Year: 2020
Type: article
Abstract: Despite genomic advances in the characterization of medulloblastoma and ependymoma subgroups, we have yet to identify effective, long-term, subgroup-specific therapies. Superimpose upon this lack of targeted agents under clinical investigation the growing interest in empowering your own immune system to recognize and destroy cancer cells, and you may have a recipe for future promising approaches. Chimeric antigen receptor T cells, known as CARs, provide an immunotherapeutic approach to reprogramming patients’ T cells with the addition of a CAR to fight the cancer in question. Donovan and colleagues attempt to harness this method through locoregional delivery for the treatment of metastatic medulloblastoma and ependymoma in the May edition of Nature Medicine.1 The authors selected a number of normal, developing brain, agnostic, cell-surface targets expressed in both medulloblastomas and ependymomas by analyzing tumor microarray data and RNA expression profiles, respectively, from representative samples. They went on to test locoregional CSF and intravenous delivery of CAR T cells targeting EPHA2, HER2, and interleukin 13 receptor α2 (IL-13Rα2) in orthotopic patient-derived cerebellar xenografts (PDX) of group 3 medulloblastoma and posterior fossa group A (PFA) ependymoma PDX models. Their data in PDX models for both disease subtypes suggest a potential role for CAR T cells. Intraventricular delivery of monovalent EPHA2 CAR T cells showed a significant increase in survival compared with CAR T cells transduced with all three epitopes in group 3 medulloblastoma, which was in contrast to that of PFA ependymoma in which TRI CAR T cells fared better with regard to model outcomes. Importantly, anatomic location of these tumors adjacent to the CSF spaces clearly benefitted from an intraventricular approach to delivery. Furthermore, the authors noted clear synergy between intraventricular CSF CAR T cell therapy and azacytidine, known to induce cancer cell death dependent upon epigenetic silencing while transiently increasing expression of tumor-associated antigens and sensitizing tumor cells to cytotoxic T cells. This combined approached may provide an effective neoadjuvant approach to therapy going forward. The presence of bacteria in tumors has been known for some time, but a comprehensive characterization of the tumor microbiome has not yet been reported. While the gut microbiome has been shown to influence malignancy parameters, the composition and role of the tumor microbiome are less well understood. A recent study published in the journal Science details the first characterization of the microbiome in several human cancers, including glioblastoma (GBM).1 The study investigated the microbiomes of seven human cancers (melanoma, lung, ovary, breast, pancreas, bone, and glioblastoma). A total of 1010 tumor samples and 516 corresponding normal tissue samples were profiled using several complementary assays and stringent controls to eliminate experimental bacterial contamination. Bacteria were detected in all tumor types in proportions varying from 14% in melanomas to 44% in GBM and 63% in breast tumors. The bacteria were found in cancer cells and immune cells and were localized intracellularly. The bacteria were at least partially alive as demonstrated by culture experiments. A total of 9190 species of bacteria were detected in tumor and normal tissues. The microbiomes of tumors of the same type were more similar to each other than they were to the microbiomes of other tumor types. There were differences in bacterial diversity between the different tumor types. Breast tumors had the most diverse microbiomes with an average of 16 bacterial species per tumor, wheres melanoma had the lowest diversity; GBM had the third highest. There were also differences in bacterial load and diversity between tumor types and corresponding normal tissues. While breast tumors had a greater load and more diverse bacteria than adjacent normal breast, this was not observed in lung and ovarian tumors (GBM was not investigated). To examine potential functional roles of the tumor microbiome, genes and pathways that the bacterial species may harbor were mapped and analyzed. A few examples of putative functions were uncovered. In lung cancer, bacteria that degrade cigarette smoke metabolites were enriched in smokers. These bacteria were enriched for pathways that degrade chemicals such as nicotine, anthranilate, toluene and phenol. In ER+ breast tumors, known to have increased oxidative stress, there was an enrichment of bacterial pathways for the synthesis of the reactive oxygen species-detoxifying molecule mycothiol. These and other findings suggest functional roles for the tumor microbiome. Altogether, the study demonstrates the presence of intracellular live bacteria in tumor tissues. This tumor microbiome varies in different tumor types and subtypes and has the potential to influence malignancy. The functions of these microbiomes and whether they can be manipulated for cancer therapy remain to be investigated. Isocitrate dehydrogenase (IDH) mutations 1 and 2 are found in a diverse group of cancers including gliomas. IDH-1 and less commonly IDH-2 mutations are present in more than 70% of low-grade gliomas (LGGs). These mutations facilitate conversion of α-ketoglutarate to 2-hydroxyglutarate that in turn, activates important oncogenic signaling pathways. IDH mutant gliomas have unique molecular features that occur early during glioma oncogenesis, but these mutations do confer a better prognosis compared with IDH wildtype gliomas. Standard treatment of LGG consists of radiation and chemotherapy. To date, there are no approved molecularly targeted therapies for LGG. Hence, IDH mutations represent a promising therapeutic opportunity. Mellinghoff et al recently reported a multicenter, open-label, phase I, dose escalation and expansion study investigating the oral IDH1 inhibitor ivosidenib in 66 patients with advanced IDH1 mutated gliomas.1 Ivosidenib was well tolerated in the study population. Although the maximum tolerated dose was not reached, a dose of 500 mg continuous daily dosing was chosen for the expansion phase. The grade 3 or higher adverse events rate was 19.7% (n = 13); 3% (n = 2) were considered treatment related. The objective response rate