SEARCH

SEARCH BY CITATION

Keywords:

  • bakers;
  • farmers;
  • natural rubber latex;
  • occupational asthma;
  • skin prick test

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References

Background

Skin prick testing (SPT) is an important step in the diagnosis of IgE-mediated occupational allergic diseases. The outcome of SPT is related to the quality of allergen extracts. Thus, the aim of the study was to assess different commercially available SPT solutions for selected occupational allergens.

Methods

SPT was performed in 116 bakers, 47 farmers and 33 subjects exposed to natural rubber latex (NRL), all with work-related allergic symptoms. The SPT solutions from different manufacturers (n = 3–5) for wheat flour, rye flour, soy, cow hair/dander, storage mites (Tyrophagus putrescentiae, Lepidoglyphus destructor, Acarus siro) and NRL were analysed with respect to their protein and antigen contents. SPT was carried out in 16 allergy centres in six European countries using standardized procedures. Specific IgE values were used as the gold standard to calculate the sensitivity and specificity of SPT solutions. The optimal cut-point for each SPT solution was determined by Youden Index.

Results

Protein and antigen contents and patterns of the SPT solutions varied remarkably depending on the manufacturer. While SPT solutions for wheat flour and soy reached overall low sensitivities, sensitivities of other tested SPT solutions depended on the manufacturer. As a rule, solutions with higher protein and antigen content showed higher sensitivities and test efficiencies.

Conclusions

There is a wide variability of SPT solutions for occupational allergens, and the sensitivity of several solutions is low. Thus, improvement and standardization of SPT solutions for occupational allergens is essential.

The diagnosis of IgE-mediated occupational allergic diseases is often difficult, because the identification of causative agents out of numerous exposures at the workplace using the common diagnostic tools and procedures is a complex issue. The usual diagnostic steps are a detailed case history, skin prick testing (SPT), in vitro diagnosis and specific inhalation challenge. In this workup, SPT in combination with work-related symptoms of the patient is relevant for compensation and further socioeconomic consequences. Therefore, guidelines for allergy diagnosis recommend using standardized SPT solutions [1-3].

Many efforts were made to standardize SPT solutions [4]. Two different systems of biological standardization were developed in the United States [5] and in Scandinavia [6]. However, there are considerable differences between these approaches. Additionally, the Nordic guidelines proposed that allergens other than common inhalant allergens should not be subjected to biological standardization [7]. Also, the EU-funded CREATE project that aimed for major allergen-based standardization did not cover occupational allergens [8]. Although some general rules for allergen standardization stated in the European Pharmacopoeia have to be followed, these are relatively vague and interpretation varies among the different countries. Thus, SPT solutions for the same occupational allergen but from different manufacturers are not necessarily comparable. The latter has been described recently for wheat and rye flour SPT solutions from German manufacturers [9, 10].

In the present study, SPT solutions from European manufacturers for a number of occupational allergens (wheat flour, rye flour, soy, cow hair/dander, storage mites, natural rubber latex (NRL)) were analysed in vitro and in vivo. Allergologists from six European countries were involved in the data collection for this multicentre study labelled STADOCA (Standard diagnosis for occupational allergy).

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References

Study design and subjects

In the STADOCA multicentre study, European allergologists from 15 allergy centres (seven German, three Italian, two Spanish, one French, one Austrian and one Polish) were involved in the data collection. Data from 196 persons exposed to occupational allergens were included. All patients suffering from work-related allergic symptoms were examined within the scope of claims for compensation due to occupational asthma. While SPTs and challenge tests were performed in the different centres, specific IgE (sIgE) measurements were carried out centrally in one facility (IPA). The study design and the protocol were reviewed and approved by the ethics committee of the Ruhr-University Bochum in accordance with the Declaration of Helsinki.

Bakers

The study included 116 bakers (81% males) with a mean age of 38.3 ± 12.0 years, 31 (26.7%) were current smokers, and 20 (17.2%) ex-smokers. While 56 bakers (48.3%) were still working at the time of the study, 60 (51.7%) had left the job due to allergic symptoms. Eighty-nine subjects (76.7%) complained of asthma, and of these, 86 also reported rhinitis. Twenty-five bakers (21.6%) reported rhinitis, but no symptoms of asthma. Based on their history, 56 bakers (48.3%) were atopic.

