Effect of elexacaftor/tezacaftor/ivacaftor on systemic inflammation in cystic fibrosis

Abstract

Background

Despite significant clinical improvements, there is evidence of persisting airway inflammation in people with cystic fibrosis (CF) established on elexacaftor/tezacaftor/ivacaftor (ETI) therapy. As CF is a multi-system disease, systemic immune profiles can reflect local inflammation within the lungs and other organs. Understanding systemic inflammation after ETI therapy may reveal important translational insights. This study aims to profile systemic inflammatory changes and relate these to the well-documented improvements observed with ETI therapy.

Methods

We conducted a single-centre longitudinal study with 57 CF subjects initiating ETI therapy. All participants were Phe508del homozygous or Phe508del/minimal function. Blood samples were collected pre-ETI and 3–12 months post-therapy initiation. Analyses included mass spectrometry-based proteomics, a multiplex immunoassay, and flow cytometry for peripheral immune cell counts and phenotype. Controls samples were provided by 29 age-matched healthy controls.

Results

Systemic inflammation reduced with ETI therapy; however, the immune profile remained distinct from healthy controls. ETI reduced neutrophil counts and was associated with a more mature, less inflammatory phenotype, as well as a shift towards an immune resolving state associated with increased CD206 expression. Cytokines known to influence neutrophil levels reduced with therapy. Despite ETI therapy, neutrophil and monocyte counts remained elevated compared with healthy controls. There was no obvious association between the ETI-related improvements in systemic inflammation and lung function.

Conclusions

Patients with CF showed evidence of persisting systemic inflammation despite ETI therapy, which may have long-term potentially adverse effects on respiratory and other organ systems.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Elexacaftor/tezacaftor/ivacaftor has led to dramatic health improvements for people with cystic fibrosis. Initial studies suggest this therapy influences systemic inflammation.

WHAT THIS STUDY ADDS

• Systemic inflammation reduced with therapy; however, the immune profile remained distinct from healthy controls. The most prominent improvements were within the levels and phenotypes of circulating neutrophils, as well as in the soluble immune mediators that regulate their activity.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The long-term consequences of persisting systemic inflammation after therapy need to be established, to identify potentially adverse effects on the lungs and other organ systems and to direct future therapeutic interventions.

Introduction

Elexacaftor/tezacaftor/ivacaftor (ETI) therapy has led to dramatic improvements in cystic fibrosis (CF) lung disease across various measures, including lung function, radiology and microbiology.¹,⁴ However, ETI reduces but does not fully resolve airway inflammation.^{4 5} This persistent airway inflammation may have implications for long-term lung health, raising concerns about the risks of progressive lung function decline and ongoing exacerbations, both of which could adversely impact quality of life and overall life expectancy.^{6 7}

Systemic immune profiles may reflect airway inflammation, as inflammatory mediators released and immune cells recruited to and from the airway are found within the peripheral circulation. Importantly, the systemic immune profile could also be influenced by inflammation within other organs affected by CF, such as the gastrointestinal and hepatobiliary systems.⁸ However, a better understanding of the relationship between systemic inflammation and CF lung disease would have obvious translational potential, aiding in the development of blood-based surrogate markers of airway inflammation and novel precision systemic therapies. Furthermore, chronic systemic inflammation in other conditions has been associated with increased risk of cardiovascular disease, cancer and neurodegenerative disorders, which can adversely affect long-term outcome.⁹ There are concerns that these complications may become more prevalent in an ageing CF population.

Systemic inflammation reduces with ETI, as evidenced by lower levels of circulating immune cells and inflammatory cytokines.¹⁰,¹⁴ However, there are inconsistencies between these studies about whether systemic inflammation fully resolves with ETI therapy. This study aimed to provide a detailed description of the effect of ETI therapy on systemic inflammation and to establish how these changes relate to improvements seen in lung function. Employing a multimodal approach, we profiled systemic inflammation, spanning both cellular and molecular dimensions.

Materials and methods

Study design

This previously described prospective observational single-centre study recruited CF subjects from a large UK specialist CF centre.⁵ CF participants possessed two Phe508del alleles, or one Phe508del and a cystic fibrosis transmembrane conductance regulator (CFTR) modulator unresponsive allele. Samples for analysis were collected from CF participants before (pre-ETI) and after ETI was commenced and ongoing at the time of follow-up (post-ETI). All participants provided written informed consent and were over 18 years, with a confirmed CF diagnosis based on genetic testing and/or sweat testing, along with typical phenotypic features, and deemed clinically stable by the medical team. Exclusion criteria included pregnancy and organ transplantation. The healthy control group comprised adult non-smokers with no significant medical history.

