Original article: Two different profiles of peach allergy in the north of Spain


Prof. (Dr) María L. Sanz
Department of Allergology and Clinical Immunology
University Clinic of Navarra
Apartado 4209
31080 Pamplona


Background:  Peach allergy has two different patterns: central Europe with oral allergy syndrome (OAS) related to a primary sensitization to birch pollen Bet v 1 and profilins and southern Europe with mostly systemic symptoms, in many cases due to sensitization to lipid-transfer proteins.

Methods:  Thirty peach-allergic patients with positive skin and food challenge tests and 29 control subjects were included. Skin prick tests (SPT) with inhalant allergens, commercial peach and apple extracts and native Pru p 3 were performed. In vitro specific immunoglobulin (Ig) E to grass pollen, birch pollen, peach, apple, rBet v 1, rBet v 2 and rPhl p 12 was determined by CAP, and rBet v 1, rMal d 1, rMal d 4, rMal d 3 and rPru p 3 using the ADVIA-Centaur platform. Basophil activation test (BAT) with commercial peach extract, commercial apple extract, nPru p 3, rMal d 3, rMal d 1 and rMal d 4 was also performed.

Results:  Pru p 3 was the major allergen in the patient group from northern Spain. Sensitization to this allergen was found in 100% of the patients with systemic symptoms or contact urticaria. Only 60% of OAS patients were sensitized to Pru p 3, being all of them sensitized to profilins and 60% of them to allergens of the Bet v 1 family. Specific IgE determination and BAT using recombinant allergens (rPru p 3) show specificity and sensitivity values close to 100%.

Conclusions:  Most peach-allergic patients coming from the north of Spain present systemic symptoms after ingestion of peach, Pru p 3 being the main allergen. Patients with OAS present profilin-Bet v 1-related sensitization. Thus, in the north of Spain our patients show a mixed central-south Europe pattern with LTP-profilin-Bet v 1 sensitization depending on the symptoms presented. The use of natural and recombinant plant allergens, allows establishing the sensitization patterns to the different allergens studied.

Peach allergy is the most frequent type of food allergy in the adult population in southern Europe (1, 2), and two clearly different clinical response patterns have been described for this food. In central Europe, oral allergy syndrome (OAS) prevails (3), whereas in southern Europe, peach allergy triggers more severe, even life-threatening reactions (4). These local differences reveal the existence of two different sensitization patterns (5, 6). In central Europe, initial birch pollen allergy influences the development of food allergy to Rosaceae, as birch allergens, mainly Bet v 1 and Bet v 2, have an important cross-reactivity with homologous proteins of Rosaceae fruits (3, 5). In southern Europe, where the presence of birch is rare, peach allergy is mediated mostly by Pru p 3, which is a lipid-transfer protein (7, 8), and sensitization to this allergen is not dependent on a previous allergy to birch pollen. Pru p 3 is a major allergen of peach in Spain (4), but it does not seem to be exclusive. Among peach-allergic patients in the north of Spain there is also a subgroup with OAS or with only contact urticaria, in an environment where the prevalence of birch pollen sensitization is rare (9). Thus, in our Atlantic climate environment (northern Spain), birch pollen represents only 2.86% of the total (589 grains of pollen/m3/year) (9). Even though this amount is low compared with the central European levels, it is higher than in other areas of our country with a continental climate, such as Madrid, where it is only around 0.10% of the total number of pollens per year (10).

The aim of this study was to analyse the different patterns of sensitization to the main Rosaceae allergens in peach-allergic patients in the southern European environment, differentiating their profile based on the clinical manifestations, as well as to validate the diagnostic potential of natural and recombinant allergens. To determine this sensitization, we used skin tests with nPru p 3 and two in vitro tests: specific immunoglobulin (Ig) E determination and flow cytometry determination of activated basophils expressing CD63, after stimulating patients’ cells with Rosaceae allergens in order to cover two different diagnostic routes, and to avoid discrepancies observed between the different in vitro diagnostic techniques for this same pathology (11). For this, both native and recombinant allergens were used, the latter not being recommended for use in vivo due to safety reasons.

