Background. Because of altered airway microbiome in asthma, we analysed the bacterial species in sputum of patients with severe asthma. Methods. Whole genome sequencing was performed on induced sputum from non-smoking (SAn) and current or ex-smoker (SAs/ex) severe asthma patients, mild/moderate asthma (MMA) and healthy controls (HC). Data was analysed by asthma severity, inflammatory status and transcriptome-associated clusters (TACs). Results. α-diversity at the species level was lower in SAn and SAs/ex, with an increase in Haemophilus influenzae and Moraxella catarrhalis, and Haemophilus influenzae and Tropheryma whipplei, respectively, compared to HC. In neutrophilic asthma, there was greater abundance of Haemophilus influenzae and Moraxella catarrhalis and in eosinophilic asthma, Tropheryma whipplei was increased. There was a reduction in α-diversity in TAC1 and TAC2 that expressed high levels of Haemophilus influenzae and Tropheryma whipplei, and Haemophilus influenzae and Moraxella catarrhalis, respectively, compared to HC. Sputum neutrophils correlated positively with Moraxella catarrhalis and negatively with Prevotella, Neisseria and Veillonella species and Haemophilus parainfluenzae. Sputum eosinophils correlated positively with Tropheryma whipplei which correlated with pack-years of smoking. α- and β-diversities were stable at one year. Conclusions. Haemophilus influenzae and Moraxella catarrhalis were more abundant in severe neutrophilic asthma and TAC2 linked to inflammasome and neutrophil activation, while Haemophilus influenzae and Tropheryma whipplei were highest in SAs/ex and in TAC1 associated with highest expression of IL-13 Type 2 and ILC2 signatures with the abundance of Tropheryma whipplei correlating positively with sputum eosinophils. Whether these bacterial species drive the inflammatory response in asthma needs evaluation.

Introduction Real-world data on Omalizumab (OMA) and Mepolizumab (MEPO) can inform their use in severe asthma (SA). We studied patients in the Wessex AsThma CoHort of difficult asthma (WATCH) to: 1. Phenotypically compare OMA or MEPO treated patients against a SA, non-biologic group (SNB). 2. Assess clinical responses to OMA and MEPO. 3. Assess the spectrum of responses to these biologics. Methods We retrospectively phenotyped biologic naïve patients from WATCH (N=478) commenced on OMA (N=105) or MEPO (N=62) compared to SNB (N=178). Biologic response was gauged using standard criteria and response features were identified using logistic regression. Results OMA and MEPO patients were phenotypically distinct. Both drugs significantly reduced exacerbations, acute healthcare encounters (emergency department or hospital admissions), maintenance oral corticosteroid dose, and improved Asthma Control Questionnaire 6 (ACQ6) scores. OMA patients with more exacerbations at baseline (P=0.024), less acute healthcare encounters (P=0.050), and no anxiety (P=0.008) were more likely to respond to it. Lower baseline ACQ6 was independently associated with higher odds of MEPO response (P=0.007). Combined (OMA or MEPO) non-responders had significantly more psychological co-morbidities and worse baseline subjective disease markers compared to responder groups. Current criteria used to measure trial outcomes for MEPO, but not OMA, missed some modalities of response. Conclusion In a difficult asthma cohort, OMA and MEPO were used for distinct SA phenotypes, yet both were multidimensionally efficacious. Among these phenotypes, some clinical features associated with response were identified which emphasized the importance of addressing treatable traits when considering biologic therapy.