Skip to content

Stable inhaled therapy in COPD

TL;DR — Long-acting bronchodilation is the pharmacologic base for symptomatic stable COPD; LABA/LAMA dual therapy generally improves bronchodilation and reduces exacerbations more than either agent alone. In exacerbation-prone populations, single-inhaler triple LABA/LAMA/ICS therapy reduced exacerbations versus dual therapy in IMPACT and ETHOS, but ICS increases pneumonia risk (Lipson 2018, PMID 29668352; Rabe 2020, PMID 32579807; Zhang 2020, PMID 32643439). Blood eosinophils and exacerbation history estimate ICS benefit; they do not create a binary diagnosis. Device usability, inspiratory flow, technique, adherence, cost and patient goals can dominate molecule-level differences. ICS withdrawal is most defensible when exacerbation risk and eosinophilic signal are low (Chapman 2018, PMID 29779416).

Treatment targets

Target Preferred measurement Why FEV1 alone is insufficient
Breathlessness mMRC, CAT, activity Hyperinflation, muscle and cardiac disease matter
Exacerbations Moderate/severe annualized rate Event history predicts future events
Health status CAT/SGRQ Captures effects not visible in spirometry
Rescue use Doses/day Technique and behavior influence signal
Safety Pneumonia, cardiovascular events, local effects Benefit–harm differs by phenotype

Drug classes

Class Main action Expected role Important harms/limits
SABA Rapid beta-2 bronchodilation Rescue Tremor, tachycardia, overreliance
SAMA Short muscarinic blockade Rescue/acute setting Dry mouth, urinary/ocular precautions
LABA Sustained beta-2 bronchodilation Maintenance Palpitations, tremor
LAMA Sustained muscarinic blockade Maintenance/exacerbation reduction Dry mouth, urinary retention risk
ICS Airway anti-inflammatory Add for exacerbation/type-2 signal Pneumonia, thrush, dysphonia, bruising
LABA/LAMA Dual bronchodilation Persistent symptoms or exacerbations Device and adherence burden
LABA/LAMA/ICS Dual bronchodilation plus steroid Selected exacerbation-prone COPD Pneumonia and steroid exposure

Bronchodilator foundation

Long-acting bronchodilators reduce dynamic hyperinflation, symptoms and exacerbations. Choice between LABA and LAMA depends on response, event history, adverse effects, device and access rather than a universal molecular ranking.

FLAME enrolled exacerbation-risk COPD and found indacaterol/glycopyrronium superior to salmeterol/fluticasone for annual exacerbation rate, supporting LABA/LAMA rather than automatic LABA/ICS escalation (Wedzicha 2016, PMID 27181606).

Escalation situation Evidence-informed option
Persistent dyspnea on one long-acting bronchodilator LABA/LAMA after technique/adherence review
Exacerbation on monotherapy LABA/LAMA or steroid-containing regimen according to eosinophils/asthma traits
Symptoms without events on LABA/LAMA Check diagnosis, device, rehabilitation and comorbidity before adding ICS
Recurrent events on LABA/LAMA Consider triple therapy; quantify eosinophils and pneumonia risk

Inhaled corticosteroids: effect modification

ICS benefit is concentrated in preventing exacerbations, not broadly reversing COPD. Prior exacerbations, blood eosinophil count and coexisting asthma features increase expected benefit; low event risk, low eosinophils, recurrent pneumonia and mycobacterial risk reduce net benefit.

Factor Direction of expected ICS net benefit
≥2 moderate or ≥1 severe recent exacerbation Higher
Blood eosinophils rising toward/above 300 cells/µL Higher
Concomitant asthma Higher; ICS usually required
Recurrent pneumonia Lower
Blood eosinophils persistently <100 cells/µL Lower
Chronic airway infection/bronchiectasis More caution

Pneumonia risk varies with dose, molecule and patient characteristics; meta-analysis confirms a class-associated signal with heterogeneity (Zhang 2020, PMID 32643439).

