Development of patien-derived cancer models of pediatric brain tumours

Introduction: Pediatric brain tumors are, despite all research efforts, still associated with high levels of morbidity and mortality. The inter- and intratumoral heterogeneity in these tumors is a major factor in the frequently unfavorable prognosis of these patients.
This research project aims to investigate the heterogeneity of pediatric brain tumors and the tumor micro-environment, as well as the heterogeneity in the responses to treatment
Methods: The different levels of heterogeneity in these tumors were characterized using literature studies and multiplex immunohistochemistry, with the latter allowing for single-cell analysis with preservation of the spatial organization. Patient-derived models were generated, characterized, and used for drug response assessment.
Results: Literature research identified the heterogeneity within the TME of pediatric glioma and illustrated how standard of care and novel therapies as well as immunotherapy can affect the TME and vice versa. Next, the multiplex immunohistochemical analysis of a large set of pediatric glioma samples again confirmed the heterogeneity. Furthermore, these analyses identified differences in T-cell activation based on their location and discovered spatial gradients in the density of T-cells, with a higher density of T-cells around the blood vessels. Gradients were discovered in the tumoral cell phenotypes depending on their location from the blood vessels in the tumor. Importantly, a p53/PD-1/PD-L1 ecosystem was identified, where the p53 status of the tumor cell influences the PD-1 status of the cytotoxic T-cells and the PD-L1 status of the macrophages and tumor cells.
Novel patient-derived models were generated and characterized, and were used for drug response assessment, illustrating the heterogeneity between these models. The generation of patient-derived models of low-grade tumors was more difficult than high-grade tumors, where the low-grade tumors frequently went into a senescent state.
Models were generated from a DLGNT, both a patient-derived cell line and a patient-derived xenograft, that recapitulated the original characteristics of the tumor of origin and were used to study possible treatment effects.
Conclusion: This study added to the characterization of the heterogeneity of pediatric brain tumors, with the recognition of vascular-related gradients in the tumoral cells and immune cells, together with the identification of a p53/PD-1/PD-L1 ecosystem. These findings still require validation but could be of interest for the use and development of novel treatment strategies, and the use of immunotherapy in selected patients. Furthermore, novel patient-derived models were generated, that can be used for the further study of these tumors and their response to treatment, including rare tumors such as DLGNT.
The identification and understanding of the heterogeneity present in pediatric brain tumors and the development of patient-derived models can aid in a more tailored and personalized treatment approach for these patients. Validation on a larger scale, combining the current characterization of the tumor during standard of care diagnostics, with novel single-cell (multiplexing) techniques and the generation of models, could be the next step towards a better understanding of these tumors and ‘personalized medicine’ for these patients.