Pathogenesis of oral allergy syndrome (pollen-food allergy syndrome)

INTRODUCTION AND TERMINOLOGY — Pollen-food allergy syndrome (PFAS or PFS) describes allergic reactions, usually limited to the oropharynx, which occur upon ingestion of certain fresh fruits, nuts, or vegetables in individuals who are sensitized to plant pollens [1,2]. These reactions are a form of immunoglobulin E (IgE)-mediated hypersensitivity. The causative allergens in these plant foods are homologous to pollen allergens [3,4]. The disorder is also called "pollen-food syndrome" and "pollen-associated food allergy syndrome" [5].

Oral allergy syndrome (OAS) is a term used variably as a synonym of PFS or to describe just the isolated oropharyngeal symptoms caused by pollen-related foods. In this review, OAS is applied to isolated oropharyngeal symptoms.

The symptoms of OAS result from contact urticaria of the oropharynx. Symptoms are usually limited to the mouth and throat and only observed with raw forms of the food because the causative allergens are rapidly inactivated by digestion and cooking, although this is not uniformly true. Systemic reactions as well as reactions to cooked foods are observed in a small proportion of patients.

This topic review will present the pathogenesis of PFS. The clinical manifestations, risk factors, diagnosis, and management of PFS are presented separately. (See "Clinical manifestations and diagnosis of oral allergy syndrome (pollen-food allergy syndrome)" and "Management and prognosis of oral allergy syndrome (pollen-food allergy syndrome)".)

PATHOGENESIS — Fruit and vegetable allergens are highly conserved and share varying degrees of homology with allergens from other fruits and vegetables as well as pollens. This structural and functional homology underlies the extensive cross-reactivity observed clinically [6-8].

Sensitization to pollen — Sensitization to inhaled pollen proteins via the respiratory tract is believed to be the initial pathogenic event. The pollen-specific immunoglobulin E (IgE) generated by this mechanism then binds to the surface of mast cells and basophils throughout the body, including those in the oropharyngeal mucosa. Upon oral contact with a related food, these IgE molecules recognize homologous conformational epitopes on proteins in the food, triggering localized release of inflammatory mediators and the symptoms of OAS. In most cases, the allergens are subsequently destroyed in the stomach, limiting any further reaction.

In support of the inciting role of pollen, studies suggest that oropharyngeal reactions caused by foods that are related to birch pollen (such as apple) are more prevalent in areas where birch trees are common, such as central and northern Europe [9]. In contrast, in birch-free Spain, allergy to apple presents differently. It is more often associated with systemic reactions (in >35 percent of patients) and has been attributed to different allergens, suggesting a distinct pathogenic mechanism [9]. Dendritic cells from birch pollen-allergic donors, when exposed to Bet v 1 in vitro, induced allergen-specific T cell proliferation and T helper type 2 (Th2) polarization (interleukin-5 [IL-5], interleukin-13 [IL-13]). In contrast, exposure to celery Api g 1 significantly enhanced the production of interferon-gamma (IFN-gamma) and downregulated IL-13 [10].

In a mouse model, animals sensitized to birch pollen and then orally challenged with apple extract exhibited OAS-like symptoms, such as snout/oral rubbing and positive skin prick test [11]. The apple extract administered with a protease inhibitor reduced the oral rubbing frequency, which was also significantly reduced in the immunized Fcer1a-/- and mast cell-deficient mice compared with the immunized control mice. The oral rubbing frequency, serum IgE levels, and Th2-cytokine production by the cervical lymph node cells were significantly reduced in sensitized mice lacking IL-33-/- and thymic stromal lymphopoietin (TSLP) receptor (Crlf2-/-), compared with sensitized wild-type mice, suggesting that IL-33 and TSLP contribute to the pathogenesis of OAS. The apple-extract stimulation did not lead to increased Th2-cytokine production in the oral mucosa or number of group 2 innate lymphoid cells or eosinophils. Thus in the mouse model, OAS induces an early-phase response by mast cell degranulation but there is no evidence of a late phase with local Th2/eosinophilic inflammation in the oral mucosa.

