Effect of individual allergen sensitization on omalizumab treatment outcomes in patients with severe allergic asthma determined using data from the Czech Anti-IgE Registry

The global prevalence of asthma is still uncertain [1, 2], and some experts suggest a prevalence of 7–8% (i.e., 700 000–800 000 people) in the Czech Republic (CR). In this particular asthma population, probably more than 80% of patients have eosinophilic (or type 2-high) and over 70% have eosinophilic allergic (or Th2- high) asthma endotype [3‐7].

Patients with severe asthma [8] (or severe refractory asthma) account for approximately 2.1% of the whole asthmatic population in the CR [3]. Currently, the Global Initiative for Asthma (GINA) recommends targeted biological therapy before the administration of systemic corticosteroids (SCSs) in step 5 of the integrated treatment strategy, wherever possible. The broad clinical effects of omalizumab have been extensively documented in controlled studies [9, 10] and real-world studies [11‐15].

However, little is known about the differences in treatment outcomes associated with a particular sensitization type or combinations of aeroallergens. In general, allergic reactivity to perennial allergens (such as house dust mites, molds, dogs, and cats) must be proved to initiate an anti-IgE monoclonal antibody omalizumab treatment [16]. Nonetheless, a previous study has indicated that sensitization to seasonal allergens can lead to a comparable treatment response [17].

In this study, we aimed to evaluate the potential differences in treatment outcomes among patients with severe allergic asthma (SAA) according to the sensitization pattern to different aeroallergens and their combinations. All patients were treated in accordance with local and international guidelines [7, 18, 19] using high-dose inhaled corticosteroids with long-acting β2-agonists or an alternative controller (leukotriene receptor antagonists or theophylline) with an add-on anti-IgE therapy with omalizumab. We used data from the Czech anti-IgE Registry (CAR).

We assessed the following treatment outcomes: (1) responding (according to the global evaluation of treatment effectiveness [GETE] at 16 weeks after treatment initiation), not responding, or withdrawing treatment; (2) a reduction in severe exacerbation rate (ER) after 12 months of treatment in responders compared with that in non-responders and patients who withdrew omalizumab treatment; and (3) an improvement in the asthma control test (ACT) result after 12 months of treatment (in all three study arms) in patients with SAA treated with omalizumab. In addition, we assessed the following secondary outcomes at baseline and after 12 months of treatment in all three study arms: (1) dependence on corticosteroid use, (2) spirometry characteristics (forced expiratory volume in 1 s FEV1), and (3) fraction of exhaled NO (FeNO). Furthermore, all treatment outcomes were analyzed in relation to the sensitization type by different aeroallergens at baseline and compared between polysensitized patients (sensitized to all tested allergens) and monosensitized or partial polysensitized patients. The main aim of the study was to test the hypothesis that different treatment outcomes may be obtained for each sensitization pattern.

In the present study, we evaluated the effect of sensitization to individual allergens or their combinations on the outcomes of anti-IgE therapy in patients with SAA. Polysensitized patients showed a higher tendency to be a responder than non-polysensitized patients, and a lower tendency to increase ACT and reduce ER.

There is strong evidence to indicate that anti-IgE therapy has an impressive immuno-modulatory effect [30, 31] and broad clinical effects [31‐33], notably a decrease in the daily dose of SCS or even its termination and an improvement in the quality of life, as documented in several real-setting studies [14]. There are also rare reports of the suppression of allergic reactivity [34‐37] or susceptibility to viral infections by anti-IgE therapy [38]. Other studies on the clinical characteristics of patients have shown optimal clinical benefits of the treatment. The biomarker-based prospective study EXTRA demonstrated that omalizumab treatment tends to be more effective in patients with elevated baseline eosinophil count, FeNO, and serum periostin level [39]. However, these results have not been confirmed in the retrospective real-life analysis of omalizumab-treated patients, the STELLAIR study, which showed a clinically significant effect irrespective of baseline eosinophil count [40]. Similar results have been reported by the prospective observational study PROSPERO, which reported that the effects of omalizumab (a reduction in ER and number of hospitalizations, and an improvement in ACT) were independent of either baseline eosinophil count or FeNO [41]. We should be aware of instability of peripheral blood immune parameters over a year in patients with stable asthma [42].

Notably, evidence regarding the influence of sensitization profile on omalizumab treatment outcomes is scarce. Currently, it is broadly accepted that the frequency of sensitization in a population to a particular allergen depends on the climate or environment; different biological features of causal allergens can determine the different clinical traits of allergy [43]. However, there are only a few studies on the clinical characteristics of patients with asthma related to the sensitization profile. A previous study in a Chinese population demonstrated that sensitization to house dust mites was associated with increased severity of asthma [44]. In contrast, another study (in Spain) reported that sensitization to different allergens was not associated with significant differences in severity and control of asthma. However, diagnostic and therapeutic approaches slightly differ according to individual allergen sensitization [45]. Moreover, there was no mention of anti-IgE treatment.

