Subphenotypes of NSAID-exacerbated respiratory disease identified by latent class analysis

Background Induced sputum (IS) allows to measure mediators of asthmatic inflammation in bronchial secretions. NSAID-exacerbated respiratory disease (NERD) is recognized as a distinct asthma phenotype, usually with a severe course, eosinophilic airway inflammation, and increased production of pro-inflammatory eicosanoids. A more insightful analysis of NERD patients has shown this phenotype to be nonhomogeneous. Objective We aimed to identify possible subphenotypes in a cohort of NERD patients with the means of latent class analysis (LCA). Methods A total of 95 asthma patients with aspirin hypersensitivity underwent sputum induction. High-performance liquid chromatography or gas chromatography coupled with mass spectrometry was used to profile eicosanoids in induced sputum supernatant (ISS). Sixteen variables covering clinical characteristics, IS inflammatory cells, and eicosanoids were considered in the LCA. Results Three classes (subphenotypes) were distinguished within the NERD cohort. Class 1 subjects had mild-to-moderate asthma, an almost equal distribution of inflammatory cell patterns, the lowest concentrations of eicosanoids, and logLTE(4)/logPGE(2) ratio. Class 2 represented severe asthma with impaired lung function despite high doses of steroids. High sputum eosinophilia was in line with higher pro-inflammatory LTE4 in ISS and the highest logLTE(4)/logPGE(2) ratio. Class 3 subjects had mild-to-moderate asthma and were also characterized by eosinophilic airway inflammation, yet increased production of pro- (LTE4, PGD(2) and 11-dehydro-TBX2) was balanced by anti-inflammatory PGE(2). The value of logLTE(4)/logPGE(2) was between values calculated for classes 1 and 3, similarly to disease control and severity. Conclusions LCA revealed three distinct NERD subphenotypes. Our results support a more complex pathobiology of aspirin hypersensitivity. Considering NERD heterogeneity, the relationship between inflammatory pathways and clinical manifestations of asthma may lead to more individualized treatment in difficult to treat patients in the future.