You are here

Article Text

Article menu
Chronic obstructive pulmonary disease
Original research
Lower airway microbiota in COPD and healthy controls
  1. Solveig Tangedal1,2,
  2. Rune Nielsen1,2,
  3. Marianne Aanerud1,2,
  4. Christine Drengenes2,
  5. Gunnar R Husebø1,
  6. Sverre Lehmann1,2,
  7. Kristel S Knudsen1,
  8. Pieter S Hiemstra3,
  9. Tomas ML Eagan1,2
  1. 1 Department of Thoracic Medicine, Haukeland University Hospital, Bergen, Norway
  2. 2 Department of Clinical Science, Faculty of Medicine, University of Bergen, Bergen, Norway
  3. 3 Department of Pulmonology, Leiden University Medical Center, Leiden, The Netherlands
  1. Correspondence to Solveig Tangedal, Department of Thoracic Medicine, Haukeland University Hospital, Bergen 5021, Norway; stangedal{at}gmail.com

Abstract

Background The lower airway microbiota in patients with chronic obstructive pulmonary disease (COPD) are likely altered compared with the microbiota in healthy individuals. Information on how the microbiota is affected by smoking, use of inhaled corticosteroids (ICS) and COPD severity is still scarce.

Methods In the MicroCOPD Study, participant characteristics were obtained through standardised questionnaires and clinical measurements at a single centre from 2012 to 2015. Protected bronchoalveolar lavage samples from 97 patients with COPD and 97 controls were paired-end sequenced with the Illumina MiSeq System. Data were analysed in QIIME 2 and R.

Results Alpha-diversity was lower in patients with COPD than controls (Pielou evenness: COPD=0.76, control=0.80, p=0.004; Shannon entropy: COPD=3.98, control=4.34, p=0.01). Beta-diversity differed with smoking only in the COPD cohort (weighted UniFrac: permutational analysis of variance R2=0.04, p=0.03). Nine genera were differentially abundant between COPD and controls. Genera enriched in COPD belonged to the Firmicutes phylum. Pack years were linked to differential abundance of taxa in controls only (ANCOM-BC (Analysis of Compositions of Microbiomes with Bias Correction) log-fold difference/q-values: Haemophilus −0.05/0.048; Lachnoanaerobaculum −0.04/0.03). Oribacterium was absent in smoking patients with COPD compared with non-smoking patients (ANCOM-BC log-fold difference/q-values: −1.46/0.03). We found no associations between the microbiota and COPD severity or ICS.

Conclusion The lower airway microbiota is equal in richness in patients with COPD to controls, but less even. Genera from the Firmicutes phylum thrive particularly in COPD airways. Smoking has different effects on diversity and taxonomic abundance in patients with COPD compared with controls. COPD severity and ICS use were not linked to the lower airway microbiota.

  • COPD Pathology
  • Respiratory Infection
  • COPD epidemiology

Data availability statement

Data are available in a public, open access repository. Data are available from the time of publication and without end date at DRYAD depository: https://doi.org/10.5061/dryad.rfj6q57ff. Submitted data include de-identified participant data (metadata) in .xlsx and .txt format, ASV and representative sequences tables, phylogenetic tree (rooted), and taxonomy in .qza format from QIIME 2, QIIME 2 code and R code. The data have been generated by ST and TME. The study protocol is published and linked to in the manuscript.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/thorax-2023-220455

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Many smokers do not develop chronic obstructive pulmonary disease (COPD), and other causative factors must exist. One potential factor is the lower airway microbiota. Inhaled corticosteroids (ICS) are both known to prevent COPD exacerbation and increase the risk of airway infections in patients with COPD. A dysbiosis of the lower airway microbiota may impact the risk of lower airway infections in COPD.

    WHAT THIS STUDY ADDS

    • This study describes a loss of evenness in the airway microbiota in bronchoalveolar lavage (BAL) in patients with COPD, with genera belonging to the Firmicutes phylum proliferating compared with healthy controls. Additionally, this study suggests that ICS do not disrupt the microbiota in BAL in patients with stable COPD.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Future studies should explore the functionality of the microbiota, both in health and disease, especially in smokers and non-smokers. ICS may not be adversely affecting the microbiota, but this needs confirmation in longitudinal studies.

    Introduction

    Chronic obstructive pulmonary disease (COPD) is an inflammatory disease of the lower airways. COPD often alternates between a stable state of chronic airflow obstruction and episodic exacerbation. Smoking is the main cause of COPD and a known irritant that affects innate and adaptive mucosal immunity in the lower airways. However, not all smokers develop COPD.1 2 Smoking has also been associated with an increased risk of pneumonia and with an altered lower airway bacterial community (microbiota).3 4 The lower airway microbiota itself has been suggested to contribute to the development and progression of several lung diseases, including COPD.5–7 Even before studies began characterising the lower airway microbiota using gene sequencing tools, it was acknowledged that acute exacerbation of COPD could be triggered by lower airway infections.8

    Treatment for COPD may include inhaled corticosteroids (ICS) with anti-inflammatory properties. ICS is now recommended for patients with frequent and severe exacerbation, and those with elevated blood eosinophils and known concurrent asthma.9 Use of ICS in patients with COPD does however come with an increased risk of pneumonia, a risk of overgrowth of Streptococcus and Haemophilus spp, and possibly an altered lower airway microbiota.10–12 Thus, investigations of the lower airway microbiota in patients with COPD and healthy controls, across smoking habits, ICS use and exacerbation history, can be important to best tailor future therapeutic strategies.

