Reviews/COPD Pharmacotherapy/Inhaled corticosteroid withdrawal and exacerbations
Ics Risk · COPD-ICS-03 · v2
Borderline: within 4 points of the Doubtful/Refuted boundary, so a re-grade could land either side.
The strongest evidence here comes from trials that sorted COPD patients by their blood eosinophil levels before stopping inhaled steroids. In those trials ([1], [2]), stopping the steroid raised flare-up rates only in patients with high eosinophil counts and not in those with low counts, and a further analysis of the same trial data ([4]) found the same pattern. A very large real-world study ([5]) found no overall rise in severe flare-ups after stopping. One trial analysis ([7]) did report an early, short-lived rise in severe flare-ups that it said did not depend on eosinophil level, but this was a secondary finding based on few events. Weighing the direct, eosinophil-sorted trial evidence against that single fragile signal, the panel concluded the claim does not hold for low-eosinophil patients.
In [1] the moderate or severe exacerbation rate rose after withdrawal only at eosinophil counts of 2% or greater (rate ratio 1·22 [95% CI 1·02-1·48]), 4% or greater (1·63 [1·19-2·24]) and 5% or greater (1·82 [1·20-2·76]), an effect 'not seen in most patients' below these thresholds. In [2] the annualized rate of moderate or severe exacerbations did not differ between treatments (rate ratio, 1.08; 95% confidence interval, 0.83 to 1.40), with the higher risk confined to patients with ≥300 blood eosinophils/μl. [4] found moderate/severe exacerbation risk was not increased overall (HR 1.05; 95% CI 0.89-1.25), rising with eosinophil count only among those with a history of frequent exacerbations. [5], a matched cohort of 42,667 discontinuers, reported a severe-exacerbation hazard ratio of 0.86 (95% CI 0.78-0.95) on discontinuation. Only [7] supplied a contrary signal: a significant early rise in severe exacerbations after ICS discontinuation (p=0.023), said to be consistent regardless of baseline blood eosinophil count, though the moderate-to-severe difference did not reach statistical significance. The contested point: The division is [7] against the rest: it reports an early rise in severe exacerbations after ICS withdrawal that it describes as consistent regardless of baseline eosinophil count, directly countering the eosinophil-stratified trials [1], [2] and [4]. The panel read this as a transient, severe-only, post-hoc signal resting on acknowledged low event counts, with a non-significant moderate/severe endpoint, and judged it unable to overturn the stratified evidence.
All 11 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from COPD Pharmacotherapy, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The two most direct eosinophil-stratified RCT analyses ([1][2]), reinforced by [4], locate the withdrawal-exacerbation effect in the high-eosinophil stratum and report its absence below threshold; [1] states the deleterious effect is 'not seen in most patients' with low counts. The claim directly contradicts the strongest primary evidence.
[7] is the only support and is fragile: post-hoc, the moderate/severe exacerbation endpoint did not reach significance, the severe-exacerbation signal had acknowledged low event counts, and the effect was transient (first quarter only). A transient, severe-only signal does not establish that withdrawal 'increases exacerbations' as a durable property in low-eos COPD.
Where it remains uncertain. [7] reports a significant early rise in severe exacerbations after ICS discontinuation (p=0.023) said to be consistent regardless of baseline eosinophil count, cutting against the eosinophil-stratified trials.
Aleatory
The severe-exacerbation signal in [7] rests on acknowledged low event counts, so chance variation could inflate or mask it.
The load-bearing papers from a corpus of 11, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
No overall exacerbation increase on withdrawal; excess risk confined to >=300 eos, so no increase in low-eos.
“The higher exacerbation risk in patients with ≥300 blood eosinophils/μl suggests that these patients are likely to benefit from triple therapy.”
Panel auditInternal validity: medium · scope of inference: Moderate-to-severe COPD without frequent exacerbations on long-term triple therapy.Key limitations: non-frequent-exacerbator population · low-eos subgroup result not reported separately · exacerbations a secondary endpoint
Large real-world cohort shows no net severe-exacerbation harm from discontinuation overall.
“The hazard ratio of a severe exacerbation with ICS discontinuation relative to ICS continuation was 0.86 (95% CI: 0.78-0.95)”
Panel auditInternal validity: medium · scope of inference: Real-world UK CPRD COPD patients aged ≥50 who initiated LAMA-LABA-ICS triple therapy.Key limitations: no eosinophil stratification · observational, residual confounding · channelling of sicker patients
Sole supporting signal: suggests a withdrawal effect that does not spare low-eosinophil patients.
“discontinuing ICS was associated with a significant early rise in severe exacerbations (p=0.023), though the difference for moderate-to-severe events did not reach statistical significance.”
Panel auditInternal validity: medium · scope of inference: Moderate-to-severe COPD with exacerbation history in FLAME, stratified by baseline therapy use.Key limitations: post-hoc, non-randomised discontinuation · eosinophil consistency stated qualitatively, not quantified · moderate-to-severe endpoint not significant
Directly against the claim: in the low-eosinophil stratum ICS withdrawal did not increase exacerbations.
“Our data suggest that counts of 4% or greater or 300 cells per μL or more might identify a deleterious effect of ICS withdrawal, an effect not seen in most patients with eosinophil counts below these thresholds.”
Panel auditInternal validity: medium · scope of inference: Severe-to-very-severe COPD with a history of exacerbations, on triple therapy before randomised ICS withdrawal.Key limitations: post-hoc subgroup analysis · severe exacerbator population only · low-eos effect described qualitatively, not quantified
4 of 11 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Predominant direct evidence shows no exacerbation increase in the low-eosinophil stratum, with one fragile contrary post-hoc signal.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
No prospective low-eosinophil withdrawal RCT. The low-eos conclusion rests on the absence of effect below threshold in [1], [2] and [4], not on a positively-defined arm. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A randomised withdrawal-versus-continuation trial enrolling only patients with low baseline blood eosinophils, powered for moderate and severe exacerbations.
Open question: Does ICS withdrawal increase exacerbations in a prospectively defined low-eosinophil COPD population? Why it is still open: Every eosinophil-stratified result in the corpus ([1], [2], [4]) infers the low-eos answer from the absence of effect below threshold rather than testing a dedicated low-eos arm.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19