Latent class analysis categorizes chronic hand eczema patients according to skin barrier impairment
Résumé
Background :
Chronic hand eczema (CHE) is the most common skin disorder affecting the hands. It causes major physical and psychological burden for patients. Classification of CHE remains challenging because of its aetiological and clinical heterogeneity. Objectives : Using latent class analysis (LCA) on a large categorical data set, our aim was to identify distinct phenotypes in a cohort of unselected CHE patients based upon clinical, genetic, molecular and physical parameters of the affected skin. Methods :
We performed two independent LCA on a cohort of 71 well‐characterized patients that initially integrated clinical severity, total immunoglobulin E plasma level, transepidermal water loss, hydration index, interleukin(IL)‐8 lesional skin level, Staphylococcus (aureus and epidermidis) colonization, FLG genotype and the expression (mRNA) of genes involved either in the filaggrin degradation and the natural moisturizing factor synthesis, the cornified envelope formation, the tight junctions’ structure and the desquamation process, or encoding antimicrobial peptides and chemokines. Results : The first LCA categorized patients into a group displaying high severity of CHE, high skin barrier impairment, high Staphylococcus colonization, high IL‐8 skin level and high frequency of mutation in the FLG gene and a second group with opposite characteristics. The second LCA identified two independent groups of patients categorized by their low or high level of skin barrier impairment and corresponding changes in the expression of the related genes.
Conclusions : Our study suggests that the degree of skin barrier dysfunction is the most important parameter to discriminate CHE patients and probably plays a pivotal role in the pathogenesis of the disease whatever the aetiological factors. As far as we know, this is the first study to address this topic using a statistical categorization method without preconception.