    The lower airway microbiota in COPD has been studied in different biological samples.13–15 Contamination confounds the use of sputum, as the samples pass unprotected through the high biomass oral cavity.16 17 Bronchoalveolar lavage (BAL) is less prone to contamination,17 and the sampling site is directly verified. Although a fairly safe procedure,18 bronchoscopy is uncomfortable and resource-intensive compared with sampling of sputum. Previous studies using BAL have tended to include small numbers of participants and controls.3 15 19–23 The largest study previously published is a substudy from the SPIROMICS cohort with 78 patients with COPD and 103 controls.13 The Bergen COPD Microbiome Study (MicroCOPD) is a single-centre study, where 100 patients with COPD and 100 controls were sampled with equipment designed for protected BAL sampling, reducing the risk of oropharyngeal contamination to a minimum.17

    With this material at hand, we looked for differences in the bacterial composition between patients with COPD and healthy controls, and for patients with COPD, we investigated the effect of disease severity, use of ICS and exacerbation history.

    Methods

    Study population

    The MicroCOPD Study protocol has been published.24 From the 249 participants enrolled at the Department of Thoracic Medicine, Haukeland University Hospital, Bergen, Norway in 2012–2015, the current study included all 97 patients with COPD and all 97 healthy controls with valid protected BAL samples (online supplemental figure 1). All bronchoscopies were performed without any other indication than study sampling. Patients with COPD were examined in a stable state defined as not having ongoing respiratory symptom exacerbation, and no use of antibiotics or oral corticosteroids within the last 2 weeks before inclusion. Subjects who met the eligibility criteria for the study but had recently used antibiotics or oral corticosteroids were requested to return to the outpatient clinic at a later date. Inclusion, collection of baseline data and the bronchoscopy procedure were scheduled within the same day.

    Supplemental material

    The structured interviews performed included information on smoking habits, medication and exacerbation history. The COPD diagnosis was verified using clinical evaluation and standardised spirometry with a forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) cut-off at 0.7 according to international guidelines.9 The Global Initiative for Chronic Obstructive Lung Disease (GOLD) staging system was used to classify disease severity.9 COPD I/II had post-bronchodilator 80%>FEV1>50% of expected, and COPD III/IV had FEV1 <50% of expected. Controls were enrolled if no known lung diseases and no significant airflow obstruction were present. To be classified as a non-smoker, participants had to be smoke-free >1 year before inclusion. A total of 22 controls had FEV1/FVC <0.7 indicating obstructive airways, but with no other objective or subjective findings supporting a COPD diagnosis. After initial inclusion of all participants, three experienced pulmonologists at our department reviewed inclusion according to available spirometry, CT scans and medical history. One person included in the COPD cohort was recategorised as a healthy control, whereas five healthy participants were reclassified as patients with COPD.

    Factors contradicting bronchoscopy and thus inclusion included oxygen saturation <90% with oxygen supplementation, partial pressure of carbon dioxide in arterial blood >6.65 kPa, known increased bleeding risk, a known allergy towards the premedication, cardiac risk and use of antibiotics within 2 weeks prior to bronchoscopy. If measures of FEV1 were both <30% predicted and <1 L, BAL sampling was not performed.24

    Sample collection

    Bronchoscopy was performed through oral access with the patient in the supine position. Protected BAL was collected from all participants with FEV1 >30% predicted and five patients with FEV1 <30% predicted, but >1 L. Sterile phosphate-buffered saline (PBS) was instilled in the right middle lobe through a sterile catheter (Combicath, Prodimed) with a sealed wax tip, followed by aspiration through the same catheter. PBS used for bronchoscopic sampling served as a negative control for the microbiota analysis. Further details on the bronchoscopic sampling procedure are published25 and are available in the online supplemental file.

    DNA extraction and target gene sequencing

    A detailed protocol for laboratory processing has been published.26 Bacterial DNA was extracted using both enzymatic and mechanical lysis methods. Samples were treated with the enzymes lysozyme, mutanolysin and lysostaphin (Sigma-Aldrich) as recommended.27 Following DNA extraction, processing with the FastPrep-24 instrument with reagents from the FastDNA Spin Kit (MP Biomedicals, Solon, Ohio, USA) was performed. Amplicon PCR (45 cycles) and index PCR were run using primers from the Nextera XT Index Kit (Illumina, San Diego, California, USA). Finally, paired-end sequencing (2×300 base pairs) of the V3–V4 region of the 16S rRNA gene as instructed by the protocol for Metagenomic Sequencing Library Preparation for the Illumina MiSeq System was performed (Part #15044223 Rev. B, MiSeq Reagent Kit V.3).

