Patients with cystic fibrosis (CF) are highly susceptible to infection and colonization of pulmonary epithelia. Repeated and chronic infections may affect disease course and efficacy of host immune protection. Higher Interleukin (IL)-7 serum levels, indicating impaired T-cell response to IL-7, have been described for chronic viral and mycobacterial infections.
Time course measures of IL-7 serum concentrations in patients with CF (n = 164; n = 78 for the second time point) and healthy controls (n = 60) were done. CF patients were characterized for disease severity parameters as well as infection status and association with IL-7 serum levels was determined.
CF patients had significantly higher IL-7 serum concentrations as compared to healthy controls (9.79 pg/ml, IQR 6.76–13.6 versus 4.55 pg/ml, IQR 2.76–9.51, p < .001). IL-7 serum levels were negatively correlated with individual CF patient's BMI (r = −0.19, p = .021) and a tendency of increased IL-7 levels in Staphylococcus aureus infected CF patients was found. Linear regression of multiple parameters revealed significant negative correlation of FEV1%pred with IL-7 serum concentrations in patients with CF (ß-coefficient: −0.04, 95% confidence interval [−0.08; −0.003], p = .034). Time course analyses after 1 year +/− 6 months showed increased IL-7 serum levels (time point 1:9.26 pg/ml, IQR 6.94–13.12 time point 2:10.86 pg/ml, IQR 9.14–14.76, p = .016) that correlated negatively with decreased FEV1%pred during CF disease course.
High IL-7 serum levels were found in CF patients and correlated with impaired lung function during CF disease course. As a candidate biomarker of T-cell dysfunction, higher IL-7 serum level may also indicate worsened immune competence of patients with CF.
Cystic Fibrosis (CF) pathology affects multiple organs but pulmonary disease predominantly influences morbidity and mortality of affected patients. Chronic pulmonary colonization e.g. by bacterial species is a typical feature of CF [
]. Immune surveillance of pulmonary tissues from CF patients is hampered by disease inherent pathology of the epithelia and increasing evidence suggested that immune cells are also affected by the lack of CFTR expression [
]. The important role of TH1 cells in host immunity against viral and intracellular bacterial pathogens is well established. However, TH1 mediated immune protection against chronic viral infections may fail as a consequence of antigen persistence and/or inflammation [
]. This rendered IL-7 serum concentrations a biomarker candidate for impaired T-cell response against persistent bacterial infection. Since an initial report indicated a role of exhausted T-cell immunity in a CF patient [
], we investigated a potential role of IL-7 serum concentrations in well-characterized cohorts of CF patients and healthy controls.
2.1 Study subjects and design
34 children and young adults diagnosed with CF were recruited at the University Children's Hospital, Duesseldorf (CFyouth cohort). 130 adult patients with CF were recruited in the Department of Pneumology, Ruhrlandklinik at the University Hospital Essen (CFadult cohort). Inclusion criteria were previous diagnosis of CF based on genetic examinations (see Supplementary Tables 2 and 3 for underlying mutations). Five patients were included for whom identification of underlying mutations failed (Supplementary Table 3) but diagnostic criteria according to published guidelines [
] confirmed CF disease for these individuals. Control donors (n = 60; control cohort) were recruited at University Children's Hospital, Duesseldorf and consist of children and adolescents initially presenting with unclear symptoms. Afterwards we excluded those diagnosed as autoimmune or infectious disease cases. For cohort characteristics see Table 1. Written informed consent was obtained from all study participants or their legal guardians. All CF patients were determined for atypical mycobacterial infections. Details for this have already been published [
]. Not all donors have been included for each analysis. The exact numbers were indicated in respective figure legends.
