Food and drug reactions and anaphylaxis

Transcription

1 The distribution of individual threshold doses eliciting allergic reactions in a population with peanut allergy Marjolein Wensing, MD, a André H. Penninks, PhD, b Susan L. Hefle, PhD, c Stef J. Koppelman, PhD, b Carla A. F. M. Bruijnzeel-Koomen, MD, a and André C. Knulst, MD a Utrecht and Zeist, The Netherlands, and Lincoln, Neb Background: Hidden peanut in consumer products can endanger patients with peanut allergy. Individual threshold doses for eliciting allergic reactions need to be elucidated to assess the risks for development of allergic reactions after accidental ingestion of peanut in a population with peanut allergy. Objective: We sought to determine the distribution of individual threshold doses in a population with peanut allergy and to correlate these thresholds to the severity of peanut-induced symptoms. Methods: Twenty-six adult patients with a convincing history of peanut-related symptoms, a specific IgE level of 0.7 ku/l or greater, or a positive skin prick test response of 2+ or greater to peanut were included. These patients underwent doubleblind, placebo-controlled food challenges with increasing doses of peanut. A threshold dose could be established when objective or repetitive subjective reactions occurred after active doses. Results: All patients had subjective oral symptoms (n = 26), prior subjective gastrointestinal symptoms (n = 14), or objective symptoms (n = 5). Reactions started within 30 minutes after ingestion of peanut, but in 2 patients additional symptoms were delayed by 1 to 2 hours. Threshold doses for allergic reactions ranged from a dose as low as 100 µg up to 1 g of peanut protein. Fifty percent of the study population (95% CI, 30%-70%) already had an allergic reaction after ingestion of 3 mg of peanut protein. Patients with severe symptoms had lower threshold doses compared with those of patients with mild symptoms (P =.027). Conclusions: A substantial part of a population with peanut allergy will react to very low amounts of peanut, requiring accurate declaration of peanut content in consumer products. This is even more important because patients with severe reactions react to lower doses than patients with mild symptoms. (J Allergy Clin Immunol 2002;110: ) From a the Department of Dermatology/Allergology, University Medical Centre Utrecht, Utrecht; b TNO Nutrition and Food Research Institute, Zeist; and c the University of Nebraska, Lincoln. Supported by Pharmacia & Upjohn. Received for publication April 22, 2002; revised August 19, 2002; accepted for publication August 20, Reprint requests: Marjolein Wensing, MD, Department of Dermatology and Allergology G02.124, University Medical Centre Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands Mosby, Inc. All rights reserved /2002 $ /87/ doi: /mai Key words: Food allergy; double-blind, placebo-controlled food challenge; peanut allergy; threshold dose Peanut is one of the most common food allergens. 1 Two recent studies estimated a prevalence of peanut allergy in adults of 0.48% in the United Kingdom and, together with tree nut allergy, of 1.1% in the United States. 2,3 Several studies and case reports have demonstrated the severe nature of allergic reactions in patients with peanut allergy. 4-6 Peanut allergy is often diagnosed in childhood and is thought to be persistent, 7 although recent studies claim that some children with peanut allergy can outgrow their allergy Because peanuts often serve as protein supplements and have good emulsifying properties and taste, they are widely used in processed foods. Because of incomplete or misleading labels on prepackaged foods, accidental ingestion frequently occurs. Hidden peanut can also end up in foods as unintended contamination caused by shared use of production lines or kitchen utensils. Despite the fact that such contamination usually concerns only minute quantities, allergic reactions have been described. 