    Statistical analyses

    Bioinformatic terms and handling of the sequencing output are described in the online supplemental file. Briefly, the bioinformatic pipeline QIIME 2 V.2022.1128 and the statistical computing environment R29 were used. The sequencing output underwent quality control, sorting of sequences in amplicon sequence variants (ASVs) and removal of contaminants. Diversity analyses were preceded by normalising all samples at 2200 reads/sample, determined from rarefaction curves (online supplemental figure 2). To capture the complexity of the data, different indexes were calculated. For alpha-diversity: observed ASVs (richness), Pielou evenness, Shannon entropy and Faith’s phylogenetic diversity (PD) (phylogenetic). For beta-diversity: Sorensen and Bray-Curtis (non-phylogenetic), weighted and unweighted UniFrac (phylogenetic) and Aitchison (compositional) distance matrices. To identify differentially abundant taxa, we used Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC) in R.30 Only ASVs found in >10% of the samples were considered.31 The handling of zero counts is described in the online supplemental file.30 32

    Data generated by ST and TMLE for this manuscript including sequencing data, relevant QIIME 2 and R code, and the metadata are available at the DRYAD data repository: https://doi.org/10.5061/dryad.rfj6q57ff.33

    Results

    Participant characteristics analysed for differences between current and non/ex-smokers are presented in table 1. Non-smoking patients with COPD were significantly older than smoking controls (Dunn’s test p=0.02). Patients with COPD had smoked more pack years than controls. In the COPD cohort, the use of ICS was more common among non-smokers than smokers. Current smoking patients with COPD had more frequent exacerbation episodes (table 1).

    View this table:
    Table 1

    Participant characteristics

    In online supplemental table 1, the overall significant differences found using Kruskal-Wallis test were supplied with Dunn’s test to show which pairs of participant groups from table 1 were significantly different.

    Sequencing results

    In BAL from patients with COPD, 1.79 million sequences were assigned to 399 ASVs. Sixty-two ASVs were unique to the COPD cohort. BAL from all 97 controls contained 1.58 million sequences assigned to 366 ASVs, of which 29 were not seen in patients with COPD (online supplemental figure 1).

    Diversity

    Alpha-diversity was lower in COPD compared with controls when considering Pielou evenness and Shannon entropy. Observed ASVs (richness) and Faith’s PD were not associated with disease (figure 1).

    Figure 1
    Figure 1

    Alpha-diversity in bronchoalveolar lavage samples from participants with and without chronic obstructive pulmonary disease (COPD) showing significantly different evenness between groups after logistic regression analyses adjusting for sex, age and smoking (smoking vs non-smoking within the last year of inclusion). (A) Counts of different amplicon sequence variants (ASVs), that is, richness; (B) Pielou evenness; (C) Shannon entropy and (D) Faith’s phylogenetic diversity (PD).

    Alpha-diversity was neither associated with FEV1 in per cent predicted or GOLD stage as a dichotomous variable, nor the use of ICS in the COPD cohort. Lower richness and PD in BAL were found in those with frequent exacerbation (observed ASVs: OR 0.95, CI 0.92 to 0.99, p=0.02; Faith’s PD: OR 0.45, CI 0.23 to 0.88, p=0.02). No significant associations remained after adjusting for disease severity, sex, age and smoking (online supplemental tables 2–4).

    Beta-diversity did not differ significantly between patients with COPD and controls, or across sex and age. Among patients with COPD, significant differences were seen between smokers and non-smokers, but only for the phylogenetic and abundance-aware weighted UniFrac matrix (permutational analysis of variance: effect size R2=4%, p=0.01).

    Taxonomy

    The relative abundances of phyla and genera in each sample, categorised by smoking status in participants with and without COPD, are presented in figure 2A,B. Taxa with an average relative abundance of less than 1% among all participants were categorised as low abundant and grouped as ‘others’.

    Figure 2
    Figure 2

    Relative abundances of (A) phyla and (B) genera in bronchoalveolar lavage from participants with and without chronic obstructive pulmonary disease (COPD) stratified by smoking. Non-smoking was defined by smoking cessation >1 year prior to inclusion. Others: taxa with mean relative abundance <1% of the total sequencing data.

    Online supplemental figure 3 shows differentially enriched ASVs in both groups, annotated with the highest resolution taxonomic information achieved from a self-trained Naive Bayes classifier and the Silva database V.138.1. Differential abundance was further tested with ANCOM-BC comparing patients with COPD with controls and comparing the two groups of participants separately. We investigated the effect of sex, age, smoking, and for the COPD cohort, COPD severity, ICS and exacerbation frequency. The results are summarised in table 2.

    View this table:
    Table 2

    Differentially abundant* phyla and genera in bronchoalveolar lavage samples from patients with COPD and controls

    Comparing the taxa in all BAL from patients with COPD against all BAL from controls, Firmicutes and nine genera were significantly differentially abundant (table 2 and figure 3A). The three genera enriched in patients with COPD all belonged to the Firmicutes phylum. Solobacterium, Prevotella, Prevotella_7 and Alloprevotella representing the Bacteroidota phylum were enriched in controls. The largest log-fold differences were found for Leptotrichia belonging to the Fusobacteriota phylum, and TM7x belonging to the Patescibacteria phylum. Figure 3B shows the mean relative abundances of the nine genera in controls, patients with mild/moderate COPD (COPD I/II) and patients with severe/very severe COPD (COPD III/IV). Granulicatella and Gemella both had their highest mean relative abundances in BAL from patients with mild/moderate COPD.