ALLERGENS IN PFS — The conformational B cell epitopes on proteins implicated in OAS and PFS are sensitive to heat, acid, and digestive enzymes and therefore usually cause reactions limited to the oropharynx. Freezing does not appear to inactivate these epitopes [12]. In contrast, the allergens responsible for isolated food allergy (in the absence of pollen sensitization) are typically resistant to both heat and digestion and therefore have a greater potential to cause systemic symptoms.

●Birch-allergic individuals appear to develop PFS more commonly than those allergic to grass or weed pollens. The allergens causing birch-related OAS have been relatively well-characterized, and most are either Bet v 1 cross-reactive antigens or profilins.

●Subjects allergic to members of the plane tree family in Spain were reported to have over a 50 percent rate of allergy to related foods, most frequently to hazelnut and peach [13]. In plane pollen-allergic adults with peach allergy, the nonspecific lipid transfer protein (nsLTP), Pla a 3, was shown to be a major allergen [14].

●Much less is known about the allergens responsible for PFS in grass- and weed-allergic patients, compared with those in tree pollen-allergic patients.

Birch pollen-related allergens — The primary allergens causing PFS in birch-sensitized patients are Bet v 1 cross-reactive antigens and profilins.

Bet v 1 cross-reactive antigens — Bet v 1 is a major allergenic protein in birch tree (Betula verrucosa) pollen. Bet v 1 belongs to the family of the pathogenesis-related protein 10 (PRP-10), which is important to plant defense [15,16]. Most PFS in birch-allergic individuals is caused by proteins that cross-react with Bet v 1, including those in Rosacea fruits (apple, cherry, apricot, pear, plum) and Apiaceae vegetables (eg, celery, carrot) (figure 1) [17]. Bet v 1 homologs include Mal d 1 (in apple), Pru av 1 (in cherry), Pru ar 1 (in apricot), Pyr c 1 (in pear), Pru p 1 (in peach), Api g 1 (in celery), and Dau c 1 (in carrot).

Bet v 1 type allergenic epitopes are conformational in nature and depend on the intact tertiary protein folding. They are highly susceptible to heating and digestion and undergo prompt disintegration in the stomach. Such treatments destroy the immunoglobulin E (IgE)-binding capacity of these allergens, although the ability to activate allergen-specific T cells persists [18-20]. In some adults with atopic dermatitis and birch pollen allergy, ingestion of cooked birch pollen-related foods results in worsening of eczema [19]. One could speculate that similar effects might occur in other conditions with T cell-mediated mechanisms, such as eosinophilic esophagitis, allergic rhinitis, or asthma. In addition, ingestion of cooked birch-pollen cross-reactive foods might drive perennially increased allergen-specific IgE levels in patients suffering from seasonal allergic disorders [21].

Profilins — The first profilin identified was an allergen in birch pollen that was named Bet v 2. The profilins are known to be ubiquitous cross-reactive plant allergens [22]. IgE of birch pollen food-allergic individuals cross-reacts with Bet v 2 homologous proteins from apple, pear, melon, carrot, celery, potato, and mugwort (figure 1) [16,23]. Sensitization to profilins can be found in 10 to 30 percent of pollen-allergic patients and most likely occurs via respiratory tract.

Risk factors for systemic reactions in patients sensitized to Bet v1 and/or Bet v2 — In a case-controlled Italian study of 89 patients with history of systemic reactions to plant foods and 81 controls, tree nuts, foods from the Rosaceae and Apiaceae families, and soymilk were the most common culprit foods [24]. Seventeen (19 percent) patients were taking a proton pump inhibitor (PPI) when the systemic reaction occurred versus 5 percent of controls. The ingestion of the offending food in liquid form (soymilk) was reported by 15 percent of patients with systemic reactions compared to 2 percent of controls. However, soy milk-induced systemic reactions were not associated with PPI treatment. Fasting and excess of allergen were other relevant cofactors for systemic reactions, but NSAIDs and exercise were not. In 39 (44 percent) of cases, systemic reactions occurred without any identifiable cofactor. For PR-10 hypersensitive patients, soybean-based beverage carry an increased risk of inducing systemic symptoms.