11,12 Not all allergic patients will react to such very low doses of peanut protein. Actually, the risk of development of an allergic reaction caused by hidden allergen is still unknown. For accurate risk assessment, information about consumption patterns and quantities of peanuts present in consumer products need to be combined with information about individual threshold doses in patients with peanut allergy. Hourihane et al 13 performed doubleblind, placebo-controlled food challenges (DBPCFCs) to determine threshold doses in 14 adult patients selected because of their high sensitivity, as shown in a previous study. They found objective reactions after a dose as low as 2 mg and subjective reactions after a dose of 100 µg of peanut protein. Five subjects passed the challenge (up to a maximum dose of 50 mg of peanut protein), leaving 9 subjects in whom individual threshold doses were revealed. This study is biased toward low threshold doses. More patients need to be challenged with higher doses to elucidate individual threshold doses in a representative population with peanut allergy to get more accurate insight into the distribution of threshold doses. This study aimed at revealing the distribution of threshold doses in an adult population with peanut aller- 915

2 916 Wensing et al J ALLERGY CLIN IMMUNOL DECEMBER 2002 Abbreviations used DBPCFC: Double-blind, placebo-controlled food challenge NOAEL: No-observed-adverse-effect level SPT: Skin prick test gy. Twenty-six patients were challenged with DBPCFCs. The observed threshold doses were further correlated with the severity of their peanut-related symptoms. METHODS Patients Adult patients of the outpatient department of Allergology of the University Medical Centre Utrecht with a history of immediate-type adverse reactions to peanut were recruited. They were included in the study if they had a positive skin prick test (SPT) response to peanut of 2+ or greater, a peanut-specific IgE level of 0.7 ku/l or greater, or both. Reported peanut-related symptoms were documented and evaluated by using an adapted version of the Mueller classification, which was originally designated for grading of insect sting induced allergic reactions. 14 Symptoms of the oral cavity and skin symptoms were classified as Mueller grade 1; gastrointestinal symptoms, laryngeal edema, or both as grade 2; dyspnea as grade 3; and cardiovascular symptoms, such as a decrease in blood pressure with additional collapse, as grade 4. Patients were classified according to their most severe reactions. Pregnant women; patients using β-blocking agents, angiotensin-converting enzyme inhibitors, or immunosuppressives; and patients with any significant concurrent disease were excluded from the study. Before the challenge, all patients were examined regarding their health condition. SPTs were performed with a commercial peanut extract (ALK-Abelló, Nieuwegein, The Netherlands), and the skin test response was documented according to the method of Dreborg and Frew. 15 Total and peanut-specific IgE levels were determined by using the CAP system FEIA (Pharmacia & Upjohn, Uppsala, Sweden). Patients entered the study when physical examination, serum chemistry and hematology, and electrocardiography revealed no abnormalities. Antihistamines were discontinued for 1 week, short-acting β-agonists for 8 hours, and long-acting β-agonists for 48 hours before the challenge. All patients accurately avoided the offending food for at least 2 weeks before the challenge. Ethical considerations This study was reviewed and approved by the ethical committee of the University Medical Centre Utrecht. All patients provided written informed consent before enrollment in the study. Peanut source Commercially available roasted peanut meal (Runner peanuts) was partially defatted to a fat content of 43.2%. The meal was stored at 20 C until use. The protein content of the peanut meal was 25.5%, which was determined by using Kjehldahl analysis, measuring the protein content of the whole food without the preceding extraction step. 16 SDS-PAGE analysis confirmed the presence of proteins migrating at the same molecular weights as the purified peanut allergens Ara h 1, Ara h 2, and Ara h 3 (data not shown) Challenge meals The hospital pharmacy weighed out a total of 10 different peanut doses: 30 µg, 100 µg, 300 µg, 1 mg, 3 mg, 10 mg, 30 mg, 100 mg, 300 mg, and 1 g of peanut protein. These doses were administered to the patients in 2 separate challenges. The first challenge consisted of the 7 lowest doses (30 µg-30 mg). Patients who did not react during this challenge were asked