Characterising the genomic landscape of two clinical risk groups in CD30+ T-cell lymphoproliferative disorders with cutaneous involvement: a multi-institutional retrospective study

Background CD30+ T cell lymphoproliferative disorders (CD30+ T LPD) with cutaneous involvement comprise a diverse group of neoplasms with variable clinical outcomes. The molecular mechanisms underlying the prognostic differences in these diseases are poorly understood. Methods In this multi-institutional retrospective study, 79 patients with CD30+ T LPD were evaluated. DNA and RNA sequencing were performed on skin biopsy samples to characterise molecular features. Patients were assigned to low- and high-risk groups based on treatment recommendations in clinical records. Pathogenic pathways, immune microenvironments, transcriptomic profiles, driver mutations, and mutation signatures were compared between the risk groups. Findings Across risk groups, there was high mutational recurrence in genes involved in epigenetic regulation and intracellular signalling pathways. Mutational signature analysis revealed a shared prevalence of 5-methylcytosine deamination, UV exposure, defective mismatch repair, and POLE-associated proofreading deficiency. In contrast, gene set enrichment analysis showed that the high-risk group was enriched in pathways governing cell cycle progression, DNA damage response, and immune activation compared to the low-risk group. Furthermore, immune microenvironment deconvolution showed a higher abundance of resting dendritic cells in the low-risk group. Finally, high-risk patients exhibited significant overexpression of clinically relevant proteins, highlighting potential alternative treatment approaches for this population. Interpretation Mutation analyses suggest diverse pathogenesis in these diseases. The high-risk and low-risk groups diverge in gene expression profiles and immune microenvironment, suggesting that these biological differences hold prognostic significance. Molecularly informed therapeutic strategies should be investigated prospectively through clinical trials. Funding This work was supported by developmental funds of the Duke Cancer Institute as part of the P30 Cancer Center Support Grant. Data generation was supported in an unrestricted fashion by the NIH, Data Driven Bioscience, Amazon and Google. SSD was supported by the Wellcome Trust Foundation and the NIH.