- Top of page
- Material and methods
The aim of the study was to develop and evaluate different methods of double-blind, placebo-controlled food challenge (DBPCFC) with apple. Three different DBPCFC models were evaluated: fresh apple juice, freshly grated apple, and freeze-dried apple powder. All challenges were performed outside the pollen season and took place from 1997 to 1999. The freeze-dried apple material was characterized by means of leukocyte histamine release (HR), skin prick test (SPT), and immunoblotting experiments. The study population consisted of birch pollen-allergic patients with a history of rhinitis in the birch-pollen season and positive specific IgE to birch. For comparison of the DBPCFC models, 65 patients with a positive open oral challenge with apple were selected. In the characterization of the freeze-dried apple material, 46 birch pollen-allergic patients were included. The IgE reactivity to apple was evaluated by measurement of specific IgE, HR, and SPT. Golden Delicious apples were used in all experiments. The results of this study showed that it was possible to perform DBPCFC with apple in birch pollen-allergic individuals. The model with freshly squeezed apple juice had a low sensitivity and displayed a high frequency of reactions to placebo, probably due to the ingredients used for blinding. The sensitivity of the models with freshly grated apple and freeze-dried apple powder was 0.74/0.60. An increase in sensitivity is desirable. The freeze-dried apple powder proved to be useful for SPT, HR, and oral challenges, but further investigation of the stability and the allergenic profile of the material is needed.
The cross-reactivity between birch pollen and various fruits, nuts, and vegetables has been intensively investigated over the last 25 years (1–6). Isolation, characterization, and cloning of major allergens from different plant foods, such as Mal d 1 from apple, have revealed a high degree of sequential homology with Bet v 1 (7–9). Different serologic and clinical aspects of the cross-reactions between birch pollen and apple have been investigated in numerous studies, but only a subgroup have included food challenges (5, 6, 10–19); of these, even fewer have evaluated patient groups of 10 or more (5, 10, 12, 17–19). Food challenge with apple may be difficult to perform, since apple allergens seem very susceptible to all types of processing of the fruit (14, 20). Moreover, the allergenicity differs between apple strains and is influenced by ripening and storage of the fruits (14, 21, 22). As opposed to classical food allergy, such as that to milk, egg, or fish, where the final diagnosis is based on the outcome of double-blind, placebo-controlled food challenges (DBPCFC) (23), open challenges are generally considered confirmatory for allergenic cross-reactions between pollen and foods. A history of allergic rhinitis in the birch-pollen season combined with the detection of specific IgE to birch increases the credibility of suspected reactions to various plant foods, but there is still a need for confirmation of the diagnosis of food allergy in some cases. Both in the investigation of severe food allergic reactions and for scientific use, a consistent DBPCFC model would be valuable.
The aim of the study was to develop and evaluate different methods of DBPCFC with apple.
- Top of page
- Material and methods
In most studies that have included challenges with apple, the challenges have been performed openly, partly because apple allergens have been considered too unstable to process, and partly because a valid DBPCFC model was not available (5, 10, 12, 17–19). The issue of diagnostic accuracy has therefore not been addressed in these studies. A DBPCFC model for fresh fruit including apple has been published, but data on the efficacy of the model regarding allergy to apple were not specified (29). In a study on double-blind challenges with hazelnuts in 86 subjects, we found a sensitivity of 0.78 (30).
The challenge models evaluated in this study seem to differ both in sensitivity and number of placebo reactions. The freshly grated apple (model II) had the highest sensitivity (0.74) and did not lead to unspecific reactions. This model has been used in a study on seasonal variation in allergy to apple (Skamstrup Hansen et al., submitted paper). In all, 52 blinded challenges were performed. Two reactions to placebo were recorded, one in a patient with no reaction to either active blinded challenge or open challenge with fresh apple. Model III (freeze-dried apple) had a comparable sensitivity (0.6) but significantly more placebo reactions, while the results of model I (freshly squeezed apple juice) with 17 reactions to active and 10 to placebo were indistinguishable from mere chance.
With a sensitivity of 0.6/0.74, the methods cannot detect the least sensitive patients. A larger amount or concentration of apple could probably increase the sensitivity, but apple has a characteristic smell and flavor and is therefore difficult to blind. We have no data on the stability of apple allergens in combination with the other ingredients of the active challenges in either of the models.
