Molecular Classifiers Outperform Histopathology in Challenging Psoriasis and Eczema Cases

Accurate differentiation between psoriasis and eczema remains a persistent diagnostic challenge, particularly in anatomically complex or clinically heterogeneous presentations such as chronic hand eczema (CHE). A new study evaluated the diagnostic concordance between conventional dermatopathology and a molecular classifier based on NOS2 and CCL27 gene expression, with particular relevance for improving diagnostic precision in ambiguous inflammatory dermatoses.¹
Methods and Materials

The investigators in Germany and the Czech Republic conducted a retrospective, multicenter analysis of formalin-fixed, paraffin-embedded (FFPE) skin biopsies, including a primary cohort of 73 cases (28 psoriasis, 45 eczema) and an independent validation cohort of 72 cases. Fourteen dermatopathologists in the primary cohort and 3 in the validation cohort independently assessed haematoxylin and eosin-stained sections. Diagnostic categories were consolidated into psoriasis, eczema, or undetermined. Molecular analysis was performed using both a manual quantitative real-time polymerase chain reaction (RT-qPCR) workflow and a fully automated platform (PsorX-LabDisk), each quantifying NOS2 and CCL27 expression to classify disease.

Results

A central finding of the study is the substantial inter-observer variability in dermatopathological diagnosis. Across both cohorts, agreement among dermatopathologists was only fair, with a Fleiss’ κ of 0.31. Pairwise agreement ranged widely, underscoring the subjective nature of histopathological interpretation in inflammatory skin disease. This variability is particularly relevant to hand eczema, where histological features such as spongiosis, acanthosis, and inflammatory infiltrates may closely resemble psoriasiform patterns, especially in chronic or hyperkeratotic subtypes. Researchers noted that in other diagnostic fields, including melanocytic tumors, almost 50% interrater variability persists even among experts.²

In terms of diagnostic performance, conventional dermatopathology achieved a mean accuracy of 76.9%, with sensitivity for psoriasis of 70% and specificity of 81.6%. Notably, eczema diagnoses were more consistently identified than psoriasis, suggesting a tendency toward underdiagnosis of psoriasis in equivocal cases. This has direct implications for CHE, as psoriasis, particularly palmoplantar variants, may be misclassified as chronic eczema, potentially delaying appropriate treatment.

In contrast, molecular diagnostics demonstrated superior and more reproducible performance. Both the manual RT-qPCR assay and the automated PsorX-LabDisk achieved a sensitivity of 92.9%, with specificities of 82.2% and 84.4%, and overall accuracies of 86.3% and 87.7%, respectively. Importantly, the automated system maintained performance comparable to the manual assay while offering a standardized, examiner-independent workflow with rapid turnaround.

The advantage of molecular testing was most pronounced in diagnostically ambiguous cases, which comprised a substantial proportion of the cohort (approximately 43%). In this subgroup, dermatopathological accuracy dropped to 61.7%, with near-random inter-observer agreement. Such scenarios are highly representative of real-world challenges, including CHE versus palmoplantar psoriasis, where clinical and histological overlap is common. In these difficult cases, both molecular approaches maintained high diagnostic accuracy (approximately 86.7%), significantly outperforming conventional pathology.

Cases and Clinical Implications

Representative cases further highlight the clinical utility of the molecular classifier. In a patient with palmoplantar disease suggestive of either pustular psoriasis or dyshidrotic hand eczema, dermatopathological interpretations were inconclusive. The molecular test supported a diagnosis of psoriasis, which was subsequently corroborated by therapeutic response to TNF- and IL-17-targeted treatments.

The study underscores several important clinical implications, according to the authors. First, it challenges the perception of dermatopathology as a definitive gold standard in inflammatory dermatoses, emphasizing its inherent subjectivity. It also supports the integration of molecular diagnostics as an adjunct tool, particularly in anatomically and histologically complex conditions such as hand eczema. By improving diagnostic accuracy, molecular classifiers may facilitate earlier initiation of targeted therapies, reduce ineffective treatment cycles, and ultimately improve patient outcomes.

References

1. Schmitt A, Proksch S, Gutzweiler L, et al. Diagnostic concordance of dermatopathology and PCR in differentiating eczema from psoriasis. J Eur Acad Dermatol Venereol. Published online March 12, 2026. doi:10.1111/jdv.70388

2. Haggenmüller S, Wies C, Abels J, et al. Discordance, accuracy and reproducibility study of pathologists' diagnosis of melanoma and melanocytic tumors. Nat Commun. 2025;16(1):789. Published 2025 Jan 17. doi:10.1038/s41467-025-56160-x