By Evan Noch, MD, PhD
Instructor, Department of Neurology, Weill Cornell Medical College
SYNOPSIS: Glioblastoma-associated immunosuppression is a significant factor associated with poor survival in this disease. Accumulating evidence suggests that mouse models of glioblastoma and other brain cancers induce systemic immunosuppression through dysregulation of a newly recognized brain-thymus axis and that targeting this pathway may promote more effective immune surveillance of these tumors.
SOURCE: Ayasoufi K, Pfaller CK, Evgin L, et al. Brain cancer induces systemic immunosuppression through release of non-steroid soluble mediators. Brain 2020;143:3629-3652.
Systemic immunosuppression is a strong deter-minant of poor prognosis in a variety of human cancers. Glioblastoma patients exhibit decreased immune surveillance, mediated by dysregulation of T-cell repertoires and reduced expression of major histocompatibility complex class II (MHCII) expression, leading to unchecked tumor growth. Consequently, although immunotherapy to promote immune-mediated attack and clearance of tumor cells has benefited other cancer patients, immunotherapy has not shown any long-term benefit in glioblastoma. The biological pathways that mediate systemic immunosuppression in glioblastoma remain incompletely understood, and therapeutic targets to promote immune cell retention and function are unknown.
Ayasoufi et al demonstrated a glioblastoma-thymus-adrenal gland axis that regulates systemic immunosuppression in response to brain tumor and brain injury models. They used a variety of brain tumor models, including the GL261 murine glioma model, a transgenic murine diffuse intrinsic pontine glioma model, and an intracranial B16 murine melanoma model. Tumors from each of these models significantly reduced thymus size and cellularity. Functionally, thymic T-cells from these mice were reduced and exhibited decreased late-stage proliferation, while B-cell numbers were increased in atrophic thymus. Next-generation sequencing of thymus from GL261-bearing mice demonstrated upregulation of T-cell proliferation programs but downregulation of deoxyribonucleic acid (DNA) replication and elongation, providing a likely cause of T-cell loss in these models. Interestingly, pathway analysis showed that ribonucleic acid (RNA) molecules associated with cancer development, particularly that of gliomas, were found among thymic bulk RNA transcripts.
To investigate systemic immunosuppression associated with brain tumor development, the authors measured immune cell numbers and phenotypes in tumor-bearing mice. They found that CD4+ and CD8+ T-cells and MHCII-expressing B-cells and monocytes/macrophages were reduced in these mice, indicating deficits in both innate and adaptive immunity. Using the technique of parabiosis to study circulating factors mediating brain tumor-induced immunosuppression, the authors joined GFP-expressing C57BL/6 mice to wild-type C57BL/6 mice. Although they implanted GL261 cells into the brain of only one parabiont, both mice exhibited thymic involution, reduced peripheral blood CD4+ T-cell counts, and decreased MHCII expression levels. Reinforcing the findings of the parabiotic model, the authors found that serum derived from GL261-bearing mice, but not from GL261 cells grown in vitro, inhibited T-cell proliferation in culture systems.
Expanding their observations from gliomagenesis to models of brain injury, the authors found that thymic involution occurs in response to experimental models of demyelination, lipopolysaccharide-induced neuroinflammation, kainic acid-induced seizures, and even mechanical injury mediated by intracranial phosphate-buffered saline injection. When mice recovered from these acute injuries, the thymus recovered also, resembling those of naïve control mice. Similar to the deleterious effects of sera from tumor-bearing mice on T-cell proliferation in vitro, the researchers found that sera from acutely injured but not recovered mice robustly inhibited T-cell proliferation. These findings indicate the presence of soluble factors that induce immunosuppression in response to brain injury.
Since stress hormones produced by the adrenal glands may mediate the sequelae of brain injury, the investigators studied tumor-induced immunosuppression in adrenalectomized mice. Surprisingly, they found that adrenalectomized mice exhibited increased numbers, but not frequencies, of immune cells in the blood, thymus, and spleen. When they implanted GL261 cells into the brains of mice, they found no evidence of thymic involution, disruptions in T-cell development, or sequestration of T-cells in the bone marrow. However, this rescue of immune function in adrenalectomized mice did not affect overall survival.
To determine the soluble factors that mediate immunosuppression, they filtered sera from glioma-bearing mice based on molecular weight and added this filtrate to cultured T-cells. Large factors with molecular weights greater than 100 kDa inhibited T-cell growth, but small molecules less than 3 kDa were not immunosuppressive. To confirm that cortisol, a small molecule stress hormone, does not induce adrenal-mediated reduction in thymic function, the researchers tested the effects of serum from wild-type and adrenalectomized mice on T-cell growth in vitro. Serum from both mouse cohorts was equally immunosuppressive, but serum from naïve adrenalectomized mice did not inhibit T-cell growth. These findings suggest that, although adrenalectomized tumor-bearing mice do not show thymic phenotypic changes, these mice exhibit thymus-independent immunosuppression from the action of a large molecular weight soluble factor associated with poor survival.
COMMENTARY
Immunosuppression in glioblastoma facilitates tumor escape from immune surveillance. By establishing a new brain-thymus-adrenal gland axis through which brain tumors and brain injury mediate thymic involution, decreased peripheral immune cell counts, and the circulation of immunosuppressive serum-derived soluble factors, the authors of this study have identified a putative target to improve immune function in these patients. Such targeted treatments also could improve the efficacy of systemic immunotherapy in this disease. The identity of the large molecular weight soluble factors mediating immunosuppression remain unknown, so future studies should use a proteomics approach to identify and modulate the effects of these factors to establish their role in brain tumor-associated systemic immune cell dysfunction.
