Abstract
Background Recent studies demonstrated that the triple combination cystic fibrosis transmembrane conductance regulator (CFTR) modulator therapy elexacaftor/tezacaftor/ivacaftor (ETI) improves lung function and reduces pulmonary exacerbations in cystic fibrosis (CF) patients with at least one F508del allele. However, effects of ETI on downstream consequences of CFTR dysfunction, i.e. abnormal viscoelastic properties of airway mucus, chronic airway infection and inflammation have not been studied. The aim of this study was to determine the longitudinal effects of ETI on airway mucus rheology, microbiome and inflammation in CF patients with one or two F508del alleles aged ≥12 years throughout the first 12 months of therapy.
Methods In this prospective observational study, we assessed sputum rheology, the microbiome, inflammation markers and proteome before and 1, 3 and 12 months after initiation of ETI.
Results In total, 79 patients with CF and at least one F508del allele and 10 healthy controls were enrolled in this study. ETI improved the elastic modulus and viscous modulus of CF sputum at 3 and 12 months after initiation (all p<0.01). Furthermore, ETI decreased the relative abundance of Pseudomonas aeruginosa in CF sputum at 3 months and increased the microbiome α-diversity at all time points. In addition, ETI reduced interleukin-8 at 3 months (p<0.05) and free neutrophil elastase activity at all time points (all p<0.001), and shifted the CF sputum proteome towards healthy.
Conclusions Our data demonstrate that restoration of CFTR function by ETI improves sputum viscoelastic properties, chronic airway infection and inflammation in CF patients with at least one F508del allele over the first 12 months of therapy; however, levels close to healthy were not reached.
Tweetable abstract
The triple combination CFTR modulator therapy ETI improves viscoelastic properties of airway mucus, chronic airway infection and inflammation in CF patients with at least one F508del allele, but without reaching levels close to healthy https://bit.ly/3OGBp3m
Introduction
Cystic fibrosis transmembrane conductance regulator (CFTR) dysfunction leads to abnormal viscoelastic properties of airway mucus, impaired mucociliary clearance and mucus plugging, which sets the stage for chronic polymicrobial infection and neutrophilic inflammation that are key drivers of the onset and progression of structural lung damage in patients with cystic fibrosis (CF) [1–6]. The triple combination CFTR modulator therapy elexacaftor/tezacaftor/ivacaftor (ETI) showed unprecedented improvements in lung function and other clinical outcomes in pivotal clinical trials and post-approval observational studies in patients with CF with at least one copy of the common F508del allele or other responsive CFTR variants [7–16]. We recently demonstrated that ETI improves F508del-CFTR function in the airways of patients with one or two F508del alleles to ∼40–50% of normal CFTR activity [15]. This degree of restoration of CFTR function is superior to previous dual drug combinations of lumacaftor/ivacaftor (LUM/IVA) and tezacaftor/ivacaftor (TEZ/IVA) in F508del homozygous patients and comparable to ivacaftor monotherapy in CF patients with a G551D mutation [15, 17–21]. In a previous study, we showed that this level of rescue of F508del-CFTR function achieved by ETI leads to substantial improvement in airway mucus obstruction, as evidenced by reduced mucus plugging detected by magnetic resonance imaging and improved ventilation homogeneity detected by multiple-breath washout [14]. However, the effects of ETI on downstream consequences of CFTR dysfunction in the lungs, i.e. abnormal viscoelasticity of airway mucus and chronic inflammation in patients with CF have not been studied, and data of effects on the sputum microbiome remain limited [22, 23].
Therefore, the aim of this study was to assess the longitudinal effects of ETI on sputum viscoelastic properties, chronic airway infection and inflammation in patients with CF with at least one F508del allele throughout the first 12 months of therapy. To achieve this goal, we conducted a prospective observational study in 79 patients with CF compound-heterozygous for F508del and a minimal function mutation or homozygous for F508del and investigated rheological properties, microbiota and key inflammation markers implicated in lung disease progression, including neutrophil elastase (NE), interleukin (IL)-1β and IL-8 [4–6], in sputum samples at baseline and at 1, 3 and 12 months after initiation of ETI therapy. Furthermore, we performed proteomics analysis as an unbiased approach to determine effects of ETI on abnormalities of the CF sputum proteome and compared the results from patients with CF to those obtained from sputum of 10 age-matched healthy controls.
Methods
Additional details on methods are provided in the supplementary material.