Farmers

Forty-seven farmers (66% males) exposed to animals and/or grains were examined. On average, they were 48.7 ± 11.2 years old, and only six of them (12.8%) were current smokers and seven (14.9%) ex-smokers. About 80% of farmers were still working at the time of the study. Twenty-eight subjects (59.6%) reported asthma, and in 24 cases (51.1%) in combination with rhinitis. Nine farmers (19.1%) reported rhinitis, but no asthma, and 41 (87.2%) suffered from cough. Twenty-four farmers (51.1%) were atopic based on case history.

Latex-exposed subjects (in the following termed ‘health care workers (HCWs)’)

Thirty-three persons (12% males) occupationally exposed to NRL were included in the study. Most were health care workers (HCWs, 72.7%), four were hairdressers (12.1%), and five had other professions (e. g. balneotherapist, cleaner). Their mean age was 40.7 ± 10.7 years; six (18.2%) were current smokers, and three (9.1%) ex-smokers. Of 10 patients (30.3%) suffering from asthma, eight also reported rhinitis, whereas twelve reported symptoms of rhinitis without asthma. Skin problems due to NRL were reported by 28 (84.8%) subjects.

In vitro analysis of SPT solutions

SPT solutions (Table 1) for wheat and rye flour, soy, cow hair/dander, storage mites and NRL were obtained from seven European manufacturers: ALK-Abelló, Hørsholm, Denmark; Allerbio, Varennes-en-Argonne, France; Allergopharma, Reinbek, Germany; Bencard, Munich, Germany; Hal, Duesseldorf, Germany; Lofarma, Milan, Italy; and Stallergènes, Antony, France.

Table 1. Test panels of commercial SPT solutions for different professions
Bakers' SPT panelFarmers' SPT panelHCWs' SPT panel
Wheat and rye flour (Wf, Rf)CowNRL
Wf/Rf1: BencardC1: Allergopharma (dander)NRL1: Allerbio
Wf/Rf2: ALK-AbellóC2: Bencard (hair)NRL2: ALK-Abelló
Wf/Rf3: LofarmaC3: ALK-Abelló (hair)NRL3: Allergopharma
Wf/Rf4: StallergènesC4: Hal (dander)NRL4: Stallergènes
C5: Lofarma (dander)NRL5: Lofarma
SoyStorage mites
S1: AllerbioTyrophagus putrescentiae (Tp)
S2: StallergènesLepidoglyphus destructor (Ld)
S3: ALK-AbellóAcarus siro (As)
Tp/Ld/As1: Allergopharma
Tp/Ld/As2: Stallergènes
Tp/Ld/As3: ALK-Abelló

The protein concentrations of SPT solutions were measured by Bradford assay [11]. Corresponding antigens were quantified by sandwich enzyme-linked immunosorbent assays (ELISAs) based on rabbit IgG for wheat flour [12], rye flour [9], cow hair [13], Tyrophagus putrescentiae, Lepidoglyphus destructor and Acarus siro [14]. Allergen concentrations of NRL SPT solutions were measured by ImmunoCAP inhibition [15]. 10 μl of each SPT solution was separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) using a precast NuPAGE® Bis–Tris gel (Invitrogen, Groningen, The Netherlands), followed by silver staining.

SPT

A test panel with defined SPT solutions from different manufacturers for each profession was created (Table 1) and provided to each participating centre. The SPT panel for bakers included flour and soy. Farmers' panel included cow hair/dander and three different storage mites, whereas subjects occupationally exposed to NRL were pricked in parallel with five NRL SPT solutions. Histamine (10 mg/ml) and saline solution (0.9%) were used as positive and negative controls, respectively.

In each allergy centre, identical batches of SPT solutions were pricked in accordance with the European position papers [1, 4]. As described earlier [9], SPTs were performed in a predetermined order, twice in opposite direction on both volar forearms, and contours of wheals were transferred with transparent tape. Assessment of SPT results was made centrally at IPA by a single person (VvK). The optimal cut-point for evaluation was chosen by Youden Index (see statistical analysis).

