Biomarker profile and disease burden associated with intermittent and long-term oral corticosteroid use in patients with severe asthma prior to biologic initiation in real-life (STAR)

Background: Asthma characterization using blood eosinophil count (BEC) (among other biomarkers and clinical indices) is recommended in severe asthma (SA), but the masking effect of oral corticosteroids (OCS), makes this challenging. Aim: Our aim was to explore the effect of OCS use (both intermittent [iOCS] and long-term [LTOCS]) prior to biologic initiation on SA phenotype and biomarker profile in real-life and to characterize the burden of SA among patients prescribed LTOCS by biomarker profile. Methods: This was a registry-based cohort study, including data from 23 countries collected between 2003 and 2023 and shared with the Internatonal Severe Asthma Registry (ISAR). Patients with SA were categorized into 3 cohorts, those with: (i) no prescription for OCS, (ii) prescription(s) for iOCS (ie, ≤90 days in previous 12-months, usually short courses for exacerbations), and (iii) prescriptions for LTOCS (ie, >90 days in previous 12-months). Biomarker distribution (ie, BEC, fractional exhaled nitric oxide [FeNO], and total Immunoglobulin E [IgE]) were quantified in the year prior to biologic initiation in patients with SA according to OCS prescription pattern. Phenotypes were characterized for those prescribed LTOCS according to BEC cut-off (<150 and ≥ 150 cells/μL). Results: Of 4305 patients included, 5.0% (n = 215), 54.1% (n = 2330) and 40.9% (n = 1760) were prescribed no OCS, iOCS, and LTOCS, respectively. The BEC distribution varied by prescription pattern and LTOCS dose (<5 mg to ≥20 mg/day); BEC was <150 cells/μL in 28.6% (n = 369/1288) of LTOCS patients, compared to 19.5% (n = 284/1460) of iOCS patients and 14.0% (n = 21/150) of those in the no OCS group. Median BEC was also significantly lower in the LTOCS versus the iOCS group (310 vs 400 cells/μL; p < 0.001). A similar pattern was noted for IgE, but not FeNO. Among LTOCS patients with BEC <150 cells/μL, 39.9% experienced ≥4 exacerbations, 75.1% had uncontrolled asthma symptoms and 55.9% had evidence of persistent airflow obstruction (compared with 40.9%, 76.2% and 59.5% of those with BEC ≥150 cells/μL, respectively). Conclusions: OCS, whether prescribed intermittently or long term, affect BEC distribution potentially leading to heightened risk of phenotype misclassification and influencing subsequent treatment decisions. FeNO appears to be less susceptible to OCS-induced suppression. Disease burden was high for those in the LTOCS group and was high independent of dose and BEC. Our findings highlight the importance of considering OCS use, even intermittent use, when characterizing SA, and suggests the need for earlier phenotyping and alternative treatment strategies for LTOCS patients with low BEC.