ALLERGENS IN ISOLATED FRUIT/VEGETABLE ALLERGY — When fruit or vegetable allergy develops in the absence of pollen allergy, patients may be sensitized to nonspecific lipid transfer proteins (nsLTPs) or to gibberellin-regulated proteins (GRPs) [25-29]. In general, sensitization to these proteins is associated with higher rates of systemic reactions and with food-dependent, exercise-induced anaphylaxis [30]. For patients in the Mediterranean region, nsLTP in peanut (Ara h 9) and in walnut (Jug r 3) have been identified as relevant allergens [31,32].

Nonspecific lipid transfer proteins — Reactions to plant foods in patients sensitized to nonspecific lipid transfer proteins (nsLTPs) are more likely to be severe and systemic [33]. The nsLTPs belong to the prolamin superfamily of plant proteins and are members of pathogenesis-related protein 14 (PRP-14). They have antifungal and antibacterial properties, and they are involved in plant defense and are abundant in the peel of plant foods [34]. In addition, nsLTPs have been found to be major allergens in various foods (eg, peach Pru p 3, cherry Pru av 3, apple Mal d 3, hazelnut Cor a 8, orange Cit s 3, strawberry Fra a 3) and in pollen.

These proteins are highly cross-reactive due to extensive sequence homology and are panallergens [35,36]. Panallergens are structurally homologous and phylogenetically conserved allergenic proteins that are frequently responsible for immunoglobulin E (IgE)-mediated polysensitization and false-positive test results in subjects without clinical reactivity.

nsLTPs are remarkably heat-stable, presumably due to four alpha-helices stabilized by four disulfide bonds. They also retain their allergenicity through processing and are found in products such as sterilized peach juice, cooked apple, beer, and fermented products, such as wine [37-39]. nsLTPs are also resistant to digestive enzymes (pepsin, trypsin) [40].

Sensitization to nsLTPs is prevalent in the Mediterranean areas (eg, Spain and southern Italy, where birch trees do not grow) and uncommon in central and northern Europe (where birches are common) [41-43]. In China, mugwort pollen nsLTP (Art v 3) has been reported as a primary sensitizer in patients with peach allergy [44].

It is assumed that nsLTPs may sensitize both by inhalation and ingestion. In fact, nsLTPs are a minor allergen in subjects with plane-pollen allergy without food allergy and a major allergen in subjects with reported food allergy to peach [14]. Sensitization via the gastrointestinal route may be more important in this type of allergy, considering the prevalence of this disorder in areas lacking birch trees. When T cells were purified from peripheral blood and stimulated with natural peach nsLTP Pru p 3, they produced mainly T helper type 2 (Th2) cytokines and very little interleukin-10 (IL-10), unlike T cells specific for other plant food and pollen allergens. The expression of gut homing integrin beta-7 on Pru p 3-specific T cells was comparable with that observed on peanut-specific T cells and higher than that seen in different pollen-specific T cells, suggesting a gastrointestinal route of sensitization to nsLTP [45].

Gibberellin-regulated proteins (GRPs)/GASA proteins — Gibberellin-regulated proteins (GRPs)/GASA proteins are members of cysteine-rich antimicrobial peptide families and are highly conserved across the plant kingdom. GRPs, also known as the Snakin/GASA protein family, are involved in plant growth and defense against pathogens. The expression of GRP is up-regulated by gibberellins, which are a class of natural plant hormones produced by plants and some fungi and bacteria.

Fruit-derived GRPs have a very high content of cysteine, rendering them resistant to heat and digestion. GRPs are considered type I food allergens that can induce severe allergic reactions. An alternative mechanism of sensitization to fruit-GRP allergens could be primary sensitization to cypress pollen GRP.

Several fruit-GRPs have been identified as allergens in Japanese and Italian patients, including peach Pru p 7, Japanese apricot Pru m 7, orange Cit s 7, pomegranate Pun g 7, as well as cypress pollen GRP. The clinical features of allergy to fruit-GRP often include systemic reactions, allergies to multiple, botanically-unrelated fruit, and dependence on cofactors, such as exercise and aspirin. Clinical cross-reactivity between fruit-GRPs has been demonstrated. GRP allergy induces peculiar clinical symptoms: laryngeal tightness and facial swelling, especially eyelid edema, which was identified as a predictive factor for sensitization to Pru p 7 [46].

