Year: 2021
Type: article
Abstract: In recent years, magnetic resonance spectroscopy (MRS) has begun to reveal its potential as a tool to help answer key problems in neuro-oncology in a non-invasive fashion. The ability to non-invasively quantitate levels of numerous compounds in and around brain tumors in vivo can address numerous challenges, with the initial advances potentially yielding diagnostic benefits. While not yet providing definitive answers in most cases, MRS has for years been helping to distinguish treatment-related necrosis from brain tumor progression. More recently, reports have indicated that MRS can non-invasively indicate the presence of IDH1/2 mutations in brain tumors through detection of elevated 2-hydroxyglutarate,1 the product of the mutated gene—potentially obviating the need for the most dangerous biopsies. A new report by Subramani et al. takes MRS in an exciting new direction, showing its potential to act as an indicator of treatment efficacy through detection of elevated glutamate as a marker of temozolomide chemotherapeutic activity against IDH-mutated gliomas.2 The authors used both genetically engineered and patient-derived models of IDH-mutant glioma (mostly lower grade) and treated with temozolomide, testing for metabolic changes with MRS and chemometrics. They showed that glutamate production rose markedly, with increased input both from glucose via the TCA cycle and from glutamine. Importantly, 1H MRS successfully detected increased glutamate levels in vivo in IDH-mutant gliomas upon treatment with temozolomide, and this elevation in glutamate preceded a decrease in tumor size. Another metabolic MRS approach, hyperpolarized 13C MRS detecting 13C flux from pyruvate and alpha-ketoglutarate to glutamate, also served to indicate temozolomide activity in IDH-mutant gliomas. Subramani et al. propose that MRS detection of glutamate could therefore serve as an early and non-invasive metabolic biomarker to gauge temozolomide effectiveness in this important subset of gliomas. These findings extend the capabilities of MRS in a new functional direction, potentially enabling its use as an early gauge of therapeutic efficacy in IDH-mutant gliomas. The work also suggests a number of questions and other applications. It begs the questions of whether temozolomide activity against IDH wild-type gliomas or other brain tumors could also be gauged similarly—and whether this has applicability to other treatments as well. The glutamate receptor inhibitor memantine has shown protective activity from the detrimental effects of whole-brain radiation therapy,3 suggesting that glutamate release from either brain tumors and/or normal brain may mediate excitotoxic radiation brain injury; one could speculate that MRS measurement of glutamate levels—assuming it is released from glioma cells as well as intracellularly—early in whole-brain or focal radiation treatment might provide an early indicator of future toxicity and a need for memantine treatment. These and other provocative questions arise from this intriguing study, suggesting the need for extensive follow-up studies. Interleukin-33 (IL-33) is a pleiotropic cytokine involved in tissue and metabolic homeostasis, infection, inflammation, cancer, and diseases of the central nervous system (CNS). IL-33 promotes these effects by translocating to the nucleus and binding to chromatin to regulate gene transcription or by binding to its cell surface receptor ST2 and activating its downstream signaling pathways. The CNS has the highest levels of IL-33 expression, where it is constitutively expressed in microglia, astrocytes, and oligodendrocytes, whereas ST2 is expressed by microglia. IL-33 is now well recognized to have a role in innate and adaptive immunity.1 In cancer, including glioma, recent evidence points to a pro-tumorigenic role for IL-33, where IL-33 levels of expression correlate with poorer prognoses and accelerated tumor growth. In a recent article from De Boeck and colleagues,2 the authors demonstrate that IL-33 acts as a major regulator of the brain tumor microenvironment. They observed that nuclear and soluble forms of IL-33 are present in a subset of glioblastoma (GBM) specimens and their derivative brain tumor-initiating cell (BTIC) cultures. Through the use of an impressive array of models and experimental approaches, the authors demonstrate that secreted IL-33 alters resident microglia and recruits and activates bone marrow-derived macrophages, natural killer cells, and T cells and that expression of IL-33 correlates with increased TAM infiltration and decreases the overall survival in GBM patients. The authors also show that nuclear IL- 33 is essential for its role in tumor growth acceleration through the recruitment of pro-tumorigenic macrophages to the GBM