Cockroach allergen component analysis of children with or without asthma and rhinitis in an inner-city birth cohort

Abstract

Background:

Cockroach is one of the most important sources of indoor allergens that can lead to IgE sensitization and to the development of rhinitis and asthma.

Objective:

To perform a cockroach allergen component analysis to determine the allergens and antibody levels and patterns of sensitization that are associated with asthma and rhinitis.

Methods:

Antibody levels (IgE, IgG and IgG4) to total cockroach and to eight cockroach allergens were determined in two groups of cockroach sensitized 10-year old children with (n = 19) or without (n = 28) asthma and rhinitis. Allergen-specific antibody levels were measured in streptavidin ImmunoCAPs loaded with each of the recombinant allergens from groups 1, 2, 4, 5, 6, 7, 9, and 11, and total cockroach-specific IgE was measured using the i6 ImmunoCap.

Results:

IgE antibody levels to cockroach allergens and extract, but not IgG or IgG4, differed between subjects with and without asthma and rhinitis. Specifically, recognition of more cockroach allergens, with higher allergen-specific IgE, were associated with disease. Variable patterns of sensitization, with no immunodominant allergens, were found in both groups. There was a good correlation between the sum of allergen-specific IgE and total cockroach IgE (r = 0.86; p <0.001).

Conclusions:

Component analysis of eight cockroach allergens revealed significant differences in IgE reactivity associated with the presence of asthma and rhinitis. Allergen-specific IgE titers and sensitization profiles were associated with asthma and rhinitis.

Graphical Abstract

CAPSULE SUMMARY:

Cockroach allergic children with asthma and rhinitis have higher IgE (but not IgG or IgG4) levels and are sensitized to more allergens than cockroach allergic children without disease.

INTRODUCTION

Cockroach allergy is an important health problem that can lead to the development of asthma and/or rhinitis. In the US, the main species of cockroach that causes allergy is the German cockroach Blattella germanica. Exposure and sensitization to cockroach allergens are two factors strongly associated with high morbidity among inner-city children with asthma.⁽¹⁾ The profile of immunoglobulin E (IgE) sensitization to cockroach allergens is unique for each patient.^(2;3) Unlike allergies to cat or mite, that are mostly associated with sensitization to the major allergens Fel d 1 or Der p 1 and Der p 2, respectively, no immunodominant allergens have been found for cockroach. Bla g 2 has been considered the major cockroach allergen in the U.S., followed by Bla g 5, with prevalences of IgE sensitization of 42–70% and 35–68%, respectively.^(2;4;5) In the last 10 years, new allergens have emerged from cockroach, classified in thirteen groups currently listed in the Allergen Nomenclature database of the World Health Organization and International Union of Immunological Societies (WHO/IUIS) (www.allergen.org). However, the relevance of sensitization to a large set of cockroach allergens for disease has not been deeply investigated. In the current study, sensitization to eight groups of cockroach allergens was analyzed among cockroach sensitized patients with disease (asthma and rhinitis) or without disease. The cockroach allergen groups are: 1 (gut microvilli-associated proteins), 2 (gut inactive aspartic proteases), 4 (lipocalins produced only in males and excreted in the spermatophore during copulation), 5 (glutathione S-transferases), 6 (troponins C involved in muscle contraction), 7 (tropomyosins), 9 (arginine kinases) and 11 (α-amylases).^(4–17) These allergens were selected based on their importance, which has been recently described.⁽³⁾

