Mepolizumab Beat Omalizumab on Asthma Attacks in Real-World Head-to-Head Cohort

Patients with severe asthma who qualify for two different biologic drugs may do better on one of them, according to a new retrospective cohort study published in Allergy, Asthma & Clinical Immunology. Japanese researchers found that mepolizumab, an antibody that blocks interleukin-5, cut clinically significant asthma exacerbations by 63 percent compared with omalizumab, an antibody that mops up immunoglobulin E, during the first year of treatment. The finding matters because a meaningful share of patients with severe asthma carry the biomarker profile that makes them eligible for both drugs, and until now there has been no head-to-head randomized trial telling doctors which agent to choose.

The study, led by Tatsuya Nagai and colleagues at Kameda Medical Center, enrolled 49 patients: 25 treated with omalizumab and 24 treated with mepolizumab. To qualify, every patient had to meet the eligibility thresholds for both therapies simultaneously. That meant a blood eosinophil count of at least 150 cells per microliter before starting treatment, or at least 300 cells per microliter at some point in the prior year, along with a total serum IgE level of at least 30 international units per milliliter and documented sensitization to at least one perennial inhalant allergen such as dust mite, mold, animal dander, moth, or cockroach. In other words, these were patients whose disease showed both the eosinophilic inflammation that mepolizumab targets and the allergic signature that omalizumab was designed to treat.

The researchers chose a primary endpoint that clinicians care about most: the rate of clinically significant asthma exacerbations during the first year on treatment, adjusted for key baseline differences between the groups. An exacerbation counted as clinically significant if it required systemic steroids for at least three days, a hospitalization, or an emergency room visit needing systemic corticosteroids. Exacerbations occurring within 14 days of each other were counted as a single event to avoid inflating the numbers. The team used Poisson regression models with a log-link function, adjusting for body mass index, blood eosinophil count, and the number of exacerbations in the previous year, and they accounted for variable follow-up times by including the natural logarithm of the treatment period as an offset term.

The headline result was striking. Mepolizumab reduced the annual exacerbation rate by 63 percent relative to omalizumab, with an adjusted incidence rate ratio of 0.37 and a 95 percent confidence interval of 0.19 to 0.69. In subgroup analyses stratified by baseline eosinophil count, the benefit was statistically significant in patients whose counts were below 300 cells per microliter, where the incidence rate ratio was 0.38 with a confidence interval of 0.14 to 0.88. In the subgroup with counts of 300 or above, the point estimate still favored mepolizumab at 0.47, but the confidence interval of 0.15 to 1.35 crossed unity, meaning the difference could not be distinguished from no effect with the available sample size.

Not every outcome favored mepolizumab. There were no significant between-group differences in emergency room visits or hospitalizations, and the adjusted incidence rate ratios for those secondary endpoints came with very wide confidence intervals, 0.23 to 11.90 for ER visits and 0.03 to 6.49 for hospitalizations, reflecting the small number of events. Both drugs did reduce oral corticosteroid doses significantly within their own groups. In the mepolizumab group, the mean daily dose fell from 6.54 milligrams at baseline to 1.62 milligrams, while in the omalizumab group it dropped from 10.55 milligrams to 4.05 milligrams. But the magnitude of that steroid-sparing effect did not differ significantly between the two drugs, with a p value of 0.473 for the between-group comparison.

The authors offer several possible explanations for why mepolizumab appeared to win, and they are careful to frame the result as hypothesis-generating rather than practice-changing. One key point is that blood eosinophil counts fluctuate over time and do not always mirror what is happening in the airway tissue. Roughly half of patients with severe asthma show eosinophilic airway inflammation, and mepolizumab targets interleukin-5, the cytokine that drives eosinophil growth and differentiation, so it may still help patients whose blood counts have temporarily dipped below the 300-cell threshold. The authors also note that some former smokers in the cohort may have had unrecognized chronic obstructive pulmonary disease or asthma-COPD overlap, conditions in which eosinophilic inflammation is a treatable trait that mepolizumab can address.

