Abstract
PURPOSE OF REVIEW: Sarcomas are a heterogeneous group of over 170 malignant tumours of mesenchymal origin. The poor prognosis highlights the need for novel therapeutic strategies. Preclinical modelling is essential, yet challenging, given that sarcomas differ substantially from carcinomas and resources are very limited. RECENT FINDINGS: GEMMs allow for the precise modelling of recurrent sarcoma genetics. The Cre-loxP system offer spatial and temporal control over the activation of oncogenes or the loss of tumour suppressors, while the CRISPR-Cas9 system enables the rapid, simultaneous editing of key drivers such as Trp53, Nf1, Kras and Pten. These models reproduce key features of human sarcomas, including their histopathology, the initiation of tumours in specific lineages and sites, and tumour-immune interactions within immune-competent hosts. GEMMs have been used to investigate hypotheses about the cells of origin, to test radiotherapy and immunotherapy, and to compare fusion-driven sarcomas with those with a complex karyotype. Despite variability, GEMMs remain essential tools for investigating the mechanisms of initiation, progression, and response to therapy. GEMMs offer mechanistic fidelity, but their use is limited by factors such as breeding burden, variability in recombination, off-target effects of CRISPR, underrepresentation of genomic complexity and inconsistent metastasis. These weaknesses reduce their predictive value, particularly with regard to advanced disease and immunotherapy. Progress will require the integration of Cre-loxP with CRISPR-Cas9, the standardisation of induction and reporting, and a closer alignment with distinct sarcoma subtypes, in order to enhance translational relevance.