Sample collection

Blood samples were collected immediately prior to starting ETI and then repeated from the first clinical follow-up visit onward, typically occurring at 3 months, but up to 1 year postcommencing ETI. This extended window was wide due to the impact of COVID-19 restrictions in our clinic. Some clinical visits were conducted at home and were not suitable for obtaining samples due to potential delays in processing. Samples were collected, processed as per the individual requirement of each analytical technique and stored frozen at −80°C.

Analytical techniques

A brief overview of the analytical techniques is provided, with more detailed descriptions available in the online supplemental methods.

Peripheral immune cell profiling

Samples were stored with a Cytodelics whole blood cell stabiliser to prevent degradation/cell death during freezing, and subsequently thawed and analysed by flow cytometry.

Plasma proteomics

Samples were depleted of the top 14 most abundant proteins. Remaining proteins were digested using beads and pepsin and then analysed by tandem mass spectrometry.

Cytokine profiling

Plasma samples were analysed using the MSD U-PLEX Biomarker Group 1 (human) 71-Plex kit (Meso Scale Discovery, catalogue # K15081K-1) as per the manufacturer’s instructions.

Clinical data

All routine clinical data were collected, including spirometry which was performed either in clinic or at home using the Bluetooth Air Next Spirometer device (Nuvoair, Stockholm, Sweden).¹⁵

Statistical and bioinformatic analysis

All data were analysed using R (V.4.1.1). Differences between groups for flow cytometry and immunoassay datasets were tested using t-test or Mann-Whitney U test, depending on the data distribution. Benjamini-Hochberg was applied to the multiplex immunoassay to correct for false discovery. Statistical significance was determined by p value or adjusted p value, with threshold 0.05. Proteomics data analyses are detailed within the online supplemental methods. In brief, data were processed including steps filtering out proteins with more than 20% missing values, for normalisation via variance stabilisation normalisation,¹⁶ batch correction using a parametric empirical Bayes framework via Combat¹⁷ and imputation of remaining missing values using the package missForest.¹⁸ Differential expression (DE) analysis was performed using linear modelling empirical Bayes via the limma package.¹⁹ Differentially expressed proteins were identified using a Benjamini-Hochberg false discovery rate (FDR) threshold of 0.05. Protein Set Enrichment Analyses (PSEAs) were undertaken to evaluate pathway enrichment of expressed proteins between groups using the Fast Gene Set Enrichment Analysis package²⁰ on a variety of databases via clusterProfiler.²¹ All code used for proteomics analyses, including packages, functions and concrete parameters used, is available in the public GitHub repository https://github.com/CBFLivUni/preprocessing_plasma_SPRINT.

Results

Clinical and demographic characteristics

We obtained samples precommencing and postcommencing ETI in 57 CF subjects (online supplemental figure E1). Table 1 presents clinical and demographic characteristics for the CF subjects and the 29 healthy controls. Individual characteristics are shown within the online supplemental table E1. The two cohorts were well balanced for age and gender. Subjects receiving either dual modulators at baseline or who were treatment naïve were combined into one cohort, as there were few differences observed in our datasets when stratified by baseline modulator status (figure 1 and online supplemental Figure E2–E4). The mean time interval between baseline and for repeat sampling on ETI was 133 days (SD=67), with sampling times categorised as follows: 19 subjects sampled at <90 days, 25 at 90–179 days, 11 at 180–269 days and 22 at ≥270 days. The mean (SD) forced expiratory volume in 1 s (FEV₁%) increased between starting ETI and the time of repeat sampling (66.0% (18.2%) vs 77.9% (21.7%), p=0.002) (online supplemental Figure E2–E4). The mean time interval between baseline and for repeat sampling on ETI was 133 days (SD=67), with sampling times categorised as follows: 19 subjects sampled at <90 days, 25 at 90–179 days, 11 at 180–269 days and 2 at ≥270 days. The mean (SD) FEV₁% increased between starting ETI and the time of repeat sampling (66.0% (18.2%) vs 77.9% (21.7%), p=0.002).

Table 1. Subject demographic and clinical characteristics.

