Multi-Omic Profiling for Molecular Diagnosis of Erythrodermic Skin Diseases

Background Erythroderma is one of the most challenging clinical scenarios in modern dermatology. It represents an acute, potentially life-threatening inflammatory condition characterised by erythema and scaling affecting more than 90% of the skin surface, with mortality rates up to 9%. Various underlying skin diseases cause erythroderma yet are difficult to distinguish clinically and histopathologically. This poses a major diagnostic challenge and delays targeted therapy, with potentially severe acute and long-term consequences. Objectives To define disease-specific molecular programs in erythrodermic skin using integrated proteomics and transcriptomics, and to translate these into a diagnostic framework. Methods Here we performed mass spectrometry (MS) - based proteomics and RNA sequencing on formalin-fixed, paraffin-embedded skin biopsies from 96 patients with erythroderma caused by pityriasis rubra pilaris (PRP, n=19), psoriasis (n=18), atopic dermatitis (AD, n=20), cutaneous T-cell lymphoma (CTCL, n=18), drug-induced maculopapular rash (MPR, n=21), and 20 healthy skin controls. Samples from four independent centers were included and MS-based proteomics was acquired on the Orbitrap Astral mass spectrometer. We performed multi-omic data integration through multi-omics factor analysis (MOFA), and disease classification was achieved using the adaptive machine-learning framework ADAPT-Mx, benchmarked against blinded clinico-histopathological assessment. Results We quantified over 9,300 proteins and 17,200 protein-coding transcripts, revealing distinct molecular signatures for each disease. The proteome showed elevated expression of type-2 immunity-associated Charcot-Leyden crystal protein in AD, establishing a potential link between eosinophil-dominant inflammation and NLRP3 inflammasome activation. Transcriptomic analysis demonstrated selective upregulation of IL-17C in PRP, correlating strongly with IL-1 family cytokine expression. Notably, only a subset of PRP patients exhibited this signature, suggesting treatment-relevant disease endotypes. Multi-omic integration identified elevated expression of T-cell regulator RASAL3 in CTCL, not previously characterized in its pathogenesis. ADAPT-Mx, our machine-learning classifier, achieved 73.9% diagnostic accuracy in the held-out test cohort, outperforming conventional clinico-histopathological assessment (59.5%), with AUC values of 0.87–1.00 across disease entities. Conclusions This study provides an integrated multi-omic framework for inflammatory skin diseases, combining RNA sequencing with large-scale MS-based proteomics. By enabling per-sample classification without imputation, our workflow offers proof of concept for multi-omic-driven diagnosis that can support clinical decision-making across inflammatory skin diseases.

Authors

Institutions

Publication Details

Journal
British Journal of Dermatology
Published
2026-11-19
DOI
https://doi.org/10.1093/bjd/ljag505
Primary Topic
Dermatology and Skin Diseases
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Multi-Omic Profiling for Molecular Diagnosis of Erythrodermic Skin Diseases

Rudolf Stadler, Anne-Sophie Boehm, Thierry M. Nordmann, Benjamin Kendziora et al.
1 links
British Journal of Dermatology
Dermatology and Skin Diseases
article

Multi-Omic Profiling for Molecular Diagnosis of Erythrodermic Skin Diseases

Rudolf Stadler, Anne-Sophie Boehm, Thierry M. Nordmann, Benjamin Kendziora, Eva Maria Oppel, Nora ASZODI, Felix Lauffer, Lars Einar French, Laurie Eicher, Pia‐Charlotte Stadler, Benjamin Maximilian Clanner-Engelshofen, Werner Kempf, Nicholas Moellhoff, Teodora Pumnea, Georg Wallmann, Anna Oschmann, Maximilian Zwiebel, Doris Helbig, M. Neulinger, Zeno Fiocco, Takashi K. Satoh, Christiane Pfeiffer, Sonja Senner, Burkhard Summer, Leonie Helene Frommherz, Rubén A. Ferrer, Matthias Mann, Sophia Czell, Irma Kupf, Valerie Glatzel, Michael Joseph Flaig, Anna Leonhardt, Katrin Kerl-French, Michaela Maurer, Mohammed Mitwalli, Annika Sander, Julia Leding, Lucas Diedrich
article en

Abstract

Background Erythroderma is one of the most challenging clinical scenarios in modern dermatology. It represents an acute, potentially life-threatening inflammatory condition characterised by erythema and scaling affecting more than 90% of the skin surface, with mortality rates up to 9%. Various underlying skin diseases cause erythroderma yet are difficult to distinguish clinically and histopathologically. This poses a major diagnostic challenge and delays targeted therapy, with potentially severe acute and long-term consequences. Objectives To define disease-specific molecular programs in erythrodermic skin using integrated proteomics and transcriptomics, and to translate these into a diagnostic framework. Methods Here we performed mass spectrometry (MS) - based proteomics and RNA sequencing on formalin-fixed, paraffin-embedded skin biopsies from 96 patients with erythroderma caused by pityriasis rubra pilaris (PRP, n=19), psoriasis (n=18), atopic dermatitis (AD, n=20), cutaneous T-cell lymphoma (CTCL, n=18), drug-induced maculopapular rash (MPR, n=21), and 20 healthy skin controls. Samples from four independent centers were included and MS-based proteomics was acquired on the Orbitrap Astral mass spectrometer. We performed multi-omic data integration through multi-omics factor analysis (MOFA), and disease classification was achieved using the adaptive machine-learning framework ADAPT-Mx, benchmarked against blinded clinico-histopathological assessment. Results We quantified over 9,300 proteins and 17,200 protein-coding transcripts, revealing distinct molecular signatures for each disease. The proteome showed elevated expression of type-2 immunity-associated Charcot-Leyden crystal protein in AD, establishing a potential link between eosinophil-dominant inflammation and NLRP3 inflammasome activation. Transcriptomic analysis demonstrated selective upregulation of IL-17C in PRP, correlating strongly with IL-1 family cytokine expression. Notably, only a subset of PRP patients exhibited this signature, suggesting treatment-relevant disease endotypes. Multi-omic integration identified elevated expression of T-cell regulator RASAL3 in CTCL, not previously characterized in its pathogenesis. ADAPT-Mx, our machine-learning classifier, achieved 73.9% diagnostic accuracy in the held-out test cohort, outperforming conventional clinico-histopathological assessment (59.5%), with AUC values of 0.87–1.00 across disease entities. Conclusions This study provides an integrated multi-omic framework for inflammatory skin diseases, combining RNA sequencing with large-scale MS-based proteomics. By enabling per-sample classification without imputation, our workflow offers proof of concept for multi-omic-driven diagnosis that can support clinical decision-making across inflammatory skin diseases.

British Journal of Dermatology
Broad Institute (US), University of Copenhagen (DK), Harvard University (US), University of Miami (US), University of Zurich (CH), Novo Nordisk Foundation (DK), LMU Klinikum (DE), Massachusetts General Hospital (US), University Hospital of Zurich (CH), Aesthetic Surgery Center (US), München Klinik (DE), Max Planck Institute of Biochemistry (DE), University Hospital Cologne (DE), Johannes Wesling Klinikum Minden (DE), Ludwig-Maximilians-Universität München (DE)
National Science Foundation, Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung, Max-Planck-Gesellschaft, Ludwig-Maximilians-Universität München
Good health and well-being
Dermatology and Skin Diseases
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon