Design: matching-adjusted indirect treatment comparison reweighting individual patient data from three trials against aggregate data from two
Population: patients with COPD, blood eosinophils 300 cells/microlitre or above, mMRC grade 2 or above, moderate-to-severe airflow limitation and chronic bronchitis
Primary outcome: annualised rate of moderate or severe exacerbations, and St George's Respiratory Questionnaire for COPD
Effect: rate ratio 0.91 (95% CI 0.60-1.37) by investigator-assessed chronic bronchitis and 1.13 (0.80-1.58) by questionnaire definition; no significant difference in quality of life

Both mepolizumab and dupilumab are licensed as add-on maintenance for inadequately controlled COPD with an eosinophilic phenotype, and they have never been compared head to head. This analysis reweighted individual patient data from three mepolizumab trials — METREX, METREO and MATINEE — to match the aggregate baseline characteristics of the two dupilumab trials, BOREAS and NOTUS, and compared the results.

The annualised rate of moderate or severe exacerbations showed no significant difference, with a rate ratio of 0.91 using the investigator-assessed chronic bronchitis definition and 1.13 using the questionnaire-based one. Quality of life on the St George's Respiratory Questionnaire likewise showed nothing between them.

The useful implication is about how the choice should be made. In the absence of a difference in efficacy, the deciding factors become the ones clinicians often treat as secondary: dosing interval, injection route and training, the patient's other type 2 conditions, and cost. Dupilumab's licensed indications extend beyond the airway; mepolizumab's do not in the same way, and for a patient with coexisting atopic dermatitis or nasal polyposis that is a real consideration.

The method is worth naming plainly because it constrains what can be concluded. This is an indirect comparison, not a trial. It is the best available evidence on the question and it is not the same kind of evidence as a randomised head-to-head.

Choose between these agents on dosing, coexisting type 2 disease and cost rather than on expected exacerbation reduction
Confirm the eosinophil threshold used locally; this analysis restricted to counts of 300 cells/microlitre or above
Check for atopic dermatitis, nasal polyposis or asthma before choosing, since one agent may cover more than one problem
Set an exacerbation-rate expectation with the patient before starting, and review at 12 months against it
Record which chronic bronchitis definition your service uses; the two definitions gave opposite point estimates here

Don't overread it

Absence of a detected difference in an indirect comparison is not evidence that the two drugs are equivalent.

The statistics, in plain English

Both rate ratios have confidence intervals that comfortably include 1.0 — 0.60 to 1.37 and 0.80 to 1.58 — and those intervals are wide, which matters more than the point estimates. A range compatible with a 40% advantage for one drug or a 58% advantage for the other is not a demonstration of equivalence; it is a failure to detect a difference. A matching-adjusted indirect comparison also depends on the assumption that all effect-modifying baseline characteristics have been measured and matched, which cannot be verified.