Discontinuing long-acting muscarinic antagonists (LAMA) or inhaled corticosteroids (ICS) may more than double the risk of moderate to severe COPD exacerbations during the 90 days following their discontinuation. The increased risk was greater than in later follow-up periods, suggesting a temporary withdrawal effect, according to a new analysis.1
Drug withdrawal effects have been documented following ICS cessation; however, this has not been previously documented following LAMA treatment cessation. Patients with COPD who are prescribed inhaled treatments miss about half their doses. The new findings suggest that missing doses may increase the risk of subsequent exacerbations.1
“Normally, when we stop a medication, we expect the treatment benefit to disappear and patients to return to their baseline risk of exacerbations or to a new baseline if their disease has progressed. However, some medications have withdrawal effects, where the risk temporarily rises above what we would expect from simply stopping the treatment,” said Alexander Mathioudakis, MD, PhD, Senior Clinical Lecturer in Respiratory Medicine at the University of Manchester, Manchester, United Kingdom. “We wanted to test whether stopping LAMA or ICS causes a temporary spike in exacerbations beyond what we would expect from simply losing the treatment benefit. And that is exactly what we observed. The effect was more pronounced than I had expected.”
Dr. Mathioudakis was the lead author on a post hoc analysis of the FLAME trial, which compared long-acting β-2 agonist (LABA) plus LAMA with LABA plus ICS. FLAME included 3,362 patients with moderate to severe COPD and a history of exacerbations.1
Patients who discontinued LAMA had a rate ratio of 2.2 for moderate to severe exacerbations (95% CI, 1.2-4.1; P = .001) during the first quarter after discontinuation compared with later periods of follow-up.
Dr. Mathioudakis said the observation was not confirmed for severe exacerbations, most likely due to a low event count. However, the data showed discontinuation of ICS was associated with a significant increase in severe exacerbations (P = .023) and a numerical increase in moderate to severe exacerbations.
“As health professionals, we need to be aware of treatment withdrawal effects,” Dr. Mathioudakis said. “We would very rarely stop LAMA in a patient with COPD once we start it clinically. That only happens if they have significant side effects, which is rare.”
ICS discontinuation, on the other hand, can be problematic.
Dr. Mathioudakis said clinical guidelines generally recommend ICS for patients with a history of exacerbations and elevated blood eosinophils. However, clinicians often see patients who started ICS without an exacerbation history or eosinophilic signal, and inappropriate ICS use can lead to infective exacerbations and pneumonia.
Clinicians should be wary of starting ICS without the appropriate indications to avoid unnecessary risk and minimize the need for later discontinuation, he said. If ICS discontinuation becomes appropriate, clinicians should also expect an increased risk of exacerbations for the 90 days following discontinuation. Exacerbations during that limited period do not necessarily mean treatment has failed and ICS treatment should resume.
Clinicians should also emphasize the benefits of remaining on inhaled medications.
“We often consider it our duty to start the right treatment, advise patients to take it, and then assume it is their responsibility to keep taking it,” Dr. Mathioudakis said. “But we have more responsibilities than that. We need to explain the importance of treatment adherence and the potentially significant risks of discontinuation. We often see patients who have an exacerbation, start their inhalers, improve and feel well, and then stop taking them or forget, only to have another exacerbation shortly afterward. It is important to proactively explain these risks.”
Dr. Mathioudakis said researchers should also consider the effects of treatment withdrawal in clinical trials. Patients entering trials frequently stop or change their existing drug regimens during a washout period. About 40% of patients recruited for the triple therapy IMPACT and ETHOS trials were receiving triple therapy at baseline, and many were assigned to LABA-LAMA or LABA-ICS at randomization, he noted. Potential treatment withdrawal effects are not being adequately considered when assessing trial results.
“Interpreting the overall results of a trial is one thing,” he said. “Looking at how an intervention works in subgroups is even more complicated, and withdrawal effects can introduce bias and have a much more significant effect… We need to develop and use appropriate methodology to identify withdrawal effects early and see how they affect practice more generally, not just for [patients with COPD].”
References
1. Mathioudakis AG, Bate S, Chatzimavridou-Grigoriadou V, et al. Disproportionate increase in COPD exacerbation risk for 3 months after discontinuing LAMA or ICS: insights from the FLAME trial. Thorax. Published online December 15, 2025. doi:10.1136/thorax-2025-223282
