Year: 2018
Type: article
Abstract: A prestigious team of researchers from Duke has tested a recombinant nonpathogenic polio–rhinovirus chimera (PVSRIPO) in an uncontrolled trial involving 61 patients with progression of a glioblastoma after standard of care plus potential salvage treatments. The virus chimera was adapted so that it infects cells via the poliovirus receptor CD155 without exerting neurotoxicity.1 CD155 has been demonstrated to be highly expressed on several malignant cells.2 Patients received an intratumoral infusion of PVSRIPO by means of convection-enhanced delivery over a period of 6.5 hours using catheters and infusion devices. The virus suspension volume was 3.25 ml. The principal mechanism is an immune reaction against virally infected cells. The primary objectives of the trial were toxicity and the optimal phase II trial dose. There was a preplanned secondary objective of comparing overall survival in the trial population with matched historic controls. A dose of PVSRIPO 5 × 107 (tissue culture infectious dose, TCID50) was identified as the phase II dose. A grade 4 intracranial hemorrhage as dose-limiting toxicity was identified at a dose of 1010 TCID50 as a potential consequence of catheter removal. Most patients in the dose-expansion cohort (involving 52/61 patients) received further anti-tumor treatment, often bevacizumab, a drug known to systematically impact progression-free survival and imaging, but not overall survival.3 Thirteen of 61 patients treated are had an overall survival of 24 months or more after treatment. The authors describe a plateau, and instead of detailing the frequency of isocitrate dehydrogenase (IDH)-mutated tumors amongst these patients, they refer to a minor impact of the IDH status at recurrence. The historical control group derived from the same side consists of 104 patients with similar but overall less favorable patient characteristics (e.g. O6-methylguanine DNA-methyltransferase and IDH status). In summary, the trial provides good evidence for the feasibility of the approach, determines a phase II dose, and nourishes initial hopes on efficacy for a (small) subgroup of patients that awaits characterization. Why would an attractive, although preliminary, approach be published in one of the most prestigious medical journals? Relevance is derived from the appeal of the potential virus-induced immune response and, most importantly, from the historical comparator. However, as compared to an active control, inclusion in a trial – e.g. the presence of real progression - is subject to uncertainty; the long interval from inclusion in one of the largest brain tumor sites in the United States also speaks for heavy selection. Until proven otherwise, a plateau in our most aggressive disease is due to favorable prognostic (molecular) factors; here the work remains too vague and fails to disclose the number of IDH-mutated tumors in the group building the plateau. Whether or not this is also an effect of one of the most commonly administered salvage treatments, bevacizumab, remains uncertain; however, the limited options for the historical controls (included most commonly after bevacizumab) and the rapid progression several patients experienced after bevacizumab failure need careful consideration. There are already rumors circulating about conditional approval based on these data. Certainly, neuro-oncology is one of the areas with the greatest medical need. On the other hand, the most prominent examples of failure to translate an uncontrolled phase II result4,5 into an overall survival benefit for patients with recurrent glioblastoma started likewise with a presumed favorable comparison to a historical control and shortcomings in exploring the potential mechanism of action in glioblastoma.3,6 For PVSRIPO the connection to immunotherapies guarantee the awareness, but the relative impact of the oncolytic properties remains to be determined. What are the next steps? PVSRIPO received a “breakthrough” designation from the Food and Drug Administration of the United States. It seems there will be a long and stormy road ahead from this label - fostering further rapid development - to a real breakthrough in the treatment of patients with recurrent glioblastoma. However, the potential may be there. The diagnostic accuracy of O-(2-[18F]fluoroethyl)-l-tyrosine (FET) PET in detecting the malignant transformation of low-grade gliomas (LGGs) has been suggested but still not proven.1 In the April issue of Journal of Neurosurgery