One of the major goals of the Inner-City Asthma Consortium (ICAC), sponsored by the National Institute of Allergy and Infectious Diseases (NIAID), is to develop an immune-based therapeutic approach to asthma by targeting cockroach allergy.⁽¹⁸⁾ It is known that the allergen content of Bla g 1 and Bla g 2 in cockroach extracts is variable.^(19;20) Recently, up to 20-fold, 728-fold and 12-fold differences for Bla g 1, Bla g 2 and Bla g 5, respectively, have been reported in cockroach extracts.⁽³⁾ The potency of the extracts was found to be dependent not only on the allergen content of the extract but also on the subject sensitization profile.⁽³⁾ Additionally, new major allergens were identified in a small group of subjects in the U.S.⁽³⁾ It is essential to evaluate the sensitization patterns to this new large set of cockroach allergens to assess their relevance to disease. In this study, a component analysis of the antibody reactivity (IgE, IgG, and IgG4) to eight cockroach allergens was performed in two groups of cockroach-sensitized children participating in the ICAC Urban Environment and Childhood Asthma (URECA) birth cohort study. One group included subjects with asthma and rhinitis whereas the subjects in the other group did not have these diseases. The goal was to identify the main cockroach allergens and to evaluate differences in allergen-specific patterns of sensitization between individuals with and without asthma and rhinitis. The analysis performed here is important for designing clinical trials and ensuring that patients are treated with cockroach extracts that contain allergens relevant to their disease.

METHODS

Study population

The Urban Environment and Childhood Asthma (URECA) birth cohort has been previously described.⁽²¹⁾ This observational birth cohort of n = 560 infants was studied after recruitment of pregnant mothers from four research centers (Johns Hopkins University, Baltimore; Boston University, Boston; Columbia University and Mt. Sinai University, New York City and Washington University, St. Louis). Inclusion criteria were as follows: residence in a census tract with at least 20% of residents below the poverty level, at least one parent with a history of allergic rhinitis, eczema and/or asthma, and birth of the enrolled child at ≥34 weeks gestation. The Human Subjects Committees at the University of Wisconsin and the four clinical sites approved the study. The total number of URECA cohort participants at age 10 years, when the samples were obtained for the current study was n = 442 children, the same number as at 7 years of age.⁽²²⁾

The population selected consisted of cockroach sensitized 10-year old children from the URECA cohort with either positive cockroach-specific IgE by streptavidin ImmunoCAP and/or positive cockroach-specific skin prick test. These 78 cockroach-sensitized patients were subdivided into four groups: 1) no asthma, no rhinitis (28/78: 35.9%), 2) rhinitis only (18/78: 23.1%), 3) asthma only (13/78: 16.7%), and 4) asthma and rhinitis (19/78: 24.4%). The groups 1 and 4 were selected for component analysis (n = 47), since their most different phenotypes were expected to facilitate the observation of differences in the component analysis. They were divided into two groups: children with rhinitis and asthma (n = 19) or without disease (n = 28). Classification of asthma (asthma symptoms or asthma medication use in the prior year) and rhinitis (presence of runny and/or stuffy nose, and frequent sneezes in the prior year) for the purpose of sample selection was determined by parental report. Twelve of the 19 children selected on the basis of reported symptoms or medication use also met the more strict criteria for asthma as defined in O’Connor et al. 2018.⁽²³⁾ Additional evidence of recent activity at age 10 (lung function testing, symptoms -including those in the past 12 months- and medication use) is shown in Table I. Patients of the same age non-allergic to cockroach (but exposed to at least 2 Units of Bla g 1 −208 ng/g- and Bla g 2 −80 ng/g-) were used as negative controls (n = 10). Positive exposure to cockroach allergens for negative controls was selected to ensure that the patients were non-allergic to cockroach.

Table I:

Description of the population. Comparison of groups with asthma and rhinitis and no disease from URECA for demographic variables.

