CXCR4-centred signalling network rewiring distinguishes primary and metastatic prostate cancer: A systems-level digital twin framework.

Metastatic progression of prostate cancer involves complex remodelling of signalling networks that cannot be fully explained by changes in individual pathway activity. In this study, we performed a systems-level analysis of signalling module interactions using transcriptomic data from primary prostate tumours in the TCGA-PRAD cohort (n = 550) and metastatic tumours from the SU2C cohort (n = 118). The framework integrates thirteen metastasis-associated signalling modules encompassing oncogenic, inflammatory, stress-response, angiogenic and mechanotransduction pathways. Module activity scores were computed for key cancer-related signalling pathways and used to construct module co-occurrence networks. While differences in median pathway activity between cohorts were modest, metastatic tumours exhibited substantial rewiring of pathway coordination. Notably, this rewiring reflects changes in pathway coordination rather than absolute activity levels, supporting a network-centric view of established metastatic disease. Network analysis identified the CXCR4 signalling axis as a central hub showing increased connectivity with inflammatory and growth-factor pathways, including NF-κB, MAPK, and hypoxia modules. Unsupervised clustering of pathway activity profiles revealed two distinct signalling states characterized by differential androgen receptor signalling and activation of metastasis-associated pathways. These signalling states were significantly associated with PTEN and TP53 alterations. A logistic regression classifier based on module activity achieved moderate discrimination between primary and metastatic tumours (AUC = 0.685), indicating that coordinated pathway activity captures biologically relevant features of disease progression. Overall, our findings highlight the importance of network-level approaches for understanding metastatic prostate cancer and identify CXCR4-centred signalling architecture as a key component of metastatic signalling rewiring.