Triple therapy trials

Trial Comparison Population Main inference
TRILOGY Single-inhaler triple vs ICS/LABA Symptomatic severe COPD with exacerbation history Added LAMA improved outcomes (Singh 2016, PMID 27598678)
FULFIL Once-daily triple vs twice-daily ICS/LABA Symptomatic advanced COPD Lung function/health status favored triple (Lipson 2017, PMID 28375647)
IMPACT Triple vs LABA/LAMA and ICS/LABA Symptomatic exacerbation-prone COPD Fewer moderate/severe events; more pneumonia vs LABA/LAMA (Lipson 2018, PMID 29668352)
ETHOS Two ICS-dose triples vs dual comparators Moderate-to-very severe exacerbation-prone COPD Higher-dose triple reduced exacerbations (Rabe 2020, PMID 32579807)

Trial interpretation is sensitive to prior ICS withdrawal at randomization, enriched exacerbation history and inclusion of asthma-like biology. Mortality analyses are hypothesis-supporting but do not make triple therapy appropriate for low-risk symptomatic disease.

De-escalation

SUNSET tested withdrawal from long-term triple therapy to LABA/LAMA in patients without frequent exacerbations. Average lung-function change was modest, while higher eosinophils identified greater risk of deterioration (Chapman 2018, PMID 29779416).

De-escalation should be planned rather than abrupt: document why ICS was started, confirm event history/eosinophils/asthma, optimize bronchodilation, agree monitoring and define a restart threshold.

Device, deposition and adherence

Device issue Consequence Response
Insufficient inspiratory flow for DPI Reduced powder dispersion Consider pMDI/soft-mist/nebulized route
Poor hand–breath coordination Low lung dose Spacer or breath-actuated alternative
Cognitive/manual limitation Errors and missed doses Simplify regimen and involve caregiver
Multiple device types Conflicting techniques Consolidate where possible
Cost/formulary change Silent nonadherence Ask directly and prescribe accessible option

Small-airway deposition is influenced by particle size and device, but theoretical deposition should not substitute for comparative patient-important outcomes (Usmani 2012, PMID 23154684).

Safety monitoring

Monitor pneumonia, oral candidiasis, dysphonia, bruising and cumulative steroid exposure with ICS. Bronchodilator cardiovascular safety is reassuring on average in randomized trials, but individual arrhythmia, ischemia and urinary/ocular risks still require context (Yang 2023, PMID 36137586).

Landmark bronchodilator and withdrawal effect sizes

Trial Comparison Quantitative result Boundary
UPLIFT Tiotropium versus placebo on background therapy Pre-bronchodilator FEV1 advantage 87–103 mL and SGRQ difference 2.3–3.3 points; no difference in post-day-30 FEV1 decline (Tashkin 2008, PMID 18836213). Control improved; disease modification was not shown.
POET-COPD Tiotropium versus salmeterol First exacerbation 187 versus 145 days; HR 0.83 (95% CI 0.77–0.90); severe-event rate ratio 0.73 (0.66–0.82) (Vogelmeier 2011, PMID 21428765). One-year exacerbation-prone population.
WISDOM Stepwise ICS withdrawal First moderate/severe exacerbation HR 1.06 (95% CI 0.94–1.19), meeting non-inferiority; trough FEV1 43 mL lower at week 52 (Magnussen 2014, PMID 25196117). Average neutrality can conceal subgroup deterioration.
SUNSET Direct de-escalation in non-frequent exacerbators Small average FEV1 loss; higher eosinophils identified greater risk (Chapman 2018, PMID 29779416). Does not apply to asthma or frequent events.

New inhaled mechanism

ENHANCE-1 and ENHANCE-2 randomized 760 and 789 participants. Ensifentrine improved 0–12-hour FEV1 AUC by 87 mL (95% CI 55–119) and 94 mL (65–124); moderate/severe exacerbation rate ratios were 0.64 (0.40–1.00) and 0.57 (0.38–0.87). Symptom and SGRQ improvements replicated incompletely (Anzueto 2023, PMID 37364283). The first-approval report documents regulatory transition, but positioning against contemporary dual/triple therapy remains unsettled (Keam 2024, PMID 39196510).