	No disease n=28	Asthma and rhinitis n=19	p-value
Site:			0.56
Baltimore	6 (21.4%)	4 (21.1%)
Boston	6 (21.4%)	5 (26.3%)
New York	11 (39.3%)	4 (21.1%)
St. Louis	5 (17.9%)	6 (31.6%)
Mother’s race:			0.69
Black	19 (67.9%)	13 (68.4%)
Hispanic	8 (28.6%)	4 (21.1%)
White	1 (3.6%)	1 (5.3%)
Other	0 (0.0%)	1 (5.3%)
Mother Atopic? Y es	25 (89.3%)	17 (89.5%)	0.99
Mother Asthma? Yes	14 (50.0%)	14 (73.7%)	0.19
Mother Ever Asthma? Yes	11 (39.3%)	13 (68.4%)	0.10
Mother Eczema? Yes	11 (39.3%)	7 (36.8%)	0.99
Child’s race:			0.61
Black	19 (67.9%)	13 (68.4%)
Hispanic	8 (28.6%)	4 (21.1%)
White	1 (3.6%)	2 (10.5%)
Child’s gender:			0.63
Female	12 (42.9%)	6 (31.6%)
Male	16 (57.1%)	13 (68.4%)
Child Characteristics at 10 years
FEV1/FVC	0.84 [0.82; 0.88]	0.81 [0.71; 0.85]	0.08
FEV1 % predicted	106 [94.7; 114]	101 [87.0; 107]	0.08
Eosinophils blood count	200 [100; 200]	500 [325; 700]	<0.01
BMI Percentile	84.1 [59.5; 99.1]	85.6 [46.9; 99.8]	0.75
Any wheeze by report or Dr. visit? Yes	1 (3.6%)	11 (57.9%)	<0.01
Report of eczema or EASI ≥ 1? Yes	5 (17.9%)	5 (26.3%)	0.50
Steroid Inhaled? Yes	0 (0.0%)	8 (42.1%)	<0.01
Steroid Oral? Yes	4 (16.0%)	9 (47.4%)	0.05

Summary statistics are frequencies (%) for categorical variables and medians (first; third quartile) for continuous variables with comparison between groups using chi-square test or Wilcoxon test, respectively. BMI Body mass index; FVC, forced vital capacity; EASI, Eczema Area and Severity Index

Expression, purification, and quantification of eight cockroach allergens

German cockroach allergens Bla g 1, Bla g 2, Bla g 4, Bla g 6, Bla g 9 and Bla g 11, and American cockroach Per a 7.0102 (> 98% amino acid identity to Bla g 7) were expressed in Pichia pastoris, and Bla g 5 was expressed in Escherichia coli. All the allergens were expressed and purified as described in the Online Repository.

Measurement of IgE, IgG, and IgG4 antibody levels

Total cockroach-specific and cockroach allergen-specific IgE, IgG and IgG4 antibody levels were measured in sera using i6 ImmunoCAPs (commercially available CAPs loaded with cockroach extract) and in-house allergen-loaded streptavidin ImmunoCAPs, respectively, in a Thermo Fisher Scientific ImmunoCAP system (Phadia™ 250 Immunoassay Analyzer) (Thermo Fisher Scientific, Portage, MI). Before measuring allergen-specific antibody levels, several steps were performed to optimize the assays that include allergen biotinylation, optimization of biotinylation, optimization of the amount of biotinylated allergen loaded into the streptavidin ImmunoCAPs, stability analysis of allergen-loaded ImmunoCAPs and evaluation of the specificity of the assay, as explained in the Online Repository. Biotinylated allergens were loaded and incubated on streptavidin ImmunoCAPS using the Phadia 100. Assays for IgE, IgG and IgG4 measurements were performed according to the manufacturer’s instructions. The lower limits of quantification were 0.1 kU_A/L for IgE, 2 mg/L for IgG and 0.07 mg/L for IgG4. A conservative value of ≥ 0.35 kU_A/L was specified as the cut-off value for a positive allergen-specific IgE response to calculate prevalences of IgE sensitization. IgE antibodies specific to other allergen sources were also measured by the ImmunoCAP system using commercially available ImmunoCAPs loaded with extracts (Aspergillus, Alternaria, mouse, timothy grass, oak, ragweed, cat, maple, dog and mite -Der p and Der f -).

Statistical analysis

Antibody levels (IgE, IgG, and IgG4) and total IgE values were right-skewed based on a diagnostic plot and the result of a Shapiro-Wilk test and were, therefore, log-transformed for statistical tests. The results of these tests are presented as geometric means and geometric standard deviations and the comparisons between groups as the ratio, 95% confidence interval and p-value for testing the effect with respect to the group comparison.

