A microsimulation model has quantified what clinicians have long suspected: sustained steroid exposure in autoimmune disease carries a measurable, and avoidable, clinical and economic burden that dose metrics alone cannot capture.
Prolonged steroid use is standard of care across a wide range of autoimmune conditions. Yet the true clinical and economic cost of that exposure has remained poorly characterized. Rigorous, long-term studies of steroid-associated adverse event burden have been sparse, leaving drug developers, payers, and health economists reliant on assumptions rather than evidence.
Research presented at ISPOR Europe 2025 moves the story from supposition to science.[1]
What the model did
The team built a microsimulation model to estimate the lifetime clinical and economic burden of adverse events associated with long-term, high-dose steroid use, benchmarked against a hypothetical steroid-sparing treatment with equivalent disease control. The model was applied across four autoimmune diseases: myasthenia gravis (MG), chronic inflammatory demyelinating polyneuropathy (CIDP), lupus nephritis, and immune thrombocytopenia. Detailed findings were reported for MG.
The model incorporated 13 adverse events: six long-term (including osteoporosis, cardiovascular disease, diabetes, depression, glaucoma, and adrenal insufficiency) and seven short-term complications. Per-cycle risk was modeled against current and cumulative dose, using published hazard ratios and a five-year dose-accumulation window. All costs were adjusted to 2024 US dollars.
The numbers
In the base case MG cohort of 3,000 simulated patients (mean age 55.4 years, average maintenance dose 15.9mg), a 50% reduction in steroid dose produced:
- A gain of 0.32 life years and 0.56 quality-adjusted life years (QALYs)
- A reduction in lifetime adverse event management costs of $24,915 per patient
Scenario analyses amplified this picture considerably. Limiting steroid use to six months resulted in savings of $38,479 and a QALY gain of 1.24; eliminating steroid exposure entirely generated cost savings of $41,368 per patient and a QALY gain of 1.41.
Sensitivity analyses identified the three model inputs with the greatest influence on cost differences: adverse event incidence hazard ratios by steroid dose, mean patient age, and adverse event-related mortality hazard ratios. Younger patients and female-only cohorts showed higher incremental benefits, driven by longer lifetime exposure.
Why age and cumulative dose matter most
The modeling makes explicit what clinical intuition suggests: patient age at disease onset and cumulative steroid dose are the primary determinants of long-term steroid-toxicity burden. The longer a patient is exposed, the greater the compound harm. This has direct implications for when steroid-sparing strategies are introduced.
Waiting until steroid-toxicity becomes clinically apparent is already too late to prevent patient harm.
The measurement gap this exposes
The new model is a significant contribution to health economic evidence for steroid-sparing strategies. It also underscores a limitation that the field is only beginning to address: the projected cost savings from dose reduction are merely inferred. They are estimates derived from adverse event rates linked to exposure levels, not direct measurements of harm in individual patients.
Steroid-toxicity manifests differently across patients receiving identical regimens. Individual susceptibility, baseline comorbidity, and accumulated prior exposure all shape the actual harm experienced. Modeling can quantify the population-level burden; it cannot, by design, distinguish who in that population bears the cost.
Implications for drug development and HEOR
For sponsors developing steroid-sparing therapies across MG, CIDP, lupus nephritis, or immune thrombocytopenia, this model provides a credible economic argument for value communication. The consistency of findings across all four disease areas strengthens the generalizability of the steroid-sparing case.
The work presented at ISPOR Europe 2025 reinforces that steroid-sparing is not simply a clinical convenience. It is a driver of meaningful quality-of-life gains and a source of substantial, quantifiable cost reduction. The next step in drug development is to translate population-level evidence into patient-level measurement, moving from modeled burden to direct measures of steroid-toxicity.
Reference
-
Stone, J.H, Neumann, P., Narayanaswami, P., et al. (2025, November 9–12). Microsimulation Model to Estimate the Clinical and Cost Burden of Adverse Events Related to Long-Term Oral Corticosteroid Usage in Autoimmune Diseases in the United States Poster. ISPOR Europe 2025, Glasgow, UK