microenvironment. Collectively, the mechanistic data generated by De Broeck et al. support the notion that IL-33 is a major orchestrator of the GBM tumor environment that facilitates tumor progression and offer a novel target toward a therapeutic strategy to improve patient outcomes. Proinflammatory cytokines or immunocytokines (antibody-cytokine fusions) may have the potential to turn cold tumors hot, but systemic toxicity limits their clinical use. Targeted delivery to the tumor site, however, may avoid these systemic toxicities. A recent study investigated three L19-based immunocytokines, L19-IL2, L19-mTNF, and L19-mIL12, that have L19 antibody tagged to human interleukin-2 (IL2), mouse tumor necrosis factor (TNF), or mouse interleukin-12 (IL12), respectively.1 These immunocytokines were administered to animal models and also human patients. The L19 antibody is specific to a tumor-associated epitope of fibronectin, an extracellular matrix protein found exclusively in the stroma of WHO grade III and IV astrocytomas. The L19-cytokine fusion molecule is designed to deliver cytokines directly to glioblastoma after systemic administration of an antibody against a glioblastoma-specific stromal protein. They selected L19-mTNF because it has already been developed to a clinical-grade level, can be safely administered to patients with extracranial malignancies, and its biological activity is attractive from a pharmacodynamic point of view. The resulting fusion proteins that were delivered intravenously appeared in the tumor microenvironment of two orthotopic immunocompetent mouse glioma models, SMA-560 and GL-261, but also in the pilot study of three human patients. Both L19-mTNF and L19-mIL12 led to an increase in other proinflammatory cytokines, such as IL1, IL6, and GM-CSF, within the tumor microenvironment. These can potentially promote the local antitumor immune response, but there have also been reports indicating immune-independent, pro-tumorigenic effects of IL6 and GM-CSF in glioblastoma. Intravenous administration of L19-mIL12 or L19-mTNF cured a proportion of tumor-bearing mice, whereas L19-IL2 did not. L19-mIL12 or L19-mTNF also promoted infiltration of tumor-infiltrating lymphocytes and increased the amounts of proinflammatory cytokines within the tumor microenvironment. This therapeutic activity was abolished in RAG−/− mice or upon depletion of CD4 or CD8 T cells, suggesting adaptive immunity. In addition, the study reported increased amounts of proinflammatory cytokines and tumor-infiltrating lymphocytes after treatment with L19-mTNF or L19-mIL12. The L19-TNF fusion protein was given to 3 patients with recurrent glioblastoma after a standard-of-care treatment comprising surgery, radiotherapy, and chemotherapy with temozolomide. After two doses of L19-TNF, they observed a regional intratumoral decrease of relative blood volume supply, as evidenced by dynamic susceptibility contrast perfusion MRI, and a necrotic area within the tumor on T1-weighted contrast-enhanced images. The reduction in blood volume supply and the necrotic area increased with subsequent doses of L19-TNF. The authors concluded that their approach converted the immunosuppressive glioma microenvironment into a proinflammatory phenotype. However, the clinical study included only three patients, and these patients were treated with one dose level of L19-TNF, which allows a safety statement only for the specific dose level and therapeutic window monitored. Also, the MRI findings, especially in such a small number of patients, can be misleading. They had tissue available from only one patient at recurrence after treatment with L19-TNF for the comparison with the tissue obtained at primary diagnosis. Despite these possible limitations, this work provides a basis for future developments using immunocytokines against malignant brain tumors, either as single agents or in combination with other treatment modalities. Of late, it seems you cannot open a journal without mention of precision medicine, an approach for disease treatment and prevention taking into account individuals’ genetic variations. So, it’s not surprising there is a movement to apply this method to tumors like medulloblastoma, for which we have yet to identify effective, long-term, subgroup-specific therapies for certain patient groups despite genomic advances in the characterization of these subgroups. Ideally, this approach would allow for more accurate treatment of a patient’s specific tumor in contrast to our historical one-size-fits-all approach of surgery, cranial-spinal irradiation (except in very young children), and multi-agent chemotherapy regimens with less consideration given to differences between individual tumors. Interestingly, Rusert and colleagues in their recent publication in Cancer Research hypothesize that each