To cite this article: Asarnoj A, Movérare R, Östblom E, Poorafshar M, Lilja G, Hedlin G, van Hage M, Ahlstedt S, Wickman M. IgE to peanut allergen components: relation to peanut symptoms and pollen sensitization in 8-year-olds. Allergy 2010; 65: 1189–1195.
Background: Allergen-specific IgE testing is often performed with crude peanut extract, but the results may be difficult to interpret because of cross-reactions between peanut and other plant allergens. The aim was to investigate IgE reactivity to peanut allergen components in children from a birch-rich region in relation to pollen sensitization and peanut symptoms.
Methods: From a birth cohort, clinical parameters were obtained through questionnaires and IgE antibody levels to peanut and birch pollen were measured. Different peanut/birch sensitization phenotypes were defined among 200 selected children. IgE reactivity to peanut and pollen allergen components was analysed using microarray technique.
Results: Peanut symptoms were reported in 87% of the children with IgE reactivity to any of the peanut allergens Ara h 1, 2 or 3 but not to Ara h 8 (n = 46) vs 17% of children with IgE reactivity to Ara h 8 but not to Ara h 1, 2 or 3 (n = 23), P < 0.001. Furthermore, symptoms were more severe in children with Ara h 1, 2 or 3 reactivity. Children with IgE reactivity both to Ara h 2 and to Ara h 1 or 3 more often reported peanut symptoms than children with IgE only to Ara h 2 (97%vs 70%, P = 0.016), particularly respiratory symptoms (50%vs 9%, P = 0.002).
Conclusions: IgE analysis to peanut allergen components may be used to distinguish between peanut-sensitized individuals at risk of severe symptoms and those likely to have milder or no symptoms to peanut if sensitized to pollen allergens and their peanut homologue allergens.
Cross-reactive carbohydrate determinant
Lipid transfer protein
Sodium dodecyl sulphate polyacrylamide gel electrophoresis
Peanut allergy has a prevalence between 0.6% and 1.8% (1, 2); it can be fatal (3–5) and is rarely outgrown (6). Therefore, the condition is not merely a diagnostic and therapeutic issue, but is also associated with decreased quality of life (7). However, many people who exhibit IgE antibodies to peanut report no symptoms at exposure (8). When such individuals are tested for peanut sensitization as part of a routine work up for other underlying allergic diseases, the presence of peanut-specific IgE antibodies may be confusing. Recently, we showed that Swedish children at school age with concomitant peanut and birch pollen sensitization reported fewer symptoms to peanuts than children with sensitization to peanut only (9). Peanut sensitization without associated symptoms has recently been observed among grass-sensitized individuals (10). We, therefore, wondered if this could be explained by differences in IgE reactivity to peanut allergen components within different groups of peanut-sensitized children.
The major peanut allergens Ara h 1, Ara h 2 and Ara h 3 are proteins considered responsible for the original sensitization to peanut in susceptible individuals (11). Ara h 2 is considered the most clinically important peanut allergen (12–15). Other peanut allergens show extensive IgE cross-reactivity between homologous allergens from various sources. Ara h 8 is homologous to the major birch pollen allergen Bet v 1 and may contribute substantially to birch pollen–peanut cross-reactivity (16). The profilins, exemplified by Ara h 5, are often involved in extensive IgE antibody cross-reactivity between various pollens and plant-derived foods (17, 18). Recently, a role for carbohydrate cross-reactive determinant (CCD) in the cross-reactivity between grass pollen and peanut has been implicated (10). Furthermore, IgE cross-reactivity between lipid transfer proteins (LTPs) in plant-derived foods (e.g. Ara h 9 in peanut and Pru p 3 in peach) has been reported (19).
The aim of this study was to investigate the IgE reactivity to different peanut, birch and grass pollen allergen components and CCD, in relation to symptoms to peanut in children from a Swedish birth cohort (BAMSE) at 8 years of age.
We included 4089 children (75% of the original target population) born 1994–1996 in Stockholm, Sweden, at birth in an unselected population-based birth cohort (the BAMSE cohort). Study design, enrolment, criteria for inclusion and procedures for data collection have been described elsewhere (20). Blood was drawn from 2480 of the children at 8 years of age. With a nested study design, two-hundred children representing four different patterns of sensitization to peanut and birch pollen were randomly selected to participate in the present study and were allocated into four equally sized groups: group A consisted of 50 of 52 children sensitized to peanut, but not to birch pollen; group B consisted of 50 of 141 children sensitized to both peanut and birch pollen; group C consisted of 50 of 237 children sensitized to birch pollen but not to peanut. Finally, group D consisted of 50 of 2012 children without sensitization to either peanut or birch pollen.
