Expression profiling of single cells and patient cohorts identifies multiple immunosuppressive pathways and an altered NK cell phenotype in glioblastoma.

Glioblastoma (GBM) is an aggressive cancer with a very poor prognosis. Generally viewed as weakly immunogenic, GBM responds poorly to current immunotherapies. To understand this problem more clearly we used a combination of natural killer (NK) cell functional assays together with gene and protein expression profiling to define the NK cell response to GBM and explore immunosuppression in the GBM microenvironment. In addition, we used transcriptome data from patient cohorts to classify GBM according to immunological profiles. We show that glioma stem-like cells, a source of post-treatment tumour recurrence, express multiple immunomodulatory cell surface molecules and are targeted in preference to normal neural progenitor cells by natural killer (NK) cells ex vivo. In contrast, GBM-infiltrating NK cells express reduced levels of activation receptors within the tumour microenvironment, with hallmarks of transforming growth factor (TGF)-ß-mediated inhibition. This NK cell inhibition is accompanied by expression of multiple immune checkpoint molecules on T cells. Single-cell transcriptomics demonstrated that both tumour and haematopoietic-derived cells in GBM express multiple, diverse mediators of immune evasion. Despite this, immunome analysis across a patient cohort identifies a spectrum of immunological activity in GBM, with active immunity marked by co-expression of immune effector molecules and feedback inhibitory mechanisms. Our data show that GBM is recognized by the immune system but that anti-tumour immunity is restrained by multiple immunosuppressive pathways, some of which operate in the healthy brain. The presence of immune activity in a subset of patients suggests that these patients will more probably benefit from combination immunotherapies directed against multiple immunosuppressive pathways.

Methylation-specific multiplex ligation-dependent probe amplification identifies promoter methylation events associated with survival in glioblastoma.

DNA methylation plays an important role in cancer biology and methylation events are important prognostic and predictive markers in many tumor types. We have used methylation-specific multiplex ligation-dependent probe amplification to survey the methylation status of MGMT and 25 tumor suppressor genes in 73 glioblastoma cases. The data obtained was correlated with overall survival and response to treatment. The study revealed that methylation of promoter regions in TP73 (seven patients), THBS1 (eight patients) and PYCARD (nine patients) was associated with improved outcome, whereas GATA5 (21 patients) and WT1 (24 patients) promoter methylation were associated with poor outcome. In patients treated with temozolomide and radiation MGMT and PYCARD promoter methylation events remained associated with improved survival whereas GATA5 was associated with a poor outcome. The identification of GATA5 promoter methylation in glioblastoma has not previously been reported. Furthermore, a cumulative methylation score separated patients into survival groups better than any single methylation event. In conclusion, we have identified specific methylation events associated with patient outcome and treatment response in glioblastoma, and these may be of functional and predictive/prognostic significance. This study therefore provides novel candidates and approaches for future prospective validation.

Profiling cytotoxic microRNAs in pediatric and adult glioblastoma cells by high-content screening, identification, and validation of miR-1300.

MicroRNAs play an important role in the regulation of mRNA translation and have therapeutic potential in cancer and other diseases. To profile the landscape of microRNAs with significant cytotoxicity in the context of glioblastoma (GBM), we performed a high-throughput screen in adult and pediatric GBM cells using a synthetic oligonucleotide library representing all known human microRNAs. Bioinformatics analysis was used to refine this list and the top seven microRNAs were validated in a larger panel of GBM cells using state-of-the-art in vitro assays. The cytotoxic effect of our most relevant candidate was assessed in a preclinical model. Our screen identified ~100 significantly cytotoxic microRNAs with 70% concordance between cell lines. MicroRNA-1300 (miR-1300) was the most potent and robust candidate. We observed a striking binucleated phenotype in miR-1300 transfected cells due to cytokinesis failure followed by apoptosis. This was also observed in two stem-like patient-derived cultures. We identified the physiological role of miR-1300 as a regulator of endomitosis in megakaryocyte differentiation where blockade of cytokinesis is an essential step. In GBM cells, where miR-1300 is normally not expressed, the oncogene Epithelial Cell Transforming 2 (ECT2) was validated as a direct key target. ECT2 siRNA phenocopied the effects of miR-1300, and ECT2 overexpression led to rescue of miR-1300 induced binucleation. We showed that ectopic expression of miR-1300 led to decreased tumor growth in an orthotopic GBM model. Our screen provides a resource for the neuro-oncology community and identified miR-1300 as a novel regulator of endomitosis with translatable potential for therapeutic application.