    Figure 3
    Figure 3

    Comparing abundances of genera across COPD and controls. (A) Analysis of Compositions of Microbiomes with Bias Correction found nine genera significantly differentially abundant comparing BAL from participants with and without COPD (positive log-fold difference=more abundant in COPD, q values for each taxon in plot: significance level q<0.05 after Holm correction). (B) Mean relative frequencies of the significantly differentially abundant genera in BAL from controls and patients with COPD with mild/moderate (GOLD stage I/II) and severe/very severe disease (GOLD stage III/IV). The right panel with cropped y-axis is not displaying Streptococcus to emphasise the remaining eight. BAL, bronchoalveolar lavage; COPD, chronic obstructive pulmonary disease; GOLD, Global Initiative for Chronic Obstructive Lung Disease.

    For the characteristics sex, age and pack years, we found no differentially abundant taxa when including all 194 participants (table 2). While Alloscardovia was absent in female patients with COPD, it was found in female controls. The Bacteroidota phylum decreased with age in the control group only, but no genera belonging to the Bacteroidota phylum were found to be differentially abundant with age. Three low-abundant genera were more abundant in smokers than in non-smokers. Of these, Filifactor was the most frequently seen with the presence in only 34 of 194 samples. Oribacterium was absent in smoking patients with COPD, whereas no significant difference was linked to smoking in the control group. Pack years were associated with differentially abundant genera only in controls. With increasing smoking history in the control group, Lachnoanaerobaculum and Haemophilus both diminished. Lautropia was never seen in BAL from smoking controls, but Lautropia was also present in only 22 of 194 samples. In the COPD cohort, disease severity and ICS were not associated with the differential abundance of any taxa. In those with frequent exacerbation, both Catonella and Actinobacillus were absent. These genera were observed in 65 and 22 samples, respectively. Leptotrichia had lower abundances in those with frequent exacerbation (table 2).

    Discussion

    This is to our knowledge the largest bronchoscopy study to date comparing the lower airway microbiota in BAL between patients with COPD and controls, and the first using protected BAL with a sterile inner catheter inserted in the working channel of the bronchoscope. We confirmed a loss in alpha-diversity in COPD compared with controls and found it linked to lesser evenness rather than a loss of richness. We identified nine genera that differed between patients with COPD and controls, of which Streptococcus is an important potential pathogenic taxon. Smoking quantified by pack years was associated with a significant reduction in Haemophilus and Lachnoanaerobaculum in healthy controls. Such associations within the COPD cohort were not found. Neither diversity nor taxonomic abundances differed in BAL fluid with the use of ICS or with increased COPD severity.

    Alpha-diversity has previously been reported to be decreased in airway samples from patients with COPD compared with controls, and to be further lowered with disease progression13 14 20 and during COPD exacerbation.23 In our study, we demonstrate that while evenness was reduced, both richness and PD were sustained in the lower airways of patients with COPD compared with healthy controls. The lower airway microbiota is proposed to be constantly modulated by the addition of bacteria through inhalation and micro-aspiration, local replication and survival, and removal through mucociliary clearance, cough, and immune responses.5 34 It is conceivable that the decrease in evenness but maintenance of richness and PD can be attributed to a continuous influx of upper airway microbiota, with certain taxa thriving more in the inflamed airways, thus diminishing evenness.

    The lack of difference in beta-diversity between BAL from patients with COPD and controls agrees with previous studies.13 35 COPD is a progressive and heterogeneous disease, with a gliding transition from a healthy state. Differences can be diluted due to this characteristic, and further due to an insufficient number of participants. The beta-diversity analyses comparing smoking and non-smoking patients with COPD showed that differences in abundance and relatedness of ASVs were not independently sufficient to separate between smokers and non-smokers. Only weighted UniFrac taking both aspects into account identified a significant difference. Meanwhile, smoking could only explain 4% of the separation of the samples, indicating a relatively small effect of smoking on beta-diversity.

    As expected from the alpha-diversity results, the current study identified several differentially abundant taxa comparing patients with COPD and controls. We observed an increase in Granulicatella, Gemella and Streptococcus in COPD, and a successive increase in Streptococcus with increased disease severity. This confirms the results from Opron et al and Wang et al in BAL and sputum, respectively.13 14 Gemella, Streptococcus, Prevotella and Leptotrichia were among several genera found to be increased in COPD compared with controls in a study by Ramsheh et al using bronchial brush samples.35 Whether the lower airway in COPD has an impaired inflammatory profile in which Firmicutes are allowed to multiply at the cost of other taxa should be investigated further.