Table 1Study cohort characteristics:
age; median ofyears (range)
sex distribution (m/f)
BMI (kg/m2); median(range)
CRP (mg/dl); median(range)
Leukocytes (x1000/μl);Median (range)
Exacerbations (n, %)
P. aeruginosa (n, %)
A. fumigatus (n, %)
S. aureus (n, %)
M. abscessus (n, %)
Pancreas insufficiency(n, %)
Diabetes (n, %)
BMI: body mass index; FEV1: % of predicted Forced Expiratory Volume; n: number of individuals; nd: not defined; na: not applicable. P.: Pseudomonas; A.: Aspergillus; S.: Staphylococcus; M.: Mycobacterium.
This study was approved by the ethics committees of the University Hospital Duesseldorf (Internal Study No. 4505 and 4844) and the University Hospital Essen (17-7365-BO).
2.3 Assessment of clinical CF features
Spirometry was performed according to ATS/ERS guidelines and expressed as percent predicted Forced Expiratory Volume (FEV1%pred). Exocrine pancreatic insufficiency was diagnosed by reduced laboratory chemical values of pancreatic elastase-1 in faeces or by the necessary use of a substitutional therapy with pancreatic enzymes in case of lasting steatorrhea. Cystic fibrosis-related diabetes was diagnosed based on the criteria of the American Diabetes Association (2016). Patients were considered as chronically infected with S. aureus, P. aeruginosa or A. fumigatus if they had >50% positive cultures of sputum, throat swabs or bronchoalveolar lavage fluid within one year before blood sampling. Exacerbation was diagnosed if patients had deteriorated symptoms requiring hospitalization and non-scheduled intravenous antibiotic treatment.
3. Serum IL-7 measurement
Serum of study participants was stored at −80 °C for concomitant IL-7 measurement at the University Childrens Hospital in Duesseldorf. Serum IL-7 concentrations were measured as described previously [
]. In brief, we used a commercially available highly sensitive ELISA (Human IL-7 Quantikine HS ELISA kit, R&D Systems) according to manufacturer's instructions. Because of low IL-7 serum concentrations, we included standards on each plate to increase the accuracy of these assays. All samples were measured at least as duplicates. Optical density was determined using an ELISA reader (Infinite M200, Tecan). IL-7 concentrations were calculated from the standard curve by applying 4-parametric logistic regression (Magellan version 7.2, Tecan).
4. Statistical analyses
Because serum IL-7 values did not pass normality test (D́Agostino & Pearson normality test), non-parametric distributions were assumed and statistical tests were chosen accordingly. Medians and interquartile ranges (IQR) or percentages were presented as descriptive statistics of continuous and categorical variables, respectively. Mann-Whitney U tests (two-sided) were used to compare the raw observed differences in IL-7 between patients and controls or between subgroups of patients. For the analysis of paired parameters (comparisons between time point 1 and 2), Wilcoxon signed rank test was used. Correlations are reported as Spearman correlation coefficients. Furthermore, a multiple linear regression model was used to estimate the difference in IL-7 between patients and controls adjusting for potential confounders. More specifically, a linear regression model with IL-7 as dependent variable and cohort, age and sex as independent variables was fitted. Thereby, the variable cohort denotes whether an individual belongs to the cohort of patients or controls, respectively. In addition, a multiple linear regression model was used to estimate the adjusted influence of the independent variables age, sex, BMI, FEV1%pred, pancreatic insufficiency, diabetes, and chronic infections with S. aureus, P. aeruginosa as well as A. fumigatus on the dependent variable (i.e. serum IL-7) in CF patients. P-values of two-sided tests ≤0.05 were considered statistically significant. SPSS-Statistics (Version 24.0.0) and Graph Pad Prism 7 (Version 7.0a) were used for statistical analyses. Graph Pad Prism 7 software was used for figure preparation.