To compare differences in baseline characteristics and demographics between no disease and asthma and rhinitis, chi-square or Fisher exact tests were used for categorical variables, and t-tests or Wilcoxon tests were used for normally distributed and non-normally distributed continuous variables, respectively.

Relative importance was calculated using the relaimpo R package with the “lmg” method.⁽²⁴⁾ The R² of a model represents the proportion of variance explained by the set of predictors included in the model. The “lmg” method calculates the relative and independent contribution of each predictor to the R² while taking care of the dependence on the ordering of predictors within the model.

Due to the exploratory nature of the analysis, no adjustments or no attempts were made to adjust for multiple comparisons. The level of significance was set at p-value < 0.05. All analyses were performed with R version 3.5, and figures were constructed with the R lattice package.

RESULTS

Description of the patient population

Subjects were 10-year old children from low-income neighborhoods in urban Baltimore, Boston, New York, and St. Louis. Most of the children were black (~68%), followed by Hispanic (21–29%) and white (3–11%), with similar percentages for their mothers (Table I). URECA is a high-risk birth cohort, so a large proportion of the n = 47 tested subjects’ mothers were atopic (~89%) and had asthma (50–74%) or eczema (36–40%). The two groups of cockroach allergic patients analyzed were similar in terms of maternal atopy (atopic, asthma, eczema), and the child’s race and gender, but significant differences were found regarding eosinophils blood count, wheeze by report or Dr. visit and the use of inhaled steroids (Table I).

Differences of IgE sensitization to allergen sources between groups with and without disease

Subjects were sensitized to other allergen sources in addition to cockroach. The sum of 12 allergen source-specific IgEs was only about 3% of the total IgE, and this proportion was similar for the group without disease (2.8%) and the rhinitis and asthma group (4.3%) (p = 0.23) (Figure E1 in the Online Repository). On the other hand, the correlation between the sum of twelve allergen source-specific IgE and total IgE was high (r = 0.82, p < 0.001), whereas the correlation between cockroach-specific IgE and total IgE was moderate (r = 0.64, p < 0.001, Figure E2 in the Online Repository). The rhinitis and asthma group were more highly sensitized by several metrics including specific IgE and by skin prick test to Alternaria, mouse, ragweed and mite (Der p) (p ≤ 0.05), the number of positive skin test to aeroallergens (p = 0.02), the sum of specific IgEs (p = 0.01) and the number of positive specific IgEs (p = 0.01, Table E1 in the Online Repository). There were no significant differences in the proportion (%) of specific IgE to twelve different sources versus the sum of twelve-specific IgE between the two groups (data not shown).

IgE, IgG and IgG4 antibody levels to cockroach extract

Statistically significant differences between groups with and without disease were found for cockroach-specific IgE (ratio, 4.32 [95% CI, 1.33−14.02]; p = 0.015) but not for IgG and IgG4 (ratio, 1.15 [95% CI, 0.85−1.55; p = 0.37] and 0.87 [95% CI, 0.45−1.68; p = 0.67], respectively (Figure 1, Table E2 in the Online Repository). The antibody levels for groups without and with disease were 1.35 versus 5.85 kU_A/L for IgE (4.3-fold), 5.06 versus 5.80 mg/L for IgG (1.1-fold) and 0.21 versus 0.18 mg/L for IgG4 (0.8 fold), respectively (Table E2 in the Online Repository). The IgG4 antibody levels to cockroach extract were low (with 28%, 16/57 of subjects under the 0.07 mg/L cut-off) and contributing only a 6 ± 7 % to the IgG antibody levels. The three correlations between two of the three variables were significant (r = 0.53 between IgE and IgG, p < 0.001; r = 0.35 between IgE and IgG4, p = 0.017; and r = 0.35 between IgG and IgG4, p = 0.015) (see Figure E3 in the Online Repository).

