Introduction
The Cystic Fibrosis Foundation Patient Registry (CFFPR) reports a high prevalence of asthma (34.6%) in people with Cystic Fibrosis (PwCF) [
1]. While our current understanding of this relationship is limited, a type 2 inflammatory (T2) eosinophilic phenotype has often been identified in CF patients [
2]. Moreover, medication targeting this phenotype has improved the clinical course of CF patients with T2 inflammation [
3].
This type 2 inflammation is characterized by high total IgE antibody titers, specific IgE sensitization and eosinophilia. Also central to this signature are type 2 cytokines (interleukin (IL)-4, IL-5, and IL-13) and type 2 cytokine producing cells (in airway or sputum), such as type 2 innate lymphoid cells [
4‐
6]. CF mouse models have shown an exaggerated IgE response to
Aspergillus fumigatus, with higher levels of IL-13 and IL-14 [
7]. Moreover, naïve CD4 + T cells from CF transmembrane conductance regulator (CFTR) deficient mice produce elevated levels of IL-4 after T cell receptor (TCR) ligation compared to wild-type CD4 + T cells [
8]. This T2 signature has also been identified in CF patients prior to and after
Pseudomonas aeruginosa infection, with a T2 inflammatory phenotype being an independent risk factor for infection [
9].
This study aimed to evaluate the relationship between the eosinophilic CF T2 inflammatory phenotype and CF-related pulmonary outcomes and microbiological data. As there are many parameters for T2 inflammation, we focused primarily on absolute eosinophil count (AEC) and IgE levels. We hypothesized that higher levels of AEC and IgE would be associated with poor lung function (FEV1), increased frequency of CF exacerbations, and higher susceptibility to microbiological infections.
Discussion
Prior to the introduction of highly effective CF modulator therapy, the CF Foundation patient registry report showed that at any given year, approximately 40% of patients had one or more exacerbations a year requiring hospitalization and/or intravenous antibiotics and approximately 10% of this group had 3 or more exacerbations a year. Each year, there is a slight change in the patient groups, some patients do not develop exacerbations annually but every other year or every 3 years and some patients suddenly develop more active disease, and more frequent yearly exacerbation. Following this cohort longitudinally allowed us to include in the analysis more patients with less (1–2 exacerbations, a less active disease) and more frequent exacerbations (more than 2) or more active disease, over a longer period of time. In addition, when a patient develops more than 2 exacerbations a year, our group initiates a frequent exacerbation protocol to identify possible new contributors to the change in phenotype which includes evaluation of ABPA/allergic or atopic phenotype based on serum IgE and AEC levels. Of note, our population has a high exacerbation rate, with 57.5% of our patients having 1–2 exacerbations per year and 24.5% of patients having 3 or more exacerbations per year. While this could be a result of our center’s geographical location it can also be a result of our study period (6 years), which allowed more incidences to be counted than otherwise i.e., even if a patient had only 1 exacerbation in the study period, they would be counted in the 1–2 Exacerbations per year cohort. In addition, our center monitors very closely the AEC and IgE levels with each exacerbation, which most likely led to an increased identification of elevated levels for these two measurements.
As per our hypothesis, higher levels of AEC were associated with more frequent exacerbations, defined as > 2 pulmonary events that required hospitalization and/or IV antibiotics. Our study shows that CF exacerbations increase the AEC levels and corelate with higher AEC variability. The correlation between CF exacerbation and AEC may be explained by exposure to intravenous antibiotics, however the correlation between CF exacerbations and AEC variability is a new finding that is worth further exploration. Based on our study however we cannot draw a conclusion on a cause and effect. It is unclear if exacerbations cause increased T2 inflammation or T2 inflammation predisposes to increased exacerbations.
When accounting for positive allergen panel positivity, as IgE increases, exacerbation events increase. In addition, each annual exacerbation is associated with an increase in IgE and this effect was significantly more pronounced in patients with atopy. This is a new finding and suggests that T2 inflammation may be partially responsible for increased exacerbations. If proven in multicenter, larger study, could lead to a change in CF management with increased use of ICS, and asthma immunological therapies. T2 inflammation leads to reduced expression of antimicrobial proteins, goblet cell hyperplasia and increased mucus production which subsequently can lead to increased bacterial burden, increased respiratory symptoms and subsequentially, increased pulmonary exacerbations [
13].