In total, 164 patients with CF and 60 healthy controls (for study group characteristics see Table 1) were included for concomitant serum IL-7 measurements. CF patients showed significantly increased IL-7 serum concentrations (median: 9.79 pg/ml, IQR 6.76–13.6) as compared to healthy controls (median: 4.55 pg/ml, IQR 2.76–9.51) (p < .001; Fig. 1A). In multiple linear regression analyses study group classification was most influential (ß-coefficient: 4.59, 95% confidence interval [2.65; 6.53], p < .001) whereas other factors, like age or sex, were less relevant (Supplementary Table 1). Controls were recruited at a childrens hospital and were similar in age distribution to the subgroup CFyouth patients from the same hospital (for details see Table 1). However, age-matched subgroups showed comparable differences between CF patients and controls (Fig. 1B). The vast majority of CF patients were stable at inclusion time point and only a minor subgroup had increased CRP values or showed symptoms of exacerbation. First, we analyzed a possible effect of inflammation-associated increased CRP levels on IL-7 serum levels. In this regard we classified CF patients according to CRP levels (CRP > 5 mg/dl: n = 13, 1 mg/dl < CRP < 5 mg/dl: n = 41). CF patients with high CRP and those with moderately increased CRP levels showed no differences in IL-7 serum concentrations as compared to CF patients with low CRP (Supplementary Fig. 1A, p = .125 and 0.151, respectively). Exacerbation of CF disease was also not associated with differential IL-7 serum levels indicating that neither acute inflammation nor acute exacerbation events confounded study group comparisons (Supplementary Fig. 1B, p = .259).
Lymphopenia has been described as an influential factor for IL-7 serum levels [
] and, therefore, we next compared leukocyte counts and IL-7 serum levels of CF patients. No association between leukocyte numbers and IL-7 were detectable (Supplementary Fig. 1C, r = 0.11, p = .168) and this did not suggest lymphopenia as the cause for increased IL-7 serum levels.
CF patients are highly susceptible to atypical mycobacterial infections and M. tuberculosis infection was recently found to affect IL-7 consumption [
]. Therefore, we stratified CF patients according to previous diagnosis of M. abscessus and identified eight CF patients (4.9%) with a history of M. abscessus infection. Although the median value of IL-7 serum concentrations was higher in M. abscessus positive (14.39 pg/ml as compared to 9.68 pg/ml in M. abscessus negative; data not shown), the subgroup was too small to reach necessary power for statistical analyses. As expected, other pulmonary infections were frequent within the study group of CF patients: 23.2% of CF patients were positive for Aspergillus (A.) fumigatus, 42.1% for P. aeruginosa, and 39.6% for S. aureus. Whereas P. aeruginosa and A. fumigatus infections had no detectable influence IL-7 serum levels (Fig. 2A ; left and middle graph, p = .896 and 0.643, respectively), S. aureus infection was accompanied by a tendency towards higher IL-7 serum concentration in CF patients (median 11.17 pg/ml, IQR 7.59–15.94 versus 9.36 pg/ml, IQR 6.42–12.85, p = .060) (Fig. 2A; right graph).
Next, we analyzed possible associations of CF disease severity on IL-7 serum levels. A moderate negative correlation of IL-7 serum concentrations with the BMI (r = −0.19; p = .021; Fig. 2B, left graph) but no association with FEV1%pred (r = −0.12; p = .135; Fig. 2B, right graph) was found. Frequent CF epiphenomena (i.e. pancreatic insufficiency, cystic fibrosis-related diabetes) were not associated with differential IL-7 serum concentrations (Fig. 2C, p = .194 and 0.285, respectively). Since several factors reflecting disease course as well as infection of CF patients may contribute to aberrant high IL-7 serum levels, we next performed multi-factorial linear regression analysis to identify key influential factors (for details see Methods section). Of note, we detected significant dependency between FEV1%pred and IL-7 serum levels of CF patients on the basis of multiple parameters (ß-coefficient: −0.04, 95% confidence interval [−0.08; −0.003], p = .034), whereas other factors were not associated (Table 2). These results indicated that impaired lung function of CF patients may be linked to increased IL-7 serum values. Since progressive worsening of lung function is a typical feature of CF, we next addressed the question if IL-7 serum concentrations change during CF disease course. We performed time course analyses (second time point (TP) after 1 year +/− 6 months (median 1.1 (0.6–1.5) years) and measured IL-7 serum concentrations for a subgroup of CF patients (n = 78). FEV1%pred and BMI values were also determined for the follow-up time point. FEV1%pred significantly decreased (TP1: median 53.5%, IQR 37.0–71.5%, TP2: median 46.5%, IQR 35.0–71.3%; p = .024; Fig. 3A , left graph) in CF patients whereas BMIs were similar during this period (p = .115; Fig. 3A, right graph). Notably, IL-7 serum concentrations increased significantly in CF patients during this period (time point 1: median 9.26 pg/ml, IQR 6.94–13.12 time point 2: median 10.86 pg/ml, IQR 9.14–14.76, p = .016) (Fig. 3B). Next, we determined possible association between FEV1%pred and IL-7 values for the second time point and found a significant negative correlation between lung function and IL-7 serum levels (Fig. 3C left graph, r = −0.28, p = .013). No correlation for the BMI of CF patients and IL-7 was detected (Fig. 3C right graph, r = −0.17, p = .135). These results strengthened our assumption about a possible role of IL-7 in CF pathogenesis contributing to worsening of lung functions.