to participate in a second challenge with 2 overlapping doses (10 and 30 mg) and 3 higher doses (100 mg, 300 mg, and 1 g). The doses were filled up with instant mashed potatoes and 0.33% of NaCl to a final weight of 15 g for the first challenge. In the second challenge oatmeal flakes were used instead of mashed potatoes, together with 0.33% NaCl and 33% brown sugar. The oatmeal was used because this better camouflaged the texture of the higher doses of peanut meal. The placebo doses consisted of mashed potatoes and salt or oatmeal flakes and brown sugar alone. A sensory test was performed in 80 healthy volunteers to investigate the masking of peanut in the used recipes. Mashed potatoes hiding 3 doses of peanut (3, 10, and 30 mg of peanut protein) were tested using the same conditions as during the challenges. Even when simultaneously administering placebo and active doses to the volunteers, no significant differences were observed (unpublished data). Just before being administered to the patients, the instant mashed potatoes or oatmeal flakes were mixed with water that had just boiled. The increasing peanut doses were randomly interspersed with an equal number of placebos by the pharmacy so that the investigator could stay blinded when administering the doses to the patient. All patients wore nose clips during preparation and ingestion of the test doses. DBPCFCs Challenges were performed in the period between February 2000 and July On the day of the challenge, only a light breakfast (crackers with tea or coffee) was allowed (>1 hour before the first test meal). Challenges were performed in a clinical research setting equipped for resuscitation and monitoring of vital signs. Intravenous access was established before the first test meal, and rescue medication was present in the room. During the challenge, the investigator or a nurse accompanied the patient. An interval of 30 minutes was allowed between each test meal when no reaction occurred or longer if symptoms were still present. Before and at regular intervals during the challenge, blood pressure, heart rate, and FEV 1 were measured for safety reasons and as objective parameters for detecting allergic reactions. Other objective symptoms indicative for an allergic reaction that were assessed during the challenge were skin manifestations, wheezing, vomiting, diarrhea, swelling in the oral cavity, and rhinoconjunctivitis. The skin and oral cavity were inspected before the challenge for preexisting lesions and during the challenge when symptoms, such as itchiness or feeling of swelling, occurred. Possible allergic reactions, such as itching of the skin or in the oral cavity (referred to as the oral allergy syndrome), 20 nausea, abdominal pain, and dyspnea (without objective decrease in FEV 1 or wheezing), were regarded as subjective. A challenge was discontinued when objective symptoms occurred or when subjective symptoms lasted for longer than 1 hour. A challenge result was regarded as positive when objective symptoms occurred or when subjective symptoms occurred repeatedly after active doses. 21 The threshold dose is determined as the lowest dose of peanut protein eliciting a convincing allergic reaction, but it is important to realize that the threshold dose for developing an allergic reaction will actually lie somewhere between the highest observed peanut dose not eliciting an allergic reaction, the no-observed-adverse-effect level (NOAEL), and the lowestobserved-adverse-effect level. 22 Patients were kept for observation for at least 2 hours after ingestion of the last test meal. All patients were contacted by telephone the day after the challenge to evaluate possible late reactions. Statistics For the cumulative distribution of threshold doses, exact 95% CIs were calculated according to a binomial distribution. Differences in threshold doses between patients with Mueller 1, 2, and 3