Demographic	Baseline ETI CF (n=57)	Healthy controls (n=29)
Age, mean (SD), years	30 (±8)	30 (±5)
Sex, male (n%)	36 (63)	15 (52)
Baseline FEV1%, mean (SD)	65.6 (±18.2)
Body mass index, mean (SD), kg/m2	19.6 (±8.7)
Genotype, n (%)
Phe508del-Phe508del	40 (70)
Phe508del-unresponsive	27 (30)
CF-related complications, n (%)
Exocrine pancreatic insufficiency	57 (100)
CF related diabetes mellitus	24 (42)
Cystic fibrosis conductance regulator (CFTR) modulation status, n (%)
Tezacaftor/ivacaftor	33 (58)
Lumacaftor/ivacaftor	4 (7)
None	20 (35)
Long-term medications, n (%)
Prednisolone*	6 (11)
Azithromycin	50 (88)
Nebulised antibiotic	45 (79)
Nebulised hypertonic saline	20 (35)
Dornase alfa	40 (70)
Sputum microbiology, n (%)
Pseudomonas aeruginosa	46 (76)
Burkholderia cepacia complex	6 (11)
Staphylococcus aureus	17 (30)
Other pathogen	6 (11)
Achromobacter xylosoxidans	3 (5)
Pandoraea apista	1 (2)
Stenotrophomonas maltophilia	2 (4)
Non-tuberculous mycobacteria	2 (4)

Data are presented as absolute number (percentage) or mean (SD).

^*For all subjects prednisolone was chronic therapy for allergic bronchopulmonary aspergillosis at doses 5–10 mg.

CF, cystic fibrosis; FEV1, forced expiratory volume in 1 s.

Figure 1. Immune profiles of pre-ETI CF subjects stratified by baseline modulator status. Absolute cell counts were measured by high-dimensional flow cytometry, with CF subjects stratified according to baseline modulator status (dual therapy vs modulator naïve). For each boxplot, individual cell count, together with median and IQRs, is plotted. Differences between cohorts are calculated by t-tests: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. CF, cystic fibrosis; ETI, elexacaftor/tezacaftor/ivacaftor.

The effect of ETI on immune cell populations

To determine immune cell profiles in healthy versus CF subjects, and the effect of ETI therapy on immune cells, we used high-dimensional flow cytometry to identify different innate and adaptive immune populations. We describe our antibodies and gating strategy within online supplemental table E2 and E3 and figure E5. Higher total numbers of immune cells (CD45+) were observed in CF patients pre-ETI compared with healthy controls, primarily due to increased levels of circulating neutrophils and monocytes (figure 2A,B). Further analyses indicated these monocyte differences were due to increased numbers of both the classical (CD14+) and intermediate (CD14+CD16+) subsets (figure 2C). There were also higher numbers of eosinophils and basophils, but not in the adaptive immune cell populations (online supplemental figure E6).

Figure 2. Quantitative profiling of circulating immune cells pre-ETI and post-ETI. Absolute cell counts were measured by high-dimensional flow cytometry. (A) Stacked bar chart displaying median values of selected circulating immune cells. (B) Box plots of the selected immune cells frequently implicated in CF lung disease pathology, with CF subjects stratified according to ETI status. (C) Box plots of the monocyte subsets. For each boxplot, individual cell count, together with median and IQRs, is plotted. Differences between cohorts are calculated by t-tests: *.

With ETI therapy, neutrophil levels reduced, leading to an overall reduction in the number of immune cells (figure 2A,B). However, ETI therapy did not lead to changes in other immune cell types (figure 2A–C, online supplemental figure E6). The post-ETI immune profile remained distinct from healthy controls, with persistently higher levels of neutrophils and monocytes (figure 2B). Together, these results indicate that ETI exposure does not normalise the systemic immune cell profile in CF patients during the period of follow-up for this study.