Table 2Multiple linear regression analysis of IL-7 in CF patients as dependent variable and subsequent parameters as independent parameters:
BMI: body mass index; FEV1%pred: % of predicted Forced Expiratory Volume; P.: Pseudomonas; S.: Staphylococcus; A.: Aspergillus; *significant difference.
This study suggested an important role of IL-7-dependent immune functions – reflected by high IL-7 serum concentrations – in CF patients. Decreased lung functions indicate potential relevance of IL-7 as a biomarker for CF disease course and T-cell based immune protection.
In the present study we found higher IL-7 serum concentrations in CF patients as compared to healthy controls. Increased IL-7 serum concentrations have been described for different pathologic conditions including lymphopenia [
]. In acute tuberculosis, increased IL-7 serum levels were accompanied by impaired IL-7 sensitivity and IL-7 mediated T-cell functions. T-cell numbers were not causative for higher IL-7 serum levels of tuberculosis patients [
], it is tempting to speculate that this may affect T-cell functions including IL-7 consumption. However, we did not find an association between CRP values and IL-7 but this may be due to the fact that only chronic inflammation would affect T-cell immunity and CRP is a marker of acute inflammatory events. We were not able to perform functional T-cell assays in the present study to investigate if high IL-7 serum levels are accompanied with impaired T-cell function. Hence, elucidation of underlying mechanisms needs to be done by future studies.
A tendency of higher IL-7 serum levels were detected in S. aureus infected CF patients suggesting that impaired IL-7 consumption is affected by this infection. In this regard high incidence rates of S. aureus infections already early during childhood of CF patients and resulting long-term colonization may be explanatory [
]. However, evidence that S. aureus infection influences IL-7 serum levels of CF patients from this study is limited and further studies need to be performed.
Multi-factorial analysis revealed an association between IL-7 serum concentrations and FEV1%pred. Higher IL-7 levels were accompanied by decreased FEV1%pred indicating worsened lung functions in CF patients with high IL-7 serum levels. Time course analyses confirmed increased IL-7 serum concentrations and decreased FEV1%pred values in CF patients. IL-7 and FEV1%pred correlated negatively at the second time point. FEV1%pred is widely used as a central surrogate marker for lung function with strong implications for treatment decisions including lung transplantation at late CF disease stages. Long-term follow-up studies may address the question if IL-7 serum levels can be used as an early marker to predict CF pathology influence on lung function.
The following are the supplementary data related to this article.
Table 2 CFTR mutation of individuals from the CF youth study group and Table 3 CFTR mutation of individuals from the CF adult study group
This study was supported in part by a grant from the Research Committee of the Medical Faculty, Heinrich-Heine-University , Duesseldorf to L. Olbrich and M. Jacobsen. The samples and data of adult patients with CF recruited in the Department of Pneumology at the University Hospital Essen were kindly provided by the West German Biobank. Special thank is granted to all participants for the willingness to contribute to this study.
Respiratory bacterial infections in cystic fibrosis.