Bashir and colleagues2 report a retrospective assessment of the diagnostic value of FET PET in patients with MRI-suspected malignant progression of LGGs that had previously been treated and the relationship between FET uptake and MRI and molecular biomarkers. Forty-two patients with histologically verified LGGs were included. Maximal and mean tumor-to-brain ratios (TBRmax and TBRmean, respectively) on FET PET as well as kinetic FET PET parameters (time to peak [TTP] and time-activity curve [TAC]) were determined. Final diagnoses were confirmed histologically. The diagnostic accuracy of FET parameters was evaluated using receiver operating characteristic (ROC) curve analysis. No significant association between FET parameters and contrast enhancement on MRI was observed. ROC curve analyses were unable to demonstrate any significant differentiation between non-transformed LGGs and LGGs that had transformed to a higher grade when evaluating FET parameters separately or combined. After excluding the oligodendroglial subgroup, a significant difference was observed between non-transformed and transformed LGGs when combining FET parameters (i.e., TBRmax > 1.6, TAC describing a plateau or decreasing pattern, and TTP <25 minutes), with the best result yielded by a combined analysis of TBRmax > 1.6 and TAC (P = 0.003). The difference was even greater when patients who had previously undergone oncological treatment were excluded (P = 0.001). The presence of an oligodendroglial component (P = 0.029), previous oncological treatment (P = 0.039), and the combined FET parameters (P = 0.027) were significant confounding factors in the detection of malignant progression. TBRmax was inversely correlated with IDH1 mutation (P = 0.006). In conclusion, FET PET imaging alone is not adequate to replace histological confirmation, but it may be useful in the detection of malignant transformation in astrocytomas. Gliomas can migrate along blood vessels as single cells, called co-option, or as a perivascular group of cells. Developmental studies have shown that the two major cell types (oligodendrocyte precursor cells [OPCs] or astrocytes) that are putative cells of origin in glioma can show differential modes of migration. Whereas OPCs utilize Wnt7-CXCR4 signaling for migration along embryonic blood vessels, astrocytes play important roles in the maintenance of tight junctions in the blood-brain barrier (BBB). To better understand the role of these glial cell types (OPCs versus astrocytes) in the regulation of the vasculature in glioma, a recent study by Griveau et al utilized isogenic mouse models in which tumors are formed that contain a functional allele of Olig2 in the backbone of Trp53 loss and EGRVIII transgene (Olig2+) or lack Olig2 in the same backbone (Olig2-).1 Gliomas formed with equal efficiency in both backgrounds, and gene expression profiling demonstrated that Olig2+ and Olig2- tumors enriched for OPC and astrocytic genes, respectively. The authors noted, however, that the tumor growth around the vasculature between the tumor types exhibited differences. Olig2+ gliomas migrated as single cells along blood vessels in the invasive edge, whereas Olig2- gliomas showed dense perivascular distribution. Importantly, Olig2- gliomas showed defects in vascular basement membrane, BBB disruption, microglial activation, and increased infiltration of macrophages. Similar to the OPCs, the Olig2+ gliomas showed higher levels of Wnt7a and Wnt7b. Deletion of Wnt7a and Wnt7b in organotypic cortical explant co-cultures showed that Wnt7 signaling is required for single-cell co-option driven by Olig2. Using chromatin immunoprecipitation techniques, the authors demonstrated that Olig2 positively regulates Wnt7 via promoter occupancy and competes with TP53, which negatively regulates Wnt7. The positive correlation of Wnt7 expression with OPC lineage and inverse correlation with TP53 was confirmed in human gliomas. Inhibition of canonical Wnt signaling using the Wnt inhibitor XAV939 or porcupine inhibitor LGK974 prevented glioma single-cell migration in mouse models. Importantly, in a patient-derived proneural xenograft model that faithfully recapitulates Olig2-driven single-cell migration, treatment with LGK974 significantly improved survival when combined with temozolamide as compared with single-agent treatments. To further strengthen the clinical relevance of these findings, the authors investigated the expression of VEGF, since it has been shown that anti-VEGF treatments can