KLF6 depletion promotes NF-κB signaling in glioblastoma.

Dysregulation of the NF-κB transcription factor occurs in many cancer types. Krüppel-like family of transcription factors (KLFs) regulate the expression of genes involved in cell proliferation, differentiation and survival. Here, we report a new mechanism of NF-κB activation in glioblastoma through depletion of the KLF6 tumor suppressor. We show that KLF6 transactivates multiple genes negatively controlling the NF-κB pathway and consequently reduces NF-κB nuclear localization and downregulates NF-κB targets. Reconstitution of KLF6 attenuates their malignant phenotype and induces neural-like differentiation and senescence, consistent with NF-κB pathway inhibition. KLF6 is heterozygously deleted in 74.5% of the analyzed glioblastomas and predicts unfavorable patient prognosis suggesting that haploinsufficiency is a clinically relevant means of evading KLF6-dependent regulation of NF-κB. Together, our study identifies a new mechanism by which KLF6 regulates NF-κB signaling, and how this mechanism is circumvented in glioblastoma through KLF6 loss.

Computational Identification of Tumor Anatomic Location Associated with Survival in 2 Large Cohorts of Human Primary Glioblastomas.

BACKGROUND AND PURPOSE: Tumor location has been shown to be a significant prognostic factor in patients with glioblastoma. The purpose of this study was to characterize glioblastoma lesions by identifying MR imaging voxel-based tumor location features that are associated with tumor molecular profiles, patient characteristics, and clinical outcomes. MATERIALS AND METHODS: Preoperative T1 anatomic MR images of 384 patients with glioblastomas were obtained from 2 independent cohorts (n = 253 from the Stanford University Medical Center for training and n = 131 from The Cancer Genome Atlas for validation). An automated computational image-analysis pipeline was developed to determine the anatomic locations of tumor in each patient. Voxel-based differences in tumor location between good (overall survival of >17 months) and poor (overall survival of <11 months) survival groups identified in the training cohort were used to classify patients in The Cancer Genome Atlas cohort into 2 brain-location groups, for which clinical features, messenger RNA expression, and copy number changes were compared to elucidate the biologic basis of tumors located in different brain regions. RESULTS: Tumors in the right occipitotemporal periventricular white matter were significantly associated with poor survival in both training and test cohorts (both, log-rank P < .05) and had larger tumor volume compared with tumors in other locations. Tumors in the right periatrial location were associated with hypoxia pathway enrichment and PDGFRA amplification, making them potential targets for subgroup-specific therapies. CONCLUSIONS: Voxel-based location in glioblastoma is associated with patient outcome and may have a potential role for guiding personalized treatment.

Thymidine kinase activation of ganciclovir in recurrent malignant gliomas: a gene-marking and neuropathological study.

OBJECT: The gene therapy paradigm of intratumoral activation of ganciclovir (GCV) following transduction of tumor cells by retroviral vectors bearing the thymidine kinase (tk) gene has produced dramatic remissions of malignant gliomas in animal models. In human trials, although the technique has been deemed safe, little antitumor effect has been demonstrated. To evaluate the basis of this inefficacy in human gliomas, the authors conducted a gene-marking trial involving neuropathological and biochemical studies of treated tumor specimens. METHODS: Five patients with malignant recurrent gliomas underwent stereotactic biopsy sampling and intratumoral implantation procedures with three aliquots of 10(6) vector-producing cells (VPCs) in columns. After 5 days, the tumor was resected and the tumor bed reimplanted with VPCs, and a course of GCV was given. Patients received clinical and radiological follow up for 6 months. Tumor specimens were analyzed neuropathologically and for tk gene expression by anti-TK immunohistochemistry and TK enzymatic activity. Four patients tolerated the treatment well but experienced tumor progression. The other developed an abscess after the second operation and died. Increased TK enzymatic activity was demonstrated in the one tumor specimen analyzed. Immunohistochemical evidence of tk gene expression was limited to VPCs. Transduction of tumor cells was not seen. Viable tumor cells were seen near VPCs containing TK. The lymphocytic immune response was mild. CONCLUSIONS: Except for the risk of infection inherent in reoperation, this tk-GCV paradigm was both feasible and safe. Pathological studies indicated that limited dissemination of VPCs and vector from the infusion site and failure to transduce tumor cells with the tk gene are major barriers to efficacy.