Study design and participants
This prospective observational post-approval study (clinicaltrials.gov identifier NCT04732910) was approved by the ethics committee of Charité – Universitätsmedizin Berlin (EA2/220/18). Written informed consent was obtained from all patients, their parents or legal guardians. Patients were eligible to participate if they were aged ≥12 years, compound-heterozygous for F508del and a minimal function mutation or homozygous for F508del, and fulfilled the inclusion and exclusion criteria detailed in the supplementary material. Patients were recruited between August 2020 and July 2022 and samples were collected between August 2020 and December 2022. Healthy control subjects were age- and sex-matched nonsmoking volunteers without any medical history of respiratory disease. Sputum rheology, microbiome analysis, cytokine measurements and proteomics were assessed if patients provided sputum at baseline and 1 month (median 25.0 days, interquartile range (IQR) 21.0–29.0 days), 3 months (median 76.0 days, IQR 69.3–91.5 days) or 12 months (median 337.0 days, IQR 303.5–410.0 days) after initiation of therapy with the approved dose of elexacaftor 200 mg and tezacaftor 100 mg every 24 h in combination with ivacaftor 150 mg every 12 h (supplementary figure S1 and table S2). Due to limited amounts of sputum in some patients, not all measurements could be performed with every sputum sample (supplementary figure S1).
Sputum collection
Sputum samples from CF patients were collected after spontaneous expectoration, directly put on ice and assessed immediately for rheology and membrane-bound NE activity on sputum neutrophils, and stored at −80°C for subsequent analyses. Sputum samples from healthy controls were collected after induction with inhaled hypertonic saline (sodium chloride 6%).
Sputum rheology
Sputum rheology was measured using a cone and plate rheometer (Kinexus Pro+; Netzsch, Selb, Germany), as described previously [24]. The elastic modulus (storage modulus, G′) and viscous modulus (loss modulus, G′′) were directly extracted from the linear viscoelastic region of the amplitude sweep.
Sputum microbiome analysis
DNA extraction and next-generation sequencing of the v4 region of the 16S rRNA gene were performed as described previously [25]. The number of 16S copies (total bacterial load) and P. aeruginosa copies was determined using quantitative (q)PCR.
Sputum inflammation markers
IL-1β, IL-6, IL-8, IL-10, IL-17A, tumour necrosis factor (TNF)-α and interferon-γ concentrations in cell-free sputum supernatants were quantified using cytometric bead array kits (BD Biosciences, San Diego, CA, USA) according to the manufacturer's instructions. Free and membrane-bound NE activity was measured as described previously [26]. Protein intensities of other neutrophil serine proteases cathepsin G (CatG, CTSG) and proteinase 3 (PR3, PRTN3), and of the antiproteases secretory leukocyte protease inhibitor (SLPI) and α1-antitrypsin (AAT, SERPINA1) were determined from sputum proteomics analysis.
Sputum proteomics
Sputum samples were solubilised by adding 4% SDS buffer and incubated at 95°C for 10 min. Proteins were reduced, alkylated and treated with Benzonase. After single-pot solid-phase-enhanced sample-preparation protein clean-up, proteins were treated with PNGase F followed by a tryptic digest. Cleaned-up peptide samples were measured on an EASY-nLC 1200 System coupled to an Orbitrap HF-X mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) running on data-dependent acquisition mode as described previously [27].
Statistical analysis
Data were analysed using R 4.0.4. Clinical data are presented as mean±sd and were tested by t-test. Data from measurements in sputum from healthy controls and patients with CF are presented as median (IQR) and were compared by Mann–Whitney test. Changes in CF sputum between baseline and 1, 3 and 12 months after initiation of ETI were tested with a linear mixed-effect model with the patient identity as random effect and are reported as median (IQR). Correction for multiple comparisons was performed using the Benjamini–Hochberg procedure. p<0.05 was accepted to indicate statistical significance.
Results
Characteristics of study population
79 patients with CF were enrolled to assess anthropometry, spirometry, sweat chloride concentration, sputum viscoelastic properties, the sputum microbiome and inflammation markers, and the sputum proteome at baseline and at 1, 3 and 12 months after initiation of ETI (supplementary figure S1). Four patients presented with pulmonary exacerbations at follow-up and were excluded from the analysis. Clinical characteristics of CF patients at baseline and after initiation of ETI, and 10 age-matched healthy controls are summarised in table 1. Patients with CF had a mean±sd age of 30.2±9.8 years; 57.3% were F508del homozygous; and 44.0% had undergone previous dual CFTR modulator therapy with either LUM/IVA or TEZ/IVA (table 1). Concomitant therapies remained unchanged during the study (table 1 and supplementary results). Improvements in forced expiratory volume in 1 s (FEV1) % predicted, body mass index (BMI) and sweat chloride concentration (table 1) by ETI therapy were consistent with previous studies [7–15, 28, 29]. CFTR genotypes are provided in supplementary table S1.