Figure 1:

Comparison of cockroach-specific IgE, IgG and IgG4 antibody levels in subjects from groups with No disease and Asthma and rhinitis. In the boxplots, the solid horizontal line and filled circle within a box represent the median and geometric mean respectively, the box margins are the interquartile range (50% of the observations), whisker lines extend for 1.5 times the interquartile range, and individual observations are marked by a hollow circle. Dash lines represent the cut-off values of quantification (values under cut-off for IgG and IgG4 were included for plotting purposes, and IgE values under cut-off were plotted as 0.1 kU_A/L). The annotations in each panel are the ratio between groups (95% CI) and p-value.

More patients had high IgE titers to cockroach in the group with asthma and rhinitis than in the group without disease (8 patients belonged to high CAP classes 4–6 in the group with disease versus only 2 in the group without disease) (Table II). Total IgE also showed significant increase between both groups (ratio 2.36, p = 0.039), but not the cockroach IgE/total IgE ratio (ratio 2.53, p = 0.064) and the skin prick test wheal size (difference 0.9, p = 0.34) (Figure E4 in the Online Repository).

Table II:

Levels of sensitization to cockroach for subjects studied.

CAP class	IgE (kUA/L)	Interpretation	Number of patients *		Total
			No disease	Asthma and rhinitis
0	<0.35	Undetectable	4	2	6
1	0.35 to < 0.7	Equivocal	8	2	10
2	0.7 to < 3.5	Positive	7	5	12
3	3.5 to < 17.5	Positive	7	2	9
4	17.5 to <50	Strong Positive	1	3	4
5	50 to < 100	Strong Positive	1	1	2
6	≥ 100	Strong Positive	0	4	4
Total			28	19	47

^*Chi-square trend test p-value 0.01

Individual pattern of IgE sensitization to eight cockroach allergens

A variable pattern of IgE antibody levels to eight cockroach allergens was found for all patients (Figure 2). Within each group, a larger cockroach-specific IgE was associated with sensitization to more allergens and higher allergen-specific IgE antibody titers (Figure 2).

Figure 2:

Individual patterns of IgE sensitization to cockroach components and cockroach extract. Groups No disease (left) and Asthma and rhinitis (right) are compared. Allergens are indicated by 1, 2, 4, 5, 6, 7, 9 and 11 and cockroach extract by cr. Increasing CAP classes ranges are indicated by a graded increase of red tones (from 0 to 6). Panels are ordered from lowest to highest levels of cockroach-specific IgE.

Subjects with disease had higher levels of IgE sensitization and recognized a broader range of allergens (Figure 2). Most patients with disease (79%) recognized up to 7 allergens, and most without disease (82%) recognized up to only 3 allergens (Figure 3). The sum of 8 allergen-specific IgE was significantly larger in the group with disease (Figure E5 in the Online Repository).

Figure 3.

Number of cockroach component allergens recognized by patients with No disease and Asthma and rhinitis (cut-off of 0.1 kU_A/L).

Component analysis of IgE, IgG and IgG4 antibodies to 8 cockroach allergens

The differences in allergen-specific IgE antibody levels between both groups were significant for most allergens tested (p < 0.05), except for Bla g 5 (p = 0.33), Bla g 9 (p = 0.076) and Bla g 11 (p = 0.34) (Figure 4, Table E2 in the Online Repository). In contrast, allergen-specific IgG and IgG4 to CR component proteins were similar in the two groups (Table E2 in the Online Repository). Most CR component IgG4 values were under the cut-off level of quantification (70 ng/L), and were similarly low in the two groups.

Figure 4.

Comparison of IgE reactivity to each individual cockroach allergen between groups without and with disease. In the boxplots, the solid horizontal line and filled circle within a box represent the median and geometric mean respectively, the box margins are the interquartile range (50% of the observations), whisker lines extend for 1.5 times the interquartile range and individual observations are marked by a hollow circle.

There was an excellent correlation between cockroach-specific IgE and the sum of eight allergen-specific IgE (r = 0.86; p < 0.001) (Figure 5). This correlation was better for the group with disease than without (r = 0.97 versus r = 0.69; p < 0.001 for both) (Figure E2 in the Online Repository).

