Clinical outcomes of melanoma brain metastases treated with stereotactic radiosurgery and anti-PD-1 therapy, anti-CTLA-4 therapy, BRAF/MEK inhibitors, BRAF inhibitor, or conventional chemotherapy.

BACKGROUND: The effect of immunologic and targeted agents on intracranial response rates in patients with melanoma brain metastases (MBMs) is not yet clearly understood. This report analyzes outcomes of intact MBMs treated with single-session stereotactic radiosurgery (SRS) and anti-PD-1 therapy, anti-CTLA-4 therapy, BRAF/MEK inhibitors(i), BRAFi, or conventional chemotherapy. PATIENTS AND METHODS: Patients were included if MBMs were treated with single-session SRS within 3 months of receiving systemic therapy. The primary end point of this study was distant MBM control. Secondary end points were local MBM control defined as a >20% volume increase on follow-up MRI, systemic progression-free survival, overall survival (OS) from both SRS and cranial metastases diagnosis, and neurotoxicity. Images were reviewed alongside two neuro-radiologists at our institution. RESULTS: Ninety-six patients were treated to 314 MBMs over 119 SRS treatment sessions between January 2007 and August 2015. No significant differences were noted in age (P = 0.27), gender (P = 0.85), treated gross tumor volume (P = 0.26), or the diagnosis-specific graded prognostic assessment (P = 0.51) between the treatment cohorts. Twelve-month Kaplan-Meier (KM) distant MBM control rates were 38%, 21%, 20%, 8%, and 5% (P = 0.008) for SRS with anti-PD-1 therapies, anti-CTLA-4 therapy, BRAF/MEKi, BRAFi, and conventional chemotherapy, respectively. No significant differences were noted in the KM local MBM control rates among treatment groups (P = 0.25). Treatment with anti-PD-1 therapy, anti-CTLA-4 therapy, or BRAF/MEKi significantly improved OS on both univariate and multivariate analyses when compared with conventional chemotherapy. CONCLUSION: In our institutional analysis of patients treated with SRS and various systemic immunologic and targeted melanoma agents, significant differences in distant MBM control and OS are noted. Prospective evaluation of the potential synergistic effect between these agents and SRS is warranted.

Attenuated reovirus displays oncolysis with reduced host toxicity.

BACKGROUND: Although the naturally occurring reovirus causes only mild symptoms in humans, it shows considerable potential as an oncolytic agent because of its innate ability to target cancer cells. In immunocompromised hosts, however, wild-type reovirus can target healthy tissues, including heart, liver, pancreas and neural structures. METHODS: We characterized an attenuated form of reovirus (AV) derived from a persistently infected cell line through sequence analysis, as well as western blot and in vitro transcription and translation techniques. To examine its pathogenesis and oncolytic potential, AV reovirus was tested on healthy embryonic stem cells, various non-transformed and transformed cell lines, and in severe combined immunodeficiency (SCID) mice with tumour xenografts. RESULTS: Sequence analysis of AV reovirus revealed a premature STOP codon in its sigma 1 attachment protein. Western blot and in vitro translation confirmed the presence of a truncated σ1. In comparison to wild-type reovirus, AV reovirus did not kill healthy stem cells or induce black tail formation in SCID mice. However, it did retain its ability to target cancer cells and reduce tumour size. CONCLUSION: Despite containing a truncated attachment protein, AV reovirus still preferentially targets cancer cells, and compared with wild-type reovirus it shows reduced toxicity when administered to immunodeficient hosts, suggesting the potential use of AV reovirus in combination cancer therapy.

Reovirus decreases azoxymethane-induced aberrant crypt foci and colon cancer in a rodent model.