ATR-dependent radiation-induced gamma H2AX foci in bystander primary human astrocytes and glioma cells.

Radiotherapy is an important treatment for patients suffering from high-grade malignant gliomas. Non-targeted (bystander) effects may influence these cells' response to radiation and the investigation of these effects may therefore provide new insights into mechanisms of radiosensitivity and responses to radiotherapy as well as define new targets for therapeutic approaches. Normal primary human astrocytes (NHA) and T98G glioma cells were irradiated with helium ions using the Gray Cancer Institute microbeam facility targeting individual cells. Irradiated NHA and T98G glioma cells generated signals that induced gammaH2AX foci in neighbouring non-targeted bystander cells up to 48 h after irradiation. gammaH2AX bystander foci were also observed in co-cultures targeting either NHA or T98G cells and in medium transfer experiments. Dimethyl sulphoxide, Filipin and anti-transforming growth factor (TGF)-beta 1 could suppress gammaH2AX foci in bystander cells, confirming that reactive oxygen species (ROS) and membrane-mediated signals are involved in the bystander signalling pathways. Also, TGF-beta 1 induced gammaH2AX in an ROS-dependent manner similar to bystander foci. ROS and membrane signalling-dependent differences in bystander foci induction between T98G glioma cells and normal human astrocytes have been observed. Inhibition of ataxia telangiectasia mutated (ATM) protein and DNA-PK could not suppress the induction of bystander gammaH2AX foci whereas the mutation of ATM- and rad3-related (ATR) abrogated bystander foci induction. Furthermore, ATR-dependent bystander foci induction was restricted to S-phase cells. These observations may provide additional therapeutic targets for the exploitation of the bystander effect.

Dose- and time-dependent changes in gene expression in human glioma cells after low radiation doses.

We have used DNA microarrays to identify changes in gene expression in cells of the radioresistant human glioma cell lines T98G and U373 after low radiation doses (0.2-2 Gy). Using Bayesian linear models, we have identified a set of genes that respond to low doses of radiation; furthermore, a hypothesis-driven approach to data analysis has allowed us to identify groups of genes with defined non-linear dose responses. Specifically, one of the cell lines we have examined (T98G) shows increased radiosensitivity at low doses (low-dose hyper-radiosensitivity, HRS); thus we have also assessed sets of genes whose dose response mirrors this survival pattern. We have also investigated a time course for induction of genes over the period when the DNA damage response is expected to occur. We have validated these data using quantitative PCR and also compared genes up-regulated in array data to genes present in the polysomal RNA fraction after irradiation. Several of the radioresponsive genes that we describe code for proteins that may have an impact on the outcome of irradiation in these cells, including RAS homologues and kinases involved in checkpoint signaling, so understanding their differential regulation may suggest new ways of altering radioresistance. From a clinical perspective these data may also suggest novel targets that are specifically up-regulated in gliomas during radiotherapy treatments.

DNA damage responses at low radiation doses.

Increased cell killing after exposure to low acute doses of X rays (0-0.5 Gy) has been demonstrated in cells of a number of human tumor cell lines. The mechanisms underlying this effect have been assumed to be related to a threshold dose above which DNA repair efficiency or fidelity increases. We have used cells of two radioresistant human tumor cell lines, one that shows increased sensitivity to low radiation doses (T98G) and one that does not (U373), to investigate the DNA damage response at low doses in detail and to establish whether there is a discontinuous dose response or threshold in activation of any important mediators of this response. In the two cell lines studied, we found a sensitive, linear dose response in early signaling and transduction pathways between doses of 0.1 and 2 Gy with no evidence of a threshold dose. We demonstrate that ATM-dependent signaling events to downstream targets including TP53, CHK1 and CHK2 occur after doses as low as 0.2 Gy and that these events promote an effective damage response. Using chemical inhibition of specific DNA repair enzymes, we show that inhibition of DNA-PK-dependent end joining has relatively little effect at low (<1 Gy) doses in hyper-radiosensitive cells and that at these doses the influence of RAD51-mediated repair events may increase, based on high levels of RAD51/BRCA2 repair foci. These data do not support a threshold model for activation of DNA repair in hyper-radiosensitive cells but do suggest that the balance of repair enzyme activity may change at low doses.