in non-enhancing gliomas (n = 35) was 2.9%, with 1 partial response. Thirty of 35 patients (85.7%) with non-enhancing glioma achieved stable disease compared with 14 of 31 (45.2%) in patients with enhancing glioma. Median progression-free survival (PFS) was 13.6 months (95% CI, 9.2 to 33.2 months) and 1.4 months (95% CI, 1.0 to 1.9 months) for the non-enhancing and enhancing LGGs, respectively. The authors noted that median PFS for patients with non-enhancing gliomas was comparable to the benefit of temozolomide in advanced IDH1 mutated LGG (approximately 7 months). Conventional 2-dimensional measurements may not capture accurate changes in the irregularly shaped, treated brain tumor volumes. The investigators explored volumetric measurements to provide greater confidence and accuracy. Using this approach, they reported tumor shrinkage in several patients. The authors cautioned that the wide applicability of volumetric assessments would require harmonization of image acquisition and analysis as well as regulatory guidance. They also postulated that ivosidenib may be more effective in non-enhancing gliomas than enhancing gliomas, because these tumors represent an earlier disease stage with fewer genetic alterations and are hence less likely to be therapy resistant. Indeed, in another exploratory analysis, changes in cell cycle genes (alterations that are associated with LGG progression) were associated with shorter PFS within the subgroup of non-enhancing gliomas. The trial was relatively small and had several other shortcomings, including a heterogeneous patient population, lack of control in a nonrandomized design, and no central pathology review. Nonetheless, these preliminary safety and efficacy results will further encourage clinical development of IDH inhibitors for IDH mutated LGGs, with more early phase, combination, and larger randomized studies. Brain metastasis is a deadly complication for many patients suffering from the most common malignancies, including lung, breast, and renal cell cancers. Recent profiling studies have revealed that significant and clinically relevant genomic heterogeneity exists across brain metastases and case-matched primaries, with precise patterns of alterations highly dependent upon tumor type. Nevertheless, specific mutations and/or DNA copy number abnormalities are likely insufficient to drive brain metastasis alone, raising the possibility that other mechanisms involving metabolic adaptation to the unique central nervous system (CNS) microenvironment play a central role. As described in their recent Cancer Discovery article, Ngo, Kim, and Osorio-Vasquez et al. investigated the extent to which constraints in the availability of key amino acids within the CNS microenvironment influence brain metastasis.1 Using multiple cell line models, followed by validation in human tissue samples, they found that brain metastases and brain metastatic cell line derivatives exhibited increased levels of the enzyme 3-phosphoglycerate dehydrogenase (PHGDH), which catalyzes the rate-limiting step of serine biosynthesis. They then confirmed that serine and glycine are among the least abundant amino acids in both brain interstitial fluid and cerebrospinal fluid. They went on to show that aggressive brain metastatic cell lines proliferate effectively in serine- and glycine-deficient media, whereas more indolent isogenic derivatives do not, a distinction dependent on PHGDH expression levels. PHGDH knockdown suppressed brain metastasis in multiple models and forced expression of PHGDH-enhanced brain metastasis, while having minimal effects on lung metastasis. Finally, the authors demonstrated that pharmacological inhibition of PHGDH suppressed cell growth in serine- and glycine-limited environments and attenuated brain metastasis, exerting toxic effects through the depletion of nucleotides, downstream metabolic products of serine biosynthesis, thereby enhancing DNA damage. This paper is highly significant in that it firmly implicates distinct and targetable metabolic pathways in the process of brain metastasis, pointing toward viable therapeutic strategies with promise across the spectrum of brain metastatic cancer variants. To accurately diagnose a brain tumor, neurosurgery is required with its associated potential risks. While surgery can be both diagnostic and therapeutic for many types of brain tumors, a noninvasive tool to diagnosis brain tumors could aid in surgical planning and longitudinal monitoring of response to treatment. Therefore, there has been a lot of interest in evaluating circulating tumor DNA (ctDNA) measured in plasma or cerebrospinal fluid. The utility of ctDNA has been limited due to its rarity in plasma, reducing the sensitivity of measuring ctDNA. In addition, distinguishing ctDNA from normal circulating DNA can be challenging. DNA methylation alterations in ctDNA have been shown to reliably identify extracranial cancers despite low plasma presence. Nassiri et al. developed an approach using cell-free methylated DNA immunoprecipitation and high-throughput sequencing (cfMeDIP-seq) to recover and profile methylated DNA fragments from plasma in patients with brain tumors.1 Using a dataset of 447 patient samples and a machine learning approach, they were able to identify patients with gliomas and distinguish them from healthy patients or patients with alternative cancers. The high sensitivity of detection was maintained regardless of IDH status or tumor grade. The gliomas were hypermethylated compared with other systemic cancers or healthy controls, and the cfMeDIP-seq signal correlated with tumor tissue methylation status, suggesting that the signal detected in plasma was coming from glioma tissue. Distinguishing individual types of brain tumors can potentially be more tricky because the cell of origin can be similar for different brain tumor types. The authors suggested that their cfMeDIP-seq approach could reliably distinguish between different intracranial tumors – something that is challenging for MRI. The role of liquid biopsies continues to evolve and these results await validation in larger studies but appear promising as one novel tool to help diagnose intracranial tumors.
Source: Neuro-Oncology
Language: English
Topic: CAR-T cell therapy research
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4251420529