Skip to content

COPD inflammation and endotypes

TL;DR — COPD inflammation is heterogeneous: macrophage/neutrophil-dominant, eosinophilic/type-2, lymphoid, infectious and low-inflammatory patterns overlap (Barnes 2016, PMID 27373322). Blood eosinophils are the most clinically developed biomarker because they predict average ICS and type-2 biologic benefit, but values vary and do not directly measure lung tissue (David 2021, PMID 33122447). Dupilumab’s replicated benefit in eosinophilic chronic-bronchitis COPD validates a selected type-2 endotype (Bhatt 2023, PMID 37272521; Bhatt 2024, PMID 38767614). IL-5 trials produced smaller or inconsistent effects, showing that one biomarker can index more than one causal pathway (Pavord 2017, PMID 28893134). Precision treatment requires prospective treatment-response validation, not molecular clustering alone.

Cells and pathways

Axis Sources/effectors Candidate consequences Treatment link
Neutrophilic Neutrophils, CXCR2 signals, proteases Mucus, tissue injury, infection-associated events No validated broad anti-neutrophil therapy
Macrophage Alveolar/interstitial macrophages Proteases, efferocytosis failure, cytokines Investigational
Type 2/eosinophilic IL-4/IL-13, eosinophils, epithelium Exacerbations, mucus, steroid response ICS, dupilumab
Adaptive immune CD8/T cells, B-cell follicles Persistent injury/autoimmune-like organization Investigational
Inflammasome/alarmin IL-33, TSLP and epithelial signals Trigger sensing and amplification Clinical trials
Infectious Dysbiosis, bacterial/viral triggers Exacerbations and neutrophilia Antibiotic/prevention pathway

Barnes synthesized macrophage, neutrophil, lymphocyte, oxidative and steroid-resistance mechanisms, emphasizing why COPD cannot be treated as uniformly steroid-responsive (Barnes 2016, PMID 27373322).

Phenotype versus endotype

A phenotype is observable—emphysema, chronic bronchitis, eosinophilia, frequent events. An endotype asserts a causal biological mechanism. A treatable trait additionally predicts benefit from a specific intervention (Christenson 2023, PMID 37353326).

Label Minimum evidence
Phenotype Reproducible clinical/physiological feature
Endotype Mechanistic pathway linked to disease
Prognostic biomarker Predicts outcome regardless of treatment
Predictive biomarker Modifies treatment effect
Treatable trait Validated test plus effective intervention

Blood eosinophils

Blood eosinophil counts correlate imperfectly with sputum/tissue eosinophilia but are accessible. Threshold analyses show a continuous relationship with exacerbation risk and steroid response rather than a natural binary split (Yun 2018, PMID 29709670).

Use Evidence maturity Limitation
ICS escalation Guideline-integrated Counts vary; event history also required
ICS withdrawal Higher counts predict greater withdrawal risk Thresholds are context-specific
Dupilumab selection Phase 3 threshold ≥300/µL Trial also required chronic bronchitis and triple therapy
Acute steroid guidance Emerging Severe-event safety/generalizability
Disease diagnosis Not appropriate Eosinophilia occurs in multiple diseases

Type-2 biologic proof and limits

METREX/METREO tested mepolizumab in eosinophilic COPD with modest/inconsistent event reduction (Pavord 2017, PMID 28893134; Pavord 2021, PMID 34163157). Dupilumab reduced annualized exacerbations and improved FEV1 in BOREAS, then replicated benefit in NOTUS (Bhatt 2023, PMID 37272521; Bhatt 2024, PMID 38767614).

The contrast suggests IL-4/IL-13 signaling and mucus/epithelial biology may matter beyond eosinophil depletion. It also warns against assuming class effects among biologics (Rabe 2023, PMID 37348121).

Neutrophils, infection and steroid resistance

Neutrophilia may reflect smoke injury, bacterial colonization, exacerbation or systemic stress. Blocking recruitment risks infection because neutrophils are host-defense cells. Oxidative stress and altered histone-deacetylase biology have been proposed in corticosteroid resistance (Barnes 2016, PMID 27373322).

The airway microbiome is dynamic and affected by antibiotics, ICS, disease severity and sampling. Association between a taxon and outcome does not establish that organism as a causal endotype.