Low-dose hypersensitivity: current status and possible mechanisms.

PURPOSE: To retain cell viability, mammalian cells can increase damage repair in response to excessive radiation-induced injury. The adaptive response to small radiation doses is an example of this induced resistance and has been studied for many years, particularly in human lymphocytes. This review focuses on another manifestation of actively increased resistance that is of potential interest for developing improved radiotherapy, specifically the phenomenon in which cells die from excessive sensitivity to small single doses of ionizing radiation but remain more resistant (per unit dose) to larger single doses. In this paper, we propose possible mechanisms to explain this phenomenon based on our data accumulated over the last decade and a review of the literature. CONCLUSION: Typically, most cell lines exhibit hyper-radiosensitivity (HRS) to very low radiation doses (<10 cGy) that is not predicted by back-extrapolating the cell survival response from higher doses. As the dose is increased above about 30 cGy, there is increased radioresistance (IRR) until at doses beyond about 1 Gy, radioresistance is maximal, and the cell survival follows the usual downward-bending curve with increasing dose. The precise operational and activational mechanism of the process is still unclear, but we propose two hypotheses. The greater amount of injury produced by larger doses either (1) is above a putative damage-sensing threshold for triggering faster or more efficient DNA repair or (2) causes changes in DNA structure or organization that facilitates constitutive repair. In both scenarios, this enhanced repair ability is decreased again on a similar time scale to the rate of removal of DNA damage.

Effects of cell cycle phase on low-dose hyper-radiosensitivity.

PURPOSE: To examine the low-dose radiation response of human glioma cell lines separated into different cell-cycle phases and to determine if low-dose hyper-radiosensitivity (HRS) differs in populations defined by cell-cycle position. To assess whether predictions of the outcome of multiple low-dose regimens should take account of cell-cycle effects. MATERIALS AND METHODS: The clonogenic survival of G1, G2 and S phase cells was measured after exposure to single doses of X-rays in two human glioma cell lines. One cell line (T98G) showed marked HRS when asynchronous cells were irradiated, while the other (U373) did not. Separation of populations and high-resolution cell counting was achieved using a fluorescence activated cell sorter. Sorted cell populations were irradiated with 240 kVp X-rays to doses between 0.05 and 5Gy. The resulting cell-survival versus dose data were comparatively fitted using the linear-quadratic and induced-repair models in order to assess the degree of HRS. RESULTS: In both cell lines the low-dose response was altered when different populations were irradiated. In T98G cells, all populations showed HRS, but this was most marked in G2 phase cells. In U373 cells, no HRS was found in G1 or S phase cells, but HRS was demonstrable in G2 phase cells. CONCLUSIONS: HRS was expressed by the whole cell population of T98G cells but the size of the effect varied with cell-cycle phase and was most marked in the G2 population. In U373 cells, the effect could only be demonstrated in G2 cells. This implies that HRS is primarily a response of G2 phase cells and that this response dominates that seen in asynchronous populations. Actively proliferating cell populations may therefore demonstrate a greater increase in radiosensitivity to very low radiation doses compared with quiescent populations.

The response of human glioma cell lines to low-dose radiation exposure.

PURPOSE: To examine the low-dose radiation response of a series of radioresistant human glioma cell lines and determine if low-dose hypersensitivity is a characteristic of these cells. MATERIALS AND METHODS: The clonogenic survival of six radioresistant human glioma cell lines was measured following exposure to graded, single, very low doses of X-rays in vitro. High resolution was achieved using either a Dynamic Microscopic Image Processing Scanner (DMIPS) or a cell sorter (CS). RESULTS: In five of the six cell lines tested, low-dose hypersensitivity (HRS) was demonstrated although in the sixth, a grade III astrocytoma line, it was not. These results are consistent with previous data indicating that low-dose hypersensitivity is more marked in more radioresistant cell lines although the difference between the glioblastoma cell lines with differing SF2 is not marked. CONCLUSION: Low-dose hypersensitivity is common in radioresistant glioma cell lines. This may have implications for the treatment of these tumours if further studies confirm that HRS translates to increased effectiveness per gray in vivo when very low doses per fraction are used.