Material and methods

Patients and challenge tests

Thirty patients were studied (12 men, 18 women, age 27.3 ± 7 years), who visited the Allergy Services of the Basurto Hospital in Bilbao or the University Clinic of Navarra due to a potential peach allergy. The clinical history of all patients regarding their potential food and respiratory allergy was thoroughly investigated. Rubbing test with peach peel was performed in all patients: peach peel was rubbed on the patient's forearm for 30 seconds and the reaction was observed after 20 min. Oral provocation test was not performed in patients with recent history (<1 year) of anaphylaxis after peach ingestion and with positive prick test and specific IgE to peach. Out of the eight patients with anaphylaxis, three were Grade II, three were Grade III and two were Grade IV (12). The rest of the patients underwent open oral challenge test with peach, starting the provocation with 20 g of peach and duplicating each intake until completing a piece of fruit (150 g approximately). The five patients who presented OAS underwent double blind oral challenge test after the open challenge test, as they did not present any clear signs (oropharyngeal pruritus) after the latter. In this double-blind oral provocation test, the peach was masked in a chocolate milk drink, to which three spoonfuls of coconut and two spoonfuls of soluble coffee were added. Finally, 24 patients were challenged.

Twenty-nine control subjects were also selected (16 men and 13 women), 10 of them healthy subjects and 19 pollen-allergic patient (age 30.9 ± 9.8 years). Among the pollen-allergic patients, 16 were monosensitized to grass pollen and three were sensitized to grass and birch pollen. All of them underwent the same in vivo and in vitro determinations as the patients and their peach tolerance was tested by open oral challenge test and rubbing test, obtaining negative results in both assays.

Prick test

All the patients and controls were prick tested according to the usual technique, with the most frequent inhalant allergens in our environment: Lolium perenne, Dermatophagoides pteronyssinus, Lepidoglyphus destructor, cat dander, dog dander, Betula verrucosa, Artemisa vulgaris, apple and peach commercial extracts (ALK-Abelló, S. A., Madrid, Spain) and nPru p 3 (20 μg/ml). Histamine hydrochloride and physiological serum were used as positive and negative controls, respectively. The skin prick test (SPT) was considered positive when the mean wheal diameter was 3 mm or greater, compared with that produced by the negative control.

In vitro tests

Specific IgE determination

In all patient and control subjects total serum and specific IgE determinations to apple, peach, L. perenne, B. verrucosa, A. vulgaris, rBet v 1, rBet v 2 and rPhl p 12 was determined by CAP (Phadia, Barcelona, Spain) following the manufacturer's instructions. In addition, specific IgE to rBet v 1, rMal d 1, rMal d 3, rMal d 4 and rPru p 3 was also determined using the ADVIA-Centaur® platform (Bayer Diagnostics, Barcelona, Spain). The purified allergens were biotin-labelled at a 7 : 1 ratio as indicated (13), and the optimal dose of allergen was determined, which was about 10–20 ng per test.

Basophil activation test

The basophil activation test (BAT) was performed as previously described (14, 15). Two final concentrations of the tested samples were assayed, according to the preliminary results: 2 and 0.5 mg/ml for peach peel, 2 and 0.3 mg/ml for apple peel, and 0.3 and 0.1 μg/ml for the purified recombinant allergens. The concentrations for the crude extracts from plant foods and pollens were the same used in a routine basis (15–17). The concentrations for the recombinant allergens were selected according to the results obtained in 10 peach-allergic patients and controls using, in addition to the ones already mentioned, one higher and one lower concentration (×4) which induced clearly lower activation percentage than those obtained with the one selected, although without unspecific activation in the controls (data not shown).

The optimal cut-off point calculated by receiver operating characteristic (ROC) curves combines sensitivity and specificity and is the one that shows as coordinates the furthest point of the curve from the diagonal. From this, we determined the cut-off point selecting preferentially the optimal specificity values over sensitivity as peach allergy is a low prevalence pathology. On that basis, results indicating a percentage of basophil activation superior to 20% with a stimulation index (SI: test value/background value) higher than 2 were considered positive for BAT.

Statistical analysis

The data were analysed with the statistical program spss 13.0. The comparison of qualitative variables was carried out by means of chi-square or Fisher's exact test. P-values of <0.05 were considered statistically significant.