Clinical characteristics of healthy controls and patients with cystic fibrosis (CF) at baseline and after initiation of elexacaftor/tezacaftor/ivacaftor (ETI)
ETI improves viscoelastic properties of CF sputum
To study the longitudinal effects of ETI on sputum viscoelastic properties, we performed rheological measurements and determined the elastic modulus (G′), viscous modulus (G″) and the mesh-pore size (ξ) of sputum samples from 10 healthy controls and 55 patients with CF at baseline and at 1, 3 or 12 months after initiation of therapy. In both groups, G′ predominated over G″, indicating a gel-like behaviour of healthy and CF sputum (figure 1a and b). In sputum from CF patients collected at baseline, G′ and G″ were increased and the mesh-pore size was decreased compared to healthy controls (figure 1). At 1 month after initiation of ETI, there was no change in G′, G″ or mesh-pore size compared to baseline. At 3 months on ETI, G′ and G″ were significantly reduced in sputum from CF patients. At 12 months on ETI, G′ and G″ remained reduced and the mesh-pore size was increased. Compared to healthy controls, G′ and G″ remained significantly elevated and mesh-pore size was still decreased in sputum from CF patients after 12 months of ETI therapy (figure 1).
Elexacaftor/tezacaftor/ivacaftor (ETI) therapy improves sputum viscoelastic properties in patients with cystic fibrosis (CF). Measurements of a) the elastic modulus (G′), b) the viscous modulus (G″) and c) the mesh-pore size (ξ) of sputum from healthy controls (n=10) and CF patients at baseline (n=55) and at 1 month (n=20), 3 months (n=40) and 12 months (n=33) after initiation of ETI therapy. Data are presented as group median (interquartile range). **: p<0.01 compared with baseline; #: p<0.05, ##: p<0.01, ###: p<0.001 compared with healthy controls.
ETI shifts airway microbiome composition towards healthy
To determine longitudinal effects of ETI on the airway microbiome, we performed 16S rRNA gene sequencing in 65 patients with CF at baseline and at 1, 3 or 12 months after initiation of therapy. At baseline, the most abundant genera were Pseudomonas, Streptococcus, Prevotella and Veillonella (figure 2a). Pseudomonas genus was detected in 43 (66%) out of 65 patients at baseline, 17 (74%) out of 23 patients at 1 month, 32 (59%) out of 54 patients at 3 months, and 13 (46%) out of 28 patients at 12 months after initiation of ETI (supplementary figure S2). The detection of P. aeruginosa by 16S rRNA gene sequencing showed a high concordance with the detection by conventional culture-based microbiology (supplementary table S3). In six patients who were P. aeruginosa positive at baseline, the pathogen was not detected by microbiome analysis, qPCR or conventional culture-based microbiology up to 12 months after initiation of ETI. At baseline, patients with CF had a higher bacterial load, a lower α-diversity, evenness and richness, a higher dominance, as well as a higher P. aeruginosa relative abundance and bacterial load compared to healthy controls (figure 2b–h). The total bacterial load did not change at any time point on ETI compared to baseline (figure 2b). However, ETI increased the α-diversity and evenness at 1, 3 and 12 months on ETI (figure 2c and d). Richness did not change at 1 month, but increased at 3 and 12 months on ETI (figure 2e). Dominance of the sputum microbiome of patients with CF was decreased at 1 and 3 months and showed a trend towards reduction at 12 months on ETI (figure 2f). Compared to baseline, the relative abundance of P. aeruginosa was unchanged at 1 month, decreased at 3 months and remained low, but was not significantly different with the number of samples available at 12 months after initiation of ETI (figure 2g and supplementary figure S1). Bacterial load of P. aeruginosa did not change at any time point (figure 2b). At 12 months on ETI, α-diversity, richness, dominance, relative abundance and P. aeruginosa bacterial load were still significantly different from healthy controls.
Elexacaftor/tezacaftor/ivacaftor (ETI) shifts the airway microbiome composition in patients with cystic fibrosis (CF) towards healthy. Paired analysis of sputum microbiome from healthy controls (n=10) and CF patients at baseline (n=65) and at 1 month (n=23), 3 months (n=54) and 12 months (n=28) after initiation of ETI therapy. a) Mean relative abundance of the most abundant genera (minimal abundance >5% in at least one sample) in sputum of healthy controls and CF patients at baseline and at 1, 3 and 12 months after initiation of ETI therapy. b) Total bacterial load was determined from the number of 16S copies measured by quantitative (q)PCR (log10 transformed). c) α-diversity was calculated by the Shannon index. d) Evenness was determined by Pielou's evenness index. e) Richness was estimated by the number of ribosomal sequence variants observed. f) Dominance was determined from the relative abundance of the most abundant amplicon sequence variant. g) Relative abundance of the species Pseudomonas aeruginosa. h) Bacterial load of P. aeruginosa was determined from the copy number of the oprL gene measured by qPCR. Data are presented as group median (interquartile range). *: p<0.05, **: p<0.01, ***: p<0.001 compared with baseline; #: p<0.05, ##: p<0.01, ###: p<0.001 compared with healthy controls.
