A multinational retrospective cohort study described persistent, treatment-associated phenotype switching between psoriasis and atopic dermatitis, a phenomenon researchers said may be mistaken for treatment failure or paradoxical worsening if not recognized.
The multicenter observational study included 148 patients treated at 17 dermatology centers in Portugal, Spain, Italy, Greece, Switzerland, and Canada. Patients had psoriasis or atopic dermatitis and developed a clinician-defined, persistent switch to the opposing inflammatory phenotype in temporal association with systemic or biologic therapy. The switch also had to lead to clinician-directed treatment modification, and histopathologic or molecular confirmation was not required.
Among the included patients, 101 developed eczematous features while receiving treatment for psoriasis, and 47 developed psoriasiform disease while receiving treatment for atopic dermatitis. Psoriasis-to-eczematous switching was most often documented during treatment with interleukin-17 inhibitors, which accounted for 58% of cases, followed by interleukin-23 inhibitors at 25%.
Atopic dermatitis-to-psoriasiform switching occurred during biologic therapies targeting type 2 inflammation. Dupilumab accounted for 91.5% of cases, and selective interleukin-13 inhibitors accounted for 8.5%.
The researchers framed the phenomenon as immune rebalancing rather than a transient paradoxical reaction. They suggested that selective cytokine blockade may disrupt reciprocal regulation between type 17/type 1 and type 2 inflammatory pathways, unmasking a subclinical or predisposed opposing phenotype.
Patients with psoriasis-to-eczematous switching were older and more likely to have cardiometabolic comorbidities. Atopic comorbidities were more common among patients with atopic dermatitis-to-psoriasiform switching. However, the groups showed overlapping clinical backgrounds: approximately 30% of patients with atopic dermatitis-to-psoriasiform switching had a family history of psoriasis, and approximately 26% of patients with psoriasis-to-eczematous switching had atopic comorbidities.
Treatment history differed between groups. All patients with atopic dermatitis-to-psoriasiform switching were biologic-naïve at the time of phenotype switch, compared with 44% of patients with psoriasis-to-eczematous switching.
Following recognition of phenotype switching, treatment was modified in all patients, consistent with the study's inclusion criteria. Janus kinase inhibitors were the most frequently documented management strategy, especially among patients with psoriasis-to-eczematous switching. In that group, 60% were switched to a Janus kinase inhibitor, most commonly upadacitinib. Among patients with atopic dermatitis-to-psoriasiform switching, 38% were switched to a Janus kinase inhibitor, also most commonly upadacitinib, although management was more heterogeneous.
Clinical severity scores improved following therapeutic adjustment in both groups over a mean treatment duration of about 17 months in the psoriasis-to-eczematous group and about 18 months in the atopic dermatitis-to-psoriasiform group. The baseline scores reflected both the treated primary disease and the emergent phenotype: among patients with psoriasis-to-eczematous switching, mean Psoriasis Area and Severity Index scores decreased from 6.6 to 1.5, and mean Eczema Area and Severity Index scores decreased from 17.5 to 3.1. Among patients with atopic dermatitis-to-psoriasiform switching, mean Psoriasis Area and Severity Index scores decreased from 8.6 to 1.4, and mean Eczema Area and Severity Index scores decreased from 13.9 to 1.0.
The researchers emphasized that the study was not designed to estimate incidence, identify predictors, or compare risk across drug classes. The higher number of cases observed during interleukin-17 inhibitor or dupilumab therapy may reflect broader use, longer availability, and greater real-world exposure rather than greater intrinsic risk. The study also could not compare the effectiveness of different management strategies.
Limitations included the retrospective observational design, lack of a comparator population, potential selection bias, incomplete data capture, and absence of systematic histopathologic or molecular confirmation. The predominantly Caucasian cohort may also limit generalizability to more diverse populations.
For physicians, the findings suggest that sustained phenotype switching should be considered when patients develop opposing psoriatic or eczematous features during targeted therapy. Recognition may help avoid misclassifying the presentation as simple treatment failure and may guide therapeutic adaptation, including broader immunomodulatory approaches such as Janus kinase inhibition in selected patients.
Disclosures: The study received no external funding. Multiple researchers reported consulting fees, advisory roles, speaker honoraria, research support, or clinical trial participation with pharmaceutical companies involved in dermatology therapeutics.
Source: Journal of the European Academy of Dermatology and Venereology
