Intranasal administration of allergen increases specific IgE whereas intranasal omalizumab does not increase serum IgE levels—A pilot study

Abstract Background Administration of the therapeutic anti‐IgE antibody omalizumab to patients induces strong increases in IgE antibody levels. Objective To investigate the effect of intranasal administration of major birch pollen allergen Bet v 1, omalizumab or placebo on the levels of total and allergen‐specific IgE in patients with birch pollen allergy. Methods Based on the fact that intranasal allergen application induces rises of systemic allergen‐specific IgE, we performed a double‐blind placebo‐controlled pilot trial in which birch pollen allergic subjects were challenged intranasally with omalizumab, placebo or birch pollen allergen Bet v 1. Total and allergen‐specific IgE, IgG and basophil sensitivity were measured before and 8 weeks after challenge. For control purposes, total, allergen‐specific IgE levels and omalizumab‐IgE complexes as well as specific IgG levels were studied in subjects treated subcutaneously with either omalizumab or placebo. Effects of omalizumab on IgE production by IL‐4/anti‐CD40‐treated PBMCs from allergic patients were studied in vitro. Results Intranasal challenge with Bet v 1 induced increases in Bet v 1‐specific IgE levels by a median of 59.2%, and this change differed significantly from the other treatment groups (P = .016). No relevant change in allergen‐specific and total IgE levels was observed in subjects challenged with omalizumab. Addition of omalizumab did not enhance IL‐4/anti‐CD40‐induced IgE production in vitro. Significant rises in total IgE (mean IgE before: 131.83 kU/L to mean IgE after: 505.23 kU/L) and the presence of IgE‐omalizumab complexes were observed after subcutaneous administration of omalizumab. Conclusion Intranasal administration of allergen induced rises of allergen‐specific IgE levels, whereas intranasal administration of omalizumab did not enhance systemic total or allergen‐specific IgE levels.


| INTRODUCTION
Immunoglobulin E (IgE) is the key antibody class responsible for allergic symptoms in sensitized patients. 1 Cross-linking of IgE bound to the surface of mast cells and basophils via its high-affinity receptor (FceRI) results in degranulation of the cells and the release of inflammatory mediators, cytokines and proteases. [2][3][4] In order to prevent IgE binding to its receptors, omalizumab-a monoclonal humanized anti-IgE antibody that specifically binds to the constant region 3 (Ce3) of IgE-has been developed and has been shown to be effective for the treatment of a variety of allergic manifestations and in particular for severe IgE-mediated asthma. [5][6][7] As the binding site of omalizumab overlaps with the binding site of IgE to FceRI, omalizumab does not induce mast cell and basophil activation or anaphylactic reactions and inhibits binding of circulating IgE to its receptor on effector cells. Administration of omalizumab therefore reduces mediator release after approximately 16 weeks of therapy, which is associated with a downregulation of FceRI expression on basophils, mast cells and antigen-presenting cells. [8][9][10][11] More recently, it has been shown that omalizumab is also effective for the treatment of chronic spontaneous urticaria. 12 Omalizumab targets the Ce3 region of IgE, which is considered to be accessible also on membrane-bound IgE in the form of B-cell receptors (BCRs) on IgE + memory B cells. It is therefore possible that cross-linking of the BCR on IgE + memory cells may contribute to the strong increases in IgE levels which are observed in patients treated with omalizumab in addition to a prolongation of the halflife of IgE within the omalizumab-IgE immune complexes. 5,13,14 In this context, it has recently been hypothesized that the binding of omalizumab to IgE may prevent degradation of IgE via binding of the resulting IgE-IgG complex to the neonatal Fc receptor (FcRn) and thus protecting it from endolysosomal degradation or that the IgE-omalizumab complexes do not undergo pinocytosis. 15 The other hypothesis that cross-linking of the BCR by omalizumab may result in increased IgE production is supported by the finding that injection of anti-IgE in previously primed mice was driving a secondary IgE response 16 and that secondary IgE responses in mice could also be boosted by repetitive B-cell epitopes which also may have crosslinked the BCR. 17 In the nasal mucosa of allergic patients, the presence of IgE + B cells as well as plasma cells 18 and local IgE production 19 have been demonstrated. In addition, a strong rise of systemic allergen-specific IgE has been observed in subjects after nasal administration of allergens [20][21][22][23] indicating the potential presence of memory IgE + B cells in the nasal mucosa. We therefore investigated the effect of intranasal administration of omalizumab, allergen or placebo on allergen-specific and total systemic IgE. We applied in a double-blind, placebo-controlled clinical study omalizumab to birch pollen allergic subjects intranasally and, as positive and negative control, included allergic subjects who received intranasal allergen challenge and buffer, respectively. The effects of intranasal omalizumab and allergen exposure on total, allergen-specific IgE and IgG levels as well as on basophil sensitivity were then analysed for a period of 8 weeks. In addition, we studied the effects of subcutaneous (s.c.) omalizumab administration in allergic subjects regarding changes in total, allergen-specific IgE and allergen-specific IgG levels.  Figure 1). Fifteen subjects fulfilled all inclusion criteria and were thus selected for the study. Those subjects were suffering from moderate-to-severe allergic rhinitis to birch pollen for at least 2 seasons according to medical history. Birch pollen allergy was confirmed by skin prick test (ALK-Abello, Wedel, Germany) and measurement of Bet v 1-specific IgE antibodies in serum (ImmunoCAP, Phadia, Uppsala, Sweden). Only patients with Bet v 1-specific IgE levels higher than 3.51 kUA/l who were not previously exposed to omalizumab were included in the study. The absence of nasal polyps or substantial deviation of the nasal septum was confirmed by anterior rhinoscopy prior to nasal challenge.
After the screening visit, patients were randomized into 3 groups at a ratio of 1:1:1 (5 patients per group, using an online randomization programme provided by the Centre of Medical Statistics, Informatics, and Intelligent Systems of the Medical University of Vienna) for intranasal challenge with either omalizumab (Xolair â , Novartis Pharmaceuticals Corporation, New York, NY), Bet v 1 (Biomay, Vienna, Austria) or placebo (0.9% sodium chloride solution). One patient in the omalizumab challenge group withdrew from the study due to illness (upper respiratory tract infection) just before the first nasal challenge. At visit 1 (t1), blood samples were taken and the first nasal challenge was performed.
Two more nasal challenges were performed on the consecutive 2 days. Further blood samples were taken 3 (t2), 5 (t3) and 8 (t4) weeks (AE4 days) after the first challenge. The clinical investigators and study subjects were blinded regarding the administered substances. Randomization and preparation of the study drugs (omalizumab, Bet v 1 or placebo) were carried out by a person who was not involved in the study and had no contact with the subjects to ensure the double-blinded nature of the study.