Symptoms of asthma, rhinitis and eczema used in the BAMSE cohort have been defined elsewhere (21). Data regarding symptoms from peanut were obtained from the questionnaire at 8 years of age prior to blood sampling. Among the 200 children, data on reported peanut symptoms were missing in two children at the 8 year follow-up. Permission for the study was obtained from the Ethics Committee of Karolinska Institutet. The parents of participating children gave informed consent for each follow-up.
Positive answer to the following question in the 8-year-questionnaire regarding the latest 12-month period: ‘Is your child allergic to any food item?’. Symptom options were ‘nose/eyes symptoms’, ‘mouth-itching’, ‘breathing problems’, ‘vomiting/diarrhoea’, ‘eczema’, ‘urticaria’. Reactions to peanut had to be indicated on at least one of these symptoms or reported as ‘excluded from the diet during the last 12 months because of previous symptoms’. Symptoms from nose/eyes or oral cavity were considered as mild symptoms, whereas wheeze or dyspnoea were considered as severe symptoms (22). Gastrointestinal symptoms, eczema and urticaria were not classified according to severity.
None of the symptoms mentioned earlier was reported for peanut by the parents when child had been eating peanut or peanut containing products during the last 12 months.
Serum samples were tested by ImmunoCAP® (Phadia AB, Uppsala, Sweden) for allergen-specific IgE antibodies to common inhalant and food allergens with Phadiatop® and the food mix fx5 test, respectively. Positive samples were further analysed for allergen-specific IgE antibodies to single food and airborne allergens, including peanut, birch and grass pollen, using extract-based ImmunoCAP tests. The measuring range was between 0.35 and 100 kUA/l. An IgE antibody concentration >100 kUA/l was given the value of 101 kUA/l in statistical evaluations.
IgE reactivity to allergen components from peanut (native Ara h 1, Ara h 2 and Ara h 3, and recombinant Ara h 8), birch pollen [recombinant Bet v 1 and Bet v 2 (profilin/Ara h 5 homologue)], timothy pollen [native Phl p 4 and recombinant Phl p 1, Phl p 5 and Phl p 12 (profilin/Ara h 5 homologue)], and peach [recombinant Pru p 3 (LTP/Ara h 9 homologue)], and to carbohydrate cross-reactive determinant (CCD) were measured using an in-house experimental semi-quantitative microarray assay based on a microspot multiplex technique (23). Native Ara h 1, Ara h 2 and Ara h 3 were affinity-purified using in-house developed monoclonal antibodies and were regarded as ∼99% pure based on SDS–PAGE and MALDI-TOF mass spectrometer analysis. All other allergen components including CCD were obtained from Phadia AB. The microarray method has recently been described by Constantin et al. (24). Results were calculated from triplicate spots. Individual cut-off levels for positive IgE antibody response for each allergen were established based on background fluorescence from negative samples spiked with up to 3000 kU/l of myeloma IgE (two times the mean background fluorescence with 3 standard deviations) and ranged between 400 and 600 FU. An IgE response higher than 5000 FU was regarded as a strong response.
Prevalence is given in total numbers and/or in percentage. Binominal exact probability was used for calculation of 95% confidence interval (95% CI) of categorical data. Background factors which may influence the development of allergen-specific IgE antibodies to peanut or birch pollen were tested to detect differences among children in this study and children in the cohort. Levels of allergen-specific IgE antibodies are expressed as geometric mean and 95% CI. Fisher’s exact test (two-tailed) was used for pairwise comparison of categorical data between groups. Spearman’s rank correlation test was used to establish the strength of relationship between specific IgE antibody responses in the microarray. P-values <0.05 were considered significant.
Comparison of children in this study compared to the study base of 4089 children
Children in the four study groups did not differ significantly in sex, parental allergy or prevalence of asthma, eczema, rhinitis or in levels of specific IgE (assessed using Phadiatop and fx5 test) at 8 years of age compared to remaining children with equal peanut- and birch sensitization of the study base or in important background factors of the whole cohort of 4089 children (data not shown).