Reovirus type 3 Dearing has demonstrated oncolytic efficacy in vitro and in vivo against a variety of cancer cell lines, tumor xenografts and syngeneic cancer models. In this study, we investigated the effectiveness of reovirus against aberrant crypt foci (ACF) and colon cancer induced by the carcinogen azoxymethane (AOM) in an immunocompetent rat model. Sprague-Dawley rats received 15 mg/kg AOM intraperitoneally once per week for 4 weeks and reovirus was administered rectally once a week for 5 weeks starting 20 weeks after the last dose of AOM. Two weeks after completion of reovirus therapy, animals were examined for tumor burden in the colon and other tissues. Reovirus-treated animals showed a decrease in total ACF numbers (P=0.014), in large ACFs (P=0.0069) and in tumor number (P=0.03) compared to vehicle-treated animals. Fewer obstructing tumors in the colon (P=0.07) and duodenum (P=0.03) and reduced hepatic metastases were also noted. In addition, a tumor cell line derived from hepatic metastases was found to be susceptible to reovirus in vitro. Our results show that repeated rectal reovirus administration had some efficacy in the treatment and prevention of AOM-induced ACFs, colon cancers and metastases.

Acquired resistance to reoviral oncolysis in Ras-transformed fibrosarcoma cells.

Reovirus shows considerable potential as an oncolytic agent for Ras-activated tumors and is currently in clinical trials. Here we ask whether such tumor cell lines can acquire resistance to reoviral oncolysis. We challenged human HT1080 fibrosarcoma cells that carry a Ras mutation by prolonged exposure to reovirus, thereby yielding highly virus-resistant HTR1 cells. These cells are persistently infected with reovirus, exhibit high Ras activity and retain the original Ras gene mutation, showing that resistance to reovirus can be displayed in cells with active Ras. The HTR1 cells also exhibit reduced cellular cathepsin B activity, which normally contributes to viral entry and activation. Persistently infected HTR1 cells were not tumorigenic in vivo, whereas immunologically cured virus-free HTR1 cells were highly tumorigenic. Thus, acquisition of resistance to reovirus may constrain therapeutic strategies. To determine whether reoviral resistance is associated with a general reduction in apoptotic potential, we challenged the HTR1 cells with apoptotic inducers and E1B-defective adenovirus, resulting in significant apoptosis and cell death following both approaches. Therefore, even if resistance to reoviral oncolysis should arise in tumor cells in vivo, other therapeutic strategies may nevertheless remain effective.

The oncolytic effect in vivo of reovirus on tumour cells that have survived reovirus cell killing in vitro.

The use of oncolytic viruses has received considerable attention in recent years and many viruses have proved to be effective against a variety of cancer models and a few are currently being used in clinical trials. However, the possible emergence and outcome of virus-resistant tumour cells has not been addressed. We previously reported the effective use of reovirus against lymphoid malignancies, including the Burkitt's lymphoma cell line Raji. Here we isolated in vitro persistently infected (PI) Raji cells, and cells 'cured' of persistent reovirus infection ('cured' cells). Both PI and cured Raji cells resisted reovirus infection and cell killing in vitro. In vivo, the PI cells were non-tumorigenic in SCID mice, but cured cells regained the parental cells' ability to form tumours. Tumour xenografts from the cured cells, however, were highly susceptible to reovirus oncolysis in vivo. This susceptibility was due to the proteolytic environment within tumours that facilitates reovirus infection and cell killing. Our results show that persistent infection by reovirus impedes tumour development and that although PI cells cleared of reovirus are tumorigenic, they are killed upon rechallenge with reovirus. Both the PI and cured states are therefore not likely to be significant barriers to reovirus oncolytic therapy.

A phase II study of carboplatin and chronic high-dose tamoxifen in patients with recurrent malignant glioma.

