Logo: to the web site of Uppsala University

Glioblastoma (GBM) is an incurable brain cancer with a median survival of less than two years from diagnosis. The tumor microenvironment plays a major role in GBM progression through sustaining immunosuppression and poor lymphocytic infiltration. Tertiary lymphoid structures (TLS) are ectopic lymphoid aggregates that form in inflamed tissues and are associated with positive prognosis in numerous cancers outside the central nervous system. Prior to this work, TLS had not been reported or studied in GBM. In this thesis, we aimed to characterize TLS in glioma patients, and to investigate immunotherapeutic approaches that could reprogram the GBM microenvironment to induce these structures, promote anti-tumor responses and prolong survival.

In Paper I, we discovered the presence of TLS in low grade and high grade glioma tissues, and found that they correlated with increased T cell infiltration inside the tumor. Moreover, we demonstrated that agonistic CD40 therapy (αCD40) induced the formation of TLS with a follicle-like organization in murine glioma models. αCD40 also promoted a population of CD11b⁺ regulatory B cells, which inhibited T cell activation. These cells were not present within the TLS, indicating that TLS formation and the induction of CD11b⁺ B cells were likely two independent processes.

In Paper II, we employed murine glioma models to study the therapeutic effect of cytokines involved in lymphoid tissue development, and selected LIGHT as the most promising candidate. To therapeutically deliver LIGHT to the tumor microenvironment, we engineered an AAV vector targeted to the brain endothelial cells to express LIGHT (AAV-LIGHT). AAV-LIGHT promoted the formation of TLS and functional high endothelial venules. Moreover, AAV-LIGHT strengthened effector and memory CD8⁺ T cell responses, and boosted a population of TCF1⁺CD8⁺ stem-like T cells. This was associated with a prolonged survival, indicating that AAV-LIGHT is a promising therapeutic candidate for the treatment of GBM. 

In Paper III, we coupled advanced spatial transcriptomics of human GBM tissue and time point experiments in murine glioma models to investigate the stages of TLS assembly. We found that TLS formation is a step-wise process, where each stage is characterized by specific cell components and pathways. Understanding the steps underlying TLS assembly will be critical to develop efficient TLS-inducing immunotherapies.

Patients diagnosed with glioblastoma (GBM) have an extremely poor prognosis despite extensive research in the recent years, with a five-year overall survival of only 7%. There is a multitude of barriers to effective treatment in GBM, including low levels of T cell infiltration. Chronic inflammation in cancer can lead to the formation of ectopic clusters of lymphoid tissue known as tertiary lymphoid structures (TLS), as well as specialised blood vessels called tumour-associated high endothelial venules (TA-HEVs), which are involved in the extravasation of different immune cell subsets. Both TLS and TA-HEVs have been associated with increased immune cell infiltration and prolonged survival in numerous peripheral cancers. In this work, we aimed to improve the anti-glioma immune response by increasing the infiltration of T cells through induction of TLS and TA-HEVs. Further, we aimed to deeply explore the characteristics and roles of these two components in the GBM setting. 

In Paper I, we pre-screened a panel of lymphoneogenic cytokines and selected LIGHT as the most promising therapeutic candidate. We then targeted LIGHT to the brain endothelial cells of glioma-bearing mice utilising an adeno-associated viral (AAV) vector (AAV-LIGHT). We found that this approach resulted in prolonged survival in association with the formation of TLS and functional MAdCAM-1⁺ TA-HEVs, as well as the promotion of effector/memory T cell responses and the TCF1⁺CD8⁺ stem-like T cell population. 

In Paper II, we correlated the presence of TLS with improved survival in GBM patients. We then combined advanced spatial transcriptomics techniques with functional studies in murine glioma to establish the developmental timeline of TLS in GBM, including the critical role of CCR7⁺CD4⁺ T cells with LTi-like features. 

In Paper III, we employed a murine knockout model of MAdCAM-1 to demonstrate its necessity in the anti-glioma immune response through the promotion of TCF1⁺CD8⁺ stem-like T cells. Furthermore, we identified the presence of TA-HEVs in human GBM in association with infiltrating T cells.

Altogether, this work uncovers previously unknown mechanisms and highlights the potential of targeting the immune/vascular interface as a therapeutic option for GBM.