Allergic sensitization within the 4 groups in relation to peanut allergen components
The four groups (A–D) differed substantially regarding the IgE reactivity to Ara h 1, 2, 3 and 8, CCD and profilin (Table 1). IgE reactivity to Ara h 1, 2 and 3 was most common in group A consisting of children sensitized to peanut but not to birch pollen. No child in group A had IgE reactivity to Ara h 8. IgE antibodies to Ara h 8 were common in children sensitized to birch pollen (group B and C; 38% and 22%, respectively). The prevalence of IgE reactivity to CCD was low in all groups, ranging from 0 to 18%. It was most prevalent in group B. Similarly, IgE reactivity to profilin (i.e. Bet v 2 and/or Phl p 12) was seen almost exclusively in group B (12%). None of the children in group A and B had IgE antibodies to LTP, i.e. Pru p 3.
|Group||A* (N = 50)||B* (N = 50)||C* (N = 50)||D* (N = 50)|
|Sensitization||n||% (95% CI)||n||% (95% CI)||n||% (95% CI)||n||% (95% CI)|
|Ara h 1||25||50 (36–64)||4||8 (2–19)||0||0||0||0|
|Ara h 2||35||70 (55–82)||18||36 (23–51)||0||0||0||0|
|Ara h 3||21||42 (28–57)||4||8 (2–19)||0||0||0||0|
|Ara h 8||0||0||19||38 (25–53)||11||22 (12–36)||0||0|
|Bet v 1||0||0||47||94 (83–99)||43||86 (73–94)||1||2 (0–11)|
|Profilin†||0||0||6||12 (5–24)||1||2 (0–11)||0||0|
|CCD||2||4 (0.5–14)||9||18 (9–31)||2||4 (0.5–14)||0||0|
Figure 1 shows IgE reactivity to Ara h 1, 2, 3 and 8, Bet v 1, profilin and CCD in relation to reported symptoms to peanut. No individual had IgE reactivity to Ara h 1 or Ara h 3 without an associated Ara h 2 sensitization. IgE reactivity to the peanut storage proteins (Ara h 1, 2 and 3) was present only in children sensitized to peanut (group A and B), whereas 11 children in group C (birch pollen sensitized only) had IgE reactivity to Ara h 8. Among the Ara h 8 sensitized children, all but two had low to moderate IgE levels to Ara h 8 (<5000 FU in the microarray assay). All (100%) sera with IgE reactivity to Ara h 8 also expressed high IgE levels to Bet v 1. Eight of 11 (73%) and 3 of 6 (50%) children with IgE reactivity to CCD and profilin, respectively, were negative for IgE to any of the tested peanut components (Ara h 1, 2, 3 and 8) in group A and B. Having IgE antibodies to CCD was instead strongly associated with IgE reactivity to Phl p 1 and/or Phl p 5 (P < 0.001), but not to Bet v 1 (P = 0.09).
When the IgE antibody levels to peanut were investigated in relation to individual IgE reactivity to the components, children with reactivity to Ara h 1, 2 or 3 but not to Ara h 8 had significantly higher IgE antibody levels (geometric mean, 95% CI) to peanut (18.9 kUA/l, 11.5–31.0 kUA/l, n = 46) than children with reactivity to Ara h 8 only (1.0 kUA/l, 0.60–1.7 kUA/l, n = 23).
Of the 100 peanut-sensitized children (group A and B), 25 were IgE negative to all tested peanut components in the microarray including CCD and profilin. In 17 of these 25 children, the peanut-specific IgE levels were below 1 kUA/l (Fig. 2). Seven of the negative children had responses to one or more peanut-related components in the microarray just below the cut-off level (data not shown).
Reported symptoms to peanut within the 4 groups and in relation to peanut allergen components
Symptoms to peanut were more common in peanut-sensitized children without concomitant birch pollen sensitization (group A: 74%, 95% CI 60–85%) than in children sensitized to both peanut and birch pollen: (group B: 43%, 95% CI 29–58%). In birch pollen-sensitized children without sensitization to peanut, reported symptoms to peanut was less common (group C: 8%, 95% CI 2–20%). None of the children sensitized neither to peanut nor to birch pollen (group D) reported symptoms to peanut (Fig. 1).