Figure 5:

Correlation between the CR-specific IgE and the sum of 8 allergen-specific IgE. The annotations in the plot area are the Pearson correlations (95% CI) and p-values.

No significant differences for allergen-specific IgG were found between IgE sensitized subjects and non-sensitized negative controls (Tables E3, E4 both in the Online Repository). No significant differences by gender were found for IgE, IgG or IgG4 to cockroach components in each of the two groups and the negative controls (data not shown).

Interestingly, eleven subjects (23% of 47; 8 from the group without disease and 3 from the group with disease) did not react to the 8 cockroach allergens tested.

Among the cockroach sensitized subjects with disease, Per a 7 was a major allergen (prevalence of IgE sensitization ≥ 50%). No major allergen was found for the group of subjects without disease. Per a 7 was the most recognized allergen in the group with disease (10/19: 52.6%) and Bla g 5 in the group without disease (11/28: 39.3%) (Table E5 in the Online Repository). No significant correlation was found between cockroach and mite (Der p) specific IgE among the 16 patients that recognized Per a 7. When considering only those participants who were highly sensitized to total cockroach (CAP class ≥ 3), Bla g 5, Bla g 9 and Bla g 11 were major in the group without disease, whereas all except Bla g 1 were major in the groups with disease (Table E5 in the Online Repository). Bla g 1 was the least recognized allergen in both groups (6/47) (Table E5 in the Online Repository).

Regarding the proportion of allergen-specific IgEs versus their sum, there was no allergen dominating the response to cockroach (i.e. > 50%). Bla g 5 showed the largest proportion (~10%) and Bla g 4 the lowest (<5%) (Figure E6, top, in the Online Repository). There were not statistically significant differences of this variable between the two groups, except for Bla g 11 (Figure E6, bottom, in the Online Repository).

Relationship of component tests to clinical disease.

To identify and quantify which IgE antibodies to 8 cockroach allergens were the most closely associated with asthma and rhinitis, a relative importance analysis was performed. The total proportion of variance explained by the model with these biomarkers was 38.4%. Bla g 6, Bla g 4 and Per a 7 were most closely associated with asthma and rhinitis, while Bla g 11 and 9 were the least important ones in explaining disease (Figure 6).

Figure 6:

Relative importance for the association of asthma and rhinitis. When combined these measurements explain 38.4% of the variance. This analysis revealed that Bla g 6, Bla g 4 and Per a 7 as the most important predictors, while Bla g 11 and Bla g 9 as the least important ones in explaining asthma and rhinitis.

DISCUSSION

Although cockroach allergy is closely associated with respiratory disease in urban centers, the specific features of cockroach sensitization that are associated with disease are unknown. In this study, the antibody reactivity to a set of eight cockroach allergens was evaluated in two groups of cockroach allergic subjects from the URECA cohort, those with or without asthma and rhinitis. The main goal was to perform a component analysis to investigate if specific allergens and/or levels of sensitization were associated with disease. IgE, IgG and IgG4 antibodies to total cockroach and each of eight different cockroach allergens from groups 1, 2, 4, 5, 6, 7, 9 and 11 were measured. Reactivity of antibodies developed against cross-reactive allergens from other insect (including cockroach) species might have also been measured (i.e. allergens from groups 1 and 7). However, this reactivity is expected to be low because German cockroach is the most common cockroach species in the US, where the studied population originated. Future studies could analyze cross-reactivity among cockroach allergens in areas such as Florida, where German and American cockroaches co-exist. One of the strengths of the study is the multicenter population, and data are representative of populations in 4 large urban areas: New York City, Boston, Baltimore and St. Louis. The demographics (70% African American, 25% Hispanic) are broadly representative of many, but not all, large urban centers.

