From Wiki Journal Club
Jump to: navigation, search
Magnussen H, et al. "Withdrawal of inhaled glucocorticoids and exacerbations of COPD". The New England Journal of Medicine. 2014. 371(14):1285-1295.
PubMedFull textPDF

Clinical Question

In patients with severe COPD receiving triple therapy (tiotropium, salmeterol, and fluticasone), is the withdrawal of inhaled corticosteroids associated with an increase in the frequency of COPD exacerbations?

Bottom Line

Among patients with severe COPD whose symptoms were stable while receiving triple therapy (tiotropium, salmeterol, and fluticasone), tapering off the inhaled corticosteroid (ICS) portion of therapy was not associated with an increase in the rate of COPD exacerbations. Withdrawal of ICS was associated with a modest decrease in the absolute FEV1 of uncertain significance.

Major Points

Initial pharmacological management of high-risk COPD (either by GOLD ≥3 severity or frequent exacerbations) is with long-acting bronchodilators plus an ICS.[1] This approach is based upon the evidence from TORCH (2007), which found that these therapies reduced the rate of lung function decline, the frequency of COPD exacerbations, and yielded symptom improvement. Importantly in TORCH, neither bronchdilators nor ICS improved overall survival, and ICS was associated with an increased incidence of pneumonia. A step-down approach in which ICS was withdrawn was subsequently proposed, but whether this step-down approach would be non-inferior to ICS continuation had not been formally evaluated.[2]

Published in 2014, the Withdrawal of Inhaled Steroids during Optimized Bronchodilator Management (WISDOM) trial studied 2,485 patients with severe to very severe COPD receiving tiotropium, salmeterol, and the ICS fluticasone propionate. Patients were randomized to continuation or slow withdrawal of the ICS over a 12-week period. The hypothesis was that ICS withdrawal would be noninferior to ICS continuation for the primary outcome of time to first moderate/severe COPD exacerbation, with a noninferiority margin of 20%. At 12 months, there was no statistically significant difference between groups for the primary outcome of time to first moderate/severe COPD exacerbation, with a hazard ratio of 1.06 (95% CI 0.94-1.19). Since the upper limit of the hazard ratio was less than the prespecified margin of 1.20, this study demonstrated that ICS withdrawal was noninferior to ICS continuation, although whether the 1.20 margin was optimal is up for debate.[3]

Interestingly, while both groups experienced reductions in FEV1, the degree of FEV1 reduction was greater in the ICS withdrawal group (between-group difference of 38 ml and 43 ml at 18 and 52 weeks, respectively). The clinical significance of such modest differences are unknown, but raises the question of whether ICS continuation may slow the pace of lung function decline.

WISDOM demonstrated that ICS withdrawal is generally safe and well tolerated, and does not appear to increase the risk of first COPD exacerbation compared to patients receiving ongoing ICS therapy. The long-term benefits of ICS withdrawal are unknown, but are likely to include reduced rates of pneumonia and minor infections such as oropharyngeal candidiasis. Ultimately the decision to continue or withdrawal ICS therapy must be made on a patient-to-patient basis, but the risk of infections on the one hand, and of potential lung function decline on the other, must be considered.


GOLD COPD (2015, adapted)[1]

  • Regular use of ICS improves symptoms, lung function, QOL, and reduces exacerbations in those with COPD and FEV1 <60% of predicted (evidence A)
  • Withdrawal of ICS may cause exacerbations in some patients, though gradual withdrawal over three months doesn't increase risk of exacerbations in the medium term
    • This withdrawal is associated with significant deterioration of lung function
  • Regular use of ICS doesn't alter FEV1 decline or mortality (evidence A)
  • Definitions of severity of airflow limitation, post-bronchodilation:
    • GOLD 1 (mild): FEV1 ≥80% of predicted
    • GOLD 2 (moderate): FEV1 50-80% of predicted
    • GOLD 3 (severe): FEV1 30-50% of predicted
    • GOLD 4 (very severe): FEV1 ≤30% of predicted


  • Multicenter, randomized, double-blind, noninferiority, randomized control trial
  • Setting: 200 centers in 23 countries
  • Enrollment: 2009-2013
  • N=2,485
    • Continuation (n=1,243)
    • Withdrawal (n=1,242)
  • Follow up: 12 months for the primary outcome
  • Analysis: Modified intention-to-treat
  • Primary outcome: Time to first moderate or severe COPD exacerbation in the first 12 months


Inclusion Criteria

  • Age ≥40 years
  • Current (≥10 PYH) or former smokers
  • Diagnosis of severe or very severe COPD
  • ≥1 COPD exacerbation in the prior year
  • Able to perform study procedures (eg, exercise testing)
  • Able to use study medication

