Year: 2016
Type: article
Abstract: Brain tumors are common, heterogeneous, and clinically challenging. Advances and clinical application of precision medical approaches in brain tumor patients have been hampered by the difficulty to readily obtain tumor tissue samples in many patients, as neurosurgical interventions are performed only in selected patients and in specific clinical situations. Importantly, a majority of patients with brain metastases and recurrent brain tumors, primary or secondary, never undergo a neurosurgical procedure. Upon this background, Pentsova et al. performed a study to investigate whether DNA from cerebrospinal fluid (CSF), which can be obtained relatively easily by a lumbar puncture, may be useful for interrogating molecular information on CNS tumors. 1 Interestingly, the authors found that indeed molecular alterations with potential clinical relevance can be found in cell-free DNA in many patients with primary or secondary brain tumors. They detected high-confidence somatic alterations in all brain metastasis patients with positive CSF cytology and were also successful in detecting molecular alterations in a sizable fraction of patients with negative CSF cytology (25% of brain metastasis patients and close to 50% of primary brain tumor patients). Importantly, genomic alterations associated with drug resistance were identified in patients undergoing therapies with several oncological biologicals. The results of the study are highly interesting and pave the way for molecular profiling and personalized medicine in patients with CNS lesions that are not amenable to biopsy or resection (e.g. tumors in eloquent brain regions, brain stem glioma, multiple brain metastases) or in patients with medical co-morbidities precluding neurosurgical interventions. Furthermore, since primary tumors and corresponding brain metastases in individual patients often harbor divergent molecular profiles 2 , genomic CSF analyses may help to personalize oncological treatment in the large and challenging population of brain metastasis patients. Another particularly important indication for molecular CSF profiling might be the search for genomic alterations associated with recurrence and treatment resistance, again both in primary and secondary brain tumors. Such data will be very useful to better understand the molecular basis of brain tumor progression on a biological level and will hopefully contribute to the development of better therapies. Overall, the results presented by Pentsova et al. may lead to a broadening of the use of CSF puncture in the oncological management of patients with neoplastic CNS disease. Thus far, CSF puncture has been mainly performed for diagnostic purposes in a small and selected group of brain tumor patients, e.g. patients with suspected leptomeningeal metastasis or CNS lymphoma. The interesting data presented here indicate that a larger group of patients may profit from comprehensive CSF analyses upon detection of CNS involvement or repetitively during the course of disease. Prospective studies will be important to evaluate the benefit of targeting DNA alterations detected in CSF samples using specific drugs. Targeting the tumor microenvironment (TME) has come to the forefront of new strategies for treating cancer. An obvious advantage is that the TME is not affected by the genetic instability that has been associated with acquisition of treatment resistance in tumors. Tumor-associated macrophages (TAMs) and microglia represent a sizable fraction of the tumor mass and are known to contribute to glioma genesis. The colony stimulating factor receptor 1 (CSF-R1) expressed on macrophages has been shown to be effectively druggable with small-molecule inhibitors in a mouse model of glioma 1 . Inhibitors of CSF-R1 are currently being evaluated in clinical trials for glioma. Anticipating acquisition of tumor resistance after prolonged exposure to the CSF-R1 inhibitor, Quail and co-workers have followed up on their previous work 1 and now present a thorough investigation of tumor recurrence in their genetic glioma mouse model resembling the proneural glioblastoma subtype, taking into account the crosstalk between the tumor and the respective TME 2 . Expression profiles of the tumors indicated the activation of the PI3K pathway at recurrence. The importance of this finding for resistance mechanisms was supported by the fact that combining the CSF-R1 inhibitor with a PI3K inhibitor at recurrence improved outcome. In accordance, Pten-deficient gliomas responded little to the CSF-R1 inhibition initially. The authors next addressed the question of the mechanism leading to activation of the PI3K pathway in the tumor cells. By transplanting the recurrent tumor cells into CSF-1R inhibitor-naïve mice, which restored sensitivity to chemical inhibition of CSF-R1, it became clear that in fact the tumor environment was the cause of resistance. Separate analysis of the tumors