AI-Driven Benefit-Risk Analysis for Severe Asthma

Our platform stratifies BRA by asthma phenotype and biomarker profile. For eosinophilic asthma (blood eos ≥300 cells/μL), we model anti-IL5 and anti-IL5R safety including eosinophil depletion consequences. For allergic asthma with elevated IgE, omalizumab anaphylaxis risk is quantified. For overlapping phenotypes, dupilumab and tezepelumab BRA accounts for broader T2 pathway modulation. Each phenotype receives tailored benefit metrics (exacerbation rate reduction, FEV1 improvement, OCS sparing) weighed against mechanism-specific adverse events.

Many severe asthma patients meet eligibility criteria for multiple biologics simultaneously (e.g., elevated eosinophils and elevated IgE). ArcaScience provides comparative BRA across eligible biologics using head-to-head data where available and network meta-analysis for indirect comparisons. The platform models class-specific risks—anaphylaxis for omalizumab versus helminth susceptibility for anti-IL5 versus transient hypereosinophilia for dupilumab—enabling clinically informed treatment sequencing decisions with quantified safety trade-offs.

Yes. Our platform includes pediatric-specific BRA models for omalizumab (approved ≥6 years), mepolizumab (≥6 years), dupilumab (≥6 years), and benralizumab (≥12 years). Pediatric considerations include growth impact from systemic corticosteroid reduction, age-appropriate injection-site reaction modeling, immune system development effects of eosinophil depletion in children, and long-term safety extrapolation where pivotal pediatric data remains limited. We integrate data from LIBERTY ASTHMA EXCURSION, MUPPITS-2, and other pediatric severe asthma trials.

Oral corticosteroid reduction is a critical benefit endpoint in severe asthma biologics. ArcaScience models cumulative OCS burden reduction alongside the steroid-associated adverse event profile being avoided—including osteoporosis, adrenal insufficiency, diabetes, cataracts, and immunosuppression. We quantify the benefit of OCS reduction using data from SIRIUS, VENTURE, and PONENTE trials, integrating this with adrenal axis recovery modeling during tapering and the risk of disease exacerbation at each step-down dose. This provides a net benefit-risk score that captures both direct biologic safety and indirect OCS reduction benefit.

Severe asthma biologics frequently show different benefit-risk profiles in real-world use versus clinical trials, due to patient heterogeneity, comorbid conditions, variable adherence, and phenotype misclassification. ArcaScience integrates registry data (ISAR, CHRONICLE, SANI), claims databases, and post-marketing surveillance alongside pivotal trial data from DREAM, MUSCA, QUEST, and NAVIGATOR. Our models adjust for selection bias, phenotype enrichment in trials, and the super-responder effect to provide BRA estimates that reflect real-world clinical practice rather than trial-optimized populations.