The effect of ETI therapy on neutrophil and monocyte phenotypes

Although levels of neutrophils and monocytes remained elevated post-ETI therapy, treatment could potentially affect the functionality of these immune cells (figure 3A–C). We observed that prior to ETI therapy, neutrophil populations in comparison to healthy controls showed reduced expression of CD15, a cell marker that reflects neutrophil maturation within the bone marrow (figure 3A).²² With ETI therapy, there was a shift towards a more mature phenotype with increased CD15 expression, as well as other markers related to neutrophil maturation, such as CD16 and CD101,²³ in keeping with a transition toward a less inflammatory phenotype. Post-ETI, CD15 expression was no longer significantly different from healthy controls, and similarly, either no significant difference or minimal differences were observed for most other markers between post-ETI and healthy controls. There was a significant increase in ETI in CD206 expression on neutrophils (figure 3B) and monocyte subsets (online supplemental figures E7 and E8), surpassing levels observed in healthy controls. Notably, pre-ETI CD206 expression in neutrophils and monocytes was comparable to healthy controls. CD206 expression was previously described on macrophages, and more latterly on neutrophils and monocytes, and is associated with immune resolution and wound repair, consistent with a more anti-inflammatory phenotype.²⁴,²⁷ There were no or minimal significant changes associated with the expression of other neutrophil cell markers between pre-ETI, post-ETI and healthy (figure 3C).

Figure 3. Immunophenotyping of neutrophils pre-ETI and post-ETI. Median fluorescence intensity (MFI) was measured by high-dimensional flow cytometry. MFI values are plotted, with CF subjects stratified according to ETI status. (A) Represents those markers that significantly changed with ETI and were related to neutrophil maturation. (B) Represents changes in the cell marker CD206. (C) These are the other cell markers that were tested. For each group, individual median cell marker MFI, together with median and IQR, is plotted. Difference between cohorts is calculated by Mann-Whitney U test: *p<0.05, **p<0.01, ***p<0.0001. CF, cystic fibrosis; ETI, elexacaftor/tezacaftor/ivacaftor.

The effect of ETI was also explored on monocyte phenotypes (online supplemental figures E7 and E8). Specifically, we focused on classical and intermediate monocyte subsets, as these were elevated pre-ETI therapy compared with healthy controls. We observed for both classical and intermediate subsets increased levels of CD101 and HLA-DR expression with ETI. There were no significant differences detected between healthy controls pretherapy or post-therapy; however, the levels, particularly of CD101, were higher post-ETI, raising the potential that there may be a difference which our current sample size did not detect. Monocytes expressing CD101 are associated with beneficial immunomodulatory effects, consistent with the immunophenotype shift related to CD206 expression.²⁸ In contrast, lower expression of HLA-DR on monocytes is associated with detrimental immunosuppression in conditions such as sepsis,²⁹ so the observed increased levels of HLA-DR suggest a positive effect on immune function.

For CXCR4 expression, levels were lower in classical monocytes but not intermediate monocytes among CF subjects. ETI therapy reduced CXCR4 in intermediate, with no significant effect on classical subtypes. Consequently, post-ETI CXCR4 expression remained significantly lower in both subsets compared with healthy controls. This reduced CXCR4 expression aligns with a more mature monocyte profile, potentially contributing to persisting inflammation. Finally, for CD11c and CCR7, minimal or no significant differences were observed between pre-ETI, post-ETI and healthy controls in both monocyte subsets.

The effect of ETI therapy on soluble immune mediator profiles

Cytokines and chemokines, collectively referred to as soluble immune mediators, influence the levels and functionality of circulating immune cells. We explored the effect of ETI therapy on the levels of 61 soluble mediators (online supplemental table E4) using multiplex immunoassays. There were 16 that differed between CF subjects and healthy controls (adjusted p<0.05), with 15 showing increased levels (figure 4A, online supplemental Figure E9). Some of these soluble mediators, such as interleukin 8 (IL8), IL6 and granulocyte-colony stimulating factor (GCSF), have well-established effects on the levels and immunophenotype of circulating neutrophils,³⁰,³² while others, such as macrophage CSF (M-CSF) and C-C motif chemokine ligand 2 (CCL2), primarily influence monocyte biology.³³

Figure 4. Quantitative profiling of systemic soluble immune mediators pre-ETI and post-ETI. (A) Volcano plot displaying pre-ETI CF versus healthy controls with log2 median fold change in soluble mediator abundance against adjusted p value calculated by Mann-Whitney U tests. Dots are coloured by thresholds based on log-fold change and adjusted p value. Horizontal dashed line indicates cut-off for adjusted p<0.05. Vertical dashed lines indicate positive and negative cut-offs for absolute log2 fold change of 1. (B) Volcano plot showing pre-ETI versus post-ETI. (C) Box plots of key altered soluble mediators. For those that were different at baseline between healthy and pre-ETI and then changed with therapy (adjusted p<0.05), the absolute concentrations are shown, with CF subjects stratified by ETI state. For each group, individual median cell marker MFI, together with median and IQRs, is plotted. Differences between cohorts are calculated by Mann-Whitney with adjustment for false discovery rate: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. CF, cystic fibrosis; ETI, elexacaftor/tezacaftor/ivacaftor; G-CSF, granulocyte-colony stimulating factor; MFI, median fluorescence intensity.

