Background Epidemics of acute asthma associated with thunderstorms occur intermittently worldwide, though airway inflammation during these acute episodes has not been characterized. The aim of this study was to characterize airway inflammation in thunderstorm asthma.
Methods Cases were recruited after presentation to the emergency room with acute asthma immediately following a thunderstorm (n = 6). They were compared to two control groups: a group of atopic asthmatics that had presented with acute asthma to the emergency room prior to the thunderstorm (n = 12), and a second group of corticosteroid naïve asthmatics who presented to the emergency room in the prior 12 months (n = 6). Subjects had spirometry, sputum induction and allergy skin tests acutely and at review 4 weeks later.
Results Thunderstorm (TS) cases were more likely to have a history of hay fever and grass pollen allergy, and less likely to be on inhaled corticosteroids (ICS) prior to presentation. Cases and control groups had a similar degree of moderate to severe acute airway obstruction (P = 1.0). TS cases had elevated sputum eosinophils (14.8% of total cell count) compared to controls (1%, 2.6%, P < 0.01). TS cases had higher sputum eosinophil cationic protein (ECP; 11 686 ng/mL) compared to controls (1883, 3300, P = 0.02) acutely. TS cases had more cells positive for IL-5 (30%) compared to controls (1, 1.5%, P = 0.02). When adjusted for ICS use, TS cases had a risk ratio for elevated sputum eosinophils of 2.4 (1.23–4.69).
Conclusion Thunderstorm asthma is characterized by airway inflammation with IL-5-mediated sputum eosinophilia and eosinophil degranulation. These results are consistent with allergen exposure as the cause of the exacerbation, and are consistent with the thunderstorm-induced grass pollen deluge as the cause of epidemic asthma after thunderstorms.