    The remodelling in COPD airways includes different degrees of emphysema, an increased bronchial mucus production, impairment of the mechanisms necessary to remove mucus, remodelling of the airway wall and altered inflammatory activity.36 37 This can create an environment in which certain bacteria thrive. One can also imagine that overgrowth of pathogenic bacteria could contribute to the remodelling seen in COPD airways. Thus, one can imagine a scenario where smokers with an adverse composition of the lower airway microbiota would be more likely to develop disease than other smokers. Madapoosi et al examined matched BAL microbiome and metabolomic data in smokers with no known COPD, and with mild and moderate COPD, and suggested that elements from both the lung microbiome and metabolome may jointly influence the pathophysiological mechanisms already in milder stage COPD.38

    Sex and age were not strongly associated with the lower airway microbiota in our study. For smoking, consistent changes in the microbiota across the COPD cohort and controls were not found, in line with findings from Opron et al.13 In a study by Pfeiffer et al specifically investigating the effect of smoking on the airway microbiota in 33 healthy smokers and 13 healthy never-smokers, a non-significant negative correlation between pack years and relative abundances of Haemophilus in BAL samples was found. The only significant differences in mean relative abundances of genera were found for Campylobacter and Neisseria.3 Although COPD is a multifactorial disease, the main risk factor for disease development is undoubtedly inhalation of harmful substances, of which tobacco smoking is the most significant. The effect of smoking on the lower airway microbiota in healthy individuals is thus an urgent research question but remains understudied.

    In the GOLD strategy documents, ICS is recommended for a carefully selected subpopulation of patients with COPD. The reason is the known increase in the risk of pneumonia and infectious exacerbation.9 Keir et al conclude their review of ICS and the lung microbiota by stressing that ICS use should be founded on leucocyte endotypes to avoid prescription to patients at a higher risk of Streptococcus and Haemophilus overgrowth.11 During COPD exacerbation, two longitudinal sputum studies have shown that alpha-diversity decreased when patients were treated with oral corticosteroids.39 40 Whether ICS use impacts the lower airway microbiota in COPD was specifically addressed in a randomised controlled trial (the DISARM Study) from Vancouver, Canada.12 41 In that study, 56 patients with COPD were sampled with unprotected cytology brushes before and 12 weeks after use or non-use of ICS.12 Lower richness and Shannon entropy were observed in the ICS-treated. However, the intraindividual differences in alpha-diversity between sampling time points resemble what we observed in a study on the microbiota in induced sputum samples collected across stable state and exacerbation in a COPD cohort.42 Whether these intraindividual differences are linked to ICS, or within what to be expected in repeated samples is arguably yet unknown. Leitao Filho et al found no links between ICS and Streptococcus, but Haemophilus was reduced after 12 weeks of treatment with fluticasone.12 It is notable that the host transcriptome was found to be affected by fluticasone, but not in the non-ICS study arm.41

    In our study, 60 of 97 patients with COPD used ICS. In stable COPD, we did not find evidence of a significant influence of ICS on the lower airway microbiota, including the abundances of Streptococcus and Haemophilus. These genera were also not linked to ICS in the two bronchoscopy studies by Opron et al and Ramsheh et al.13 35

    Combined, there is yet little current evidence of a link between ICS treatment and unfavourable changes in the lower airway microbiota in patients with COPD.

    Frequent exacerbation in COPD was linked to reduced abundance of Leptotrichia, a genus belonging to the Fusobacteriota phylum. A publication focusing on its pathogenic potential concludes that it rarely causes severe infections, and mainly in immunosuppressed patients. The Fusobacteriota phylum is commonly found as a commensal bacterium in the oral cavity,43 and its role in those with frequent exacerbation is yet elusive.

    Despite being one of the most extensive bronchoscopy studies on the lower airway microbiota in COPD, our study has several limitations. First, the cross-sectional design does not permit causal inferences. Future longitudinal and mechanistic studies are needed to determine whether the differences in microbiota between patients with COPD and controls are a cause or a result of disease development and progression. Additionally, the sequencing of target genes like the 16S rRNA only provides information at the genus level, limiting species-level evaluation.44 Second, our study relied on self-reported smoking habits, which could have been improved by using biomarkers such as cotinine. Third, we sampled BAL from the right middle lobe in all participants ensuring standardisation. However, this precludes analysing geographical variation within the airway microbiota, which is likely to exist.15 20 Ventilation-to-perfusion ratio differs from the upper to lower parts of the lungs, and some airways may receive more micro-aspiration than others. Whether this potential variation is differentially impacted by ICS use is unknown. Fourth, a 2-week wash-out of antibiotics before inclusion may have been too narrow. All medication use was recorded, and out of the 194 study participants, 89 patients with COPD and 90 healthy controls had not used any antibiotics the previous 12 weeks to inclusion. Finally, the control group was not age or smoking matched to the patients with COPD.

    Conclusion

    Differences in the lower airway microbiota related to COPD are evident. The difference lies in an altered microbial balance in which bacteria from the Firmicutes phyla appear favoured. This could not be linked to smoking or to the use of ICS in patients with COPD. Whether differences in the lower airway microbiota observed in patients with COPD compared with healthy controls are pathogenic in COPD or are the results of COPD must be further investigated in longitudinal studies.

    Data availability statement

    Data are available in a public, open access repository. Data are available from the time of publication and without end date at DRYAD depository: https://doi.org/10.5061/dryad.rfj6q57ff. Submitted data include de-identified participant data (metadata) in .xlsx and .txt format, ASV and representative sequences tables, phylogenetic tree (rooted), and taxonomy in .qza format from QIIME 2, QIIME 2 code and R code. The data have been generated by ST and TME. The study protocol is published and linked to in the manuscript.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and was approved by the regional ethical committee (REK Vest case number 2011/1307). All participants provided oral and written consent.

