Asthma and most chronic airway diseases are heterogeneous entities. Even severe asthma does not represent a single phenotype of asthma. This heterogeneity of phenotypes is the starting point of new approaches for characterisation, understanding and management of asthma in the near future. The ultimate approach will be to identify new phenotypes sharing coherent underlying biological mechanisms (i.e. the concept of endotypes [1]) to better predict future risks. Ultimately, new specific, targeted or personalised therapeutic avenues and management will be developed based on these new groupings of patients and used [2].
The main challenge is to avoid pre-established hypotheses. An unbiased approach of phenotyping is appealing in view of these ambitious goals. The increased popularity of clusters analysis in asthma benefits from the development of cohorts of asthmatic patients worldwide [1, 3–5]. For example, clusters derived from the Severe Asthma Research Program (SARP) initiative in the USA have been well received and disseminated within the severe asthma community [4]. This phenomenon can be explained by different factors: …