Diagnosis of Food Allergy by RAST

Diagnosis of Food Allergy by RAST Donald R. Hoffman, Ph.D. Objective The purpose of this paper is to relate experience with RAST in the diagnosis of food allergy mediated by specific IgE antibodies. The diagnosis of food allergy is an extremely difficult problem for the practitioner. The problem is further complicated by common usage in which any adverse effect related, no matter how tenuously, to the ingestion of a food is called "food allergy." Even among those reactions which can be clearly or even unequivocally related to the ingestion of specific foods, many are not allergic or even immunologic. For example, if I eat four dozen raw oysters, I develop a stomach ache for an hour or two; my wife does also. This is not allergy. This is due to some weak toxins in the algae the oysters eat. Other types of adverse reactions to foods involve either biochemical mechanisms like the disaccharidase deficiency-related milk intolerance, or pharmacologic mechanisms like strawberry-induced urticaria. For the purposes of this discussion the term food allergy will be reserved for those reactions in which the role of specific IgE antibodies has been or can be demonstrated. All other adverse reactions to foods will be classified as food intolerances. IgE-mediated reactions are known to include the immediate skin test or wheal and flare, anaphylaxis, allergic asthma, Donald R. Hoffman, Ph.D., Associate Professor of Pathology, Department of Pathology and Laboratory Medicine, East Carolina University School of Medicine, Greenville, N.C. 123

- 124 D. R. HOFFMAN allergic rhinitis-conjunctivitis, some urticaria and some angioedema; there is an unknown role but important association of IgE with atopic eczema and some role of IgE in gastrointestinal disease, as will be shown later. There is no positive evidence for a role of IgE in headache, tension-fatigue syndrome or urinary tract symptoms. In studies with foods we have never found a positive IgE antibody associated with any of the above three syndromes. RAST is the method of choice for studying IgE antibodies since it does not exhibit interference from other classes of antibodies like short-term sensitizing IgG, and also since prick skin testing with foods presents the risk of inducing systemic reactions in extremely sensitive individuals. To establish whether RAST could be used in food allergy studies, a group of patients were collected with unequivocal histories of food allergy. In order to qualify for this study the patients had to be able to identify the food suspected; ingestion of the food had to produce the reaction every time it was consumed, and without consumption of an excessively large quantity; and the reaction had to have occurred within the previous six months. Some of the patients with non-life-threatening symptoms were placed on a strict elimination-challenge-elimination-challenge regimen to verify the clinical histories. Prick skin testing was performed until two severe systemic reactions caused us to discontinue this phase of the study. This group, collected and reported on by Hoffman and Haddad [1], showed reactions to 14 common foods. Age-matched control sera were collected and tested by RAST and used to determine the normal range for each allergen. Some of the results from this study are given in Table 1. Good allergen systems in which allergen is more than a contaminant of the extract, such as codfish, peanut and egg, gave excellent correlations for those patients with the most severe symptoms and less good correlations for those with milder symptoms. Correlations with less defined allergens like orange and chocolate were not as good. Table 2 shows the correlation of various symptoms with positive RAST for patients with unequivocal histories. This again shows the higher correlation for the more severe symptoms. It should also be noted that cases of allergic rhinitis only, without systemic

mber DIAGNOSIS 0/20 a/is 3/20 OF Skin 8/8 Control11/11 IS/IS 6/9 6/10 4/8 7/8 9/14 6/7 7/9 1/5 3/3 Eczema Symptoms a/i 0/2 2/8 0/4 5/5 4/5 Other FOOD Test 11/12 5/7 5/5 8/9 3/3 7/12ALLERGY Urticaria, 125 Anf(ioedema, Anaphylaxis. Tested symptoms, and tension-fatigue syndrome are always RAST negative to the suspected food. Studies of a further group of patients have also demonstrated that as the history becomes more vague or the incidence of reaction sporadic, the chance of finding IgE antibodies decreases. Finding specific IgE antibodies in cases of chronic urticaria is extremely rare. In order to determine the potential value and pitfalls of using RAST to diagnose food allergy, several representative systems were investigated in detail. Results with milk RAST are given in Table 3 [2]. The cases of anaphylaxis were all RAST positive; verified cases of urticaria and asthma also gave high correlations between RAST and history. In the studies of Table 2. Correlation of RAST in Symptom on Food Ingestion in a Group of Patients with Unequivocal Histories Symptom No. of Patients Anaphylaxis 10 Bronchospasm 23 Angioedema 13 Atopic eczema 30 Urticaria 26 Diarrhea * 12 Acute gastrointestinal upset 6 Allergic rhinitis only 25 Tension-fatigue syndrome 15 % RAST Positive 100 *Cases of disaccharidase deficiency, infection, etc. are not included. 96 92 87 62 92 33 a a