Treatment-selection controversies

Question Evidence for escalation Evidence for restraint
Add ICS? Exacerbation history and increasing eosinophils predict more benefit (Yun 2018, PMID 29709670) Pneumonia risk and low benefit at low eosinophils (Zhang 2020, PMID 32643439)
Withdraw ICS? WISDOM/SUNSET support selected withdrawal (Magnussen 2014, PMID 25196117; Chapman 2018, PMID 29779416) Lung-function loss and eosinophilic deterioration require monitoring
Add drug for dyspnea? Dual bronchodilation can reduce hyperinflation Comorbidity, deconditioning and device failure may dominate (Welling 2023, PMID 37039738)
Infer mortality benefit? IMPACT/ETHOS analyses are compatible with benefit (Lipson 2018, PMID 29668352; Rabe 2020, PMID 32579807) Prior-ICS withdrawal, enrichment and multiplicity complicate causality

Mortality, exacerbations and pneumonia do not move together

SUMMIT provides an important neutral comparator to exacerbation-enriched triple-therapy trials. Among 16,485 people with moderate COPD and cardiovascular risk, fluticasone furoate/vilanterol did not significantly reduce all-cause mortality (HR 0.88, 95% CI 0.74–1.04) or cardiovascular events (HR 0.93, 0.75–1.14), although it reduced exacerbations and slowed FEV1 decline by 8 mL/year (1–15) (Vestbo 2016, PMID 27203508). A meta-analysis restricted to 52-week single-inhaler trials estimated mortality RR 0.69 (0.53–0.90) for triple therapy versus LABA/LAMA, no difference versus ICS/LABA (RR 0.94, 0.72–1.24), and pneumonia RR 1.43 (1.21–1.68) versus LABA/LAMA (Lai 2022, PMID 35207460). These data support benefit–harm stratification; they do not justify extrapolating a mortality claim to people without recent exacerbations or to comparisons where the steroid component is unchanged.

Formal recommendations encode some of that uncertainty differently. ATS strongly recommends LABA/LAMA over monotherapy for dyspnea/exercise intolerance, but makes triple therapy and ICS withdrawal conditional on exacerbation history and makes no general recommendation for adding ICS solely for eosinophilia outside an exacerbation-prone group (Nici 2020, PMID 32283960). ERS conditionally supports withdrawal in people without frequent exacerbations but strongly recommends against withdrawal when eosinophils are at least 300/µL (Chalmers 2020, PMID 32366483).

Oral anti-inflammatory add-on evidence shows a narrower responder phenotype and substantial discontinuation. RE2SPOND’s overall exacerbation rate ratio was 0.92 (95% CI 0.81–1.04), not significant, while adverse-event discontinuation was 11.7% with roflumilast versus 5.4% with placebo (Martinez 2016, PMID 27585384). A prespecified pooled REACT/RE2SPOND analysis estimated a 12.3% overall reduction (rate ratio 0.88, 0.80–0.97), larger after prior hospitalization (0.74, 0.63–0.88) and with eosinophils at least 300/µL plus prior hospitalization (0.57, 0.37–0.88) (Martinez 2018, PMID 29763572). The subgroup signal requires prospective confirmation and must be weighed against tolerability.

Technique and mortality evidence have different failure modes

A COPD-specific review found 19 studies and 6,487 participants relating objectively assessed inhaler errors to outcomes, but 15 studies were low quality, adherence was usually unmeasured, and the publications reported too little detail to allow quantitative pooling; evidence linking fewer errors to better FEV1, symptoms, health status or exacerbations was therefore weak (Halpin 2024, PMID 38626929). Technique checks remain sensible, but a device-training intervention needs outcome evidence rather than assuming every observed error is critical.

Across 60 randomized trials and 103,034 participants, ICS-containing treatment was associated with lower all-cause mortality than non-ICS inhaled therapy (Peto OR 0.90, 95% CI 0.84–0.97), with a larger triple-therapy estimate (OR 0.73, 95% CI 0.59–0.91); eosinophils ≥200/µL were the strongest reported subgroup predictor (OR 0.58, 95% CI 0.36–0.95) (Chen 2023, PMID 35921883). This aggregate signal competes with pneumonia harm and with concerns about withdrawal-enriched comparators and trial selection; it does not justify ICS for low-exacerbation, low-eosinophil COPD.

