부산대학교 도서관

Abstract Gliomas, the most frequent type of primary tumor of the central nervous system in adults, results in significant morbidity and mortality. Despite the development of novel, complex, multidisciplinary, and targeted therapies, glioma therapy has not progressed much over the last decades. Therefore, there is an urgent need to develop novel patient-adjusted immunotherapies that actively stimulate antitumor T cells, generate long-term memory, and result in significant clinical benefits. This work aimed to investigate the efficacy and molecular mechanism of dendritic cell (DC) vaccines loaded with glioma cells undergoing immunogenic cell death (ICD) induced by photosens-based photodynamic therapy (PS-PDT) and to identify reliable prognostic gene signatures for predicting the overall survival of patients. Analysis of the transcriptional program of the ICD-based DC vaccine led to the identification of robust induction of Th17 signature when used as a vaccine. These DCs demonstrate retinoic acid receptor-related orphan receptor-γt dependent efficacy in an orthotopic mouse model. Moreover, comparative analysis of the transcriptome program of the ICD-based DC vaccine with transcriptome data from the TCGA-LGG dataset identified a four-gene signature (CFH, GALNT3, SMC4, VAV3) associated with overall survival of glioma patients. This model was validated on overall survival of CGGA-LGG, TCGA-GBM, and CGGA-GBM datasets to determine whether it has a similar prognostic value. To that end, the sensitivity and specificity of the prognostic model for predicting overall survival were evaluated by calculating the area under the curve of the time-dependent receiver operating characteristic curve. The values of area under the curve for TCGA-LGG, CGGA-LGG, TCGA-GBM, and CGGA-GBM for predicting five-year survival rates were, respectively, 0.75, 0.73, 0.9, and 0.69. These data open attractive prospects for improving glioma therapy by employing ICD and PS-PDT-based DC vaccines to induce Th17 immunity and to use this prognostic model to predict the overall survival of glioma patients.