| Nasal allergen challenge
Purified recombinant Bet v 1 produced according to good manufacturing practice (GMP) (Biomay AG) or Xolair (Novartis) was diluted in sterile 0.9% sodium chloride freshly before use. Bet v 1 was applied at a concentration of 50 lg/mL as this concentration has previously been shown to lead to increase in systemic serum IgE levels after intranasal administration. 21 Omalizumab was applied at an equimolar concentration (438 lg/mL), and 0.9% sodium chloride was applied as placebo. Patients were administered 1 puff per nostril using a metered pump delivering 15 lL per puff.
2.3 | Sera from subjects who had received s.c.

administration of omalizumab or placebo
Recruitment, study population, administration of the drug as well as dosage and blood draw schedule are described in detail in Gevaert et al. 24 Sera from this study were obtained and tested in our study as follows: sera "Before s.c. administration" were tested at visit V1 (V1: screening visit, at least 2 weeks before visit V2 where first dose of omalizumab was administered) except for subject S9 where serum was from V2. Sera "After s.c. administration" were tested at visit V8 (12 weeks after first subcutaneous administration) except for subject S18 where serum from visit V6 (8 weeks after first subcutaneous administration) and for subject S16 where serum from visit V10 (16 weeks after first subcutaneous administration) was tested. Skin prick tests were performed using a grass mix, 2 different tree mixes (Betulaceae and Fagaceae), as well as mugwort, mould, cat, dog, D. pter., D. far. extract all purchased from Stallergenes (Antony, France).

| Measurements of total IgE, allergen-specific
IgE and allergen-specific IgG levels in sera Total IgE and IgE specific to birch extract (t3), Bet v 1 (t215), grass (g6), olive (t9), cat (e1) and house dust mite (d1) were F I G U R E 1 Design of the intranasal challenge study. A, Subjects were challenged intranasally with either omalizumab, Bet v 1 or placebo on 3 consecutive days (d1-d3, arrows, top). Blood samples (arrows, bottom) were taken before the first (t1) challenge and on days 21 (t2), 35 (t3) and 56 (t4) (AE4 days) after the first challenge. B, Flow chart depicting number of subjects who were screened (n = 24), randomized (n = 15), and who completed (n = 14) the study ECKL-DORNA ET AL. | 1005 measured by ImmunoCAP (Phadia, Uppsala, Sweden). Allergen-specific IgG levels were analysed using a microarray based on Immu-noCAP ISAC technology (Phadia-Thermofisher, Uppsala, Sweden) containing more than 170 different allergen molecules 25 according to the manufacturer's instruction. Degranulation was induced by addition of rBet v 1 (0.01 lg/mL) or buffer as control. Released b-hexosaminidase was measured, and percentage release was calculated as described. 28 All measurements were performed in triplicates. Percentage changes in b-hexosaminidase release between t1 and t3 were calculated.  Percentage change in total IgE levels compared to supernatants from unstimulated cells was calculated.  Figure 1). A group challenged intranasally with Bet v 1 was included as a positive control because it has been shown that intranasal challenge with Bet v 1 induces systemic rises of Bet v 1-specific IgE levels. [20][21][22][23] One subject of the omalizumab group dropped out 1 day before intranasal challenge due to a viral upper respiratory tract infection. All other participants completed the study. Demographic, clinical and serological data of the subjects are displayed in Tables 1 and Table S1 indicating a balanced distribution of mean Bet v 1-specific and mean total IgE levels in the 3 groups.