Low-dose hypersensitivity after fractionated low-dose irradiation in vitro.

PURPOSE: It was demonstrated previously that some radioresistant tumour cell lines respond to decreasing single, low radiation doses by becoming increasingly radiosensitive. This paper reports the response of four radioresistant human glioma cell lines to multiple low-dose radiation exposures given at various intervals. Three of the cell lines (T98G, U87, A7) were proven already to show low-dose hyper-radiosensitivity (HRS) after single low doses; the fourth, U373, does not show HRS after acute doses. MATERIALS AND METHODS: Clonogenic cell-survival measurements were made in vitro using the Dynamic Microscopic Image Processing Scanner (DMIPS) or Cell Sorter (CS) following exposure to 240kVp X-rays one or more times. RESULTS: A consistent, time-dependent hypersensitive response to a second, or subsequent, dose was observed in the cell lines that demonstrated HRS. This time-dependent change in radiosensitivity did not occur in the radioresistant cell line that did not show HRS (U373). In one cell line that demonstrated strong HRS, T98G, a similar time-dependent hypersensitive response was also seen when the cells were irradiated whilst held in the G1-phase of the cell cycle. In this same cell line, significantly increased cell kill was demonstrated when three very low doses (0.4 Gy) were given per day, 4 h apart, for 5 days, compared with the same total dose given as once-daily 1.2Gy fractions. CONCLUSIONS: These data demonstrate the possibility that a multipledose per day, low-dose per fraction regimen, termed 'ultrafractionation', could produce increased tumour cell kill in radioresistant tumours compared with the same total dose given as conventional-sized 2 Gy fractions.

Ultrafractionation in A7 human malignant glioma in nude mice.

PURPOSE: Low-dose hyperradiosensitivity (HRS) has been demonstrated in numerous cell lines in vitro, including a number of radioresistant human malignant glioma cell lines such as A7. The aim of our experiment was to show whether HRS can be exploited by using ultrafractionated irradiation (UF) to improve local control of A7 tumours growing in nude mice. Extrapolation of the in vitro results predict a 3.7-fold difference in the efficacy of UF compared with conventional fractionation (CF). MATERIAL AND METHODS: Subcutaneuously growing A7 tumours were irradiated either with UF (126 fractions in 6 weeks, 0.4 Gy per fraction) or CF (30 fractions in 6 weeks, 1.68 Gy per fraction). The total dose was 50.4 Gy in both experimental arms. Fractionated irradiations were given under ambient conditions and followed by graded top-up doses under clamp hypoxia. Endpoints were tumour growth delay and local tumour control 180 days after the end of treatment. RESULTS: UF resulted in a significant decrease of tumour growth delay and in a significant increase of the top-up TCD(50) compared with CF (40.0 Gy [95% CI 29; 61 Gy] versus 28.3 Gy [24; 35 Gy], p=0.047). CONCLUSIONS: Despite a pronounced HRS phenomenon in vitro, UF was significantly less effective than CF in A7 human malignant glioma in nude mice. These results neither disprove the existence of HRS nor do they exclude a possible clinical value of UF. The findings rather indicate that simplistic extrapolation from results obtained after single-dose exposure or few fractions in vitro is not sufficient to predict outcome of UF in vivo and that comprehensive evaluation of novel treatment options in animal models continues to be an essential requirement for clinical translation.

Primary central nervous system lymphoma with testicular relapse.

We report the case of a 54-year-old man with primary central nervous system lymphoma (PCNSL) who achieved a radiological complete response to high-dose methotrexate and 6 months later had a simultaneous local and testicular relapse.

A phase II study using retinoids as redifferentiation agents to increase iodine uptake in metastatic thyroid cancer.