Whereas all 116 bakers and 33 HCWs were tested on both forearms with all SPT solutions from the bakers' and the HCWs' panels, respectively, only 31 farmers were tested in parallel with SPT solutions for cow and for storage mites. Twelve additional farmers were exclusively pricked with cow SPT solutions and four with storage mite SPT solutions.

Specific IgE determination

Specific IgE antibodies (sIgE) to wheat and rye flour, soybean, cow dander, Tyrophagus putrescentiae, Lepidoglyphus destructor, Acarus siro and NRL were measured by ImmunoCAP (Thermo Fisher Scientific formerly Phadia, Uppsala, Sweden) according to the manufacturer's recommendations. CAP values ≥0.35 kU/l were considered positive.

Specific challenge test

Depending on anamnestic data and on the decision of the medical examiner, specific challenge tests with allergens were performed in the different allergy centres. The information whether the challenge test was performed nasal, bronchial or as workplace simulation as well as the test result (positive or negative) was given to IPA by the examiner. Seventy bakers were challenged with wheat flour and 54 with rye flour.

Statistical analyses

For the evaluation of SPT solutions, the mean of both SPT determinations was used. Specific IgE results were taken as gold standard because challenge tests were not performed in all cases. True positives (tp) were subjects with positive sIgE and positive SPT; true negatives (tn) were subjects with negative sIgE and negative SPT; false positives (fp) were subjects with negative sIgE and positive SPT; false negatives (fn) were subjects with positive sIgE and negative SPT. Two-by-two tables were used to calculate sensitivities [tp/(tp + fn)], specificities [tn/(tn + fp)], positive (PPV) [tp/(tp + fp)] and negative [tn/(tn + fn)] predictive values (NPV) as well as test efficiencies [(tp + tn)/(tp + fp + tn + fn)]. For the evaluation of SPT results, the wheal size that leads to the maximum Youden Index (sensitivity + specificity-1) was chosen as optimal cut-point. Youden Index considering equally sensitivity and specificity was used because this index should be independent of selection bias. Receiver-operating characteristic (ROC) curves were constructed to obtain the area under the curve (AUC), which is a criterion for test quality.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References

In vitro analysis of the SPT solutions demonstrated high variability in protein and antigen content between SPT solutions from different manufacturers (Table 2). Also, in SDS PAGEs, there were differences in intensity and numbers of protein bands (Fig. 1).

Table 2. Protein and antigen contents of commercial SPT solutions for different professions [Correction made after online publication on 20 March 2013: units of protein content changed to μg/ml.]
SPT solutionProtein content [μg/ml]Antigen content [U/ml]
Bakers
Wf1538706
Wf2411959
Wf319188
Wf44961845
Rf1495419
Rf249162
Rf318321
Rf410372721
S11114n.d.
S21364n.d.
S31631n.d.
Farmers
C1433
C27225
C312177
C4436313
C56591439
Tp186109
Tp2b.d.37
Tp35881597
Ld178193
Ld2b.d.15
Ld3205266
As190154
As2b.d.34
As3192429
HCWs  Allergen content [μg/ml]
  1. b.d., below detection limit; n.d., not done.

NRL1658.3
NRL2554.0
NRL34411.7
NRL4b.d.2.5
NRL5b.d.0.6
image

Figure 1. Equal volumes of SPT solutions were separated using SDS-PAGE and silver staining. (A) Wheat flour (Wf), (B) rye flour (Rf), (C) soy (S), (D) cow hair/dander (C), (E) Tyrophagus putrescentiae (Tp), (F) Lepidoglyphus destructor (Ld), (G) Acarus siro (As), (H) natural rubber latex (NRL). Manufacturers: see Table 1.

Download figure to PowerPoint

For most of the tested allergens, the number of subjects reacting in SPT (mean wheal size >0 mm) varied according to the manufacturer of the SPT solution (Table 3). There were no significant differences between SPT results obtained with the same solution at different positions (near wrist or antecubital fossa; data not shown). Approximately 58% of the bakers showed a wheal reaction with the rye flour SPT solution Rf4; however, only 19% of bakers showed a reaction with Rf3. Further remarkable differences were seen between the SPT solutions Wf3 and Wf4, C1 and C5, Tp2 and Tp3, Ld2 and Ld3 as well as As2 and As3.