Tissue-resolved biology

Spatial transcriptomics identified emphysema-associated B-cell communities (Rojas-Quintero 2024, PMID 37934672). Experimental work suggests eosinophils can promote matrix destruction through cathepsin L, complicating the simple view that eosinophilia is only treatment-responsive inflammation (Xu 2023, PMID 37816708).

Systemic inflammation

CRP, fibrinogen and leukocyte measures associate with outcomes but are influenced by obesity, infection, cardiovascular disease and smoking. A systemic marker can be prognostic without being causal or lung-specific.

Longitudinal inflammatory states

In ECLIPSE, 30% of participants with COPD had no measured systemic inflammatory signal and 16% had persistent inflammation. The persistently inflamed group had 13% versus 2% mortality and 1.5 versus 0.9 exacerbations/year despite similar pulmonary abnormalities (Agustí 2012, PMID 22624038). At eight years, changes in exacerbations, BODE, SGRQ, systemic inflammation and later FEV1 decline were independently associated with mortality and poorly predicted by one baseline measure (Celli 2021, PMID 33303557).

Biologics as mechanistic perturbations

Pathway/trial Enrichment Quantitative result Interpretation
IL-5Rα, GALATHEA/TERRANOVA Frequent exacerbations; primary stratum eosinophils ≥220/µL No dose met the primary significance threshold; GALATHEA 100 mg rate ratio 0.83, P=0.05, while TERRANOVA rate ratios were 0.85–1.04 (Criner 2019, PMID 31112385). Eosinophil depletion did not ensure average success.
IL-4Rα, BOREAS/NOTUS Eosinophils ≥300/µL, chronic bronchitis, triple therapy Replicated event and FEV1 benefit (Bhatt 2023, PMID 37272521; Bhatt 2024, PMID 38767614). Supports IL-4/IL-13 biology only in the enrolled population.
IL-5, MATINEE Eosinophils ≥300/µL, exacerbations despite triple therapy Exacerbations 0.80 versus 1.01/year; rate ratio 0.79 (95% CI 0.66–0.94), without significant symptom/QoL differences (Sciurba 2025, PMID 40305712). Selection produced an event signal, not uniform multidomain benefit.
TSLP, COURSE Exacerbation-prone COPD on triple therapy Overall primary result did not establish a broad effect; prespecified eosinophil subgroups motivated further study (Singh 2025, PMID 39653044). Upstream blockade remains stratified.

Acute versus chronic eosinophil use

CORTICO-COP reduced median systemic-steroid exposure from 5 to 2 days while preserving days alive and out of hospital (difference −0.4 days, 95% CI −1.3 to 0.5); 30-day treatment failure was 26% in both arms (Sivapalan 2019, PMID 31122894). STARR2 extends biomarker-directed steroid reduction to primary care (Ramakrishnan 2024, PMID 37924830). Neither proves that the same threshold should determine long-term ICS or biologic selection.

Infection–inflammation tension

Neutrophils, macrophages and adaptive immune aggregates can promote proteolysis, mucus and remodeling, yet antimicrobial defense is indispensable (Barnes 2016, PMID 27373322; Kheradmand 2023, PMID 36201635). ICS-associated pneumonia shows that lowering inflammation is not itself net benefit (Zhang 2020, PMID 32643439). Mucus plugs may bridge epithelial type-2 biology, infection and prognosis, but CT plug burden remains associative (Diaz 2023, PMID 37210745).

Measurement and classification failure modes

Signal Useful role Failure mode
Blood eosinophils Estimates probability of ICS/biologic benefit Varies with time, infection and systemic steroid exposure
Sputum differential Proximal airway inflammatory pattern Induction and processing limit scale; samples may not represent distal lung
CRP/fibrinogen Systemic risk association Nonspecific to COPD and affected by comorbidity
CT mucus plugs Spatial epithelial/mucus phenotype Association does not prove type-2 causation or treatment response
Tissue transcriptomics Resolves local cell states and niches End-stage tissue and dissociation/spatial biases affect inference
Microbiome profiles Community structure and instability Antibiotics, ICS and sampling method can dominate signal

Endotype validation standard

An endotype should satisfy more than clustering: reproducibility across cohorts; temporal stability or a specified transition rule; biological coherence; prognostic value beyond routine clinical variables; and, most importantly, a prospective treatment interaction. Dupilumab and MATINEE provide pathway-specific perturbation evidence in narrowly selected populations (Bhatt 2023, PMID 37272521; Sciurba 2025, PMID 40305712). Persistent inflammation and spatial immune niches remain prognostic or mechanistic associations until an intervention changes outcomes through the proposed pathway (Agustí 2012, PMID 22624038; Rojas-Quintero 2024, PMID 37934672).

