Gene Expression Profile Helps Predict Atopic Dermatitis Therapy Response

Scientists have identified gene expression profiles that could help clinicians more effectively treat patients with atopic dermatitis (AD).

Writing in the Journal of the American Academy of Dermatology, investigators said they identified a genetic profile that suggests a patient will respond to Janus kinase (JAK) inhibitors, as well as a T helper 2 (Th2) molecular profile that responds similarly to both JAK inhibitors and Th2-targeted therapies.¹

The treatment landscape for AD has evolved significantly in recent years with the development of biologic therapies that interfere with Th2 inflammatory cytokine signaling, the authors noted.

In addition to Th2-targeting therapies, selective JAK inhibitors have also been developed and successfully used to treat patients with moderate-to-severe AD. Yet, the authors explained that the considerations used to decide which patient gets have little to do with the underlying disease itself.

“The current clinical approach to treatment does not account for the molecular drivers of an individual patient’s AD, leading to a trial-and-error approach during therapy selection,” they wrote.

Without a biomarker-driven approach, things like comorbidities, monitoring requirements, and administration preference generally guide treatment decisions.

That may be why many patients do not respond well to therapy. The investigators noted that the Eczema Area and Severity Index (EASI) is typically used to assess response to therapy. A 75% improvement from baseline (known as EASI-75) is generally considered to indicate a successful response to therapy.

Yet, less than 60% of patients treated with Th2-targeted therapies achieve EASI-75, the authors noted.² When the efficacy bar is raised to 90% improvement (EASI-90), response rates dip to 18-26% for patients on Th2-targeted therapies and 25% to 42% for patients taking JAK inhibitors at 12 to 16 weeks.¹

The new study explored whether performing gene expression profiling on patients with AD might make it easier to more precisely choose an appropriate therapy. In the new study, they sought to validate a test designed to perform RNA sequencing on skin scrapings from AD or psoriasis lesions. The test, which is abbreviated as GEP-497, is based on an ensemble algorithm that uses 12 neural networks with 487 genes in order to classify patients by the probability that they would respond to certain types of treatment. The investigators enrolled 192 patients with AD or psoriasis in the training data set. The validation cohort was composed of 110 patients with AD who were at least 12 years old and were taking either a Th2-targeted therapy or a JAK inhibitor.

The investigators found that 3 in 10 patients (30.4%) were classified as having a JAK inhibitor responder profile. Those patients had a far higher rate of EASI-90 at 3 months if they were treated with a JAK inhibitor (45.5%) compared to a Th2-targeted therapy (8.3%; P = .021). Those patients also achieved EASI-90 3.8 times faster than patients receiving Th2-directed therapy; they also reported “no itching” 5.5 times more frequently than those treated with Th2-targeted therapy.

Among patients with a Th2 molecular profile, the EASI-90 rate was not statistically different between the 2 therapies (26.5% for Th2-targeted therapy and 33.3% for JAK inhibitor therapy; P = .728).

The authors concluded their findings support the idea that, in patients with a JAK inhibitor responder profile, inflammatory processes outside the Th2 pathway provide an important contribution.

The investigators added that the use of a tool like the GEP-487 could make it easier to more quickly match patients with meaningful therapy. They noted that other attempts to use biomarkers to guide research require “semi-invasive” methods like biopsies or tape strips.

“These approaches are challenging because they are relatively time intensive, require sample processing or immediate freezing of samples with equipment not typically found in dermatology clinics,” they wrote.

The new test is noninvasive and uses simple tools and techniques, they noted. They said these new results show the test can play a meaningful role in the clinic.

“Use of the 487-GEP provides a clear framework that enables individualized systemic treatment selection and improved health outcomes for patients with AD,” they wrote.

References

1. Silverberg JI, Eichenfield LF, Armstrong AW, et al. The 487-gene expression profile test guides systemic therapy selection to improve outcomes for patients with atopic dermatitis: Results from a prospective trial. J Am Acad Dermatol. Published online February 16, 2026. doi:10.1016/j.jaad.2026.02.034

2. Silverberg JI, Gooderham M, Katoh N, et al. Combining treat-to-target principles and shared decision-making: International expert consensus-based recommendations with a novel concept for minimal disease activity criteria in atopic dermatitis. J Eur Acad Dermatol Venereol. 2024;38(11):2139-2148. doi:10.1111/jdv.20229