Cluster Analysis of Allergic Poly-Sensitizations in Urban Adults with Asthma

Introduction While reliable, quantitative in vitro testing for sensitivity to aeroallergens has been available for decades, if and how asthma severity markers might be predictably expressed in clusters matched for comparable multiple sensitizations is unknown. Objective Our aim is to use machine learning techniques to explore how allergic poly-sensitization (APS) clusters may serve as precision markers in adult urban patients with moderate to severe asthma. Methods We constructed a database of sensitizations to the 25 aeroallergens in the Zone 1 Northeastern US ImmunoCAP® assay. We used the Scikit-Learn® machine learning library to perform model-based clustering to identify APS clusters. Clusters were compared for differences in common clinical markers of asthma. Results The database consisted of 509 patients. Unbiased mixture modeling identified ten clusters of increasing APS of varying size (n = 1 to 339) characterized by significant increases in mean serum immunoglobulin E (p<.001), peripheral blood eosinophil count (p<.001), and D LCO (p=.02). There was a significant decline in mean age at presentation (p<.001), FEV 1 /FVC (p=.01), and FEF 25-75 (p=.002), but not FEV 1 (p=.29), nor RV/TLC (p=.14) with increasing APS by simple linear regression. Finally, we identified two divergent paths for the poly-atopic march, one driven by perennial and the other by seasonal allergens. Conclusion We conducted a pilot study for a novel machine learning understanding and approach to the classification of APS and potential influences if included in asthma cluster analyses. The methods used here can be easily applied to other geographic regions with different allergens