The selection of patients for this comparative study could pose a bias. The patients in models I and III were selected by history of allergy to birch pollen, and, depending on the result of the open challenge, included in the comparison of challenge models. Patients in model II were selected by history of allergy to birch pollen and concomitant food allergy to apples. They might therefore be more focused on this aspect of their allergic disease, more aware of symptoms, and perhaps also more clinically sensitive, even though they did not respond differently to fresh apple in the open challenge judged by the symptom scores. Moreover, the background data on open challenges indicate that the patients in group III were less clinically responsive than groups I and II. Some of the patients in this group had received immunotherapy before this investigation. The importance of this selection bias is unknown.
The symptoms experienced by the patients in all three models were very mild, and most patients had OAS (oral allergy syndrome) (31) without any objective signs. A problem in all models could be the fact that the patients received more than one active challenge a day in the form of more active doses or an active blinded challenge followed by an open challenge with fresh apple. In two patients included in model II, the open challenge was repeated on a separate day because of conflicting results. One of the patients had a mild reaction to the active blinded challenge but no reaction to the following open challenge. The second patient had no reaction upon either the blinded challenge or the open challenge in spite of a history of clinical symptoms and a positive open challenge performed earlier. Both challenges were positive. Development of tachyphylaxis is obviously a problem in interpretation of challenge results, determination of threshold doses, etc.
In model I, different regimens regarding doses and concentration of apple juice were tested (data not shown). Although the patient groups were too small for statistical analysis, increasing concentration seemed to be more important than increasing the amount – an observation that also could be ascribed to the development of tachyphylaxis during the challenge procedure. The liquid challenge of model I was included since it has proven to be useful for different food allergens (32, 33). The addition of concentrated blackcurrant juice to disguise the taste of apple gave the mixture a very strong taste and produced a sensation in the mouth that could probably mimic OAS. Moreover, the use of commercial juices in a very high concentration could pose a problem, since it has been found that heat-stable IgE-binding lipid transfer proteins are present in commercial fruit juices (34). An Italian study showed that 28% of patients with allergy to apple and birch pollen were sensitized to a stable 9-kDa allergen in apple, identified as a lipid transfer protein that does not cross-react with Bet v 1 (19). These allergens do not seem to be equally important in northern Europe, where birch pollen is predominant (35). Several groups have shown that the clinically most important allergen in apples in northern Europe is Mal d 1, mainly due to the high degree of sequential homology between Bet v 1 and Mal d 1 (7–9, 14). Only one patient complained of occasional symptoms to apple juice without specifying whether the problem was fresh juice or commercial juice. This patient did not react to placebo. There was a significant relation between the onset of seasonal rhinitis and plant food allergy in this study (data not shown), indicating that the food allergy was related to birch pollen. Furthermore, the commercial apple juice in normal 1:5 dilution has been used for SPT in 21 birch- and apple-allergic patients (unpublished data) and did not lead to skin wheals of ≥7 mm2.
Freshly grated apple, as used in model II, gave an acceptable sensitivity, but it was difficult to perform a precise dose titration; furthermore, grated fruit is very susceptible to oxidative degradation.
Lyophilization of foods is acceptable for DBPCFC and has been used for different allergen sources (36), but the process may change the allergenic profile of some foods (37). Use of freeze-dried material for challenges has some practical advantages over fresh apples. It allows a dose titration, the freeze-dried material can be stored, and the same batch of freeze-dried apple can be used for some time without the change of allergenicity seen during ripening and storage of apples (21, 22). Evaluated on six different drying sessions performed over 6 weeks, the described preparation of apples by direct freezing in liquid nitrogen before lyophilization seems to be practicable and reproducible. Since the process avoids addition of toxic compounds, the freeze-dried material can be used for both in vivo and in vitro tests. The freeze-dried material was able to elicit specific skin reactions and HR, as well as clinical symptoms, when used for oral challenges. Evaluated by HR and SPT, extracts of the freeze-dried apple were stable for at least 24 h.
Immunoblotting results indicated that Mal d 1 was preserved during the lyophilization and extraction processes.
In conclusion, the results of this study showed that it was possible to perform DBPCFC with apple. Three different models were compared, fresh apple juice, grated apple, and freeze-dried apple, respectively. The first model had an unacceptably low sensitivity combined with a high rate of reactions to placebo. The sensitivity of the last two challenge models was comparable (0.74/0.60), and unspecific reactions to placebo were seen in 0/19 and 4/25 patients. It is highly desirable that the sensitivity of the DBPCFC with apple is increased further. Lyophilization of apples produces material useful for SPT, HR, and oral challenges, but an investigation of the long-time stability and the allergenic profile of the freeze-dried material is needed.