patient with medulloblastoma is sensitive to different therapies and seek to tailor those treatments based on genetic and cellular variations of patients’ tumors to improve outcomes.1 Using a panel of patient-derived xenografts (PDXs), the authors used a combinatorial approach including DNA sequencing, RNA sequencing, and empirical drug screening to identify potential therapeutic options in real time. Notably, they were able to show that PDX responses varied to current standard-of-care chemotherapeutic agents and in doing so, suggested actinomycin D, a drug not generally used for medulloblastoma, is active in PDXs representing group 3 medulloblastoma, a subgroup with one of the worst outcomes. Moreover, they were able to show such an approach is feasible in the clinical setting. Patients underwent surgical resection of their tumors. Excess tissue not required for diagnostic purposes was processed for sequencing and drug screening both for newly diagnosed or recurrent tumors. Sequencing data was not available immediately, but drug screening was completed successfully within a few days. With respect to precision medicine, it is all very well to believe that a patient’s histological and genomic features may suggest an optimal individualized treatment plan from available polypharmacy, but the well-described heterogeneity of this disease at diagnosis or progression, the complexity of how these genetic changes interact, and our inability to predict antineoplastic treatment responses using traditional approaches pose a number of challenges in the development of therapeutic algorithms to provide truly individualized care. At the same time, we depend upon population-level data to suggest how different therapies may interact with each other, as it is not yet possible to recognize these effects in an individual patient. Moreover, combining two agents identified as potential treatments may lead to antagonistic as opposed to synergistic responses. The question whether this personalized approach to treating medulloblastoma will impact not only our therapeutic options but also outcomes for patients in the future remains to be answered. Meningiomas traditionally have taken a backseat to gliomas in terms of both clinical and basic science research. Consequently, our knowledge about the molecular drivers in meningioma has been limited. Recently, several papers exploring molecular drivers of meningioma have been published,1,2 and at least one national clinical trial using different genomically driven chemotherapy options (Alliance A071401) has been launched. While most meningiomas are considered benign tumors, life expectancy can still be diminished, particularly in higher-grade meningiomas, so a better understanding of the biological landscape of meningiomas is important. Magill et al combined genomic profiling with MRI to identify mechanisms underlying intratumor heterogeneity in meningioma.3 In 13 human meningiomas, they performed molecular profiling in 86 spatially distinct samples. The samples included all meningioma grades (I-III), and there were distinct patterns of gene enrichment across the 3 grades, with Grade III meningiomas having more significant transcriptomic and epigenomic intratumoral heterogeneity. In particular, copy number variation was uniform across Grade I meningiomas but was more heterogeneous in higher-grade meningiomas. Biopsy areas were registered to ADC maps, and the authors found that areas with high ADC values were characterized by developmental gene expression programs (ex. Wnt, ectoderm, pluripotency, and neural crest gene expression programs). Whole-tumor ADC was more heterogeneous in high-grade meningiomas as well, suggesting increased variability in tumor cytoarchitecture. The authors went on to validate these findings in a human cerebral organoid model of meningioma and found that the genes CDH2 and PTPRZ1 were enriched in high-ADC regions and can drive meningioma cell proliferation and tumorigenesis. In sum, this study sheds more light on the intratumoral heterogeneity of meningiomas using both genomic and radiographic features and provides evidence for additional therapeutic targets. More patients are needed to confirm these results as well as tumors from more diverse locations; all the tumors were along the midline, supratentorially. Prior studies have shown regional predilection of genetically distinct meningiomas, so these results may not generalize to tumors in other locations. Regardless, this bed-to-bench work combining genomic characterization, a routine MRI sequence, and human organoid models is a paradigm that can be repeated with other biological questions in meningiomas and other brain tumors.
Source: Neuro-Oncology
Language: English
Subfield: Genetics
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4210583333