Table 3. SPT reactions and maximum wheal sizes obtained with commercial SPT solutions for different professions. Youden Index of the SPT solutions at different cut-points was calculated with the mean wheal size according to the gold standard sIgE
SPT solutionMean wheal size >0 mmn (%)Maximum wheal size [mm]Youden Index at cut-pointb
≥1.5 mm≥2.0 mm≥2.5 mm≥3.0 mm
  1. a

    In six farmers, SPT was not performed twice. Here, the single values were considered.

  2. b

    For each SPT solution, results with highest Youden Index are indicated in boldface.

Bakers (n = 116)
Wf139 (33.6%)9.5 0.419 0.3700.3580.316
Wf246 (39.7%)9.0 0.382 0.3760.3050.308
Wf333 (28.5%)9.0 0.313 0.2850.2880.254
Wf450 (43.1%)9.5 0.489 0.4540.4120.401
Rf148 (41.4%)9.5 0.462 0.4070.4040.376
Rf232 (27.6%)8.0 0.336 0.3280.3070.227
Rf322 (19.0%)6.5 0.164 0.1240.1110.109
Rf467 (57.8%)13.0 0.691 0.6910.6620.674
S128 (24.1%)7.50.2640.287 0.298 0.269
S231 (26.7%)7.50.253 0.275 0.2130.161
S327 (23.3%)11.0 0.153 0.1390.1240.087
Farmers (n = 43a)
C15 (13.5%)6.0 0.467 0.4670.4670.384
C210 (27.0%)10.0 0.634 0.5500.4670.500
C313 (35.1%)10.0 0.733 0.7330.6830.683
C414 (37.8%)11.5 0.850 0.7500.7500.750
C515 (40.5%)11.50.750 0.817 0.6500.650
Farmers (n = 35a)
Tp18 (27.6%)11.0 0.769 0.6980.6980.698
Tp26 (20.7%)5.50.405 0.445 0.4450.333
Tp311 (37.9%)22.0 0.689 0.6890.6890.689
Ld111 (37.9%)12.00.840 0.880 0.8800.658
Ld25 (17.2%)9.00.6980.6980.738 0.778
Ld312 (41.4%)11.00.7200.760 0.800 0.729
As18 (27.6%)9.00.4330.4330.433 0.510
As23 (10.3%)3.50.462 0.500 0.5000.500
As311 (37.9%)8.50.6050.6450.683 0.721
HCWs (n = 33)
NRL111 (33.3%)14.50.7220.764 0.847 0.847
NRL212 (36.4%)10.00.7220.7640.764 0.806
NRL312 (36.4%)13.00.7220.7220.764 0.806
NRL412 (36.4%)14.5 0.806 0.8060.8060.806
NRL510 (30.3%)8.00.542 0.584 0.4730.514

To identify the optimal cut-point for the different SPT solutions, the Youden Index was calculated for SPT solutions at four different wheal sizes. As shown in Table 3, for all tested flour SPT solutions, the maximum Youden Index was reached at small wheal diameters (≥1.5 mm). This was also true for 80% of SPT solutions for cow hair/dander (C1, C2, C3, C4) and two-thirds of Tp SPT solutions (Tp1, Tp3). For the other allergens, the highest Youden Index was reached at different wheal sizes (Table 3). Altogether, even the highest Youden Index of the SPT solutions for one allergen varied between manufacturers, especially for the SPT solutions Rf3 and Rf4 (0.164 vs 0.691).

Seventy-one of 116 bakers (61.2%) showed sIgE to wheat flour (range 0.38–82.6 kU/l, median 2.52 kU/l), 75 (64.7%) to rye flour (range 0.36–91.5 kU/l, median 2.18 kU/l) and 27 (13.3%) to soybean (range 0.37–17.0 kU/l, median 0.68 kU/l). Twelve of the 43 farmers (27.9%) showed cow-sIgE (range 0.40–23.3 kU/l, median 3.20 kU/l), while nine of 35 (25.7%) were positive to Tyrophagus putrescentiae (range 0.40–21.5 kU/l, median 3.24 kU/l) and Lepidoglyphus destructor (range 0.40–32.4 kU/l, median 2.78 kU/l), and eight (22.9%) to Acarus siro (range 0.46–15.7 kU/l, median 2.52 kU/l). Nine of the 33 HCWs (27.3%) had sIgE to NRL (range 0.54–42.4 kU/l, median 1.66 kU/l).