Controversy map

Claim Evidence supporting Evidence limiting
Eosinophilic COPD is one disease Repeated ICS/dupilumab/mepolizumab event signals Benralizumab neutrality and incomplete symptom effects
Neutrophils are a therapeutic target Protease and mucus biology Host-defense loss and infection risk
Systemic inflammation is spillover Longitudinal mortality association Many patients lack persistent inflammation; causality unproven
Immune follicles drive emphysema Spatial localization and adaptive-immunity biology Reverse causation and end-stage sampling

NETs and the gut–lung axis: mechanism versus tractable target

Neutrophil extracellular traps (NETs) combine antimicrobial DNA/protein scaffolds with a liability: excess proteases, histones and extracellular DNA can injure tissue, impede mucociliary clearance and paradoxically impair bacterial killing (Keir 2022, PMID 35197267). Experimental and natural rhinovirus exacerbation studies found amplified airway NET formation in COPD, correlation with inflammatory and clinical severity, and protection from immunopathology after NETosis inhibition or DNase in mice, supporting causality in that model but not yet a human treatment effect (Katsoulis 2024, PMID 38982052). A separate gut–lung literature proposes circulating mediators, dysbiosis, barrier dysfunction, hypoxia and oxidative stress as bidirectional links, but currently mixes human association, animal and in-vitro evidence (Wang 2023, PMID 36883945). These mechanisms therefore enlarge the target space while also raising a stricter validation requirement: demonstrate target engagement in the airway, preserved host defense and a patient-important outcome.

Longitudinal microbiome data show why a single sputum sample is an unstable endotype label. Across 1,706 samples from 510 participants, neutrophilic disease separated into a relatively stable Haemophilus/IL-1β/TNFα state and a dynamic balanced-microbiome/IL-17A state that could switch toward eosinophilic inflammation (Wang 2021, PMID 33332995). In a smaller extreme-trajectory cohort, 156 mL/year FEV1 loss was accompanied by higher MUC5AC/MUC5B, altered mucus mechanics and emergence of Achromobacter/Klebsiella, with later severe disease associated with Haemophilus, Moraxella and Pseudomonas (Meldrum 2024, PMID 38315959). Antibiotic, ICS and exacerbation effects remain inseparable from much of this observational biology.

The microbiome signal is prognostic, dynamic and not yet causal

In 200 people with severe COPD followed for three years with 1,179 sputum samples, viral detection was strongly linked to exacerbation; rhinovirus, coronavirus and influenza were detected in 13.1%, 5.1% and 3.6% of events, while Moraxella and Haemophilus were respectively 5-fold and 1.6-fold more likely to dominate during exacerbation (Bouquet 2020, PMID 32228581). This supports a dynamic virus–bacteria model rather than a stable single-pathogen endotype.

In a separate 253-person stable-COPD cohort, Proteobacteria dominance and lower diversity tracked neutrophilic/low-eosinophil disease and more frequent exacerbations; Proteobacteria dominance predicted mortality versus Firmicutes-dominant profiles (HR 2.58, 95% CI 1.43–4.66), but the estimate versus balanced communities was imprecise (HR 7.47, 95% CI 1.02–54.86) (Dicker 2021, PMID 32353489). Antibiotics, severity and airway ecology can each produce the observed state, so prognostic association cannot establish that microbiome manipulation will help.

