Improving evaluation of drugs in atopic dermatitis by combining clinical and molecular measures

[...]expanding on a molecular AD meta-analysis,2 we evaluated both clinical improvements and molecular skin changes (using Affymetrix U133Plus 2.0 gene-arrays) of different systemic and topical therapeutics in patients with AD.2 Seven publicly available microarray data sets from patients with AD were found in Gene Expression Omnibus,1,3-7,9 as previously reported.2 Outcomes of interest included a comparison of the clinical and molecular improvements across the meta-analysis–derived AD transcriptome (MADAD) (see the Methods section in this article's Online Repository at https://www.jaci-inpractice.org).1,3 Improvement was defined as the percent change over time (eg, difference in lesional skin expression at a particular time point with the lesional baseline expression, divided by the difference in lesional and nonlesional expression at baseline, all multiplied by 100). Clinical improvements from outcome measures by each study (eg, SCORing of AD, Eczema Area and Severity Index, and total sign score) were also calculated for each patient as available. Because of the inherent normalization of percent improvement, the percent improvements were directly compared across treatments by parametric and nonparametric tests (eg, t test and Wilcoxon test; see Table E1 in this article's Online Repository at www.jaci-inpractice.org). [...]the baseline expression of AD-related markers was correlated with the clinical improvement at various time points across treatments to assess whether individual skin biomarkers could potentially predict future clinical responses.2 Overall, most clinical improvements were aligned with molecular improvements (Figure 1; Table E1). [...]the individual biomarkers here could help predict therapeutic responses to various treatments, which is helpful when targeting a heterogeneous disease with varied treatment responses.