Clinical symptoms

Of the 30 patients studied, 19 had systemic symptoms (11 had urticaria/angioedema, seven of them associated with OAS, eight with anaphylaxis), five had OAS following the ingestion of peach and six had urticaria after contact with peach, although they tolerated it when ingested. Seventeen had positive rubbing test. Altogether, these clinical data allowed to differentiate three groups of patients: patients with OAS, patients with contact urticaria and patients with systemic symptoms (Table 1). Within the subgroups of peach-allergic patients, four patients with OAS had nasal and/or bronchial symptoms during the grass pollen season and only one of them had nasal symptoms during birch pollination (patient no. 2; Table 1). Of the 19 patients with systemic symptoms after ingestion of peach, eight had symptoms during the grass pollen season and only one had symptoms during the birch pollen season (patient no. 8). Among the patients with contact urticaria, three of them had symptoms during the grass pollen season, and one also had symptoms with birch pollen (patient no. 27).

Table 1.   Clinical symptoms, associated pollinosis and SPT responses to Pru p 3 of the selected and control patients
Patient no.SexAge (years)SymptomsLolium pollenBirch pollennPru p 3Peach fruit
  1. M, man; F, female; U, urticaria; AN, anaphylaxis; CU, contact urticaria; SPT, skin prick test; BAT, basophil activation test; OAS, oral allergy syndrome; Ig, immunoglobulin.

Summary OAS patients (positive results; n = 5; %)100100100801006010060
Summary systemic symptoms patients (positive results; n = 19; %)57.968.473.721.110.5100100100
Summary contact urticaria patients (positive results; n = 6; %)10010010016.750100100100
Total patients (positive results; n = 30; %)73.38083.33033.393.310093.3
Summary healthy controls (positive results; n = 10; %)00000000
Summary pollinic controls (positive results; n = 19; %)100100100026.30015.8
Total controls (positive results; n = 29; %)65.565.565.5017.20010.3

Skin tests

The skin test responses are summarized in Table 1. All the patients had positive skin tests with peach and 28 with nPru p 3 (two OAS patients had negative prick test). In 28 cases, the prick test was positive to apple (negative in patients nos 16 and 20). Twenty-two had a positive prick test to L. perenne, nine to B. verrucosa and 12 to A. vulgaris. All the pollen-allergic control subjects showed a positive skin test and specific IgE to grass pollen. This fact is explained because more than 90% of our pollen-allergic patients in our environment are sensitized to grass pollen (9). All of them had a negative skin test with peach, apple, nPru p 3, and birch and mugwort pollen. Regarding the group of healthy controls, all of them had negative skin tests with the tested allergens.

Joint consideration of the data described above indicated that SPT with nPru p 3 was the best in vivo diagnostic tool for our peach-allergic patients, showing 93% sensitivity (100% in patients with systemic symptoms or contact urticaria) and 100% specificity.

In vitro tests: sIgE and BAT

The results of several in vitro tests are summarized in Table 1 and Figs 1 and 2, for both the patient and control groups.

Figure 1.

 Specific immunoglobulin E level to the different allergens tested in the control group and in the different groups of patients.

Figure 2.

 Basophil activation test to the different allergens tested in the control group and in the different groups of patients.

Pollen and peach extracts

Twenty-four patients had positive specific IgE to Lolium pollen and 10 to birch pollen. In the control group, 16 subjects had positive specific IgE to L. perenne (all of them in the pollen-allergic group) and five to birch pollen.

Twenty-eight patients and three pollen-allergic controls had positive specific IgE to peach (93% sensitivity, 90% specificity). Twenty-six patients had positive BAT to peach, whereas nine pollen-allergic controls were positive (BAT to peach shows 87% sensitivity, and 69% specificity).

Purified allergens

LTP family

Twenty-seven patients had positive specific IgE to LTPs (27 to rPru p 3 and 24 to rMal d 3). The specific IgE to rPru p 3 presented 90% sensitivity (100% for the patients with systemic symptoms or contact urticaria) with a specificity of 100%. The BAT showed a sensitivity of 77% for rPru p 3 (84% for patients with systemic symptoms or contact urticaria) and a specificity of 97% (24 positive patients to rMal d 3 and 23 to rPru p 3). Only one pollen-allergic control showed a positive BAT to rPru p 3.