Another layer of interpretation involves omalizumab itself. The predictors of response to omalizumab remain poorly established, and total serum IgE levels turn out to be unreliable indicators of who will benefit. Post-hoc analyses have suggested that a baseline eosinophil count of at least 260 cells per microliter predicts a good response to omalizumab, and recent biomarker-focused literature argues that type 2 inflammatory markers such as blood eosinophils and fractional exhaled nitric oxide may predict responsiveness to anti-allergic therapies better than IgE alone. That could explain why omalizumab performed relatively better in the high-eosinophil subgroup, narrowing the gap with mepolizumab. The authors also point to baseline imbalances: the omalizumab group had significantly more exacerbations and ER visits in the prior year, suggesting worse disease severity at the start.

The study has important limitations that temper how far the findings can be generalized. It was a small, single-center retrospective analysis with no formal sample size calculation, which limits statistical power. The retrospective design means selection bias cannot be ruled out, and asthma diagnoses rested on the clinical judgment of treating pulmonologists rather than on systematically performed bronchodilator reversibility or methacholine challenge testing. The study period spanned different treatment eras, and because omalizumab became available in Japan before mepolizumab, differences in the timing of treatment may have influenced which drug patients received. Some patients also discontinued the index biologic before one year, five in the omalizumab group and three in the mepolizumab group, and although follow-up differences were handled with an offset term, heterogeneity in treatment persistence may still have affected the results. Residual confounding remains possible despite adjustment for eosinophil count, body mass index, prior exacerbations, and prior healthcare use.

There are also quirks in the patient population worth noting. The cohort included patients with total IgE levels above 1500 international units per milliliter, which falls outside the typically approved dosing range for omalizumab, potentially limiting generalizability to populations managed under strict prescription guidelines. The definition of perennial allergen sensitization relied on a specific IgE threshold of 0.35 kilounits per liter, which may have captured patients with clinically minor sensitization. Fractional exhaled nitric oxide, a biomarker increasingly discussed in biologic selection, was available in only four patients, precluding any meaningful analysis of its role. Documented COPD was present in three patients in each group, and bronchiectasis in one patient in each group, adding further clinical complexity to the comparison.

What the study does offer is preliminary real-world evidence in a clinically relevant subgroup that randomized trials have largely ignored. Previous indirect comparisons found no significant differences in major exacerbations or hospitalizations between the two drugs, and one prospective study restricted to patients with eosinophil counts of 300 or above did find mepolizumab superior, but that population may have been biased in mepolizumab’s favor. By including patients with lower eosinophil counts, this analysis suggests the benefit may extend more broadly than that earlier work implied. The authors conclude that mepolizumab may be more effective than omalizumab for reducing clinically significant exacerbations in patients eligible for both, but they stress that larger, multicenter, prospective, biomarker-stratified head-to-head trials are needed before these results should guide routine biologic selection. Until such trials arrive, the choice between these two antibodies for overlap-eligible patients remains an open question that this small study helps to sharpen rather than settle.

Subject of Research: Comparative real-world effectiveness of mepolizumab versus omalizumab in severe asthma patients eligible for both biologic therapies

Article Title: Comparative effectiveness of mepolizumab and omalizumab in patients eligible for both: a retrospective cohort study

Article References: Nagai, T., Inoshima, N., Matsui, H., Kubota, N., Funaki, Y., Otsuki, A., Ito, H., & Nakashima, K. (2026). Comparative effectiveness of mepolizumab and omalizumab in patients eligible for both: a retrospective cohort study. Allergy, Asthma & Clinical Immunology, 22(1), Article 56. https://doi.org/10.1186/s13223-026-01055-9

Image Credits: AI Generated

DOI: 10.1186/s13223-026-01055-9

Keywords: severe asthma, mepolizumab, omalizumab, biologic therapy, asthma exacerbations, blood eosinophil count, immunoglobulin E, retrospective cohort study, oral corticosteroids, incidence rate ratio, type 2 inflammation, real-world evidence