Another novel finding of our study is that longitudinal analysis showed that ppFEV1 decline correlated with increased AEC and IgE even when adjusting for age, gender, presence of
Pseudomonas spp., MRSA, other bacterial spp.,
Aspergillus spp, and other fungi. Our population is older, with a mean of 35 years, and also has a higher prevalence of
Pseudomonas (70%), MRSA (32%), and Mycobacteria (21%) than previously reported in other population cohorts. Despite the high infection rates, only 23% of the patients had frequent (> 2) exacerbations/year at any time during the study period. Moreover, neither AEC nor IgE were statistically associated with any bacterial strain. We believe this indicates that the relationship between AEC and pulmonary exacerbations is independent of bacterial presence. While the exact nature of this relationship is unclear, research has shown that eosinophils participate in pulmonary pathology [
14]. This could indicate a direct impact of the T2 phenotype on pulmonary structure and performance. Moreover, some of these CF pulmonary exacerbations could have been triggered by eosinophil-mediated changes in mucus production and airway expression of inflammatory mediators.
While bacteria were not correlated with either AEC or IgE, the E/R/S fungi were almost statistically-significantly correlated (p = 0.076) with elevated AEC. We decided to evaluate these fungi despite their low sample size as
Exophiala spp
, Rasamsonia spp
, and Scedosporium spp. have been shown to cause pulmonary disease in CF patients [
15,
16]. The lack of statistical significance could be due to grouping the fungi during analyses or can be due to fungi-induced diseases, as well as underpowering due to the small cohort included in our analysis. On univariate analysis, there was no correlation between AEC and IgE and
Aspergillus spp despite the high prevalence of
Aspergillus in our population (69%). In the models with covariates, exacerbation rates did not correlate with bacterial colonization but did correlate with
Aspergillus colonization suggesting that perhaps frequent antibiotic and steroids exposure increases the risk of
Aspergillus colonization. This is an interesting finding as
Aspergillus can generate a T2 response and has been associated with ABPA in the CF population [
7,
17] observations that warrant further studies.
UVA is situated in Charlottesville, Virginia which is on the asthma belt. Most of our subjects come from rural Virginia areas and work on farms or own farmlands and frequently work with animals and are exposed to multiple allergens. Virginia is a hot-humid, forest-heavy state. Our center’s geographical location and humid weather exposes our patients to a large variety and concentration of pollen and mold allergens. We recognize that a weakness of this study is the utilization of self-reported or physician-diagnosed asthma, and it is possible that the true prevalence of asthma is smaller. To avoid possible bias when using asthma as a variable in the analysis, we chose not to explore this diagnosis. According to the Virginia Department of Health, the 2016 asthma prevalence in Virginia was 8.6%, compared to the national average asthma prevalence of 7.7%. Environmental exposures and/or asthma exacerbations could underlie both correlations between AEC and pulmonary exacerbations and E/R/S respectively. The geographical environment may skew our patient population towards a T2 phenotype. Our patient population is also older with a mean age at study time of 35 which could also be a confounder as there is an increase the risk of sensitization in our patient cohort with increasing age. We found age to be a statistically significant covariate in some models with covariates and, as expected, age correlated with ppFEV1 decline.
There are several limitations to this study and further research is required to better understand the implications of elevated peripheral absolute eosinophils counts in CF disease. We are limited by our study’s retrospective design, unique geographical distribution, and being a single center study with a limited number of patients. Given the retrospective limitations of the study, we could not account for oral steroid use during the study period. These are frequently prescribed in our patient population by outside providers. Highest average AECs were used for univariate analysis to account for this limitation. The AECs obtained during exacerbations were included in the analyses as we think they are a marker of T2 inflammation regardless of trigger (environmental allergens, antibiotics, or idiopathic T2 inflammation itself). Another limitation of our study is the paucity of IgE data. In addition, as mentioned earlier, our study grouped together Exophiala spp, Rasamsonia spp, and Scedosporium spp due to the low number of cases, which may impact findings.
In conclusion, this study has shown that in CF patients, AEC correlated with frequent pulmonary exacerbations requiring hospitalizations and/or intravenous antibiotics, independent of bacterial airway colonization. Treatments that target eosinophilia, such as Mepolizumab (anti-human IL-5 monoclonal antibody), have positive effects in CF patients with T2 inflammation [
3]. With this treatment and other biologics that target T2 inflammation in mind, this study’s findings provide an opportunity to address the eosinophilic phenotype in CF patients with current and new immunomodulators approved for asthma treatment as having the potential to lower the health care costs of hospitalization and decrease exposure to intravenous antibiotics.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.