3 J ALLERGY CLIN IMMUNOL VOLUME 110, NUMBER 6 Wensing et al 917 TABLE I. Patient characteristics and results of DBPCFCs Subjective Objective CAP MPD symptoms symptoms Age Peanut-related Age at SPT peanut Total (mg peanut during during Patient no. Sex (y) symptoms diagnosis (y) peanut (ku/l) IgE protein) DBPCFC DBPCFC 1 F 18 Oas, cu, dys Oas, gi Ls 2 F 22 Oas 4 4+ < Oas 3 F 41 Oas, cu, rc Oas 4 F 25 Oas, gi, le < Oas, gi 5 F 33 Oas, gi, cu < Oas Ls 6 F 35 Oas, gi, dys Oas, gi 7 M 29 Oas, cu Oas 8 F 20 Oas, gi, cu Oas, gi Ls 9 F 26 Oas < Oas 10 F 24 Oas, cu Oas, gi 11 F 19 Oas, gi, dys, le, gu, rc 4 4+ > Oas, gi Le 12 F 20 Gi, dys, le, gu, rc < Oas, gi 13 F 31 Oas, cu Oas 14 F 33 Oas, gi, dys < Oas, gi Vo 15 M 17 Oas Oas 16 M 18 cu, le Oas 17 F 17 cu, le 4 4+ > Oas, gi 18 M 28 Oas, gi Oas 19 F 33 Gi Oas, gi 20 F 26 Oas, gi, dys, le, cu, as Oas, gi 21 F 24 Oas, gi, dys, le, gu, cu Oas, gi 22 M 22 Oas, gi, cu, le, dys, gu, as Oas 23 F 28 Oas, gi, dys, gu 2 4+ > Oas, gi 24 F 32 Oas, gi, rc Oas 25 M 28 Oas, gi, le, dys, as Oas 26 F 20 Oas, gi, cu, le, dys 3 4+ > Oas, gi Vo, le* MPD, Minimum provoking dose; oas, oral allergy syndrome; cu, contact urticaria; dys, dyspnea; ls, lip swelling; rc, rhinoconjunctivitis; gi, gastrointestinal symptoms; le, laryngeal edema, gu, generalized urticaria, vo, vomiting; as, anaphylactic shock. *One patient had hoarseness and difficulty in swallowing and started vomiting 1 hour after ingestion of 30 mg of peanut protein. reactions were calculated by using Kruskal-Wallis 1-way analysis (P <.05 was regarded as significant), after which Mann-Whitney U tests were performed for pairwise intergroup comparisons (with significance levels adjusted to the Bonferroni correction). RESULTS Patient characteristics Twenty-six patients (20 female and 6 male patients) with a mean age of 25 years (range, years) were included in the study. Patient characteristics are summarized in Table I. All patients had a positive SPT response to peanut, and peanut-specific IgE levels were 0.7 ku/l or greater in all but 2 patients (patients 2 and 15). Medical history revealed itching of the oral cavity (Mueller grade 1) as the most severe reaction in 7 patients and laryngeal edema and gastrointestinal symptoms in 8 patients (Mueller grade 2). Eight patients had dyspnea, which was accompanied by generalized urticaria in 4 patients (Mueller grade 3). Three patients (patients 20, 22, and 26) had cardiovascular symptoms with decrease in blood pressure, resulting in faintness (Mueller grade 4). The age of onset of peanut-related symptoms (and diagnosis of peanut allergy) was at 8 years or less in 18 patients. Eight patients developed peanut allergy later in life. Eighteen (69%) patients had several allergic reactions caused by accidental ingestion of peanut in the previous year. In 8 patients the last peanut-related reactions had occurred more than 1 year ago. Thirteen patients had asthma (well controlled), 21 had atopic dermatitis, and 23 had symptoms of seasonal rhinoconjunctivitis. Symptoms observed during DBPCFCs All 26 patients included had a positive challenge result. Nine patients did not react up to a dose of 30 mg of peanut protein and were enrolled in a second challenge with higher doses (with an interval of at least 2 weeks). The first symptoms reported during all challenges were symptoms of itching or a feeling of swelling in the oral cavity, sometimes extending to the ears (oral allergy syndrome). This started within 5 minutes after ingestion of peanut and recurred in all but 4 patients after subsequent higher doses (2 times in 2 patients and 3 times in 20 patients). The remaining 4 patients had oral symptoms only after the last dose of 1 g of peanut protein. Fourteen patients also had gastrointestinal symptoms. Gastrointestinal symptoms started within 20 to 30 minutes after ingestion of peanut and mostly resolved quickly. These symptoms nevertheless resulted in discontinuation of the challenge in 7 patients. Two patients had difficulty in swallowing, hoarseness, or both and facial flushing several minutes (patient 11) to 1 hour after ingestion of peanut (patient 26). Of the 3 patients with a history of cardiovascular symptoms, 2 had gastroin-