ALLERGENS IN SPECIFIC FOODS — Specific allergens have been identified in some plant foods that are associated with either oropharyngeal or systemic reactions. However, there is no reliable way to discern which allergen an individual is sensitized to outside of research settings. While the commercially available allergen microarray ImmunoCAP ISAC detects several major fruit and vegetable allergens, predominantly Bet v 1 cross-reactive, the component immunoglobulin E (IgE) profiling does not seem to enhance the diagnostic accuracy in PFS, suggesting that additional allergens may be relevant [47]. This information is presented to help clinicians understand the known associations between pollens and foods and to be aware of the foods that have been implicated in systemic reactions.

Apple — The apple allergen Mal d 1, a common cause of OAS, is a Bet v 1 homolog that shares 64 percent amino acid sequence similarity and common IgE epitopes with Bet v 1. A study utilized Bet v 1-Mal d 1 chimeric protein and showed that birch-allergic adults with PFS to apple had significantly higher IgE reactivity to the Mal d 1 regions grafted on Bet v 1 than birch-allergic adults without PFS [48]. This suggests that different B cell epitopes may be important for the expression of PFS to apple. Mal d 1 is highly unstable and can be disrupted by heating, processing, and digestion. In contrast, the nonspecific lipid transfer protein (nsLTP) Mal d 3 is highly resistant to heating and retains the ability to bind IgE following heating at 90ºC (194ºF) for 20 minutes [39,49].

Multiple reports documented that in the areas where birch trees are common, such as in central and northern Europe (Netherlands, Austria, northern Italy), allergic symptoms to apple are usually mild (>90 percent report oral symptoms) and associated with birch pollen allergy and sensitization to Mal d 1 [9,50]. Also, in Korean adults and children with atopic dermatitis and sensitization to birch pollen, apple is the most commonly reported OAS trigger [51,52]. In rare patients with anaphylactic reactions to apple, no IgE to nsLTP has been detected [53]. In contrast, in birch-free southern Europe (Spain), allergy to apple is more often severe (>35 percent report systemic reactions) and associated with peach allergy and sensitization to the nsLTP Mal d 3 [54].

The content of these allergens in individual apples is highly variable and affected by the specific cultivar, ripeness at the time of harvesting, and duration and conditions of storage. Both Mal d 1 and Mal d 3 are expressed throughout fruit development, and their levels increase with ripening. Certain apple cultivars were reported to contain less Mal d 1, such as Santana, Braeburn, Pinova, or Elise compared with high Mal d 1 content in Golden Delicious and Jonagold [55]. These differences were confirmed with prick skin testing and double-blind, placebo-controlled food challenges (DBPCFC) for Santana and Golden Delicious [50]. Mal d 1 content was shown to increase during storage at 2ºC (35.6ºF) and was higher in stored apples than in freshly picked apples [56,57].

Hazelnut — Hazelnut contains allergens homologous to Cor a 1 and Bet v 1, the major allergens in hazel and birch tree pollens respectively, as well as profilins, Cor a 2 [58]. Other allergens, including the lipid transfer protein, Cor a 8, 11S seed storage globulin (legumin-like), Cor a 9, and 2S albumin, Cor a 14 may be responsible for systemic reactions to hazelnut [59-62]. Sensitization to Cor a 1 is highly prevalent in hazelnut-allergic patients from western Europe (Switzerland/Denmark), while sensitization to Cor a 8 is more common in patients from Spain. Sensitization to profilin Cor a 2 is detected in approximately 40 percent of the patients with hazelnut allergy, regardless of geographic location [63].