After ETI treatment, levels of six soluble mediators—IL6, IL8, G-CSF, IL3, C-X-C motif chemokine ligands 1 and 5 (CXCL1 and CXCL5)—significantly reduced, all of which are known to influence levels of circulating neutrophils by a range of mechanisms (figure 4B,C).^{34 35} There were no differences before commencing ETI in the levels of IL3, CXCL1 or CXCL5 in pwCF compared with healthy controls. Notably, while IL-6 levels reduced with ETI, they remained significantly elevated compared with healthy controls. In contrast, G-CSF and IL8 levels were very similar (online supplemental figure E9A). Of the other cytokines that did not display a significant change with ETI, eight remained significantly different from healthy controls, including M-CSF and CCL2 (online supplemental E10). The remaining five were no longer significant, suggesting an anti-inflammatory effect restoring towards normal levels (online supplemental figure E9B,C). These changes were not visible within principal component analysis (online supplemental Figure E11).

The effect of genotype and duration of therapy on the ETI response was explored across all datasets for those variables that changed with ETI therapy. This included variables from the immune cell counts, immune cell markers and soluble mediators. We did not identify any significant influence of genotype (online supplemental Table E5) or duration of ETI therapy on the changes observed in these variables (online supplemental table E6).

The effect of ETI on the plasma proteome

Next, we investigated differences in the plasma proteome, complementing the profiling of soluble mediators, as many cytokines or chemokines are too low in abundance to be detected by plasma proteomics. A total of 426 proteins were identified, meeting the criteria of at least two unique peptides and an FDR of less than 1%. Only four proteins were differentially expressed between CF and healthy controls (figure 5A and C), and no differential protein expression was observed with ETI therapy (figure 5B).

Figure 5. Quantitative profiling of cystic fibrosis (CF) plasma proteomes pre-ETI and post-ETI. (A) Volcano plot for pre-ETI CF versus healthy displaying fold change as calculated by limma against adjusted p values. Dots are coloured by thresholds based on log-fold change and adjusted p value. Horizontal dashed line indicates cut-off for adjusted p<0.05. Vertical dashed lines indicate positive and negative cut-offs for absolute log2 fold change of 1. (B) Volcano plot of pre-versus post-ETI CF. (C) Box plots of differentially expressed proteins. For those that were different at baseline between healthy and pre-ETI (adjusted p<0.05), the absolute concentrations are shown, with CF subjects stratified by ETI state. For each group, individual median cell marker MFI, together with median and IQRs, is plotted. Differences between cohorts are calculated by limma with adjustment for false discovery rate: *p<0.05. (D) The protein set enrichment analysis for CF versus healthy. X-axis indicates the Net Enrichment Scores (NES) of a pathway, with a negative score indicating that the pathway was downregulated in the CF group compared to the healthy group and a positive score indicating that the pathway was upregulated. The colour of the dot reflects p adjusted by Benjamini-Hochberg correction. The size of the dot represents the number of proteins present in the pathway that were also present in the data. (E) Protein set enrichment analysis of pre-ETI versus post-ETI therapy. ETI, elexacaftor/tezacaftor/ivacaftor; ITIH2, inter-alpha-trypsin inhibitor heavy chain H2; LRG1, leucine rich alpha-2-glycoprotein 1; PEX11B, peroxisomal membrane protein 11B; SPP2, secreted phosphoprotein 2.

To gain further functional insights, we undertook PSEA, a ranked overrepresentation method that identifies significant pathways/functions that differ between two conditions. In this case, we examined fold change between CF subjects and healthy controls, and the effect of ETI therapy (pre vs post). Unlike other enrichment approaches, PSEA uses the entire list of proteins rather than only those that have been identified a priori based on a statistical threshold. Results showed a significant overrepresentation of innate immune proteins, including those related to neutrophil degranulation in CF patients pre-ETI compared with healthy controls (figure 5D). The individual proteins responsible for these enrichment signals were identified (online supplemental table E7), and this included calprotectin (S100A8/S100A9) and C reactive protein (CRP). ETI therapy led to decreased expression of innate immune proteins, particularly within the complement pathway (figure 5E). The proteins responsible for these enrichment signals also included CRP and calprotectin (online supplemental table E8). These changes were not visible within principal component analysis (online supplemental figure E12).