4 918 Wensing et al J ALLERGY CLIN IMMUNOL DECEMBER 2002 FIG 1. Cumulative number of responding patients (in percentages) according to peanut dose (in milligrams of peanut protein; n = 26). testinal symptoms. The third patient only reported oral symptoms. Breathing difficulty or dyspnea was not reported during the challenges, and no significant decrease in FEV 1 was observed during the challenges. In 6 patients oral antihistamines were required. One patient (patient 26) received intravenous antihistamines and prednisolone as treatment for severe laryngeal edema. This was the only patient who needed to stay overnight for observation of possible protracted or late reactions. No late or delayed reactions ( 2 hours) occurred, apart from the reaction in one patient, who started vomiting 2.5 hours after the last ingestion of peanut. During 35 performed peanut challenges, a total of 216 placebo doses were administered to 26 patients. Only 8 mild placebo responses were observed in 6 (4%) patients. These placebo responses consisted of mild itching of the lips or subjective gastrointestinal symptoms. Because of the majority of negative placebo responses and the repetitive occurrence of symptoms to the active doses, this did not interfere with the possibility of establishing threshold doses in these patients. Threshold doses Responses during the challenges were subdivided into subjective (oral symptoms, abdominal pain, and nausea) and objective (lip swelling, vomiting, and hoarseness) categories. Threshold doses for subjective responses varied from 100 µg up to 1 g of peanut protein (Table I). Extrapolation of the achieved dose-response curve (Fig 1) shows that 50% (95% CI, 30%-70%) of a population with peanut allergy will have (subjective) allergic reactions after a dose of 3 mg of peanut protein and that no patient (95% CI, 0%-13%) will react to the lowest dose of 30 µg of peanut protein. Threshold doses inducing objective symptoms were only established in 6 patients after 10 mg (n = 1) and 30 mg (n = 5) of peanut protein. Correlation between severity of symptoms and individual threshold doses In Fig 2 individual threshold doses are correlated to the severity of peanut-related symptoms (as documented by medical history and DBPCFCs). Of 11 patients with a history of severe peanut-related reactions (Mueller grade 3 or 4), 8 patients had moderate-to-severe reactions (grade 2 or 3) during the challenges. Of 8 patients with a history of moderate reactions to peanut (grade 2), 5 had moderate reactions during the challenge, and 3 had only mild reactions (grade 1). All patients with a history of mild peanut-related symptoms had mild reactions during the challenge. Patients with a history of severe peanutrelated reactions (grade 3 or 4) had significantly lower threshold doses than patients with mild reactions (P =.0001). When the severity of reactions as observed during DBPCFCs was scored and analyzed, patients with moderate-to-severe reactions had significantly lower threshold doses compared with those of patients with mild reactions (P =.027). DISCUSSION It is important to determine threshold doses of food allergens that are able to induce symptoms in patients with food allergy. More information about these threshold levels of different food allergens will enable us to assess risks for patients with food allergy to experience allergic

5 J ALLERGY CLIN IMMUNOL VOLUME 110, NUMBER 6 Wensing et al 919 FIG 2. Differences in threshold levels with respect to severity of peanut-related symptoms (as determined by medical history; n = 26). Filled dots, Mild reactions (n = 13) during DBPCFCs (Mueller grade 1); open dots, moderate (n = 12) and severe (n = 1) reactions during DBPCFCs (Mueller grade 2 or 3). reactions after ingestion of contaminated consumer products. 22,23 Several case reports on accidental ingestion of hidden allergens yield some information about minute quantities of food allergens that induced allergic symptoms. 11,12,24 In several studies DBPCFCs have been performed, and threshold doses have been reported, 13,25,26 but these did not aim at establishing the distribution of threshold doses in a population with food allergy. The main objective of this study was to determine individual threshold doses for eliciting allergic reactions in a population with peanut allergy. The lowest threshold doses found in this study equal those found by Hourihane et al. 13 Fig 1 shows the dose-response relationship of 26 patients with peanut allergy. From Fig 1, one can conclude that 50% of a population with peanut allergy (95% CI, 30%-70%) will have an allergic reaction after ingestion of 3 mg of peanut protein (about 1 50 peanut). As described in earlier case reports, this amount of peanut can be hidden in a mouthful of several consumer products. 