Abstract
Glioblastoma (GB) is one of the most aggressive brain tumors, characterized by high infiltrative capacity that enables tumor cells to invade healthy brain tissue and evade complete surgical resection. This invasiveness contributes to resistance against conventional therapies and a high recurrence rate. Strategies capable of eliminating residual tumor cells are urgently needed. Photodynamic therapy (PDT) using 5-aminolevulinic acid (5-ALA), an FDA- and EMA-approved compound, induces selective accumulation of the photosensitizer protoporphyrin IX (PpIX) in metabolically active tumor cells, enabling targeted cytotoxicity through light activation. A major limitation to its clinical application is the unclear variation in the cytotoxic effect of PDT according to individual tumoral differences. In this study, we propose and validate an in vivo model of patient-derived GB initiating cells (GICs) and brain organoids to test the effects of PDT. First, patient-derived GICs were molecularly characterized by flow cytometry and copy number variation profiling using OncoScan CNV Assays, then co-cultured with human brain organoids to generate a hybrid model recapitulating key aspects of the tumor microenvironment. 5-ALA photodynamic therapy (PDT) efficacy was assessed in vitro by GFP-based viability measurements, LDH release assays, and TUNEL staining. Then, a murine model was generated to study PDT in vivo, based on a heterotopic (renal subcapsular engraftment) xenograft of the GICs-human brain organoid co-culture. PDT was tested in the model; in each subject, one kidney tumoral engraftment was treated and the contralateral served as a control. Immunofluorescence analysis was used to study the cell composition of the brain organoid-tumoral engraftment after PDT, and the effects on non-GIC cells. The antitumoral effect was determined by the degree of cell death analysis with the TUNEL technique. The GICs-brain organoid co-culture resulted in tumoral growth and infiltration both in vitro and in vivo. The pattern of growth and infiltration varied according to the tumoral genetic profile. 5-ALA PDT resulted in a reduction in the number of GICs and an increase in apoptotic cells in all four lines tested in vitro. A correlation was found between the induced phototoxicity in vivo with the molecular typification of GICs cell lines in vitro. There were no changes in the number or distribution of neuronal cells after the application of PDT, while a reduction in active astrocytes was observed. 5-ALA PDT could be effective in eradicating GICs with a heterogeneous molecular profile. The hybrid human-murine model presented here could be useful in investigating adjuvant therapies in GB, under the concept of personalized medicine.