Withdrawal of Inhaled Glucocorticoids and Exacerbations of COPD

Magnussen H, et al. for the WISDOM Investigators.

N Engl J Med. 2014 Oct 2;371(14):1285-94. doi: 10.1056/NEJMoa1407154. [Full text]

Combination inhalers, and the evidence for them in COPD is a crowded (and lucrative) space (see for example ETHOS, FULFIL, TRILOGY, TRIBUTE). We know that in COPD most therapies are for symptom management and to attempt to decrease exacerbations, but they may not have much impact on disease progression or mortality.

When WISDOM was published in 2014, Long-acting muscarinic antagonists (LAMAs) were relatively new and had been shown to reduce exacerbations. LAMAs potentially have a better safety profiled than inhaled corticosteroids (ICS). With this in mind WISDOM sought to see if patients with severe to very severe COPD on triple inhaled therapy (ICS+LABA+LABA) could be stepped down off of ICSs without increased exacerbation risk via a non-inferiority study over 1 year [1].

As noted there have been many major trials and new combined triple therapy inhalers introduced since WISDOM. We will review other studies in future 52in52s and this study may help put those in context

Patient Population

A total of 2,485 adult current or former smokers (≥10 pack-years) with severe / very severe COPD (as defined by post-bronchodilator FEV1 < 50% and FEV1 <70% of FVC) that had at least 1 exacerbation in last 12 mo were recruited from 200 centers across 23 countries. Most were male (~80%) and were GOLD3 with a mean FEV1 of about 34. Key exclusion criteria were a:

  • “Significant diseases other than COPD”
  • History of thoracotomy
  • Asthma dx on ICS
  • On daytime home O2 (>1 hr / day)
  • Chronic PO steroids use (>5 mg /day)
  • Recent MI, HF exacerbation, or arrhythmia
  • URI or recent COPD exac (inclusion delayed by 6 wk)
Study Design

There was a 6 week run-in period in which all patients were on Tiotoprium 18 mg daily, Salmeterol 50 mg BID, and Fluticasone 500 mg BID. Then 1,242 patients in the intervention arm stepped down their ICS every 6 weeks as follows: total 1000 mg -> 500 -> 200 -> 0 (placebo), such that they were off ICS at 18 weeks.


The primary outcome was the time to a first moderate to severe COPD exacerbation. A moderate exacerbation was defined as an increase in lower respiratory track symptoms or ≥ 2 new COPD related symptoms for at least 3 days requiring antibiotics and/or oral GCs.

Important secondary outcomes included the change in baseline FEV1 over time and symptom assessment, for example with St. George’s Respiratory Questionnaire (SGRQ). Adverse events were tracked, including the incidence of penumonia and cardiovascular events.


Over the year long study period, the time to first moderate to severe exacerbation was similar between groups, HR 1.06 (CI 0.94-1.19, P0.35). This was the primary outcome. The secondary outcomes suggest that over that year, FEV1 and symptom control seemed to suffer. By 18 weeks the worsening of FEV1 in the ICS withdrawal group was lower statistically, and at 52 weeks it had dropped by 39 (p = 0.001). Similarly, at 52 weeks symptoms as measured by the SGRQ were worse in the ICS withdrawal group (+1.15 vs. -0.07, p = 0.047). Additionally, when looking at severe exacerbations at a year there was a trend to worsening HR 1.20 (0.98–1.48, P=0.08). There were no difference in the rates of adverse events between the groups, including similar frequency of hospitalization, death, pneumonia or cardiovascular event.


While the study showed non-inferiority at one year regarding no change in time to first exacerbation it has been criticized that the follow up period was too short. On average COPD exacerbation happen every 1.3-2 years [2], hence, 1 year may not adequately capture the “next” exacerbation, Additionally, there seemed to be a trend in the difference in time to first severe exacerbation in the ICS withdrawal group.

Importantly, since the goal of inhaled therapy is to reduce exacerbation AND aid in symptom management. It is troubling that FEV1 declined and symptoms were reported more poorly controlled in the ICS withdrawal group. Based on this study it is likely that the decision to continue or taper ICSs should be symptom drive.

As noted, since the publication of WISDOM there have been numerous trials looking at combined triple therapy that will be explored in a future review.

  1. Reilly JJ. Stepping down therapy in COPD. N Engl J Med. 2014 Oct 2;371(14):1340-1. doi: 10.1056/NEJMe1409219.
  2. Cosio M, Baraldo S. Inhaled Glucocorticoids and COPD Exacerbations. N Engl J Med 2015; 372:92-94. doi: 10.1056/NEJMc1413308