Peanut — Peanuts, even after roasting, can be another source of OAS in birch-allergic patients (figure 1). In a series of 20 adult patients with birch pollen and peanut allergy who underwent DBPCFC to lightly roasted peanut flour, 60 percent developed only OAS, whereas 40 percent had additional systemic symptoms [64]. Specific immunoglobulin E (IgE) antibodies against the peanut allergen Ara h 8, homologous with birch Bet v 1, were detected in sera from 85 percent of subjects. Ara h 8 is also homologous to Api g 1 (celery), Gly m 4 (soy), and Pru av 1 (cherry) [65]. Ara h 8 was partially disrupted by roasting and completely destroyed by gastric enzymes. These data suggest that peanuts may cause OAS in a subset of birch pollen-allergic individuals. However, clinicians should be aware that the majority of peanut- and tree nut-allergic individuals react to nonpollen-related stable proteins that result in more systemic and severe reactions. Thus, diagnosis of PFS in subjects with peanut- and or tree nut-induced oral symptoms should be made cautiously. We usually consider patients with peanut or tree nut allergy, even if symptoms have been limited to the oropharynx, to be at risk for more severe reactions and manage them accordingly (ie, prescribe self-injectable epinephrine and provide appropriate counseling). (See "Management and prognosis of oral allergy syndrome (pollen-food allergy syndrome)".)

Component resolved diagnostics technology that detects IgE sensitization to individual allergens may be used in the future to define the risk for more severe symptoms, such as detectable IgE antibodies to stable seed storage proteins (eg, Ara h 1, Ara h 2 a, and Ara h 3) or lipid transfer protein (Ara h 9). Exclusive IgE sensitization to pollen cross-reactive Ara h 8 may be associated with mild oral symptoms and low risk for anaphylaxis (table 1). (See "Future diagnostic tools for food allergy".)

Celery — OAS caused by celery tuber (celeriac) can be seen in patients who are sensitized only to birch (figure 1). Cross-reactivity between Bet v 1 and celery Api g 1 underlies this reaction. Systemic reactions to celery are also reported, especially in highly sensitized patients. (See 'Celery-mugwort-birch-spice syndrome' below.)

Peach — Peach is one of the most frequent causes of food allergy in Europe, both in the northern regions and in Mediterranean areas. Profilins are cross-reactive proteins in pollen and peach and are the likely cause of OAS, which is more common in central and northern Europe [66]. In one report, up to 86 percent of patients with pollinosis reported isolated oropharyngeal symptoms in response to peach [67]. Peach is also the most commonly reported cause of OAS in young Mexican adults [68].

Peach allergy in the Mediterranean (eg, Spain) appears to be particularly prevalent and is associated with more systemic reactions [69]. One Spanish study showed that 73 percent of peach-allergic patients did not have pollen allergy, and these subjects reacted with anaphylaxis. Peach nsLTP (Pru p 3) was the responsible allergen [35]. Severity of the reactions was augmented in the presence of cofactors, including nonsteroidal anti-inflammatory drugs (NSAIDs) and exercise [70]. By comparison, in a study of Spanish patients who had both peach allergy and pollinosis, the nsLTP Pla a 3, also found in the pollen of the plane tree, was a major allergen, and Pru p 3 was a minor allergen [14].

Cultural differences in food procurement may also contribute to the high rates of systemic peach reactions in Spain. Peach nsLTP is highly concentrated in the peach fuzz [71]. Among children with reported allergic reaction to ingestion or contact with peach, 93 percent tolerated peach pulp during an oral food challenge [72]. When peaches are bought from small local farmers, as is typical in the Mediterranean, they typically have abundant fuzz. In contrast, when peaches are grown by large producers, the fruit is brushed and washed prior to packaging, and most of the peach fuzz is removed. Peaches in central and northern European countries are generally imported from large producers.

Peach-allergic patients are frequently reactive to other related foods (figure 2) [2,70,73,74].

Carrot — Carrot-specific IgE antibodies can cross-react with celery, watermelon, apiaceous spices (fennel, coriander, caraway, aniseed), and birch and mugwort pollens (figure 1). The responsible carrot allergens are Dau c 1 (a Bet v 1 homolog), carrot profilin, a Bet v 6 cross-reactive allergen, and cross-reactive carbohydrate determinants [75-78].

In a case series of 20 patients with carrot allergy documented with DBPCFC to raw carrot, 16 patients developed OAS, 3 patients had additional symptoms outside the oropharynx, and 1 patient had no oral symptoms but developed flushing, conjunctivitis, and dyspnea [76]. Among 12 patients challenged with raw carrot during DBPCFC, two reacted with oral symptoms at the first dose of 10 mg, and the first objective symptoms were observed at the cumulative dose of 35 mg [79]. (See 'Celery-mugwort-birch-spice syndrome' below.)