Open questions

  • What is the net mortality effect of triple therapy after accounting for ICS withdrawal and event enrichment? (Lipson 2018, PMID 29668352; Rabe 2020, PMID 32579807)
  • Can repeated eosinophil trajectories guide ICS initiation and withdrawal better than one count? (Chapman 2018, PMID 29779416)
  • Which device-matching intervention improves outcomes rather than technique scores alone? (Usmani 2012, PMID 23154684)
  • How should pneumonia history and bronchiectasis alter ICS thresholds? (Zhang 2020, PMID 32643439)
  • In Japanese claims data, single-inhaler triple therapy had higher adherence than multiple-inhaler therapy at 6, 12 and 18 months, but absolute persistence remained short and the study did not establish a causal reduction in admissions (Joksaitė 2024, PMID 39632104).

References

  1. Wedzicha JA, et al. Indacaterol-glycopyrronium versus salmeterol-fluticasone for COPD. N Engl J Med. 2016. PMID 27181606
  2. Lipson DA, et al. Once-daily single-inhaler triple versus dual therapy in COPD. N Engl J Med. 2018. PMID 29668352
  3. Rabe KF, et al. Triple inhaled therapy at two glucocorticoid doses in COPD. N Engl J Med. 2020. PMID 32579807
  4. Zhang Q, et al. Pneumonia risk with different inhaled corticosteroids in COPD: meta-analysis. COPD. 2020. PMID 32643439
  5. Chapman KR, et al. SUNSET: triple therapy de-escalation to indacaterol/glycopyrronium. Am J Respir Crit Care Med. 2018. PMID 29779416
  6. Singh D, et al. TRILOGY single-inhaler triple versus ICS/LABA. Lancet. 2016. PMID 27598678
  7. Lipson DA, et al. FULFIL trial: once-daily triple therapy. Am J Respir Crit Care Med. 2017. PMID 28375647
  8. Usmani OS, et al. Treating the small airways. Respiration. 2012. PMID 23154684
  9. Yang M, et al. Long-acting bronchodilator combinations and major cardiovascular events: systematic review/meta-analysis. Eur Respir J. 2023. PMID 36137586
  10. Tashkin DP, et al. A 4-year trial of tiotropium in COPD. N Engl J Med. 2008. PMID 18836213
  11. Vogelmeier C, et al. Tiotropium versus salmeterol for prevention of COPD exacerbations. N Engl J Med. 2011. PMID 21428765
  12. Magnussen H, et al. Withdrawal of inhaled glucocorticoids and COPD exacerbations. N Engl J Med. 2014. PMID 25196117
  13. Anzueto A, et al. Ensifentrine phase 3 ENHANCE trials. Am J Respir Crit Care Med. 2023. PMID 37364283
  14. Keam SJ. Ensifentrine: first approval. Drugs. 2024. PMID 39196510
  15. Yun JH, et al. Blood-eosinophil thresholds and COPD exacerbations. J Allergy Clin Immunol. 2018. PMID 29709670
  16. Welling JBA, et al. Dynamics of hyperinflation. Respirology. 2023. PMID 37039738
  17. Jokšaitė S, et al. Comparative adherence and persistence of single-inhaler and multiple-inhaler triple therapies among patients with COPD in Japan. BMJ Open. 2024. PMID 39632104
  18. Vestbo J, et al. Fluticasone furoate and vilanterol and survival in COPD with heightened cardiovascular risk. Lancet. 2016. PMID 27203508
  19. Lai CC, et al. Single-inhaler triple versus dual therapy and mortality in COPD: systematic review and meta-analysis. Life (Basel). 2022. PMID 35207460
  20. Nici L, et al. Pharmacologic management of COPD: an official ATS clinical practice guideline. Am J Respir Crit Care Med. 2020. PMID 32283960
  21. Chalmers JD, et al. Withdrawal of inhaled corticosteroids in COPD: an ERS guideline. Eur Respir J. 2020. PMID 32366483
  22. Martinez FJ, et al. Roflumilast added to ICS/LABA in RE2SPOND: randomized trial. Am J Respir Crit Care Med. 2016. PMID 27585384
  23. Martinez FJ, et al. Determinants of response to roflumilast: pooled randomized-trial analysis. Am J Respir Crit Care Med. 2018. PMID 29763572
  24. Halpin DMG, et al. Systematic review of the effects of patient errors using inhaled delivery systems on clinical outcomes in COPD. BMJ Open Respir Res. 2024. PMID 38626929
  25. Chen H, et al. Association of inhaled corticosteroids with all-cause mortality risk in patients with COPD: a meta-analysis of 60 randomized controlled trials. Chest. 2023. PMID 35921883