Multicolumn infusion of gene therapy cells into human brain tumors: technical report.

OBJECTIVE: Effective gene therapy for brain tumors may require saturation of the tumors with tumoricidal doses of the therapeutic gene. Safe, precise, and efficient delivery of gene therapy vectors is required. Most reported cases of and published protocols for gene therapy for brain tumors involve freehand injection of retroviral vector-producing cells (VPCs) into the brain. Major disadvantages of this method include the inaccuracy of hand-guided needle placement and limited control of injection parameters. These factors can result in failure to deliver the viral vectors to specifically targeted sites within the brain, extensive tissue disruption resulting from excessively forceful injection, and reflux of the injectate along the needle tract. METHODS: We describe a novel stereotactic strategy for saturating tumor volumes with tumoricidal doses of gene therapy vectors and a new, more precise method of infusing VPCs. With our new instrument, the multicolumn stereotactic infusion system, needle placement is stereotactically guided and both VPC infusion and needle withdrawal are mechanically controlled. RESULTS: This technique, which has been used effectively for six patients, permits precise deposition of columns of VPCs throughout the targeted tumor volume. CONCLUSION: This technique should facilitate saturation of tumors with tumoricidal doses of gene therapy vectors and should improve the results of gene therapy protocols that rely on intraparenchymal injection for delivery.

Accelerated fractionated proton/photon irradiation to 90 cobalt gray equivalent for glioblastoma multiforme: results of a phase II prospective trial.

OBJECT: After conventional doses of 55 to 65 Gy of fractionated irradiation, glioblastoma multiforme (GBM) usually recurs at its original location. This institutional phase II study was designed to assess whether dose escalation to 90 cobalt gray equivalent (CGE) with conformal protons and photons in accelerated fractionation would improve local tumor control and patient survival. METHODS: Twenty-three patients were enrolled in this study. Eligibility criteria included age between 18 and 70 years, Karnofsky Performance Scale score of greater than or equal to 70, residual tumor volume of less than 60 ml, and a supratentorial, unilateral tumor. Actuarial survival rates at 2 and 3 years were 34% and 18%, respectively. The median survival time was 20 months, with four patients alive 22 to 60 months postdiagnosis. Analysis by Radiation Therapy Oncology Group prognostic criteria or Medical Research Council indices showed a 5- to 11-month increase in median survival time over those of comparable conventionally treated patients. All patients developed new areas of gadolinium enhancement during the follow-up period. Histological examination of tissues obtained at biopsy, resection, or autopsy was conducted in 15 of 23 patients. Radiation necrosis only was demonstrated in seven patients, and their survival was significantly longer than that of patients with recurrent tumor (p = 0.01). Tumor regrowth occurred most commonly in areas that received doses of 60 to 70 CGE or less; recurrent tumor was found in only one case in the 90-CGE volume. CONCLUSIONS: A dose of 90 CGE in accelerated fractionation prevented central recurrence in almost all cases. The median survival time was extended to 20 months, likely as a result of central control. Tumors will usually recur in areas immediately peripheral to this 90-CGE volume, but attempts to extend local control by enlarging the central volume are likely to be limited by difficulties with radiation necrosis.

Oncolytic virus therapy of multiple tumors in the brain requires suppression of innate and elicited antiviral responses.

The occurrence of multiple tumors in an organ heralds a rapidly fatal course. Although intravascular administration may deliver oncolytic viruses/vectors to each of these tumors, its efficiency is impeded by an antiviral activity present in complement-depleted plasma of rodents and humans. Here, this activity was shown to interact with complement in a calcium-dependent fashion, and antibody neutralization studies indicated preimmune IgM has a contributing role. Short-term exposure to cyclophosphamide (CPA) partially suppressed this activity in rodents and humans. At longer time points, cyclophosphamide also abrogated neutralizing antibody responses. Cyclophosphamide treatment of rats with large single or multiple intracerebral tumors substantially increased viral survival and propagation, leading to neoplastic regression.

Regression of experimental brain tumors with 6-thioxanthine and Escherichia coli gpt gene therapy.

