Emily Kallen

Background food allergy
Food allergy has a major global health impact. Among adults, the prevalence based on self-reported symptoms varies between 2% and 37% across Europe, whereas the true prevalence, as confirmed by an oral food challenge, ranges between 0.2% and 4.1%.

A food allergy is an abnormal immune response to specific proteins in food. Before someone becomes allergic, they must first become sensitized. This means that the immune system recognizes a specific food protein and produces specific antibodies against it, known as IgE antibodies. However, not everyone who is sensitized will actually develop a food allergy.

In principle, any food can cause an allergic reaction. Previous research has mainly focused on the eight most common allergenic foods: cow’s milk, hen’s eggs, wheat, soy, peanut, tree nut, fish and shellfish. However, little is known about the frequency and severity of allergies to other foods. Our research has demonstrated that patients can be allergic to as many as 200 different foods in daily clinical practice, although the majority of reactions are caused by approximately 30 foods (Chapter 2). Notably, allergic reactions to fruit are common, with apple and kiwi being among the most frequently reported, followed by tree nut and peanut.

The symptoms of a food allergy can range from mild reactions, such as itching in the mouth, to severe and potentially life-threatening reactions, including drop of blood pressure and loss of consciousness. Our study showed that seeds, including sesame, sunflower and pine nut, most frequently cause severe symptoms. Additionally, exotic fruit such as lychee and papaya were regularly associated with severe reactions (Chapter 2).

These findings demonstrate that a much broader range of foods than has been investigated to date can cause allergic reactions and that some of these foods are frequently associated with severe symptoms.

Certain specific allergenic proteins are important predictors of the presence and severity of food allergy
When there is a suspicion of food allergy based on the clinical history, it is common practice to assess whether patients have specific IgE (sIgE) to the foods that cause their symptoms (sensitization). This can be measured by skin prick tests and blood tests.

Foods contain allergenic proteins that often share a structural similarity because they belong to the same protein families. IgE antibodies directed to certain protein families appear to be better predictors of food allergy than those directed to whole food extracts.

For example, those directed to so-called 2S albumins, storage proteins abundantly present in legumes, nuts and seeds, may be more accurate in confirming or excluding allergy than standard extract-based testing. In adults with suspected cashew allergy, we demonstrated that sIgE to Ana o 3 (the 2S albumin protein in cashew) correctly classified 72% of patients as allergic or non-allergic (Chapter 6). In contrast, sIgE against cashew extract correctly classified only 49% of patients.

In light of this favorable outcome, we also explored this approach in adults with suspected almond allergy (Chapter 5). Surprisingly, 2S albumin of almond was not found to play a similar role as reported for peanut, several tree nuts and sesame seed. Consequently, this protein cannot be used as a predictor of almond allergy. Also none of the other almond proteins and extract tested proved to be a reliable predictor either (Chapter 4 and 5).

In addition to establishing the diagnosis, IgE directed to specific proteins may also help to assess the potential severity of a food allergy. For example, sIgE to the peach lipid transfer protein Pru p 3 has been recognized as a risk factor for severe symptoms in patients with peach allergy. We have shown that the presence of sIgE to the more recently identified peach allergen Pru p 7 is a risk factor for severe peach allergy in adults in both Europe and, even more clearly, in Japan (Chapter 3).

Regional differences in patterns of allergen recognition
It is well established that sensitization to proteins from the lipid transfer protein (LTP) family is more common in the Mediterranean region, whereas PR-10 proteins predominate in Northern European countries, mainly due to cross-reactivity with highly prevalent birch pollen. In this thesis, we confirmed these patterns for peach (Chapter 3) and almond (Chapter 5). In addition, we extended these findings.

Despite the low concentration of birch pollen in the Mediterranean region and Japan, patients in these areas were frequently sensitized to PR-10 proteins (Chapter 3). This suggests that cross-reactivity also plays a role in these areas, possibly driven by other tree species such as oak.

Peach is often considered the primary sensitizing source for lipid transfer proteins. However, this is not what we observed in the Netherlands (Chapter 5). Other food sources, or perhaps pollen such as mugwort or plane tree pollen, may act as the initial sensitizing trigger in the Netherlands.

These findings demonstrate that patterns of allergen recognition differ between Northern and Southern Europe. Although the same allergens are recognized, albeit in different proportions, is likely due to different origins of sensitization.

Improving the diagnosis of food allergy by monitoring or conducting home-based oral food challenges
The gold standard for diagnosing a food allergy is an oral food challenge, during which the patient is given gradually increasing amounts of the suspected food in a hospital setting. A food challenge provide valuable information about the presence of food allergy, potential severity of a reaction and the dose at which symptoms occur.

Food challenge outcomes vary in severity and severe reactions can occur. We investigated whether continuous monitoring during an oral food challenge could predict a severe reaction before objective allergic symptoms occurred using the following key vital parameters: heart rate, blood pressure, QT interval, respiratory rate and body temperature (Chapter 7). Our study demonstrated that a combination of these parameters could predict an allergic reaction with approximately 60% accuracy before it became clinically apparent. If these results can be confirmed, oral food challenges could potentially be discontinued earlier, thereby reducing the frequency or severity of serious reactions.

An oral food challenge is labor- and time-intensive, and is therefore associated with substantial costs. In addition, high demand and limited capacity often result in long waiting lists. Our results show that, for certain patients, a home-based oral food challenge could be a suitable alternative (Chapter 9). Based on careful patient selection and with thorough instruction, we determined whether the test could be safely performed at home. The criteria were as follows: 1) Only patients with a history of mild symptoms were selected. 2) These patients received clear verbal and written instructions outlining the procedure and how to act in case of a reaction. 3) The home-based oral food challenge was always initiated with the lowest dose. 4) Clear stop criteria were defined in advance and provided in writing. 5) Emergency medication was supplied and 6) direct, easily accessible contact with hospital staff was ensured throughout the process.

