Clinical phenotypes in alpha-1 antitrypsin deficiency: A cluster analysis on EARCO data

Alpha-1 antitrypsin deficiency (AATD) is a rare, monogenic disorder predisposing individuals to liver and lung diseases. Yet, even within the same genotype, AATD is very heterogeneous in clinical presentation. This heterogeneity suggests a complex interplay of environmental and (epi-)genetic factors, highlighting the need for a detection and study of clinical phenotypes. Here, we performed a cluster analysis using baseline data from the European Alpha-1 Research Collaboration (EARCO) to identify distinct clinical phenotypes in AATD. We identified six clusters of AATD clinical phenotypes using K-Prototypes with a cross-validated weighted F1 score of 0.926. Moreover, using a trained Random Forest classifier, we identified age at diagnosis, lung function and tobacco consumption as most important features in distinguishing these clusters, in addition to the AATD-associated genotype. Overall, these findings underscore that, beyond genotype, a multitude of other factors significantly influence AATD clinical phenotypes. Accordingly, this emphasizes the importance of integrating (epi-)genetic, environmental and behavioral data in future fundamental and translational research in order to unravel the complex mechanisms underlying patient heterogeneity. Moreover, longitudinal follow-up of individuals allocated to these clusters will help to better understand disease progression and to assess the potential benefits of specific therapies in different phenotypes.