In the CR, patients with severe uncontrolled asthma are treated in specialized sites of NCTA [46, 47]. Only these centers are eligible to indicate biological treatment with anti-IgE antibody (omalizumab) in the CR [7, 47]. This careful approach may contribute to the higher number of responders (according to GETE) in the eXpeRience registry (88.9%) [15] than in data from other countries (69.9%) [48]. Extensive supervision of patients in the NCTA centers can result in better treatment outcomes with respect to some parameters in all groups of patients.

In our analysis of patients, whose data are included in the CAR registry, we reported similar response rates according to the GETE analysis at 16 weeks after treatment initiation (82.8%). However, the treatment outcomes were remarkable in all groups of patients (including non-responders and omalizumab-withdrawn patients). This might be due to the fact that all patients were treated with omalizumab for at least 16 weeks (non-responders), 12 months (responders), and between 16 weeks and 12 months (omalizumab-withdrawn patients). Another reason could be the unequal number of patients in all study arms. In addition, there may be a bias in results of responders compared with that of omalizumab-withdrawn patients, because all patients in the withdrawn group were a subgroup of responders. Moreover, 14 patients (20%) in this group discontinued treatment because of the stabilization of disease. Other reasons for withdrawal were diverse, but none of them was worsening of the disease. Although Namazy et al. [49, 50] proved that the use of omalizumab in pregnant women was safe, we preferred to terminate treatment by a mutual agreement with pregnant women in some centers. The outcomes were not influenced by the fact that non-responders appeared to have more severe disease than other patients (Table 2).

Lombardi et al. [43] reported that some aeroallergens, especially molds, pet dander, cockroach, and ragweed, were found to be more strongly associated with severe asthma. The association between mold sensitization and severe asthma is well known and conceptualized as severe asthma associated with fungal sensitization [51]. The identification of related causal allergen(s) is important for optimal complex therapeutic strategies, including specific allergen immunotherapy [52]; however, evidence of allergen avoidance is under debate [53].

In our study, we assessed mites instead of cockroaches because the representation of patients with sensitization to cockroach in the Czech population is weak compared with that in other countries (e.g., the USA). In the subgroup of monosensitized or polysensitized patients, there was no significant difference in treatment outcomes regarding GETE (Table 1) and no difference in asthma severity. Nevertheless, polysensitized patients (as defined above) had higher odds (OR = 2.217, p = 0.02) of being responders than all other subgroups of non-polysensitized patients. In contrast, they had reduced odds of ACT improvement (OR = 0.503, p = 0.032) among all other subgroups (OR = 0.398, p = 0.023 in the responder subgroup) and a reduction in exacerbation (OR = 0.431, p = 0.034 in the responder subgroup) compared with the non-polysensitized patients. This can be explained by the tendency (although nonsignificant) of polysensitized patients to have a higher baseline ACT score and lower rate of severe exacerbations than non-polysensitized patients. In addition, these patients tend to have a better clinical status at treatment initiation than non-polysensitized patients (Fig. 8). We suggest that these polysensitized patients with asthma may have a different subgroup of allergic diseases that may share some features with the “Th2-ultrahigh” concept suggested by Peters [54]. However, we do not have enough data to confirm this possible connection.

There were certain limitations to our study. First is the low number of patients with a distinct sensitization profile, which led to a loss of statistical power. However, a substantial number of patients from the initial pool (n = 389) had to be excluded from the assessment (n = 110, 28.3% of all enrolled patients in the registry) due to incomplete data in the registry. The used per-protocol analysis was not designed to treat censored (or missing) data. Second, we lost some information owing to the semiquantitative evaluation of categorized levels of sensitization. Third, we could not assess differences in treatment outcomes among patients sensitized to perennial or seasonal allergens (including weeds) because virtually all patients were concurrently sensitized to at least one (or mostly more) seasonal allergen and at least one (or more) perennial allergen, which was a necessary pre-condition for treatment initiation according to the omalizumab Summary of Product Characteristics. Fourth, there are some conflicting results regarding baseline ACT, an ER in responders; this could be caused only by sampling error. However, due to the exploratory nature of the study, we believe that it is important to remark all interesting data configuration to highlight potential variables for future confirmatory studies. Thus, we are unable to suggest a simple and clear clue to deal with the kind of sensitisation of patients with asthma receiving omalizumab therapy. We rather focused on seeking potential variables and parameters that could be used as putative biomarkers to estimate the supposed disease evolution and possible treatment outcomes. Finally, because of the observational nature of the study, we did not exploit randomization and placebo control group.

In summary, this is the first study to provide data on the effects of allergen sensitization of patients with allergic asthma on omalizumab treatment response. We believe that there may be some biological differences among distinct subpopulations of patients with asthma in terms of polyvalence of atopic sensitization. These differences may lead to distinct objective treatment effects (e.g., reduction in severe exacerbations) and different subjective disease perception (e.g., improvement in the ACT result).

We found that polysensitized patients exhibited a higher tendency to be responders than non-polysensitized patients but a lower tendency of showing an improvement in the ACT results and a reduction in ER. We suggest that these polysensitized patients with asthma may constitute a separate subgroup of individuals with different allergic diseases and distinct clinical characteristics. Prospective data from controlled trials are needed to confirm these observations.