S37 A novel computational image analysis method gives comprehensive insight into airway remodelling and resolution in the ovalbumin model of asthma

Background The pathophysiological changes occurring during the onset and progression of airway remodelling in asthma are intensively researched. However, protocols examine only a small fraction of airways and resolution of remodelling is barely examined. A mathematical model developed by our group suggests that rate of inflammation resolution post-exacerbation is a critical determinant of long term airway remodelling.1 We therefore sought to assess remodelling in unprecedented detail and its resolution across a large number of airways in a well-characterised model of chronic ovalbumin (OVA)-induced airway remodelling.2 Methods 43 BALB/C mice underwent OVA sensitisation and 10 OVA or PBS control airway challenges over 34 days. Animals were sacrificed on days 34, 35, 37, 39 and 41, and their lungs analysed by immunohistochemistry and differential bronchoalveolar lavage cell counts. A custom MATLAB program was developed to quantify airway size, airway smooth muscle (ASM) and total collagen area fractions relative to basement membrane perimeter for approximately 30–80 airways/mouse lung. Results At maximal remodelling (day 34), ASM and collagen area fractions increased in OVA challenged mice compared to controls (ASM 0.15±0.07 vs 0.08±0.07, p To determine how remodelling varies with airway size, airways were categorised by epithelial basement membrane perimeter. ASM and collagen area fraction increased in all airways. Large airways (1000 mm–1500 mm) had the largest increase followed by medium (500 mm–1000 mm), then small ( Conclusion The technique allows, for the first time, comprehensive assessment of airway remodelling and reveals heterogeneity in remodelling, and resolution across airway sizes. The different anatomical distributions and time-courses of components of the remodelling response has implications for use of mouse models and future prevention of airway remodelling. References Biomech Model Mechanobiol2018. https://doi.org/10.1007/s10237-018-1037-4 J Immunol2011. https://doi.org/10.4049/jimmunol.1003507