When children in groups A–D were regrouped according to whether they reported symptoms to peanut, 94% of those reporting peanut allergy, and 30% of the tolerant children, had peanut-specific IgE levels >0.35 kUA/l (Table 2). IgE reactivity to Ara h 1, 2 and 3 was much more common in children who reported peanut allergy than in tolerant children (P < 0.0001). IgE reactivity to Ara h 2 was most common, being found in 73% of the children with reported allergic symptoms to peanut (Table 2). Only seven of the peanut-tolerant children (5%) had IgE antibodies to Ara h 2, all but one with low/moderate IgE reactivity (<5000 FU) (Fig. 1).
|Reported symptoms to peanut||Yes (N = 62)||No (N = 136)||P-value|
|Sensitization||N||% (95% CI)||n||% (95% CI)|
|Peanut||58||94 (84–98)||41||30 (23–39)||<0.0001|
|Birch pollen||25||40 (28–54)||73||54 (45–62)||0.093|
|Ara h 1||28||45 (32–58)||1||1 (0–4)||<0.0001|
|Ara h 2||45||73 (60–83)||7||5 (2–10)||<0.0001|
|Ara h 3||24||39 (27–52)||1||1 (0–4)||<0.0001|
|Ara h 8||9||15 (7–26)||21||15 (10–23)||1.0|
|Bet v 1||25||40 (28–54)||64||47 (38–56)||0.44|
|Profilin†||2||3 (0.4–11)||5||4 (1–8)||1.0|
|CCD||2||3 (0.4–11)||11||8 (4–14)||0.35|
Thirty of the 53 Ara h 2-sensitized children had concomitant IgE reactivity to Ara h 1 or 3, and 97% of them reported peanut symptoms. The corresponding symptom prevalence among the 23 children with IgE reactivity only to Ara h 2 was 70% (P = 0.016). Respiratory symptoms were reported in 50% and 9% of the cases in the former (53 children) and latter (23 children) group, respectively (P =0.002).
Reported current symptoms to peanut, i.e. symptoms during the last 12 months, were investigated among the 46 children with IgE reactivity to Ara h 1, 2 or 3 but not Ara h 8 and the 23 children with IgE reactivity to Ara h 8 only and not to Ara h 1, 2 or 3 (Fig. 3). Eighty-seven per cent of those with Ara h 1, 2 or 3 IgE reactivity reported at least one symptom to peanut, whereas only 18% of those with IgE reactivity to Ara h 8 reported symptoms (P < 0.001). It is worth noting the trend towards a similar difference regarding oral cavity symptoms. However, the differences were significant only for upper and lower respiratory symptoms, possibly because of the limited number of children (Fig. 3). Eczema after exposure to peanut was not reported in any of the groups.
IgE reactivity to profilin (Bet v 2 and Phl p 12), any grass pollen allergen (including both major allergen components and whole extract) or CCD was not significantly associated with reported symptoms to peanut (data not shown).
In this study of children from a population-based birth cohort, we found that 83% of those with IgE reactivity to the peanut storage protein Ara h 2 also reported allergic symptoms to peanut, whereas only 18% of children with IgE reactivity exclusively to Ara h 8 reported symptoms. Furthermore, the symptoms of the latter group were considerably milder. Interestingly, IgE reactivity to Ara h 1 or Ara h 3 was always combined with IgE reactivity to Ara h 2, and was associated with reported peanut allergy in more than 95% of the children. Half of the children with reactivity to both Ara h 2 and Ara h 1 or 3 reported respiratory symptoms after exposure to peanut.
Specific IgE antibodies can be induced directly by allergens in food and often cause more severe reactions than cross-reacting IgE antibodies induced by pollen allergens (25, 26). The major allergens Ara h 1, Ara h 2 and Ara h 3 are considered as more clinical relevant peanut allergens. They are highly resistant to heating (roasting) and enzymatic degradation and are therefore potentially dangerous allergens (27–29). Previous studies report a prevalence of IgE reactivity to Ara h 2 in allergic individuals of about 80-100% (12–15). In our nested study, the prevalence was 73%. The slightly lower prevalence might be explained by differences in age but is most likely because of the selection of study subjects. The children in the BAMSE cohort were recruited shortly after birth on a population basis, whereas the subjects from the other studies were all recruited from clinics (12, 14, 15).
