In recent years, CFTR modulators have changed the paradigm of cystic fibrosis (CF). The latest is a triple combination of two correctors (elexacaftor and tezacaftor) and a potentiator (ivacaftor) for people with at least one F508del mutation [1]. This treatment has led to an unprecedented improvement, significantly enhancing lung function, nutritional status, and quality of life compared to previous modulators [2]. However, its impact on the gut and lung microbiota, which in CF show reduced richness and diversity compared to healthy controls [3,4], remains understood, despite altered microbiota being linked to systemic inflammation. Most available microbiota studies have focused on composition, even though functionality is equally important. Addressing this requires integrating metagenomics, metaproteomics, and metabolomics to obtain a more comprehensive view of host–microbe interactions.

Our main hypothesis is that individual clinical changes (ppFEV1, AST/ALT/GGT, calprotectin, elastase) are associated with targeted changes in microbiological taxa and functions. Therefore, we first described compositional and functional profiles, and second, whether clinical improvements in lung, hepatic, intestinal, and pancreatic functions after one year of ETI are associated with specific bacterial changes.

CFTR modulators have transformed CF management since ETI became available [7]. In this study, we assessed the impact of one year of ETI on gut and lung microbial communities and whether individual clinical improvements were associated with specific bacterial changes. ETI led to clear clinical improvement, particularly in lung function [8], with almost 20% of our patients shifting from moderate–severe to normal–mild stage. Gastrointestinal improvement was more modest, although we also observed signs compatible with recovery from pancreatic insufficiency to sufficiency in some patients [9,10]. Fecal elastase levels increased significantly, whereas calprotectin did not change. The absence of a significant change in fecal calprotectin may be explained by the clinically stable status of the cohort at baseline and by the fact that paired calprotectin values were not available for all patients, reducing statistical power for this analysis. Both adults (significantly) and children (trend) improved BMI, as previously reported [11,12]. Transaminases did not change significantly, although AST showed a tendency to increase, possibly attributed to therapy [13].

Eradication of the main CF-colonizing pathogens was also not observed after 12 months of ETI [14,15]. In our cohort, the P. aeruginosa load tended to decline, whereas S. aureus load tended to increase. Most patients whose S. aureus load increased tended to improve lung function. In mixed colonization, the reduction of P. aeruginosa may diminish ecological competition, facilitating the growth of S. aureus. Evidence suggests that S. aureus may play different roles depending on age and disease stage in CF: it has been associated with airway inflammation and reduced lung function in children [16], but in adults may indicate a less severe state, particularly when P. aeruginosa colonization decreases. Eradication of non-tuberculous Mycobacteria has been reported in a multicenter study [17], including M. abscessus. We observed this in two patients; however, another remained colonized. Although the main pathogens were not eliminated, other atypical microorganisms may increase. Despite ETI-related improvements in mucociliary clearance, residual bronchial structural damage may still impair airway defenses and facilitate colonization by upper respiratory tract microorganisms.

Despite the clinical improvements, overall gut microbiota diversity remained relatively stable. Alpha diversity did not change significantly, and beta-diversity shifts explained only a small fraction of variance. However, these findings reflect the 12-month post-treatment state and do not exclude earlier transient microbial changes after ETI initiation. Similar findings were reported by Gutiérrez-Díaz et al. [18], whereas Reasoner et al. [19] described increased alpha diversity in patients aged 6–18 years. We observed specific compositional changes, especially in feces, where enrichment of the butyrate-producing bacterium Anaerostipes hadrus was significant after one year of treatment. Butyrate-producing bacteria have been linked to reduced gut inflammation [20], and SCFAs, particularly butyrate, have beneficial local and systemic effects, especially on the respiratory tract [21,22]. These findings suggest that ETI may promote an increase in beneficial taxa. In sputum, compositional changes were more discrete and mainly involved commensal oropharyngeal species. The high proportion of host reads in sputum samples reflects the difficulty of obtaining high-quality sputum samples after ETI, when sputum production may decrease. This may also be consistent with the increased detection of oropharyngeal-associated taxa at T12. Previous studies have reported ETI-induced shifts in the respiratory microbiota, with increased evenness, patient-specific community signatures, and early increases in commensal taxa shortly after treatment initiation [23,24].