PURPOSE: To determine the response rate, time to disease progression, survival, and toxicity of intravenous carboplatin and chronic oral high-dose tamoxifen in patients with recurrent malignant gliomas. PATIENTS AND METHODS: Patients with histological confirmation of recurrent malignant gliomas were eligible for this multicenter phase II trial. Treatment consisted of 400 mg/m2 carboplatin intravenously every 4 weeks and oral high dose chronic tamoxifen (80 mg bid in women and 100 mg bid in men). RESULTS: Twenty seven patients met the eligibility criteria and were evaluable for response. The histological subtypes were: 16 (59%) glioblastoma multiforme (GBM), malignant astrocytoma (5 patients), malignant mixed glioma (5 patients), and glioblastoma/gliosarcoma (1 patient). Twenty-two patients (82%) had an ECOG performance status of 0 or 1. No complete responses were observed, 4 patients (15%) achieved a partial response, and 14 patients (52%) had stable disease. Median time to progression was 3.65 months (95%CI 2.56, 4.83). Median overall survival was 14.09 months (95%CI 7.06, 19.91). One patient with a recurrent GBM had a sustained partial response and is progression free 81 months since starting treatment. Another patient with mixed malignant oligoastrocytoma also had a prolonged partial response (lasting 63 months) and is alive 84 months after treatment for recurrence. The most frequently reported grade 3 or 4 toxicities were fatigue (19%), nausea (11%) and anorexia (11%). CONCLUSIONS: Carboplatin and high dose tamoxifen has similar response rates to other regimens for recurrent malignant gliomas and are probably equivalent to those found using tamoxifen as monotherapy. Long-lasting periods of disease free survival in some patients (particularly those with malignant mixed oligo astrocytomas) were found.

Reovirus as an experimental therapeutic for brain and leptomeningeal metastases from breast cancer.

Brain and leptomeningeal metastases are common in breast cancer patients and our current treatments are ineffective. Reovirus type 3 is a replication competent, naturally occurring virus that usurps the activated Ras-signaling pathway (or an element thereof) of tumor cells and lyses them but leaves normal cells relatively unaffected. In this study we evaluated reovirus as an experimental therapeutic in models of central nervous system (CNS) metastasis from breast cancer. We found all breast cancer cell lines tested were susceptible to reovirus, with > 50% of these cells lysed within 72 h of infection. In vivo neurotoxicity studies showed only mild local inflammation at the injection site and mild communicating hydrocephalus with neither diffuse encephalitis nor behavioral abnormalities at the therapeutically effective dose of reovirus (intracranial) (ie 10(7) plaque-forming units) or one dose level higher. In vivo, a single intratumoral administration of reovirus significantly reduced the size of tumors established from two human breast cancer cell lines and significantly prolonged survival. Intrathecal administration of reovirus also remarkably prolonged survival in an immunocompetent racine model of leptomeningeal metastases. These data suggest that the evaluation of reovirus as an experimental therapeutic for CNS metastases from breast cancer is warranted.

How often are nonenhancing supratentorial gliomas malignant? A population study.

The presence of contrast enhancement in a brain tumor is often regarded as a sign of malignancy. The authors identified 314 patients with malignant and low-grade supratentorial glial neoplasms in an unselected population, 58 of which lacked contrast enhancement on preoperative neuroimaging. Nonenhancing gliomas were malignant in approximately one third of cases, especially in older patients. Histologic confirmation of the diagnosis is therefore important in all patients suspected of harboring a primary glial neoplasm.

Differential expression and localization of TIMP-1 and TIMP-4 in human gliomas.

Studies have suggested that an imbalance of matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) may contribute to the malignant phenotype of gliomas. In this study, we have undertaken a detailed analysis of expression of the TIMP family in normal human brain and malignant gliomas at both the mRNA and protein level. Reverse transcription-PCR (RT-PCR) analyses of total RNA from surgical tumour specimens revealed unique expression patterns for the 4 members of the TIMP family, with TIMP-1 and -4 showing positive and negative correlations, respectively, with glioma malignancy. By RT-PCR, TIMP-2 and TIMP-3 expression did not change with tumour grade. In situ hybridization localized TIMP-1 to glial tumour cells and also to the surrounding tumour vasculature. TIMP-4 transcripts were predominantly localized to tumour cells, though minor expression was found in vessels. Recombinant TIMP-4 reduced invasion of U251 glioma cells through Matrigel, and U87 clones overexpressing TIMP-4 showed reduced invasive capacity in vitro. TIMP-4, but not TIMP-1, blocked Membrane Type-1-MMP-mediated progelatinase-A (MMP-2) activation in human umbilical vein endothelial cells. The differential expression and localization of individual TIMPs may contribute to the pathophysiology of human malignant gliomas, particularly with regard to tumour vascularization.

