Measuring the corticosteroid responsiveness endophenotype in asthmatic patients

Inhaled corticosteroids (ICSs) constitute the most commonly prescribed therapies for asthma. They are effective, but there are up to 24% of asthma patients who do not achieve significant improvement with them. ICSs produce treatment responses in six clinical phenotypes: lung function, bronchodilator response, airway responsiveness, symptoms, need for oral steroids, and frequency of emergency department visits or need for hospitalization. For the past 15 years and in an escalating prevalence of asthma, researchers have considered these phenotypes to be guided by separate mechanisms.

Clemmer et al propose a move away from the focus on single phenotypes to a more holistic approach. They suggest that there is a corticosteroid responsiveness endophenotype that modulates the asthma disease process, is latent in ICS-untreated patients, and is active in ICS-treated patients. Under this hypothesis, the corticosteroid responsiveness endophenotype influences the asthma disease process to produce the treatment effect observed in all the clinical phenotypes (J Allergy Clin Immunol 2015; 136(2): 274-281).

As such, the authors present a composite phenotype responsiveness model that combines the six clinical phenotypes and measures the endophenotype. They used principal component analysis (PCA) to determine the model in a study population of both ICS-treated and non-ICS-treated patients with mild to moderately severe asthma. The model was then tested in four replication populations. Using treatment effect area under the receiver operating characteristic curve (AUC), they demonstrate that a composite phenotype measures corticosteroid responsiveness with greater accuracy and stability across populations than the individual clinical phenotypes do.

The potential applications of the composite phenotype are many. It should enable asthma pharmacogenetic studies with more power for a given sample size or that require a smaller sample to achieve a given power. Given that it collapses multiple longitudinal clinical observations into a corticosteroid response metric and that it is easily implemented in a single computer program, it could allow a clinical practitioner to more accurately estimate ICS response. Finally, the model could be used to characterize the many asthma patients who do not respond to ICS treatment with better accuracy.