Profilin family

Six patients had specific IgE to at least one of the profilins tested (six to rBet v 2, six to rPhl p 12 and five to rMal d 4), whereas seven controls (one healthy and six pollen-allergic subjects) showed specific IgE to one or several of them. Ten patients and six pollen-allergic controls had positive BAT with rMal d 4.

Bet v 1 family

Only four patients had specific IgE to rMal d1 and/or rBet v 1, whereas none of the control subjects showed specific IgE to rMal d 1. Five patients had positive BAT to rBet v 1 and six to rMal d 1. One control subject had a positive result for each of them.

Differences depending on the symptoms

Analysis of the differences between the groups of patients based on their symptoms indicate that there were no significant differences with any of the allergens tested in vivo or in vitro between the patients with systemic symptoms and those with contact urticaria. Therefore, in the subsequent comparison with patients showing OAS, we have considered them as only one group.

Birch pollen, Bet v 1 and homologous allergens

The group of patients with OAS presented a higher percentage of positive specific IgE to birch pollen (100% OAS, 20% rest of patients) and positive BAT to rMal d 1 (60% OAS, 12% rest of patients; P < 0.05) than patients with contact urticaria and/or systemic symptoms considered altogether.

LTP allergens

Patients with systemic symptoms or contact urticaria had a greater percentage of positive IgE to rPru p 3 (100%vs 40% OAS; P < 0.005) and to rMal d 3 (92%vs 40% OAS; P < 0.05) than patients with OAS. In BAT, this higher percentage was observed only with rMal d 3 (P < 0.05; 88%vs 40% OAS).


Patients with OAS presented a greater percentage of positive specific IgE to rBet v 2 (100%vs 4%), rMal d 4 (80%vs 4%) and rPhl p 12 (100%vs 4%) than patients with contact urticaria or systemic symptoms considered altogether (P < 0.005 in all the comparisons). These data were similar for rMal d 4 (100% OAS, 20% rest of patients; P < 0.005) in BAT (Fig. 2).

Seven of the patients with systemic symptoms, had associated OAS. Nevertheless, there were no significant differences for any of the food allergens tested between patients with systemic reactions who developed OAS and those who did not. Unlike patients who had OAS exclusively, only one patient with systemic and oral symptoms (also with associated birch allergy) had positive specific IgE to rBet v 2, rPhl p 12, rMal d 4 and rMal d 1, as well as positive BAT with rMal d 4 (patient no. 8).

The presence of a positive prick test to grass or Artemisa pollen among the patients did not show significant differences with the in vitro sensitization values to the different allergens, except for a higher frequency of positive specific IgE to Phleum profilin in the group of patients with a positive SPT to grass pollen. Patients with a positive SPT to birch pollen had more frequent positive IgE test to rBet v 2, rPhl p 12 and rMal d 4.


The main symptom (OAS) presented by patients allergic to peach in central Europe is different from that presented by the Mediterranean patients, who show a high percentage of systemic symptoms. These data could be explained by the existence of two different patterns of sensitization to profilins/Bet v 1 homologous or LTPs, respectively, among allergic patients from the two above-mentioned geographical areas.

In the southern European environment, and particularly in the northern of Spain (Bilbao/Pamplona), patients are environmentally exposed to birch pollen, as corresponds to an Atlantic climate, with 589 grains/year of birch pollen and approximately 25% of pollen-allergic patients had a positive prick test to birch (9). However, the clinical expression of this sensitization is irrelevant. In fact, over the last 10 years, in the Allergy Service of the Basurto Hospital (Bilbao, Spain), only three desensitization treatments to this pollen have been prescribed. In contrast, more than 90% of our peach-allergic patients showed a positive skin test to Pru p 3, which was supported by positive specific IgE levels and BAT to this allergen. These results confirm the data from previous studies that point to Pru p 3, the major peach allergen in Spain and other Mediterranean areas (4, 6, 18). Sensitization to rMal d 3 was present in 90% of our group of patients. This could be because of the high structural similarity between these LTP allergens (18, 19). In fact, 50% of our patients tolerated the ingestion of apple (information obtained from the clinical interview), which corroborates this structural relationship as the cause of rMal d 3 sensitization, rather than an actual primary sensitization to apple LTP. These data are similar to the ones obtained by other authors in apple-allergic patients, being in Spain Mal d 3 the predominant allergen for this sensitization, unlike the central Europe countries where the predominant allergen in apple-allergic patients is Mal d 1. In these series, like in ours, the first triggering food is peach and subsequently it can associate other Rosaceae fruits such as apple due to the cross-reactivity of their LTPs (6). Furthermore, the sensitization to LTP is observed in 60% of our patients with OAS and 100% of patients with contact urticaria, thus suggesting that the sensitization itself is not synonymous of systemic symptoms. In fact in some central European patients monosensitized to Pru p 3 only OAS have been registered (5).