Perivascular cells regulate endothelial membrane type-1 matrix metalloproteinase activity.

Angiogenic stimuli selectively induced expression of membrane type-1 matrix metalloproteinase (MT1-MMP) transcripts and protein in human umbilical vein endothelial cells (HUVECs). Pro-MMP-2 activation was blocked by treatment with tissue inhibitor of metalloproteinases-2 (TIMP-2), but not by TIMP-1 or inhibitors of other proteinase classes. Anti-MT1-MMP antibodies abrogated recombinant pro-MMP-2 activation by plasma membranes, indicating that MT1-MMP is the main mediator of pro-MMP-2 activation in HUVECs. Cocultures of HUVECs with smooth muscle cells (SMC) or pericytes (PC) resulted in the suppression of HUVEC pro-MMP-2 activation. Treatment of A10 SMC conditioned media with a neutralising anti-TIMP-2 antibody prevented the suppression of HUVEC pro-MMP-2 activation. Inhibition of HUVEC MT1-MMP function by PC and SM3 SMC correlated with elevated TIMP-3 expression. Thus, perivascular supporting cells regulate the functions of proangiogenic MMPs elaborated by endothelial cells via selective expression of TIMPs. This interplay may be important for maintenance of blood vessel architecture and neovascularisation.

Reovirus as an oncolytic agent against experimental human malignant gliomas.

BACKGROUND: Reovirus is a naturally occurring oncolytic virus that usurps activated Ras-signaling pathways of tumor cells for its replication. Ras pathways are activated in most malignant gliomas via upstream signaling by receptor tyrosine kinases. The purpose of this study was to determine the effectiveness of reovirus as an experimental treatment for malignant gliomas. METHODS: We investigated whether reovirus would infect and lyse human glioma cell lines in vitro. We also tested the effect of injecting live reovirus in vivo on human gliomas grown subcutaneously or orthotopically (i.e., intracerebrally) in mice. Finally, reovirus was tested ex vivo against low-passage cell lines derived from human glioma specimens. All P values were two-sided. RESULTS: Reovirus killed 20 (83%) of 24 established malignant glioma cell lines tested. It caused a dramatic and often complete tumor regression in vivo in two subcutaneous (P =.0002 for both U251N and U87) and in two intracerebral (P =.0004 for U251N and P =.0009 for U87) human malignant glioma mouse models. As expected, serious toxic effects were found in these severely immunocompromised hosts. In a less immunocompromised mouse model, a single intratumoral inoculation of live reovirus led to a dramatic prolongation of survival (compared with control mice treated with dead virus; log-rank test, P<.0001 for both U251N and U87 cell lines). The animals treated with live virus also appeared to be healthier and gained body weight (P =.0001). We then tested the ability of reovirus to infect and kill primary cultures of brain tumors removed from patients and found that it killed nine (100%) of nine glioma specimens but none of the cultured meningiomas. CONCLUSIONS: Reovirus has potent activity against human malignant gliomas in vitro, in vivo, and ex vivo. Oncolysis with reovirus may be a potentially useful treatment for a broad range of human cancers.

On the selection of primary variables in numerical formulation for modeling multiphase flow in porous media.