    Acknowledgments

    We thank Eli Nordeide, Lise Monsen, Hildegunn Fleten, Ingvild Haaland, Harald G Wiker, Tuyen Hoang, Randi Sandvik, Tharmini Kalananthan, Elise O Leiten, Einar Marius Hjellestad, Ane Aamli, Øistein Svanes and Per Bakke (Department of Thoracic Medicine, Haukeland University Hospital, Bergen, Norway; and Department of Clinical Science, Faculty of Medicine, University of Bergen, Norway), and all the participants making this study possible.

    References

    1. ↵
      2. Løkke A ,
      3. Lange P ,
      4. Scharling H , et al
      . Developing COPD: a 25 year follow up study of the general population. Thorax 2006;61:935–9. doi:10.1136/thx.2006.062802
      OpenUrlAbstract/FREE Full Text
    2. ↵
      2. Crotty Alexander LE ,
      3. Shin S ,
      4. Hwang JH
      . Inflammatory diseases of the lung induced by conventional cigarette smoke: a review. Chest 2015;148:1307–22. doi:10.1378/chest.15-0409
      OpenUrlCrossRefPubMed
    3. ↵
      2. Pfeiffer S ,
      3. Herzmann C ,
      4. Gaede KI , et al
      . Different responses of the oral, nasal and lung microbiomes to cigarette smoke. Thorax 2022;77:191–5. doi:10.1136/thoraxjnl-2020-216153
      OpenUrlAbstract/FREE Full Text
    4. ↵
      2. Baskaran V ,
      3. Lim WS ,
      4. McKeever TM
      . Effects of tobacco smoking on recurrent hospitalisation with pneumonia: a population-based cohort study. Thorax 2022;77:82–5. doi:10.1136/thoraxjnl-2020-216494
      OpenUrlAbstract/FREE Full Text
    5. ↵
      2. Mathieu E ,
      3. Escribano-Vazquez U ,
      4. Descamps D , et al
      . Paradigms of lung microbiota functions in health and disease, particularly, in asthma. Front Physiol 2018;9:1168. doi:10.3389/fphys.2018.01168
    6. ↵
      2. Einarsson GG ,
      3. Comer DM ,
      4. McIlreavey L , et al
      . Community dynamics and the lower airway microbiota in stable chronic obstructive pulmonary disease, smokers and healthy non-smokers. Thorax 2016;71:795–803. doi:10.1136/thoraxjnl-2015-207235
      OpenUrlAbstract/FREE Full Text
    7. ↵
      2. Garcia-Nuñez M ,
      3. Millares L ,
      4. Pomares X , et al
      . Severity-related changes of bronchial microbiome in chronic obstructive pulmonary disease. J Clin Microbiol 2014;52:4217–23. doi:10.1128/JCM.01967-14
      OpenUrlAbstract/FREE Full Text
    8. ↵
      2. Sapey E ,
      3. Stockley RA
      . COPD exacerbations . 2: aetiology. Thorax 2006;61:250–8. doi:10.1136/thx.2005.041822
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Global Initiative for Chronic Obstructive Lung Disease (GOLD)
      . Global Initiative for Chronic Obstructive Lung Disease. 2022.
    10. ↵
      2. Rønn C ,
      3. Sivapalan P ,
      4. Eklöf J , et al
      . Hospitalization for chronic obstructive pulmonary disease and pneumonia: association with the dose of inhaled corticosteroids. A nation-wide cohort study of 52 100 outpatients. Clin Microbiol Infect 2023;29:523–9. doi:10.1016/j.cmi.2022.11.029
      OpenUrl
    11. ↵
      2. Keir HR ,
      3. Contoli M ,
      4. Chalmers JD
      . Inhaled corticosteroids and the lung microbiome in COPD. Biomedicines 2021;9:1312. doi:10.3390/biomedicines9101312
    12. ↵
      2. Leitao Filho FS ,
      3. Takiguchi H ,
      4. Akata K , et al
      . Effects of inhaled corticosteroid/long-acting β2-agonist combination on the airway microbiome of patients with chronic obstructive pulmonary disease: a randomized controlled clinical trial (DISARM). Am J Respir Crit Care Med 2021;204:1143–52. doi:10.1164/rccm.202102-0289OC
      OpenUrl
    13. ↵
      2. Opron K ,
      3. Begley LA ,
      4. Erb-Downward JR , et al
      . Lung microbiota associations with clinical features of COPD in the SPIROMICS cohort. NPJ Biofilms Microbiomes 2021;7:14. doi:10.1038/s41522-021-00185-9
    14. ↵
      2. Wang Z ,
      3. Liu H ,
      4. Wang F , et al
      . A refined view of airway microbiome in chronic obstructive pulmonary disease at species and strain-levels. Front Microbiol 2020;11:1758. doi:10.3389/fmicb.2020.01758
      OpenUrl
    15. ↵
      2. Erb-Downward JR ,
      3. Thompson DL ,
      4. Han MK , et al
      . Analysis of the lung microbiome in the “healthy” smoker and in COPD. PLoS ONE 2011;6:e16384. doi:10.1371/journal.pone.0016384
    16. ↵
      2. Ditz B ,
      3. Christenson S ,
      4. Rossen J , et al
      . Sputum microbiome profiling in COPD: beyond singular pathogen detection. Thorax 2020;75:338–44. doi:10.1136/thoraxjnl-2019-214168
      OpenUrlAbstract/FREE Full Text
    17. ↵
      2. Grønseth R ,
      3. Drengenes C ,
      4. Wiker HG , et al
      . Protected sampling is preferable in bronchoscopic studies of the airway microbiome. ERJ Open Res 2017;3:00019-2017. doi:10.1183/23120541.00019-2017
    18. ↵
      2. Leiten EO ,
      3. Martinsen EMH ,
      4. Bakke PS , et al
      . Complications and discomfort of bronchoscopy: a systematic review. Eur Clin Respir J 2016;3:33324. doi:10.3402/ecrj.v3.33324
      OpenUrlCrossRefPubMed
    19. ↵
      2. Zakharkina T ,
      3. Heinzel E ,
      4. Koczulla RA , et al
      . Analysis of the airway microbiota of healthy individuals and patients with chronic obstructive pulmonary disease by T-RFLP and clone sequencing. PLoS One 2013;8:e68302. doi:10.1371/journal.pone.0068302
    20. ↵
      2. Mika M ,
      3. Nita I ,
      4. Morf L , et al
      . Microbial and host immune factors as drivers of COPD. ERJ Open Res 2018;4:00015-2018. doi:10.1183/23120541.00015-2018
    21. ↵
      2. Pragman AA ,
      3. Kim HB ,
      4. Reilly CS , et al
      . The lung microbiome in moderate and severe chronic obstructive pulmonary disease. PLoS One 2012;7:e47305. doi:10.1371/journal.pone.0047305
    22. ↵
      2. Segal LN ,
      3. Clemente JC ,
      4. Wu BG , et al