Open questions

  • Why did dupilumab produce clearer benefit than anti-IL-5 strategies at similar eosinophil thresholds? (Pavord 2017, PMID 28893134; Bhatt 2024, PMID 38767614)
  • How stable are eosinophil-based treatment assignments over time? (David 2021, PMID 33122447)
  • Can neutrophilic inflammation be modified without increasing serious infection? (Brightling 2019, PMID 31073084)
  • Do tissue B-cell communities cause emphysema progression? (Rojas-Quintero 2024, PMID 37934672)
  • Which acute events should avoid systemic steroids? (Ramakrishnan 2024, PMID 37924830)

References

  1. Barnes PJ. Inflammatory mechanisms in COPD. J Allergy Clin Immunol. 2016. PMID 27373322
  2. Brightling C, et al. Airway inflammation in COPD: progress to precision medicine. Eur Respir J. 2019. PMID 31073084
  3. David B, et al. Eosinophilic inflammation in COPD: marker to treatable trait. Thorax. 2021. PMID 33122447
  4. Yun JH, et al. Blood eosinophil thresholds and exacerbations in COPD. J Allergy Clin Immunol. 2018. PMID 29709670
  5. Pavord ID, et al. Mepolizumab for eosinophilic COPD. N Engl J Med. 2017. PMID 28893134
  6. Pavord ID, et al. Mepolizumab for eosinophil-associated COPD: METREX/METREO analysis. Int J Chron Obstruct Pulmon Dis. 2021. PMID 34163157
  7. Bhatt SP, et al. Dupilumab for COPD with type-2 inflammation. N Engl J Med. 2023. PMID 37272521
  8. Bhatt SP, et al. Dupilumab for COPD with blood-eosinophil evidence of type-2 inflammation. N Engl J Med. 2024. PMID 38767614
  9. Rabe KF, et al. Targeting type-2 inflammation and epithelial alarmins in COPD. Am J Respir Crit Care Med. 2023. PMID 37348121
  10. Christenson SA. COPD phenotyping. Respir Care. 2023. PMID 37353326
  11. Xu X, et al. Eosinophils promote pulmonary matrix destruction via cathepsin L. Signal Transduct Target Ther. 2023. PMID 37816708
  12. Rojas-Quintero J, et al. Spatial transcriptomics in emphysema. Am J Respir Crit Care Med. 2024. PMID 37934672
  13. Agustí A, et al. Persistent systemic inflammation and poor COPD outcomes. PLoS One. 2012. PMID 22624038
  14. Celli B, et al. Markers of COPD disease activity and eight-year mortality. Eur Respir J. 2021. PMID 33303557
  15. Criner GJ, et al. Benralizumab for prevention of COPD exacerbations. N Engl J Med. 2019. PMID 31112385
  16. Sciurba FC, et al. Mepolizumab to prevent exacerbations of eosinophilic COPD. N Engl J Med. 2025. PMID 40305712
  17. Singh D, et al. Tezepelumab in COPD: COURSE. Lancet Respir Med. 2025. PMID 39653044
  18. Sivapalan P, et al. Eosinophil-guided corticosteroids in hospitalized COPD exacerbations. Lancet Respir Med. 2019. PMID 31122894
  19. Zhang Q, et al. Pneumonia risk with different inhaled corticosteroids in COPD patients: a meta-analysis. COPD. 2020. PMID 32643439
  20. Diaz AA, et al. Airway-occluding mucus plugs and mortality in COPD. JAMA. 2023. PMID 37210745
  21. Kheradmand F, et al. Adaptive immunity in COPD and experimental emphysema. Physiol Rev. 2023. PMID 36201635
  22. Ramakrishnan S, et al. Blood-eosinophil-guided prednisolone for COPD exacerbations: STARR2. Lancet Respir Med. 2024. PMID 37924830
  23. Keir HR, et al. Neutrophil extracellular traps in chronic lung disease: implications for pathogenesis and therapy. Eur Respir Rev. 2022. PMID 35197267
  24. Katsoulis O, et al. Neutrophil extracellular traps promote immunopathogenesis of virus-induced COPD exacerbations. Nat Commun. 2024. PMID 38982052
  25. Wang L, et al. The bidirectional gut–lung axis in COPD. Am J Respir Crit Care Med. 2023. PMID 36883945
  26. Wang Z, et al. Inflammatory-endotype-associated airway microbiome in stable and exacerbated COPD. Am J Respir Crit Care Med. 2021. PMID 33332995
  27. Meldrum OW, et al. Accelerated lung-function decline and mucus–microbe evolution in COPD. Am J Respir Crit Care Med. 2024. PMID 38315959
  28. Bouquet J, et al. Microbial burden and viral exacerbations in a longitudinal multicenter COPD cohort. Respir Res. 2020. PMID 32228581
  29. Dicker AJ, et al. The sputum microbiome, airway inflammation, and mortality in COPD. J Allergy Clin Immunol. 2021. PMID 32353489