A proteome-based classification of paediatric adrenocortical tumours links functional tumour states to clinical outcome and therapeutic vulnerabilities.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND: Paediatric adrenocortical tumours (pACT) are rare endocrine neoplasms characterised by marked biological heterogeneity and variable clinical behaviour. Current risk stratification relies largely on histopathology and clinical staging, which incompletely capture underlying tumour biology. We aimed to define proteome-based subtypes of pACT and evaluate their biological, clinical, and therapeutic relevance. METHODS: In this retrospective cohort study, we performed mass spectrometry-based proteomic profiling of 83 pACT and adjacent non-tumour adrenal tissues. Differential abundance analysis, pathway enrichment, consensus clustering, survival analyses, and clinicopathological correlations were performed. A minimal protein classifier was derived using forward feature selection and leave-one-out cross-validation. Cluster-specific therapeutic vulnerabilities were assessed using a curated druggability framework. FINDINGS: Tumours displayed profound proteomic differences compared with adjacent non-tumour adrenal cortex, characterised by activation of cholesterol biosynthesis, steroidogenic, MYC, mTORC1, and cell-cycle programs and suppression of extracellular matrix, adhesion, and immune-related pathways. Unsupervised clustering identified four proteome-defined tumour states that only partially overlapped with histopathological diagnosis. These subtypes differed significantly in endocrine phenotype, vascular invasion, Ki67 index, necrosis, and overall survival (log-rank p = 0.00036). Distinct biological programs included stromal-immune-enriched, mitochondrial steroidogenic, IGF/mTOR-driven anabolic, and highly proliferative chromatin-remodelled states. A five-protein classifier (DAAM2, CIP2A, TSC2, PALS2, and P3H1) reproduced cluster assignments with 92.8% leave-one-out cross-validation accuracy. Cluster-specific analyses identified distinct therapeutic target landscapes, including IGF-axis and epigenetic vulnerabilities. INTERPRETATION: Proteomic profiling defines biologically and clinically distinct pACT subtypes that provide prognostic information beyond histopathology. This framework enables proteome-based stratification, identifies candidate therapeutic vulnerabilities, and can be approximated by a five-protein classifier. FUNDING: German Childhood Cancer Foundation, Mitteldeutsche Kinderkrebsforschung, European Union's EU4Health Programme, Max Planck Society, Onassis Foundation, and Joachim Herz Stiftung.