Evaluation of sensitivities, specificities, test efficiencies as well as positive (PPV) and negative predictive values (NPV) was performed using the optimal cut-point of each SPT solution (Table 4). SPT solution specificities were between 80% and 100%; however, sensitivities, test efficiencies, PPVs and NPVs showed partially extreme discrepancies. SPT solutions for some allergens such as wheat flour and soy reached overall low sensitivities (all ≤58%). In contrast, sensitivities of SPT solutions for rye flour, cow hair/dander and storage mites varied depending on the manufacturer. With one exception, NRL SPT solutions reached a comparably high quality. In this context, it should be mentioned that one HCW with positive sIgE to NRL, who reacted with none of the NRL SPT solutions, reported wasp venom allergy and had sIgE to horseradish peroxidase (HRP) containing crossreactive carbohydrate determinants (CCDs).

Table 4. Evaluation of different SPT solutions based on the gold standard sIgE
SPT solutionSensitivityaSpecificityaTest efficiencyaPPVaNPVaAUCb
  1. PPV, positive predicted value; NPV, negative predicted value; AUC, area under curve.

  2. a

    Evaluation was performed using the optimal cut-point of each SPT solution (see Table 3).

  3. b

    Data were obtained by ROC (receiver-operating characteristic) curves.

Bakers (n = 116)
Wf10.4790.9330.6550.9190.5320.726
Wf20.4930.8890.6470.8750.5260.719
Wf30.3800.9330.5950.9000.4880.672
Wf40.5780.9110.7070.9110.5780.767
Rf10.5600.9020.6810.9130.5290.759
Rf20.3600.9760.5780.9640.4550.697
Rf30.2130.9510.4740.8890.3980.609
Rf40.8130.8780.8360.9240.7200.872
S10.4440.8540.7590.4800.8350.638
S20.4440.8310.7410.4440.8310.630
S30.3330.8200.7070.3600.8020.607
Farmers (n = 43)
C10.5000.9670.8330.8570.8290.774
C20.6670.9670.8810.8890.8790.903
C30.8330.9000.8810.7690.9310.925
C40.9170.9330.9290.8460.9660.944
C50.9170.9000.9050.7860.9640.933
Farmers (n = 35)
Tp10.8890.8800.8820.7270.9570.924
Tp20.4451.0000.8531.0000.8330.689
Tp30.8890.8000.8240.6150.9520.922
Ld11.0000.8800.9120.7501.0000.973
Ld20.7781.0000.9411.0000.9260.880
Ld31.0000.8000.8530.6431.0000.964
As10.6250.8850.8260.6250.8850.757
As20.5001.0000.8821.0000.8670.731
As30.8750.8460.8530.6360.9570.901
HCWs (n = 33)
NRL10.8890.9580.9390.8890.9580.924
NRL20.8890.9170.9090.8000.9570.921
NRL30.8890.9170.9090.8000.9570.917
NRL40.8890.9170.9090.8000.9570.919
NRL50.6670.9170.8480.7500.8800.796

Using not only the optimal cut-point, but all measured wheal sizes, ROC plots were performed for all SPT solutions. Resulting AUCs were also different between the SPT solutions (Table 4). ROC curves for rye flour, soy, cow hair/dander and NRL are exemplarily shown in Figure 2.

image

Figure 2. ROC (receiver-operating characteristic) curves including area under the curve (AUC) for the rye flour (Rf), soy (S), cow (C) and natural rubber latex (NRL) SPT solutions. Evaluation was performed using sIgE as gold standard. Manufacturers: see Table 1.

Download figure to PowerPoint

Evaluation of SPT results in a subgroup of 70 (wheat flour) and 54 (rye flour) bakers in relation to the two gold standards, sIgE and challenge test, resulted in the same ranking of flour SPT solutions (data not shown). Thus, the great differences between the flour SPT solutions were independent of the gold standard.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References

In vitro analysis of 30 SPT solutions for occupational allergens showed striking differences in protein and antigen content and the pattern of protein bands. As a rule, the allergen solutions with higher antigen concentrations gave higher numbers of positive test results and greater wheal sizes.