Our results demonstrate that home-based oral food challenges performed in accordance with the above-mentioned criteria was safe for all patients who underwent the procedure. Future research should focus on optimizing the implementation of home-based oral food challenges and evaluating their impact on reducing workload and healthcare costs.

Birch pollen-related food allergy: underestimated impact and potential for treatment
Birch pollen-related food allergy is the most common food allergy in Northern and Central Europe. It affects individuals who are already allergic to birch pollen. The PR-10 protein found in birch pollen shows strong similarity with related proteins found in many different foods, which can lead to cross-reactivity. Consequently, patients often exhibit allergies to multiple foods simultaneously. We demonstrated that this can involve many foods, in our study up to 16 (Chapter 9).

The foods that most frequently cause symptoms are fruits such as apple and peach, nuts such as hazelnut and walnut, legumes such as soy and peanut, and vegetables such as carrot and celery (Chapter 9).

The symptoms of birch pollen-related food allergy are generally mild, such as oral itching. However, this is not always the case. This thesis found that 13% of birch pollen allergic patients undergoing an oral food challenge experienced severe reactions (Chapter 9). Specifically, patients who underwent a food challenge for soy milk in our study experienced severe symptoms. Fresh cherry and raw walnut were the next most frequent causes of severe reactions (both 33%). In absolute numbers, raw hazelnut most frequently caused severe symptoms.

Patients with a birch pollen-related food allergy experience a reduction in quality of life (Chapter 9). Mostly because they are afraid of having a food allergic reaction, they constantly monitor their diet and find it difficult to maintain a healthy lifestyle. Furthermore, the foods they are allergic to are essential components of a nutritious diet.

Currently, there is no treatment available for patients with a birch pollen-related food allergy. As birch pollen-related food allergy is associated with birch pollen allergy, which can be treated with birch pollen immunotherapy, we conducted a literature review to investigate whether this treatment could also be effective for birch pollen-related food allergy (Chapter 10). The review revealed that few high-quality studies have been conducted and that their results were inconsistent. Therefore, we investigated whether immunotherapy using a sublingual tablet containing birch pollen could be an effective treatment for patients with birch pollen-related food allergy (Chapter 11). Already after one year of treatment, half of the patients reported a significant improvement in their food allergy-related quality of life. Additionally, two-third of patients reported a reduction in symptom severity or were able to tolerate a larger amount of the causative food. Given these positive results, patients will be followed for further evaluation.

Personalized dietary advice for food allergy
There is no standard treatment available for patients with food allergy. Therefore, guidance is provided on how to manage allergic reactions using an emergency kit, as well as on how to follow an appropriate diet. As patients with birch pollen-related food allergy often have to avoid many foods or can only tolerate them in a specific processed form, personalized dietary advice is needed. It is known that patients usually do not experience symptoms from processed fruits and vegetables (such as apple sauce), but react primarily to fresh products. However, it is unclear whether this also applies to nuts and soy.

In this thesis, we demonstrate that patients who experienced adverse reactions to raw almond did not react to roasted almond. In such cases, roasted almond does not need to be avoided (Chapter 4). Of patients allergic to raw hazelnut, one-third did not report symptoms when consuming roasted hazelnut or hazelnut paste (Chapter 9).

For soy, we created a ranking of different products based on the frequency and severity of symptoms they caused. From most to least frequent, the order was: soy dairy, tofu, soybean and tempeh, meat substitute, soy sauce, bread and soy containing biscuit “Biscoff”. The order for severity of symptoms was quite similar: soy dairy caused the most severe reactions, followed by tofu, meat substitute, soybean and tempeh, soy sauce, bread and “Biscoff”. Additionally, 53% of patients who reported symptoms to soy milk did not experience reactions to other soy products.

The fact that reactions to different processed (soy) products vary in both frequency of causing symptoms and severity of symptoms, and that many patients tolerate certain products, suggests that the diet can be personalized.

New candidates for allergen labelling
To enable individuals with a food allergy to properly follow dietary advice, allergen labelling provides important guidance. Allergen labelling provides important guidance for individuals with a food allergy. In Europe, labelling is currently mandatory for 14 foods or food groups, including tree nut, peanut, cow’s milk, soy, hen’s egg, shellfish, sesame, fish, cereal, celery, mustard, mollusc, lupin and sulphur dioxide/sulphites.

However, foods that are currently not labelled can still cause frequent and severe reactions. Our research showed that pine nut was responsible for 5% of food-allergic reactions, 51% of which were severe (Chapter 2). For sunflower seed, the respective figures were 2% and 43%. Based on these results, it may be advisable to consider labelling pine nut in the future and placing sunflower seed on a ‘watch list’ for potential future allergen labelling.

Key conclusions of this thesis
- Almost 200 different foods can cause a food allergy
- The 2S albumin protein of cashew is a reliable predictor of cashew allergy, whereas that of almond is not a reliable predictor of almond allergy
- The peach allergen Pru p 7 is the most reliable indicator of severe peach allergy in both Europe and Japan
- Continuous monitoring of heart rate, blood pressure, QT interval, respiratory rate and body temperature shows promising results in predicting allergic reactions and may enable earlier conclusions from oral food challenges, preventing severe allergic reactions during OFC
- A home-based oral food challenge is safe, provided certain conditions are met
- Birch pollen-related food allergy generally causes mostly mild symptoms, but can lead to severe reactions in more than 10% of patients
- Sublingual immunotherapy with birch pollen may be an effective treatment for patients with birch pollen-related food allergy
- Pine nut and sunflower seed cause frequent and severe reactions, and may therefore be candidates for labeling