      . Randomised, double-blind, placebo-controlled trial with azithromycin selects for anti-inflammatory microbial metabolites in the emphysematous lung. Thorax 2017;72:13–22. doi:10.1136/thoraxjnl-2016-208599
      OpenUrlAbstract/FREE Full Text
    23. ↵
      2. Gupta S ,
      3. Shariff M ,
      4. Chaturvedi G , et al
      . Comparative analysis of the alveolar microbiome in COPD, ECOPD, Sarcoidosis, and ILD patients to identify respiratory illnesses specific microbial signatures. Sci Rep 2021;11:3963. doi:10.1038/s41598-021-83524-2
    24. ↵
      2. Grønseth R ,
      3. Haaland I ,
      4. Wiker HG , et al
      . The Bergen COPD microbiome study (MicroCOPD): rationale, design, and initial experiences. Eur Clin Respir J 2014;1. doi:10.3402/ecrj.v1.26196
    25. ↵
      2. Leiten EO ,
      3. Eagan TML ,
      4. Martinsen EMH , et al
      . Complications and discomfort after research bronchoscopy in the MicroCOPD study. BMJ Open Respir Res 2020;7:e000449. doi:10.1136/bmjresp-2019-000449
    26. ↵
      2. Hoang T ,
      3. Wiker H ,
      4. Mikal L Eagan T , et al
      . 16S amplicon PCR for the V3-V4 region for the microcopd samples v1. [Preprint] 2019. doi:10.17504/protocols.io.2sygefw
    27. ↵
      2. Yuan S ,
      3. Cohen DB ,
      4. Ravel J , et al
      . Evaluation of methods for the extraction and purification of DNA from the human microbiome. PLoS One 2012;7:e33865. doi:10.1371/journal.pone.0033865
    28. ↵
      2. Bolyen E ,
      3. Rideout JR ,
      4. Dillon MR , et al
      . Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol 2019;37:852–7. doi:10.1038/s41587-019-0209-9
      OpenUrlCrossRefPubMed
    29. ↵
      1. R Foundation for Statistical Computing
      . R: A language and environment for statistical computing. 2020.
    30. ↵
      2. Lin H ,
      3. Peddada SD
      . Analysis of compositions of microbiomes with bias correction. Nat Commun 2020;11. doi:10.1038/s41467-020-17041-7
    31. ↵
      2. Wallen ZD
      . Comparison study of differential abundance testing methods using two large Parkinson disease gut microbiome datasets derived from 16S amplicon sequencing. BMC Bioinformatics 2021;22:265. doi:10.1186/s12859-021-04193-6
    32. ↵
      2. Kaul A ,
      3. Mandal S ,
      4. Davidov O , et al
      . Analysis of microbiome data in the presence of excess zeros. Front Microbiol 2017;8:2114. doi:10.3389/fmicb.2017.02114
    33. ↵
      2. Tangedal S ,
      3. Aanerud M , et al
      . Data from the lower airway microbiota in COPD and healthy controls. Dryad Digital Repository 2023. doi:10.5061/dryad.rfj6q57ff
    34. ↵
      2. Bassis CM ,
      3. Erb-Downward JR ,
      4. Dickson RP , et al
      . Analysis of the upper respiratory tract microbiotas as the source of the lung and gastric microbiotas in healthy individuals. mBio 2015;6:e00037-15. doi:10.1128/mBio.00037-15 pmid:25736890
      OpenUrlAbstract/FREE Full Text
    35. ↵
      2. Ramsheh MY ,
      3. Haldar K ,
      4. Esteve-Codina A , et al
      . Lung microbiome composition and bronchial epithelial gene expression in patients with COPD versus healthy individuals: a bacterial 16S rRNA gene sequencing and host transcriptomic analysis. Lancet Microbe 2021;2:e300–10. doi:10.1016/S2666-5247(21)00035-5 pmid:35544166
      OpenUrlPubMed
    36. ↵
      2. Kim V ,
      3. Rogers TJ ,
      4. Criner GJ
      . New concepts in the pathobiology of chronic obstructive pulmonary disease. Proc Am Thorac Soc 2008;5:478–85. doi:10.1513/pats.200802-014ET pmid:18453359
      OpenUrlCrossRefPubMed
    37. ↵
      2. Persson LJP ,
      3. Aanerud M ,
      4. Hardie JA , et al
      . Antimicrobial peptide levels are linked to airway inflammation, bacterial colonisation and exacerbations in chronic obstructive pulmonary disease. Eur Respir J 2017;49:1601328. doi:10.1183/13993003.01328-2016 pmid:28298400
      OpenUrlAbstract/FREE Full Text
    38. ↵
      2. Madapoosi SS ,
      3. Cruickshank-Quinn C ,
      4. Opron K , et al
      . Lung microbiota and metabolites collectively associate with clinical outcomes in milder stage chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2022;206:427–39. doi:10.1164/rccm.202110-2241OC pmid:35536732
      OpenUrlPubMed
    39. ↵
      2. Huang YJ ,
      3. Sethi S ,
      4. Murphy T , et al
      . Airway microbiome dynamics in exacerbations of chronic obstructive pulmonary disease. J Clin Microbiol 2014;52:2813–23. doi:10.1128/JCM.00035-14 pmid:24850358
      OpenUrlAbstract/FREE Full Text
    40. ↵
      2. Wang Z ,
      3. Bafadhel M ,
      4. Haldar K , et al
      . Lung microbiome dynamics in COPD exacerbations. Eur Respir J 2016;47:1082–92. doi:10.1183/13993003.01406-2015 pmid:26917613
      OpenUrlAbstract/FREE Full Text
    41. ↵
      2. Yip W ,
      3. Li X ,
      4. Koelwyn GJ , et al
      . Inhaled corticosteroids selectively alter the microbiome and host transcriptome in the small airways of patients with chronic obstructive pulmonary disease. Biomedicines 2022;10:1110. doi:10.3390/biomedicines10051110 pmid:35625847
      OpenUrlPubMed
    42. ↵
      2. Tangedal S ,
      3. Nielsen R ,
      4. Aanerud M , et al
      . Sputum microbiota and inflammation at stable state and during exacerbations in a cohort of chronic obstructive pulmonary disease (COPD) patients. PLoS One 2019;14:e0222449. doi:10.1371/journal.pone.0222449 pmid:31527888
      OpenUrlCrossRefPubMed
    43. ↵
      2. Eribe ERK ,
      3. Olsen I
      . Leptotrichia species in human infections II. J Oral Microbiol 2017;9:1368848. doi:10.1080/20002297.2017.1368848 pmid:29081911
      OpenUrlCrossRefPubMed
    44. ↵
      2. Jeong J ,
      3. Yun K ,
      4. Mun S , et al
      . Publisher correction: the effect of taxonomic classification by full-length 16S rRNA sequencing with a synthetic long-read technology. Sci Rep 2021;11:10861. doi:10.1038/s41598-021-90067-z pmid:34011958
      OpenUrlPubMed