Functional metaproteomic analyses revealed more distinct differences than compositional data. After 12 months of ETI, the intestinal and airway environments appeared to shift from a pro-inflammatory to a more balanced state. In fecal samples, neutrophil-associated human proteins and RETN, a myeloid-derived proinflammatory cytokine [25], were significantly more abundant at baseline. After treatment, ITLN2 was the only human protein significantly more abundant; this Paneth cell–specific protein, reduced in inflammatory bowel disease [26], is consistent with a more regulated mucosal environment. In sputum, we observed a global downregulation of inflammatory and innate immune proteins after ETI. By contrast, extracellular matrix-associated proteins remained upregulated, which may be compatible with residual airway structural remodeling in the context of chronic CF lung disease. This structural remodeling may contribute to the persistence of pathogenic populations and the immigration of oropharyngeal organisms. Consistently, systemic inflammation is reduced with ETI, but the immune profile of these patients remains different from that of healthy controls [27].

The fecal microbial metaproteome increased in proteins associated with carbohydrate and lipid metabolism. Although fecal butyrate did not increase significantly, enzymes in butanoate metabolism, particularly 3-hydroxyacyl-CoA dehydrogenase (EC 1.1.1.35), were significantly elevated. These findings are consistent with the enrichment of butyrate-producing bacteria and indicate that a stable microbial community can shift its functional output depending on the surrounding physiological context. In sputum, no individual microbial protein reached significance, but grouping by enzyme activity revealed a significant increase in formaldehyde dehydrogenase (EC 1.2.1.46). Non-methylotrophic bacteria such as P. aeruginosa can oxidize formaldehyde directly to formate via glutathione-independent pathways [28], and the upregulation of this enzyme may reflect a shift toward more oxidative microbial metabolism, possibly driven by reduced mucus obstruction and better oxygen availability after treatment.

Several fecal taxa were associated with clinical status. Before treatment, A. muciniphila and B. pseudocatenulatum were enriched in individuals with better lung function, and these differences disappeared after one year of ETI. Regarding pancreatic function, V. atypica and V. rogosae were more abundant in individuals with lower elastase levels at baseline; V. atypica remained enriched in this group after 12 months. Although this genus has not been directly related to pancreatic insufficiency, it has been reported as enriched in the fecal microbiome of pancreatic cancer patients [29]. Transaminase levels did not change significantly overall, but associations with fecal taxa were species-specific: several Clostridium species were more abundant in individuals with lower ALT and AST after treatment, whereas C. symbiosum was enriched in patients with higher GGT.

In sputum, no specific taxa were correlated with improved lung function before or after treatment, which is expected, as the clinical effects are likely driven by community-wide rather than single-species changes. However, S. aureus was more enriched in patients with lower elastase and calprotectin and with higher AST and GGT at baseline, although all these associations disappeared after one year, despite an overall increase in S. aureus counts. A less inflamed airway environment and reduced competition with P. aeruginosa may allow S. aureus to persist without necessarily indicating worse clinical status. As in feces, associations between microbial taxa and transaminases were species-specific. Fusobacterium spp. were more abundant in patients with higher ALT and AST after treatment, but F. nucleatum was enriched in those with lower GGT. Similar genus- and species-specific patterns were observed for Veillonella and Prevotella. These context-dependent relationships suggest that anaerobes in CF act as adaptable members of a dynamic microbial network rather than being strictly beneficial or harmful. Overall, these associations should be considered exploratory and intended to identify potential clinical–microbiome relationships rather than causal effects.