The identification of transgenes with antitumor activity is critical to the development of gene therapy of cancer. Retrovirus-mediated transfer of the Escherichia coli gpt gene into rat C6 glioma cells without subsequent selection still inhibited the proliferation of this mixed polyclonal population upon addition of the prodrug, 6-thioxanthine, with an ID50 of 4.1 microM, whereas parental C6 cells were not affected at a concentration of 500 microM. In a time-course assay, effects of the prodrug on the mixed polyclonal cell proliferation required at least 10 days of exposure. In mixed co-cultures, a bystander effect was not present over the first 4 days of prodrug exposure, but required trypsinization of the co-cultures and replating at lower densities. This "modified" bystander assay thus revealed a 50% decrease in C6 cell proliferation, even when the initial ratio of gpt-expressing to parental C6 cells was as low as 1:19. In a nude mouse model of subcutaneous tumors, co-grafts of C6 glioma and gpt-retrovirus producer cells displayed retarded growth upon exposure to 6-thioxanthine (6-TX). In a nude mouse model of intracerebral tumors, grafting of the gpt-retrovirus producer cells leads to an 80% reduction in intracerebral tumor volumes after 6-TX treatment. This reduction results in a 28% increase in the mean time of survival of animals that harbor intracerebral tumors (p < 0.0005). These antitumor effects indicate that the gpt/6-TX enzyme/prodrug pair is a promising alternative to the thymidine kinase gene and ganciclovir combination in the gene therapy of cancer.

Giant epidermoid involving 3 cranial bones.

We describe a patient with an incidental finding of an expansile osteolytic lesion centered in the sphenoid involving 3 cranial bones-frontal, temporal, and sphenoid-and invading the brain parenchyma and orbit. Biopsy demonstrated an epidermoid. The lesion was excised through frontotemporal craniotomy. Despite the benign nature of the tumor, the epidermoid had a destructive pattern of growth. Complete resection was needed to adequately address the inflammatory mass effect of the large lesion.

Interstitial fluid pressure in intracranial tumours in patients and in rodents.

Fluid transport parameters in intracranial tumours influence the delivery of therapeutic agents and the resolution of peritumoral oedema. The tumour and cortex interstitial fluid pressure (IFP) and the cerebrospinal fluid pressure (CSFP) were measured during the growth of brain and pial surface tumours [R3230AC mammary adenocarcinoma (R3230AC) and F98 glioma (F98)] in rats. Intratumoral and intracranial pressures were also measured in rodents and patients treated with dexamethasone, mannitol and furosemide (DMF), and hypocapnia. The results show that (1) for the R3230AC on the pial surface, IFP increased with tumour volume and CSFP increased exponentially for tumours occupying a brain volume of 5% or greater; (2) in F98 with volumes of approximately 10 mm3, IFP decreased from the tumour to the cortex, whereas for tumour volumes > 16 mm3 IFP equilibrates between F98 and the cortex; (3) DMF treatment reduced the IFP of intraparenchymal tumours significantly and induced a pressure gradient from the tumour to the cortex; and (4) in 11 patients with intracranial tumours, the mean IFP was 2.0 +/- 2.5 mmHg. In conclusion, the IFP gradient between intraparenchymal tumours and the cortex decreases with tumour growth, and treatment with DMF can increase the pressure difference between the tumour and surrounding brain. The results also suggest that antioedema therapy in patients with brain tumours is responsible in part for the low tumour IFP.

Procarbazine, lomustine, and vincristine (PCV) chemotherapy for grade III and grade IV oligoastrocytomas.

The authors provided procarbazine, lomustine (CCNU), and vincristine (PCV) chemotherapy to 32 patients whose tumors contained varying mixtures of oligodendroglial and astrocytic cells. Twenty-five patients had oligodendroglioma-astrocytoma (oligoastrocytoma) with a histological Grade of III (19 patients) or IV (six patients); seven had anaplastic oligodendroglioma. The PCV therapy was administered every 6 weeks for a total of at least 124 cycles. The median duration of follow-up review from the start of chemotherapy was 19.3 months. Nineteen patients were treated before receiving radiation therapy and 12 after receiving it (one patient received concurrent radiotherapy and chemotherapy). Grade 3 or 4 hematological toxicity was experienced by nine (31%) of 29 patients. Ten patients had delayed treatment due to treatment-related toxicities (34.5%). Ninety-one percent of the 32 patients responded to the therapy. These included 10 patients with a complete response and 19 with a partial response. The median time to progression was 15.4 months for all patients and 23.2 months for those with Grade III tumors. The median time to progression for patients with Grade III oligoastrocytomas was 13.8 months; for those with Grade IV oligoastrocytoma it was 12.4 months and for those with anaplastic oligodendrogliomas it was 63.4 months (p = 0.0348). These patients survived a median of 49.8 months, 16 months, and 76 or more months, respectively, from the start of chemotherapy (p = 0.0154). The PCV therapy provides durable responses in patients with Grade III or IV oligoastrocytomas.