The main finding was that the cockroach allergic subjects with asthma and rhinitis were overall IgE sensitized to more allergens and at higher levels than the subjects without disease. No differences regarding site, mother’s and child’s race and child’s gender between the two groups accounted for these observations (as previously also reported for race regarding allergic sensitization and asthma severity^(25;26)). Even with a relatively small cohort of subjects in control and disease groups, the differences found between both groups were significant. Future studies in larger cohorts will need to address differences between asthma and rhinitis patients. The previously reported importance of Bla g 2 among cockroach-sensitized patients⁽²⁾ was questioned by the results that show additional important allergens (i.e. from groups 5, 7, 9, 11). However, no component-specific IgEs were associated exclusively with the group that developed disease. Equivalent findings have been reported for only few other allergen sources. A recognition of a larger number of house dust mite allergens (up to 7) with higher IgE levels was observed in patients with asthma versus nonasthmatic subjects.⁽²⁷⁾ Similar results were found for atopic dermatitis: patients with severe disease had a significantly greater frequency of IgE reactivity to allergens from cat (rFel d 1) and house dust mite (rDer p 4 and rDer p 10), and had an IgE reactivity profile that was more spread towards different allergens compared to patients with moderate atopic dermatitis. On the other hand, there were no significant differences in the frequencies of IgE reactivity to grass pollen allergens between both groups. ⁽²⁸⁾ Finally, greater levels of IgE to Fel d 1 were also observed in cat-allergic children with asthma compared with those with rhinoconjunctivitis.⁽²⁹⁾

The differences observed at the level of allergen-specific and cockroach-specific IgE are not reflected in cockroach skin prick test reactivity, which was similar between the two groups. In this sense, the total cockroachIgE and the component analysis provide an additional level of information about the subject’s reactivity. Total cockroach IgE and the percentage of total IgE accounted for by cockroach-specific IgE were significantly greater for the group with asthma and rhinitis.

The total IgE levels to perennial allergens have been reported to be associated with asthma.⁽³⁰⁾ The median level of total IgE is typically 200–400 kU_A/L in atopic disease.⁽³⁰⁾ The values found here were around this range (197.35 versus 464.86 kU_A/L for the groups without and with disease, respectively). Interestingly, the sum of the IgE specific to 12 different allergen sources, including cockroach, was not statistically different for the two groups of cockroach sensitized subjects. The sum of the IgE specific to 12 different allergen sources accounted for <5% of the total IgE. Despite this small proportion, there was an excellent correlation between that sum of source-specific IgEs and total IgE. For the two groups, cockroach-specific IgE was the greatest among all the source-specific IgE tested, followed by mite Der p and Der f (all three indoor allergen sources). Overall, both groups showed a similar proportion of IgE sensitization to different sources. The magnitude of the sensitization (assessed by IgE titers and the number of allergens to which patients were sensitized) was the main difference between both groups.

The panel of eight allergens covers most of the IgE reactivity to cockroach extract, as shown by the high correlation between cockroach-specific IgE and the sum of the 8 allergen-specific IgEs. None of the allergens seemed to dominate the response to cockroach, since the proportion of IgE to each allergen versus the sum of all cockroach allergen-specific IgEs was as high as ~10% for Bla g 5 and as low as < 5% for Bla g 4, but none > 50%. This is in contrast with mite or cat allergy, where a large proportion of the IgE reactivity is directed to immunodominant allergens (i.e. anti-Der p 1 and anti-Der p 2 IgE together accounted on average for 85% of the mite-specific IgE).⁽³¹⁾

Some cockroach allergic patients (11/47: 23%) failed to recognize the 8 allergens tested, which suggests that other proteins are involved in cockroach sensitization. This phenomenon is not unusual and has also been observed for mite allergy in a cohort of patients from which 11.3% (11/97) did not have IgE reactivity to a panel of 13 mite allergens in a microarray.⁽³²⁾ Most of the cockroach allergic subjects who did not recognize any allergen had low IgE antibody levels, and might be sensitized to additional, most likely minor allergens, that were either not tested or are yet to be identified. Efforts in this direction, to test potentially relevant new cockroach allergens, are currently ongoing in our laboratory, and could lead to an expanded component test with increased sensitivity versus the current one.