and different types of cells constituting the TME uncovered increased expression of IGF-1 by TAMs that activated the IGF-1R expressed on the tumor cells, leading to the observed upregulation of the PI3K pathway. Again, the importance of this identified crosstalk between microenvironment and the recurrent tumor was supported by improved outcome when combining an IGF-R1 inhibitor with the CSF-R1 inhibition. This timely and rigorous molecular dissection of inherent and acquired resistance to CSF-R1 inhibitors identified several critical hubs of signaling loops for which targeted drugs are already available and may be used for combination therapies. It will be of utmost importance to determine whether the insights of resistance in the mouse model will be informative for explaining the inherent resistance observed in recurrent glioblastoma against CSF-R1 inhibition 3 in order to design rational combination therapies and biomarkers for patient selection. Glioblastomas exhibit remarkable intratumoral heterogeneity that makes treatment design difficult. Complicating this further is the fact that these tumors can exhibit phenotypic plasticity and genetic evolutionary dynamics. Understanding the molecular basis of this heterogeneity can lead to newer strategies for treatment. In a recent study Lu et al. describe a phenotypic shift of molecular subtypes of glioma that is dependent on Olig2. The authors use two independent models of proneural (PN) glioma to test the influence of Olig2 on tumorigenesis. In the first model, they use ganciclovir-mediated ablation of Olig2 with HSV-TK knocked in at the Olig2 locus in mice. These mice also overexpressed PDGFB, and deletion of Pten and Trp53 in the white matter cortex makes this a clinically relevant model. In a second model they use Cre virus to delete floxed Olig2, Pten, and Trp53. In both models Olig2 ablation or deletion caused delayed onset of tumor growth, featuring reduced mitotic cells and improved survival. When primary cells from these tumors were cultured as neurospheres, only the control mice that were Olig2+ formed free-floating spheres, whereas Olig2-knockout tumor cells attached as monolayers and showed decreased tumorigenic capacity when re-implanted in syngeneic mice. Combining RNA-seq and CHiP-seq, the authors found that the loss of Olig2 leads to a PN to classical (CL) transition with concomitant loss of Olig2 at the enhancers of cell cycle regulators and oncogenes. Importantly, they found that loss of Olig2 not only leads to substantially downregulated PDGFRa but consequently increases EGFR pathway signaling. Treating Olig2 null tumor-bearing mice with gefitinib, an inhibitor of EGFR, leads to a significant reduction in tumor size, delayed tumor growth, and improved lifespan of the mice. While the exact mechanism of Olig2 downregulation of EGFR is unclear, this study has immediate clinical implications. With Olig2 inhibitors in clinical development, patients with a PN signature can potentially be treated with an Olig2 inhibitor and gefitinib for maximal response. The combination of focused ultrasound (FUS) with intravenous injection of microbubbles can be used to transiently overcome the blood-brain barrier in order to deliver intravascular therapeutics to brain parenchyma. This approach has been shown to be effective mostly for smaller molecules including temozolomide, but experience with biologics such as monoclonal antibodies is limited. Adding to a body of previous work, Liu and colleagues 1 recently examined the impact of FUS on delivery of bevacizumab (BEV) to a U87 glioma mouse xenograft model. Specifically they compared animals treated with FUS plus BEV (50mg per kg body weight per week x 5 weeks), with those receiving BEV or FUS only. BEV penetration into the brain was assessed by three methods: liquid chromatography/mass spectrometry (LCMS), BEV immunoblotting, and by PET-CT following injection of Ga-68-labeled BEV. To monitor disruption of the BBB the authors used dynamic contrast-enhanced (DCE) perfusion MRI, calculating parameters including the volume transfer constant Ktrans. As expected, FUS targeted to the tumor region resulted in significantly increased penetration of BEV into the brain. The numbers were impressive: with intermediate FUS exposure (0.4MPa) BEV concentration increased 5.7 fold (as measured by LCMS). With aggressive exposure (0.8MPa) BEV concentration increased 57 fold. Results were confirmed with PET-CT and anti-BEV immunoblots. DCE parameters correlated well with BEV concentration and also correlated with FUS exposure levels. The combination of BEV and FUS controlled tumor growth up to 5 weeks, unlike BEV alone. Median and mean survival were increased in the combination arm (73d v 46d median survival for BEV + FUS compared with BEV only). On histopathologic examination with immunstaining for CD31, the BEV + FUS tumors showed a larger necrotic core and reduced peripheral vasculature compared with those