The relationship between ETI-related changes in systemic inflammation and lung function

Finally, we investigated the relationship between changes with ETI therapy observed for systemic inflammation with change in absolute FEV1% predicted. Given the neutrophilic nature of the effects observed with ETI, we focused on this signature in the cell counts, cell markers and soluble mediators and then only on those variables that showed a significant change with therapy (table 2). There were only inverse correlations found between changes in two of the neutrophil cell markers—CXCR4 and CD16—and lung function. As would be expected if these changes were influencing lung function, CXCR4 expression significantly decreased with ETI therapy, whereas in contrast, CD16 expression increased with ETI. Furthermore, there were no differences between pre-ETI and healthy for either neutrophil cell marker.

Table 2. Relationship of systemic neutrophilic inflammation with lung function.

Dataset	Variable	Relationship with change in absolute FEV1%
		Spearman correlation	Unadjusted p value
Immune cells	Neutrophils	0.041	0.775
Neutrophil cell surface markers	CD101	0.056	0.713
Neutrophil cell surface markers	CXCR4**	−0.444	0.001
Neutrophil cell surface markers	CD206	−0.162	0.260
Neutrophil cell surface markers	CD15	−0.156	0.278
Neutrophil cell surface markers	CD16*	−0.311	0.028
Neutrophil cell surface markers	CD66b	−0.272	0.056
Cytokines	IL3	0.026	0.860
Cytokines	IL6	−0.144	0.322
Cytokines	IL8	−0.219	0.131
Cytokines	G-CSF	−0.132	0.368
Cytokines	CXCL5	0.252	0.081
Cytokines	CXCL1	0.009	0.951

^*p<0.05, **p<0.01.

FEV1, forced expiratory volume in 1 s; G-CSF, granulocyte-colony stimulating factor.

Discussion

In this study, we determined the effects of ETI on systemic inflammation in PwCF with at least one Phe508del allele. We observed prior to commencing ETI therapy that the immune profile was characterised predominantly by excessive neutrophilic inflammation. This excessive neutrophilic inflammation, while reducing with ETI, did not fully resolve during the follow-up period of this study. Circulating neutrophil levels decreased but remained elevated compared with healthy controls. In contrast, monocyte levels appeared unaffected and similarly remained elevated. Despite levels of immune cells remaining elevated in post-ETI CF subjects, there was a shift towards a less inflammatory immunophenotype for both neutrophils and monocytes. Changes were also observed in the soluble immune mediator profile and plasma proteome, with levels of six inflammatory immune mediators reduced, and downregulation of the innate immune and the complement protein sets. Finally, although we identified weak associations between the changes in the expression of two neutrophil cell markers and lung function, the pretherapy expression of these markers relative to healthy controls, and the changes associated with ETI therapy did not suggest a plausible direct influence on CF lung disease.

It is well established that PwCF prior to commencing ETI therapy have excessive systemic neutrophilic inflammation.^{10 11} Accordingly, we observed immune cell profiles in CF subjects in comparison to healthy controls that had elevated levels of circulating neutrophils and expression of cell markers consistent with a less mature, more inflammatory phenotype. In concordance with other studies, we found that ETI reduces neutrophil levels.^{10 36} In addition, and in slight contrast to previous work, we found that these levels do not fully normalise post-therapy. The most likely explanation of this difference is the definition of ‘normal’ in each of the studies. Other studies have used clinical reference ranges, a comparator cohort of non-CF bronchiectasis or a smaller number of healthy controls than that used in this study. Additionally, this discrepancy could be due to longer ETI exposure in other studies, but we found no significant correlation was observed between time on therapy and neutrophil counts. This finding of persisting inflammation is consistent with the findings of Casey et al who report that systemic inflammatory cytokines and innate immune proteins in PwCF on ETI resemble levels seen in non-CF bronchiectasis, remaining elevated rather than normalising.¹²