5,11 The NOAEL in our study is 30 µg of peanut protein. Within the constraints of statistical uncertainty, the NOAEL can be used as an indication for safe levels of the potential presence of allergens. It is important to investigate a randomly recruited group of patients to obtain threshold levels representative for the general population with peanut allergy. Although in most studies conducting DBPCFCs patients with a history of anaphylactic shock were excluded for ethical reasons, it is important to determine threshold doses, especially in these patients, because they are most at risk of developing life-threatening reactions. At the same time, it is hard to avoid a selection bias when including patients for DBPCFCs because patients with more severe symptoms will be less prepared to participate in such a study. In this study 11 patients with severe symptoms were included, of whom 3 had a history of anaphylactic shock. Of course, ultimate safety precautions were applied. By starting at low doses and avoiding large increments, severe reactions were prevented. It also allowed discontinuation of the challenge before the development of more severe reactions. In fact, all peanut-related reactions, as observed during the challenges, were milder than those revealed by case history. DBPCFCs in 3 patients with a history of anaphylactic shock resulted in mild symptoms strictly localized to the oral cavity in 1 patient, with additional gastrointestinal symptoms in 2 others. None of these reactions required treatment. These 3 patients were not the most sensitive (as determined by threshold doses of 300 µg and 1 mg of peanut protein). Overall, however, patients with a history of severe symptoms (Mueller grade 3 or 4) were significantly more sensitive to peanut compared with patients with mild symptoms (Mueller grade 1). These findings point out the importance of revealing threshold doses of patients with a history of severe peanut-related symptoms to avoid underestimation of the minimum provoking doses. Of course, DBPCFCs constitute only an approximation of exposure in everyday life. Patients have to eat a lot of meals in a formal hospital setting and have to register carefully all kinds of symptoms. This can have 2 results. An underestimation of threshold doses can occur because patients feel comfortable and safe and are not panicking. On the other hand, thresholds can be overestimated because very mild allergic symptoms would remain unno-

6 920 Wensing et al J ALLERGY CLIN IMMUNOL DECEMBER 2002 ticed in everyday life. Moreover, we did not investigate the reproducibility or intraindividual variability of threshold doses. Because of various environmental or personal factors, the individual threshold dose probably varies over time. 27 Overall, these sources of variability will be leveled out when investigating the distribution of threshold doses among a group of patients with peanut allergy. As described in the literature, the age at onset of peanut-related symptoms mostly is at child age. 7 Also, in this study most patients reported peanut-related symptoms from child age (n = 18). However, 8 patients who used to eat peanuts without symptoms when they were young had peanut allergy at higher ages (range, years). Whether the route of sensitization in these patients is different is not known. It is possible that crossreactive IgE, initially directed to pollen, plays a role. All patients except one were sensitized to pollen (not shown). It is known that peanut allergens implicated in severe peanut allergy are not the typical pollen-related allergens but rather seed storage proteins The high prevalence of pollen sensitization might point toward a bias in patient selection in favor of those with less severe symptoms linked to pollen cross-reactive allergens. The nature of potential cross-reactive allergens other than profilin and cross-reactive carbohydrate determinants is still under investigation. In summary, threshold doses for eliciting allergic reactions in 26 patients with peanut allergy appeared to vary from doses between 30 and 100 µg up to a dose between 300 and 1000 mg of peanut protein. Furthermore, patients with potentially life-threatening reactions appeared to be more sensitive than patients