receiving FUS or BEV only. Thus, the authors demonstrated that FUS increases BEV penetration of the BBB, leading to prolonged survival in a mouse glioma model, and that this increased penetration occurs despite the vascular normalization effect of BEV. Because vascular normalization following BEV therapy may impede prolonged angiogenic suppression by BEV in glioblastoma, this method may potentiate the effects of BEV, generating a more sustained response that could translate into longer patient survival. Substantial progress has been made in the treatment of patients with various cancers with immune checkpoint inhibitors. Recently, two PD-1 inhibitors, pembrolizumab and nivolumab, have been approved for treatment of metastatic melanoma and non-small-cell lung cancer (NSCLC) after assessment in phase III trials that showed improvement in overall survival compared with standard of care in both diseases. In a single institution, two-cohort, phase II trial, Goldberg and co-workers 1 enrolled patients with melanoma or NSCLC and brain metastases to receive pembrolizumab. Key eligibility criteria were stage IV melanoma or NSCLC, age 18 years or older, Eastern Cooperative Oncology Group performance status (ECOG PS) of 0–1, life expectancy longer than 3 months, and adequate organ function. Patients had at least one brain metastasis between 5mm and 20mm that was untreated or was unequivocally progressing after radiation. Tumor tissue from a brain metastasis before enrollment was required for melanoma patients, while tumor PD-L1 positivity was required for enrollment of NSCLC patients. Patients with neurological symptoms attributable to brain metastases or requiring corticosteroids and patients with leptomeningeal disease were excluded. Any number of previous systemic therapies was allowed. The efficacy of systemic therapy was assessed every 8 weeks. Cerebral and systemic metastases were assessed with MRI and CT, respectively, using Response Evaluation Criteria in Solid Tumors (RECIST) (version 1.1). The primary endpoint was brain metastasis response, defined as the percentage of patients with a complete or partial response per modified RECIST, with a confirmed response requiring repeat imaging at least 4 weeks later. Secondary endpoints were toxicity, overall response in the brain and body, progression-free and overall survival, and predictive biomarkers. In terms of results, 4 of 18 patients (22%) with melanoma had confirmed brain metastasis partial responses. CNS responses were durable for up to 10.0 months, and all responses were ongoing at the time of data analysis. Of patients with NSCLC, 6 of 18 (33%) had brain metastasis responses, which included 4 complete responses and 2 partial responses. The CNS responses were durable for up to 7 months with 4 patients continuing to respond at the time of data analysis. None of the patients with either melanoma or NSCLC who had target lesions previously treated with stereotactic radiosurgery or whole brain radiotherapy had a CNS response. Overall, there was strong concordance between CNS and systemic responses. Median overall survival in patients with melanoma was not reached, and in patients with NSCLC survival was 7.7 months. The authors did not observe treatment-related deaths, and adverse events were as seen in the registration trials without significant neurological disturbances. This trial, specifically designed to study pembrolizumab in patients with brain metastases, is the first to assess the activity of a PD-1 axis inhibitor in the CNS in patients with melanoma or NSCLC. The results suggest that systemic PD-1 inhibition can be effective in treating non-irradiated brain metastases. Two points deserve particular attention. First, patients were allowed to continue in the trial despite disease progression if they were deriving clinical benefit according to investigator assessment or if progressive lesions (brain or systemic) could be controlled with local therapy, as recently suggested by the RANO Group 2 . Second, biopsy of one lesion revealed inflammation, suggesting that tumor growth in this setting might not always represent progression 3 . In this regard, advanced imaging studies, including PET, could be of value. Limitations of this study include the immature survival data, lack of biomarker analysis, and small sample size. Moreover, the heterogeneity in terms of previous local and systemic therapies, volume of CNS disease, and variability of histological and genetic subtypes limit the conclusions. Future studies should investigate whether pembrolizumab is effective or safe in patients with symptomatic or larger brain metastases; moreover, combination treatment strategies and biomarker development will be important.
Source: Neuro-Oncology
Authors: Matthias Preusser, Monika Hegi, Krishna Bhat, Whitney Pope, Riccardo Soffietti, Kenneth Aldape
Language: English
Subfield: Genetics
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4255368355