It could be shown for flour and also for some other SPT solutions that already wheal diameters ≥1.5 mm displayed the highest Youden Index. Although such a small wheal size is unusual as cut-off, Bousquet et al. [16] confirmed our finding by showing that in epidemiological studies, a cut-off level >0 mm is the most appropriate definition of positive SPT to assess allergic sensitization.

As shown for Youden Index, sensitivities, test efficiencies, PPVs, NPVs and AUCs also showed partially high discrepancies between SPT solutions, whereas specificities were overall greater than 80%. In this context, it should be noted that the ranking of the SPT solutions depends on which of the above-mentioned parameters is considered. Whereas predictive values (PPV and NPV) depend largely on the prevalence rates of cases under study, sensitivity and specificity are presumably inherent properties of the test. For some extracts, there is, for example, a discrepancy between the results of PPV and the Youden Index. This is due to the fact that PPV may be enhanced by increasing the specificity while Youden Index considers both sensitivity and specificity equally. Because test efficiency and AUC also considered both sensitivity and specificity, the ranking of these parameters is related to Youden Index.

While none of the tested SPT solutions for wheat flour and soy showed a consistently high quality compared to sIgE results, qualities of other tested SPT solutions depended on the manufacturer.

We are aware that specific IgE test is not the universal gold standard for allergic sensitization and that problems with the quality of the allergen extracts for sIgE test could be the same as for SPT and also for challenge test. However, in the present study, the number of ‘false-positive’ results (positive SPT, negative sIgE) was very small, and according to our experience, the sIgE test we used is sensitive. In a former study [9], it could be shown that in relation to the gold standard challenge test, the sIgE test and the SPT – when using a potent SPT solution – showed comparable sensitivities. Because it was the aim of the study to compare the quality of different commercial SPT solutions not only in vitro but also in vivo, a consistent standard was necessary. Therefore, the sIgE test was used as gold standard, because the sIgE results were available for each patient and for each relevant allergen.

Comparative studies of SPT solutions from different manufacturers have rarely been performed, especially with occupational allergens. In a paediatric study, sensitivities of three commercial NRL SPT solutions varied from 65% to 96% [17]. Recently, it could be shown that diagnostic extracts of Dermatophagoides pteronyssinus used in Europe and Mexico are less potent than those used in the United States [18]. Our data indicate that solutions with higher protein and antigen content showed higher sensitivity and test efficiency. Thus, SPT solutions should contain all relevant allergens in a sufficient amount. However, regarding the protein content, it has to be guaranteed that no nonallergenic proteins like serum albumin had been added to the SPT solutions, which was sometimes done for stabilization [9].

The observed differences between the flour SPT solutions appeared to be independent of the gold standard, as previously shown [9].

In contrast to SPT solutions for bakers and farmers, with one exception, NRL SPT solutions for the diagnosis of latex allergy showed concordantly good quality. The sensitivities would have reached 100% upon exclusion of one patient suffering from wasp venom allergy with sIgE to NRL and HRP who reacted with none of the NRL SPT solutions. It is known that insect venoms and NRL share IgE-binding CCDs [19, 20]. Therefore, in those cases, supplementary screening tests with at least one CCD-containing allergen should be performed [21, 22].

In conclusion, our data could help in the selection of the most suitable SPT solution for the diagnosis of occupational allergies. Unfortunately, we were not able to analyse all potential occupational allergens from all potential manufacturers. Additionally, we cannot exclude batch-to-batch variability among SPT solutions. However, there is a need for allergen standardization and improvement of SPT solutions for occupational allergens.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References

This study was supported by the DGUV (German Social Accident Insurance; project IPA-60-STADOCA), St. Augustin, Germany. We thank Yvonne von der Gathen for the quantification of NRL allergens and Ursula Meurer, Sabine Eckel and Anne Flagge for excellent technical assistance.

Conflict of interest

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References

All authors have no conflict of interest to declare.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References