The prevalence of allergen-specific IgE sensitization depended on the total cockroach-specific IgE titers of the subjects chosen. For example, no major allergens (recognized by more than 50% of patients) were found for the whole population of cockroach-sensitized subjects tested (n = 47). In contrast, among a sub-group of highly allergic patients (CAP class > 3), most allergens (all except Bla g 1) were major in the disease group, whereas Bla g 5, Bla g 9 and Bla g 11 were major in the subjects without disease. This lack of immunodominant allergens regarding prevalence of IgE sensitization highlights an important difference between cockroach and other allergen sources. Mite, cat, and birch-allergic subjects are mainly sensitized to only one or few major allergens with high IgE prevalences (Der p 1, Der p 2 and Der p 23 for mite, Fel d 1 for cat and Bet v 1 for birch). This observation has implications for immunotherapy, which might be effective for more patients when dominant allergens are administered. For example, Pauli et al. showed that the administration of only recombinant Bet v 1 was as effective as administration of birch extracts, for the specific treatment of birch pollen allergy. ⁽³³⁾ For cockroach, the administration of more allergens might be necessary for a vaccine to be effective for more patients.

Only a few studies report associations between IgE sensitization to specific cockroach allergens and disease. In Taiwan, a component analysis revealed that eight American cockroach allergens did not have equal importance in terms of pathogenicity among subjects of a wide range of ages (8–86).⁽³⁴⁾ IgE reactivity to allergens from groups 1–7 and Per a 9 were compared between patients with asthma and rhinitis or rhinitis alone. A high proportion of subjects with persistent asthma and rhinitis (81%; 17/21) had IgE reactivity to Per a 2. In contrast, 80% (16/20) of individuals with only allergic rhinitis had IgE reactivity to Per a 9 compared with only 29% (6/21) of asthma and rhinitis patients. The highest prevalences of IgE sensitization in patients with asthma and rhinitis were for Per a 2 (81%), followed by Per a 5 (67%). These two allergens also dominated the IgE antibody responses to five cockroach allergens in a US cohort of 118 highly cockroach allergic patients. Among sera with high IgE to cockroach extract (3.5–100 IU/mL), the prevalence of IgE antibodies to Bla g 2 and Bla g 5 were the greatest (71% and 58%, respectively).⁽²⁾

Unlike the Taiwanese study, groups of 10-year old children with and without asthma and rhinitis were compared here. Prevalence of IgE sensitization to any one specific allergen was under 40% in the cockroach allergic subjects without disease, in contrast with a higher prevalence in the group with disease (above 40% for most allergens except Bla g 1 and Bla g 4, using a conservative cut-off of 0.35 kU_A/L instead of 0.1 kU_A/L). Among highly allergic patients with disease (CAP ≥ 3), IgE prevalence reached 60–80% for all allergens except for Bla g 1 (30%). Interestingly, differences in IgE prevalence between both groups were the highest for Bla g 1 (7.4-fold) which might indicate that this allergen is an important contributor to the development of disease for patients sensitized to this allergen (26%).

Currently, cockroach extracts are not standardized and have high variability in allergen content.⁽³⁾ Identifying patterns of sensitization to specific CR antigens could help to guide allergen immunotherapy, in efforts to ensure that the patient receives allergens relevant to disease. The ultimate goal is to develop a component diagnostic test to increase the accuracy of the prediction of asthma and rhinitis based on cockroach component analysis. During the ongoing cockroach immunotherapy trials of the Inner City Asthma Consortium, the cockroach component analysis will allow evaluation of the effectiveness of immunotherapy based on the antigen content of the therapeutic cockroach extract and individual patterns of sensitization.

Clinical Implications:

Cockroach allergic subjects with asthma and rhinitis are sensitized at higher levels to more cockroach allergens than patients without disease. Sensitization profiles should be considered for diagnostic and therapeutic purposes.

ABBREVIATIONS:

CCD

Cross-reactive carbohydrate determinants

ICAC

Inner-City Asthma Consortium

NIAID

National Institute of Allergy and Infectious Diseases

URECA

Urban Environment and Childhood Asthma

WHO/IUIS

World Health Organization and International Union of Immunological Societies