Soybean — Soybean is a prominent cause of systemic food allergy in children. Reactions are attributed to heat-stable proteins that induce sensitization via the gastrointestinal tract, such as seed storage proteins (legumins, Gly m 6, and vicilins [Gly m 5] or 2S albumins) [80,81]. In adults, soybeans may cause OAS or systemic symptoms.

One study examined 20 adults without previous reactions to soy who developed OAS and/or systemic symptoms following ingestion of a soy protein-containing food supplement [82]. IgE binding to Gly m 4 (previously called starvation-associated message 22 or SAM22) was demonstrated in the majority. Gly m 4 is homologous to Bet v 1, although it appears to be more stable than other Bet v 1 homologues (eg, such as Mal d 1 in apple that only rarely induces systemic reactions) and therefore more likely to cause systemic symptoms [83,84].

In another study, 22 birch-allergic adults with histories of soy reactions underwent DBPCFC, 10 patients experienced symptoms localized to the oral cavity, and 6 patients had a more severe reaction [85]. Gly m 4-specific IgE was found in all but one of the patients. The Gly m 4 content of foods was highly dependent on processing, and it was lowest in unripe green soybeans and highest in ripe soybeans after three years of storage. Gly m 4 content of soybeans was reduced after 30 minutes of cooking. Among food products, Gly m 4 content was highest in dietary soy powder (used in foods such as protein energy bars), much lower in tofu and soy drink, and undetectable in fermented soy products, such as soy sauce and miso.

Kiwi — Kiwi fruit has been reported to cause both OAS (especially in Japan and Southern Europe) and severe systemic reactions and cross-reactions with celery, rye, birch, mugwort, timothy grass pollen, and latex (figure 1) [86,87].

The major allergen specific for kiwi appears to be actinidin (Act c 1), a member of the cysteine protease family [88]. Actinidin does not cross-react with pollen proteins, is assumed to cause sensitization via the gastrointestinal tract, and is associated with more severe symptoms [89]. The Bet v 1 homolog, Act d 8, and the profilin, Act d 9, are important allergens in pollen-related kiwi allergy manifesting with less severe symptoms. Act d 11, of the ripening-related protein family (from the Bet v 1 superfamily), Act d 12, 11S globulin and Act d 13, 2S albumin were identified as major allergens [90,91].

Melons — The majority of patients reporting allergy to melons (Cucurbitaceae) (watermelon, cantaloupe, and honeydew) have weed pollen allergy (figure 1). Melon-allergic patients are often reactive to other related foods (avocado, banana, kiwi, and peach) (figure 2) [1,92,93].

Melon allergy can also arise from cross-reactivity with grass pollens. Clinical studies from a ragweed-free area showed that melon allergy still occurred mainly in patients with pollinosis, although Plantago and orchard grass pollens were the cross-reactive pollens [1]. Sera from 17 adults from Spain with oral symptoms to melons bound profilin Cuc m 2, and the majority also recognized profilins from timothy grass Phl p 12 and birch Bet v 2. The Cuc m 2 mimotope was defined and was almost identical to the sequences in Phl p 12 and Bet v 2 [23]. Allergy to zucchini, another plant in the Cucurbitaceae family, has been reported and confirmed with DBPCFC with two of four patients having OAS. IgE inhibition experiments showed cross-reactivity with the birch pollen profilin (Bet v 2) [94].

SYNDROMES ASSOCIATED WITH SYSTEMIC REACTIONS — Certain types of PFS are associated with more severe reactions. Several of these have been recognized as specific syndromes.

