The ability of biomarkers to assess the severity of atopic dermatitis

Abstract

Background: To develop precision medicine for atopic dermatitis (AD), it is critical to establish relevant biomarkers. However, the characteristics of various biomarkers have not been fully understood. We previously carried out the Biomarkers to Predict Clinical Improvement of AD in Patients Treated with Dupilumab (B-PAD) study, a comprehensive nationwide study in Japan, to explore biomarkers for AD.

Objective: The aim of this study is to find biomarkers associated with objective and subjective clinical findings in patients with moderate-to-severe AD based on the B-PAD study and to identify biomarkers sensitive enough to assess the severity of AD.

Methods: We performed the B-PAD study as a consortium composed of 19 medical facilities in Japan, enrolling 110 patients with moderate-to-severe AD. We evaluated the Eczema Area and Severity Index (EASI) for objective assessment as well as the Patient-Oriented Eczema Measure (POEM) and a numeric rating scale for pruritus (pruritis-NRS) for subjective assessment, measuring 19 biomarkers at baseline.

Results: We found that 12, 6, and 7 biomarkers showed significant and positive associations with the EASI, POEM, and pruritis-NRS, respectively. Most of the biomarkers associated with either the POEM or the pruritis-NRS were included among the biomarkers associated with EASI. Of the biomarkers examined, CCL26/eotaxin-3 and SCCA2 were the most capable of assessing severity for EASI, as shown by the 2 kinds of receiver operating characteristic analyses, respectively, whereas lactate dehydrogenase was the best for both the POEM and pruritis-NRS, again using the 2 analyses.

Conclusion: We found biomarkers associated with the EASI, POEM, and pruritis-NRS, respectively, based on the B-PAD study. Moreover, we identified CCL26/eotaxin-3 and/or SCCA2 as the biomarkers having the greatest ability to assess severity in the EASI; lactate dehydrogenase did the same for the POEM and pruritis-NRS. These findings will be useful in treating patients with moderate-to-severe AD.

Keywords: Atopic dermatitis; B-PAD; Biomarkers to Predict Clinical Improvement of AD in Patients Treated With Dupilumab; EASI; LDH; POEM; SCCA2; biomarker; eotaxin-3; precision medicine; pruritus-NRS.

Fig 1

Association of 14 biomarkers with the EASI. Dot plot analysis of the correlation between each biomarker and EASI in the subjects (N = 110) is depicted.

Fig 2

Summary of the associations of 14 biomarkers with clinical findings. Summaries of the associations of each biomarker with EASI (A), the POEM (B), and the pruritis-NRS (C) are depicted. Biomarkers highlighted in gray show significant statistical difference. (D) The relationships of the biomarkers associated with the EASI that are also associated with the POEM or the pruritis-NRS are depicted.

Fig 3

Association of 14 biomarkers with the POEM. Dot plot analysis of the correlation between each biomarker and POEM in the subjects (N = 110) is depicted.

Fig 4

Association of 14 biomarkers with the pruritis-NRS. Dot plot analysis of the correlation between each biomarker and the pruritis-NRS in the subjects (N = 110) is depicted.

Fig 5

Abilities of the biomarkers to assess clinical severity of AD. The abilities of the biomarkers associated with the EASI (A and D), the POEM (B and E), and the pruritis-NRS (C and F) to assess the severity of each clinical finding by the ROC analysis based on the logistic regression analysis (A-C) or by backward stepwise linear regression analysis (D-F). G-I, AUCs according to the ROC analysis based on the backward stepwise linear regression model, using the data from the backward stepwise linear regression analysis (D-F).