ETI reduces airway inflammation
To determine the longitudinal effects of ETI on airway inflammation, we measured IL-1β, IL-6, IL-8 and TNF-α concentrations, i.e. key inflammation markers previously shown to be associated with CF lung disease progression [4], in sputum from 10 healthy controls and 62 CF patients at baseline and at 1, 3 or 12 months after initiation of ETI. At baseline, IL-1β, IL-8, IL-6 and TNF-α levels were elevated in CF compared to healthy sputum (figure 3). ETI significantly reduced IL-1β at 1 month and all subsequent time points after initiation of therapy (figure 3a). IL-8 was unchanged at 1 month, reduced at 3 months, and remained low, but was not significantly different with the number of samples available at 12 months (figure 3b). IL-6 showed a trend towards increase at 1 and 3 months and returned towards baseline at 12 months on ETI (figure 3c). TNF-α did not change at any time point after initiation of ETI (figure 3d). Compared to healthy controls, IL-1β, IL-8, IL-6 and TNF-α were still elevated in sputum from CF patients treated with ETI (figure 3).
Elexacaftor/tezacaftor/ivacaftor (ETI) therapy reduces airway inflammation in patients with cystic fibrosis (CF). Measurements of a) interleukin (IL)-1β, b) IL-8, c) IL-6 and d) tumour necrosis factor (TNF)-α in sputum from healthy controls (n=10) and CF patients at baseline (n=62) and at 1 month (n=24), 3 months (n=43) and 12 months (n=31) after initiation of ETI therapy. Data are presented as group median (interquartile range). *: p<0.05, **: p<0.01, ***: p<0.001 compared with baseline; ##: p<0.01 and ###: p<0.001 compared with healthy controls.
ETI reduces airway protease burden
Protease/antiprotease imbalance with increased NE activity in the airways plays an important role in the onset and progression of structural lung damage in CF [5, 6]. To determine the longitudinal effects of ETI on this protease/antiprotease imbalance in CF airways, we measured free NE activity in sputum and membrane-bound NE activity on sputum neutrophils, and determined the abundance of the neutrophil serine proteases CatG and PR3, as well as the antiproteases SLPI and AAT by label-free shotgun proteomics in sputum of healthy controls and CF patients at baseline and at 1, 3 and 12 months after initiation of therapy. At baseline, free and membrane-bound NE activity was elevated in CF compared to healthy sputum (figure 4a and b). ETI significantly reduced the activity of free NE in sputum from CF patients at 1, 3 and 12 months compared to baseline (figure 4a). Membrane-bound NE showed a trend towards reduction at 3 months and was significantly reduced at 12 months on ETI (figure 4b). CatG and PR3 were elevated in sputum of CF patients at baseline compared to healthy controls. At 1 month after initiation of ETI, there was a trend towards reduction of CatG and PR3. At 3 and 12 months on ETI, both CatG and PR3 were reduced (figure 4c and d). Measurements of antiproteases showed that SLPI was reduced and AAT was increased in sputum of CF patients at baseline compared to healthy controls (figure 4e and f). Under ETI therapy SLPI was increased at 3 and 12 months on ETI. AAT was significantly decreased in sputum from CF patients from 1 month onwards after initiation of ETI (figure 4e and f). Compared to healthy controls, free NE activity, as well as CatG, PR3 and AAT levels were still elevated in sputum from CF patients after 12 months of ETI therapy (figure 4).
Elexacaftor/tezacaftor/ivacaftor (ETI) therapy reduces airway protease burden in patients with cystic fibrosis (CF). a) Measurement of free neutrophil elastase (NE) activity in cell-free sputum supernatants from healthy controls (n=10) and CF patients at baseline (n=62) and at 1 (n=24), 3 (n=43) and 12 months (n=31) after initiation of ETI therapy. b) Measurement of membrane-bound NE activity on neutrophils in sputum from healthy controls (n=10) and CF patients at baseline (n=40) and at 3 (n=29) and 12 months (n=16) after initiation of ETI therapy. c–f) Log2 protein intensity of the neutrophil serine proteases c) cathepsin G (CatG) and d) proteinase 3 (PR3), and the antiproteases e) secretory leukocyte peptidase inhibitor (SLPI) and f) α1-antitrypsin (AAT) in sputum from healthy controls (n=10) and CF patients at baseline (n=45) and at 1 (n=14), 3 (n=43) and 12 months (n=13) after initiation of ETI therapy. Data are presented as group median (interquartile range). *: p<0.05, **: p<0.01, ***: p<0.001 compared with baseline; #: p<0.05, ##: p<0.01, ###: p<0.001 compared with healthy controls.