Diagnosis of suspected intracranial metastases. Role of direct tissue examination.

Accurate histologic diagnosis is often critical to the appropriate choice of treatment of patients suspected of harboring brain metastases. In most cases, frame based stereotactic biopsy can provide an accurate histologic diagnosis with minimal risk of morbidity. In addition, newer modalities allow for efficient coupling of stereotactic biopsy with treatment in a single session. These include frameless stereotactic systems that facilitate the combination of open biopsy and resection as well as biopsy procedures that are combined with focal interstitial or external beam irradiation. These techniques provide increasingly convenient and effective means of achieving both diagnosis and treatment of cerebral metastases.

Recommendations for treatment of brain metastases.

As treatment options for cerebral metastases have increased, so has controversy regarding relative indications for their use. The therapeutic options, criteria for choice, and recommendations for treatment require careful consideration of available data.

Therapy of primary CNS lymphoma with methotrexate-based chemotherapy and deferred radiotherapy: preliminary results.

Disease-free survival in primary CNS lymphoma has improved with the advent of methotrexate-based pre-irradiation chemotherapy. Prolonged response durations have been noted in six of eight patients refusing radiation therapy in two of our prior series. We have treated an additional 11 patients with methotrexate-based chemotherapy without subsequent planned irradiation. Some received maintenance chemotherapy. Most have had durable responses with little or no toxicity. Prolonged responses can be maintained without radiation therapy, thus avoiding potential long-term radiation toxicity.

Echo-planar MR determination of relative cerebral blood volume in human brain tumors: T1 versus T2 weighting.

PURPOSE: Maps related to relative cerebral blood volume (rCBV) were generated with the use of the T1 effects produced by a low-dose bolus passage of gadopentetate dimeglumine. The T1 maps were evaluated in a tumor population and compared with rCBV maps obtained with T2-weighted measurements. METHODS: Imaging was performed in 19 patients with suspected intraaxial brain tumors. For the T1 rCBV maps, a low-dose bolus of contrast material was given during T1-weighted interleaved spin-echo echo-planar MR imaging. This was followed by a second injection during serial T2-weighted imaging for generation of the T2 rCBV maps. RESULTS: Among patients with low-grade lesions (n = 9), T1-based and T2-based rCBV maps showed comparably low rCBV in 7 subjects. In the other 2 patients, with confirmed tumor dedifferentiation, elevation of rCBV values was seen on maps obtained with both techniques. Among patients with high-grade tumors (n = 10), 4 had no evidence of recurrence and 6 did have tumor recurrence (confirmed by follow-up and positron emission tomography). In patients with the high-grade lesions exhibiting conventional contrast enhancement, lesions tended to have higher estimated values on T1 rCBV maps than on the T2 rCBV maps. CONCLUSION: Although the T1 rCBV maps showed less contrast as compared with the T2 rCBV maps, they provided diagnostic information that was comparable to the T2 rCBV maps in our series of 19 patients with primary brain tumors.

Retrovirus-mediated gene therapy of experimental brain neoplasms using the herpes simplex virus-thymidine kinase/ganciclovir paradigm.