| Statistics
In addition, we investigated sera obtained from subjects before and after receiving omalizumab (n = 15) or placebo (n = 6) subcutaneously 24 because it has been shown that subcutaneous injection of omalizumab induces rises of total serum IgE. These patients have been previously described by Gevaert et al, 24 and their demographic and clinical characteristics are listed in Table S2.

| Intranasal challenge with Bet v 1 increases
allergen-specific IgE levels while intranasal omalizumab challenge has no effect on allergenspecific or total IgE levels Subjects who were challenged intranasally with Bet v 1 on 3 consecutive days outside of the birch pollen season showed significant increases in Bet v 1-specific serum IgE levels as compared to subjects who were challenged with placebo or omalizumab (Kruskal-Wallis test: P = .0053, two-sided Wilcoxon rank-sum test Bet v 1 vs omalizumab and Bet v 1 vs placebo: P = .016) (Figure 2A, Table S3). The median increase in Bet v 1-specific serum IgE observed at t3 in the Bet v 1-challenged group was 59.23% (range: 3.55%-124.60%). No relevant changes in total or Bet v 1-specific IgE levels were observed after nasal challenge with omalizumab, and there were no statistically significant differences in changes of total IgE levels between t1 and t3 between the 3 treatment groups (Kruskal-Wallis Test: P = .0521) ( Figure 2B). There were also no relevant changes observed for grass, olive, cat or house dust mitespecific IgE levels between t1 and t3 between the 3 groups (Table S3). 3.3 | Loading of basophils with sera after Bet v 1 challenge increases sensitivity but patient's basophil sensitivity is not increased In a first set of experiments, we investigated whether loading of RBL cells expressing the human FceRI with sera from subjects obtained 5 weeks after intranasal challenge with Bet v1 increases Bet v 1-specific sensitivity. We found significant group differences between the intranasally challenged groups (Kruskal-Wallis: P = .0463). The sensitivity of basophils loaded with sera from Bet v 1-challenged subjects was significantly higher than that of basophils loaded with sera from omalizumab-challenged subjects (pairwise Wilcoxon test: P = .032) ( Figure 3A). The sensitivity of basophils loaded with sera from Bet v 1challenged subjects was also greater than that of basophils loaded with serum from placebo-challenged subjects but did not reach significance (P = .095).
No differences regarding Bet v 1-specific sensitivity was found when basophils from challenged subjects were obtained at t1 and t3 and compared in vitro for allergen-specific sensitivity ( Figure 3B).
Basophil activation in PBMCs isolated from all subjects was mea-  When we analysed allergen-specific IgE levels, a fourfold to sevenfold rise in allergen-specific IgE was observed after subcutaneous administration of omalizumab but not of placebo regardless of the nature of the tested allergen (Table S4). Interestingly, in 3 patients, Using an ELISA based on antibodies recognizing specifically IgEomalizumab complexes but neither noncomplexed IgE nor noncom-

| Neither intranasal nor subcutaneous application of omalizumab affects allergen-specific IgG levels
We also analysed allergen-specific IgG levels in sera obtained from subjects before and after intranasal or subcutaneous application of omalizumab using a microarray containing more than 170 different allergens. 25 There were no consistent and no significant changes in allergen-specific IgG levels after intranasal or after subcutaneous application of omalizumab, Bet v 1 or placebo for Bet v 1, Phl p 1, Phl p 5, Ole e 1, Fel d 1 and Der p 1 (Table S5; Fig. S1).

| Omalizumab does not increase IgE production induced in allergic patients' PBMCs stimulated with anti-CD40/IL-4 in vitro
To further investigate the potential effect of omalizumab on IgE production in vitro, we stimulated PBMCs derived from allergic subjects (Subject 1 and 2) with anti-CD40 and IL-4 in the presence or absence of omalizumab as well as an isotype control for omalizumab.
Anti-CD40 and IL-4 stimulation induced an increase in total IgE levels in culture supernatants of both allergic subjects as previously described 30,31 (Fig. S2). Addition of omalizumab had no relevant effect on IgE production in the PBMC cultures.

| DISCUSSION
In the present study, we investigated the effect of intranasal administration of Bet v 1, omalizumab and placebo on the levels of total and allergen-specific IgE in patients with birch pollen allergy. The IgE in the nasal mucosa. 19 We therefore hypothesized that IgE-pro-  (Table S5). Finally, it is possible that at the time points investigated (ie, t1 and t3) the newly produced IgE was not yet fully loaded onto the patients basophils which were preoccupied with the IgE which was there before.
The results can thus be summarized as follows: intranasal application of the major birch pollen allergen Bet v 1 led to an increase in allergen-specific but not total systemic IgE levels, while intranasal omalizumab application had no effect on total or allergen-specific