Relationship between sputum rheology, microbiome, inflammation and protease burden
To assess the relationship between sputum viscoelastic properties, the microbiome, airway inflammation and protease burden at baseline and in response to therapy, we determined correlations for all measured parameters 1) at baseline (supplementary figure S8a); 2) between baseline and change on ETI therapy (supplementary figure S8b); and 3) change on ETI therapy (supplementary figure S9). The results of these analyses are provided in the supplementary material.
ETI shifts the CF sputum proteome towards healthy
Finally, we performed label-free shotgun proteomics of sputum from 10 healthy controls and 45 patients with CF at baseline and at 1, 3 or 12 months after initiation of ETI as an unbiased approach to assess the effects of ETI on abnormalities of the CF sputum proteome. In total, 2986 proteins were quantified, of which 1644 (55%) were significantly different between healthy controls and patients with CF at baseline (supplementary figure S5a). The complete list of all identified proteins is provided in the supplementary material. At 1 month after initiation of ETI, 208 (7%) proteins were significantly different (supplementary figure S5b). At 3 months on ETI, 250 (8%) proteins and at 12 months 129 (4%) proteins were different compared to baseline (supplementary figure S5c and 5d). Among the 100 top differentially regulated proteins at 3 months, we found 27 to be upregulated and 73 to be downregulated in response to ETI (supplementary table S4). The increase of the most upregulated protein uteroglobin (SCGB1A1), as well as one of the most downregulated proteins (resistin, RETN) by proteomics analysis was confirmed by ELISA (supplementary figure S7). Overall, 1025 (34%) proteins were different between all five groups and hierarchical cluster analysis revealed two major clusters (figure 5a). Cluster 1 contains proteins with a lower abundance in CF compared to healthy sputum. Gene set enrichment analysis (GSEA) of Gene Ontology (GO) terms showed that these proteins are involved in processes such as protein targeting to the endoplasmic reticulum (ER) and the plasma membrane (figure 5a). Cluster 2 includes proteins that are expressed at higher levels in sputum from CF patients compared to healthy individuals, and is enriched in proteins linked to immunological and inflammatory processes, such as neutrophil activation and positive regulation of immune system processes (figure 5a). After initiation of ETI, both proteome clusters in CF sputum shifted towards the protein signature pattern of healthy controls (figure 5a). However, at 12 months after initiation of ETI, there were still 1264 significantly differentially regulated proteins present compared to healthy controls (supplementary figure S5e). GSEA of GO terms showed that ETI treatment increased the abundance of proteins belonging to the GO terms signal recognition particle (SRP)-dependent co-translational protein targeting to the membrane, protein targeting to the ER and epithelial cell differentiation (figure 5b). Furthermore, proteins involved in inflammatory processes, such as granulocyte activation, neutrophil-mediated immunity including neutrophil activation and degranulation showed increased levels in sputum of patients with CF compared to healthy controls and were reduced after initiation of ETI (figure 5b). Although these changes are sustained or even increased over 12 months of ETI therapy, the GO terms at 12 months are still different compared to healthy controls (figure 5b). The complete list of significantly different GO terms is provided in supplementary figure S6.
Elexacaftor/tezacaftor/ivacaftor (ETI) therapy shifts the sputum proteome of patients with cystic fibrosis (CF) towards healthy. a) Heatmap of significantly differentially expressed proteins (cut-off p<0.01) in sputum from healthy controls (n=10) and CF patients at baseline (n=45) and at 1 (n=14), 3 (n=43) and 12 months (n=13) after initiation of ETI therapy clustered by hierarchical clustering in a cluster with downregulated proteins (cluster 1) and a cluster with upregulated proteins (cluster 2) in CF patients at baseline compared to healthy controls. Proteins that contribute most to representative enriched Gene Ontology (GO) terms are highlighted in the first column. b) Gene set enrichment analysis of GO terms in CF sputum at baseline compared to healthy controls (first column), in CF sputum at 1 (second column), 3 (third column) and 12 months (fourth column) after initiation of ETI compared to baseline and in CF sputum at 12 months after initiation of ETI compared to healthy controls (fifth column) ordered by normalised enrichment score (NES) (false discovery rate <0.01). SRP: signal-recognition particle; ER: endoplasmatic reticulum.