Selecting the proper primary variables is a critical step in efficiently modeling the highly nonlinear problem of multiphase subsurface flow in a heterogeneous porous-fractured media. Current simulation and ground modeling techniques consist of (1) spatial discretization of mass and/or heat conservation equations using finite difference or finite element methods; (2) fully implicit time discretization; (3) solving the nonlinear, discrete algebraic equations using a Newton iterative scheme. Previous modeling efforts indicate that the choice of primary variables for a Newton iteration not only impacts computational performance of a numerical code, but may also determine the feasibility of a numerical modeling study in many field applications. This paper presents an analysis and general recommendations for selecting primary variables in simulating multiphase, subsurface flow for one-active phase (Richards' equation), two-phase (gas and liquid) and three-phase (gas, water and nonaqueous phase liquid or NAPL) conditions. In many cases, a dynamic variable switching or variable substitution scheme may have to be used in order to achieve optimal numerical performance and robustness. The selection of primary variables depends in general on the sensitivity of the system of equations to the variables selected at given phase and flow conditions. We will present a series of numerical tests and large-scale field simulation examples, including modeling one (active)-phase, two-phase and three-phase flow problems in multi-dimensional, porous-fractured subsurface systems.

Localization of gelatinase-A and gelatinase-B mRNA and protein in human gliomas.

Malignant gliomas maintain a poor prognosis and survival rate due to their marked local invasive growth and neovascularization. Matrix metalloproteinases (MMPs) have been implicated in glioma invasion and angiogenesis, but it is unknown whether they are produced by the tumor cells or surrounding stroma. Using in situ hybridization and immunohistochemistry, we found expression of mRNA for both gelatinase-A (MMP2) and gelatinase-B (MMP9) localized to tumor cells and vascular structures in glioma sections. Gelatinase-A protein expression was detected most prominently in tumor cells, with very little signal seen in vasculature. Gelatinase-B protein expression was prominent in vascular structures but was also expressed in tumor cells. Our data show that these proteases are produced by glioma cells and vascular structures and suggest that synthetic MMP inhibitors might be useful in this disease.

Gelatinase-A (MMP-2), gelatinase-B (MMP-9) and membrane type matrix metalloproteinase-1 (MT1-MMP) are involved in different aspects of the pathophysiology of malignant gliomas.

Matrix metalloproteinases (MMPs) have been implicated as important factors in gliomas since they may both facilitate invasion into the surrounding brain and participate in neovascularization. We have tested the hypothesis that deregulated expression of gelatinase-A or B, or an activator of gelatinase-A, MT1-MMP, may contribute directly to human gliomas by quantifying the expression of these MMPs in 46 brain tumour specimens and seven control tissues. Quantitative RT-PCR and gelatin zymography showed that gelatinase-A in glioma specimens was higher than in normal tissue; these were significantly elevated in low grade gliomas and remained elevated in GBMs. Gelatinase-B transcript and activity levels were also higher than in normal brain and more strongly correlated with tumour grade. We did not see a close relationship between the levels of expression of MT1-MMP mRNA and amounts of activated gelatinase-A. In situ hybridization localized gelatinase-A and MT1-MMP transcripts to normal neuronal and glia, malignant glioma cells and blood vessels. In contrast, gelatinase-B showed a more restricted pattern of expression; it was strongly expressed in blood vessels at proliferating margins, as well as tumour cells in some cases. These data suggest that gelatinase-A, -B and MT1-MMP are important in the pathophysiology of human gliomas. The primary role of gelatinase-B may lie in remodelling associated with neovascularization, whereas gelatinase-A and MT1-MMP may be involved in both glial invasion and angiogenesis.

Is primary CNS lymphoma really becoming more common? A population-based study of incidence, clinicopathological features and outcomes in Alberta from 1975 to 1996.