with milder symptoms, stressing even more the need for a careful declaration of peanut content. We thank all the patients who participated in this study, W. J. Koers for his help in patient recruitment, Frank Rijnja and Julie Nordlee for preparing and arranging the ingredients for the test meals, and Ronald van Ree for careful reading of the manuscript. REFERENCES 1. Bruijnzeel-Koomen C, Ortolani C, Aas K, Bindslev-Jensen C, Björkstén B, Moneret-Vautrin D, et al. Adverse reactions to food. Allergy 1995;50: Emmett SE, Angus FJ, Fry JS, Lee PN. Perceived prevalence of peanut allergy in Great Britain and its association with other atopic conditions and with peanut allergy in other household members. Allergy 1999;54: Sicherer SH, Munoz-Furlong A, Burks AW, Sampson HA. Prevalence of peanut and tree nut allergy in the US determined by a random digit dial telephone survey. J Allergy Clin Immunol 1999;103: Bock SA, Lee WY, Remigio LK, May CD. Studies of hypersensitivity reactions to foods in infants and children. J Allergy Clin Immunol 1978;62: Sampson HA, Mendelson L, Rosen JP. Fatal and near-fatal anaphylactic reactions to food in children and adolescents. N Engl J Med 1992; 327: Hourihane JO, Kilburn SA, Dean P, Warner JO. Clinical characteristics of peanut allergy. Clin Exp Allergy 1997;27: Bock SA, Atkins FM. The natural history of peanut allergy. J Allergy Clin Immunol 1989;83: Hourihane JO, Roberts SA, Warner JO. Resolution of peanut allergy: case-control study. BMJ 1998;316: Spergel JM, Beausoleil JL, Pawlowski NA. Resolution of childhood peanut allergy. Ann Allergy Asthma Immunol 2000;85: Skolnick HS, Conover-Walker MK, Barnes Koerner C, Sampson HA, Burks AW, Wood RA. The natural history of peanut allergy. J Allergy Clin Immunol 2001;107: McKenna C, Klontz KC. Systemic allergic reaction following ingestion of undeclared peanut flour in a peanut-sensitive woman. Ann Allergy Asthma Immunol 1997;79: Kemp SF, Lockey RF. Peanut from food cross contamination. JAMA 1996;275: Hourihane JO, Kilburn SA, Nordlee JA, Hefle SL, Taylor SL, Warner JO. An evaluation of the sensitivity of subjects with peanut allergy to very low doses of peanut protein: A randomized, double-blind, placebo-controlled food challenge study. J Allergy Clin Immunol 1997;100: Mueller HL. Diagnosis and treatment of insect sensitivity. J Asthma Res 1966;3: Dreborg S, Frew A, editor. Allergen standardisation and skin tests. EAACI position paper. Allergy 1993;48: Cunniffe P, editor. Official methods of analysis of AOAC international. 16th ed. Arlington (VA): AOAC International; Chapter 32, p. 23 and chapter 39, p Burks AW, Williams LW, Helm RM, Connaughton C, Cockrell G, O Brien TJ. Identification of a major peanut allergen, Ara h I, in patients with atopic dermatitis and positive peanut challenges. J Allergy Clin Immunol 1991;88: Burks AW, Williams LW, Connaughton C, Cockrell G, O Brien T, Helm RM. Identification and characterization of a second major peanut allergen, Ara h II, with use of the sera of patients with atopic dermatitis and positive peanut challenge. J Allergy Clin Immunol 1992;90: Rabjohn P, Helm EM, Stanley JS, West CM, Sampson HA, Burks AW, et al. Molecular cloning and epitope analysis of the peanut allergens Ara h 3. J Clin Invest 1999;103: Ortolani C, Ispano M, Pastorello E, Bigi A, Ansaloni R. The oral allergy syndrome. Ann Allergy 1988;61(suppl): Sampson HA. Immunologically mediated food allergy: the importance of food challenge procedures. Ann Allergy 1988;60: Madsen C. Where are we in risk assessment of food allergens? The regulatory view. Allergy 2001;56: Bousquet J, Björkstén B, Bruijnzeel-Koomen CAFM, Huggett A, Ortolani C, Warner JO, et al. Scientific criteria and the selection of allergenic foods for product labelling. Allergy 1998;53: Gern JE, Yang E, Evrard HM, Sampson HA. Allergic reactions to milkcontaminated non-dairy products. N Engl J Med 1991;324: Hansen TK, Bindslev-Jensen C. Codfish allergy in adults. Identification and diagnosis. Allergy 1992;47: Sicherer SH, Morrow EH, Sampson HA. Dose-response in double-blind, placebo-controlled oral food challenges in children with atopic dermatitis. J Allergy Clin Immunol 2000;105: Hourihane JO. The threshold concept in food safety and its applicability to food allergy. Allergy 2001;56:86-90.