Discussion
This is the first study to assess the downstream effects of restoration of CFTR function by the triple combination CFTR modulator therapy with ETI on key abnormalities of sputum viscoelastic properties, the airway microbiome and chronic airway inflammation in adolescent and adult patients with CF with at least one F508del allele throughout the first 12 months of therapy. In this post-approval study in a real-world setting, the clinical responses to ETI on lung function, BMI and sweat chloride concentration were in line with results of controlled clinical trials and other observational studies [7–15, 28, 29]. Our rheological measurements show that treatment with ETI improved the viscoelastic properties by reducing the elastic (G′) and viscous (G″) moduli of sputum from patients with CF from 3 months onward. Furthermore, initiation of ETI led to a rapid and sustained shift of the microbiome composition towards a healthy airway microbiome and to a sustained reduction in airway inflammation and protease burden in patients with CF. In addition, using proteome analysis as an unbiased approach, we found that under ETI therapy alterations in the sputum proteome of CF patients were shifted towards that of healthy controls. However, in comparison to sputum from a healthy control group, our data also demonstrate that sputum rheology, the microbiome, inflammation markers and protease burden, as well as proteome signatures, remain abnormal in CF patients during the first 12 months of ETI therapy. Collectively, our results obtained with a multi-method approach provide novel insights into the impact of improvement of CFTR function by ETI on mucus properties, chronic airway infection and inflammation, and protease burden that constitute key risk factors in the onset and progression of CF lung disease.
Abnormal viscoelasticity of airway mucus, caused by increased mucin concentration due to airway surface dehydration, increased concentration of DNA released from inflammatory cells, and increased disulfide cross-linking of mucins caused by reactive oxygen species generated in neutrophilic airway inflammation, is a key feature of CF that correlates with the severity of lung function impairment as well as airway infection and inflammation in patients [1–3, 30–33]. Previous studies showed that LUM/IVA improves viscoelastic properties of CF mucus [34], and that ETI reduces mucin concentration and improves mucociliary transport in CF human bronchial epithelial cells in vitro [35]. We show for the first time that ETI improves viscoelastic properties of sputum from patients with CF in vivo. However, as improvement in mucus hydration is expected to occur shortly after initiation of ETI, it is surprising that the elastic (G′) and viscous (G″) moduli were not changed compared to baseline at 1 month on therapy, and were only improved at the later time points. We speculate that irreversible cross-linking of mucins may persist with improvement of airway surface hydration, that improvement in mucus rheology may require clearance of cross-linked mucus gels, as well as secretion of newly formed mucins under conditions of less oxidative stress in the airways, and that this process may take longer than 1 month of ETI therapy. Despite the substantial improvement in the viscoelastic properties at 3 and 12 months on ETI, they remained substantially elevated compared to sputum from healthy controls indicating that novel mucolytics [3, 36] may still be needed to fully restore mucus properties in patients with CF.
Pathological CF mucus causing impaired mucociliary clearance sets the stage for chronic polymicrobial infection and inflammation, and improvement of viscoelastic properties of the mucus has therefore the potential to limit these abnormalities in CF airways [1, 2, 4, 5, 37]. However, studies investigating the effects of prior CFTR modulator drugs on airway infection and inflammation reported mixed results [38–43]. One study investigating the effects of LUM/IVA that is less effective in restoring CFTR function than ETI showed an increase in α-diversity of the microbiome accompanied by improved IL-1β, while another study observed no change in α-diversity, but a trend towards a reduction in the relative abundance of P. aeruginosa [20, 38, 39]. Treatment with IVA in patients with a G551D mutation, which is similarly effective in restoring CFTR function as ETI in patients with at least one F508del allele [15, 18], led to a reduction in the relative abundance of P. aeruginosa and an increase in the α-diversity of the airway microbiome, which was also associated with improvements in airway inflammation [40–43]. In our study, we found a similar change of the microbiome composition including a reduction in the relative abundance of P. aeruginosa in response to ETI therapy in patients with one or two F508del alleles. These findings are consistent with a small previous study showing an increase in the microbiome α-diversity and evenness, and a trend towards the reduction of Pseudomonas spp. with ETI [22]. Despite the reduced relative abundance, the absolute amount of P. aeruginosa did not change after initiation of ETI. We speculate that this finding may be explained by an increase in the richness of commensal bacteria also contained in the microbiota of healthy controls rather than an absolute reduction in P. aeruginosa. Of note, long-term studies with IVA also reported a rebound of improvements of the microbiome after the first year of treatment [41]. It is therefore encouraging that we did not observe a rebound of microbiome parameters at 12 months after initiation of ETI. Furthermore, in six (14%) out of 43 patients who were P. aeruginosa positive at baseline, the pathogen was not detected by microbiome analysis, qPCR or conventional culture-based microbiology up to 12 months indicating potential eradication. These results are in line with recent studies showing that ETI leads to partial restoration of CFTR function and bacterial killing in CF macrophages that play an important role in host defence against CF pathogens [44, 45]. However, long-term studies over several years will be necessary to investigate the microbiome trajectory and determine whether the beneficial effects of ETI observed in our study persist over time.