The neutrophil immunophenotypes revealed several novel findings about the influence of ETI therapy on systemic inflammation. We observed a shift towards a more mature neutrophil phenotype with therapy, characterised by increased expression of CD15, CD16 and CD101. While previous studies have not examined the effect of ETI on immune cell phenotypes, ivacaftor has been shown to promote a less active neutrophil phenotype in CF subjects with the G551D variant, though via examining different cell markers.³⁷ Additionally, we observed CD206 expression increased to levels significantly higher than those observed in healthy controls, consistent with a transition to an immune-resolving phenotype.^{26 27}

Although monocyte levels remain unchanged with ETI, notable phenotypic shifts were observed. Unlike neutrophils, changes in monocytes are less consistent across studies, although typically showing less dramatic reductions.^{10 11} Interestingly, monocytes also displayed an increase in CD206 expression, similar to neutrophils, suggesting an immune resolving phenotype. Previously, this increased CD206 expression has only been reported in CF for cultured and stimulated monocytes.³⁸ While immune resolving phenotyping has not been described with CFTR modulation, our observed shifts in markers, including CD101 and HLA-DR, are consistent with a less inflammatory monocyte phenotype that has already been observed with ETI exposure,³⁹ as well as in G551D and R117H monocytes exposed to ivacaftor.^{40 41} These prior studies have reported that CFTR modulation may directly influence proinflammatory tendencies, as well as improving monocyte function, potentially driving these observed phenotypic changes.³⁷⁴⁰,⁴²

The multimodal and multi-dimensional assessment of systemic inflammation undertaken here has allowed the soluble immune mediator profile and plasma proteome to also be characterised in pwCF, and then further interrogated for the effects of ETI therapy. We observed increased levels of soluble mediators in CF samples that have been previously implicated in CF lung disease pathogenesis, including IL6, IL8, G-GSF and IL1-beta.⁴³ However, there were also differences seen in other lesser-studied blood soluble mediators, such as those relating to Th2-mediated inflammation (Eotaxin and Eotaxin-2) and regulating circulating monocyte levels (M-CSF and CCL2). There were six soluble mediators that reduced with ETI, all known to influence circulating neutrophil levels. ETI has been shown to reduce the levels of IL6, IL8, CXCL1 and CXCL5, whereas IL3 and G-CSF have not been described previously.^{12 14} Notably, among the soluble mediators that remained persistently elevated after ETI therapy was CCL2, which is also a novel finding.

Our plasma proteomics revealed only four proteins that were differentially expressed in PwCF compared with healthy controls, with no significant changes observed after ETI therapy. Prior studies on the blood transcriptome with lumacaftor/ivacaftor, which provides less CFTR restoration than ETI, identified 36 differentially expressed genes.⁴⁴ This discrepancy may reflect that not all transcript changes translate into protein expression, and that transcriptomics, unlike proteomics, can detect low-abundance proteins such as cytokines. Despite minimal protein DE, PSEA revealed that innate immune proteins were overexpressed in PwCF, which includes within this protein set many that are categorised as related to neutrophil degranulation. Elevated levels of some innate immune proteins, such as calprotectin and CRP, have well-described correlations with inflammation and have been proposed as biomarkers, but their roles are not well understood. ETI therapy reduced the expression of innate immune proteins, including those involved in the complement system and fibrin formation. While complement activity and hypercoagulability have been implicated in CF pathophysiology,^{45 46} their response to ETI therapy has not been previously reported. Given their links to systemic inflammation, these pathways warrant further investigation.

The persisting inflammation seen after ETI therapy is likely multifactorial. The proinflammatory effects of CFTR deficiency on immune cells may not be fully mitigated by ETI, potentially compounded by the irreversible bronchiectasis, as indicated by the persisting lung function impairment after ETI in many of our subjects. This structural lung disease may explain the ongoing airway infection observed post-ETI with gram-negative organisms, in particular Pseudomonas aeruginosa,¹¹ further sustaining inflammation. Our immune profiling suggests that ETI exerts a stronger influence on neutrophils than on monocytes, as reflected in the immune cell and soluble mediator profiles. Those mediator profiles that influence neutrophil levels, such as G-CSF and IL8, decreased with ETI therapy, while CCL2, which drives monocyte levels, remained elevated. Persisting systemic inflammation, in particular the soluble mediators not fully addressed by ETI, may play a key role in maintaining abnormal monocyte levels and phenotypes.⁴⁷ One possible explanation for ongoing airway infection, which may drive persistently elevated levels of IL6.⁴⁸ Although IL6 levels reduced with ETI, they remained above healthy controls. IL6 has a broad range of proinflammatory functions, including promoting secretion of CCL2, which may contribute to ETI’s limited impact on circulating monocytes compared with neutrophils.