AIMS: Radio-iodine is effective in treating metastatic differentiated thyroid cancers. In 20% of cases, however, these tumours fail to take up radio-iodine, and treatment options are then limited. Failure of iodine uptake might be reversible using redifferentiating agents. Retinoids redifferentiate a variety of cell types and increase iodine uptake in thyroid tumour cells in vitro. The aim of this study was to assess whether oral isotretinoin could increase radio-iodine uptake in patients with iodine-uptake-negative metastatic thyroid cancer. METHODS: Patients who had iodine-uptake-negative metastatic papillary or follicular thyroid cancers were selected from the thyroid database at The Royal Marsden Hospital and enrolled to an open-label, non-randomised phase II trial. Sites of metastatic disease were assessed using computed tomography or magnetic resonance imaging, and absence of iodine uptake was confirmed using a diagnostic radio-iodine scan before study entry. In eligible patients, isotretinoin was prescribed at 1.5 mg/kg/day orally for 8 weeks. Response was assessed within 2 weeks of completing treatment with repeat radio-iodine scan. All patients were reviewed every 2 weeks during treatment for assessment of toxicity. RESULTS: Sixteen patients were treated with isotretinoin between January 2001 and July 2002: nine with metastatic papillary thyroid cancer, five with metastatic follicular cancer and two with Hurthle cell carcinoma. Median age was 57 years. All patients tolerated 8 weeks of oral isotretinoin. Mucocutaneous side-effects and minor changes in biochemical or lipid profiles were documented in most patients. In one patient, radio-iodine uptake increased after retinoid administration; however, this was not large enough to permit a significant dose of iodine to be given to sites of metastatic disease. In the other 15 patients, no radio-iodine uptake was documented. CONCLUSION: Treatment with isotretinoin does not reliably increase radio-iodine uptake in patients with metastatic thyroid cancer. This treatment alone does not enable radio-iodine to be used for further treatment.

External beam and conformal radiotherapy in the management of gliomas.

External beam radiotherapy remains an important local treatment modality in both high and low grade gliomas, however its contribution to outcome remains modest. In high grade tumours this is because of their extreme clinical radioresistance, with local recurrences occurring even after doses over 70 Gy. In low grade tumours radiation does not seem to alter the overall pattern of disease progression significantly. Therefore despite use of the new technologies now available that allow radiotherapy to be delivered more accurately and to higher doses, local control of these tumours is still rarely achieved. Unfortunately these tumours have not proven sensitive to changes in radiotherapy fractionation or to the addition of radiosensitising agents. Novel approaches to these tumours are needed, based on an improved understanding of both tumour and normal tissue response to radiation.

The treatment of malignant cerebral tumours.

Malignant cerebral tumours are uncommon. While a large proportion are resistant to conventional therapies there are a significant number of curable malignant brain tumours that require recognition and appropriate therapy.

Benchmarking of a treatment planning system for spot scanning proton therapy: comparison and analysis of robustness to setup errors of photon IMRT and proton SFUD treatment plans of base of skull meningioma.

PURPOSE: Base of skull meningioma can be treated with both intensity modulated radiation therapy (IMRT) and spot scanned proton therapy (PT). One of the main benefits of PT is better sparing of organs at risk, but due to the physical and dosimetric characteristics of protons, spot scanned PT can be more sensitive to the uncertainties encountered in the treatment process compared with photon treatment. Therefore, robustness analysis should be part of a comprehensive comparison between these two treatment methods in order to quantify and understand the sensitivity of the treatment techniques to uncertainties. The aim of this work was to benchmark a spot scanning treatment planning system for planning of base of skull meningioma and to compare the created plans and analyze their robustness to setup errors against the IMRT technique. METHODS: Plans were produced for three base of skull meningioma cases: IMRT planned with a commercial TPS [Monaco (Elekta AB, Sweden)]; single field uniform dose (SFUD) spot scanning PT produced with an in-house TPS (PSI-plan); and SFUD spot scanning PT plan created with a commercial TPS [XiO (Elekta AB, Sweden)]. A tool for evaluating robustness to random setup errors was created and, for each plan, both a dosimetric evaluation and a robustness analysis to setup errors were performed. RESULTS: It was possible to create clinically acceptable treatment plans for spot scanning proton therapy of meningioma with a commercially available TPS. However, since each treatment planning system uses different methods, this comparison showed different dosimetric results as well as different sensitivities to setup uncertainties. The results confirmed the necessity of an analysis tool for assessing plan robustness to provide a fair comparison of photon and proton plans. CONCLUSIONS: Robustness analysis is a critical part of plan evaluation when comparing IMRT plans with spot scanned proton therapy plans.