Proteomics and transcriptomics reveal molecular subtypes and biomarkers of advanced cutaneous T-cell lymphoma.

Cutaneous T-cell lymphomas (CTCL) are a group of non-Hodgkin T-cell lymphomas, with mycosis fungoides and Sézary syndrome being the most common subtypes. Advanced-stage CTCL are typically aggressive, exhibiting interpatient heterogeneity in treatment response and prognosis. The underlying pathogenesis remains incompletely elucidated, posing challenges for the selection of appropriate therapies. We obtained tumor cell-enriched regions of 33 advanced CTCL samples from 31 patients using laser capture microdissection, followed by integrated proteomic and transcriptomic profiling. Selected biomarkers were further validated via immunohistochemistry. We identified three molecular subtypes of advanced CTCL, which exhibited significant differences in clinical phenotypes, signature proteins, and pathways. These subtypes were designated as intracellular signaling subtype, metabolic subtype, and extracellular matrix remodeling subtype, respectively. Within intracellular signaling subtype, the PI3K-AKT-mTOR pathway was characteristically upregulated, and we found the expression level of phospho-AKT was associated with response to PI3Kδ inhibitor therapy. Comparative proteomic analysis of patients with varying treatment responsiveness and disease progression identified CTSB, GSTO1, and WDFY4 as potential biomarkers for predicting treatment responsiveness, and GOLGA1 and STIP1 as potential biomarkers for progression prediction. This study explored a potential molecular subtyping framework for advanced-stage CTCL associated with different clinical phenotypes. Our findings provided preliminary evidence suggesting that certain biomarkers may be associated with treatment response to PI3K inhibitors. Additionally, we screened and preliminarily identified candidate biomarkers that may be associated with treatment responsiveness and progression risk, which may assist clinicians in the management of advanced-stage CTCL. Notably, these molecular differences may also correlate with clinical characteristics and require validation in larger cohorts.