In addition to these improvements of the sputum microbiome, initiation of ETI also led to improvements in key inflammation markers such as IL-1β and IL-8 that have been implicated in CF lung disease progression [4]. The effects on these pro-inflammatory cytokines observed in our study are in line with those reported for IVA in CF patients with a G551D mutation, and greater than those observed for LUM/IVA in F508del homozygous patients [39–42]. Besides elevated pro-inflammatory cytokines, neutrophilic airway inflammation is associated with sustained proteolytic activity that is caused by a protease/antiprotease imbalance and constitutes a key driver of lung damage and disease progression in CF [4–6]. In this context, increased activity of NE, but also of the other neutrophil serine proteases CatG and PR3, are important contributors to the protease burden of the CF lung [4–6, 46–50]. In our study, we observed a decrease in NE activity and reduced sputum levels of CatG and PR3 after initiation of ETI. These effects on neutrophil serine proteases were durable and associated with improvements of AAT and SLPI levels [4, 6, 46, 51], suggesting an overall improvement of protease/antiprotease imbalance in the lungs of CF patients under ETI therapy. However, long-term studies will be required to determine the impact of the residual protease burden on progression of structural lung damage in patients with CF.
Our studies of established markers of CF airway inflammation were complemented by longitudinal sputum proteomics analyses as an unbiased approach to assess effects of ETI on molecular pathways altered in CF lung disease. In line with reduced inflammation markers and neutrophil serine proteases, we observed a global reduction of proteins linked to inflammatory processes, such as neutrophil activation and degranulation that was durable at 12 months after initiation of ETI. Furthermore, proteins involved in airway epithelial cell differentiation that were reduced in sputum from CF patients at baseline were upregulated under ETI, which may reflect improved epithelial homeostasis. In addition, we found that ETI increased the abundance of proteins related to SRP-dependent co-translational protein targeting to the ER that were previously identified as high-confidence members of the F508del–CFTR interactome [52]. We speculate that pharmacologically corrected F508del–CFTR may have specific binding partners during processing and trafficking that are only part of the CFTR interactome when the protein is folded and not prematurely degraded. This notion is supported by a recent study showing large differences in the wild-type and F508del–CFTR interactomes [53].
Using a multi-method approach, our study also provides insights into the relationship between different downstream effects of CFTR dysfunction in the airways and the impact of pharmacological restoration of CFTR function in patients. First, we found that FEV1 % pred, the airway microbiome and airway inflammation correlated across our study population with a large range in lung disease severity. Second, our data show that the magnitude of change in sputum viscoelastic properties, the microbiome and inflammation markers after initiation of ETI is dependent on the baseline level of each parameter. Third, regarding the level of change within and across the different sputum outcome measures, we generally observed strong correlations within parameters of rheology, microbiome and inflammation, and changes in the microbiome correlated with changes in inflammation markers including protease activity. Of note, we did not observe correlations between changes in FEV1 % pred or sweat chloride with any of the sputum parameters, suggesting that more sensitive outcome measures are needed to assess effects of CFTR-directed therapeutics on airway inflammation and infection in patients in the clinical care setting. The physiological and clinical consequences of these data are two-fold. First, our results link the rescue of CFTR function by ETI to effects on immediate downstream consequences of CFTR dysfunction in the airways of patients with CF. The improvements observed in the sputum outcome measures under ETI therapy provide a potential mechanism underlying the improvement in lung function and reduction in pulmonary exacerbations observed in the clinical trials and real-world observational studies [7–15, 28, 29]. Conversely, our data also show substantial residual abnormalities under ETI that provide a rationale for novel mucolytic, anti-inflammatory and anti-infective therapies for patients with CF.
This real-world observational study has limitations. First, we included patients with a large range in age and baseline lung function, possibly influencing the observed treatment effects. Second, microbiome analysis was based on 16S rRNA sequencing instead of metagenomics, which would have increased the resolution, possibly detecting further differences in microbiome composition and response to ETI therapy. While inclusion of a healthy control group enabled us to determine residual abnormalities in CF patients on ETI, the sputum induction required in healthy individuals may have an impact on sputum properties such as rheology. Furthermore, all measurements were performed in CF patients who expectorated sputum at baseline and after initiation of ETI, biasing our study towards patients with more advanced lung disease. In this patient population, despite the substantial improvements of sputum outcomes observed, sputum viscoelasticity, inflammation markers and protease burden remained elevated, and abnormal proteome signatures were only partially restored compared to healthy controls during the first 12 months on ETI. Whether long-term treatment in patients with established lung disease, or early initiation of ETI therapy in children with CF with preserved lung function [54], can achieve a more complete resolution of abnormal mucus properties, airway dysbiosis and inflammation remains to be determined in future studies. However, our data support the use of the sputum outcome measures utilised in our study for the assessment of effects of novel therapeutic interventions targeting increased mucus viscoelasticity, airway infection and inflammation in patients with CF and possibly other muco-obstructive lung diseases.