Recent results in experimental brain tumors indicate that transfer of sensitizing genes to tumor cells in vivo with subsequent drug treatment can reduce tumor masses and prolong the survival of rodents. In the present study, the 9L rat gliosarcoma model was used to evaluate the therapeutic effectiveness of the herpes simplex virus-thymidine kinase (HSV-tk) gene, delivered by a retrovirus vector, against tumor cells in the rat brain after systemic application of the nucleoside analogue ganciclovir (GCV). The HSV-tk gene was inserted into a retroviral vector (pMFG), which was produced using the amphotropic packaging cell line CRIP-MFG-S-HSV-TK. Packaging cells were implanted into established 9L tumors in the brains of syngeneic rats to effect gene delivery to tumor cells, followed by intraperitoneal GCV injections. Treated animals survived significantly longer (more than twice as long) than did the control groups. Brains from GCV-treated and nontreated animals were examined immunohistochemically at different time intervals after grafting of CRIP-MFG-S-HSV-TK cells and GCV treatment. Tumors in GCV-treated animals were significantly smaller as compared with nontreated animals at all time points. Sections stained immunohistochemically for HSV-TK confirmed gene transfer to tumor cells, which could be distinguished from packaging cells by different morphology and immunohistochemical staining for the retroviral envelope protein gp70. Approximately 45% of the cells in tumors implanted with CRIP-MFG-S-HSV-TK cells, but not treated with GCV, showed immunocytochemical staining for HSV-TK, demonstrating a high-efficiency of retrovirus-mediated gene transfer. Tumors in rats treated with packaging cells and GCV showed only 9% HSV-TK-positive cells after treatment, indicating that most cells expressing the HSV-tk gene were killed. The success of this therapeutic modality in experimental animals depends in large parts on the high efficiency of gene delivery and on the immune response against tumor cells.

Anterior midline approaches to the central skull base.

Anterior midline approaches are safe and appropriate for extradural lesions of the central brain base. They are occasionally warranted for intradural lesions as well. Transnasal routes expose the clivus well. They are readily expanded superiorly, inferiorly, and laterally. Recent innovations are reductive; they expand exposure with less facial disassembly. Lateral and most intradural extensions of lesions warrant more lateral approaches.

Chromosome 19q deletions in human gliomas overlap telomeric to D19S219 and may target a 425 kb region centromeric to D19S112.

Chromosome 19q harbors a tumor suppressor gene that is involved in astrocytoma, oligodendroglioma and mixed glioma tumorigenesis. We had previously mapped this gene to an approximately 5 megabase region of chromosome 19q13.2-13.3 between APOC2 and HRC. To narrow the location of this tumor suppressor further, we studied 138 gliomas for loss of allelic heterozygosity at six microsatellite polymorphisms between APOC2 and HRC, including a newly described polymorphism in the ERCC2 gene. Allelic loss occurred in 48 gliomas (35%), including 25 of 41 oligodendroglial tumors (61%). Four cases had proximal breakpoints within the APOC2-HRC region, two telomeric to ERCC2 and two telomeric to D19S219. In addition, one of the latter tumors had an interstitial deletion between D19S219 and D19S112, a distance of only 425 kilobases surrounding the DM (myotonic dystrophy) gene. These findings suggest that the glioma tumor suppressor on chromosome 19q maps to 19q13.3, telomeric to D19S219 and perhaps centromeric to D19S112. The data exclude a number of candidate genes from 19q13.2-13.3, including a putative phosphatase gene and the DNA repair/metabolism genes ERCC1, ERCC2 and probably LIG1.

The multiple tumor suppressor 1/cyclin-dependent kinase inhibitor 2 gene in human central nervous system primitive neuroectodermal tumor.

The recently described multiple tumor suppressor 1/cyclin-dependent kinase inhibitor 2 (MTS1/CDKN2) gene, encoding the cyclin-dependent kinase 4 inhibitor p16, is mutated in a wide variety of tumor cell lines, including gliomas. To investigate the possible role of this gene in the genesis of the central nervous system primitive neuroectodermal tumor (PNET), four established PNET cell lines and 18 PNET surgical specimens were studied for deletions and mutations of the MTS1/CDKN2 gene. One of the four cell lines had homozygous deletion of the gene. No mutation in any of the three MTS1/CDKN2 exons was detected in the other three cell lines by single strand conformational polymorphism analysis. Eighteen surgical PNET specimens were studied for allelic and homozygous deletion at chromosome 9p21, the location of the MTS1/CDKN2 gene. No loss of heterozygosity was noted in 11 of the tumors, and no homozygous loss was noted in any tumor. Single strand conformational polymorphism analysis of the entire coding region of the MTS1/CDKN2 gene revealed no mutation within MTS1/CDKN2 in any tumor. Although deletion of MTS1/CDKN2 may occur in some PNET cell lines, neither deletion nor mutation of the gene is found in tumors before culture. The genesis of the human central nervous system PNET does not involve deletion or mutation of the MTS1/CDKN2 gene.