Research which advances the current understanding of the immunoglobulin and inflammatory profile within the BAL and serum of asthmatic patients, in particular, highlighting differences between local airway and systemic profiles, may offer real potential therapeutic insights. We therefore sought to interrogate and compare the immune response localized to the asthmatic airway with that observed systemically, with a particular reference to immunoglobulin profiling.

Overall, absolute immunoglobulin levels were higher in serum. This was not unanticipated and is potentially accounted for by the dilution effect which occurs as part of the nature of the BAL sampling process. However, several cytokine levels were higher in BAL samples (IL1β, IL-6, IL-8), while others were higher in plasma (IL-4, IL-13, IL-17, IL-10, and TNFα); therefore, the BAL dilution effect is not wholly responsible for varying levels. Instead, this probably represents the function of the mucosal, epithelia rich, airway site, and the natural distribution of immunoglobulins. Our analysis detected 28% of IgE in BAL, while most other immunoglobulins were detected at 100%. We propose that this is not just due to dilution alone. To further investigate why 28% of patients had detectable IgE signals and the rest did not, we compared the groups and found that there was a significant difference in FEV1 values with a lower % predicted in those with a detectable IgE (mean 80.6% v 68.5%, P = 0.03), there were more males and a greater proportion of ex-smokers in the IgE-detected group (52% v 18%). Therefore, the detection of IgE may reflect a more inflammatory clinical phenotype. Furthermore, in the IgE-detected group, there was a correlation between %eosinophils in BAL and serum eosinophils (P = 0.001, r = 0.69) and plasma IL-5 (P = 0.01, r = 0.55) levels which may further reflect this.

Our study demonstrated weak relationships between BAL and serum levels of IgM, IgA, IgG2, and IgG4,(Fig. 1B–E), while no correlation was observed between airway and systemic levels of IgD, IgE, IgG1, and IgG3, suggesting systemic and local airway immunoglobulin levels are not fully in concordance. Furthermore, with the exception of TNF-α, weak relationships were also found between BAL and plasma levels of IL-10, IL-1β, IL-4, IL-5, IL-6, and IL-13 (Fig. 2B–G). Overall, the best predictor of the BAL analyte level was serum TNF-α with a coefficient of determination of 0.73 meaning that 73% of TNF-α BAL levels can be deduced from its serum levels. The same could not be said for any of the other analytes examined in this study. Treatment decisions, therefore, should be made with great caution when only serum immunoglobulin and cytokine levels other than those of TNF-α are available, as systemic and BAL levels of both immunoglobins and cytokines are not fully reflective of each other and measurements in one compartment may not be an accurate representation of the pathophysiology occurring in the other.

IgE is a potent monomeric immunoglobulin associated with hypersensitivity and allergic asthma and is the most clinically relevant of the immunoglobulins analyzed in our study [11]. IgE is a target of omalizumab, a humanized anti-IgE monoclonal antibody used to treat severe asthma [5]. The prescription of omalizumab is based on serum IgE levels which our data suggest may not be reflective of the airway inflammatory profile. Studies which have compared induced sputum levels of IgE to that of serum IgE have shown contradictory reports with some showing a correlation between the two levels, while others have shown no correlation between the two compartments [19,20,21,22]. Furthermore, studies which use more invasive measures of airway sampling would suggest that IgE levels are higher in the airways which may indicate localized production of IgE not reflected in the serum [23]. Therefore, the use of serum IgE as a marker for commencing biological therapy in severe asthma may not be appropriate. In fact, these discrepancies may explain why some patients meeting the current clinical criteria for treatment with omalizumab therapy fail to display a beneficial clinical response [24,25,26].

Our study highlights immunoglobulin variation is not only present between different individuals but also within individuals, with variation in immunoglobulin levels existing within the airway and systemically. The lack of an observed relationship between airway and systemic IgE is of clinical relevance to asthma, having possible implications for drug efficacy within an asthmatic cohort. As mentioned, important therapeutic interventions with omalizumab are made based on systemic rather than airway immunoglobulin profiles. The response to this therapy is not universally positive. This suggests that such an approach for anti-IgE monoclonal therapy may not be optimal in all cases and that perhaps further defining the airway inflammatory profile may afford a more sophisticated, targeted, individualized approach to patient treatment with such agents. We believe that our findings may partly explain the clinical variation in response to omalizumab and given the costs associated with such treatment in ‘non-responders’ may justify further individual patient profiling prior to prescription of these medications. We believe that studies assessing clinical response based on airway profiling, in addition to the current systemic profiling are now warranted.

There are some potential limitations to our study. The lack of correlation between serum and BAL IgE could possibly be because IgE was analyzed using an ImmunoCAP system while the other immunoglobulins were analyzed with sandwich ELISA which slightly vary in target detection. Furthermore, only 28% of samples had detectable BAL IgE levels compared with nearly 100% detection for other immunoglobulins, but we did explore this further and believe it may reflect a group of patients with elevated inflammation within the airways. Dilution because of BAL sampling may also have affected results when compared to serum levels of immunoglobulins, but this was not the case with cytokine levels. However, our study has several strengths. We have a large patient number of clinically well-defined asthma patients. Furthermore, both BAL and blood samples were acquired on the same day and analyzed in a similar manner to minimize potential confounders. We acknowledge that bronchoscopy may not be viable in all centers or in all patients who are being screened prior to starting monoclonal therapy; however, studies on the relationship between sputum and blood IgE have been contradictory [19,20,21,22]. Bronchoscopy offers an advantage in that it allows for visual inspection and guarantees sample acquisition to fully phenotype patients prior to starting monoclonals.