Asthma severity and biological therapy selection
To manage the disease, severe asthma requires rigorous pharmaceutical therapy. Although oral corticosteroids are effective, there are significant adverse effects associated with their administration. Although biologics that target the distinct inflammatory mechanisms that underlie the disease have been shown to be beneficial, not all individuals respond to them in the same way. Because biologics are expensive drugs, not anticipating responders has a significant impact on healthcare costs because we treat more patients than those who can respond. Therefore, it would be better to more accurately choose the “right patients” to provide the “right biologics”.
It was possible to improve the ability to anticipate outcomes in patients treated with biologics, as demonstrated by machine learning approaches. Using cluster analysis, researchers recently discovered 4 distinct clusters with varied responses to benralizumab in the high T2 phenotype. The highest response rate (80-90%) was observed in two of these groups, characterized by higher levels of inflammatory markers.
For asthma research, machine learning holds promise because it would allow them to predict which patients will respond to which drugs. The methods could speed up the diagnostic process and increase the likelihood that the best possible course of action will be chosen for each patient, thereby improving patient care and satisfaction.