How clinical phenotypes can translate into taxonomy: can we judge atopic dermatitis by its cover?

Abstract

Atopic dermatitis (AD) is a heterogeneous disease with a diverse clinical picture. Correspondingly, its severity varies from mild disease with flexural eczema, which may be successfully treated with a short-term topical treatment, to very severe universal disease requiring systemic immunomodulatory treatment. In the current literature the terms phenotype and endophenotype are often used. The first describes the clinical picture met by the clinician in the clinic, and the second is mostly used to encompass different biomarkers found in biological samples such as blood and skin. A significant part of the research published in recent years has centred on correlating endophenotype to the prognostic risk of comorbidities and response to treatment, especially as AD has entered the age of biologics. However, in the clinical situation with the patient and patient’s family, it would be optimal to be able to make risk assessments and predictions from the severity and clinical history of the patient’s disease. In this issue of the BJD, Mulick et al. report on their study of 11 866 children from the Avon Longitudinal Study of Parents and Children birth cohort. The study aimed to classify AD into robustly identifiable subtypes, using data generated from questionnaires on the presence of flexural dermatitis and severity within the previous year, collected at 11 time points over the child’s first 14 years of age. The study is performed elegantly, and should in many ways serve as guidance for future epidemiological studies. The authors divide the population into an explorative (development) cohort, on which they perform latent class analysis, and a validation cohort to confirm the results obtained in the first group. From the latent class analysis models, they found the optimal number of phenotype groupings is five, comprising four disease phenotypes – Unaffected/Rare, Severe–Frequent AD, Moderate–Frequent AD, Moderate–Declining AD and Mild– Intermittent AD – and that there are clinically relevant associations between the different subgroups and asthma. They also show that family history of AD increases the risk of having one of the severe phenotypes. Altogether, the study demonstrates that there are phenotypic subtypes of AD, and that they are clinically meaningful. The approach is based on clinical subtypes, and together with studies performed on endophenotypes, we are moving towards a better taxonomy within the large and heterogeneous group of patients with AD, the ultimate goal being a precise and personalized risk assessment, intervention and treatment.