Celery-mugwort-birch-spice syndrome — The "celery-mugwort-birch-spice syndrome" is a potentially severe form of celery allergy seen in adults and children who are sensitized to both birch and mugwort pollen. Patients with this syndrome react to celeriac (celery tuber). This allergy is especially important in Switzerland, Germany, and France. Patients may react to the Apiaceae family (carrot, caraway, parsley, fennel, coriander, fenugreek, cumin, dill, and aniseed), as well as paprika, pepper, mango, garlic, leek, and onion (figure 1) [95-97]. This cross-reactivity is believed to be due to celery profilin (Api g 4), which is highly cross-reactive with profilins from both mugwort (Art v 4) and birch (Bet v 2), as well as latex Hev b 8 [97]. In two subjects with systemic reactions to fennel, sensitization to an allergen homologous to Api g 5 (high molecular weight allergen) has been detected [98]. Some patients may also react to latex. (See 'Latex-fruit syndrome' below.)

Patients with celery-mugwort sensitization had immunoglobulin E (IgE) antibodies that recognized heated celery extracts, in contrast to those with OAS to celery [95]. In addition, 46-kDa to 60-kDa bands cross-reacting with IgE against mugwort pollen extract have been described in celery and identified as cross-reactive carbohydrate determinants (CCDs) on glycoproteins [99]. Patients with reactions to cooked celery confirmed by double-blind, placebo-controlled food challenges (DBPCFC) were sensitized to profilin and/or CCDs [100].

Mugwort pollen-food allergy syndrome — Patients sensitized to mugwort (Artemisia vulgaris) may develop systemic food allergy (eg, to mustard). The responsible allergens have not been identified. Among 148 Chinese adults with mugwort pollen allergy, 72 percent reported food allergy and among them, 48 percent reported at least one episode of food-induced anaphylaxis [101]. Food allergy was associated with sensitization to Art v 3 but not Art v 1. Among individuals reporting food allergy, 80 percent were sensitized to Pru p 3, 69 percent to Ara h 9, and 63 percent to Cor a 8; those with systemic reactions had higher concentrations of sIgE to these allergens compared to those with OAS.

Latex-fruit syndrome — Approximately 30 to 50 percent of individuals who are allergic to natural rubber latex (NRL) show an associated hypersensitivity to some plant-derived foods, especially fresh fruits [102]. In one series, over 50 percent of reported reactions to foods in latex-allergic individuals were anaphylactic [103]. An increasing number of plant foods, such as avocado, banana, chestnut, kiwi, peach, tomato, white potato, and bell pepper, have been associated with this syndrome (figure 3) [103-107]. The allergens involved in the latex-fruit syndrome include hevein (Hev b 6.02), Hev b 7, and the panallergen profilin Hev b 8. The evaluation of latex-fruit syndrome is reviewed elsewhere. (See "Latex allergy: Epidemiology, clinical manifestations, and diagnosis", section on 'Testing for allergy to latex-related foods'.)

SUMMARY

●Pollen-food allergy syndrome (PFAS or PFS) describes immunoglobulin E (IgE)-mediated allergic reactions, usually limited to the oropharynx, which occur upon ingestion of certain fresh fruits, nuts, vegetables, or spices, in individuals who are sensitized to plant pollens. The causative allergens in these plant foods are homologous to pollen allergens. The term "oral allergy syndrome" (OAS) is used in this review to describe PFS reactions that are limited to the oropharynx. (See 'Introduction and terminology' above.)

●The pathogenic mechanism begins in the respiratory tract with sensitization to inhaled pollen. Symptoms then result from oral contact with plant foods containing homologous allergens. In most cases, although not all, the allergens are subsequently destroyed in the stomach, limiting any further reaction. (See 'Pathogenesis' above.)

●The allergens implicated in OAS and PFS are sensitive to heat, acid, and digestive enzymes. The primary allergens causing PFS in birch-sensitized patients are Bet v 1 cross-reactive antigens and profilins. (See 'Allergens in PFS' above.)

●In contrast to the allergens that cause PFS, the allergens responsible for isolated food allergy (in the absence of pollen sensitization) are typically resistant to both heat and digestion and therefore have a greater potential to cause systemic symptoms. Sensitization is believed to occur through the gastrointestinal tract. (See 'Allergens in isolated fruit/vegetable allergy' above.)

●Specific allergens have been identified in some plant foods, which are associated with either oropharyngeal or systemic reactions. However, there is no way to discern which allergen an individual is sensitized to, outside of research settings. (See 'Allergens in specific foods' above and 'Syndromes associated with systemic reactions' above.)