IgE reactivity to Ara h 8, on the other hand, was only seen among children sensitized to birch pollen, and almost exclusively in sera with strong IgE response to the major birch pollen allergen Bet v 1. Furthermore, a number of birch pollen-sensitized children in group C had IgE reactivity to Ara h 8 without positive results to peanut in ImmunoCAP, reflecting the fact that peanut extract usually contains low amounts of Ara h 8 (16). The lack of correlation in our study between IgE reactivity to the birch pollen-related Ara h 8 and reported symptoms to peanut, is in accordance with the finding that co-sensitization to birch pollen and peanut less often is associated with peanut allergy than sensitization to peanut only (9). Our results are in line with previous findings that co-sensitization to peanut and timothy grass is related to milder, if any, clinical symptoms to peanut (30). A speculative explanation for the low prevalence of symptoms to peanut among children mono-sensitized to Ara h 8 is that most IgE-binding epitopes on Ara h 8 are destroyed during the roasting process or by gastric digestion (16).
Ten peanut-sensitized children without IgE antibodies to any of Ara h 1, 2, 3 or 8 had IgE antibodies to CCD and/or profilin, indicating a role for these two highly cross-reactive components in peanut sensitization. However, only one of these ten children reported any symptoms to peanut. Our results agree with another study showing a role for CCD in grass pollen-allergic patients with positive test results to peanut (10).
In our material, we did not find any IgE reactivity to Pru p 3, probably because all the children had lived in Sweden since birth. To measure IgE reactivity to Ara h 9 might have been more appropriate in our study, even if IgE reactivity to LTP is more common in the Mediterranean area (31, 32).
Twenty-five of the 100 peanut-sensitized 8-year-old children (defined as having a peanut-specific IgE level above 0.35 kUA/l) did not show IgE reactivity above the microarray cut-off to any peanut-related component in our study. However, only eight of them exhibited IgE levels to peanut above 1 kUA/l in ImmunoCAP, and only nine reported symptoms to peanut, generally mild ones. A possible explanation is that they might have IgE antibodies to other relevant peanut allergens not present on the microarray, i.e. the 2S albumins Ara h 6 and Ara h 7 (18), or peanut oleosin (33). Furthermore, the detection limit of the microarray might not have been low enough to identify all peanut-specific IgE antibodies at concentrations close to the ImmunoCAP cut-off at 0.35 kUA/l. This bias in our material would only tend to dilute our results, not the opposite.
As we have only studied 8-year-old children, we cannot exclude the possibility that our results would have been slightly different if other age groups had been included. In two smaller studies of selected patients between 3–20 years, the importance of IgE to Ara h 1, 2 and 3 has been demonstrated (13, 14). However, our study is the first investigation of IgE reactivity to peanut components in relation to symptoms and birch pollen sensitization in a population-based cohort from a birch-tree-rich region employing a nested design, i.e. with randomly selected children within a cohort. As our results are in agreement with the ones by Flinterman et al. and Astier et al. on IgE reactivity to Ara h 1, 2 and 3 these results are likely to be significant and transferable to the population of children at school age.
One weakness of this study is that no oral challenges were performed to verify clinical allergy and types of symptoms. However, it is worth noting that the likelihood of reported symptoms to peanut in relation to quantitative IgE levels to peanut in our BAMSE cohort closely resembles the pattern seen in children whose peanut allergy has been confirmed by oral challenge (9, 13, 14, 34). As over-reporting of peanut symptoms is the most likely potential error (35), we suggest that the parental reported peanut ‘tolerant’ children, i.e. no report of symptoms to peanut among children who eat peanut, really are tolerant. A strength in this context is that symptoms were reported prior to blood sampling and analysis of peanut-specific IgE as well as IgE to peanut allergen components.
In conclusion, we consider the use of native and recombinant allergen components as a major step forward in the diagnosis of IgE-mediated diseases and in the study of the mechanisms behind allergy. Knowledge about component-specific IgE profiles in peanut-sensitized individuals, in particular when pollen sensitized, can provide a tool to distinguish clinical relevant and potentially dangerous peanut allergy from peanut sensitization without clinical relevance. It remains to be seen whether this will reduce the need for oral peanut challenges.
We thank all the participating families and all the staff working within the BAMSE project as well as laboratory collaborators at Phadia AB.
Sources of funding
Supported by the Swedish Asthma and Allergy Foundation, the Swedish Heart and Lung Foundation, the Vardal Foundation for Health Care Sciences and Allergy Research, Stockholm County Council, The Swedish Research Council, Sweden.