a possible therapeutic intervention on eosinophil migration (47). A synthetic molecule (UCB 35625) blocking CCR3 and CCR1 has been recently developed, and might be soon in clinical trials. Future clinical trials in various allergic diseases will hopefully reveal a place for therapeutic agents acting at this level of inflammation in allergic diseases.
The IgE-mediated food allergy is a common disease with potentially severe or fatal reactions in some patients. The life of these patients is mostly impaired by a strict avoidance diet and a risk of accidental ingestions because of contamination in processed foods. An efficient pro-active treatment of food allergies is desperately needed in these patients. Although no treatment is currently available in daily clinical practice, options studied include the use of anti-IgE antibodies, immunomodulatory treatment such as Chinese traditional herbal medicine. Furthermore, the potentially severe side effects of immunotherapy could be controlled by using modified or recombinant allergens.
We might hope that in the near future we should be able to provide our patients with food allergy with an efficient and safe treatment of their condition.
The author is supported in part by grants from the Swiss National Research Foundation (3200-065203.01), the Sir Jules Thorn Charitable Trust, and the Helmut Horten Foundation. Sabine Cuenod is acknowledged for excellent secretarial assistance.