Developing 3D Models of Paediatric Solid Tumours

Abstract

High-grade gliomas (HGG) are aggressive brain tumours with no curative therapies, and survival rates have remained unchanged for decades. Although treatments often appear promising in preclinical studies using conventional 2D assays and animal models, they frequently fail at clinical trials due to a lack of efficacy or unacceptable toxicity. More representative 3D in vitro models of the tumour and healthy brain microenvironment are urgently required to address these discrepancies and to improve patient outcomes. Initially, we developed a functional assay for assessing neurotoxicity using dorsal cortical organoids mimicking the structure and cellular populations of the developing human neocortex. Spontaneous electrical activity in organoids, measured by micro-electrode array (MEA), was significantly altered upon chronic exposure to kinase inhibitors lorlatinib and gefitinib. Next, we evaluated the potency of a novel EphA2-targeted CAR-T cell therapy against HGG models. Treatment sensitivity decreased from 2D to 3D spheroid models, necessitating increased dosage and duration. When treatment was administered to assembloid co-culture models comprising both tumour and organoid components, CAR-T cells targeted EphA2-expressing tumour cells with diminished sensitivity, more closely reflecting in vivo responses. Lastly, patient-derived organoids (PDOs) were established from paediatric glioblastoma multiforme (pGBM) patient tumour samples, and circulating tumour DNA (ctDNA) was isolated from culture medium. ctDNA exhibited hotspot mutations consistent with clinical reports from the primary tumour, offering an alternative resource for diagnosis, preclinical therapeutic screening, and personalised treatment selection. Collectively, this research underscores the importance of 3D in vitro models in drug development and clinical workflows, demonstrating their utility in predicting neurotoxicity, assessing treatment efficacy, and enhancing diagnostic characterisation.