Exclusion Criteria

  • Existing medical conditions other than COPD which will influence the results (opinion of the investigator)
  • Prior thoracotomy/pulmonary resection
  • Current clinical diagnosis of asthma requiring corticosteroid therapy
  • Use of oxygen for more than 1 hour per day (night-time oxygen is allowed)
  • Use of systemic corticosteroids at doses in excess of 5mg/day in prior 6 weeks
  • Respiratory tract infection or COPD exacerbation in the 6 weeks prior to the initial screening visit or during the screening period prior to visit 2
  • Recent MI
  • Prior unstable arrhythmia requiring intervention or change in drug therapy in the prior 3 months
  • Patients who have been hospitalized for cardiac failure in the prior year
  • Malignancy with active XRT or chemotherapy
  • Clinical diagnosis of bronchiectasis
  • Pregnant, nursing women, or inadequate contraception
  • Participation in another trial with investigation drug recently
  • Known hypersensitivity to a study medication
  • Active enrollment in pulmonary rehabilitation or enrollment in the prior four weeks
  • History of alcohol or drug abuse
  • Currently participating in another investigational study
  • Previously participated in this study

Baseline Characteristics

From the continuation group.

  • Demographics: Male 82%, age 64 years
  • COPD data: Former smoker 65%, duration of disease 8 years
    • GOLD 3: 61%
    • GOLD 4: 38%
    • Lung function: FEV1 34% of predicted
  • SGRQ score: 46 (out of 100, higher is worse overall health functional status)
  • mMRC score: 1.8 (out of 4, higher is worse dyspnea)
  • Medications: LAMA 47%, LABA 65%, ICS 70%, all three 38%


  • Participants received triple therapy during a 6 week run-in, consisting of:
    • Tiotropium 18 mcg once a day
    • Salmeterol 50 mcg twice a day
    • Fluticasone propionate 500 mcg twice daily
  • Participants were then randomized to a group:
    • Continuation - Continue fluticasone propionate
    • Withdrawal - Withdrawal of fluticasone propionate in three steps over a 12-week period:
      • STEP 1: beginning week 7 reduce 1000µg to 500µg, then
      • STEP 2: beginning week 13 reduce 500µg to 200µg, then
      • STEP 3: beginning week 19 reduce 200µg to 0µg (placebo) for the remaining 40 weeks.


Presented as withdrawal vs. continuation. Analyses are for noninferiority.

Primary Outcomes

Time to first moderate or severe COPD exacerbation in the first 12 months
HR 1.06; 95% CI 0.94-1.19; P=0.35

Secondary Outcomes

Rate of moderate or severe COPD exacerbations
0.95 vs. 0.91/patient-year (NS)
Time to first severe COPD exacerbation
HR 1.20; 95% CI 0.98-1.48
Change in FEV1 at from baseline to week 18
38 mL greater loss in the withdrawal group (P<0.001)
Change in health assessment scores from baseline
Week 18: -0.001 vs. -0.030 points (P=0.36)
Week 52: +0.035 vs. -0.028 points (P=0.06)
Week 27: +0.55 vs. -0.42 points (P=0.08)
Week 52: +1.15 vs. -0.07 points (P=0.047)

Adverse Events

24.2% vs. 23.5%
Fatal: 3.2% vs. 2.7%
5.5% vs. 5.8%
Cardiac event
2.2% vs. 2.0%
Major: 1.5% vs. 1.1%
0.5% vs. 0.7%


  • Low ethnic and sex diversity
  • The study population consisted of smokers and former smokers with COPD and may not be applicable to patients with other causes of COPD
  • Statistical analysis was performed by the sponsor; all medications were supplied by the sponsor; editorial assistance was provided by a medical writer employed by a company paid by the sponsor
  • The duration of follow-up was probably too short to make assumptions about the use of inhaled corticosteroids in patients with COPD[4]
  • A subgroup analysis by disease severity would have been informative[4]
  • Individuals without prior COPD exacerbation would have been an interesting group to include[4]
  • Sputum eosinophils may have helped identify those more likely to do poorly with the withdrawal of steroids[4]


  • Boehringer Ingelheim Pharma

Further Reading

  1. 1.0 1.1 GOLD writers. "Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease." www.goldcopd.org. Published 2015-04. Accessed 2015-04-30.
  2. Reilly JJ. "Editorial: Stepping down therapy in COPD." The New England Journal of Medicine. 2014;371:1340-1341.
  3. Erbland ML. Role of inhaled glucocorticoid therapy in stable COPD. In: UpToDate, Post TW (Ed), UpToDate, Waltham, MA. (Accessed on May 26, 2015.)
  4. 4.0 4.1 4.2 4.3 Multiple authors. "Correspondence: Inhaled glucocorticoids and COPD exacerbations." The New England Journal of Medicine. 2015;372;92-94.