As limitations, the unadjusted T0–T12 compositional and functional analyses are descriptive and do not identify adjusted causal effects. These analyses were intended to capture broad patterns over one year of ETI, rather than age-stratified responses, which were not assessed because further subgrouping would have reduced statistical power in this moderate-sized longitudinal cohort. Clinical–microbiome association models were then used to examine individual clinical variables, with age and sex included as covariates. However, we did not adjust for prior CFTR modulator therapy or CFTR genotype, as Reasoner et al. [19] reported similar community composition in previously treated and modulator-naïve individuals, and Duong et al. [30] found genotype-related differences explaining only ∼2% of the variance. Similarly, antibiotic exposure was not included as an adjustment variable because treatment patterns changed during follow-up and were highly unbalanced at T12, particularly for azithromycin. Such additional adjustment or stratification in our cohort would have further reduced statistical power due to our limited sample size. Moreover, some differentially abundant taxa, particularly in feces, were of low relative abundance and should be interpreted cautiously. Our cohort consisted of clinically stable patients at baseline, which may have limited the magnitude of microbiota or inflammatory changes observed. Other systemic inflammatory markers, such as CRP, which may decrease after ETI initiation, and cardiometabolic risk variables were not incorporated because they were not systematically available at paired T0 and T12 timepoints. Additionally, sputum quality, especially at 12 months, may have been suboptimal and more representative of upper than lower respiratory airway microbiota. Finally, the absence of intermediate paired fecal and sputum multi-omics samples prevented the assessment of early transient microbial shifts, including possible short-term changes in alpha- or beta-diversity that may have stabilized before T12. A longer follow-up is also needed to determine whether the observed changes are sustained over time.

In summary, this longitudinal multi-omics study shows that one year of ETI therapy produces substantial clinical improvement in people with CF, accompanied by targeted microbiome and metaproteome functional shifts rather than global community restructuring. Specific gut taxonomic and functional adaptations appear linked to butyrate-producing bacteria, butyrate-related pathways, and downregulation of inflammatory human proteins in both compartments. Following respiratory improvement, oxidative metabolism gains prevalence over fermentation in the lungs. Species-specific roles of anaerobic genera must be considered, although further studies are needed to determine causality. Despite persistent pathogenic colonization, probably due to permanent fibrotic damage, our results support that CFTR modulation delivers its systemic benefits not by resetting the microbiota but by functionally reprogramming the host–microbe ecosystem toward a less harmful state.

N.B.P. conceptualization and writing original draft; N.B.P., C.G.D., E.P.G., A.M.B. experimental processing and writing; N.B.P., D.B., and J.T. metagenomic analysis; N.B.P. and C.G.D. metaproteomic analysis; J.T. and A.C.S. supervision in metagenomic analysis, writing – review, and editing; M.G., L.M., and S.V.S. providing resources and validation; A.O., J.O.I., R.C. and C.G. writing – review and editing; J.D.D.C., M.G., and R.d.C. conceptualization, writing – review and editing; A.O., J.O.I., R.d.C., S.V.S., and J.D.D.C. funding acquisition.

The Hospital Ramón y Cajal Ethics Committee approved in 2020 (Ref. number 328/2020). The research was conducted ethically, in accordance with the World Medical Association Declaration of Helsinki. All participants and their parents/legal guardians signed a written informed consent form.

Artificial intelligence has not been used in the design of the article or in the creation of images, but the manuscript has been grammatically and orthographically corrected using the deepL website (https://www.deepl.com/es/translator).

This work was supported by the Instituto de Salud Carlos III and the European Union – NextGenerationEU (PI22/01045 to J.D.D.C. and S.V.S., and PI19/01043 to R.C. N.B.P. was supported by a research contract from Instituto de Salud Carlos III (Rio Hortega program, grant number CM23/00037). This research was also supported by Personalized and precision medicine grant from the Instituto de Salud Carlos III (MePRAM Project, PMP22/00092), Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación, funded by NextGenerationEU funds from the European Union that finance the actions of The Resilience and Recovery Facility.

L.M., S.V.S., and R.d.C. have received funding from Vertex S.L. The rest of the authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.