The sensitization pattern of patients with systemic symptoms and of patients with only contact urticaria, but who tolerate the ingestion of the allergen, both in vivo and in vitro, is almost the same. All of them showed a positive prick test and in vitro test (specific IgE and BAT) to Pru p 3. In contrast, sensitization to other allergens, such as Bet v 1 or profilins, is practically irrelevant in both groups (5% of the cases).

In patients with OAS, the sensitization pattern shows clear differences compared with the other two groups. Although three of the patients with OAS had a positive prick test to Pru p 3, sensitizations to two allergen families that are practically not present in the other patients were found in this group. First, all of them had sensitization to profilins (100% of the cases to rPhl p 12 and Bet v 2, 80% of the cases to Mal d 4), which indicates a high cross-reactivity between profilins of different plant foods and pollens (20, 21). In fact, in two of the profilins used by our group, Bet v 1 and Phl p 12, the sequence homology in three of their epitopes ranges from 89% to 94% (18). Secondly, an important number of patients (60%) were sensitized to Bet v 1 and its homologues. This percentage is slightly lower than the one reported in the central European population (22, 23), but clearly higher than the one presented in Central Spain in patients with Rosaceae allergy (4, 6). Moreover, only one patient had birch pollen-related symptoms. Therefore, this sensitization does not seem to be because of birch pollen exposure, which is very low in our area in contrast with central Europe. At least the sensitization to profilins is likely due to the grass pollen sensitization presented by our patients (80% had grass pollen-related rhinoconjunctivitis) (24).

The above-mentioned sensitization pattern indicates that our population is different from the one in central Europe, as demonstrated by the high percentage of LTP sensitization, which is characteristic of the Mediterranean area. However, it also shows differences with other regions within Spain, where profilin sensitization is more important (4) (20%vs 34%). On the other hand, sensitization to Bet v 1 in our patients is higher than the one registered in Central Spain (14%vs 7%), which could be because of the greater birch pollen levels in the northern region of Spain. Only five patients in our series had OAS. Because of the low prevalence of these manifestations in our environment, the conclusions obtained are limited, although it is evident that 100% of them showed positive specific IgE determination to profilins (rBet v 2, rMal d 4, rPhl p 12). The results pose only specificity problems with some grass pollen-allergic patients, who are also sensitized to profilins but present no symptoms after the ingestion of peach.

In conclusion, and according to the results of this study, we conclude that most peach-allergic patients coming from the north of Spain suffer from systemic symptoms after the ingestion of peach, Pru p 3 being the main allergen. In our series, the peach-allergic patients who present OAS show an intermediate sensitization pattern between the one presented in central and the one presented in south Europe, being the profilins the predominant allergen; furthermore 60% of them also present a sensitization to Bet v 1 or its homologues.

The future availability of recombinant or native allergens for in vitro and probably in vivo diagnosis of allergic diseases will provide tools which will improve the diagnostic reliability of these patients (25, 26). In the case of peach allergy, the possibility of determining specific IgE and/or BAT to Pru p 3 and to one of the profilins used and/or Bet v 1 homologues will allow confirmation, together with the clinical history, not only of peach allergy, but also of the responsible allergen(s). Thus, it will be able to explain the skin-systemic symptoms or OAS related to this LTP or profilin-Bet v 1 sensitization.


The finantial support was provided by Ministerio de Educación y Ciencia (DGI, BIO2006-07473).