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors ST—conceptualisation, data curation, formal analysis, investigation, methodology, visualisation, writing of original draft, editing, and directly accessed and verified the data used in the manuscript. RN—conceptualisation, data curation, investigation, methodology, project administration, supervision, reviewing original draft, and directly accessed and verified the data used in the manuscript. MA—data curation, supervision and reviewing original draft. CD—investigation, methodology and reviewing original draft. GRH—investigation and reviewing original draft. SL—investigation and reviewing original draft. KSK—investigation and reviewing original draft. PSH—conceptualisation and reviewing original draft. TMLE—responsible for the overall content as the guarantor, conceptualisation, data curation, formal analysis, funding acquisition, investigation, reviewing original draft, methodology, project administration, resources, and directly accessed and verified the data used in the manuscript. All authors reviewed, contributed to and approved the final version of the article.

    • Funding The study was funded by the Bergen Medical Research Fund and Helse-Vest (Western Norway Regional Health Authorities; no award/grant number available).

    • Competing interests TMLE—support for the present manuscript; grant from Bergen Medical Research Fund and Helse-Vest (Western Norway Regional Health Authorities; no award/grant number available); other unrelated grants: GlaxoSmithKline. RN—support for the present manuscript: Novartis, Boehringer Ingelheim, GlaxoSmithKline, AstraZeneca and Timber Merchant Delphin’s Endowment; other grants: AstraZeneca. MA—payment for lectures and support for attending meetings: Roche, AstraZeneca and Pfizer. GH—payment for lectures: Boehringer Ingelheim; participation in advisory board for AstraZeneca. KSK—payment for lectures: AstraZeneca and Boehringer Ingelheim. PSH—other grants (paid to department): Boehringer Ingelheim; payment for lectures: AstraZeneca; licensed patent on synthetic antimicrobial peptides.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.