126 O. R. HOFFMAN Lactalbumin 30 077% 49 36 I11 0112 10 189 12 534 68 3 2 36 Lactoglobulin 2I 303 89 11 50% 85% 010% TestedTable 3. 59% RAST Casein 18 7 0 NO Total 53 71 to44% 10% Milk Antigens Challenge Milk No disaccharidase allergytested deficiency Number RAST Positive to: intractable diarrhea in infants, a group of five with disaccharidase deficiency proved by lactose challenge were all RAST negative; 11 of the 12 in the non-enzyme-deficient group were RAST positive. In all of these patients, diagnoses of infection or gastrointestinal disease of known etiology (e.g. celiac disease) had been excluded. The nonatopic control group was all RAST negative; however, 11 out of 89 children with severe respiratory allergy but no food allergy were RAST positive. About half of children with atopic eczema were RAST positive to milk [3]. Positive RAST to milk is:found in an eightfold higher incidence (88%) in those patients who clearly have atopic symptoms following ingestion than in an atopic control group with no known milk allergy. RAST results on sera from patients with egg white allergy are given in Table 4. Only two of 27 patients with unequivocal histories gave a negative RAST when tested with ovalbumin or ovomucoid, the two known allergens of egg white. No additional patients were positive to egg yolk, but one patient who reacted only to fried eggs was only positive to denatured

DIAGNOSIS Ovalbumin+ 40 13 67 OFOvomucoid+ 2611 910 3 1Total FOOD 62 3 1 Number Poritive3 ALLERGY 21 12 10 63 127 Allergy Symptom ovomucoid. RAST results for control groups are given in Table 5. Those in the nonatopic group over 24 months of age were all RAST negative. Some RAST-positive sera were found in nonatopic infants about the time they were first introduced to egg; this appears to be a transient phenomenon and has also been observed by Johansson et al in Sweden [4]. Occasionally these infants will express symptoms, but these will disappear along with the IgE antibodies in a matter of months. About half of the children with atopic eczema had IgE antibodies against egg and 12% of the children with severe respiratory allergy but no food allergy were RAST positive to egg. This contrasts with 93% positive RAST in the group expressing symptoms upon egg ingestion. RASTs with milk and egg show a reasonable level of discrimination between symptomatic groups, with the exception of patients with atopic eczema. Cereal grains present much more of a problem [5}. Some selected results of RAST with Table 5. RAST to Egg White in Patients with No Overt Symptoms on Egg Ingestion and in Patients Symptomatic After Egg Ingestion Group 33 53 93 012 RAST 25 3(±) 19 120 % Positive Number98 36 27 219

128 D. R. HOFFMAN cereal grains and the related-system grass are shown in Table 6. The patients symptomatic by either inhalation (baker's asthma) or ingestion were RAST positive to the offending cereal among other cereals - but so were a large number, about 56%, of the atopic control group consisting of patients with respiratory allergy but no histories of food allergy. Those patients consumed cereals regularly but with no adverse effects. An extensive series of studies was initiated to determine if the sera from the symptomatic patients could be distinguished from those from asymptomatic patients. Cross-reactivity among the cereals and grass pollen was investigated by RAST inhibition and RAST absorption. Antigen specificity was investigated using RAST with various fractions of wheat, and the relationship of IgE antibodies against grass pollen to those against cereals was studied. None of these methods provided any differentiation of the symptomatic and asymptomatic groups, with all combinations of cross-reactivities and specificities being found in each group. Recently Aas [6] has reported similar results with both RAST and skin test using cereals; in addition, he notes a possibly similar situation with soybeans. RAST with cereals is positive in those patients with adverse reactions, but - = Negative; Wheat 222 Rice 548Asthma, II2 3I3 ±-2 467 ±I 323 I4 610 2 8OatCorn 10 2 56 73 I Angioedema, Asthma, Symptoms ± = None3 questionable; wheat wheat None 53 wheat Table corn Rye with inhalation corn ingestion Grass Cereals 26. ± I 2I RAST = p>95% to positive; Cereal Grains 2-10 = increasingly positive. RAST Score *