In summary, our study demonstrates that ETI leads to improvements in sputum viscoelastic properties, chronic airway infection and inflammation as well as abnormalities in airway proteome signatures in patients with CF aged ≥12 years with one or two F508del alleles that are sustained throughout the first 12 months of therapy. Our data on residual disease activity under this highly effective CFTR modulator therapy also indicate that additional therapeutic strategies may be needed to control airway infection and inflammation in adolescent and adult CF patients with chronic lung disease.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material ERJ-02153-2022.Supplement
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-02153-2022.Shareable
Acknowledgements
The authors thank the patients with CF for their participation in this study; M. Albrecht (University Hospital Schleswig-Holstein/Campus, Lübeck, Germany), S. Mayer (University of Heidelberg, Heidelberg, Germany), M. Daniltchenko, M. Drescher, A. Rohrbach, K. Seidel and J. Tattersall-Wong (Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Berlin, Germany) for excellent technical assistance; C. Labitzke and E. Halver (Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin) for excellent documentation; and our clinical colleagues for clinical care of study participants.
Footnotes
Clinical trial registered with www.clinicaltrials.gov (NCT04732910). Pseudonomysed microbiome data and scripts will be shared. Due to data safety constrictions, for proteomics only pooled data will be shared. Data will become available upon acceptance of the manuscript.
This article has an editorial commentary: https://doi.org/10.1183/13993003.01008-2023
Conflict of interest: J. Röhmel reports payment for presentations at educational events from Vertex Pharmaceuticals, outside the submitted work. D. Lauster has received funding from Deutsche Forschungsgemeinschaft for work related to this manuscript. M. Stahl reports funding from Deutsche Forschungsgemeinschaft related to this manuscript, and grants from Vertex Pharmaceuticals and Mukoviszidose e.V. (German CF Foundation) outside the submitted work, payment for work on an advisory board from Vertex Pharmaceuticals, and is elected, unpaid secretary of the group on paediatric CF of Assembly 7 of the ERS. P. Mertins received funding from Deutsche Forschungsgemeinschaft and the German Federal Ministry of Education and Research related to this work. S.Y. Graeber reports grant support from Mukoviszidose e.V. (German CF Foundation) and Vertex Pharmaceuticals Incorporated outside the submitted work, with payments made to the author's institution. Chiesi GmbH and Vertex Pharmaceuticals Incorporated have provided personal fees for presentations and advisory boards. M.A. Mall declares that he has received funding from Deutsche Forschungsgemeinschaft and German Ministry for Education and Research for work related to this manuscript. In the past 36 months, he has received grants from Vertex Pharmaceuticals, personal fees for consultancy from Boehringer Ingelheim, Arrowhead Pharmaceuticals, Vertex Pharmaceuticals, Santhera, Sterna Biologicals, Enterprise Therapeutics, Antabio and Abbvie, lecture fees from Boehringer Ingelheim, Arrowhead Pharmaceuticals and Vertex Pharmaceuticals, travel reimbursement from Boehringer Ingelheim and Vertex Pharmaceuticals, and personal fees for participation in advisory boards from Boehringer Ingelheim, Arrowhead Pharmaceuticals, Vertex Pharmaceuticals, Santhera, Enterprise Therapeutics, Antabio, Kither Biotech, Abbvie and Pari. Additionally, he served as a member of the Board and Vice-President of the European Cystic Fibrosis Society from 2014 to 2020. All other authors have nothing to disclose.
Support statement: This study was supported by grants from the German Research Foundation (CRC 1449 – project 431232613; sub-projects A01, B03, C03, C04, Z01, Z02; and project 450557679) and the German Federal Ministry of Education and Research (82DZL009B1). The funders had no role in the design, management, data collection, analyses or interpretation of the data or in the writing of the manuscript or the decision to submit for publication. S. Thee, M. Stahl and S.Y. Graeber are participants of the BIH-Charité Clinician Scientist Program funded by the Charité – Universitätsmedizin Berlin and the BIH. Funding information for this article has been deposited with the Crossref Funder Registry.
- Received November 9, 2022.
- Accepted May 21, 2023.
- Copyright ©The authors 2023. For reproduction rights and permissions contact permissions{at}ersnet.org
References