Persistent systemic inflammation may adversely affect long-term outcomes for PwCF. Chronic systemic inflammation has been associated with an increased risk of developing a range of complications including cardiovascular disease, malignancy and neurodegenerative conditions.⁹ The changes observed in immunophenotype with ETI might mitigate some potential neutrophil-related toxicity. CF neutrophils are a heterogeneous population of immune cells, with different subsets exhibiting distinct phenotypic and functional characteristics.⁴⁹ The shift observed towards a more mature phenotype would result in less active neutrophils with a shorter lifespan, while the immune-resolving phenotypes of neutrophils and monocytes could further reduce neutrophil activity. The increased CD206 expression seen with ETI on both neutrophils and monocytes is particularly intriguing. CD206 is a mannose receptor, typically associated with alternatively activated macrophages of M2-like phenotype.^{24 25} However, there is increasing evidence that CD206 is also an important marker of neutrophil and monocyte function,^{26 27} suggesting that an immune resolving state may still be ongoing when sampling was repeated and that it may fall with additional time on therapy. Further study is required to better understand the role of these immune cell phenotypes in both local and systemic inflammation in PwCF, as well as how they might impact long-term clinical outcome.

Finally, we did not identify a clear systemic inflammation signature associated with ETI-related improvements in absolute change in FEV1%. Although two markers, CD16 and CXCR4, showed significant inverse correlations with FEV1%, their baseline expression levels in CF relative to healthy controls, along with their changes following ETI therapy, do not suggest a direct relationship to lung function outcomes in CF. This failure to identify systemic markers that correlated with change in FEV₁% may reflect the restrictive approach, whereby we only evaluated neutrophil markers that changed significantly with ETI therapy to limit the number of comparisons given our relatively small sample size. In addition, not observing a strong correlation between systemic inflammation and lung function does not rule out that local airway inflammation plays a direct role in influencing lung function in CF but suggests that if this is the case, we are unable to identify this signature utilising systemic inflammatory markers. Therefore, the next crucial step is to gain a more comprehensive understanding of the relationship between systemic inflammation, airway inflammation and CF lung disease. This is an important goal, as identifying specific systemic immune mechanisms related to airway inflammation may produce novel therapeutic targets or potential surrogate measures of airway inflammation.

This study has encountered several notable limitations. This is a single-centre study involving a relatively small number of PwCF, and therefore, our findings need replicating in a broader population to confirm their generalisability. There was unavoidable variation in the time between initiating ETI and repeat sampling, caused by COVID-19-related restrictions on clinical contact for PwCF. The period of follow-up was also not sufficiently long to ensure that we observed all ETI-related resolution of inflammation. However, when tested, we found no influence of timing of sampling on our results. We acknowledge that, due to having only 57 subjects an effect of time on some variables could have been overlooked. We did not record compliance for other chronic medication which has been widely observed in clinical practice to reduce with ETI, which may influence inflammation and thus represent another potential confounder. Unfortunately, we were not able to assess airway inflammation or infection as most participants were unable to produce spontaneous sputum once exposed to ETI. In those that did produce sputum, many were unpaired with blood sampling. Further studies seeking to validate our finding in patients established on ETI therapy should ideally attempt simultaneous assessment of systemic and airway inflammation, potentially obtaining respiratory samples from sputum induction or bronchoscopy with lavage. Although plasma proteomics offered valuable insights, including highlighting the potential importance of complement activity, insufficient subject numbers hindered the full utilisation of this approach. Our findings suggest that much of the proteome is not influenced by CF, and there is significant overlap between disease and healthy protein levels. As such, future studies might be more successful identifying individual proteins with larger cohort sizes and by taking a targeted approach, for example, focussing on proteins related to innate immunity.

In conclusion, while ETI reduced excessive systemic inflammation in PwCF, it did not fully resolve. The most prominent improvements were within the levels and phenotypes of circulating neutrophils, as well as in the soluble immune mediators that regulate their activity. However, there was no clear relationship between ETI-related changes in systemic inflammation and acute changes in lung function. The long-term consequences of persisting systemic inflammation after ETI therapy need to be established for PwCF, in order to identify potentially adverse effects on organ systems and direct future therapeutic interventions.

The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the North West Lung Centre Charity or the Department of Health.

Data availability statement

Data are available on reasonable request.