An eosinophil count that comes and goes still identifies a treatable COPD phenotype

Design

Pooled analysis of three phase 3, randomised, double-blind, placebo-controlled trials (METREX NCT02105948, METREO NCT02105961, MATINEE NCT04133909)

Population

Patients with COPD receiving mepolizumab 100 mg or placebo, subgrouped by blood eosinophil counts in the 12 months before randomisation, at screening and at baseline

Primary outcome

Annualised rate of moderate or severe exacerbations, and of exacerbations needing an emergency visit or admission

Effect

21% reduction with any count 300 or above pre-randomisation; 12-27% with persistently raised counts; 22-36% with variable counts; no benefit persistently below 150 cells/microlitre

Blood eosinophil count guides biologic selection in COPD, and it moves. The practical consequence has been a habit of demanding a persistently raised count before treating, on the assumption that a fluctuating one represents noise rather than a phenotype. This pooled analysis of the phase 3 METREX, METREO and MATINEE trials tested that assumption directly.

Patients on mepolizumab 100 mg or placebo were grouped by their eosinophil counts in the 12 months before randomisation, at screening and at baseline. Moderate or severe exacerbations were reduced, or trended down, wherever type 2 inflammation was present: a 21% reduction in anyone with a count of 300 or more at any pre-randomisation time point, 12-27% with persistently raised counts, and 22-36% with counts that varied over time. Across subgroups from 150 cells per microlitre upwards, mepolizumab reduced exacerbations needing an emergency visit or admission. Only patients whose counts stayed below 150 gained nothing.

The fluctuating group did at least as well as the persistent group, which is the finding. Intermittent elevation is not a failed attempt at a stable phenotype; it is the phenotype, seen at different moments.

So change what you do with the record. Instead of waiting for two consecutive high counts, look back at every count in the last twelve months and treat a value of 300 or more at any point as evidence of type 2 inflammation. The corollary is equally useful: a patient whose counts have consistently sat below 150 is unlikely to benefit, and that is a defensible reason not to start an expensive drug - which in Indian practice is the decision that actually needs the evidence.

• Review all eosinophil counts from the past 12 months, not just the most recent one
• A count of 300 or more at any point supports type 2 inflammation, even if later counts are low
• Consistently below 150 across time points predicts no benefit - use it to avoid futile treatment
• Exacerbations needing emergency care fell across subgroups from 150 upwards
• This is a pooled post-hoc subgroup analysis, so it supports selection, it does not redefine a licence

The statistics, in plain English

These are subgroup analyses pooled across three trials, so they inherit each trial's randomisation but not its protection against chance findings: the more subgroups examined, the more likely one looks impressive by luck. Reductions described as 'trending toward' were not statistically significant, and the ranges quoted (12-27%, 22-36%) are spans across several subgroups rather than confidence intervals. The clearest result is the negative one - no benefit below 150 cells per microlitre - because it is consistent across every way the data were cut.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.