P173 Clinical utility of sputum cell count in severe asthma

<h3>Background</h3> Sputum cell count can be used for phenotyping airway inflammation. However, its clinical use remains limited due to technical challenges and limited evidence for its role in severe asthma. At present, sputum analysis is predominantly used as a research tool. In this context, we developed a clinical sputum service for patients with severe asthma. <h3>Aim</h3> Assess performance of a clinically delivered sputum cytology service in phenotyping airway inflammation and guiding clinical decision-making in severe asthma. <h3>Methods</h3> Patients with severe asthma uncontrolled on standard and biologic treatment provided supervised spontaneous sputum samples. Those who failed underwent sputum induction. Samples were processed and reported by hospital pathology department using standardised protocol. The asthma multidisciplinary team used results to guide treatment decisions. Phenotypic categories included eosinophilic (eosinophil count ≥3%), neutrophilic (neutrophil count ≥60%), mixed eosinophilic &amp; neutrophilic, or pauci-granulocytic. <h3>Results</h3> Of 126 patients, sputum quality was sufficient for differential count in 73 (57.9%): 46 (63%) female, mean age 52.1±16.2 years, mean blood eosinophil count 0.23±0.50 x10⁹/L, mean FeNO 36.52±38.91ppb. Spontaneous samples were obtained from 102 patients, of which 59 (57.8%) were successful. There were 24 sputum inductions performed, 14 (58.3%) of which worked. Of sputum tests that were successful (n=73), mean eosinophil sputum count was 9.38±19.11, and neutrophil count was 60.14±27.44. Samples were eosinophilic in 20.5% (n=15), neutrophilic in 35.6% (n=26), mixed eosinophilic and neutrophilic in 23.3% (n=17), and pauci-granulocytic in 20.5% (n=15). Phenotypic distribution amongst samples obtained from patients on biologics (n=34) are shown in table 1. Correlation between blood and sputum eosinophils (r=0.2258, p=0.0602), and FeNO and sputum eosinophils (r=0.2140, p=0.1036) was negligible. Following sputum results, 15% (n=10) commenced biologic treatment. Of those already on biologics, treatment was switched for 6 patients. Prophylactic antibiotics were started in 13 (19.4%), extended for 1 (1.9%), and stopped for 1 (1.9%) patient following sputum results. <h3>Conclusion</h3> Sputum analysis as part of clinical service was successful in 57.9% of samples, allowing airway phenotyping that guided management where alternative biomarkers proved inconclusive. Further standardisation of sputum analysis as part of clinical service, and greater exploration of inducible sputum analysis with larger patient dataset is required.