DIAGNOSIS OF FOOD ALLERGY 129 also in a high percentage of atopic patients who tolerate cereals by ingestion. One possible explanation could derive from studies of patients with baker's asthma who express symptoms on inhalation of wheat flour but can eat cake or bread without incident. The patients with false-positive RAST may express respiratory symptoms on inhalation challenge, but this hypothesis is as yet unproved. In our laboratory we are currently carrying out a detailed study of shrimp allergy. As part of this study we examined our bank of sera for all those known to be RAST positive for shrimp. The distribution of clinical histories is shown in Table 7. The laboratory studies used serum from one of the anaphylactic reactors who had his last anaphylactic reaction sitting beside a physician in a restaurant. This patient is sensitive to shrimp by contact; he is a grocery clerk and develops urticaria when he stacks frozen shrimp boxes. He is also sensitive to crab and lobster. When investigated by twodimensional crossed immunoelectrophoresis, raw shrimp extract contains over 50 discernible proteins, several of which have allergenic activity. When the extract is prepared from cooked shrimp, many fewer proteins are visible and they are less discrete because of denaturation and aggregation. At present we have isolated, in relatively pure form, two shrimp allergens, both of which are heat-stable and found in extract of cooked shrimp as well as raw shrimp. Allergenic activity of one of these is comparable to unfractionated extract when measured by RAST; the second is weaker. These materials are not yet characterized. Table 7. Clinical Histories of 61 Patients RAST Positive to Shrimp Symptom No. Patients Anaphylaxis Angioedema Urticaria Respira tory distress Wiskott-Aldrich syndrome Eosinophilic granuloma Atopic eczema Unspecified, but shrimp RAST requested 4 5 8 1 1 1 34 9

130 D. R. HOFFMAN In order to show the use of RAST in diagnosis of food allergy when the history is unclear but specific, I will describe an unusual case. A 22-year-old woman with minimal symptoms of respiratory allergy developed sudden-onset hives, flushing and angioedema of the face and eyelids within minutes after eating at fast-food restaurants, particularly those selling tacos and certain brands of hamburgers. RAST was performed to a panel of 15 foods. Most common allergens, including nuts, milk, egg, wheat, cod and shrimp, were negative. Corn was positive with a RAST score of 1. Peanut gave a RAST score of 4 (31 % binding) and soybean gave a 2. Inquiries at the fast-food restaurants confirmed that tacos were cooked in peanut oil and that some hamburger chefs used peanut oil as a grill lubricant. On further questioning the patient indicated that she did not eat peanuts or peanut products because of a reaction that occurred at about age 5. RAST testing can provide a useful, valuable and safe method for evaluating food allergy histories. With allergens like milk and egg, the discrimination between symptomatic and nonsymptomatic is about eightfold. RAST results with other food allergens like codfish, shrimp, nuts and peanut are even better. However, interpretation of positive RAST to cereal grains and soybean is extremely difficult due to the high incidence of IgE antibodies against these allergens in the atopic population not expressing ingestant sensitivity to these allergens. RAST can be a useful adjunct to the diagnosis of food allergy in many cases without presenting risk to the patient. References 1. Hoffman, D.R. and Haddad, Z.H.: Diagnosis of IgE mediated immediate hypersensitivity reactions to food antigens by radioimmunoassay. J. Allergy Clin. ImmunoL 54:165, 1974. 2. Hoffman, D.R.: Food allergy in children: RAST studies with milk and egg. In Evans, R. III (ed.): Advances in Diagnosis of Allergy: RAST. Miami:Symposia Specialists, 1975, p. 165. 3. Hoffman, D.R., Yamamoto, F.Y., Geller, B.D. and Haddad, Z.H.: Specific IgE antibodies in atopic eczema. J. Allergy Clill. ImmunoL 55:526,1975. 4. Johansson, S.G.O., Bennich, H., Berg, T. and Foucard, T.: The clinical significance of IgE antibody in serum as "determined by RAST. In Goodfriend, L., Sehon, A.H. and Orange, R.P. (eds.): Mechanisms in

DIAGNOSIS OF FOOD ALLERGY 131 Allergy Reagin-Mediated Hypesensitivity. New York:Marcel Dekker, 1973. 5. Hoffman, D.R.: The specificities of human IgE antibodies combining with cereal grains. Immunochemistry 12: 535, 1975. 6. Aas, K.: The diagnosis of hypersensitivity to ingested foods. Clin. Allergy 8:39, 1978.