promote vessel co-option. In mouse models, isogenic U87 models that are sensitive and resistant to bevacizumab (a VEGF inhibitor), as well as matched pre- and post-treatment human samples, anti-VEGF therapy caused enrichment of OPC features and increased Olig2 expression and Wnt7 signaling. This study not only clarifies the molecular underpinnings of glioma-vascular interactions but also identifies Wnt inhibition as a therapeutic target for tumors that escape anti-angiogenic therapies. Of note, the Wnt pathway inhibitor LGK974 was shown to cross the BBB in this study, and future clinical studies are warranted for this agent. Large-scale analyses of genomic alterations in glioblastoma (GBM) have been performed and published by The Cancer Genome Atlas (TCGA) and others. Subsequent studies investigated the genomic heterogeneity in spatially distinct tumor fragments and single cells. However, genomic heterogeneity in distinct anatomical areas of GBM has not been systematically investigated. A recent study appearing in the journal Science addressed this issue and reported the first anatomic transcriptional atlas of human GBM.1 The study used laser microdissection and RNA-seq to analyze the genomes of distinct anatomic regions in 42 GBM tumors. The analyzed regions were the tumor leading edge (LE), infiltrating tumor (IT), cellular tumor (CT), pseudo-palisading cells around necrosis (PAN), and microvascular proliferation (MVP). Differential gene expression analysis revealed 3627 genes that were enriched in either of LE, CT, PAN, or MVP, whereas IT genes ranged in a continuum between LE and CT. Samples from the same anatomic region across different tumors resembled each other more than samples from different anatomic regions within the same tumor, suggesting anatomic region-specific expression profiles. CT, PAN, and to a lesser extent MVP resembled the corresponding bulk tumors with regards to copy number variation and single nucleotide variants of select key GBM mutated genes (TP53, PTEN, EGFR, ATRX, IDH1 NF1, PIK3R1, PIK3CA). LE differed from the bulk tumor, an observation that was explained by the presence of a large proportion of non-neoplastic cells in this anatomic region. No mutation associated with any of the anatomic regions predicted overall survival better than MGMT promoter methylation status in the bulk tumor. The atlas and associated database for clinical and genomic data were designated “Ivy Glioblastoma Atlas” and can be found at http://glioblastoma.alleininstitute.org/. They represent a useful platform for testing new hypotheses associated with anatomic genomic heterogeneity and other basic, preclinical, and clinical questions in human GBM. Determining progression in slow-growing tumors such as low-grade glioma can be challenging. Anaplastic transformation of these tumors into higher-grade neoplasms typically results in faster growth, treatment resistance, and poor outcomes. Early markers of progressions include the development of contrast enhancement, change in growth rate, and development of increased perfusion.1 Recent work by Chen and colleagues2 suggests that changes in diffusion metrics may also signal patient progression earlier than that identified by RANO criteria. The authors examined untreated biopsy-proven low-grade gliomas, following changes in apparent diffusion coefficient (ADC) generated from areas of FLAIR signal abnormality. By analyzing 15 stable and 13 progressive tumors with a mean follow-up time of 41 months, the authors defined an unfavorable change in diffusion as an interval decrease in ADC values of ≥20%. Of various metrics, a change in 10th percentile ADC (ADC10) was the most accurate in predicting progression, correctly categorizing 12/13 patients with tumor progression, with 3/15 false positives in the stable disease group leading to an overall diagnostic accuracy of 86%. Furthermore, this method identified progression, on average, 8 months before conventional imaging using RANO criteria. This report adds to the growing body of evidence that advanced MRI can yield metrics predicative of low-grade glioma progression many months before progression is defined by standard imaging. An interesting follow-up will be to combine perfusion and diffusion imaging to see if even greater accuracy can be achieved.
Source: Neuro-Oncology
Authors: Wolfgang Wick, Riccardo Soffietti, Krishna Bhat, Roger Abounader, Whitney Pope
Language: English
Subfield: Genetics
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W4213079199