BACKGROUND: The incidence of primary CNS lymphoma (PCNSL) is believed to be increasing in immunocompetent patients but this may not be universally true. The objective of this study was to determine in a population if the incidence of PCNSL is increasing, if the histologic subtypes are changing, and to describe the clinicopathologic and outcome characteristics of PCNSL. PATIENTS AND METHODS: We identified all Alberta residents with a histologic diagnosis of PCNSL from 1 January 1975 to 31 December 1996 using the Alberta Cancer Registry. Annual age-standardized incidence rates (ASIR), clinicopathologic and outcome characteristics were determined. RESULTS: There were 50 immunocompetent PCNSL patients; the median age was 64 and 30 were male. Their median survival was 10.15 months. Histology was available for review in 37 (74%) patients: 19 (51%) were diffuse large cell, 16 (43%) were immunoblastic and 2 (5%) were unclassifiable malignant lymphomas. The ASIR ranged from 0.178-1.631/10(6) and no change in ASIR was found (test for trend, P = 0.26) for gender or age. The ASIR of malignant gliomas did not change either but increased for all other non-Hodgkin's lymphoma (94.95-138.7610(6); test for trend, P = 0.0001) The number of brain biopsies increased from 1979-1985 (test for trend, P < 0.0001) but remained stable from 1986-1996 (test for trend, P = 0.99). CONCLUSIONS: Unlike several other populations, PCNSL is not becoming significantly more common in Alberta. If this difference is real (i.e., not due to differences in cancer registry coding practices etc.) comparisons between Albertans and other populations in whom the incidence is rising may provide clues regarding the etiology of PCNSL.

Cellular proliferation in pilocytic and diffuse astrocytomas.

Using quantitative image analysis, we evaluated the MIB-1 labeling index (LI) in a large population of pilocytic (n = 131) and diffuse astrocytomas (n = 140), explored its significance as a prognostic predictor of survival, and compared it to other commonly accepted predictors, including grade and its histologic determinants, atypia, mitoses, endothelial proliferation, and necrosis. Diffuse astrocytomas were graded according to the St Anne-Mayo scheme and included 45 grade 2, 50 grade 3, and 45 grade 4 astrocytomas. In pilocytic astrocytomas, mean, median, and range of MIB-1 LIs were 1.1, 0.9, and 0-3.9%, respectively. In diffuse astrocytomas, these values were 2.3, 2, and 0-7.6% in grade 2; 6, 4.4, and 0.1-25.7% in grade 3; 9.1, 6, and 0.3-36% in grade 4. There was a significant difference in the distribution of MIB-1 LIs between pilocytic and diffuse grade 2 astrocytomas (p < 0.001), between grade 2 and grade 3 (p < 0.001), and between tumors of grade 3 and 4 (p = 0.014). Among pilocytic astrocytomas there was no association between survival and MIB-1 LI or any histologic parameter. In diffuse astrocytomas, MIB-1 LI was significantly correlated with grade as well as with mitotic activity (<0.001) and survival. While in diffuse astrocytomas of all grades, necrosis was the strongest factor associated with survival, in tumors of grades 2 and 3 the MIB-1 LI preceded other histologic parameters and, on multivariate analysis, remained the only feature predictive of survival. Grade 3 astrocytomas with a single "solitary" mitosis had a significantly lower MIB-I LI than did grade 3 tumors with >1 mitosis and, compared to the latter, had a significantly longer survival (p = 0.013), one not significantly different from patients with grade 2 astrocytomas. These findings suggest that the cutoff point between grade 2 and 3 in the St. Anne-Mayo scheme may not be optimal and may need to be revised.

Reovirus therapy of tumors with activated Ras pathway.

Human reovirus requires an activated Ras signaling pathway for infection of cultured cells. To investigate whether this property can be exploited for cancer therapy, severe combined immune deficient mice bearing tumors established from v-erbB-transformed murine NIH 3T3 cells or human U87 glioblastoma cells were treated with the virus. A single intratumoral injection of virus resulted in regression of tumors in 65 to 80 percent of the mice. Treatment of immune-competent C3H mice bearing tumors established from ras-transformed C3H-10T1/2 cells also resulted in tumor regression, although a series of injections were required. These results suggest that, with further work, reovirus may have applicability in the treatment of cancer.